Document [original]

The Mitochondrial Protein Profile Changes during

the Aging Process

Proteomic aspects and mathematical modeling

Vorgelegt von

Diplom-Ingenieur

Lei Mao

aus Pekin, China

angefertigt am Institute für Humangenetik

Charité Campus Virchow-Klinikum

Unter der Beteuerung von

Prof. Dr. Dr. Joachim Klose

Zur Erlangung des akademischen Grades

Doktor der Naturwissenschaften

- Dr. rer.nat. -

der Fakultät III-Prozesswissenschaften

Institut für Biotechnologie

der Technischen Universität Berlin

Univ.-Prof. Dipl.-Ing. Dr. U. Stahl

Genehmigte Dissertation

Promotionsausschuss:

Vorsitzender: PD Dr. rer. nat. Roland Lauster

Gutachter: Univ.-Prof. Dipl.-Ing. Dr. Ulf Stahl

Gutachter: Prof. Dr. Dr. Joachim Klose

Tag der wissenschaftliche Aussprache: 15.02.2005

Berlin 2005

D 83

Table of Contents

1 Introduction......................................................................................................................1

1.1 Disposable soma theory of aging...................................................................................1

1.2 Free radical theory of aging ...........................................................................................1

1.3 The role of mitochondria in free radical production......................................................2

1.4 Involvement of mitochondrial mutation in aging...........................................................2

1.4.1 Possible mechanism of mtDNA mutation..........................................................4

1.4.2 mtDNA supposed to be especially vulnerable ...................................................4

1.5 The mitochondrial theory of aging.................................................................................4

1.5.1 Experimental evidences of mitochondrial theory of aging ................................5

1.5.2 Theoretical supports of the mitochondrial theory of aging................................6

1.5.3 Controversial observations.................................................................................6

1.6 Possible causes of these controversial observations ......................................................7

1.7 Proteomic analyses in aging research become obligatory..............................................8

1.8 Current state of mitochondrial proteomic research........................................................9

1.8.1 Why there has been a lack of protein-level analysis........................................10

1.8.2 Recent development of proteomic technology makes this study tractable ......10

1.9 Complexity of the system requires modeling...............................................................11

2 Aim of the Study.............................................................................................................12

3 Animals, Materials and Methods..................................................................................13

3.1 Animals and ethnical agreements.................................................................................13

3.2 Materials and methods .................................................................................................13

3.2.1 Organ obtainment.............................................................................................13

3.2.2 Enzymatic histochemical staining of COX-activity.........................................14

3.2.3 Mitochondria isolation.....................................................................................15

3.2.3.1 Tissue homogenisation and crude mitochondria collection.........................15

3.2.3.2 Purification of mitochondria using gradient centrifugation.........................16

3.2.4 Electronic microscopic control of mitochondrial morphology........................17

3.2.5 Sequential extraction of mitochondrial proteins..............................................18

3.2.5.1 Tris-buffer extraction...................................................................................19

3.2.5.2 Triton buffer extraction................................................................................20

3.2.5.3 Methanol-chloroform protein extraction......................................................20

3.2.6 Protein concentration analysis..........................................................................20

3.2.7 Sample preparation for 2D-electrophoresis......................................................22

3.2.8 Sample preparation for Western immunoblotting............................................23

3.2.9 Sample preparation for Blue-native electrophoresis........................................23

3.2.10 Large-gel 2D-electrophoresis...........................................................................23

3.2.10.1 Principle of 2D-electrophoresis................................................................24

3.2.10.2 Procedure of isoelectric focussing electrophoresis ..................................24

3.2.10.3 Second dimension of 2D-electrohporesis.................................................26

3.2.11 Silver staining...................................................................................................26

3.2.11.1 Analytical silver staining..........................................................................26

3.2.11.2 Preparative silver staining........................................................................27

3.2.12 2D-gel evaluation.............................................................................................28

3.2.13 Protein identification........................................................................................29

3.2.13.1 In-gel trypsin digestion.............................................................................30

3.2.13.2 MALDI-TOF-MS analysis.......................................................................31

3.2.13.3 ESI-MS analysis.......................................................................................31

3.2.13.4 Databank-based protein annotation..........................................................32

3.2.13.5 Membrane protein prediction and protein sequence alignment...............32

3.2.14 Western immunoblotting..................................................................................33

3.2.15 Blue-native electrophoresis..............................................................................35

3.2.16 Mathematical modeling....................................................................................35

3.2.16.1 Modeling concept.....................................................................................36

3.2.16.2 Differential equations and numerical solutions........................................38

4 Statistics...........................................................................................................................41

5 Result...............................................................................................................................43

5.1 Organ weight difference during development and aging.............................................43

5.2 Aging causes COX-activity deficiency on mouse muscle tissue.................................44

5.3 Result of mitochondria isolation..................................................................................46

5.3.1 Morphology of isolated mitochondria..............................................................47

5.3.2 Yield of mitochondria......................................................................................50

5.3.3 Comparison of mitochondria isolation from fresh and frozen material...........51

5.3.4 Comparison of mitochondria isolation from young and old organism............52

5.4 Result of protein pre-fractionation...............................................................................53

5.5 Result of 2D-electrophoresis analyses .........................................................................55

5.5.1 Reproducibility of 2D-electrophoresis carried out in this study......................57

5.5.2 Comparison of whole tissue 2D-gel to mitochondrial 2D-gel.........................59

5.5.3 Effect of protein pre-fractionation on protein resolution.................................60

5.5.4 Access of possible proteomic changes based on protein spot pattern..............62

5.6 Result of protein identification.....................................................................................63

5.6.1 Quality of MS-spectra......................................................................................63

5.6.2 Databank-based protein identification .............................................................64

5.7 Prediction of hydrophobicity of identified proteins.....................................................65

5.8 Quantitative changes of protein spots observed...........................................................66

5.8.1 Alpha-synuclein increased with age in brain mitochondria.............................69

5.8.2 COX subunit Vb decreased with age ...............................................................71

5.8.3 10kDa heat shock protein decreased with age .................................................72

5.8.4 Two complex I subunits decreased with age....................................................74

5.8.5 Peroxiredoxin 1 decreased with age in liver mitochondria..............................75

5.8.6 Regucalcin decreased with age in liver mitochondria......................................77

5.8.7 Increase of a mitochondrial ribosomal protein in liver....................................78

5.8.8 Alteration of ubiquinol-cytochrome c reductase binding protein and ATP

synthase subunit................................................................................................80

5.9 Result of Western immunoblotting..............................................................................82

5.10 Preliminary result of Blue-native electrophoresis........................................................83

5.11 Mathematical simulation and model fitting .................................................................84

6 Discussion........................................................................................................................89

6.1 Investigation on mitochondrial theory of aging needs proteomic approaches.............89

6.2 Mouse has been proven to be pertinent model organism for aging study....................89

6.3 The choice of organs....................................................................................................90

6.4 The choice of strategy..................................................................................................90

6.5 Mitochondrial isolation was successful........................................................................91

6.6 Protein fractionation was effective...............................................................................92

6.7 Satisfactory result from 2D-PAGE analysis was obtained...........................................93

6.8 Gel image evaluation was successful...........................................................................94

6.9 Protein identification was efficient ..............................................................................94

6.10 2D-electrophoresis combined with mass spectrometry is an efficient proteomic

strategy .........................................................................................................................95

6.11 Analysis of membrane proteins remains a problem.....................................................95

6.12 Mitochondrial protein profile change during the aging process ..................................96

6.12.1 Down-regulation of complex I and complex IV subunits indicates mtDNA

mutation............................................................................................................96

6.12.2 Increase of complex III and complex V subunits suggest feedback regulation97

6.12.3 MtDNA-encoded COX subunit I showed only moderate change....................97

6.12.4 Decrease of a mitochondrial heat-shock protein could suggest the increased

consumption of heat shock protein...................................................................98

6.12.5 Down-regulation of peroxiredoxin suggest elevated oxidative stress in aged

individual..........................................................................................................98

6.12.6 Down-regulation of regucalcin in liver mitochondria indicates a lowered

buffering capacity of calcium...........................................................................99

6.12.7 Up-regulation of alpha-synuclein in brain mitochondria resembles neuronal

degenerative diseases......................................................................................100

6.12.8 Difference between brain and liver respecting mitochondrial aspect of aging....

........................................................................................................................102

6.13 Potential of Blue-native electrophoresis analysis.......................................................103

6.14 The accumulation of defective mitochondria with age was simulated......................104

6.15 Mutation rate of mouse mtDNA was estimated.........................................................105

6.16 Result of current study is consistent with the mitochondrial theory of aging............107

7 Conclusion.....................................................................................................................109

8 Outlook..........................................................................................................................112

9 Zusammenfassung........................................................................................................114

10 Reference.......................................................................................................................116

III

Summary

The accumulation of mitochondria bearing mutated genomes was proposed to be an

important factor involved in aging (Wallace, 2001). In order to investigate the effect of

mtDNA mutation at the protein level, we studied the mitochondrial proteome during aging

with a mouse model (C57/BL6).

For the validation of the mouse model, histochemical staining was carried out to compare

the cytochrome c oxidase (COX) activity on mouse muscle tissues of young (newborn) and

old animals (24-months). No COX-negative myocyte was found in young mouse muscle

tissue, whereas a significant part of the myocytes (43%) in the old-aged mouse muscle

showed lowered COX-activity compared to the remaining cells in the same tissue. The

senescent muscle tissue displayed typical “mosaic” pattern, similar to that has been

described in previous studies using human material. This indicates that the mouse could be

a valid model for human aging.

Mitochondria were isolated from mouse brain and liver at six different ages (newborn to 24-

months, n=8 to 13) using continuous gradient centrifugation. Mitochondrial proteins were

sequentially extracted using Tris-buffer (tris-(hydroxymethyl)-aminomethane buffer)

(resulting “Fraction I”) and Triton-containing buffer (resulting “Fraction II”), while the

remaining pellet underwent methanol-chloroform extraction (resulting “Fraction III”). Large-

gel 2D-electrophoresis analysis and a modification of 2D-electrophoresis (employing Triton-

X100) were utilized for the analysis of “Fraction I” and “Fraction II” proteins, respectively.

Western immunoblotting was carried out on “Fraction III” samples to elucidate the changes

in mtDNA-encoded protein COX subunit I.

The expression of two respiratory chain complex I subunits (NADH-ubiquinone

oxidoreductase 13 kDa-A subunit and NADH-ubiquinone oxidoreductase 1 alpha

subcomplex 5) and one complex IV subunit (COXVb) decreased with age. One subunit of

complex III (ubiquinol-cytochrome c reductase binding protein), one subunit of complex V

(ATP F0 subunit) and a mitochondrial ribosomal protein increased in expression during

aging. Together, these data indicate that complex I and IV deficiency in aged tissue is

accompanied by feedback regulation of other protein complexes in the respiratory chain.

This is consistent with the previous prediction that accumulation of mtDNA deletion affect

predominantly complex I and complex IV genes (Vu et al., 2000).

Furthermore, the observed down-regulation of the 10 kDa mitochondrial heat shock protein

indicated an elevated level of oxidative stress in aged mouse brain and liver tissue, which

could be a common aspect in aging and neuronal degenerative diseases (Cottrell et al.,

2000; Richter et al., 1988). The up-regulation of mitochondria-associated alpha-synuclein in

brain tissue might indicate an enhanced susceptibility to protein aggregation with advanced

age (Goedert, 1997; Ueda et al., 1993). The decrease of mitochondria-associated

regucalcin in liver tissue indicates a lowered mitochondrial buffering capacity of calcium

(Takahashi and Yamaguchi, 2000; Xue et al., 2000).

A mathematical model was developed to simulate the accumulation of defective

mitochondria during aging. When we applied our quantitative data observed by 2DE to this

model, the mtDNA mutation rate was estimated to be 1.2x10-8 per gene per day. This

mutation rate is high enough to lead to the accumulation of defective mitochondria during

the biological time scale.

The experimental data gained from our proteomic study were in accordance with the

hypothesis that mitochondrial somatic mutations accumulate with age. This may explain the

progression of mitochondrial dysfunction and increasing level of oxidative stress during the

aging process. Future investigation will focus on the mtDNA-encoded protein changes and

the in-depth mechanism of protein interaction during the aging process. This work would

contribute to a better understanding of the mechanism of aging process, would also find

application in the development of mitochondria-targeting therapies to prevent from insidious

accompanies of aging process such like age-related degenerative diseases.

Keyword:

Mitochondria, aging, degenerative diseases, proteome, 2D-PAGE, mouse model,

mathematical modeling.

List of Abbreviations

°C degree Celsius

2D-electrophoresis two-dimensional electrophoresis

8-OH-dG 8-hydroxydeoxyguanosine

A (mA) ampère (milliampère)

ACN acetonitrile

ADP adenosine diphosphate

ATP adenosine triphosphate

BCA Bicinchoni acid

Bisacrylamide N, N’-methylene-bis-acrylamide

BisTris 2-[Bis-(2-hydroxyethyl)amino]-2-(hydroxymethyl)propane-1,3-diol

bp (kbp) base pair (kilo base pairs)

BSA bovine serum albumin

CHAPS 3-[(3-cholamidopropyl)-dimethylammonio]-propan-sulfonate

cm centimeter

CoA coenzyme A

Complex I NADH-ubiquinone oxidoreductase

Complex II succinate-ubiquinone oxidoreductase

Complex III ubiquinol-cytochrome c reductase

Complex IV cytochrome c oxidase

Complex V ATP synthase

CoQ Coenzyme Q (ubiquinone)

COX cytochrome c oxidase

Da (kDa) dalton (kilodalton)

dAMP deoxyadenosine 5'-monophosphate

dCMP deoxycytidine 5'-monophosphate

dGMP deoxyguanosine 5'-monophosphate

DHB 2,5-dihydroxybenzoic acid

DNA desoxyribonucleic acid

Dpi dots per inch

dpi dots per square-inches

DTT dithiotreitol

EDTA ethylenediaminetetraacetic acid

EGTA ethylene glycol-bis(beta-aminoethyl ether)-N,N,N',N'-t etraacetic acid

EMBL European Molecular biology laboratory

ESI electro spray ionisation

FADH2 reduced flavin adenine dinucleotide

g (mg, ng) gram (milligram, nanogram)

h hour

HPLC high performance liquid chromatography

HTML Hyper Text Markup Language

IEF isoelectric focussing

IgG immunoglobulin G

IPG immobilised pH-gradient

l (ml, ml) litre (millilitre, microliter)

M (mM) molar (mill molar)

m/z ratio of mass to charge

MALDI-TOF matrix assisted laser desorption/ionisation-time of flight

min minute

mRNA messenger ribonucleic acid

VII

MS mass spectrometry

mtDNA mitochondrial DNA

MPTP 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine

MW molecular weight

NADH reduced nicotinamide adenine dinucleotide

PAGE polyacrylamide gel electrophoresis

PBS phosphate buffer saline

PCR polymerase chain reaction

pI isoelectric point

PMSF phenylmethylsulphonyl fluoride

RNA ribonucleic acid

ROS reactive oxygen species

rpm rotations per minute

rRNA ribosomale RNA

SDH succinate:ubiquinone oxidoreductase

SDS sodium deodecylsulfate

SDS-PAGE sodium dodecylsulfate-polyacrylamide gel electrophoresis

SOD superoxide dismutase

SPF Specific pathogen Free

TEMED N, N, N’, N’ – tetramethylethylenediamine

TFA trifluoroacetic acid

TIM translocases of the inner-membrane

TOM translocases of the outer-membrane

Tricine N-[2-Hydroxy-1,1-bis(hydroxymethyl)ethyl]glycine

Tris-base tris-(hydroxylmethyl)-amoniomethane

Tris-HCl tris (hydroxymethyl) aminomethane, Hydrochloride

Triton-X100 Polyethylene glycol tert-octylphenyl ether

tRNA transfer RNA

Tween 20 Polyoxyethylene (20) sorbitan monolaurate

V(mV) volt (mill volt)

v/v volume per volume

w/v weight per volume

VIII

1 Introduction

“Aging is the progressive accumulation of changes with time that are associated with or

responsible for the ever-increasing susceptibility to disease and death which accompanies

advancing age“ (Harman, 1981). As a biological phenomenon, aging exist in all sexually

reproductive life forms, including unicellular organisms (Bowen and Atwood, 2004). The

functional pathways involved in aging process may include responses to endogenous and

exogenous changes, such like hormonal changes and damage accumulation (Finch, 1993).

Aging and longevity are also influenced by genes (Finch and Tanzi, 1997).

1.1 Disposable soma theory of aging

Although no consensus of aging mechanism exist, it has been noticed that only species

with a clear distinction between soma and germ line undergo somatic senescence

(Kirkwood and Holliday, 1979; Le Bourg, 1998). Based on their observation, Lindop &

Rotblatt proposed the disposable somatic theory (Henshaw, 1947; Lindop and Rotblat,

1961b), which states that aging may result from the accumulation of unrepaired somatic

defects due to the reduced investigation for the somatic maintenance. For instance,

telomerase, the enzyme that is responsible for the maintenance of the proper length of the

chromosomal DNA does not function in somatic part (Harley et al., 1990; Hastie et al.,

1990; Lindsey et al., 1991). This is also in accordance with the somatic theory of aging.

1.2 Free radical theory of aging

Respecting the reason of somatic damage, it is generally accepted that cellular DNA is

constantly exposed to exogenous and endogenous DNA-damaging agents, with reactive

oxygen species (ROS) as the most important contributors. This was deduced from the early

observation of Lindop and Henshaw, who showed in their experiments that irradiation

damage of DNA shortened lifespan in animal models and induced features of premature

aging (Henshaw, 1947; Lindop and Rotblat, 1961a).

The direct relation of free radical and gene mutation has been proposed by Lindop &

Rotblatt, who intensively studied the correlation between radiation and the concentration

increase of 8-hydroxydeoxyguanosine (8-OH-dG), a marker of DNA oxidative damage.

Later, Feeney suggested that free radicals are also involved in reactions that led to the

damage of the biological membrane and proteins (Feeney and Berman, 1976).

1.3 The role of mitochondria in free radical production

As the major site of free radical generation in the eukaryotic cells, mitochondria have been

pushed to the middle of the stage. Mitochondria are complex organelles bound by an inner

and an outer membrane. They are involved in multiple cellular processes such as citric acid

cycle, oxidative phosphorylation, β-oxidation, calcium homeostasis, urea cycle, heme

biosynthesis, apoptosis and cell signalling (Green and Amarante-Mendes, 1998).

Phylogenetic data support an origin of mitochondria from the bacteria of the Order

Rickettsiales (Emelyanov, 2001).

The respiratory chain localized on the mitochondrial inner membrane is composed of five

multi-polypeptide enzyme complexes. These include complex I (NADH-ubiquinone

oxidoreductase), complex II (succinate-ubiquinone oxidoreductase), complex III (ubiquinol-

cytochrome c reductase), complex IV (cytochrome c oxidase, or COX) and complex V (ATP

synthase). Additionally, two mobile electron carriers (ubiquinone and cytochrome c) are

also involved in the oxidative phosphorylation. The respiratory chain oxidizes electrons

distributed from reduced nicotinamide adenine dinucleotide (NADH) or flavin adenine

dinucleotide (FADH2) to build the electro-chemical gradient across the mitochondrial inner

membrane. This potential energy is utilized by complex V to synthesize ATP from ADP and

orthophosphate.

Noteworthylly, the components positioned early in the respiration chain (Complex I and

Coenzyme Q) leak some fraction of the electrons directly to molecular oxygen to form

superoxide anion (O2•-). Under physiological conditions, up to 2% of the total oxygen

consumption was estimated to form superoxide radicals (Joenje, 1989). Besides, in the

presence of reduced transition metals, H2O2 can also be converted to the highly reactive

hydroxyl radical (OH•) by Fenton reactions. These are only partially detoxified by a variety

of enzymes and free radical scavengers, including catalase, glutathione and peroxiredoxin

(Joenje, 1989).

1.4 Involvement of mitochondrial mutation in aging

Except for Complex II, all respiratory chain complexes contain protein subunits that are

encoded by mitochondrial genes on the circular mtDNA. These proteins are synthesized

inside mitochondria by their own protein synthesis machinery (fig.1). The influence of

mtDNA mutation on aging thus raised attention (Wallace, 1992).

Fig.1: The human mitochondrial genome is a 16,569 bp molecule of double-stranded DNA.

This mtDNA encodes for 13 essential components of the respiratory chain. DN1-ND6 and

ND4L encode seven subunits of complex I, Cytb is the only mtDNA encoded complex III

subunit. COXI-III encode for three of the complex IV subunits, whereas the ATP6 and

ATP8 genes encode for two subunits of complex V. OH and OL are origins of replication

from heavy and light strand mtDNA, respectively. Area marked in green shows the

sequence range that is frequently been deleted (cited from web site of University of

Minnesota, www.chem.umn.edu/groups/arriaga/mitochondria.htm).

Numerous mtDNA mutations have been observed in diverse mitochondrial diseases. These

include large deletions (Lee et al., 1994), short duplications (Brockington et al., 1993) and

point mutations (Munscher et al., 1993).

Interestingly, many mtDNA abnormalities that had previously been linked to a variety of

clinical syndrome were also found to accumulate in normal aged individuals (Corral-

Debrinski et al., 1992; Linnane et al., 1992). Particularly, a 4977bp large mtDNA deletion

was most commonly found in aged individuals (Cortopassi and Arnheim, 1990; Linnane et

al., 1990). This so-called “common deletion” occurs between two 13bp direct repeats

located at nucleotide position 8470 to 8482 and at 13447 to 13459, respectively (Lee et al.,

1994; Yen et al., 1991; Zhang et al., 1997a).

In 1997, Tengan observed the positive correlations between age and “common deletion”

1.4.1 Possible mechanism of mtDNA mutation en proposed to be the exposure to

he mutagenic potential of 8-OH-dG is reflected by its miscoding properties: instead of

1.4.2 mtDNA supposed to be especially vulnerable higher than in comparison to

dditionally, since mtDNA molecules in mammals have a high information density with

.5 The mitochondrial theory of aging

Based on the observation that accumulation of mitochondrial somatic mutation is

associated with the mitochondrial dysfunction that occur during the aging process, the

levels in both non-diseased controls (r=0.80) and patients with mitochondrial diseases

(r=0.69) (Tengan et al., 1997). This led to the suggestion that there is an important

connection between aging and the accumulation of mitochondrial DNA mutations (Wallace,

1992).

The most likely cause of mtDNA rearrangement has be

ROS (O2•-, H2O2 and OH•) (Ames et al., 1993; Richter et al., 1988). In 1992, Hayakawa

observed the clear correlation between the 8-OH-dG content in human heart and the

amount of mtDNA with a deletion (r=0.93) (Hayakawa et al., 1992).

dCMP, dAMP can be incorporated opposite the modified base 8-OhdG during replication

(Shibutani et al., 1991), thus introducing point mutation. In turn, illegitimate recombination

or strand slippage could happen during replication, which results in mtDNA deletion

(Baumer et al., 1994; Taylor et al., 2001). The expression of mtDNA might also be

compromised through interference in RNA polymerase bound with damaged nucleotides

(Nagley and Wei, 1998). Furthermore, mtDNA molecules subjected to extensive oxidative

damage can become cross-linked to other macromolecules such as proteins and lipids and

this too could interfere with efficient mtDNA replication and transcription in mitochondria

(Nagley and Wei, 1998).

The mutation rate of mtDNA is proposed to be 10 to 20 fold

nuclear genes (DiMauro et al., 2000; Osiewacz and Hamann, 1997; Zeviani et al., 1998).

This is due to the lack of histone protection on the mtDNA molecule, the proposed

insufficient DNA repair mechanism in mitochondria (Cullinane and Bohr, 1998), and the

extensive exposure of mtDNA molecules to ROS.

essentially no sequence redundancy, large-scale deletions and point mutation often cause

the loss or truncation of not only structural genes, but also rRNA and tRNA genes. This will

in turn have deleterious effects on mitochondrial protein production.

mitochondrial theory of aging was subsequently deduced by Harman as an extension of the

free radical theory (Harman, 1972) (fig.2).

Reactive oxygen

speceis

Mitochondrial

dysfunction

mtDNA mutation

Age-related changes

Reactive oxygen

speceis

Mitochondrial

dysfunction

mtDNA mutation

Age-related changes

Fig.2: Mitochondrial theory of aging proposes that free-radical-induced mtDNA mutation is

able to accumulate along with age. This could be the driving force leading to mitochondrial

ysfunction and the phenomenon of aging.

tation is the driving force of the aging process

owald and Kirkwood, 1999; Linnane et al., 1989; Richter, 1988). The theory foundations

94; Brierley et al., 1998; Kopsidas et

al., 1998). Histochemical staining has shown that aged muscle tissue show alternating

The mitochondrial theory of aging states that the slow accumulation of impaired

mitochondria due to free-radical induced mu

of mitochondrial theory of aging include two assumptions: firstly, mtDNA supposed to be

more vulnerable than nuclear DNA respecting free radical induced mutations. Compared to

nuclear DNA, human fibroblast mtDNA is damaged three times more by H2O2, 16 times

more by ROS in general. More importantly, mitochondrial mutations have the potential of

disrupting whole cellular physiology (Wallace, 1997).

1.5.1 Experimental evidences of mitochondrial theory of aging

Several lines of evidence support the view that the bio-energetic function of the

mitochondria deteriorates with age (Boffoli et al., 19

cytochrome c oxdiase (COX)-positive and COX-negative regions, displaying a mosaic

pattern (Kopsidas et al., 2000; Mita et al., 1989). In parallel, there exists a corresponding

“mosaic pattern” respecting mtDNA mutation. Those COX-negative regions contain high

concentrations of the deleted mtDNAs (Shoubridge et al., 1990; Zhang et al., 1997b).

Accumulation of mtDNA mutations are widespread processes in various human tissues

during the aging process. Age-dependent deterioration of mitochondrial respiration activity

nd age-associated mitochondrial DNA deletion were observed in human liver, heart, skin

ormal skin tissues harbour higher level of the 4977bp deleted mtDNA than the faster

hysiology of the cell and thus cannot play a

le in the aging process (Coller et al., 2002). To affect cellular physiology, the nascent

sing a

itochondria, the

l model constructed by Kowald and

irkwood (Kowald and Kirkwood, 2000) was able to simulate such dynamic processes

scle fibres were a real finding in senescent tissues, the pivotal role

of mtDNA mutations to the aging process is still controversial (Brierley et al., 1997). Some

other investigations failed to detect age-dependent changes in mitochondria (Bodenteich et

and brain tissues (Corral-Debrinski et al., 1992; Kovalenko et al., 1998; Yen et al., 1994).

Furthermore, it was demonstrated that mtDNA mutation does not occur uniformly in

different tissues (Kovalenko et al., 1998). Liu and Pang observed that the slower growin

growing skin tissues in cancerous and precancerous skin tissues (Liu et al., 1998; Pang et

al., 1994). MtDNA rearrangement in tissues from aged human subjects occur in levels

ranging from very low in liver, to considerable in cardiac muscle, to almost total in skeletal

muscle (Kopsidas et al., 1998; Liu et al., 1998). Generally, it gives consensus that age-

related mtDNA mutation occur more frequently and accumulate much faster in tissues of

high energy-demand and low mitotic activity.

However, since most cells contain hundreds to thousands of mtDNA molecules, A single

mutant molecule is unlikely to influence the p

somatic mutants must somehow accumulate in the cell to significant levels.

1.5.2 Theoretical supports of the mitochondrial theory of aging

In 1976, the mitochondrial theory of aging was refined by De Grey (de Grey, 1997), who

reasoned the theoretical possibility of mitochondrial mutation accumulation. By propo

slower degradation rate of defected mitochondria compared to wild type m

defective mitochondria gain a selective advantage.

According to DeGrey’s idea, the mutated mtDNA molecules can eventually lead to a

homoplasmic status of the cell. A mathematica

theoretically.

1.5.3 Controversial observations

Whilst COX-deficient mu

al., 1991; Manzelmann and Harmon, 1987). Jazin found in the brain tissue that while the

t al., 2003). They showed that mtDNA deletions co-locate not only with

itochondrial abnormalities, but also with thinned and degenerated fibre morphology.

came from genetic analysis, mostly polymerase chain

proteins

high incidence of mutations (Jazin et al., 1996; Nekhaeva

t al., 2002). This implies that results of PCR are largely influenced by whether the applied

ne is sufficient to clarify complete mechanisms of aging. The mtDNA genes

eed a whole repertoire of nuclear proteins for their protein transcription, translation and

occurrence of sequence variation in mtDNA was significantly higher in the non-coding (D-

loop) region of mtDNA of the aged individuals compared with that of the younger subjects,

a very low occurrence of variation was found in coding regions of mtDNA segments (Jazin

et al., 1996).

Kraytsberg argued that accumulation of defective mitochondria in cells is probably not the

only and not the most important mechanism potentially relating mtDNA mutations to aging

(Kraytsberg e

Respecting the proposed lack of DNA repair mechanism in mitochondria, several repair

pathways have been recently described for mtDNA, including double strand breakage

repair (Bohr and Dianov, 1999).

1.6 Possible causes of these controversial observations

Reasons for the existence of these controversies could be two-fold: Firstly, the majority of

tions described above the observa

reaction (PCR) analysis on mtDNA and nuclear DNA encoded mitochondrial

(Cortopassi and Arnheim, 1992).

PCR methods have the advantage of selectively amplifying only mtDNA bearing certain

range. However, mtDNA mutations appear to concentrate in certain “hotspot” areas, i.e.,

small regions of the genome with

primer contains certain mutation hotspots or not. Scoring mutations on a fragment that is

too short can lead to the overestimation or underestimation of mutation events (Kraytsberg

et al., 2003).

Furthermore, even though the recent development of whole range PCR is able to amplify

the complete mitochondrial DNA molecule, it is still doubtful whether the genomic

information alo

assembly (Sickmann et al., 2003; Zhao et al., 2000). The presence of a mitochondrial gene

on the mtDNA molecule does not mean that this respective protein will be properly

produced and assembled into functional protein complexes.

Expression levels of a protein depend not only on transcription rates of the gene, but also

on additional control mechanisms, such as transcript stability, translational regulation and

protein degradation. Moreover, both the activity and the function of proteins can be altered,

ainly through post-transnational modification (glycosylation, phosphorylation) or

of expressed and modified protein

etworks. Transcription or translation products of genes in oxidative damaged or mutated

mbly, and in turn

o elucidation of the integrated functions of genes,

nzymes, membranes and metabolites in the whole organism.” (Pirt, 1991).

research also

ssesses protein activities, modifications, localization and interactions. By studying global

al., 1999) (fig.3).

proteolytic cleavage (Amson et al., 1996; Boguski and Schuler, 1995; Harry et al., 2000).

All these points could contribute to the presence of largely controversial observations using

genetic analysis (Rustin et al., 2000; Storm et al., 2002).

1.7 Proteomic analyses in aging research become obligatory

Based on the above consideration, the need of protein-level analysis arises because

phenotypes of senescent cells appear through functions

mtDNA might deleteriously affect protein synthesis, protein complex asse

respiratory enzyme function. An insufficient respiratory chain activity would consequentially

further increase oxidative stress in the cells. Such phenomenon could be largely observed

at the protein level, rather than in gene-based analysis. Therefore, research on age-

dependent protein alterations in the cells is necessary in clarifying the involvement of

mitochondria in the aging process.

As has been pointed out by Pirt, “The limiting factor is understanding not only the gene

structure but also the gene expression, which varies in a most complex way. The new

paradigm should be addressed t

The term “proteome” was first advocated by Marc Wilkins in 1995, defined as the entire

protein complement in a given cell, tissue or an organism (Anderson and Anderson, 1996;

Wasinger et al., 1995; Wilkins et al., 1996). In its wider sense, proteomic

patterns of proteins and their changes dynamically, proteomic research can improve our

understanding of system-level cellular behaviour.

High throughput two-dimensional protein electrophoresis coupled with peptide mass

fingerprinting analysis by mass spectrometry (MS) have become the most powerful

techniques for morden proteome analysis (Gras et

Protein separation with 2D

electrophoresis

2D-Gel pattern evaluation

Protein identificaton

Protein function annotation and

discussion

Protein separation with 2D

electrophoresis

2D-Gel pattern evaluation

Protein identificaton

Protein function annotation and

discussion

Fig.3: Strategy of differential protein display employing 2D-electrophoresis. Proteins are

separated by 2D-electrohporesis, and proteomic profiles are displayed as spots on the gel

slab. Quantitative and qualitative differences of corresponding spots among gels are

accessed according to differences of protein spot patterns. Subsequently, proteins in the

teome has been characterized

ochondrial cybrids

toulakis and Schlaeger,

roteome (Sickmann et al., 2003; Westermann and Neupert, 2003). Unfortunately, there is

spot of interest are digested with an endoproteinase and subjected to mass spectrometry

for identification by peptide mass fingerprinting.

1.8 Current state of mitochondrial proteomic research

The total number of different proteins or polypeptides in a mitochondrion was estimated to

be around 1000 (Lopez and Melov, 2002). Mitochondrial pro

on the human placental cells (Rabilloud et al., 1998), trans-human-mit

(Lopez and Melov, 2002), human neuroblastoma cell line (Foun

2003), human lymphoblastoid cell line (Xie, 2003) and on human heart mitochondria

(Taylor et al., 2002). Blue-native electrophoresis has greatly contributed to the

mitochondrial protein complex investigations (Schagger and von Jagow, 1991), while

Western immunoblotting remains to be an effective hypothesis-driven proteomic strategy.

So far, the largest proteomic study of purified mitochondria was performed by Sickmann on

yeast mitochondria, leading to the identification of 750 mitochondrial or mitochondria-

associated proteins with a coverage of up to 90% of predicted yeast mitochondria

still a lack of protein-level evidence of the accumulation of mutated mitochondria during the

aging process (Wei, 1998; Zhang et al., 1998).

1.8.1 Why there has been a lack of protein-level analysis

The current lack of proteomic level investigation of the mitochondrial theory of aging is

largely due to the lack of technical feasibility. First, it is likely that mutation rate of mtDNA is

very small respecting a certain type of mutation, thus, more sensitive methods will be

Until now, it has been

ficulty of such projects.

lly, large-gel 2D-

ibility to

reveal the majority of the cellular proteins (Klose et al., 2002).

protein identification. Other

ethods such like Blue-native electrophoresis and protein chip micro-arrays also provide

is study, we dared to challenge

e mitochondrial theory of aging with an in-depth proteomic study. The goal of this

needed (such as pre-fractionation, mitochondrial sub-proteomics).

difficult to accurately assess spontaneous mtDNA mutation rate in vivo in various organs

and tissues during aging (Linnane et al., 1990).

Secondly, proteomic field has not been as powerful as the genetic method since very

recently. Furthermore, many mitochondrial proteins of interest could be low abundant or

membrane protein. This further escalates the dif

1.8.2 Recent development of proteomic technology makes this study

tractable

Only recently, sensitivity and effectiveness of proteomic analysis is rising through a whole

new repertoire of high-throughput technical developments. Especia

electrophoresis analysis has now reached a technical state that offers the poss

New protein analytical methods such like mass spectrometry compatible for

macromolecules, computational tools, comprehensive databases for characterization of

molecular structures of proteins led to large-scale strategies in

the opportunity to directly test various proteomic aspects.

These made proteomics an attractive strategy of studying complex biology problems such

like aging, in order to gain additional knowledge of protein localization, protein interaction

and their influence on protein structure and function. In th

proteomic study was to obtain a more global and integrated view of aging biology by

studying dynamic protein networks, rather than certain protein individually.

1.9 Complexity of the system requires modeling

Upon obtaining the experimental data from proteomic studies, it would be necessary to

determine a sequence of events that can elucidate the primary factors responsible for the

cascade of complex events that accompany aging.

However, there is a range of factors that can potentially influence the abundance of mtDNA

mutation. These include mtDNA mutation rate, metabolic rate, efficiency of mtDNA repair,

propagation ability of mutant relative to that of normal mtDNA, influences of mutant mtDNA

on cell proliferation and function, as well as mitochondrial degradation rates. For instance,

a mitochondrial DNA mutation rate that is too small could make the influence of

mitochondrial mutation negligible in the aging process, thus counteract the hypothesis of

mitochondrial theory of aging.

Such a dynamic system defies understanding by verbal arguments alone. Quantitative

tools would be necessary to probe reliably into the details of the system. From a

mathematical point of view, the state of a dynamical system is specified by the

concentration values of all biochemical species in the reaction network. After formulating

the network into mathematical terms, its qualitative features can be demonstrated for

comparison with experimental data, as well as for the generation of new parameter values.

Sophisticated computational methods could be beneficial to interpret the complexity of

biological information in aging studies.

2 Aim of the Study

The objective of this study was to access age-related changes of mitochondrial proteome.

This could in turn reveal the influence of the proposed mitochondrial somatic mutation at

the protein level.

In the first stage of this study, an inbreed mouse strain (C57BL/6) was to be validated as

the model of human aging; the mitochondria isolation was to be optimized on the mouse

liver and bran tissues. In addition, the influence of donor organism age and sample

handling on the mitochondria isolation was to be investigated. In the mean time, proteomic

analysis tools and protein pre-fractionation methods were to be optimized, in order to

improve the resolution capacity of hydrophobic proteins in the 2D-electrophoresis analysis.

In the second stage, the mitochondrial proteins were to be analyzed by 2D-electrophoresis.

Mass spectrometry was employed for databank-based protein identification. Special

respect was paid on the mitochondrial-DNA-encoded proteins employing Western

immunoblotting. Since all 13 mitochondrial-DNA-encoded proteins are localized in the

mitochondrial respiratory chain protein complexes, preliminary application of Blue-native

electrophoresis was to be conducted for protein complex analysis. A question of particular

interest was whether there is tissue-specific property respecting mitochondrial protein

profile change in the aging process. This was to be investigated through the parallel study

of both brain and liver mitochondria.

Mathematical model has been employed previously to simulate the accumulation of

mutated mitochondria in the aging process. However, validation of such model was

hampered due to the lack of mtDNA mutation data. A parallel goal of this study was to

utilize the data gained in experiment to estimate the mtDNA mutation rate. For this

purpose, the hypothetical mechanisms involved in the mitochondrial theory of aging were to

be converted into a mathematical model. By fitting the experimental data into this model,

the mtDNA mutation rate of mouse tissue was to be calculated under the current modeling

setting.

3 Animals, Materials and Methods

3.1 Animals and ethnical agreements

Experimental protocol of this study was approved by the Charité institutional review

committee for the care of animal subjects and was performed in accordance with national

animal care guidelines (Tierschutzgesetz, Germany).

Specific pathogen Free (SPF) C57BL/6 mice (Mus musculus domesticus) were provided by

Charles River Germany (Sulzfeld, Germany) and Charles River France (Cedex France). All

experimental animals were treated humanely. They received standard feed ad libitum, free

access to water and human care. Animals were kept at constant temperature (22-24°C)

and humidity and had a 12:12 hour light-dark cycle before entering the study.

For the current aging study, healthy mice with mixed gender of the following aging stages

were used: newborn (0 to 2 weeks), 5-months, 10-months, 15-months, 20-months and 24-

months. Three 22-months mice were used for the obtainment of muscle biopsies and the

morphological control of mitochondria isolation from senescent tissues.

3.2 Materials and methods

All chemicals and reagents were purchased from Merck (Darmstadt, Germany) if not

otherwise indicated.

3.2.1 Organ obtainment

The animals were sacrificed by swift de-capitalization with a sturdy dissecting scissors. The

use of narcotic was avoided to eliminate possible influence on protein profile. Mouse liver

perfusion was based on the method of Seglen (Seglen, 1976) with modifications. The

abdomen was opened and the intestines gently moved aside to expose the hepatic portal

vein and inferior vena cava caudalis. The portal vein was cannulated by a cannula (i.d.

diameter 0.4mm, Terumo Leuven Belgium), the liver was perfused in situ with of 0.9% (w/v)

NaCl until slightly distended. The vena cava was then cut and the perfusion continued until

the liver was completely blanched. After 5ml of perfusion, the liver was freed from the

connective membranes, surgically removed from the body and rinsed with cold (0°C) 0.9%

(w/v) NaCl. Gall bladder was removed.

The obtainment of mouse brain tissue was conducted according to Klose (Klose, 1999).

The muscle and membranous tissue from the posterior part of the skull was removed over

the cerebellum. Using a pair of small surgical scissors, the skull was cut from the foramen

magnum to the olfactory bulbs and the flaps of skull were removed. Caution was exercised

to keep the tips of the scissors away from the midline of the cerebellum and cerebral

cortex. Subsequently, the mouse brain was gently pried from the skull, and immersed into a

beaker of cold saline (0.9% w/v NaCl). The trigeminal and optic nerves were trimmed away,

the bulbi olfactorii left intact. Spinal cord was cut off at the border to the rhombencephalon.

For the obtainment of muscle biopsy, hind limb muscles were carefully excised surgically.

Organ weight of brain and liver tissues was controlled directly after organ obtainment for

the calculation of mitochondrial yield. The extracted liver and brain tissues were either

directly used for mitochondria isolation or shock-frozen in liquid nitrogen for later uses.

Freshly extracted muscle tissue specimens without fixation were shock-frozen in

isopentane cooled by liquid nitrogen and were stored at -80°C for the histochemcal

demonstration of COX-activity.

3.2.2 Enzymatic histochemical staining of COX-activity

For the validation of the C57BL/6 strain as model of human aging, enzymatic histochemical

staining of cytochrome c oxidase activity was carried out on the mouse muscle biopsies.

Staining was performed on air-dried serial sections of muscle biopsies of young mouse (2

weeks) and old mice (22-months), according to Seligman (Seligman et al., 1968). The

oxidation of 3,3’-diaminobenzidine (DAB, Sigma-Aldrich Steinheim Germany) at the site of

cytochrome c oxidase activity results in a brown compound insoluble to ethanol.

8µm transverse frozen sections were cut with microtom Cyrostat (Microm, Walldorf

Germany) and sections were attached to a cover slip (SuperFrostPlus, R.Langenbrink,

Emmendingen Germany).

The incubation medium for the COX-staining consisted of 5mg 3,3’-diaminobenzidine

tetrahydrochloride, 9ml of 50mM PBS (pH 7.6), 20µg/ml catalase (Sigma-Aldrich Steinheim

Germany) solution, 10mg of practical grade cytochrome (Sigma-Aldrich Steinheim

Germany) and 750mg sucrose. Sections were incubated in the incubation medium in a

Columbia staining dish (Thomas Scientific, Swedesboro, NJ USA) for 60 minutes at room

temperature. After washing with three changes of deionized water (Millipore Schwalbach,

Germany), sections were dehydrated in alcohols with ascending concentrations: 50%,

70%, 80%, twice of 95% and twice of 100%, and subsequently cleared with 3 changes of

RotiClear (Carl Roth, Karlsruhe, Germany). Section mounting was conducted in synthetic

organic mounting medium (Permount, Fisher Scientific; Pittsburgh, PA).

Microscopic examination was carried out on a light microscope conjugated with digital

camera under bright field illumination (Carl Zeiss, Oberkochen, Germany). Fibre numbers

were determined on photographs (Magnification 100x and 200x) of 21 sections. Between

48 and 114 fibres were examined from each muscle biopsy.

3.2.3 Mitochondria isolation

The method of mitochondria isolation was adopted from Jungblut (Jungblut and Klose,

1985), with slight modifications. All procedures were carried out at 4°C, if not otherwise

indicated, to minimize protease activity. Except for newborn tissue, mitochondria of each

organ (either brain or liver) were isolated separately, without the pooling of different tissue

samples.

Since part of the materials (24-months mice sample) used in this study were collected

beforehand and frozen stored in our laboratory, a comparison of mitochondria isolation

from fresh and snap frozen materials was carried out.

All solutions used in the centrifugal separations were iso-osmotic solutions (260 mOsm) to

that of physiological condition of intact mitochondria, because indications of buoyant

density of the organelle are meaningful only in conjunction with the osmolality of the

medium employed.

3.2.3.1 Tissue homogenisation and crude mitochondria collection

In order to rupture the tissue and lyses the cell membrane without affecting most of the

organelle structure, homogenization was carried out at 0°C with a gentle hand-held teflon-

glass Dounce homogenizer (clearance 0.1mm) avoiding vacuum. For this purpose, the

freshly removed or swiftly thawed brain or liver tissues were first suspended in three

volumes of homogenisation medium (100mM KCl, 0.5M Tris-HCl, 5mM MgCl2, 1mM ATP-

Mg, 1mM ethylene glycol-bis(beta-aminoethyl ether)-N,N,N',N'-t etraacetic acid (EGTA),

0.08 v/v protease inhibitor cocktail CompleteTM table stock solution (solved in 2 ml 50mM

PBS, Molecular Biochemicals Roche, Mannheim, Germany), pH 7.5). Protease inhibitors

were added to prevent from protein degradation through biogenic proteases such like

serine proteases, cysteine proteases and metalloproteases. The substances were added

directly into the protein extracts of proper volume, and intensively stirred for 15 minutes. Up

and down movement were carried out until no significant big tissue pieces could be seen

by raw eye.

After homogenisation, the homogenate was centrifuged at 800xg (2400rpm, Kendro,

Hanau, Germany) for ten minutes to sediment nuclei and debris. The supernatant was

collected. This step was repeated once. Subsequently, the final “post-nuclear supernatant”

was collected and subjected to 10,000xg (10,600rpm, Type 50 Ti rotor, Beckman Coulter,

Krefeld, Germany) for 15 minutes to obtain the crude microsomal fraction.

The crude microsomal fraction that contain multiple kinds of microsomes (mitochondria,

lysosome, peroxisome, Golgi complex, etc.) was diluted with 7ml of 0.25M sucrose

solution, and subjected to a serial of three wash steps: 8000xg, 5000xg and 3500xg

(9600Upm, 7600Upm and 6300Upm, respectively, each for 15 minutes, Type 50 Ti rotor,

Beckman Coulter, Krefeld, Germany). This was for the remove of most microsomal bodies

other than mitochondria and lysosome.

After the last washing step, the supernatant was carefully removed by a Pasteur pipette

without disturbing the pellet. Generally, only lysosome co-pellet with mitochondria at

3500xg since their similar density to that of mitochondria (Jungblut and Klose, 1985).

3.2.3.2 Purification of mitochondria using gradient centrifugation

In order to further purify mitochondria (separate lysosome from mitochondria), Percoll

continuous gradient centrifugation was employed. The density gradient media Percoll

(Sigma-Aldrich Steinheim Germany) is an inert colloidal suspension of silica particles

coated with polyvinyl pyrrolidone (15-30nm in diameter), with virtually no osmotic effects

(Sims, 1990). The low viscosity of Percoll ensures a fast establishment of continuous

gradient body and quick centrifugal separation (Pascale et al., 1998; Sims, 1990).

For this purpose, approximately 700µl of suspended crude mitochondrial pellet (1:1 v/v,

suspended with 0.25M sucrose with 0.1mM EGTA, pH7.2) was carefully topped on 7ml of

30% (v/v) Percoll solution (in 0.25M sucrose with 0.1mM EGTA) in a polycarbonate ultra

centrifugation tube (Nalgene, Rochester, NY USA), avoiding any disturbance of the bottom

phase. The centrifugal tube was subjected to 100,000xg (37,400rpm, Type 50 Ti rotor,

Beckman Coulter, Krefeld, Germany) of ultra centrifugation for 15 minutes, with no

mechanic brake applied upon deceleration.

During the centrifugation, the Percoll solution builds a continuous gradient body due to the

migration of silicon particles in the strong gravity field. Meanwhile, the pellet materials co-

migrate to their corresponding density layer. Only the intact mitochondria migrate to the

density layer of 1.09-1.13 g/ml, while most lysosome and some broken mitochondria to the

density layer of 1.05 g/ml above the mitochondrial layer (Jungblut and Klose, 1985).

After the gradient centrifugation, the lower band was aspirated with care and was washed

with 0.25 M sucrose solution (10,000xg, 10 minutes Type 50 Ti rotor, Beckman Coulter,

Krefeld, Germany) to obtain the mitochondrial pellet. One sample was taken from the upper

band for microscopic control. An extensive dilution (1:5 v/v dilution of Percoll) using 0.25 M

sucrose solution was necessary in order to sediment mitochondria effectively. This was due

to the presence of remaining Percoll, which slowed down the centrifugation process by

increasing the density of the surrounding medium.

If required for morphological control analysis using electronic microscopy, ca. 20µg of the

pellet was immediately immersed in fixation solution of electronic microscopy. The

remaining pellet of intact mitochondria was snap-frozen in liquid nitrogen for further

proteomic analysis.

3.2.4 Electronic microscopic control of mitochondrial morphology

The electronic microscopic analysis was a collaboration with the Charité Institute of

Anatomy. The mitochondria isolated from young (5-months) and old (22-months) organism

(brain and liver tissues), as well as from fresh and frozen materials (brain and liver tissues)

were controlled for their morphology and purity.

Portions of isolated mitochondria were fixated with fixation solution containing 60mM PBS

and 2.5% glutaraldyhyde (v/v) at room temperature for 15 minutes. Mitochondria were then

centrifuged at 15,800xg (20,000rpm, Ti50 rotor) for 5 minutes and the supernatant was

discarded. The pellet was layered with fresh fixation solution and left overnight at 4°C. Two

additional washing steps were carried out in 60mM PBS (15,800xg, 10 min each, Type 50

Ti rotor, Beckman Coulter, Krefeld, Germany). The specimen were post-fixed overnight in

2% (w/v) osmium tetraoxide (in PBS, pH7.2), and washed twice in water (five minutes

each, Millipore Schwalbach, Germany).

Because the mitochondrial pellet displayed a diffused structure, it was necessary to embed

it in agar before further processing. For this purpose, the pellets were resuspended in 4%

(v/v) agar solution (Hobot et al., 1984). After the solidation of agar, the agar-embedded

specimen was cut into 1mm3 blocks.

Dehydration of the specimen was carried out in a graded series of ethanol: 50% for 10 min,

70% for 10min, 95% for 10 min and twice of 100% for 30min. After dehydration, specimens

were incubated in pure propylene oxide for 15 minutes, Epon/propylene oxide 1:2 for one

hour, Epon/propylene oxide 1:1 for one hour and in pure Epon overnight. The resin-

embedded specimen was heat polymerised at 60°C for 48 h.

Ultra-thin sections were cut on a Leica Ultramicrotome III (Leica, Wetzlar Germany), and

counterstained with 4% (w/v) aqueous uranyl acetate for 5 minutes, followed by 30

seconds of incubation with a 1:5 dilution of lead acetate solution. Electronic micrographs

were obtained using a transmission electron microscope (Carl Zeiss, Oberkochen,

Germany) operating at 80 kV. The incidence of mitochondrial morphology was calculated

by counting 100 mitochondrial cross-sections on six separate micrographs (10,000X

original magnification).

3.2.5 Sequential extraction of mitochondrial proteins

The goal of sample preparation was to introduce proteins into a solution compatible to

down-stream protein analyses, so as to enable high-resolution separation of the proteins.

During this procedure, proteins should be kept intact, preventing from adverb modifications

and degradations. In order to enrich low-abundant proteins and membrane proteins in the

down-stream analysis, mitochondrial proteins were first separated into three different

fractions.

The procedure for the sequential extraction of mitochondrial proteins was adopted from

previous studies with modifications (Molloy et al., 2001; Ramsby et al., 1994; Weiss et al.,

1992) (fig.4).

Mitochondrial

isolation mitochondrial

disruption

Modified 2D-

PAGE

Fraction III

(methanol/chlo-

roform extraction)

Fraction I

(Tris-buffer

extraction)

Western

Immuno-

blotting

2D-PAGE

analysis

Fraction II

(Triton-buffer

extraction)

Mitochondrial

isolation mitochondrial

disruption

Modified 2D-

PAGE

Fraction III

(methanol/chlo-

roform extraction)

Fraction I

(Tris-buffer

extraction)

Western

Immuno-

blotting

2D-PAGE

analysis

Fraction II

(Triton-buffer

extraction)

Fig.4: After isolation of mitochondria from mouse brain and liver tissues, the pooled

mitochondrial samples were pre-fractionated into three different fractions using sequential

extraction strategy. Different fractions were analysed with different analysis methods.

For the simultaneous processing of different mitochondrial samples, mitochondrial pellets

from the same aging stage and tissue origin (brain or liver) were pooled into mixed

samples. The mitochondrial membrane envelope was first disrupted by the addition of 0.5

volume of distilled water (Millipore Schwalbach, Germany), which acts as hypoosmotic

medium. Afterwards, three times of freeze-thaw cycle (37°C to liquid nitrogen -156°C) were

carried out to ensure the lyses of mitochondria (Jungblut and Klose, 1985).

3.2.5.1 Tris-buffer extraction

As the first step, 0.5 volume of 100mM Tris buffer (pH7.3) was added to extract soluble

proteins. The mitochondrial whole lysate then underwent ultracentrifugation at 100,000xg

for 45 minutes (Type 50 Ti rotor, Beckman Coulter, Krefeld, Germany) to sediment crude

mitochondrial membrane pellet. Supernatant out of this centrifugation step was designated

as “Fraction I”.

It was expected that the most soluble proteins were released from mitochondrial matrix and

intermediate space by hypotonic and mechanic disruption, and that this “Fraction I”

contained mainly hydrophilic proteins from mitochondrial matrix and membrane-associated

hydrophilic protein subunits (Molloy et al., 2001; Weiss et al., 1992). This fraction was

further analyzed using large-gel 2D-electrophoresis method developed in our laboratory.

3.2.5.2 Triton buffer extraction

The crude mitochondrial membrane pellets were re-suspended with one volumes of 50mM

Tris buffer (pH 7.4) containing 1mM dithiotreitol (DTT) and 0.1% (v/v) Triton-X100, stirred

for 2 hours at 4°C, and centrifuged for one hour at 226,200xg (50,000rpm, Type 50 Ti rotor,

Beckman Coulter, Krefeld, Germany) at 4°C. This was for the purpose of releasing the

membrane-associated proteins of intermediate solubility (Santoni et al., 1999). The

resulting supernatant was named “Fraction II” in the context of this study.

3.2.5.3 Methanol-chloroform protein extraction

Finally, the remaining pellet was subjected to methanol-chloroform extraction in order

extract membrane-bounded proteins (Yerushalmi et al., 1995). For this purpose, four

volume of deionised water (Millipore Schwalbach, Germany) was first added to the

membrane pellet and suspended vigorously (Wessel and Flugge, 1984). This suspension

was then aliquoted at 150µl portions in 1.5ml tubes. 600µl of methanol and 150µl of

chloroform were added to each aliquot. After incubation on ice for 20 minutes under

occasional vortexing, 450µl water (Millipore Schwalbach, Germany) was added to assist

the phase separation.

The tubes then underwent centrifugation at 14,000xg (12,600rpm) for 2 minutes in order to

separate different phases. After the centrifugation, the proteins were in the intermediate

layer, with chloroform layer under it and the methanol-rich layer above it, which was

carefully removed.

This protein pellet was washed once by the addition of 600µl methanol, so that protein

pellet was left at the bottom of the tube. The ultimate pellet was designated as the “Fraction

III” throughout this work. It supposed to contain mainly hydrophobic membrane proteins

(Molloy et al., 1998). The protein pellets were allowed to dry under cold Argon for 10

minutes, before it was resuspended in either Laemmli buffer or in deionised water (Millipore

Schwalbach, Germany) for protein concentration analysis.

3.2.6 Protein concentration analysis

Protein concentration measurement was carried out for “Fraction I”, “Fraction II” and

“Fraction III”, accomplished before the addition of detergents or catropes. Bininchonicic

acid (BCA, Perbio, Rockford, IL, USA) protein assay was employed.

This method combines the reduction of Cu2+ to Cu+1 by protein in an alkaline medium (the

biuret reaction) with a colorimetric detection of the cuprous cat ion (Cu+1) using a reagent

containing bicichoninic acid. The chelating of two molecules of bininchonicic acid with one

cuprous ion forms the purple-coloured reaction product of this assay (fig.5).

Fig.5: Chemical reactions involved in the protein concentration assay using bicinchoninic

acid (cited from http://brahms.chem.uic.edu/~chem455/frames.html, University of Illinois at

Chicago). Biuret reaction generates single-valent cuprous cat ion. Single valent copper ion

reacts with two molecules of bicinchoninic acid to form BCA-Cu+1 complex, which has a

maximum absorption at the wavelength of 562nm.

The water-soluble chromophore exhibits a strong absorbance at 562 nm that is linear with

increasing protein concentration over a working range of 20µg/ml to 2000 µg/ml (Smith et

al., 1985).

The protein standards were prepared by diluting a 2.0mg/ml bovine serum albumin (BSA)

stock reagent serially with deionised water (Millipore Schwalbach, Germany), yielding a

working range from 20 µg/ml to 1000µg/ml. Deionised water was used as blank.

The BCA working reagent was prepared by mixing 50 volumes of “reagent A” (containing

BCA) with one volume of “reagent B” (containing CuSO4). In a 96-well micro array plate,

two times of 20µl of either sample (diluted 1:10 and 1:100 with deionised water (Millipore

Schwalbach, Germany) or standard was mixed with 300µl working reagent, and incubated

for 30 minutes at 37°C. After cooling to room temperature, the plate was read at 570 nm

with a spectrophotometer (Amersham Biosciences Freiburg, Germany).

3.2.7 Sample preparation for 2D-electrophoresis

A highly standardardized sample preparation protocol introduced by Klose was employed

for the sample preparation of 2D-electrophoresis (Klose, 1999; Klose and Kobalz, 1995).

The key feature of this protocol is to solubilize proteins and keep them soluble by the

addition of zwiter-ion detergent and high concentrations of catropes.

200µg of protein extraction (“Fraction I”, “Fraction II” and mitochondria pellet) was

accommodated with the following substrates:

• 50mM phosphate buffer, pH7.1

• 100mM KCl

• 10% glycerol (v/v)

• 4% (w/v) 3-[(3-cholamidopropyl)-dimethylammonio]-propan-sulfonate (CHAPS) for

mitochondrial pellets

• 1µM Pepstatin A (pre-solved in ethanol, Sigma-Aldrich, Steinheim Germany)

• 1.4µM phenylsemthylsulfonylfluorid (PMSF, Sigma-Aldrich, Steinheim Germany)

(pre-solved in ethanol)

• 0.08 v/v protease inhibitor cocktail CompleteTM table stock solution (solved in 2 ml

50mM PBS) (Molecular Biochemicals Roche, Mannheim, Germany).

For mitochondrial total protein extract, a sonification treatment was carried out to maximize

the membrane structure disruption and protein solubilization (Klose, 1999). Sonification

was performed at 0°C for six times of 10 seconds (Ultrasonic bath, Bandelin, Berlin

Germany), with 40-45 seconds of intensive stirring between each application and one

minute keeping on ice. To enhance the sonification strength, 0.034 part (v/w) of glass

beads with low sodium content (2.5 ± 0.05 mm diameter, Wolf Glaskugeln GmbH, Mainz,

Germany) was added during the treatment (Klose, 1999). The glass beads were removed

following the final sonification step.

Nucleic acids in the sample has been indicated to interfere with the isoelectric focusing

process (O'Farrell, 1975). Upon further stirring at 4°C, 0.025 part (v/w) of Dnase

(Benzonase) was added into the mitochondrial total protein extracts in order to degradate

the mitochondrial DNA.

Subsequently, protein homogenates of all fractions were adjusted with 6M urea and 3M

thiourea. After dissolving, 0.01 part (v/w) of 70mM DTT was added and stirring was

continued at room temperature for another 30 minutes. A high concentration of catropes,

such as urea or thiourea, function through the partial breaking of hydrogen bunds in the

solution. The reducing agent DTT was essential for ensuring the breakage of disulfide

bridges, which would otherwise oppose protein denaturation and prevent saturation of the

polypeptide with sodium deodecylsulfate (SDS) in the following procedure. These protein

denaturants could improve the unfolding of the protein secondary structure and thus keep

the proteins in solution (Herskovits et al., 1985).

Finally, 1% (v/w) of ampholyte mixture Servalyte 2-4 (Serva, Heidelberg, Germany) was

added as carrier ampholyte. The samples were stored at –80°C until Isoelectric focusing

analysis.

3.2.8 Sample preparation for Western immunoblotting

20µg of “Fraction III” protein pellet obtained from methanol-chloroform extraction was

suspended with 70µl Laemmli sample buffer (25mM Tris-buffer, 2% (v/v) SDS, 192mM

DTT, pH 8.3) accommodated with10% (w/v) glycerol (Laemmli et al., 1976). The samples

were heated at 95°C for 5 minutes to ensure the protein denaturation. After quick cooling

on ice, the sample was ready for SDS-PAGE resolutions.

3.2.9 Sample preparation for Blue-native electrophoresis

Mitochondrial pellet corresponding to 1mg protein was suspended in 100 µl of extraction

buffer comprising 0.75M aminocaproic acid and 50mM 2-[Bis-(2-hydroxyethyl)amino]-2-

(hydroxymethyl)propane-1,3-diol (BisTris). 12.5µl n-dodecyl-β-D-maltoside (10% w/v) was

added to the suspension, in order to solubilize mitochondrial inner membrane proteins

(Schagger and von Jagow, 1991).

Following incubation on ice for 20 minutes with occasional vortexing, samples were

centrifuged at 14,000xg (Heraeus, Hanau, Germany) for 10 minutes though a 100kDa

molecular filter (Microcon, Millipore Schwalbach, Germany). 6.3µl of 5% (w/v) suspension

of Coomassie brilliant blue G-250 in aminocaproic acid (0.5M) was added. Samples were

kept on ice until Blue-native electrophoresis analysis.

3.2.10 Large-gel 2D-electrophoresis

Large-gel 2D-PAGE analysis was applied on whole mitochondrial protein extract,

mitochondrial protein “Fraction I” and “Fraction II” protein samples. At least three times of

2D-PAGE analysis were carried out for each sample in order to access the reproducibility

of the method and to exclude the artefact-deduced gel-to-gel differences. The large-gel 2D-

electrophoresis analysis of mouse brain total protein extract were carried out in our

laboratory in the frame of a parallel project.

3.2.10.1 Principle of 2D-electrophoresis

The current modern 2D-electrophoresis was developed independently by Klose and

O’Farrel (Klose, 1975; O'Farrell, 1975). It is a combination of isoelectric focusing with SDS-

polyacrylamide gel electrophoresis. This method provides the opportunity to separate

proteins from a highly complex protein mixture and make proteins accessible for further

biochemical analysis. The modern large-gel 2D-electrohporesis allows the visualization of

as many as 10,000 of protein-spots on a single gel (Klose and Kobalz, 1995).

In the isoelectric focusing electrophoresis, pH gradient can be established by applying a

mixture of ampholytes (which is a mixture of synthesized oligoamino, oligoacarboxylic

acids) with different isoelectric points to a polyacrylamide gel in an electric field. Upon the

application of voltage, the carrier ampholytes stack according to their pI, and an increasing

pH gradient is established within the gel. Migration of the proteins continue until all the

components of the system reach a steady state, i.e., their isoelectric points. The rates of

protein migration depend on the charge density (the ratio of charge to mass) of the

individual proteins.

In the second dimension separation, the SDS-polyacrylamide gel electrophoresis (SDS-

PAGE) utilizes SDS (CH3(CH2)11SO4-Na+) as an anionic detergent. Polypeptides bind SDS

to the main chains in a constant weight ratio (1.4g SDS per gram protein) (Lottspeich and

Zorbas, 1998). Due to the much higher anionic charge of SDS comparing to that of

proteins, the charge difference between proteins can be ignored. All proteins have the

similar charge to mass ratios at pH8.4.

In addition, upon binding to SDS, polypeptide chains are forced into extended cylindrical

conformations with a constant diameter. SDS treatment thus eliminates the effect of

differences in protein shape. So that chain length, which reflects mass, is the sole

determinant of the migration rate of proteins in SDS polyacrylamide electrophoresis. Thus,

the electrophoresis mobility of the proteins is only dependent on their molecular weight

(Lottspeich and Zorbas, 1998).

3.2.10.2 Procedure of isoelectric focussing electrophoresis

The isoelectric focusing was performed according to protocol of Klose (Klose and Kobalz,

1995). High precision capillary glass tubes with internal diameter of 1.5 mm and a length of

20cm were used for the isoelectric focussing separation of “Fraction I” and “Fraction II”

proteins, while the 40cm IEF gels (0.9mm diameter i.d.) were used for mitochondrial total

protein reference gel.

The gel solution contained 3.5% (w/v) acryamid, 0.3% (w/v) N, N’-methylene bisacrylamide

(BisAcryamide), 9M urea and 2% (v/v) carrier ampholyte mixture, which was a mixture of:

• one part of Pharmalye pH3.5-10.0 (Pharmacias, Uppsala, Schweden),

• one part of Servalye pH2.0-11.0 (Serva, Heidelberg Germany),

• three part of Pharmalye pH4.0-6.5 (Pharmacias, Uppsala, Schweden),

• two part of Pharmalye pH5.0-8.0 (Pharmacias, Uppsala, Schweden), and

• one part of Pharmalyte pH6.5-9.0 (Pharmacias, Uppsala, Schweden).

In order to resolve more hydrophobic proteins on the 2D-PAGE gels, we included 2% (v/v)

Triton-X100 in the isoelectric focusing gel solution (Stephenson et al., 1980) for the

analysis of “Fraction II” protein extracts in this study.

Gel solution was filled into the tubes using accurately fitted nylon strings as plungers. The

gels were left undisturbed for 30 minutes before the nylon strings were removed, allowing

for polymerisation at room temperature. A further polymerisation for 3-4 days at room

temperature was allowed before use.

Isoelectric focussing running buffer were made of purest water (Millipore Schwalbach,

Germany) with electric resistance of 18.2mΩ. The anode buffer contained 4.25% (v/v)

orthophospho acid and 2M urea (pH2); the cathode buffer contains 5% (v/v) ethylendiamin,

9M urea and 5% (w/v) glycerol (pH11).

200µg protein sample (“Fraction I” and “Fraction II”) was applied on each 20cm 1.5mm

diameter gel for preparative silver staining, whereas 100µg of total mitochondrial protein

was applied on 40cm 0.9mm diameter gel for analytical silver staining. Prior to the loading

of protein sample on the anodic end of IEF gel, a Sephadex (Sigma-Aldrich, Steinheim

Germany) mixture (with 2% (v/v) carrier ampholyte mixture described above, 70mM DTT

and 9M urea) was loaded to a height of 2mm. This Sephadex layer has shown to have

positive effect on the isoelectric focussing, as it sieves up the possible granules in the

sample, and in turn improves the entry of soluble samples into the IEF gel.

The isoelectric focusing was carried out at a series of increasing voltage: 100V for 1 hour,

200V for 1 hour, 400V for 17.5 hour, 650 V for 1 hour, 1000V for 30 minutes, 1500 V for 10

minutes and 2000V for 5 minutes. Immediately after isoelectric focusing electrophoresis,

the gels were expelled into the equilibration solution containing 125mM tris (hydroxymethyl)

aminomethane, hydrochloride (Tris-HCl, pH 6.8), 40% (w/v) glycerol, 65mM DTT and 3%

(w/v) SDS. Gel strips with the length of 40cm were cut at the middle. After 10 minutes of

incubation under gentle shaking, the gel strips were stored at –80°C until SDS-PAGE

analysis.

3.2.10.3 Second dimension of 2D-electrohporesis

The sodium dodecylsulfate polyacryamid gel electrophoresis was carried out in the format

of 23.2cm x 30 cm x 1.0 cm gel cassette. The gel solution consisted of Laemmli buffer

(25mM Tris, 192mM glycine, pH 8.3) containing 15% acrylamide (w/v) and 0.2%

BisAcrylamide. Laemmli buffer was also used as electrophoresis buffer. One drop of

bromophenol blue dye was added in the anode buffer as running state indicator.

Thawed at room temperature, the first dimension gel was gently transferred onto the

surface of the SDS-PAGE gel, preventing the stretching of the gel and the introduction of

air. Running buffer containing 1% (w/v) agarose was overlaid to restrict the movement of

the IEF gel.

Electrophoresis was carried out at 15°C, first at 85mA for 15 minutes and then at 120mA

for approximately five hours. This procedure was stopped when the bromophenol blue dye

front reached 2cm short of the lower edge of the gel plate. After SDS-PAGE, the gels were

transferred into fixation solution for staining procedure.

3.2.11 Silver staining

The 2D-gels bearing proteins separated by isoelectric focusing and SDS-PAGE were

stained using either analytical silver staining or preparative silver staining protocol.

3.2.11.1 Analytical silver staining

In the silver staining procedure employing silver nitrate, silver ion binds to the amino acid

side chains, primarily the sulfhydryl and carboxyl groups of proteins (Coligan et al., 1995).

Since silver ions complexed with the protein undergo faster reduction than free silver ions,

particles of colloid silver (between 20 and 80 nm in diameter) form preferentially at the site

of proteins on the surfaces of the gel, leading to the deposition of silver grains.

The analytical silver staining is a sensitive staining method with a detection limit between 1

and 10 ng protein (Rabilloud et al., 1992; Switzer et al., 1979). This protocol was used to

produce a master 2D-PAGE gel pattern for brain and liver mitochondrial total protein

extract (Heukeshoven and Dernick, 1988; Jungblut and Seifert, 1990).

Directly after the SDS-PAGE separation, the proteins on the gels were fixed overnight with

fixation solution containing 10% (v/v) acetate and 50% (v/v) ethanol under continuous

shaking. This step immobilizes the proteins in the gel or retards their diffusion to a large

extend, while at the same time removes substances such as SDS or glycerol that might

interfere with the staining procedures.

Afterwards, the gels were incubated in a solution containing 30% (v/v) ethanol, 0.25M

Sodium acetate, 0.8mM sodium thiosulfate and 20ml/l glutaraldehyde. Sodium thiosulfate

was added to create latent images of protein spots by the precipitation of micro granules of

silver sulfide, while the inclusion of glutaraldehyde promotes silver reduction (Rabilloud,

1990).

After the subsequent two washing steps in water (each for 10 minutes), the gels were

incubated in silver-containing solution (0.1% w/v) for 45 minutes, with the addition of 0.01%

(v/v) formaldehyde, which assists the silver ion reduction and the precipitation of metallic

silver.

After a short rinse (one minute) with water, the gels were allowed to develop color (allow

the reduction of silver ion to silver) in the presence of 2.5% (w/v) sodium carbonate, 0.02%

(v/v) formaldehyde. 0.5mM Thimerosal (Mercury-[(o-carboxyphenyl)thio]ethyl sodium salt,

Sigma-Aldrich, Steinheim Germany) was added as a preservative and colour enhancer.

Since it was essential that the color development be carried out in an absolutely

transparent solution, one change of development solution was carried out after one minute.

After the desired intensity of staining was achieved, this process was interrupt by the

addition of 50mM of tetrasodium ethylenediamine tetraacetate (Titriplex III). Depending on

the colour intensity, the colour development duration ranged between 3 to10 minutes.

3.2.11.2 Preparative silver staining

Mass preparative silver staining was performed using the protocol of Schevshenvo

(Shevchenko et al., 1997) with slight modifications (Giavalisco, 2003). Compared to the

traditional silver staining protocol (Klose and Kobalz, 1995; Rabilloud, 1990; Rabilloud et

al., 1992; Swain and Ross, 1995), the Schevchendo method retains the possibility of

diverse down-stream protein microanalysis. This is due to its omit of the sensitisation

treatment with glutaraldehyde, which is known to attach covalently to the proteins through

Schiff-base formation with the α and ε-amino groups (Shevchenko et al., 1997).

The fixed gels were first customerized in 30% ethanol (v/v) for 10 minutes, and then

sensitised by a short incubation (one minute) with sodium thiosulfate (0.01% w/v). Two

rinses with deionized water (Millipore Schwalbach, Germany, one minute each) were

carried out to remove the unbound sodium thiosulfate. The incubation of silver (0.15% w/v)

was carried out for 45 minutes at the absence of formaldehyde. The colour development

procedure was identical to that of analytical silver staining.

3.2.12 2D-gel evaluation

The gel image evaluation remains a critical point of 2D-electrophoresis analysis. In this

study, both visual gel evaluation and the gel image processing software were employed, in

order to reduce artifacts.

For this purpose, the gel images were first digitalized using a densitometer (Umax Mirage-II

DIN A3 scanner, Willich, Germany) with the resolution of 300 dots per square-inches (dpi).

Respecting the similar gel patterns, the majority of spots were manually assigned to their

counterparts in each of the age group according to their relative position on the gel pattern.

Each set of homologous spot on different gels built a “super spot”. Comparison in each

super spot was carried out respecting their intensity, their outward appearance and the

variation type. Four categories of variation types were evaluated according to previous

study (Kaindl, 2001):

• Presence/absence variants: whether a spot is visible;

• Amount variants: changes in size and intensity of protein spots;

• Mobility variants: alteration of spot position on the gel caused by possible changes

of protein charge, molecular weight and conformation;

• Splitting variants: one of the two spots involved being split into to two or more spots.

• According to these findings, the protein spot changes during the aging process was

postulated.

The comparison of brain mitochondrial 2D-gels to the total brain protein 2D-gels was

carried out by Dr. Sagi in our laboratory.

For the quantitative analysis of protein concentration difference in different age groups, all

gels were analysed using Proteomweaver (version 2.1, Definiens, Munich, Germany),

which is a commercially available software for two-dimensional gel image analysis. Manuel

correction was carried out after automatic spot recognition.

For this purpose, one average gel was generated for each age group by the software

package in order to reduce the fluctuation of gel-to-gel variation inside the same group.

Average gels from six different age groups were matched to each other automatically, and

the gel-matching pattern was controlled and edited manually to minimize software-induced

artifacts.

Spot volumes were calculated with build-in feature for the spot quantification, which applied

a fixed multiple of the Gaussian radius of the spot as background intensity function (Users’

manual, Definiens, Munich, Germany). After spot volume calculation, the spot information

was extracted and output by the ProteinWeaver. Properties of each “super spot”, which

was a serious of matched spot on different gels, was presented in a role, showing the

super spot identity, Cartesian location, spot surface area and spot volume respecting the

integration of intensity value of each pixel.

Spot intensities were normalized by calculating the “relative intensity”, which was defined

as percentage of the total spot volume on its parent gel. Coefficients of variance

(CV=standard deviation/mean) of spot intensities in “super spots” were used to control the

quality of automatic gel evaluation procedure (Challapalli et al., 2004). Scatter plots and

correlation coefficients of relative intensities from different gel-pairs were calculated using

StatView software package (Abacus, NC, USA). Ultimately, spot differences detected by

visual gel evaluation were taken to statistical tests to determine the statistical significance

(α<5%).

3.2.13 Protein identification

Presently, the most commonly used technique for protein identification is “peptide mass

fingerprinting” employing mass spectrometry. This involves the generation of peptides

using specific proteolytic enzymes (such as trypsin, chrymotrypsin), the determination of

peptide masses and the matching against a theoretical spectrum of peptide fragments

calculated from databases of known protein sequences (Mann and Wilm, 1994). Positive

ionisation is generally used for protein and peptide analyses, because peptide possesses

functional groups that readily accept protons (H+):

R-NH2 + H+

→

R-NH3+ (R: residual of protein or peptide) (eq.1)

The Matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry (MALDI-

TOF-MS) (Fernandez et al., 1998) and the electrospray ionization iontrap mass

spectrometry (ESI-Iontrap-MS) analysis were carried out by Mass spectrometry core facility

of our institute.

MALDI is based on the bombardment of sample molecules with a laser light to bring about

sample ionisation (Hillenkamp et al., 1991; Mann et al., 2001). The matrix substance

transforms the photonic energy into excitation energy, which leads to sputtering of analyte

and matrix ions from the surface of the co-crystal.

Electrospray Ionisation (ESI) is one of the Atmospheric Pressure Ionisation (API)

techniques suitable for the analysis of polar molecules ranging from less than 100 Da to

more than 1,000,000 Da in molecular weight (Wilm and Mann, 1996). The liquid sample is

physically sprayed into an acidic environment. Rapid desolvation of sample droplet leads to

the ionisation of sample molecules before their entering into a mass analyser.

The ions are then accelerated in an electric field and fly towards a detection board. The

speed of flight depends on the mass-to-charge ratio. The detector monitors the ion current,

amplifies it and the signal is then transmitted to the data system where it is recorded in the

form of mass spectra. The mass spectrum gives information about mass-to-charge ratio

(m/z) of each component, and the relative abundance of the various components in the

sample.

3.2.13.1 In-gel trypsin digestion

The aims of in-gel tryptic digestion were two fold. The first is to eliminate the chemical

substances (salts, detergents) remained from previous steps of sample preparation, 2D-

electrophoresis and staining procedure, which could otherwise disturb the mass spectra

and sensitivity. Another goal of tryptic digestion is to cut the protein into certain peptide

mixtures according to the amino sequence pattern, so that the peptide mass fingerprinting

analysis can be carried out.

Protein spots of interests were manually excised from the 2D-gel with a hand-held spot

picker before proteolytic digestion using trypsin. Trypsin has the specific proteolytic activity

for peptide bund at C-terminal lysine or arginine, provided that prolin is not the subsequent

amino acid. It is useful for mass spectrometric studies because each proteolytic fragment

contains a basic arginine or lysine amino acid residue, and thus is suitable for positive

ionisation of mass spectrometric analysis.

The spots were desalted by the pendelling addition of 100µl acetonitrile (CH3CN) and 100µl

of 100mM ammoniumbicarbonate (NH4HCO3) for the total of six times, each with an

incubation period of 10 minute at 37°C. This ensures the comprehensive removal of salt

and detergents in the gel pieces.

The liquid phase was then removed and the gel pieces completely dried in a vacuum

concentrator (Eppendorf, Hamburg, Germany). This ensures the complete permeation of

the trypsin into the gel piece and thus prevent from extensive self-digestion of trypsin.

Trypsin in-gel digestion was carried out in the trypsin solution (Seq. Grade modified

Porcine trypsin, Promega WI, USA) containing 12.5ng/µl Trypsin in 50mM NH4HCO3 (pH8),

at 37°C overnight, to ensure complete digestion.

After trypsin digestion, peptides were recovered by combining the liquid phases from

several extractions of gel pieces with 50% aqueous acetonitrile containing 5% formic acid.

Another desalting treatment was carried out using ZipPlate (Millipore Schwalbach,

Germany) in case MALDI-TOF-MS analysis was carried out as the next step.

3.2.13.2 MALDI-TOF-MS analysis

Anchor Chip technology was applied for MALDI-TOF analysis. AnchorChips are equipped

with hydrophilic patches (“anchors”) in a hydrophobic surrounding causing the relatively

hydrophilic analyte to concentrate on the anchors (Karas and Hillenkamp, 1988). For this

purpose, 1.5µl matrix material 2,5-dihydroxybenzoic acid (DHB) solution (3.3mg/ml DHB

solubilzed in 90% (v/v) acetone and 0.05% (v/v) trifluoroacetic acid) was mixed with 1.5µl of

the sample and allowed to co-crystallize on the anchored position.

MALDI-TOF mass spectra of the peptide mixture were obtained using the Bruker Reflex IV

MALDI-TOF mass spectrometer (Bruker, Bremen, Germany), operated in the reflector

mode in order to increase mass resolution. Signals corresponding to mass-to-charge (m/z)

ranging from 0 to 3500 were monitored.

Each spectrum was the cumulative average of 50 to 100 laser shots (nitrogen laser,

337nm). An internal calibration using the monoisotopic peaks of trypsin self-digestion

products (residues 108-115, [M+H]+=842.509 Da and residues 58-77, [M+H]+=2211.104

Da) as internal standards was carried out for each measurement.

The XMASS/NT software package (Bruker, version 5.1.16) was used for the data

processing. Peptide peaks were de-isotoped and those exceeding 5% of full scale were

submitted for database searching.

3.2.13.3 ESI-MS analysis

Nano-electrospray ionisation (Wilm and Mann, 1996) coupled to a reversed phase high

pressure liquid chromatography (HPLC) was employed in this work (LCQ Deca ion trap

mass spectrometer, ThermoFinningan).

10µl of the sample (ca. 75 fmol of each peptide) was injected into a pre-packed C-18

column (75µm inner diameter, LCPacking, Amsterdam, Netherlands). The separation of the

peptides was performed with a gradient of 2-50% acetonitril containing 0.1% formic acid at

a flow rate of 0.2µl/min. After the HPLC separation, a voltage of 1000 to 1200 V was

applied on the gold-plated vial situated within the ionisation source of the mass

spectrometer, resulting in sample spraying and ionisation.

3.2.13.4 Databank-based protein annotation

The obtained spectra of peptide masses were analysed by searching though databases to

match the corresponding proteins. This is achieved through the comparison of in silico

digestion product of protein in the non-redundant sequence database (NCBInr) with the

help of online search engine Mascot (http://www.matrixscience.com). Several parameters

were set before the Mascot search, including: the taxonomy of the specimen (mammalian),

the used protease (trypsin), the number of accepted missed cleavages (one), the mass

deviation tolerance (1.0 Da), and possible modifications (e.g., oxidation of methionine

residue in a polypeptide increases its mass by 18 Da).

A protein was considered to be directly identified, in case at least four measured peptides

match the in silico digestion peptide mixture and the probability based Mowse score was

higher than the threshold value corresponding to p<0.05 (Pappin et al., 1993).

Respecting the spot homology in the gel evaluation, counterpart spots in a “super spot”

were considered to be indirectly identified, if one of them had been directly identified

(Kaindl, 2001).

3.2.13.5 Membrane protein prediction and protein sequence alignment

In order to identify putative membrane proteins, all the identified proteins were subjected to

the Gravy value calculation and the prediction of putative trans-membrane domain using

SOSUI tool (http://sosui.proteome.bio.tuat.ac.jp).

Gravy score, or general average hydrophobicity score, or, is a theoretical measurement of

protein hydrophobicity according to their primary amino acid sequence. It is calculated as

an arithmetic mean of the hydrophobicity of all amino acids of a protein sequence (Kyte

and Doolittle, 1982). Integral membrane proteins typically have higher GRAVY scores than

do soluble proteins. By definition, the Gravy factor does not consider any influence of

protein secondary or higher-level structure. The SOSUI tool predicts the trans-membrane

helices by calculating the hydrophobicity of the amino acid charges and the sequence

length of a candidate peptide (Hirokawa et al., 1998). Protein-protein sequence alignment

was performed using Basic Local Alignment Search Tool (protein-protein, BLASTp

(www.ncbi.nlm.nih.gov/BLAST/).

3.2.14 Western immunoblotting

Western immunoblotting employed in the current study involves transferring protein bands

from the polycrylamide gel onto a nitrocellulose (Amersham Hybond ECL) membrane by

electrophoresis (Towbin and Gordon, 1984), and the subsequent immuno-detection of

protein of interest.

SDS-PAGE separation for Western immunoblotting analysis was carried out using the

buffer system of Laemmli (25mM Tris, 192mM glycine, 2% SDS (v/v), pH 8.3) (Laemmli et

al., 1976). Protein content was normalized by translocase of outer mitochondrial membrane

20kDa subunit (TOM20), which is part of the mitochondrial protein translocase machinery

with one trans-membrane domain.

Samples were run though 4% loading gel (pH 6.8) and 15% running gel (pH 8.4) in the

presence of 0.1% (v/v) SDS. Molecular weight markers (Rainbow, Amershan Biosciences,

Freiburg, Germany) was employed for the confirming of protein transfer and the molecular

weight orientation.

For the electro-blotting of proteins onto a nitrocellulose membrane (Amersham Hypond C),

a two-pH Tris-glycine “semi-dry” electrophoretic transfer system containing 20% methanol

was used. The pH value of the cathode buffer was 9.5 and the pH at the anode was 10.4,

thus creating a pH-gradient which facilitates the electro-transfer of proteins. The use of

methanol was shown to increase the binding capacity of nitrocellulose for proteins

(Timmons and Dunbar, 1990).

The gel and membrane were saturated with transfer buffer and were stacked together

horizontally between buffer-saturated filter paper pads (10mm thickness), then sandwiched

between both planar electrodes. The electrodes were separated solely by the thickness of

the stack, creating a uniform strength for protein transfer. Blotting was performed at

0.8mA/cm2 for two hours. Analogue to SDS-PAGE, The protein mobility is a function of

molecular weight, with the larger proteins being transferred more slowly.

After transfer of the proteins from the gel to the membrane, the remaining protein-binding

sites on the membrane were blocked overnight at 4°C in 3% bovine serine albumin in a

solution containing 10mM Tris-HCl, 133mM NaCl, 0.1% (v/v) Tween 20 (pH7.4) to avoid

non-specific binding of the antibodies or detection reagents in subsequent steps. BSA free

of endogenous peroxidases was used for this purpose.

Immunoblots were incubated for one hour at room temperature with monoclonal mouse-

anti-human cytochrome c oxidase subunit I IgG (dilution 1:100, Molecular Probes,

Göttingen Germany), and polyclonal rabbit-anti-mouse TOM20 IgG antibody (Santa Cruz

biotechnology, dilution 1:200) under gentle wagging. After intensive washing procedure

(200mM Tris-Base, 9% (w/v) NaCl, 1% (w/v) Tween 20), one hour incubation was carried

out with horseradish-peroxidase-conjugated secondary antibody (goad anti-mouse-IgG and

goad anti-rabbit IgG, dilution 1:2000), which was included in the ECL detection kit

(Amershan Biosciences, Freiburg, Germany).

Horseradish peroxidase (HRP) was used as a reporter enzyme to catalyse the oxidation of

luminol in the presence of hydrogen peroxide (H2O2) (Motsenbocker, 1988; Whitehead et

al., 1979). Immediately following oxidation, luminol is in an excited state that may decay to

the ground state via a light-emitting pathway (fig.6).

Fig.6: Chemical reaction involved in the oxidation of luminal and light emission in the

presence of hydrogen peroxide (H2O2) and horseradish peroxidase (the asterisk denotes

excited state of 3-aminophthalate). This reaction was involved for the fluorescent detection

in Western immunoblotting analysis. Cited from website of Wageningen University

(www.ftns.wau.nl/oc/research/phytochemistry/Antioxidants/lunteren/lunteren.htm)

For the fluorescent detection, the membrane was soaked briefly in the chemiluminescense

detection reagent containing luminol and hydrogen peroxide (H2O2). The light output, which

had a maximum emission at 428nm, peaked for 15 to 20 minutes before decaying. The

resulting light was fluorographed using standard X-ray film (X-Omat, Kodak, Stuttgart,

Germany). The approximate exposing time was 15 seconds.

Spot intensity of Western immunoblots was accessed using ImageQuant software package

(Amershan Biosciences, Freiburg, Germany) after digitalizing the fluorographs (Umax

Mirage-II DIN A3 scanner, Willich, Germany). The intensity ratio of COX I to TOM20 was

calculated.

3.2.15 Blue-native electrophoresis

Blue-native electrophoresis is a charge shift method, where the electrophoresis mobility of

protein complex is determined by the negative charge of the bound Coomassie dye, as well

as the size and shape of the complex (Schagger and von Jagow, 1991). In a two-

dimensional Blue-native/SDS-PAGE analysis system, monomeric proteins migrate within

the hyperbolic diagonal, whereas protein spots below the diagonal indicate their protein

complex nature.

Blue-native electrophoresis was scaled down to mini-gel system (0.75 x 70 x 82 mm, Bio-

Rad, Munich, Germany). A non-linear 6-15% polyacrylamide gradient slab gel was used

and 300µg protein (35µl) was loaded per slot (Brookes et al., 2002; Klement et al., 1995).

The anode buffer contained 50mM BisTris, the cathode buffer comprised 50mM N-[2-

Hydroxy-1,1-bis(hydroxymethyl)ethyl]glycine (Tricine), 15mM BisTris and Coomassie

brilliant blue G-250 (0.02% w/v).

The gel was run for 30 minutes at 40V and then at 110V until the Comassie dye reached

the end of the gel. Strips of first-dimension gel containing separated protein complexes

were cut out, incubated with 3% SDS (w/v) and 1% β-Mercaptoethanol (w/v) for 20 minutes at

room temperature. The gel stripe was subsequently subjected to 15% SDS-PAGE for the

separation in the second dimension. The resulting second dimension gel was stained with

preparative silver staining, as described in 3.2.11.2.

3.2.16 Mathematical modeling

We devised an abstract mathematical model in order to investigate the mechanism of wild

type and mutated mitochondrial competition inside a post-mitotic cell during the aging

process. The main purpose of employing this model in this study was to apply the model on

our experimental data to calculate the mtDNA mutation rate of mouse.

3.2.16.1 Modeling concept

This model was based on the following prerequisites: Free radicals induce both mtDNA

deletions and mitochondrial membrane damage; Mitochondria with an intact genome have

a growth advantage over mitochondria bearing genome damage; Mutated mitochondria

produce less amount of free radicals compared to wild type mitochondria; Accumulation of

membrane damage is proportional to free radical level. Besides, it was assumed that the

degradation rate of mitochondria is positively correlated to their membrane damage level.

Figure 7 gives an overview of the current model.

Wild-type

mitochondria

Defected

mitochondria

Rad

Degradation

k*Rad

f2f1

Degradation

ϕ1ϕ2

Decay

Ao*Rad

Wild-type

mitochondria

Defected

mitochondria

Rad

Degradation

k*Rad

f2f1

Degradation

ϕ1ϕ2

Decay

Ao*Rad

Fig.7: Biochemical reactions described by the current model. Two different classes of

mitochondria were considered, those with intact genome (Mw) and those with mutated

genome (Mm). Different classes produce different amounts of free radicals, which cause

the transition of mitochondria from wild type to mutated type. Mitochondria are replicated

and degraded with rate constants that are positively correlated to the amount of available

mitochondria.

Mitochondria replication process is independent of the cell cycle in the relaxed condition.

However, cell mitosis makes the replicate obligatory. In order to reflect the difference of

post-mitotic tissue and mitotic tissue, the replication rate for mouse brain and liver were set

to 14 and 9.5 days, respectively. This is based on the consideration that cell division force

mitochondria to replicate (Dallman, 1967; Neubert et al., 1966).

Since the mitochondria are not fully autonomous entities, but depend on the cytosolic

protein synthesis, the investment of a cell for the production of mitochondria has an upper

limit (Brown, 1991; Kowald and Kirkwood, 2000). The mathematical construct used to

simulate this behaviour is given as the following control term:

Current proliferation rate = maximum proliferation rate x (A-Mw-Mm), (eq.2)

where “A” is the maximum number of mitochondria in a cell (we supposed here that there

are a maximum of 1000 intact mitochondria in a cell, so A=1000), and the maximum

proliferation rate for intact and mutated mitochondria were α and β, respectively. This

construct reflects the dependency of mitochondrial proliferation rate on the remaining

resource in the cell. It is a combination of product inhibition mechanism for the

mitochondrial production reaction and a mass-dependence law (Brown, 1991).

Mitochondria are continuously degraded in the cell (Stryer, 1995). The rate of mitochondrial

degradation is proportional to the amount of available mitochondria (Kowald and Kirkwood,

2000). Since the mitochondria bearing mutated genetic material also have defected

respiratory chain, they generate less amount of free radicals. This deduces that compared

with wild type mitochondria, defective mitochondria accumulate membrane damage more

slowly. Therefore, they have a slower degradation rate (ϕ1> ϕ2, where ϕ1 and ϕ2 are the

degradation rate of wild type and mutated mitochondria, respectively). This is de Grey’s

idea of “survival of the slowest” (de Grey, 1997).

Through the continuous exposure to free radicals in the environment, mitochondrial

genomes are cumulatively damaged during the aging process. This phenomenon was

described by mutation reaction in the modeling system. Free radicals can damage the

mtDNA with a rate “k”, converting intact mitochondria into defective ones. The number of

mutation event is proportional to the free radical level and the amount of intact

mitochondria. For simplicity, a rate constant was set throughout this system for mtDNA

mutation reaction, supposing that the strength of radical-induced damage remains the

same.

As side-product of the respiratory chain enzyme activity, free radicals are produced by all

mitochondria during the oxidative phosphorylation process. It is assumed that wild type

mitochondria generate more radicals than mutated mitochondria (f1>f2, where f1 and f2 are

the free radical production rate of wild type and mutated mitochondria, respectively).

This model contains a constant number of antioxidant molecules that destroys radicals at a

rate of ϕ3. In the radical clearance process, antioxidants act as enzymes that are not

consumed in the reaction. The removal of free radicals is proportional to the existing

amount of radicals and antioxidant level.

3.2.16.2 Differential equations and numerical solutions

To assist the deduction of numerical solutions from the assumptions described above, the

ordinary differential equations (ODE) were developed. This model consists of two

equations for the different mitochondrial sub-populations (wild type and mutated type), and

one for the free radical level in a wild-type mitochondrion. The full set of differential

equations reads:

()

dMw

w A Mw Mm k Rad Mw Mw

dMm

m A Mw Mm k Rad Mw Mm

dRad fMwfMm Rad

=α⋅ ⋅ − − − ⋅ ⋅ −ϕ ⋅

=β⋅ ⋅ − − + ⋅ ⋅ −ϕ ⋅

=⋅ + −ϕ⋅

(eq. 3, 4, 5)

In these equations, “Mw”, “Mm” and “Rad” are the concentrations (number of mitochondria

per cell) of wild type mitochondria, mutated mitochondria and free radical level (in the unit

of 106), respectively.

The first and third terms in eqn (3) and (4) are the synthesis and degradations of variables,

respectively, with rates dependent on the Mw and Mm current concentration. The second

term in eqn (3) and (4) describes the mutation reaction that rendering wild type

mitochondria into pool of mutated mitochondria. Table 1 shows the parameters chosen to

reflect the known features in the frame of the model:

Tab.1: Standard parameters used for the mathematical simulation:

Parameter name Value Description

α Brain: 0.07 d-1

Liver: 0.073 d-1

Replication rate of wild type mitochondria was set

at 14 days for post-mitotic cells, 9.5 days for non-

post-mitotic cells. According to Dallman (Dallman,

1967; Neubert et al., 1966).

β 0.035 d-1 Replication rate of defective mitochondria was set

at 28 days.

A 1000 Maximum number of total mitochondria in a cell.

ϕ1 0.0693 d-1

Degradation rate of wild type mitochondria. It

corresponds to a half-life of 10 days (Menzies and

Gold, 1971).

ϕ2

0.0231 d-1

Degradation rate of defective mitochondria, set to

25% of that of wild type mitochondria

ϕ3 7000 d-1 Rate of free radical removal by antioxidants (in the

unit of 106), according to Rotilio (Rotilio et al.,

1972).

f1 900 d-1 Free radical production rate of intact mitochondria

(in the unit of 106), according to Joenje (Joenje et

al., 1985).

f2 300 d-1 Free radical production rate of mutated

mitochondria was set to one third that of intact

mitochondria (in the unit of 106). According to

Kowald (Kowald and Kirkwood, 2000).

k to be calculated Mutation rate of mouse mitochondrial DNA.

The initial values for wild type mitochondria, mutated mitochondria and free radical were

set to 1000, 0 and 0, respectively:

(0)

1000;

Rad

(eq. 6, 7, 8)

For the numerical integration of time evolution of different variables (wild type mitochondria,

mutant type mitochondria and free radicals), the Runge-Kutta method was used. Based on

the Euler method, the basic idea of the Runge-Kutta method is the adding up of the

combination of the error terms of numerical integration, in order to decrease the error terms

order by order. This method is suitable for non-oscillating functions and weak oscillating

functions (Cartwright and Piro, 1992).

Parameter scanning respecting value “k” was carried out to determine the mouse mtDNA

mutation rate. For this purpose, “k” values ranging from 10-3 to 10-12 per gene per day was

simulated in a batch, the values of Mw(t=newborn), Mw(t=5-months), Mw(t=10-months), Mw(t=15-months),

Mw(t=20-months) and Mw(t=24-months) calculated from these mathematical simulations were

extrapolated. These sets of data were compared to the set of the experimental data gained

from the 2D-electrophoresis through linear regression analysis.

As the experimental data set, the mean relative intensity values of mitochondrial NADH-

ubiquinone oxidoreductase 13 kDa-A subunit, NADH-ubiquinone oxidoreductase 1 alpha

subcomplex 5 and the COX Vb obtained through 2D-electrophoresis analysis were used.

This was based on the reduced assumption that only wild type mitochondria contained

respiratory complex I and complex IV subunits, wile mutated mitochondria were totally

depleted of this subunits (di Rago et al., 1997). The “k” value bearing the simulated data

set that best fits the experimental data set (with correlation coefficient nearest to 1) was

considered as the mtDNA mutation rate of our mouse model. For this purpose, the linear

regression with the highest correlation coefficients values (closest to one) was treated as

the optimal fit.

4 Statistics

Sample dimension for all age groups in this study was eight (n=8), except for newborn

group, which had a sample dimension of thirteen (n=13) (tab.2). The Western

immunoblotting analysis was repeated six times. Histochemical staining of COX-activity

was carried out for three times on muscle biopsies from three mouse individuals. Each

large-gel 2D-electrophoresis analysis experiment was repeated at least three times (n=3 to

6, tab.3).

Tab. 2: Number of biological materials employed in this study:

Tissue

Type Newborn 5-months 10-months 15-

months 20-

months 22-

months 24-

months

Brain 13* 8 8 8 8 -- 8§

Liver 13* 8 8 8 8 -- 8§

Muscle

biopsy 3 -- -- -- -- 3 --

Note: * Sample pooled before mitochondria isolation; § these samples entered the current study as

frozen materials.

Tab. 3: Sample dimension of different analysis carried out in this study:

Analysis Sample type Sample dimension

COX-activity staining Muscle biopsy n=3

2D-electrophoresis Brain, liver n=3 to 6

Western immunoblottong Brain, liver n=6

Blue-native electrophoresis Liver n=3

All metric values measured in this study were reported as mean ± standard error of mean

(SEM) of the given number of experiments, while ordinal values were given as median

together with quartile distance (∆0.50). Median values were used for quantitative change of

spot intensity.

For the gel image evaluation, relative spot intensity was used for the spot volume

normalization, calculated as the ratio of individual spot volume to the sum of all spot

volumes on its parent gel:

Relative intensity =

Intensity of this spot

Σ(Intensity of each spot)

Relative intensity =

Intensity of this spot

Σ(Intensity of each spot)

(eq.9)

For the access of reproducibility of 2D-electrophoresis, correlation coefficient of different

gel pairs was calculated using the following formula:

()()

()(

XXYY

XYY

−−

∑

∑∑

(eq.10)

where x and y values were the relative intensities of matched spots on each gel profile, and

sums were taken ranging over all matched spots in each gel profile.

For all experiments besides 2D-gel spot evaluation, group-to-group difference investigation

between more than two groups was performed with analysis of variance (ANOVA)

(p<0.05), with the null-hypothesis that no difference exists. Difference analysis between two

groups were performed with student’s t-test (p<0.05).

Since the sample dimension of 2D-electrophoresis was small (n=3 to 6), Kruskal-Wallis test

was employed for the access of variation in multiple-groups. In case significant variance

was detected among the six age groups (p<0.05), non-parametric Mann-Whitney U test

was further conducted for the pair-wise comparisons among different age groups to gain

group-to-group difference information.

For the protein concentration assay, the calculation of average alteration rates of diverse

proteins and for the model fitting, linear regression analysis according to “least squares”

model was employed. The average change rates per day were determined by the slope of

linear regression. Correlation coefficients were calculated for the control of regression

quality.

5 Result

In this aging study, a mouse model (C57BL/6) was first validated as model of human aging

by histochemical analysis of COX-activity. Then mitochondria from brain and liver tissues

were isolated from mouse of different aging stages. The change of mitochondrial proteome

during the aging process was accessed using proteomic approaches, including large-gel

2D-electrophoresis and Western immunoblotting. Preliminary studies employing Blue-

native electrophoresis was carried out to investigate the possibility of accessing further

protein subunits of respiratory chain complexes.

In parallel, a mathematical model was constructed to simulate the accumulation of

defective mitochondria during the aging process through free-radical-induced mtDNA

mutation. With the help of the mathematical model, the mouse mitochondrial DNA mutation

rate was estimated.

5.1 Organ weight difference during development and aging

The brain, liver and muscle tissues of different mouse age groups ranging from newborn to

24-months were obtained. The average weight of brain and liver organs were 0.44 ± 0.08 g

(n=43) and 1.25 ± 0.39 g (n=43), respectively, with a ratio of liver to brain as 2.84 (fig.8).

Except for newborn stage, no significant difference of organ weight among all aging stages

was observed.

Brain

Liver

Age (month)

502410 15 20

1.2

1.4

1.6

1.8

Organ Weight (g)

Brain

Liver

Age (month)

502410 15 20

Age (month)

502410 15 20

1.2

1.4

1.6

1.8

1.2

1.4

1.6

1.8

Organ Weight (g)

Fig.8: Weight-age relationship of the mouse brain and liver was evaluated in all

experimental groups. No significant difference between all groups except for newborn

stage was revealed.

5.2 Aging causes COX-activity deficiency on mouse muscle tissue

Mouse muscle biopsies with 79 ± 4 (n=21) muscle fibres per section underwent COX-

activity staining and were examined under light microscope (100x to 200x magnification,

fig.9). In sections from normal elderly subjects (22-months), there were muscle fibres with

very low enzyme activity of cytochrome c oxidase (red arrow), displaying beige colour,

while the other muscle fibres in the same section displayed dark brown colour, indicating

normal or elevated COX-activity (blue arrow). The whole section showed “mosaic pattern”

of alternating light and dark brown staining. The COX-positive ratio of 22-months-old mice

muscle accounted 58.7 ± 9.2% (n=21), while that of young muscle (2 weeks) was 100%

(n=21).

Histochemical staining of COX activity

Young muscle tissue Old muscle tissue

COX-negative

COX-positive

Histochemical staining of COX activity

Young muscle tissue Old muscle tissue

COX-negative

COX-positive

Fig.9: Histochemical staining of COX-activity on young (5-months, left image) and old

mouse (22-months, right image) muscle sections. Abnormal muscle fibres with low or

absent of COX-activity were detected in the rectus femoris muscle of a 22-months old

mouse. Red arrow indicates a representative COX-negative myocyte displaying beige

colour, the blue arrow shows a COX-positive phenotype stained brown.

Significant difference was observed between young and old muscle (student t-test,

p<0.0001). However, histochemical staining of COX activity gives only ordinal results

(positive or negative). No quantitative measurement could be conducted (fig.10).

100

2 22

Percent of COX-positive (%)

Age (month)

100

2 22

Percent of COX-positive (%)

Age (month)

Fig.10: Bar chart showing the percentage of COX-positive muscle fibres measured in the

COX-activity staining experiment. Significant difference between young (5-months) and old

(22-months) muscle were observed. 41% of senescent myocytes were COX-negative,

while no COX-negative myocytes were detected in young counterparts.

5.3 Result of mitochondria isolation

Mitochondria were isolated from mouse brain and liver tissues using the combination of

gravity centrifugation and gradient centrifugation. After gradient centrifugation using 30%

Percoll solution, two faintly white distinct bands could be observed in the continuous

gradient body, with diffuse floating material between the two bands (fig.11). The two bands

appeared at densities of 1.09-1.13 g/ml and 1.05 g/ml, respectively (Jungblut and Klose,

1985). Both the low-density fraction and the higher-density fraction were collected and

investigated using electronic microscopy.

Membranous structure

Pure mitochondria

Membranous structure

Pure mitochondria

Membranous structure

Pure mitochondria

Fig.11: Image of a centrifuge tube after Percoll gradient centrifugation during the

mitochondrial isolation procedure. Two bands were found, with some diffuse materials in

the intermediate area. The upper band contained membranous structure and lysosomes,

whereas the lower band was proved later to be highly purified intact mitochondria.

At the end of the washing step, the resulting mitochondrial fraction was diffuse, with pale

white colour. Care was taken not to disturb the pellet during the aspiration of washing

medium. Agar embedding procedure was necessary in the electronic microscopic analysis.

5.3.1 Morphology of isolated mitochondria

Through the electronic microscopic investigation, the upper layer in the Percoll gradient

body was observed to contain mainly membranous structures, with a broad distribution of

small vesicles and lysosomes, as well as damaged mitochondria (result not shown). The

lower fraction with a narrow distribution of density range was highly enriched in intact

mitochondria (fig.12). This shows that the intact mitochondria with higher density migrated

to the lower layer during the continuous gradient centrifugation, while most lysosome and

part of the broken mitochondria migrated to the upper layer. Based on this observations,

only the lower fractions were used in the for subsequent proteomic analysis.

Mouse Liver MitochondriaMouse Brain Mitochondria Mouse Liver MitochondriaMouse Brain Mitochondria

Fig.12: Electronic micrographs of the lower fraction material obtained after Percoll gradient

centrifugation. Left image: brain mitochondria, Right image: liver mitochondria. 10,000x

original magnification.

On the electronic micrograph of isolated brain and liver mitochondria, the whole field

demonstrated a highly homogeneous population of mitochondria (94.2 ± 0.7%, n=6) with

inner and outer membranes intact. Only a small number (5.9 ± 1.4 %, n=6) of objects was

other than mitochondria, among which mostly lysosomes (3.0 ± 0.6%, n=6) displaying

regular round shape, homogeneously dyed grey. This showed that the mitochondria

isolation from both mouse brain and liver tissues has been effective (fig.13).

Mitochondria

Lysosomes

Others

Mitochondria

Lysosomes

Others

Fig.13: Percentage of mitochondria, lysosome, and debris in the mitochondrial fraction after

Percoll gradient purification. 94% of particles were intact mitochondria, 3% were lysosome,

others were undefined cell debris.

Both mitochondria isolated from brain and liver have the form of either round, or tube-like

structure with a length of 1.8 ± 0.1 µm (n=100), and a width of 0.9 ± 0.03 µm (n=100). They

consisted of double membranous structure that has been made contrast by osmium. The

outer membrane built a round or oval envelope, while the inner membrane possessed

pronounced cristae structure. In some electronic micrograph, the mitochondria were slightly

rounded, indicating either a lower osmotic pressure of the surrounding medium, or the

slight damage of mitochondria structure.

5.3.2 Yield of mitochondria

Together with the organ weight data collected in this study, the yield of mitochondria could

be calculated as the “mg mitochondria per gram of wet tissue”:

Yield of mitochondria = Amount of mitochondria [mg]

Weight of wet tissue [g]

Yield of mitochondria = Amount of mitochondria [mg]

Weight of wet tissue [g]

(eq.11)

The yield of mitochondria was 4.21 ± 1.46 mg/g (n=41) from mouse liver tissue and

2.33 ± 0.97 mg/g (n=41) from brain tissue, respectively. Mitochondrial yield of liver was 1.8

times to that of brain. In both brain and liver tissues, the yields of mitochondria from

newborn mice were significantly lower than that of the remaining aging stages (fig.14).

Brain

Liver

Yield of Mitochondria (mg/g)

Age (month)

502410 15 20

Brain

Liver

Yield of Mitochondria (mg/g)

Age (month)

502410 15 20

Age (month)

502410 15 20

Fig.14: Mitochondrial yields from mouse brain and liver tissues of different age groups.

Both brain and liver tissues from the newborn stage gave significant small yield of

mitochondria.

Interestingly, for brain tissue, there was in addition an age-correlated difference of

mitochondrial yield: significantly smaller yields of mitochondria (p<0.05) was obtained from

the brain tissue of 20-months and 24-months mice. No similar tendency was observed in

data of liver mitochondrial yields (fig.15).

Age (month)

502410 15 20

Liver

Brain

Yield of Mitochondria (mg/g)

Age (month)

502410 15 20

Age (month)

502410 15 20

Liver

BrainBrain

Yield of Mitochondria (mg/g)

Fig.15: Yields of mitochondria respecting different age groups. For mouse liver tissue, only

the newborn group had smaller yield compared to other aging stages. In brain, 20-months

and 24-months aging groups gave significantly smaller yield of mitochondria compared to

that of other stages aging (5-months, 10-months and 15-months). The asterisks indicate

statistical significance (p<0.05).

5.3.3 Comparison of mitochondria isolation from fresh and frozen material

Since it was not always possible to obtain fresh materials for mitochondria isolation, frozen

mouse organs were also employed in this study (24-months age group). The potential

influence of this initial sample handling condition on the outcome of mitochondrial quality

was taken into consideration. Upon comparing the mitochondria isolation result from fresh

and frozen materials, part of the mitochondria isolated from frozen tissue showed slightly

lower membrane integrity (fig.16). The yield of mitochondria gained from frozen material

was smaller.

Mitochondria from fresh

mouse muscle

Mitochondria from frozen

mouse muscle

Mitochondria from fresh

mouse muscle

Mitochondria from frozen

mouse muscle

Fig.16: Comparison of mitochondria isolation from fresh and frozen mouse muscle tissues.

No substantial difference in quality of isolated mitochondria could be observed, although

some of the mitochondria isolated from frozen tissue had slightly lower outer membrane

integrity.

5.3.4 Comparison of mitochondria isolation from young and old organism

The subject of this study was to elucidate the protein profile changes of mitochondria in

different aging stages. For this purpose, it was to be prerequisited at the first line that the

mitochondria isolated from all aging stages were of comparable quality. Thus, a

comparison of mitochondria isolation from young and old tissues was conducted.

Figure 17 shows the electronic microscopy of mitochondria isolated from young (5-months)

and old (22-months) mouse brain tissue. No significant difference in mitochondrial

morphology and purity could be observed from this comparison. This showed that it was

tractable to isolate pure intact mitochondria from both young and old mouse liver and brain

tissue.

Mitochondria from

young mouse brain

Mitochondria from

old mouse brain

Mitochondria from

young mouse brain

Mitochondria from

old mouse brain

Mitochondria from

old mouse brain

Fig.17: Mitochondria isolated from young (left image, 5-months) and old (right image, 22-

months) mouse organs (brain and liver) showed comparable quality under electro

micrograph.

5.4 Result of protein pre-fractionation

In order to enrich low-abundant proteins and membrane proteins in the mitochondria

proteomic analysis, we have separated the mitochondrial proteins into three fractions using

the sequential extraction strategy. Tris-buffer was first used to extract soluble proteins,

which resulted in “Fraction I”. This obtained supernatant after centrifugation was

transparent, lightly pale white. Afterwards, Triton-X100-containing buffer was used to

extract membrane-associated proteins, which resulted in “Fraction II”. Noticeably, this

fraction had a darker colour compared to that of “Fraction I”. The remaining pellet fraction

had a high viscosity, indicates the high content of lipids.

According to the protocol of Molly (Molloy et al., 1999), we carried out the methanol-

chloroform extraction using the pellet obtained after Triton buffer treatment. As expected, a

substantial amount of protein interface was precipitated between the methanol/water phase

and chloroform phase, forming a solid pale white layer. This demonstrates that there was

still considerable amount of protein in the membranous structures that could not be

extracted by detergent-containing aqueous buffer system.

The extracted protein pellet was partially dried in cold Argon to reduce the amount of

remaining organic solutions, and then resolved with Laemlli sample buffer or water.

However, even after overnight stirring, not all pellet materials were resolved in the solution.

The protein concentrations of all three fractions (“Fraction I”, “Fraction II” and “Fraction III”)

were measured by the BCA protein assay. The concentration of the protein “Fraction I” was

measured to be 10.27 ± 2.17 mg/ml (n=12), while that of “Fraction II” was 5.76 ± 0.96

mg/ml (n=12). Noticeably, the protein concentration of “Fraction III” could not be measured

successfully, possibly due to the presence of remaining organic solution, or due to the

membrane protein property that was not compatible to the BCA assay or protein standard.

Respecting the total volume of both “Fraction I” and “Fraction II” samples, together with the

assumption that there was approximately 280mg/ml protein in the mitochondrial pellet

(Brown, 1991), Tris-buffer extraction proteins (“Fraction I”) and Triton-containing buffer

extracted protein (“Fraction II”) accounted for 25.6 ± 2.8 % (n=11), 56.2 ± 5.5 % (n=11) of

total protein present of the mitochondrial pellet, respectively. This led to our prediction of

the percentage of “Fraction III” (methanol-chloroform extracted) to be around 17%. This,

again, demonstrate that a considerable part of protein remained in the protein pellet

extracted using methanol-chloroform. Figure 18 shows the percentage of protein present in

three different fractions.

100

Fraction I Fraction II Fraction III

Percent of total protein (%)

100

Fraction I Fraction II Fraction III

Percent of total protein (%)

Fig.18: Mitochondrial proteins were fractionated into three different fractions using

sequential extraction strategy. Tris-buffer extracted proteins (“Fraction I”), Triton-containing

buffer extracted protein (“Fraction II”) account for 25.6 ± 2.8% (n=11) and 56.2 ± 5.5 %

(n=11) of the total protein present of the mitochondrial pellet. The percentage of “Fraction

III” (methanol-chloroform extracted) was estimated to be 17.0 ± 3.1% (n=11) respecting the

starting material amount.

Due to the lack of protein concentration data of “Fraction III” samples, weight of protein

pellet instead of protein amount was used as the sample amount orientation in the

subsequent Western immunoblotting analysis.

5.5 Result of 2D-electrophoresis analyses

2D-electrophoresis analysis was employed for the quantitative analysis of “Fraction I” and

“Fraction II” mitochondrial proteins of different age groups. A broad carrier ampholyte

mixture with pH value ranging from 2-11 was used as carrier ampholites in order to get a

panorama view of the mitochondrial proteins. Figure 19 shows the reference gel patterns of

brain and liver mitochondrial total protein extract. By convention, the resulting high-

resolution spot patterns of a 2D-electrophoresis were oriented with the low, acidic

isoelectric points on the left and the lowers molecular weight proteins at the bottom.

a) 2D-gel pattern of mouse brain mitochondrial total proteins

b) 2D-gel pattern of mouse liver mitochondrial total proteins

a) 2D-gel pattern of mouse brain mitochondrial total proteins

b) 2D-gel pattern of mouse liver mitochondrial total proteins

Fig.19: The mouse brain (upper image) and liver (lower image) mitochondrial protein total

extract (from 5-months old mouse brain and liver tissues) were separated by large-gel 2D-

electrophoresis (40x30 cm). Shown are representative analytical silver stained gels at

protein load of 100

After the preparative silver staining, ampholyte running front gave hill-form background at

the lower side of the gels (see fig.20). This is specific for preparative silver staining due to

the short rinsing period. Since the carrier ampholyte contain both amino- and carboxyl

residues, they behave like small proteins that also bind to SDS. Due to their small size, the

vast majority migrate at the dye front (O'Farrell, 1975). In the analytical silver staining

method, these small ampholytes were washed out of gels though the application of long

incubation period.

5.5.1 Reproducibility of 2D-electrophoresis carried out in this study

In order to study primary variability of the 2D-electrophoresis method carried out in this

experiment, the gel-to-gel difference of the large-gel 2D-electropheresis analysis was first

accessed with six different gels generated from a single sample, run on different days

spreading in the time span of six months.

Although the contrast level of the gels was slightly different, the majority of analogous

protein spots showed only minimal differences in relative location and shape. The average

correlation coefficient of these six 2D-gels was 0.98 ± 0.05. Figure 20 shows an example of

wrapped image of two 2D-gels from this same sample, created by the image software

ProteinWeaver. On this wrapped image, single gel image bears either blue or orange,

which is the complementary colour of blue. Analogous spots on different gels with similar

spot pattern and location display black colour, which is the optical addition of blue and

orange.

One gel

Another gel

One gel

Another gel

Fig.20: The wrapped image of two 2D-gels (20x30 cm) run from the mouse liver

mitochondrial protein sample (5-months of age, “Fraction I”) displays minimal spot pattern

difference of the two gels. The matched spots were displayed in black colour, which

account for the vast majority of spots. This indicates a high reproducibility of the 2D-

electrophoresis method. The spots that could not be matched were either orange or blue.

Using the relative intensity of one gel as X-axis, and that of the other gel as Y-axis, the

majority of the matched spots of the gel pairs were localized close to the diagonal line

(fig.21). This indicates that the matched spots show correlated intensity.

Relative spot intensity of one gel (log)

Relative spot intensity of another gel (log)

Relative spot intensity of one gel (log)

Relative spot intensity of another gel (log)

Fig.21: Scatter plot of the relative spot intensities of two 2D-gel patterns of brain

mitochondria (Fraction I protein) from a same mitochondrial sample. The x-axis shows the

intensity of the gel image on one gel, the y-axis shows the spot intensities on another 2D-

gel. Spots with the same relative intensity on both gels were found on the diagonal of the

scatter plot. The distance from the diagonal is a measure of the intensity difference of the

protein spots in the two gels.

The same tendency was observed in 2D-gels from different age groups in our experiment.

In order to access the comparability of 2D-gels generated in our current experiment, we

defined “spot matching rate” as the percent of corresponding spots on two different 2D-gels

from different mitochondrial protein samples. The average spot matching rate of liver

mitochondrial 2D-gels among different age groups was 91 ± 4% (n=15), that of brain

mitochondrial gels were 79 ± 8% (n=15). The average spot matching rate of the whole

study accounted 80 ± 13% (n=66) (fig.22). The 2D-gels from brain mitochondria had a

slightly less satisfactory spots matching quality compared to that of liver.

Whole study

Brain

Liver

Correlation coefficient

Whole study

Brain

Liver

Correlation coefficient

Fig.22: Box chart showing the value distribution of “spot matching rates” of 2D-gels in this

aging study. The mean value of “spot matching rates” for liver-to-liver gel matching, brain-

to-brain gel matching were 0.91 and 0.79, respectively. The average “spot matching rates”

of the 2D-gels throughout this study was 0.80.

Respecting the matching quality of 2D-gels from different age groups, the best match of

brain mitochondrial 2D-gels was obtained between the 10-months age group and the 24-

months age group (spot matching rate 99.3%). In liver, the 2D-gels of newborn stage

matched best with 20-months aging stage (spot matching rate 98.3%). There was no

correlation of gel image matching quality to the age distance. Larger age difference does

not correlate to a smaller matching quality. Specifically, there was no significant difference

in protein spot patterns between 2D-gels of 24-months samples (which entered this study

as frozen materials) compared to that of all other age groups.

5.5.2 Comparison of whole tissue 2D-gel to mitochondrial 2D-gel

In order to investigate whether the isolation of mitochondria allowed us to access changes

of additional protein spots on the 2D-gels, the 2D-electrophoresis gel patterns of the brain

total protein extract and brain mitochondria were compared to each other (fig.23). The two-

dimensional protein spot pattern of mitochondria was largely reduced compared to that of

brain whole tissue spot pattern. While the 2D-gel of mouse brain total protein extract

contains over 6000 spots, 626 protein spots were found on the 2D-gel map of mitochondrial

total protein extract.

Fig.23: Comparison of brain mitochondrial total protein extract to brain total protein using

2D-electrophoresis (40x30 cm). All protein spots shown in black colour were also visible on

the 2D-gel of mouse brain total protein extract (shared spots). Red dots indicate spots that

are additional or demonstrated enhanced intensity on the mitochondrial 2D-gel (courtesy of

Dr. Sagi).

Despite the reduced protein spot pattern on the mitochondrial gel, 90 spots, which were

barely visible on the whole organ gel, showed up intensely on the mitochondrial gel.

Another 26 spots were completely additional on the mitochondrial 2D-gel, supporting the

effect of sub-cellular fractionation on the enrichment of low-abundant proteins (courtesy of

Dr. Dijana Sagi in our research institute).

5.5.3 Effect of protein pre-fractionation on protein resolution

Through the 2D-electrophoresis separation carried out in our current study, 556 protein

spots were resolved from the mitochondrial “Fraction I” protein samples (fig.20), while 149

spots were resolved from “Fraction II” protein samples (fig.24).

In order to target membrane proteins, we employed Triton-X100 in the 2D-electrophoresis

separation of “Fraction II” proteins obtained through the sequential protein extraction

procedure. Consequently, it was our interest to investigate how many hydrophobic proteins

were resolved on the “Fraction II” 2D-gels.

As shown in fig.24, very different protein spot pattern was obtained from “Fraction II”

proteins (compared to fig.19, fig.20). Due to the drastic difference of spot patterns, it was

not feasible to compare the 2D-gel patterns of “Fraction II” proteins to other 2D-gel pattern

of our study. Thus, all 80 spots with relatively high intensity on the 2D-gels of “Fraction II”

were taken to mass spectrometric protein identification.

94kda glucose-

regulated protein

Vesicle-associated

membrane protein UDP-glucuronosyl-

transferase precursor

Voltage-

dependent anion

channel protein 3

Progesteron-

receptoer membrane

component

Apocytochorome b5

ATP synthase

F0 subunit 8

94kda glucose-

regulated protein

Vesicle-associated

membrane protein UDP-glucuronosyl-

transferase precursor

Voltage-

dependent anion

channel protein 3

Progesteron-

receptoer membrane

component

Apocytochorome b5

ATP synthase

F0 subunit 8

Fig.24: An example of 2D-gels (20x30 cm) from “Fraction II” sample of mouse liver

mitochondria. Very different pattern was gained compared to that of “fraction I” or standard

pattern. Seven proteins that later predicted to be membrane proteins are marked on the gel

image.

Though the comparison of 2D-gel patterns generated from pre-fractionated mitochondrial

proteins to that of mitochondrial total protein extract, six additional protein spots were

gained though the fractionation of brain mitochondrial proteins. Among these, two

additional proteins were only visible on 2D-gels of “Fraction I” samples, while four protein

spots were only visible on 2D-gels of “Fraction II”.

Analogously, four protein spots were additionally gained though fractionation of liver

mitochondrial proteins. Among these, two protein spots were only visualized on 2D-gels of

“Fraction I” proteins (one of them later identified to be mitochondrial ribosomal protein L12);

one protein spot was not visible on 2D-gels of liver mitochondrial total protein extract, but

visible in both 2D-gels of both “Fraction I” and “Fraction II” (later identified to be NADH

dehydrogenase 1 beta subcomplex 7). The other protein spot was only visible on the 2D-

gels of “Fraction II” sample (later identified to be mitochondrial DNA topoisomerase).

5.5.4 Access of possible proteomic changes based on protein spot pattern

Pursuing our initial interest of age-related changes in mitochondrial proteome, we

evaluated two-dimensional mitochondrial protein patterns of six different age groups,

generated from either “Fraction I” and “Fraction II” samples of mouse brain or liver

mitochondria. Through visual comparison of the mitochondrial protein spot patterns of

different age groups bearing the same fraction number and organ type, numerous

differences among the different age groups were detected in both brain and liver

mitochondria.

Intensities of some of the spots were observed to decrease or increase with age, indicating

the change of steady state concentration of the corresponding proteins in mitochondria

during the aging process. Specifically, 26 changed spots were detected in brain “Fraction I”

gels; 16 changed spots were found in liver “Fraction I” gels; 19 spot changes were

observed in brain “Fraction II” 2D-gels, while 34 spot changes were annotated in liver

“Fraction II” 2D-gels (tab.4). Noticeably, eight protein spots showed only decreased profile

from 20-months age group to 24-months group.

Tab.4: Protein spot changes from newborn to 24-months group detected on 2D-gels of

different mitochondrial protein fractions:

Brain mitochondria Liver mitochondria

Alteration “Fraction I” “Fraction II” “Fraction I” “Fraction II”

Up-regulation 3 3 5 5

Down-regulation 9 (4)* 6 9 (1)* 10 (3)*

Other changes 14 0 2 19

Not visible in

newborn

8 8 2 8

Visible only in

newborn

12 4 1 6

Note: * among the down-regulated protein spots, number of spots showing only a decrease from 20-

months to 24-months of age are listed in the parenthesis.

For the access of tissue-specifity respecting mitochondrial aging profile, we compared the

cross differences of 2D-gels respecting brain or liver mitochondrial sample origin. Eight

protein spots appeared to be enhanced in brain or liver mitochondrial 2D-gels, respectively.

Seven of these protein spots were on the liver mitochondrial protein 2D-gels, while one was

brain mitochondria-specific. Together with the 80 selected protein spots on the “Fraction II”

2D-gels, 173 protein spots underwent mass spectrometric identification (tab.5).

Tab.5: Protein spots from 2D-gels that underwent mass spectrometric protein identification:

Categories Number of spots

Changed spots in brain “Fraction I” 26

Changed spots in brain “Fraction II” 9

Changed spots in liver “Fraction I” 16

Changed spots in liver “Fraction II” 34

Spots from “Fraction II” proteins 80

Possible liver mitochondria-specific 7

Possible brain mitochondria-specific 1

TOTAL 173

5.6 Result of protein identification

For protein identification, we employed both MALDI-TOF-MS and ESI-Iontrap-MS analyses.

MALDI-TOF-MS is fast and effective method, but needs larger amount of peptide sample.

Besides, MALDI-TOF-MS is not suitable for the identification of protein mixture. For this

reason, intensive spots on the 2D-gels were identified with MALDI-TOF-MS, whereas ESI-

Iontrap-MS was employed for protein spots with less intensity, especially those spots that

were partially merged to adjacent ones.

5.6.1 Quality of MS-spectra

Figure 25 shows an example of mass spectra of the MALDI-TOF-MS experiment.

Comparing to the in silico digested peptide list, measured peptide was highlighted on the

peptide sequence. The methionine-containing peptides and their corresponding oxidized

derivates, which leads to a (m/z) mass increase of 18 Dalton, were marked blue with their

non-oxidized counterparts at the left side. Although it was not expected that all methionine

residuals have their oxidized counterparts, this pair-wise representation of methionine-

containing peptides strongly confirmed the identity of these peptides.

m/z

500 1000 1500 2000 2500 3000 3500

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

[Abs. Int. * 1000]

837.575

402-409

935.479

430-437

985.464

371-378

1067.477

466-474

1170.585

59-68

1233.530

234-245

1268.512

507-517

1461.738

718-730

1463.636

412-424

1500.657

522-534

1565.615

32-44

1570.655

592-605

1581.632

32-44

1601.674

634-648

1667.658

657-671

1725.882

724-739

1731.676

535-549

1753.743

145-160

1765.772

410-424

1778.823

702-717

1859.745

379-395

1868.842

69-84

1875.746

379-395

1952.909

143-160

2274.071

118-138

2362.032

535-555

m/z

500 1000 1500 2000 2500 3000 3500

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

[Abs. Int. * 1000]

837.575

402-409

935.479

430-437

985.464

371-378

1067.477

466-474

1170.585

59-68

1233.530

234-245

1268.512

507-517

1461.738

718-730

1463.636

412-424

1500.657

522-534

1565.615

32-44

1570.655

592-605

1581.632

32-44

1601.674

634-648

1667.658

657-671

1725.882

724-739

1731.676

535-549

1753.743

145-160

1765.772

410-424

1778.823

702-717

1859.745

379-395

1868.842

69-84

1875.746

379-395

1952.909

143-160

2274.071

118-138

2362.032

535-555

Fig.25: An example of MS-spectra obtained in this study. 34 mass peak signals were

measured. Among these, 20 peaks (labelled black) were able to be assigned as tryptic

digested peptide sequences of aconitase (gi 18079339). Another six masses (labelled blue)

could be orientated as modification of existing peptide peaks. Three of the peaks were

identified as endo-digestion product of trypsin.

5.6.2 Databank-based protein identification

The obtained mass spectrometry raw data (peak list and relative intensities) were used to

search against the National Centre for Biotechnology Information non-redundant protein

database NCBInr (all mammalian species) with the Mascot search engine.

Mascot incorporates a probability-based implementation of the Mowse (mol. Wt. Search)

algorithm (Pappin et al., 1993). It calculates a p-value describing the absolute probability

that the observed match is a random event (with the assumption that the entries in the

sequence databases are random sequences). The Mowse score displayed in the result

sheet was reported as –10log10(p). The significant threshold value for Mowse score is thus

a function of the size of the sequence database being searched. In the current study, a

threshold value of 61 corresponded to a match with statistical significance (p<0.05) for

searching all mammalian species of NCBInr database.

For the majority of cases in our study, four or five measured peptides were sufficient to

confirm the identity of the protein satisfactorily. Of the 173 protein-spots-of-interest

prepared from the 2D-electrophoresis analysis, 123 spots were successfully identified,

which account for 71% of all the 173 protein identifications. The number of matched

peptide masses was 13.8 ± 0.6 (n=123). The average percentage of the sequence

coverage was 51.3 ± 1.5 % (n=123). The median of Probability Mowse score accounted as

102, with a quartile distance ∆0.50=69 (n=123). This corresponds to an expected probability

value (p) as 1.15 E-5. Nonetheless, some spot could not be identified despite multiple

efforts.

Among the 50 proteins that were not identified successfully, 10 protein spots were from the

2D-gels of brain mitochondrial “Fraction I” protein; 13 spots were from the 2D-gels of brain

mitochondrial “Fraction II” protein. Two protein spots were from the 2D-gels of liver

mitochondrial “Fraction I” protein, while 25 spots were from the 2D-gels of liver

mitochondrial “Fraction II” protein.

5.7 Prediction of hydrophobicity of identified proteins

In order to access the efficiency of membrane protein resolution of 2D-electrophoresis

analysis and sequential protein pre-fractionation strategy carried out in this study, all 123

identified proteins were subjected to Gravy score prediction. Only seven of the identified

proteins in this study were predicted to contain trans-membrane domains (ranging from one

to two), with Gravy scores ranging from 0.139 to -0.72. Among them, the ATP synthase F0

subunit 8 is one of the 13 mtDNA-encoded proteins. Noticeably, all these seven spots were

additional spots on the 2D-gel pattern of “Fraction II” proteins. The annotation of these

membrane proteins, together with their number of trans-membrane domain (TMD) and

Gravy factor are listed in the following table (tab.6).

Tab.6: Properties of membrane proteins identified on the 2D-gels of “Fraction II” samples:

Protein name gi number TMD number Gravy factor

ATP synthase F0 subunit 8 (mtDNA

encoded)

gi 5834958 1xTMD +0.139

UDP-glucuronosyltransferase 1-1

precursor

gi 2501472 2xTMD +0.087

Vesicle associated membrane

protein 2

gi 2253399 1xTMD +0.003

Voltage-dependent anion-selective

channel protein 3

gi 12643945 1xTMD -0.280

Cytrochorme b5 outer mitochondrial

membrane precursor

gi 31542438 1xTMD -0.602

Progesterone receptor membrane

component

gi 31980806 1xTMD -0.604

Endoplasmin precursor (94kDa

glucose-regulated protein)

gi 729425 1xTMD -0.72

5.8 Quantitative changes of protein spots observed

Among the 123 identified spots (see table A1 for summary), 27 spots were determined to

be cellular contaminations. These include hemoglobin, alpha globin, cytoplasmic ribosomal

proteins, myosin, major urinary proteins, albumin, beta tubulin and mouse keratin. Since

these cellular proteins are not mitochondria-specific, they were not further analyzed in the

current study. Five protein spots that were additional on the 24-months age group 2D-Gels

were identified to be QIL1; Heat-responsive protein 12; annexin A2; dimethglycine DH and

citron-kinase K.

Compared to other age stages, 23 protein spot changes were observed only in the

newborn sample, while 26 protein spots were absent in newborn state. Since these

changes could reflect possible influence of pre- and post-natal switch, these 49 protein

spots were excluded in our downstream analysis.

Specifically, inconsequence of visual gel evaluation and computer-aided gel evaluation was

observed. According to protein quantification using ProteinWeaver software, seven protein

spots that were evaluated visually as either with age-related decrease or increase profile

showed fluctuating behavior inside certain age groups without detectable rules. These

seven proteins were not further analyzed in detail. Table 7 shows a summary result of

protein spots that underwent mass spectrometric identification.

Tab.7: Identification results of protein spots that underwent MS-identification:

Category Number of spots

Not identified 50

Cytosolic contamination 27

Additional protein spots in 24-months

sample

Changes only in newborn 23

Absent in newborn 26

With fluctuating profile 7

Not statistical significance 25

Statistical significant 10

The concentration profile changes of the remaining 35 protein spots were taken to

statistical test as described in chapter 4. Among these observations, alterations of only ten

protein profile changes demonstrated statistical significance. Table 8 is a summary of

proteins that demonstrated reproducibly quantitative and qualitative changes along with

age (p<0.05 during the aging process). To access a general alteration profile during the

whole time range, linear regression analysis was carried out to obtain the percent change

per day respecting the protein concentration of newborn samples.

Tab.8: Proteins on mitochondrial 2D-gels that showed significant alteration among different

age groups. The average change rates were determined by linear regression analysis:

Change

Protein name gi number in Brain [d-1] in Liver [d-1]

NADH-ubiquinone oxidoreductase 13 kDa-

subunit 38075371 -0.429% -0.028%

NADH-ubiquinone oxidoreductase 1 alpha

subcomplex 5 13386100 -0.657% ND

Cytochrome c oxidase, subunit Vb 6753500 -0.342% -0.022%

Ubiquinol-cytochrome c reductase binding

protein 133885726 ND 0.557%

TP synthase, H+ transporting, mitochondrial

F0 complex, subunit F 7949005 0.314% ND

Mitochondrial ribosomal protein L12 22164792 ND 0.357%

10kDa Mitochondrial heat shock protein 6680309 -0.057% -0.714%

Regucalcin 6677739 ND -0.057%

Alpha-synuclein 6678047 0.071% ND

Peroxiredoxin 1 6754976 ND -0.014%

Note: Only proteins with statistically significant changes in different age groups were listed (p<0.05). Negative

sign indicates decrease tendency. ND: not detected.

In the following chapters, the concentration profile of these ten proteins against age were

graphically presented, that of brain and liver mitochondria were presented separately with

different curves.

5.8.1 Alpha-synuclein increased with age in brain mitochondria

Alpha-synuclein, which represents itself as a prominent spot on the acidic side of brain

mitochondrial 2D-gels (pI: 4.74, MW: 14kDa), was hardly detectable in the newborn group.

However, the relative intensity of this protein spot increased significantly with age. Upon

evaluating the protein spot of alpha-synuclein on the 2D-gels of brain total proteins, this

protein was present during the whole time range on 2D-gels of brain total proteins. Less

prominent increase tendency along with time was observed (fig.26).

1 week 2 week 4 week

22 weeks14 week

100 weeks

75 weeks

8 week

42 weeks

24 months

Newborn 5 months 10 months

15 months 20 months

0.5

1.5

Brain mitochondria

Total brain

Age (month)

502410 15 20

Ratio to newborn

a) Change of alpha-synuclein in brain mitochondria

b) Change of alpha-synuclein in total brain

c) Time plot of alpha-synuclein

1 week 2 week 4 week

22 weeks14 week

100 weeks

75 weeks

8 week

42 weeks

24 months

Newborn 5 months 10 months

15 months 20 months

0.5

1.5

Brain mitochondria

Total brain

Age (month)

502410 15 20

Ratio to newborn

a) Change of alpha-synuclein in brain mitochondria

b) Change of alpha-synuclein in total brain

c) Time plot of alpha-synuclein

1 week 2 week 4 week

22 weeks14 week

100 weeks

75 weeks

8 week

42 weeks

24 months

Newborn 5 months 10 months

15 months 20 months

0.5

1.5

Brain mitochondria

Total brain

Age (month)

502410 15 20

Ratio to newborn

a) Change of alpha-synuclein in brain mitochondria

b) Change of alpha-synuclein in total brain

c) Time plot of alpha-synuclein

Fig.26: Time-dependent alpha-synuclein level in brain mitochondria compared with that of

total brain. No significant difference of protein level was observed in brain total protein

extract throughout the lifespan (new born to 100 weeks, b). On the contrary, there was a

significant increase in mitochondria-associated alpha-synuclein level (a). The time-plot of

synuclein is shown in c.

5.8.2 COX subunit Vb decreased with age

The relative intensity of COX subunit Vb (pI: 8.69, MW: 14kDa) in brain mitochondria was

not changed significantly from newborn stage to 10-months stage. However, pronouced

decrease was observed in the 15-months groupe. In liver mitochondria, the COX Vb level

first showed an increase at 5-months group. The decrease tendency of relative intensity

sustained until 24-months of age (fig.27).

Brain

Liver

Age (month)

502410 15 20

Ratio to newborn

Change of COX subunit Vb

1.2

1.4

Brain

Liver

Age (month)

502410 15 20

Ratio to newborn

Change of COX subunit Vb

1.2

1.4

Brain

Liver

Age (month)

502410 15 20

Age (month)

502410 15 20

Ratio to newborn

Change of COX subunit Vb

1.2

1.4

1.2

1.4

Fig.27: Time-dependent level change of cytochrome c oxidase subunit Vb protein spot on

the 2D-gels. Shown are relative spot intensity ratios compared to that of newborn stage. A

decrease tendency was observed in both brain and liver mitochondrial 2D-gels (a and b,

respectively).

5.8.3 10kDa heat shock protein decreased with age

Spot intensity of the 10kDa mitochondrial heat shock protein (pI:7.93, MW: 11kDa) showed

decreased profile in both brain and liver mitochondria throughout the time range measured

in this study. The decrease was more pronounced in liver mitochondria (fig.28).

24 months

a) Change of 10 kDa heat shock protein in brain mitochondria

b) Change of 10 kDa heat shock protein in liver mitochondria

Newborn 5 months 10 months

15 months 20 months

24 months

Newborn 5 months 10 months

15 months 20 months

c) Time plot of 10 kDa heat shock protein

Brain

Liver

Age (month)

502410 15 20

Ratio to newborn

1.2

1.4

24 months

a) Change of 10 kDa heat shock protein in brain mitochondria

b) Change of 10 kDa heat shock protein in liver mitochondria

Newborn 5 months 10 months

15 months 20 months

24 months

Newborn 5 months 10 months

15 months 20 months

c) Time plot of 10 kDa heat shock protein

Brain

Liver

Age (month)

502410 15 20

Ratio to newborn

1.2

1.4

24 months

a) Change of 10 kDa heat shock protein in brain mitochondria

b) Change of 10 kDa heat shock protein in liver mitochondria

Newborn 5 months 10 months

15 months 20 months

24 months

Newborn 5 months 10 months

15 months 20 months

c) Time plot of 10 kDa heat shock protein

Brain

Liver

Age (month)

502410 15 20

Ratio to newborn

1.2

1.4

1.2

1.4

Fig.28: Two-dimensional gel image insets of mitochondrial 10 kDa heat shock protein in

brain and liver mitochondria (a and b, respectively). The partial 2D-gel images are shown

and the protein spot corresponding mitochondrial 10kDa heat shock protein were indicated

with arrows (a: brain; b: liver). The corresponding time-plot is shown in c.

5.8.4 Two complex I subunits decreased with age

Two protein subunits of NADH-ubiquinone oxidoreductase were observed to decrease with

age. The decrease tendency of NADH-ubiquinone oxidoreductase 13 kDa-A subunit (pI:

10.32, MW: 18kDa) was observed in both brain and liver mitochondria (fig.29), whereas the

alteration of NADH-ubiquinone oxidoreductase 1 alpha subcomplex 5 (pI: 7.82, MW:

13kDa) was detected only in brain mitochondria (fig.30).

Liver

Brain

Age (month)

502410 15 20

Ratio to newborn

Change of NADH 13kDa subunit

1.2

1.4

Liver

Brain

Age (month)

502410 15 20

Ratio to newborn

Change of NADH 13kDa subunit

1.2

1.4

Liver

Brain

Age (month)

502410 15 20

Age (month)

502410 15 20

Ratio to newborn

Change of NADH 13kDa subunit

1.2

1.4

1.2

1.4

Fig.29: Time-dependent level change of NADH-ubiquinone oxidoreductase 13 kDa-A

subunit in brain and liver mitochondria. Data were obtained through evaluation of 2D-

electrophoresis gels of mouse brain and liver mitochondria of six different ages.

Age (month)

502410 15 20

Ratio to newborn

Change of NADH subunit 5 in brain mitochondria

1.2

1.4

Age (month)

502410 15 20

Age (month)

502410 15 20

Ratio to newborn

Change of NADH subunit 5 in brain mitochondria

1.2

1.4

Fig.30: Time-dependent level change of NADH-ubiquinone oxidoreductase 1 alpha

subcomplex 5 in brain mitochondria. Data were obtained through evaluation of 2D-

electrophoresis gels of mouse brain mitochondria of six different ages.

5.8.5 Peroxiredoxin 1 decreased with age in liver mitochondria

On the basic side of the mitochondrial 2D-gel, a prominent decrease of protein spot was

detected in the liver mitochondrial 2D-gels. This spot was identified to be peroxiredoxin 1

(pI: 8.26, MW: 22kDa, fig.31). Since peroxiredoxin 1 is an abundant antioxidant in

cytoplasm, we performed protein-protein sequence alignment in order to verify the protein

annotation result. Peroxiredoxin 1 could be aligned to two mitochondria-associated proteins

with high similarity scores: substrate protein of mitochondrial ATP-dependent proteinase

[bovine] (gi627764, score 260, 1e-68), and the putative mitochondrial peroxiredoxin (gi

16751316, score 245, 5e-64).

Age (month)

502410 15 20

Ratio to newborn

24 months

Newborn 5 months 10 months

15 months 20 months

a) Change of peroxiredoxin 1 in liver mitochondria

b) Time plot of peroxiredoxin 1 in liver mitochondria

1.2

1.4

Age (month)

502410 15 20

Ratio to newborn

24 months

Newborn 5 months 10 months

15 months 20 months

a) Change of peroxiredoxin 1 in liver mitochondria

b) Time plot of peroxiredoxin 1 in liver mitochondria

1.2

1.4

Age (month)

502410 15 20

Age (month)

502410 15 20

Ratio to newborn

24 months

Newborn 5 months 10 months

15 months 20 months

a) Change of peroxiredoxin 1 in liver mitochondria

b) Time plot of peroxiredoxin 1 in liver mitochondria

1.2

1.4

1.2

1.4

Fig.31: Two-dimensional gel images of peroxiredoxin 1 of liver mitochondria. The gels were

prepared and protein spots were analyzed as described in experimental procedure. The

partial 2D-gel images are shown and the corresponding protein spots were indicated with

arrow. The corresponding time-plot is shown in sub-figure b.

5.8.6 Regucalcin decreased with age in liver mitochondria

Again on the 2D-gels of liver mitochondria, a significant decrease of regucalcin protein spot

(pI: 5.15, MW: 33kDa) was observed (fig.32). The relative intensity of regucalcin dropped

along with age. In the 24-months group, only 65% of the regucalcin amount was available

compared to that of newborn group.

a) Change of regucalcin in liver mitochondria

Age (month)

502410 15 20

Ratio to newborn

5 month

24 months

Newborn 5 months 10 months

15 months 20 months

b) Time plot of regucalcin in liver mitochondria

1.2

1.4

a) Change of regucalcin in liver mitochondria

Age (month)

502410 15 20

Ratio to newborn

5 month

24 months

Newborn 5 months 10 months

15 months 20 months

b) Time plot of regucalcin in liver mitochondria

1.2

1.4

a) Change of regucalcin in liver mitochondria

Age (month)

502410 15 20

Age (month)

502410 15 20

Ratio to newborn

5 month

24 months

Newborn 5 months 10 months

15 months 20 months

b) Time plot of regucalcin in liver mitochondria

1.2

1.4

Fig. 32: Two-dimensional gel image insets of regucalcin protein spot in liver mitochondria

(a). The protein spot corresponding to regucalcin were indicated with arrow. The time-plot

of synuclein is shown in b.

5.8.7 Increase of a mitochondrial ribosomal protein in liver

Further rightward on the liver mitochondrial 2D-gels, increase of relative intensity of a

protein spot was detected. This protein spot was identified to be a mitochondrial ribosomal

protein L12 (pI: 9.34, MW: 22kDa). The increase level was especially pronounced at the

20-months age group, which accounted for more than three times compared to the

newborn age stage (fig.33). A significant decrease from 20-months to 24-months was

followed.

24 months

Newborn 5 months 10 months

15 months 20 months

Age (month)

502410 15 20

Ratio to newborn

a) Change of mitochondrial ribosomal L12 protein in liver mitochondria

b) Time plot of mitochondrial ribosomal L12 protein in liver mitochondria

24 months

Newborn 5 months 10 months

15 months 20 months

Age (month)

502410 15 20

Age (month)

502410 15 20

Ratio to newborn

a) Change of mitochondrial ribosomal L12 protein in liver mitochondria

b) Time plot of mitochondrial ribosomal L12 protein in liver mitochondria

Fig.33: Two-dimensional gel images of mitochondrial ribosome L12 protein in liver

mitochondria. The partial 2D-gel images are shown and the corresponding protein spots

were indicated with arrow. The time-plot of this spot is shown in b.

5.8.8 Alteration of ubiquinol-cytochrome c reductase binding protein and

ATP synthase subunit

On the liver mitochondria 2D-gels, one of the respiratory chain complex III subunit, the

ubiquinol-cytochrome c reductase binding protein (pI: 9.1, MW: 14kDa) showed significant

increase profile at 20-months age stage (fig.34a). In brain mitochondrial 2D-gels, the

subunit F of ATP synthase F0 complex (pI: 9.36, MW: 12kDa) showed increased level at

both 5-months and 20-months age groups (fig.34b).

Age (month)

Ratio to newborn

Age (month)

502410 15 20

a) Change of ubiquinol-cytochrome c reductase binding protein in liver mitochondria

b) Change of ATP synthase F0 subunit F in brain mitochondria

502410 15 20

Ratio to newborn

Age (month)

Ratio to newborn

Age (month)

502410 15 20

Age (month)

502410 15 20

a) Change of ubiquinol-cytochrome c reductase binding protein in liver mitochondria

b) Change of ATP synthase F0 subunit F in brain mitochondria

502410 15 20

Ratio to newborn

502410 15 20

Ratio to newborn

Fig.34: Time-dependent level change of ubiquinol-cytochrome c reductase binding protein

in liver mitochondria (a) and ATP synthase F0 subunit F in brain mitochondria (b). Data

were obtained through evaluation of 2D-electrophoresis gels of mouse brain and liver

mitochondria of six different ages.

5.9 Result of Western immunoblotting

Using the Western immunoblotting method, we tested all three commercially available

antibodies against mtDNA-encoded proteins: anti-human COX subunit I, II and III

(Molecular Probes, Göttingen Germany). Among these three antibodies, only anti-COX I

specifically cross-react with mouse mtDNA-encoded COX subunit I. The possible change of

COX subunit I, a mtDNA encoded protein, was analysed using the TOM20 as an internal

control for protein amount. The quantitative results are shown in figure 35.

1,2

1,4

1,6

COXI/TOM20 Ratio

Age (month)

502410 15 20

1,2

1,4

1,6

COXI/TOM20 Ratio

Age (month)

502410 15 20

Fig.35: Immunoblotting analysis of the brain mitochondrial “Fraction III” (20

g) from pooled

sample of different age groups. The anti-COX subunit I antibody recognized specifically the

mtDNA-encoded COX I (57 kDa) protein in the mitochondrial inner membrane.

Quantification was performed by reference to mitochondrial translocase of outer membrane

TOM20. An asterisk showing the significant increase at 15-months (p<0.05).

As can be seen in fig.35, there was no significant change in COX I level among all aging

stages except for 15-months stage. A pronounced increase at 15-months aging stage

compared to other aging stages was observed. This was followed by a drop-back at 20-

months group. This phenomenon was observed in Western immunoblots of both brain and

liver mitochondria. This increase of COX I level at 15-months aging stage was shown to be

significant comparing to other groups (p<0.05).

5.10 Preliminary result of Blue-native electrophoresis

In the last stage of this study, we explored the possibility of quantitatively study of further

respiratory chain subunits in the frame of protein complex. For this purpose, we first

estimated the mitochondria complexes with the help of a 100kDa molecular sieve using

Blue-native electrophoresis analysis.

After the first dimension of Blue-native electrophoresis, proteins bands with molecular

weights similar to those of complexes I, V, III and IV (complex I >910 kDa; complex V > 550

kDa; complex III 490 kDa; complex IV 204 kDa (Arnold et al., 1998; Ludwig et al., 1998))

were clearly visible on the stained gel slab (result not shown). After the second dimension

separation using SDS-PAGE, protein subunits of different complexes were resolved into

individual subunits (fig.36).

Over 34 protein spots were resolved on the mini-scale gel. Among these, 26 protein spots,

corresponding to 37 proteins were successfully identified using MALDI-TOF mass

spectrometry (tab.A2 in appendix). This suggests a complex protein mixture property of

protein spots. At least 28 of the 37 identified proteins are known mitochondrial proteins.

Among them, seven proteins were mitochondrial respiratory chain subunits, including one

protein (COX subunit IV) that was a membrane protein containing two trans-membrane

domains. Another calcium-binding protein calreticulin (with two trans-membrane domains)

was also identified. However, no mtDNA-encoded protein was detected.

Complex I

910kDa

Complex V

550kDa

Complex III

490kDa

Complex IV

204kDa

32 ?

44?

Complex I

910kDa

Complex V

550kDa

Complex III

490kDa

Complex IV

204kDa

32 ?

44?

Fig.36: Two-dimensional Blue-native electrophoresis of the mitochondria fraction from

mouse muscle mitochondria. Protein (1mg mitochondrial pellet as starting material) was

separated in the first dimension by Blue-native electrophoresis and in the second

dimension by SDS-PAGE as detailed in chapter 3.2.15. The protein spots were visualised

using preparative silver staining. The position of complexes I, III, IV and V were indicated.

The protein identification results of marked protein spots are shown in table A2.

5.11 Mathematical simulation and model fitting

We used a mathematical model that describes the relationship between the mtDNA

mutation and the mitochondria population dynamics. This model was proposed whereby

the rates of mitochondrial replication were under feedback control of the available

mitochondria in the cell (Kowald and Kirkwood, 2000). Given initial values and standard

parameters, the equations could be solved numerically for different variables against time

(fig.37).

200 400 600 800 1000 1200

200

400

600

800

1000

Number of

mitochondria

Time (day)

Mw (intact

mitochondria)

Rad (radical

level)

Mm (mutated

mitochondria)

200 400 600 800 1000 1200

200

400

600

800

1000

Number of

mitochondria

Time (day)

Mw (intact

mitochondria)

Rad (radical

level)

Mm (mutated

mitochondria)

Fig.37: One example of a serious of mathematical simulation results of the model. The

population of wild type mitochondria decrease with age, while the number of mutated

mitochondria in a cell increase with age. Standard parameters described in table 1 and

mutation rate value k=1.2x10-8 per gene per day was used to produce these curves. Mw:

number of intact mitochondria in a cell; Mm: number of mutated mitochondria in a cell; Rad:

concentration of free radicals in the cell.

From the time-dependent curves of wild type mitochondria (Mw), mutated mitochondria

(Mm) and free radical (Rad), an expected tendency of the decrease of wild type

mitochondria and the accumulation of mutated mitochondria in a cell could be deduced.

Varying the mutation rate value “k” from 10-3 to 10-12 per gene per day, the alteration

velocity of intact mitochondria and defective mitochondria could be altered accordingly

(fig.38).

Time (day)

Number of intact mitochondria

Time (day)

Number of intact mitochondria

Fit.38: A summary of simulation results by varying the mutation rate “k” from 10-3 to 10-12

per gene per day. Only the number of intact mitochondria along with time is plotted on the

diagram.

In the following table (tab.9), the experimental data are the average concentration ratios of

NADH-ubiquinone oxidoreductase 13 kDa-A subunit, NADH-ubiquinone oxidoreductase 1

alpha subcomplex 5 and Cytochrome c oxidase, subunit Vb (see tab.8) in mouse brain and

liver mitochondria. These data were obtained from 2D-electrophoresis experiments. As

reference time point, the protein concentrations at newborn group were set as one. The

values accounted alteration of -0.476% per day and -0.025% per day for mouse brain and

liver mitochondria, respectively. In the same table, values calculated from the mathematical

model were listed at the right side of the experimental data, with the mtDNA mutation rate

at 1.2x10-8 per gene per day.

Holding the assumption that the experimental data and modeling data could be closely

associated, we applied linear regression analysis to model the relationship between these

two time-dependent variables.

Tab.9: Linear regression analysis of experimental data and values calculated from the

mathematical model:

Brain mitochondria Liver mitochondria

Experimental

data Modeling data Experimental

data Modeling data

Newborn 1.0 1.0 1.0 1.0

5-months 0.778 ± 0.104 0.998 0.981 ± 0.060 0.998

10-months 0.911 ± 0.113 0.994 0.995 ± 0.023 0.995

15-months 0.746 ± 0.070 0.973 0.942 ± 0.017 0.983

20-months 0.767 ± 0.064 0.860 0.859 ± 0.026 0.929

24-months 0.610 ± 0.072 0.546 0.854 ± 0.014 0.780

r=0.76 r=0.83 Linear

regression

analysis p<0.1 p<0.05

Alteration rate

[per day]

-0.476% -0.319% -0.025% -0.023%

After linear regression analysis using the least square model, the data set built from mtDNA

mutation rate as 1.2x10-8 per gene per day gave the best fit (r=0.76, p<0.1 for brain

mitochondria, or r=0.83, p<0.05 for liver mitochondria). To view this fit, the computed

regression line was plotted, with the experimental data on the Y-axis and the simulation

result on the X-axis (fig.39). As can be deduced from the figure, most of the data points

were clustered around the regression line. Thus, the mtDNA mutation rate of mouse

mitochondrial DNA was determined to be 1.2x10-8 per gene per day, which was the most

suitable value according to our experimental data obtained from 2D-electrophoresis.

Linear fit

Data points (Liver)

Linear fit

Data points (Brain)

r =0.76, p<0.1 r =0.83, p<0.05

Experimental data

Simulation results

Mouse liver

1.1

.5 .6 .7 .8 .9 11.1.7

.7 .8 .9 1

Simulation results

Mouse brain

Scale unit: Protein level compared to newborn

Linear fit

Data points (Liver)

Linear fit

Data points (Brain)

r =0.76, p<0.1 r =0.83, p<0.05

Experimental data

Simulation results

Mouse liver

1.1

.5 .6 .7 .8 .9 11.1

1.1

.5 .6 .7 .8 .9 11.1.7

.7 .8 .9 1

Simulation results

Mouse brain

Scale unit: Protein level compared to newborn

Fig.39: Linear fitting of experimental data gained from the 2D-electrophoresis and modeling

result, which was calculated with standard parameter and mutation value of 1.2x10-8 per

gene per day. As reference time points, the protein concentrations at newborn group were

set as one.

6 Discussion

6.1 Investigation on mitochondrial theory of aging needs proteomic

approaches

As more and more age-related degenerative disease being linked to mitochondrial

pathology, mitochondria become viable pharmacological targets for the treatment of

numerous diseases, as well as for the possible retardation of aging process (Grad et al.,

2001; Morin et al., 2001). In order to develop reliable therapeutic intervention, the age-

related mitochondrial change and its mechanism need to be investigated and understood at

the first line.

The current study undertaken was for the purpose of getting an insight of the mitochondrial

proteome change during the aging process. Driven by the hypothesis of mitochondrial

theory of aging, the emphasis of this study was put on the possible effect of mitochondrial

somatic mutations which could be responsible for some aspects of the aging process

(Harman, 1972; Linnane et al., 1989).

6.2 Mouse has been proven to be pertinent model organism for aging study

For this purpose, we chose an inbred mouse strain as the model of human aging.

Advantages of using mouse model include the similar genetic background of inbred strain,

the ability to control precisely the developmental and aging stages, as well as the existence

of large body of reference data. However, because the phylogenic lineages leading to the

human being and the mouse are thought to have diverged about 90 million years ago

(Copeland et al., 2002), we first strengthen to validate the mouse model.

Abundant references have verified the close correlation of cytochrome c oxidase activity

deficiency in muscle fibre to the extent of mtDNA rearrangements in aged individuals

(Brierley et al., 1998; Cooper et al., 1992; Trounce et al., 1989). To the best of our

knowledge, however, no respective data on mouse exist.

Thus, we have chosen the histochemical staining of COX-activity on young and old mouse

muscle tissue as our criterion of model validation (Sugiyama et al., 1993). As expected, we

observed the prominent “mosaic pattern” on COX-stained old muscle tissue, similar to that

been described for aged human muscle. This suggests that C57BL/6 strain mouse could

be a pertinent model organism for human aging.

6.3 The choice of organs

As to organs of choice in this study, we chose both brain and liver tissues. The brain tissue

consume about 20% of the total oxygen in the body (Wade and Bishop, 1962). The

relatively high metabolic activity of brain and the low mitotic rate of neuron could possibly

accelerate the free radical induced damages (Brierley et al., 1998).

Many age-related diseases affect brain tissue but not liver (Wallace, 1992). Respecting the

mitochondrial aspect involved, it is of interest to investigate such tissue-specifity. Using

both liver and brain tissues, we could be able to investigate the influence of mitotic rate on

mitochondria mutation accumulation.

6.4 The choice of strategy

Characterization of proteins in whole tissues is sometimes difficult to accomplish

particularly for low abundant proteins or hydrophobic proteins (Murayama et al., 2001).

Sub-cellular fraction of mitochondria permits the isolation of our target organelles in pure

form in order to perform relevant studies (Jung et al., 2000; Lopez et al., 2000; Xie, 2003).

In order to further increase sensitivity of our proteomic investigation, we combined sub-

cellular fractionation to sub-fractionation of mitochondrial proteins, in order to reveal low

abundant proteins, especially to enhance the capacity of membrane protein resolution.

In this study, a total of 116 novel spots or spots with enhanced intensity were present on

the mitochondrial 2D-gels compared to 2D-gels of brain whole protein extract. This

observation indicates that the sub-cellular fractionation procedure utilized in this study

largely depleted the pure mitochondria fraction from other cellular proteins. The reduced

spot pattern of mitochondrial gel also facilitated efficient gel evaluation procedure. This

demonstrated that sub-cellular fractionation could intensify the low-abundant proteins and

facilitate their easier detection and evaluation.

More importantly, the change of mitochondria-associated alpha synuclein that otherwise

could have escaped identification demonstrated that the isolation of cell organelle brought

us additional cellular location information. This sub-cellular fractionation technique would

also be a strategy for the initial identification of previously unknown proteins and for their

assignment to particular sub-cellular localization or interaction (Bell et al., 2001; Neubauer

et al., 1998).

In the same line, pre-fractionation of proteins using sequential extraction strategy could

enrich certain hydrophobic proteins, which were not pronounced in total cell lysates. The

detection of seven membrane proteins in “Fraction II” demonstrated the effect of protein

pre-fractionation strategy.

6.5 Mitochondrial isolation was successful

It is important for sub-proteomic studies to obtain intact cellular components, so that

contamination of irrelevant proteins could be minimized and cellular localization information

preserved. The electronic micrographs of isolated mitochondria in this study showed that

isolated mitochondria were largely intact, with purity over 94%. The yields of mitochondria

from brain and liver tissue were comparable to that of other authors (Fernandez-Vizarra et

al., 2002; Jungblut and Klose, 1985). Probably due to the higher water-content of newborn

tissue (Holland et al., 1986), the newborn mice materials gave lower mitochondrial yield

compared to the remaining aging stages. On the other hand, the detection of 27 protein

spots that representing cytosolic contamination showed that minimal cellular

contaminations can be revealed by more sensitive analysis methods such like the 2D-

electrophoresis analysis.

Comparing the mitochondria isolation from fresh and frozen materials, it was observed that

the yield of mitochondria from frozen material was lower compared to that of fresh material,

indicating the damage of mitochondrial membrane through freeze-thaw effect. This could

have contributed to the possible contamination of 24-months mitochondria samples. On the

other hand, the aspiration of only intact mitochondrial band after the gradient centrifugation

enabled our investigation of non-damaged mitochondria in the down-stream analyses. The

isolation of mitochondria from old organism has been described as difficult in previous

literature (Frese and Stahl, 1992). However, this was not validated in our study respecting

the quality of mitochondria isolated from young and old mouse tissue.

Notice, however, that there was a much smaller yield of mitochondria from brain tissue, as

was also observed by other authors (Jungblut and Klose, 1985). This could be due to the

small amount of starting material of brain tissue, as well as the special tissue structure of

brain organ (personal communication with Dr. Wallace DC).

Nevertheless, for mouse brain tissue which has a mean weight of about 0.4g, an average

amount of 1mg pure mitochondria could be obtained, which corresponds to 280µg total

protein (Brown, 1991). This was sufficient for a large-gel 2D-electrophoresis analysis, which

needs about 200µg of proteins. Taken together, the successful isolation of mitochondria of

sufficient yield in this study demonstrates that it is feasible to carry out sub-cellular

fractionation for proteomic studies.

6.6 Protein fractionation was effective

The dynamic range of protein abundance within the cells has been estimated to be as high

as 107 (Lopez and Melov, 2002). As approximately 90% of the total protein of a typical cell

is made up of only 10% of the 10,000 to 20.000 abundant protein species (Miklos and

Maleszka, 2001; Zuo et al., 2001), many low-abundance proteins may not be detectable by

conventional methods. In such a complex protein mixture, pre-fractionation has shown be

beneficial to facilitate the identification of low-abundant proteins (Lopez and Melov, 2002)

In this study, we have employed a sequential extraction strategy to increase the relative

concentration of low abundant proteins in different fractions. Based on Klose (Klose, 1999),

combined with different disciplinarians such as from Molloy (Molloy et al., 1999), we

separated mitochondrial proteins into three fractions.

Since the capacity of protein solvation in a buffer system is not unlimited (Brown, 1991), the

high representation of highly soluble proteins in the solution makes the solution of

hydrophobic protein more difficult. Thus, we first applied Tris-buffer to extract most soluble

proteins. This “Fraction I” obtained through Tris-buffer extraction contained mostly water-

soluble proteins.

By pellet suspension in a buffer containing 0.1% Triton-X100, proteins of intermediate

solubility could be released and collected as “Fraction II”. The use of Triton-X100 was

borrowed by Villa (Villa et al., 1998), originally aimed at the isolation of a mitochondrial

inner membrane protein with calcium transport activity. Combined with a nonionic detergent

Triton-X100 in the isoelectric focusing separation (Stephenson et al., 1980), we detected

seven membrane proteins with trans-membrane domain on the 2D-gels of “Fraction II”.

This suggests that the sequential extraction strategy could be useful for the separation of

mitochondrial membrane proteins.

Due to the presence of organic solutions, methanol-chloroform extraction was a selective

method favouring hydrophobic proteins, with the aim that membrane protein could be

extracted from the phospholipids bilayers (Molloy et al., 1999). Yerushima has described

similar method to extract only highly hydrophobic membrane proteins from bacterial

membrane (Yerushalmi et al., 1995).

The amount of protein in three different fractions showed that there is still a substantial part

(17%) of proteins remained after Triton-buffer extraction (remained in the phospholipids).

This justified the use of methanol-chloroform extraction. However, irreversible precipitation

of hydrophobic proteins did happen during the resolvation procedure in this study, which

resulted in insoluble pellet. This indicates that organic solvents does not provide good

results for the resolution of membrane proteins (Kashino, 2003). Together, this sequential

extraction protocol allowed us to increase the relative abundance of less soluble proteins,

albeit to a limited extent.

6.7 Satisfactory result from 2D-PAGE analysis was obtained

Different proteomic approaches are suitable for different categories of proteins. The 2D-

electrophoresis method developed by Klose and O’Farrel has been proved to be superiors

for separating soluble proteins (Klose, 1975; O'Farrell, 1975). This method has been widely

used for the investigation of the proteomic and genetic changes in a global way. In this

study, large-gel 2D-electrophoresis method was successfully applied on “Fraction I” and

“Fraction II” protein analysis.

In order to monitor technical variations of the 2D-electrophoresis method undertaken in this

study, we controlled the reproducibility of 2D-gels in this study using same samples as well

as the experimental samples.

Most of the protein spots achieved satisfactory quantitative reproducibility. This could be

concluded through the following points: First, the presence or absence of the majority of

spots of different gels was the same, indicating a stable resolution of the gels. Second, the

relative positions of the spots were similar on different gels. Thirdly, the relative intensity of

the majority of spots was similar on all gels. Theses three criteria could also be deduced

from the high value of correlation coefficients.

Although gels from same sample beard higher correlation coefficients over the gels from

experimental samples, the variation of all 2D-gels carried out in this study were well under

20%. This indicates that the gel-to-gel variation in our 2D-electrophoresis system was small

enough to allow one to analyse changes in mitochondrial protein in the aging process

(Chang et al., 2003). This constitutes a solid prerequisite for the subsequent gel image

evaluation. Combined with multiple gel runs and appropriate statistical methods, we were

able to identify differentially expressed spots at different age groups with statistical

confidence.

On the other hand, it was noticed that many spots could not be matched automatically,

although they show apparent counterpart property in manual comparison. This indicates

that the current software for 2D-gel analysis was still not mature for the automatic gel

image evaluations. Thus, manual gel evaluation was conducted before software

quantifications in this study.

6.8 Gel image evaluation was successful

Quantitative and qualitative modifications can frequently be reflected on the protein spot

pattern change in the 2D-electrophoresis system (O'Farrell, 1975). Electrophoretic mobility

changes of proteins in the horizontal direction could be due to difference of amino acid

substitutions involving change in charge (Klose et al., 2002), or different protein

modifications, such as phosphorylation (Robinson and Pauling, 1974). The vertical shift

could result from either molecular mass alterations or changes in protein conformation that

affects the shape of the SDS-protein complex (de Jong et al., 1978; Klose et al., 2002).

Different post-translational modifications of the same protein commonly lead to spot slitting

patterns, which represent different isoforms.

According to these theoretical guidelines and previous experiences, numerous protein spot

differences were successfully detected. The underlying cause for protein spot pattern

variations may be somatic mutation in the structural gene, or in regulatory sequences

(promoter, enhancer etc). Increase or decrease of cellular concentration of certain proteins

may either be the result of a decreased synthesis or an increased degradation rate.

6.9 Protein identification was efficient

Protein identification is another factor largely influencing a proteomic study. Although

MALDI-TOF-MS identification has been considered to be an effective high-throughput

method, it has a certain signal capacity. If two or more proteins present in the same protein

spots, the spectra from those overlapping protein peptide mixtures may confuse the

database search and largely decrease the search score (Karty et al., 2002). Combined with

the ESI Iontrap mass spectrometry which pre-fractionate the peptide mixture using reverse-

phase HPLC, this disadvantage can be avoided.

In this study, a satisfactory 72% of the spots were identified, with an average of sequence

coverage over 50%. This also validates the effectively of mass spectrometric protein

identification and the usefulness of preparative silver staining.

Respecting proteins that were not successfully identified in this study, there are two major

points of considerations: either the protein has not been annotated in the data bank, or the

MS-spectra data were not accurate or comprehensive enough to distinguish between

several entries in the database. If the database search is not fruitful, then further

information would be required. This could be achieved by tandem MS experiments to

determine the amino acid sequences of the individual proteolytic peptides contained in the

digest mixture (Johnson and Biemann, 1987).

6.10 2D-electrophoresis combined with mass spectrometry is an efficient

proteomic strategy

The 2D-electrophoresis analysis has undergone a long way of establishing period and is

presently a relative matured method (Klose, 1975; O'Farrell, 1975). As an effective

hypothesis-less method, large-gel 2D-electrophoresis facilitates the global proteomic

comparison through gel evaluation.

Using differential proteomic analysis, it is possible to carry out classical perturbation

analysis through the comparison of normal condition to stimulated condition, or the

comparison of the time stages. Thus, this system is especially suitable for aging studies.

Our data showed that the reproducibility of 2D-PAGE analysis is reliable enough that it can

be used to detect protein level variation by demonstrating the qualitative and quantitative

changes of protein spots (Klose et al., 2002).

Although the 2D-electrophoresis method is highly effective for soluble proteins, the under-

representation of trans-membrane-proteins, especially integrate membrane proteins on the

2D-gel was a serious drawback of the method (Santoni et al., 2000). For this reason, we

have applied other proteomic approaches such like Western immunoblotting and Blue-

native electrophoresis in parallel.

6.11 Analysis of membrane proteins remains a problem

Big efforts have been paid in the modification of the 2D-electrophoresis method trying to

improve the resolution of membrane proteins (Molloy, 2000). Based on previous literatures,

we applied Triton-X100 in the isoelectric focusing gel (Stephenson et al., 1980). Including

this modification on “Fraction II” proteins, seven membrane proteins were detected, which

beard Gravy score ranging from 0.139 to -0.72, with one to two trans-membrane domains.

The fact that these spots were available only on 2D-gel patterns of “Fraction II” proteins

suggests the beneficial effects of protein pre-fractionation procedure and/or the use of

Triton-X100 in the electrophoresis system.

However, none of the protein identified in this study contained more than two trans-

membrane domain, neither were we able to resolve more than one mtDNA-encoded

proteins using this system. This indicates that 2D-electrophoresis method is still not

powerful enough for the analysis of highly hydrophobic integral membrane proteins, as also

been shown by previous studies (Santoni et al., 1999).

The loss of hydrophobic proteins probably occurs during the whole experimental process. It

is believed to be due to unable to extract the proteins from membrane double layers;

unable to solve these proteins into the solution; and protein precipitation during the transfer

from first to second dimension (Adessi et al., 1997; Gygi et al., 2000).

6.12 Mitochondrial protein profile change during the aging process

Respecting our initial goal of accessing age-related changes in mitochondrial proteome, we

discuss in the following the ten proteins that were observed to change significantly during

the aging process.

6.12.1 Down-regulation of complex I and complex IV subunits indicates

mtDNA mutation

In the current study, two nuclear-DNA-encoded subunits of complex I (NADH-ubiquinone

oxidoreductase 13 kDa-A subunit and NADH-ubiquinone oxidoreductase 1 alpha

subcomplex 5) and one subunit of complex IV (COX Vb in brain) were observed to down

regulate during the aging process. As has been suggested by Remacle, this could indicate

deficiency of respiratory chain complex subunits induced by abnormalities of mitochondrial

DNA (Remacle et al., 2004). A decreased representation of mtDNA-encoded complex I and

complex IV subunits could in turn lead to a reduced requirement of other subunits for the

protein complex assembly. Nevertheless, other influences such as possible nuclear DNA

mutation or post-translational modification are not to be excluded.

Previous works on human material showed that most mtDNA deletions ablate the region

between the genes encoding ATP8 and cytochrome b, which codes predominantly for

subunits of complex I and complex IV subunit (Cortopassi and Wong, 1999; Vu et al., 2000)

(see fig.1). This could imply that complex I and complex IV are the most affected protein

complexes through mtDNA mutation. Here we showed that similar phenomenon could also

be observed on mouse model at protein level.

In a concert, Bowling showed that there was an age-associated progressive impairment of

complex I (over 22% by 80 years) and complex IV activity (Bowling et al., 1993; Yen et al.,

1991). Cooper reported activity decrease of complex I (59%) and complex IV (47%)

comparing donor groups of 20-30 and 60-90 years (Cooper et al., 1992). In the current

study, we observed significant COX-activity deficiency in part of the old mouse myocytes.

Together with our current result, these indicate that such activity decrease of complex I and

complex IV could be directly linked to the protein steady state concentration decrease.

6.12.2 Increase of complex III and complex V subunits suggest feedback

regulation

In a chain of enzyme reaction, if the activity of an up-stream (complex I) enzyme changes,

then all of the down-stream metabolite pools and enzyme activities could be altered (Stryer,

1995). As the cell responds to the decrease of complex I and complex IV functionality,

increased synthesis of protein of down-stream enzyme complexes in the respiratory chain

would soon follow, intending to maintain the required functional levels of the whole

respiratory chain.

Two nuclear-encoded proteins localized in the down-stream of respiratory chain reaction,

ubiquinol-cytochrome-c binding protein (complex III subunit in liver mitochondria) and ATP

F0 (complex V subunit in brain mitochondria) were observed to increase with age. This

indicates such a negative feedback mechanism, which is common in metabolic pathway

control.

6.12.3 MtDNA-encoded COX subunit I showed only moderate change

In order to verify the proposed decrease of nuclear-DNA-encoded respiratory chain

subunits due to mtDNA mutation, it was our interest to investigate the possible alteration of

mtDNA-encoded subunits. As a hypothesis-driven investigation, we employed Western

immunoblotting to probe the concentration change of an mtDNA-encoded complex IV

subunit, COX I, in brain and liver mitochondria.

Interestingly, only a transient up-regulation of COX subunit I at 15-months age group was

significant in the WB analysis, the concentration of COX I dropped back to average level at

20-months aging stage. This result did not support our hypothesis. However, as can be

deduced from figure 1, COX I is not inside the most deleted region on the mtDNA molecule.

In effect, several authors described that mtDNA encoded COX I and COX II (but not COX

III) present at elevated levels in Alzheimer disease patients compared to normal control

individual (Hirano et al., 1997; Wallace, 1997). This could indicate enhanced protein

synthesis in the mitochondrial matrix as a response to mtDNA mutation. In this respect, our

observation of the up-regulation of mitochondrial ribosomal protein L12 could be

considered as such compensatory attempt to enhance the mitochondrial protein expression

due to decreased mitochondrial respiratory chain function.

On the other hand, it need be noticed that decrease of protein concentration at 15-months

age group is a seldom case in the range of our experimental results. Several other proteins

showed also specific lower protein concentration at 15-months age group. This could

suggest possible systematic error.

6.12.4 Decrease of a mitochondrial heat-shock protein could suggest the

increased consumption of heat shock protein

Respecting the respiratory chain complex I deficiency in old age, Bandy hypothesized that

the mitochondrial genome damage would increase the steady-state concentration of

reduced intermediate of the respiratory chain, leading to formation of more free radicals

through their auto-oxidation (Bandy and Davison, 1990).

Heat-shock responses are a fundamental and widespread type of cellular defense against

environmental stress (Hansen et al., 2003). The down-regulation of mitochondrial heat

shock protein in both brain and liver mitochondria could be linked to the increased

consumption of heat shock proteins. This in turn suggests that mitochondria in aged

individual may not be able to deal with oxidative stress in a sufficient scale. Similar results

were also reported by other authors (Zabel et al., 2002).

6.12.5 Down-regulation of peroxiredoxin suggest elevated oxidative stress in

aged individual

Peroxiredoxins are enzymes catalyzing the destruction of peroxides. The decrease of

peroxiredoxin level was commonly discussed as increased oxidation or aggregation of

these proteins due to enhanced oxidative stress (Rabilloud et al., 2002). The down-

regulation of peroxiredoxin 1 was also observed in this present study. However, so far, no

indication of mitochondria-association of peroxiredoxin 1 has been reported.

Because peroxiredoxin is an abundant antioxidant in the cell, we cannot exclude the

possibility of contamination of cytoplasmic proteins. Upon performing protein-protein

sequence alignment, peroxiredoxin 1 could be aligned to two mitochondria-associated

proteins with high similarity scores. We speculate that we could have observed the

decrease of certain mitochondria-specific peroxiredoxin, or the protein detected could be a

substrate of mitochondria-associated proteinase. More detailed studies in this respect

would be obligatory before further discussion.

Several proteins that had previously been located to other cellular compartments were

found in yeast or human heart mitochondria (Sickmann et al., 2003; Taylor et al., 2002).

This suggests that such proteins may be specifically associated with mitochondria. They

could be linked to the mitochondria though either binding to the outer membrane of

mitochondria (beta actin), or transport material to the mitochondria. It is not clear how

closely these proteins are associated in vivo and whether the connection is biologically

relevant.

6.12.6 Down-regulation of regucalcin in liver mitochondria indicates a lowered

buffering capacity of calcium

In our 2D-electrophoresis analysis of “Fraction I” proteins, we have observed that a

calcium-binding protein regucalcin decreased significantly with age in the liver

mitochondria. Interestingly, regucalcin, also called senescent marker protein 30 (SMP30),

has been intensively investigated since 1992, when it was first identified as an androgen-

independent marker protein that is down-regulated during the aging process (Fujita et al.,

1998; Fujita et al., 1992). The amount of regucalcin in aged rat liver decreased to 40% of

that in adult rat liver (Fujita et al., 1998). In the current study, a similar decrease range of

42% was observed comparing the regucalcin level in mouse liver mitochondria of 5-months

group to that of 24-months group.

This decrease of regucalcin was observed only in liver mitochondria, not in brain. This is

probably due to the much lower normal concentration of regucalcin in brain compared to

liver (Yamaguchi and Isogai, 1993). Andreyev and Fiskum reported that brain and liver

possess different mechanism for the calcium induced cytochrome c release (Andreyev and

Fiskum, 1999). The treatment of Ca2+ induced the mitochondrial membrane permeability

transition (MPT) in liver but not in brain. They suggested that the membrane permeability

transition is responsible for liver cytochrome c release, while certain MPT-independent

mechanism is responsible for the release of cytochrome c from brain mitochondria.

As a second messenger for the common signal transduction elements in a cell, intracellular

calcium is a key player in regulation of various cellular functions such as proliferation,

differentiation and adhesion (Stryer, 1995). Prolonged elevation of calcium has been

suggested to induce apoptosis by stimulating Ca2+ Ma2+ dependent endonucleases (Kaiser

and Edelman, 1977). A low cytoplasmic Ca2+ concentration of living cells is maintained by

energy-requiring pumps. Theses pumps either remove calcium to the extra-cellular space

by transporting it across the plasma membrane or accumulate it inside intracellular

organelles such as the mitochondria and endoplasmic reticulum (Carafoli and Zurini, 1982;

van Os, 1987).

Regucalcin has been shown to enhances ATP-dependent calcium pump enzyme activity in

both plasma membrane and isolated mitochondria (Takahashi and Yamaguchi, 2000; Xue

et al., 2000). Ishigami found that hepatocytes from regucalcin-knock-out mouse were highly

susceptible to tumour necrosis factor-alpha- and Fas-mediated apoptosis (Ishigami et al.,

2002). The specific binding of regucalcin on a glycoprotein located on the mitochondrial

outer membrane has been verified by Panfili (Panfili et al., 1980).

Based on the pure fraction of isolated mitochondria utilized in our proteomic study, we

reason that we have measured the change of mitochondria-associated regucalcin

concentration (Fountoulakis and Schlaeger, 2003). This could suggest that less amount of

regucalcin is bound on liver mitochondrial outer membrane in the aged stages.

Recently, Ishigami A (2004) reported that regucalcin-deficient mice accumulate lipid

droplets in liver, while the liver mitochondria were abnormally enlarged. This could suggest

that regucalcin-deficiency profoundly affect mitochondrial function or transport of metabolite

into mitochondria (which could be ion-dependent).

Taken together, down-regulation of mitochondria-associated regucalcin could be

associated with alterations in the mitochondrial buffering capacity of Ca2+ and calcium

signalling in the aged. However, whether this is a up-stream or down-stream event of

mitochondrial dysfunction during aging is still unknown. Further analysis would be required

to investigate the detailed mechanism.

6.12.7 Up-regulation of alpha-synuclein in brain mitochondria resembles

neuronal degenerative diseases

Another interesting finding using the 2D-electrophoresis is the significant increase of alpha-

synuclein level in the brain mitochondrial proteins during the aging process. This finding

raised our interest because alpha-synuclein gives also a prominent spot on the 2D-gel

pattern of brain total protein extract. Less prominent increase tendency was observed upon

comparing the 2D-gels of whole brain proteins in different aging stages (1 week to 25-

months, courtesy to Mrs. Herrmann). The differentially pronounced pattern of the same

protein on mitochondrial 2D-gels indicates that we have observed the increase of

mitochondria-associated alpha-synucleins.

100

Alpha-synuclein is predominantly a neuron-specific presynaptic protein (Hsu et al., 1998).

Among others, it has the function of protecting against oxidative stress via interaction with

the stress-signalling pathway in neuronal cells (Hashimoto et al., 2002). Since the

identification of alpha-synuclein as a prominent protein component in the aggregates of

amyloid bodies (Goedert, 1997; Ueda et al., 1993), it has been intensively studied in the

fibrogenesis processes of brain tissue.

The relation of alpha-synuclein to mitochondria first drew attention when it was noticed that

MPTP, a toxin specifically inhibiting respiratory chain complex I, was depleted completely of

its toxicity in alpha-synuclein knock-out mice (Dauer et al., 2002; Kuhn et al., 2003).

Focusing on the role of mitochondria in the neuronal fibrogenesis procedure, Perrin and

Dawson & Dawson observed that alpha-synuclein is specifically associated with membrane

compartments in cultured cells and brain tissue through interactions with acidic head

groups of phospholipids (Dawson and Dawson, 2003; Perrin et al., 2001).

Lately, Song links over-expression of alpha-synuclein to mitochondrial dysfunction in vivo

(Song et al., 2004). Their morphological study showed that alpha-synuclein-treated mice

had significantly greater mitochondrial abnormalities than either saline-treated controls or

MPTP-treated wild-types. Hence, they speculated that mitochondria play a role in early

stage of neuronal fibrogenesis that could be induced by alpha-synuclein-related protein

aggregation.

Our current study showed that the increased level of mitochondria-associated alpha-

synuclein could be a common phenomenon in the normal aging process. The translocation

of increasing amount of alpha-synuclein proteins to the mitochondrial membrane with time

might play an important role in either triggering or perpetuating age-related neuronal

fibrogenesis. Our observation thus links the mechanism of aging closer to that of neuronal

degenerative disease.

Integrating other observations of this current study, we take one step ahead to speculate

that Complex I function deficiency in old age is probably up-stream to aggregation of alpha-

synuclein in the aging process. Deficiency of complex I induced by mtDNA mutation could

create an elevated oxidative stress that ultimately leads to aggregation of alpha-synuclein

and the neuronal apoptosis.

101

6.12.8 Difference between brain and liver respecting mitochondrial aspect of

aging

In this study, the change of NADH-ubiquinone oxidoreductase 13 kDa-A subunit, COXVb

and 10kDa heat shock protein were observed in both brain and liver mitochondria, while

the change of other proteins were observed in either liver or brain (see tab.8). These

observation could suggest tissue-specifity.

The decrease rate of 10kDa mitochondrial heat shock protein was more pronounced in liver

than in brain mitochondria. This could be related to the higher metabolic rate of liver tissue

compared to brain, which lead to a higher production of reactive oxygen species.

The average decrease rate of both NADH-ubiquinone oxidoreductase 13 kDa-A subunit

and COX Vb in brain were higher than that of in liver. The decrease of NADH-ubiquinone

oxidoreductase 1 alpha subcomplex 5 was observed in brain tissue only. Taken together,

brain mitochondria showed higher decrease rate of protein steady state concentration in

both complex I and complex IV subunit. This could indicate that brain is more susceptible to

the secondary loss of nuclear DNA encoded complex I and IV protein subunits.

Another phenomenon observed in this study is that the feedback mechanism functions

more efficient in liver mitochondria than in brain mitochondria in the respect of protein

synthesis machinery. This could be deduced from the following observation: First, there

was an increase of mitochondrial ribosomal protein in liver mitochondria, but not in brain;

Second, at the presence of complex I deficiency, the up-regulation of complex III subunit

was observed in liver mitochondria, while the up-regulation of more down-stream complex

V was observed in brain tissue. This could suggest that liver has higher capacity of

feedback regulation compared to brain tissue. This less efficient feedback mechanism of

brain mitochondria could further contribute to its higher vulnerability.

One possible explanation for the phenomenon mention above could be the influence of

mitotic rate. Brain has much lower mitotic rate, which means a reduced rate of

mitochondrial renewal. While the liver with higher mitotic rate is able to dilute mutated

mitochondria more effectively (Kowald and Kirkwood, 2000).

Respecting the tissue-specific protein profile changes of regucalcin and alpha-synuclein

observed in this study, it could be deduced that brain is more susceptible to cellular

fibrogenesis, while liver more vulnerable to abnormal calcium buffering.

102

Brown suggested that the cellular protein concentrations are generally close to the

maximum solubility in cells (Brown, 1991). The extraordinary complex cellular structure of

brain tissue could mean a frequent violation to the maximal protein concentration. This in

turn gives a high susceptibility of brain cells to protein aggregation.

As the central processing field of cellular metabolism, liver cells have been designed

against this fault, since the majority of the metabolic functions require ion-formed

macromolecules (Stryer, 1995). However, large amount of ions in the cell could means a

high ionic strength, which in turn makes the hepatocytes vulnerable for ion management.

Another difference of brain and liver observed in this study is the finding that old-aged

mouse brain tissues (20-months and 24-months stage), but not old-aged mouse liver tissue

gave significant smaller yield of mitochondria compared to their young counterparts (all

other aging stages). Notice that only intact mitochondria with density between 1.09 and

1.13 g/ml could be focussed during the continuous gradient purification method. Damaged

mitochondria migrate to other density layer as diffused material. Thus, the possibility

remains that the defective mitochondria were not isolated in our sub-cellular fractionation

since they bear densities other than that of intact mitochondria.

Based on this reasoning, we speculate that there could be less amount of intact

mitochondria in old brain tissue, due to increased representation of defective mitochondria.

If this should be the case, its influence should be considered seriously upon interpreting the

result of down-stream proteomic analysis.

6.13 Potential of Blue-native electrophoresis analysis

Valuable results have been gained from the 2D-electrophoresis analysis in this study.

However, only one mtDNA-encoded subunit has been identified so far. This is largely due

to the highly hydrophobic properties of the mtDNA-encoded proteins, that could precipitate

at the basic pole during IEF (Hanson et al., 2001; Lopez et al., 2000; Santoni et al., 2000).

Although Western immunoblotting remains an effective hypothesis-driven alternative

against specific protein targets, the scarity of commercially available antibody against

mtDNA-encoded protein stronly obstacate this efforts.

Confronting this serious shortness, we employed Blue-native electrophoresis method as an

additional strategy to investigate mitochondrial proteins in the frame of intact protein

complex. Another potential advantage of Blue-native electrophoresis is that functional

103

information regarding protein interactions within the complex could be retained due to the

non-denaturing conditions of the first dimension electrophoresis system.

In the preliminary experiment of Blue-native electrophoresis, we were able to resolve four

mitochondrial respiratory chain complexes at the first dimension. Only the Complex II,

which is the smallest respiratory complex (130 kDa), eluded detection after preparative

silver staining. An additional subunit of respiratory chain Complex IV containing two trans-

membrane domains was also resolved. This demonstrates the potential usefulness of Blue-

native electrophoresis as an additional strategy in further studies focusing on the alteration

of mitochondrial respiratory chain proteins. The detection of another calcium binding

protein on the Blue-native gel could suggest its functional association with mitochondria.

Due to the small gel format used in this prelimeary experiment, protein subunits in the

same protein complex were not optimally resolved at the second dimension (SDS-PAGE).

This could be seen from the complex protein mixture property which overstrained the

MALDI-TOF-MS measurement. A larger gel format combined with pre-fractionation of MS

measurement could lead to the identification of more mitochondrial-specific proteins.

Traditionally, Blue-native electrophoresis is not a quantitative analytical method. However,

it could be possible to combine modern techniques, such as the florescent staining method

in order to improve the ability of this method respecting protein quantification.

6.14 The accumulation of defective mitochondria with age was simulated

The recognition that aging is mechanistically complex has brought the need for in silico

approaches. These approaches are following the spirit of Joel Keizer, who first recognized

that many problems in molecular biology could be formulated as physical-chemical

processes and studied by modern tools of nonlinear dynamical systems (Hastings, 2001).

Among the most controversial hypothesis of aging are those involved in the progressive

accumulation of error-bearing or altered macromolecules with advancing age. The model of

Kowald has been a theoretical prove of the mitochondrial theory of aging (Kowald and

Kirkwood, 2000). However, since the presence of numerous parameters of unknown

values, it has not been tractable to prove it experimentally.

Based on the model proposed by Kowald, we have employed a strongly reduced

mathematical model for the dynamic discussion of mitochondrial theory of aging. In our

current model, the accumulation of mutated mitochondria in the aging process can be

104

treated as dynamic system containing a collection of components (wild type mitochondria,

mutated mitochondria and free radicals), the properties of which change with time as they

response to interactions among the components. Our model also included the protein

maximum concentration hypothesis by Brown (Brown, 1991) in order to further reduce

parameter dimension. Such construct also reflects the mitochondria-host-cell dependency.

As can be seen in figure 37, our model faithfully simulated the accumulation of mutated

mitochondria in the aging process. Notice that there exist a “quasi-stable phase” of both

curves of intact and defective mitochondria at the beginning of simulation period.

Subsequently, the defective mitochondria accumulated dramatically, accompanied by the

rash decrease of intact mitochondria. The model allowed us to test the hypothesis of

mitochondrial theory of aging by comparing the simulated behaviour of the model with the

observed behaviour of the biological subject.

6.15 Mutation rate of mouse mtDNA was estimated

The somatic mutation rate of mtDNA is an important aspect directly influencing the

validation of mitochondrial theory of aging. In case the mutation rate is smaller than a

certain threshold value, the strength of mutation could be neglected, so the mitochondrial

mutation will not be a driving force of the aging process. Unfortunately, It is currently

difficult to directly measure the level of mutant mtDNA within living cells (Chinnery and

Samuels, 1999; Linnane et al., 1989). The major goal of the utilization of model in this

study was to calculate the mouse mtDNA mutation applying our experimental data obtained

using proteomic approach.

For the calculation of mtDNA mutation rate, we used data of nuclear-encoded complex I

and complex IV subunit alterations. This was based on two considerations: first, the mtDNA

mutation affects predominantly respiratory complex I and complex IV genes (Cortopassi

and Wong, 1999; Vu et al., 2000); second, the physiological stoimetry of complex assembly

deduce that the reduction level of mtDNA-encoded subunit of complex I should be

proportional to the reduction of nuclear-encoded subunits (Remacle et al., 2004).

Notice here that we did not use complex III and complex V subunits. This was based on the

observation of feedback phenomenon in our experiment, which caused up-regulation in

down-stream enzyme complexes in the respiratory chain. Furthermore, the strength of

feedback inhibition is not known.

105

For efficient parameter scanning, we assumed that the change of wild type mitochondria

was directly correlated with the change of mean value of the respiratory chain complex I

and IV subunits. This was based on the simplified assumption that only wild type contained

respiratory complex I and complex IV subunits, wile mutated mitochondria were totally

depleted of this subunits (Remacle et al., 2004).

Assembly of the functional respiratory chain complexes requires the coordinated

contribution of subunits synthesized in both the cytoplasm and the mitochondria. In case

there is a lack of mtDNA-encoded subunits, there should be correlated lack of nuclear

DNA-encoded subunits in the same protein complex (di Rago et al., 1997).

Under the current setting of both experiments and mathematical simulation in this study,

mtDNA mutation rate of mouse (this strain C57BL/6) could be calculated as 1.2x10-8 per

gene per day. This corresponds to 1.7x10-7 per gene per mitochondrial genome replication.

This result is consistent with to the data of Shenkar, who predicted the mutated rate of

4977bp deletion to be 5.95x10-8 per mitochondrial genome replication (Shenkar et al.,

1996).

With this mutation rate, dramatic accumulation of mutated mitochondria happens at

approximately 600 days of age in our mouse model. This ensures that the vast majority of

mitochondria remain intact throughout the developmental and reproduction phase of mouse

life. According to our simulation result, at the time point of 800 days, which is the average

lifespan of the C57BL/6 mouse (Rowlatt et al., 1976), the mitochondria population is

consisted of 30% of intact mitochondria and 70% of defective mitochondria. This correlates

well with the observation of mitochondrial dysfunction in old age. It also indicates that the

mitochondrial somatic mutation is a factor that is not to be neglected respecting aging

process.

Interestingly, both brain and liver got the same mutation rate value, in case the

mitochondrial turnover rate “α” for liver was set to 0.073 d-1. This corresponds to a

replication rate of liver mitochondria to be 9.5 days. This indicates that even with the same

mutation rate, the brain tissue could accumulate mutated mitochondria at a higher rate

compared to liver. Thus, the faster accumulation of mutated mitochondria in brain could be

largely due to the lower mitotic rate.

The purpose of modeling in this study was to use mathematical simulation to test abstract

biological hypothesis of the mitochondrial theory of aging. Using the current model, we

106

were able to estimate the mutation rate of mtDNA of mouse, which is the key player in the

whole system. Through more advanced experiments, we could be able to determine

additional factors that influence the length of this “quasi-stable phase”. In this way, it would

be possible to find intervention strategy in order to retard aging.

6.16 Result of current study is consistent with the mitochondrial theory of

aging

About 15 years ago, it was proposed that aging is caused by life-long accumulation of

mitochondrial mutations (Linnane et al., 1989), which comprise cellular energy metabolism

and increase intracellular oxidative stress. This could result in the development of the

multiple degenerative changes in tissues at old age.

The experimental results gained in this study have been consistent with this view. Indirect

suggestion of mtDNA mutation was observed by the observation of complex I and complex

IV protein down-regulation and the feedback effects of down-stream complex subunits, as

well as the up-regulation of mitochondrial protein synthesis machinery. This indicates that

the key initiating event of aging and age-related degeneration could be a decline in

mitochondrial function, which leads to progressive oxidative damage that is exacerbated

along with time.

To the author’s personal opinion, a basic underlying ground for this pivotal role of

mitochondria in aging process could lie on the symbiotic nature of the modern eukaryotic

cells. Symbiotic of proto-eukaryotic cell and the Rickettsiales-like aerobic bacteria was

beneficial for the host cell, since it gained increased energy supply through extensive

oxidation of nutrient. This eminently promoted the subsequent evolution. However, this

benefit is accompanied by two major disadvantages: First, genome of microorganisms

have generally much higher mutation rates than that of eukaryotic organism. MtDNA

mutation was observed to occur on a frequent basis (Ferguson and von Borstel, 1992). As

a facultative parasitic microorganism, this frequent mutation contributed to their fast

adaptation to the environment. However, after becoming an obligatory symbiotic object, this

property turns catastrophe for their host cell due to the low selective pressure in the cellular

environment.

As can be seen in the mathematical simulation of this study, instead of being depleted,

defective mitochondria could achieve homoplasmy state under normal physiological

condition, taking the advantage of their slower degradation rate. This accumulation of

defective mitochondria constitute as a potential hazard for the cell that manifest with time.

107

Respecting this problem, the host eukaryotic cell has been transferring the mitochondrial

genes into the nuclei, in order to better perpetuate them. However, due to its inability of

doing it in a complete manner, this problem has been inherited to date.

Another problem of symbiotic with an aerobic microorganism is their constant generation of

free radical. Persistent oxidative stress could in turn cause further damage to DNA,

membrane and proteins. It seems that oxidative stress cannot be totally prevented despite

the cellular construction of diverse heat shock proteins and antioxidants. The decrease of

mitochondrial heat shock protein and peroxiredoxin profiles observed in the current study

suggests an elevated oxidative stress level in the mitochondria of the aged individuals.

Oxidative stress commonly acts as signal initiating apoptosis (Chen et al., 2001). It has

been suggest that cells harbouring mutant mtDNA are more prompt to apoptosis (Khrapko

et al., 1999). In this sense, mitochondrial DNA mutation could play an important role in

neurological and other age-related diseases, sharing apoptosis as common feature (Horton

et al., 1995).

The observed change of regucalcin and alpha-synuclein suggest that some genetic and

environmental factors that increase the susceptibility of cells to apoptosis might interact

with common molecular pathways. These include calcium signalling pathway or red-ox cell

signalling. The influence of genetic, tissue-specific and epigenetic factors could contribute

to the development of different pathologies. The insights obtained from the current

proteomic characterization of the aging process may also be applied to the role of

mitochondria in other age-related disorders.

Taking together, these finding indicate that, far from being merely an energy supplier of the

cell, mitochondria play key roles in red-ox cell signalling and apoptosis (Bogoyevitch et al.,

2000; Lai et al., 1996; Levonen et al., 2001; Liu et al., 1996). The emerging paradigm is a

complex cross-talk between mitochondrial and cellular functions, with reactive oxygen

species playing a key role.

108

7 Conclusion

In this study, we have put the mitochondrial theory of aging on trial at the proteomic aspect.

The mouse model (C57BL/6) was first validated as a suitable model for human aging, and

the effect of senescence on brain and liver mitochondria were assessed by proteomic

comparisons of different age groups.

Sub-cellular fractionation has been shown to enrich part of the low-abundant mitochondrial

proteins in the proteomic analysis. Sequential extraction of proteins, together with the

employment of additional detergent improved the resolution of membrane proteins on the

2D-gels to a certain extent. This shows that sub-cellular fractionation and protein pre-

fractionation are feasible and effective strategies for proteomic studies. Differential

proteomic studies combined with hypothesis-driven method such like immunoblotting was

shown to be powerful method to detect global protein profile changes. However, membrane

protein remains a problem in proteomic studies.

Numerous protein profile changes during the aging process were observed. They suggest

progressive mitochondrial dysfunction and increasing oxidative stress with advancing age.

Respecting the difference between post-mitotic and mitotic tissue, an increased

susceptibility to protein aggregation in the aged mouse brain was observed; while aged

liver tissue showed decreased capacity of calcium buffering. There was indication for more

severe mitochondrial abnormality in old brain compared to old liver.

Preliminary experiment showed that Blue-native electrophoresis could be potentially useful

for the investigations of respiratory complex subunits and other mitochondria-associated

proteins. Further experiments are on the way to improve the quantification capacity of this

method.

A mathematical model has been useful for the estimation of mtDNA mutation rate of mouse

based on experimental data. The result of the mathematical simulation confirms the

mitochondrial somatic mutation as an important factor in the aging process. Further

experimental data would be needed to improve the model, so as to further our

understanding of aging as a dynamic process. Only theoretical and experimental works in

concert can push forward our understanding of the dynamic aging process.

Aging is essentially a gradual decline in an organism’s capacity of responding to

environmental stress and return to the resting state. Central to the restoring of resting state

are a series of inter-related signaling pathways such as calcium signaling and controlled

109

protein or organelle degradation. Deficiency of such cellular mechanism will lead to

significant alterations in their restoring capacity. Energy deficiency will further exacerbate

this situation since most cellular processes are energy-dependent.

The evidence gained from this study was in concert to the mitochondrial theory of aging. By

summarizing the major evidences gained in this current study, we conclude that mutated

mitochondria can accumulate under normal physical conditions along with time. The

mitochondrial somatic mutation could directly lead to respiratory chain complex deficiency.

Mitochondrial dysfunction, in turn, causes the elevation of oxidative stress level in the cell.

Oxidative damage manifests itself partially through protein aggregation and calcium

buffering breakdown (fig.40).

Accumulation of

mutated mitochondria

Elevated

oxidative stress

Respiratory deficiency

Disturbance of

calcium buffering

Protein

aggregation

Reactive oxygen

speceis

Accumulation of

mutated mitochondria

Elevated

oxidative stress

Respiratory deficiency

Disturbance of

calcium buffering

Protein

aggregation

Reactive oxygen

speceis

Fig.40: The current study expanded our knowledge respecting mitochondrial theory of

aging. The mitochondrial somatic mutation could directly lead to mitochondrial dysfunction.

This, in turn, causes the elevation of oxidative stress level in the cell, which manifests itself

through protein aggregation and disturbed calcium buffering.

Through this study, we were able to expand our knowledge respecting the mitochondrial

theory of aging. This study also demonstrated that aging and age-related diseases could

be related through common pathways. It also provides a precondition for further differential

proteomic experiments and mathematical simulation. Together, this could ultimately provide

110

a rationale for mitochondria as a target for preventing from the adverb side effect of aging,

as well as to provide valid therapies in age-related diseases.

111

8 Outlook

Upon completion of this study, some further investigations would be required in the future

respecting the following issues. Diverse protein profile changes during the aging process

were observed in the current study. As the next step, analysis would be required to

investigate whether there exist underlying gene expression alternations, alteration in post-

transcriptional modifications or protein degradation control. Gene level control experiments

(RT-PCR, northern blot) and protein modification analysis would be helpful in answering

these questions. Blue-native gel electrophoresis combined with fluorescent staining could

be useful to gather additional information regarding the possible changes in protein

complex property.

More importantly, it deserves our understanding of detailed mechanism causing these

protein level alternations, to determine the up-stream factors and system interplay. For

example, how the proposed mitochondrial somatic mutations possibly lead to a high

susceptibility of brain to fibrogenesis, or the vulnerability of liver mitochondria to calcium

buffering.

Detailed investigations of the biochemical changes are still hampered by the lack of means

for the analysis of hydrophobic membrane proteins. Further improvement will be needed in

both protein extraction and protein separation strategy, in order to facilitate the

investigation of mtDNA-encoded proteins directly. Alternatively, the employment of

profound mass spectrometric and chromatographic strategies, such as multidimensional

liquid chromatography coupled online to tandem mass spectrometry could be potentially

useful to give indirect quantification of hydrophobic membrane proteins (Rabilloud et al.,

2003; Sickmann et al., 2003).

Upon investigating the proteomic alteration of mouse brain and liver mitochondria in this

study, the question arose whether and to what extent protein alterations that occur in one

organ are valid in other organs. Additionally, whether there is a homogeneous behaviour in

the same organ. Especially for highly differentiated tissue like brain, different anatomical

and functional areas could be largely heterogeneous (Itoh et al., 1996; Soong et al., 1992).

In the future, additional tissue types and tissue fractionations should be analyzed for their

proteomic characterization.

Respecting the modest observation of mtDMA-encoded protein alteration in this study,

another problem could be the low level of mtDNA mutation of a certain type which requires

a high sensitivity of the measurement method. A different picture could emerge if

112

techniques were used that allow one to investigate individual cells or pure cell

subpopulations. Laser capture micro dissection techniques could be methodology of choice

in the future.

Studies on whole organism model have been useful to access global changes related to

the aging process. Hypothesis-less differential proteomic analysis further reduces the

range of key factors in the dynamic system in our study. In the future, hypothesis-driven

experiments employing in vitro model could be useful to deduce the intrinsic relationship

between free-radical induced mitochondrial mutation and oxidative stress in the frame of

cellular senescence.

As an example, different oxidative stress accelerator (tert-butyl hydroperoxide (BHP)) or

inhibitors, respiratory chain enzyme inhibitors, heat shock treatment could be applied on

cell culture model to manipulate the cellular aging process, so as to measure the factors of

interests (Rabilloud et al., 2002). Genotoxic stress could be simulated by genetically

manipulation of mouse model with increased mtDNA mutation rate (Trifunovic et al., 2004).

Such experiments combined with advanced mathematical simulation will help to delineate

the contribution of mitochondria in aging process and age-related diseases. In the same

context, it would also be of interest to investigate the Influence of epigenetic factors on the

heterogeneous outcome of common pathway involved in the aging process, as well as

diversity of age-related diseases.

113

9 Zusammenfassung

Man nimmt an, dass die Anreicherung von Mitochondrien mit mutiertem Genom

maßgeblich am Alterungsprozess beteiligt ist (Wallace, 2001). Um die Auswirkung der

mtDNA-Veränderungen bei der Alterung auf Proteinebene zu erforschen, wurde das

mitochondriale Proteom während des Alterns anhand eines Mausmodells (C57/BL6)

untersucht.

Vor der Untersuchung des Proteoms wurde durch histochemische Färbung nachgewiesen,

dass die Zytochrome C Oxydase (COX) Aktivität bei alten (24 Monate) im Gegensatz zu

jungen Mäusen (2 Woche) im Muskelgewebe abnimmt. Während bei jungen Mäusen keine

COX-negativen Muskelzellen gefunden wurden, zeigte ein bedeutender Teil der Myozyten

(43%) im Muskel von alten Mäusen eine reduzierte COX-Aktivität. Dies bestätigt, dass die

Lebensspanne der Maus für die Untersuchung der mitochondrialen Alterung ausreicht.

Aus dem Gehirn und der Leber von Mäusen wurden Mitochondrien an sechs

unterschiedlichen Zeitpunkten (von Neugeboren bis zu einem Alter von 24 Monaten, n=8

bis 13 pro Zeitpunkt) mit Hilfe einer kontinuierlichen Gradienten-Zentrifugation isoliert.

Mitochondriale Proteine wurden mit einer sequenziellen Extraktionsstrategie basierend auf

Tris-Puffer ("Fraktion I"), Triton-Puffer ("Fraktion II") und Methanol-Chloroform ("Fraktion

III") extrahiert. Großgel 2D-Elektrophorese (2DE) und eine modifizierte 2D-Elektrophorese

Prozedur (unter Verwendung von Triton-X100) wurden für die Analyse der Proteine aus

„Fraktion I", beziehungsweise "Fraktion II", verwendet. Bei "Fraktion III" wurde mit Hilfe von

Westernblots Veränderungen im Expressionsniveaus des mtDNA-kodierten Protein COX

Untereinheit I untersucht.

Die Expression von zwei Untereinheiten des Komplexes I (NADH-Ubiquinone

Oxidoreductase 13 kDa-A Untereinheit und NADH-Ubiquinone Oxidoreductase 1 alpha

Subcomplex 5) und einer Untereinheit des Komplexes IV (COX Untereinheit Vb) der

Atmungskette nahm mit zunehmendem Alter ab. Eine Untereinheit von Komplex III

(Ubiquinol-Zytochrome c Reductase bindendes Protein), eine Untereinheit von Komplex V

(ATP F0 Untereinheit F) und ein mitochondrial-ribosomales Protein zeigten ein erhöhtes

Expressionsniveau im Alter. Diese Ergebnisse zeigen, dass das Fehlen von Komponenten

der Komplexe I und IV der Atmungskette in alterndem Gewebe durch Feedbackregelung

anderer Proteinkomplexe in der Atmungskette begleitet wird. Diese Beobachtungen stützen

die Hypothese, dass die Ansammlung von mtDNA Mutationen überwiegend Gene für die

Komplexe I und IV beeinflusst (Vu et al., 2000).

114

Die Herunterregulierung des 10 kDa mitochondrialen Hitzeschock-Proteins deutet auf ein

erhöhtes Niveau an oxidativem Stress im alternden Mausgehirn und in der Leber hin.

Oxidativer Stress tritt sowohl bei der Alterung als auch bei neurodegenerativen

Erkrankungen auf (Cottrell et al., 2000, Richter et al., 1988). Die Erhöhung der Expression

des mitochondrien-assozierten alpha-Synuclein im Gehirn könnte auf eine erhöhte

Anfälligkeit zur Bildung von Proteinaggregaten mit fortschreitendem Alter hindeuten

(Goedert, 1997; Ueda et al., 1993). Diese sind ebenfalls einen Kennzeichen von einigen

neurodegenerativen Erkrankungen. Die Abnahme des mitochondrien-assoziierten

Regucalcin in der Leber weist auf eine gesenkte mitochondriale Pufferungsfähigkeit für

Kalziumionen hin (Takahashi und Yamaguchi, 2000; Xue et al., 2000).

Ein mathematisches Modell wurde entwickelt, um die Anhäufung von defekten

Mitochondrien während des Alterns zu simulieren. Mit den experimentell erhobenen

quantitativen Daten aus der Großgel 2DE wurde eine mtDNA Mutationsrate von 1.2x10-8

pro Gen und Tag abgeschätzt. Diese Mutationsrate ist groß genug, um eine Anreicherung

von defekten Mitochondrien während des Alterns innerhalb der Lebensspanne einer Maus

zu verursachen.

Die experimentellen Daten, die durch die Untersuchung des mitochondrialen Proteoms

gewonnen wurden, unterstützen die Hypothese, dass sich mitochondriale Veränderungen

mit dem Alter anreichern. Dies erklärt mitochondriale Funktionsstörungen und die Zunahme

von oxidativen Stresses während des Alterns. Zukünftige Untersuchungen werden sich auf

die Identifizierung weiterer, mtDNA-kodierter Proteinexpressionsveränderungen und die

Proteininteraktionen während des Alterns konzentrieren.

Schlüsselwörter:

Mitochondiren, Alterung, Proteom, zweidimensionale Proteinelektrophorese,

Mathematischemodellierung

115

10 Reference

Adessi, C., Miege, C., Albrieux, C., and Rabilloud, T. (1997). Two-dimensional electrophoresis of membrane

proteins: a current challenge for immobilized pH gradients. Electrophoresis 18, 127-135.

Ames, B. N., Shigenaga, M. K., and Hagen, T. M. (1993). Oxidants, antioxidants, and the degenerative diseases

of aging. Proc Natl Acad Sci U S A 90, 7915-7922.

Amson, R. B., Nemani, M., Roperch, J. P., Israeli, D., Bougueleret, L., Le Gall, I., Medhioub, M., Linares-Cruz,

G., Lethrosne, F., Pasturaud, P., et al. (1996). Isolation of 10 differentially expressed cDNAs in p53-induced

apoptosis: activation of the vertebrate homologue of the drosophila seven in absentia gene. Proc Natl Acad

Sci U S A 93, 3953-3957.

Anderson, N. G., and Anderson, N. L. (1996). Twenty years of two-dimensional electrophoresis: past, present

and future. Electrophoresis 17, 443-453.

Andreyev, A., and Fiskum, G. (1999). Calcium induced release of mitochondrial cytochrome c by different

mechanisms selective for brain versus liver. Cell Death Differ 6, 825-832.

Arnold, I., Pfeiffer, K., Neupert, W., Stuart, R. A., and Schagger, H. (1998). Yeast mitochondrial F1F0-ATP

synthase exists as a dimer: identification of three dimer-specific subunits. Embo J 17, 7170-7178.

Bandy, B., and Davison, A. J. (1990). Mitochondrial mutations may increase oxidative stress: implications for

carcinogenesis and aging? Free Radic Biol Med 8, 523-539.

Baumer, A., Zhang, C., Linnane, A. W., and Nagley, P. (1994). Age-related human mtDNA deletions: a

heterogeneous set of deletions arising at a single pair of directly repeated sequences. Am J Hum Genet 54,

618-630.

Bell, A. W., Ward, M. A., Blackstock, W. P., Freeman, H. N., Choudhary, J. S., Lewis, A. P., Chotai, D., Fazel,

A., Gushue, J. N., Paiement, J., et al. (2001). Proteomics characterization of abundant Golgi membrane

proteins. J Biol Chem 276, 5152-5165.

Bodenteich, A., Mitchell, L. G., and Merril, C. R. (1991). A lifetime of retinal light exposure does not appear to

increase mitochondrial mutations. Gene 108, 305-309.

Boffoli, D., Scacco, S. C., Vergari, R., Solarino, G., Santacroce, G., and Papa, S. (1994). Decline with age of the

respiratory chain activity in human skeletal muscle. Biochim Biophys Acta 1226, 73-82.

Bogoyevitch, M. A., Ng, D. C., Court, N. W., Draper, K. A., Dhillon, A., and Abas, L. (2000). Intact

mitochondrial electron transport function is essential for signalling by hydrogen peroxide in cardiac

myocytes. J Mol Cell Cardiol 32, 1469-1480.

Boguski, M. S., and Schuler, G. D. (1995). ESTablishing a human transcript map. Nat Genet 10, 369-371.

Bohr, V. A., and Dianov, G. L. (1999). Oxidative DNA damage processing in nuclear and mitochondrial DNA.

Biochimie 81, 155-160.

Bowen, R. L., and Atwood, C. S. (2004). Living and dying for sex. A theory of aging based on the modulation

of cell cycle signaling by reproductive hormones. Gerontology 50, 265-290.

Bowling, A. C., Mutisya, E. M., Walker, L. C., Price, D. L., Cork, L. C., and Beal, M. F. (1993). Age-dependent

impairment of mitochondrial function in primate brain. J Neurochem 60, 1964-1967.

Brierley, E. J., Johnson, M. A., James, O. F., and Turnbull, D. M. (1997). Mitochondrial involvement in the

ageing process. Facts and controversies. Mol Cell Biochem 174, 325-328.

Brierley, E. J., Johnson, M. A., Lightowlers, R. N., James, O. F., and Turnbull, D. M. (1998). Role of

mitochondrial DNA mutations in human aging: implications for the central nervous system and muscle. Ann

Neurol 43, 217-223.

Brockington, M., Sweeney, M. G., Hammans, S. R., Morgan-Hughes, J. A., and Harding, A. E. (1993). A

tandem duplication in the D-loop of human mitochondrial DNA is associated with deletions in mitochondrial

myopathies. Nat Genet 4, 67-71.

Brookes, P. S., Pinner, A., Ramachandran, A., Coward, L., Barnes, S., Kim, H., and Darley-Usmar, V. M.

(2002). High throughput two-dimensional blue-native electrophoresis: a tool for functional proteomics of

mitochondria and signaling complexes. Proteomics 2, 969-977.

116

Brown, G. C. (1991). Total cell protein concentration as an evolutionary constraint on the metabolic control

distribution in cells. J Theor Biol 153, 195-203.

Carafoli, E., and Zurini, M. (1982). The Ca2+-pumping ATPase of plasma membranes. Purification,

reconstitution and properties. Biochim Biophys Acta 683, 279-301.

Cartwright, J. H. E., and Piro, O. (1992). The Dynamics of Runge-Kutta Methods. Int J Bifurcations Chaos 2,

427-449.

Challapalli, K. K., Zabel, C., Schuchhardt, J., Kaindl, A., Klose, J., and Herzel, H. (2004). High reproducibility

of large gel two dimensional electrophoresis. Electrophoresis accepted.

Chang, J., Van Remmen, H., Cornell, J., Richardson, A., and Ward, W. F. (2003). Comparative proteomics:

characterization of a two-dimensional gel electrophoresis system to study the effect of aging on

mitochondrial proteins. Mech Ageing Dev 124, 33-41.

Chen, Y., Stanford, A., Simmons, R. L., Ford, H. R., Hoffman, R. A., Cai, J., Anders, M. W., Stevens, J. L., and

Jones, D. P. (2001). Nitric oxide protects thymocytes from gamma-irradiation-induced apoptosis in

correlation with inhibition of p53 upregulation and mitochondrial damage. Cell Immunol 214, 72-80.

Chinnery, P. F., and Samuels, D. C. (1999). Relaxed replication of mtDNA: A model with implications for the

expression of disease. Am J Hum Genet 64, 1158-1165.

Coligan, J. E., Dunn, B. M., and Ploegh, H. L. (1995). Current Protocols in Protein Sciences. In, D. W.

Speicher, and P. T. Wingfield, eds. (New York, John Willey & Sons), pp. 94-99.

Coller, H. A., Bodyak, N. D., and Khrapko, K. (2002). Frequent intracellular clonal expansions of somatic

mtDNA mutations: significance and mechanisms. Ann N Y Acad Sci 959, 434-447.

Cooper, J. M., Mann, V. M., and Schapira, A. H. (1992). Analyses of mitochondrial respiratory chain function

and mitochondrial DNA deletion in human skeletal muscle: effect of ageing. J Neurol Sci 113, 91-98.

Copeland, N. G., Jenkins, N. A., and O'Brien, S. J. (2002). Genomics. Mmu 16--comparative genomic

highlights. Science 296, 1617-1618.

Corral-Debrinski, M., Horton, T., Lott, M. T., Shoffner, J. M., Beal, M. F., and Wallace, D. C. (1992).

Mitochondrial DNA deletions in human brain: regional variability and increase with advanced age. Nat

Genet 2, 324-329.

Cortopassi, G. A., and Arnheim, N. (1990). Detection of a specific mitochondrial DNA deletion in tissues of

older humans. Nucleic Acids Res 18, 6927-6933.

Cortopassi, G. A., and Arnheim, N. (1992). Using the polymerase chain reaction to estimate mutation

frequencies and rates in human cells. Mutat Res 277, 239-249.

Cortopassi, G. A., and Wong, A. (1999). Mitochondria in organismal aging and degeneration. Biochim Biophys

Acta 1410, 183-193.

Cottrell, D. A., Blakely, E. L., Borthwick, G. M., Johnson, M. A., Taylor, G. A., Brierley, E. J., Ince, P. G., and

Turnbull, D. M. (2000). Role of mitochondrial DNA mutations in disease and aging. Ann N Y Acad Sci 908,

199-207.

Cullinane, C., and Bohr, V. A. (1998). DNA interstrand cross-links induced by psoralen are not repaired in

mammalian mitochondria. Cancer Res 58, 1400-1404.

Dallman, P. R. (1967). Cytochrome oxidase repair during treatment of copper deficiency: relation to

mitochondrial turnover. J Clin Invest 46, 1819-1827.

Dauer, W., Kholodilov, N., Vila, M., Trillat, A. C., Goodchild, R., Larsen, K. E., Staal, R., Tieu, K., Schmitz,

Y., Yuan, C. A., et al. (2002). Resistance of alpha -synuclein null mice to the parkinsonian neurotoxin

MPTP. Proc Natl Acad Sci U S A 99, 14524-14529.

Dawson, T. M., and Dawson, V. L. (2003). Molecular pathways of neurodegeneration in Parkinson's disease.

Science 302, 819-822.

de Grey, A. D. (1997). A proposed refinement of the mitochondrial free radical theory of aging. Bioessays 19,

161-166.

de Jong, W. W., Zweers, A., and Cohen, L. H. (1978). Influence of single amino acid substitutions on

electrophoretic mobility of sodium dodecyl sulfate-protein complexes. Biochem Biophys Res Commun 82,

532-539.

117

di Rago, J. P., Sohm, F., Boccia, C., Dujardin, G., Trumpower, B. L., and Slonimski, P. P. (1997). A point

mutation in the mitochondrial cytochrome b gene obviates the requirement for the nuclear encoded core

protein 2 subunit in the cytochrome bc1 complex in Saccharomyces cerevisiae. J Biol Chem 272, 4699-

4704.

DiMauro, S., Hirano, M., Schon, E. A., and Andreu, A. L. (2000). Mitochondrial encephalomyopathies:

therapeutic approaches. Neurol Sci 21, S901-908.

Emelyanov, V. V. (2001). Evolutionary relationship of Rickettsiae and mitochondria. FEBS Lett 501, 11-18.

Feeney, L., and Berman, E. R. (1976). Oxygen toxicity: membrane damage by free radicals. Invest Ophthalmol

15, 789-792.

Ferguson, L. R., and von Borstel, R. C. (1992). Induction of the cytoplasmic 'petite' mutation by chemical and

physical agents in Saccharomyces cerevisiae. Mutat Res 265, 103-148.

Fernandez, J., Gharahdaghi, F., and Mische, S. M. (1998). Routine identification of proteins from sodium

dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels or polyvinyl difluoride membranes

using matrix assisted laser desorption/ionization-time of flight-mass spectrometry (MALDI-TOF-MS).

Electrophoresis 19, 1036-1045.

Fernandez-Vizarra, E., Lopez-Perez, M. J., and Enriquez, J. A. (2002). Isolation of biogenetically competent

mitochondria from mammalian tissues and cultured cells. Methods 26, 292-297.

Finch, C. E. (1993). FRAR course on laboratory approaches to aging. Theories of aging. Aging (Milano) 5, 277-

289.

Finch, C. E., and Tanzi, R. E. (1997). Genetics of aging. Science 278, 407-411.

Fountoulakis, M., and Schlaeger, E. J. (2003). The mitochondrial proteins of the neuroblastoma cell line IMR-

32. Electrophoresis 24, 260-275.

Frese, D., and Stahl, U. (1992). Oxidative stress and ageing in the fungus Podospora anserina. Mech Ageing

Dev 65, 277-288.

Fujita, T., Shirasawa, T., Inoue, H., Kitamura, T., Maruyama, N., Sato, N., and Shimosawa, T. (1998). Hepatic

and renal expression of senescence marker protein-30 and its biological significance. J Gastroenterol

Hepatol 13, S124-131.

Fujita, T., Uchida, K., and Maruyama, N. (1992). Purification of senescence marker protein-30 (SMP30) and its

androgen-independent decrease with age in the rat liver. Biochim Biophys Acta 1116, 122-128.

Giavalisco, P. (2003) Proteomanalyse von Pfanzen, Dissertation, Freien Universität Berlin, Berlin.

Goedert, M. (1997). Familial Parkinson's disease. The awakening of alpha-synuclein. Nature 388, 232-233.

Grad, J. M., Cepero, E., and Boise, L. H. (2001). Mitochondria as targets for established and novel anti-cancer

agents. Drug Resist Updat 4, 85-91.

Gras, R., Muller, M., Gasteiger, E., Gay, S., Binz, P. A., Bienvenut, W., Hoogland, C., Sanchez, J. C., Bairoch,

A., Hochstrasser, D. F., and Appel, R. D. (1999). Improving protein identification from peptide mass

fingerprinting through a parameterized multi-level scoring algorithm and an optimized peak detection.

Electrophoresis 20, 3535-3550.

Green, D. R., and Amarante-Mendes, G. P. (1998). The point of no return: mitochondria, caspases, and the

commitment to cell death. Results Probl Cell Differ 24, 45-61.

Gygi, S. P., Corthals, G. L., Zhang, Y., Rochon, Y., and Aebersold, R. (2000). Evaluation of two-dimensional

gel electrophoresis-based proteome analysis technology. Proc Natl Acad Sci U S A 97, 9390-9395.

Hansen, J. J., Bross, P., Westergaard, M., Nielsen, M. N., Eiberg, H., Borglum, A. D., Mogensen, J.,

Kristiansen, K., Bolund, L., and Gregersen, N. (2003). Genomic structure of the human mitochondrial

chaperonin genes: HSP60 and HSP10 are localised head to head on chromosome 2 separated by a

bidirectional promoter. Hum Genet 112, 71-77.

Hanson, B. J., Schulenberg, B., Patton, W. F., and Capaldi, R. A. (2001). A novel subfractionation approach for

mitochondrial proteins: a three-dimensional mitochondrial proteome map. Electrophoresis 22, 950-959.

Harley, C. B., Futcher, A. B., and Greider, C. W. (1990). Telomeres shorten during ageing of human fibroblasts.

Nature 345, 458-460.

Harman, D. (1972). The biologic clock: the mitochondria? J Am Geriatr Soc 20, 145-147.

118

Harman, D. (1981). The aging process. Proc Natl Acad Sci U S A 78, 7124-7128.

Harry, J. L., Wilkins, M. R., Herbert, B. R., Packer, N. H., Gooley, A. A., and Williams, K. L. (2000).

Proteomics: capacity versus utility. Electrophoresis 21, 1071-1081.

Hashimoto, M., Hsu, L. J., Rockenstein, E., Takenouchi, T., Mallory, M., and Masliah, E. (2002). alpha-

Synuclein protects against oxidative stress via inactivation of the c-Jun N-terminal kinase stress-signaling

pathway in neuronal cells. J Biol Chem 277, 11465-11472.

Hastie, N. D., Dempster, M., Dunlop, M. G., Thompson, A. M., Green, D. K., and Allshire, R. C. (1990).

Telomere reduction in human colorectal carcinoma and with ageing. Nature 346, 866-868.

Hastings, A. (2001). In memoriam: Joel E. Keizer. J Theor Biol 210, 133-139.

Hayakawa, M., Hattori, K., Sugiyama, S., and Ozawa, T. (1992). Age-associated oxygen damage and mutations

in mitochondrial DNA in human hearts. Biochem Biophys Res Commun 189, 979-985.

Henshaw, P. S. (1947). The biological effects of pile radiations. Radiology 9, 349-364.

Herskovits, T. T., Carberry, S. E., and Villanueva, G. B. (1985). Subunit dissociation of Busycon canaliculatum

hemocyanin. Biochim Biophys Acta 828, 278-289.

Heukeshoven, J., and Dernick, R. (1988). Improved silver staining procedure for fast staining in PhastSystem

Development Unit. I. Staining of sodium dodecyl sulfate gels. Electrophoresis 9, 28-32.

Hillenkamp, F., Karas, M., Beavis, R. C., and Chait, B. T. (1991). Matrix-assisted laser desorption/ionization

mass spectrometry of biopolymers. Anal Chem 63, 1193A-1203A.

Hirano, M., Shtilbans, A., Mayeux, R., Davidson, M. M., DiMauro, S., Knowles, J. A., and Schon, E. A. (1997).

Apparent mtDNA heteroplasmy in Alzheimer's disease patients and in normals due to PCR amplification of

nucleus-embedded mtDNA pseudogenes. Proc Natl Acad Sci U S A 94, 14894-14899.

Hirokawa, T., Boon-Chieng, S., and Mitaku, S. (1998). SOSUI: classification and secondary structure prediction

system for membrane proteins. Bioinformatics 14, 378-379.

Hobot, J. A., Carlemalm, E., Villiger, W., and Kellenberger, E. (1984). Periplasmic gel: new concept resulting

from the reinvestigation of bacterial cell envelope ultrastructure by new methods. J Bacteriol 160, 143-152.

Holland, B. A., Haas, D. K., Norman, D., Brant-Zawadzki, M., and Newton, T. H. (1986). MRI of normal brain

maturation. AJNR Am J Neuroradiol 7, 201-208.

Horton, T. M., Graham, B. H., Corral-Debrinski, M., Shoffner, J. M., Kaufman, A. E., Beal, M. F., and Wallace,

D. C. (1995). Marked increase in mitochondrial DNA deletion levels in the cerebral cortex of Huntington's

disease patients. Neurology 45, 1879-1883.

Hsu, L. J., Mallory, M., Xia, Y., Veinbergs, I., Hashimoto, M., Yoshimoto, M., Thal, L. J., Saitoh, T., and

Masliah, E. (1998). Expression pattern of synucleins (non-Abeta component of Alzheimer's disease amyloid

precursor protein/alpha-synuclein) during murine brain development. J Neurochem 71, 338-344.

Ishigami, A., Fujita, T., Handa, S., Shirasawa, T., Koseki, H., Kitamura, T., Enomoto, N., Sato, N., Shimosawa,

T., and Maruyama, N. (2002). Senescence marker protein-30 knockout mouse liver is highly susceptible to

tumor necrosis factor-alpha- and Fas-mediated apoptosis. Am J Pathol 161, 1273-1281.

Itoh, K., Weis, S., Mehraein, P., and Muller-Hocker, J. (1996). Cytochrome c oxidase defects of the human

substantia nigra in normal aging. Neurobiol Aging 17, 843-848.

Jazin, E. E., Cavelier, L., Eriksson, I., Oreland, L., and Gyllensten, U. (1996). Human brain contains high levels

of heteroplasmy in the noncoding regions of mitochondrial DNA. Proc Natl Acad Sci U S A 93, 12382-

12387.

Joenje, H. (1989). Genetic toxicology of oxygen. Mutat Res 219, 193-208.

Joenje, H., Gille, J. J., Oostra, A. B., and Van der Valk, P. (1985). Some characteristics of hyperoxia-adapted

HeLa cells. A tissue culture model for cellular oxygen tolerance. Lab Invest 52, 420-428.

Johnson, R. S., and Biemann, K. (1987). The primary structure of thioredoxin from Chromatium vinosum

determined by high-performance tandem mass spectrometry. Biochemistry 26, 1209-1214.

Jung, E., Heller, M., Sanchez, J. C., and Hochstrasser, D. F. (2000). Proteomics meets cell biology: the

establishment of subcellular proteomes. Electrophoresis 21, 3369-3377.

Jungblut, P., and Klose, J. (1985). Genetic variability of proteins from mitochondria and mitochondrial fractions

of mouse organs. Biochem Genet 23, 227-245.

119

Jungblut, P. R., and Seifert, R. (1990). Analysis by high-resolution two-dimensional electrophoresis of

differentiation-dependent alterations in cytosolic protein pattern of HL-60 leukemic cells. J Biochem

Biophys Methods 21, 47-58.

Kaindl, A. (2001) Genetic analysis of mouse proteome in two-dimensional protein patterns of heart and liver,

Dissertation, Humboldt Universität zu Berlin, Berlin.

Kaiser, N., and Edelman, I. S. (1977). Calcium dependence of glucocorticoid-induced lymphocytolysis. Proc

Natl Acad Sci U S A 74, 638-642.

Karas, M., and Hillenkamp, F. (1988). Laser desorption ionization of proteins with molecular masses exceeding

10,000 daltons. Anal Chem 60, 2299-2301.

Karty, J. A., Ireland, M. M., Brun, Y. V., and Reilly, J. P. (2002). Defining absolute confidence limits in the

identification of Caulobacter proteins by peptide mass mapping. Artifacts and unassigned masses

encountered in peptide mass mapping. J Proteome Res 1, 325-335.

Kashino, Y. (2003). Separation methods in the analysis of protein membrane complexes. J Chromatogr B

Analyt Technol Biomed Life Sci 797, 191-216.

Khrapko, K., Bodyak, N., Thilly, W. G., van Orsouw, N. J., Zhang, X., Coller, H. A., Perls, T. T., Upton, M.,

Vijg, J., and Wei, J. Y. (1999). Cell-by-cell scanning of whole mitochondrial genomes in aged human heart

reveals a significant fraction of myocytes with clonally expanded deletions. Nucleic Acids Res 27, 2434-

2441.

Kirkwood, T. B., and Holliday, R. (1979). The evolution of ageing and longevity. Proc R Soc Lond B Biol Sci

205, 531-546.

Klement, P., Nijtmans, L. G., Van den Bogert, C., and Houstek, J. (1995). Analysis of oxidative

phosphorylation complexes in cultured human fibroblasts and amniocytes by blue-native-electrophoresis

using mitoplasts isolated with the help of digitonin. Anal Biochem 231, 218-224.

Klose, J. (1975). Protein mapping by combined isoelectric focusing and electrophoresis of mouse tissues. A

novel approach to testing for induced point mutations in mammals. Humangenetik 26, 231-243.

Klose, J. (1999). Large-gel 2-D electrophoresis. Fractionated extraction of total tissue proteins from mouse and

human for 2-D electrophoresis. Methods Mol Biol 112, 147-172.

Klose, J., and Kobalz, U. (1995). Two-dimensional electrophoresis of proteins: an updated protocol and

implications for a functional analysis of the genome. Electrophoresis 16, 1034-1059.

Klose, J., Nock, C., Herrmann, M., Stuhler, K., Marcus, K., Bluggel, M., Krause, E., Schalkwyk, L. C., Rastan,

S., Brown, S. D., et al. (2002). Genetic analysis of the mouse brain proteome. Nat Genet 30, 385-393.

Kopsidas, G., Kovalenko, S. A., Heffernan, D. R., Yarovaya, N., Kramarova, L., Stojanovski, D., Borg, J.,

Islam, M. M., Caragounis, A., and Linnane, A. W. (2000). Tissue mitochondrial DNA changes. A stochastic

system. Ann N Y Acad Sci 908, 226-243.

Kopsidas, G., Kovalenko, S. A., Kelso, J. M., and Linnane, A. W. (1998). An age-associated correlation

between cellular bioenergy decline and mtDNA rearrangements in human skeletal muscle. Mutat Res 421,

27-36.

Kovalenko, S. A., Kopsidas, G., Kelso, J., Rosenfeldt, F., and Linnane, A. W. (1998). Tissue-specific

distribution of multiple mitochondrial DNA rearrangements during human aging. Ann N Y Acad Sci 854,

171-181.

Kowald, A., and Kirkwood, T. B. (1999). The mitochondrial theory of aging: do damaged mitochondria

accumulate by delayed degradation? Exp Gerontol 34, 605-612.

Kowald, A., and Kirkwood, T. B. (2000). Accumulation of defective mitochondria through delayed degradation

of damaged organelles and its possible role in the ageing of post-mitotic and dividing cells. J Theor Biol

202, 145-160.

Kraytsberg, Y., Nekhaeva, E., Bodyak, N. B., and Khrapko, K. (2003). Mutation and intracellular clonal

expansion of mitochondrial genomes: two synergistic components of the aging process? Mech Ageing Dev

124, 49-53.

Kuhn, K., Wellen, J., Link, N., Maskri, L., Lubbert, H., and Stichel, C. C. (2003). The mouse MPTP model:

gene expression changes in dopaminergic neurons. Eur J Neurosci 17, 1-12.

120

Kyte, J., and Doolittle, R. F. (1982). A simple method for displaying the hydropathic character of a protein. J

Mol Biol 157, 105-132.

Laemmli, U. K., Amos, L. A., and Klug, A. (1976). Correlation between structural transformation and cleavage

of the major head protein of T4 bacteriophage. Cell 7, 191-203.

Lai, Y. K., Lee, W. C., Hu, C. H., and Hammond, G. L. (1996). The mitochondria are recognition organelles of

cell stress. J Surg Res 62, 90-94.

Le Bourg, E. (1998). Evolutionary theories of aging: handle with care. Gerontology 44, 345-348.

Lee, H. C., Pang, C. Y., Hsu, H. S., and Wei, Y. H. (1994). Differential accumulations of 4,977 bp deletion in

mitochondrial DNA of various tissues in human ageing. Biochim Biophys Acta 1226, 37-43.

Levonen, A. L., Patel, R. P., Brookes, P., Go, Y. M., Jo, H., Parthasarathy, S., Anderson, P. G., and Darley-

Usmar, V. M. (2001). Mechanisms of cell signaling by nitric oxide and peroxynitrite: from mitochondria to

MAP kinases. Antioxid Redox Signal 3, 215-229.

Lindop, P. J., and Rotblat, J. (1961a). Long-term effects of a single whole-body exposure of mice to ionizing

radiations. I.Life shortening. ProcRoySoc 154, 332-349.

Lindop, P. J., and Rotblat, J. (1961b). Shortening of life and causes of death in mice exposed to a single whole-

body dose of radiation. Naturwissenschaften 189, 645-648.

Lindsey, J., McGill, N. I., Lindsey, L. A., Green, D. K., and Cooke, H. J. (1991). In vivo loss of telomeric

repeats with age in humans. Mutat Res 256, 45-48.

Linnane, A. W., Baumer, A., Maxwell, R. J., Preston, H., Zhang, C. F., and Marzuki, S. (1990). Mitochondrial

gene mutation: the ageing process and degenerative diseases. Biochem Int 22, 1067-1076.

Linnane, A. W., Marzuki, S., Ozawa, T., and Tanaka, M. (1989). Mitochondrial DNA mutations as an important

contributor to ageing and degenerative diseases. Lancet 1, 642-645.

Linnane, A. W., Zhang, C., Baumer, A., and Nagley, P. (1992). Mitochondrial DNA mutation and the ageing

process: bioenergy and pharmacological intervention. Mutat Res 275, 195-208.

Liu, V. W., Zhang, C., Pang, C. Y., Lee, H. C., Lu, C. Y., Wei, Y. H., and Nagley, P. (1998). Independent

occurrence of somatic mutations in mitochondrial DNA of human skin from subjects of various ages. Hum

Mutat 11, 191-196.

Liu, X., Kim, C. N., Yang, J., Jemmerson, R., and Wang, X. (1996). Induction of apoptotic program in cell-free

extracts: requirement for dATP and cytochrome c. Cell 86, 147-157.

Lopez, M. F., Kristal, B. S., Chernokalskaya, E., Lazarev, A., Shestopalov, A. I., Bogdanova, A., and Robinson,

M. (2000). High-throughput profiling of the mitochondrial proteome using affinity fractionation and

automation. Electrophoresis 21, 3427-3440.

Lopez, M. F., and Melov, S. (2002). Applied proteomics: mitochondrial proteins and effect on function. Circ

Res 90, 380-389.

Lottspeich, F., and Zorbas, H. (1998). Bioanalytik (Heidelberg. Berlin, Spektrum Akademischer Verlag).

Ludwig, J., Kerscher, S., Brandt, U., Pfeiffer, K., Getlawi, F., Apps, D. K., and Schagger, H. (1998).

Identification and characterization of a novel 9.2-kDa membrane sector-associated protein of vacuolar

proton-ATPase from chromaffin granules. J Biol Chem 273, 10939-10947.

Mann, M., Hendrickson, R. C., and Pandey, A. (2001). Analysis of proteins and proteomes by mass

spectrometry. Annu Rev Biochem 70, 437-473.

Mann, M., and Wilm, M. (1994). Error-tolerant identification of peptides in sequence databases by peptide

sequence tags. Anal Chem 66, 4390-4399.

Manzelmann, M. S., and Harmon, H. J. (1987). Lack of age-dependent changes in rat heart mitochondria. Mech

Ageing Dev 39, 281-288.

Menzies, R. A., and Gold, P. H. (1971). The turnover of mitochondria in a variety of tissues of young adult and

aged rats. J Biol Chem 246, 2425-2429.

Miklos, G. L., and Maleszka, R. (2001). Integrating molecular medicine with functional proteomics: realities

and expectations. Proteomics 1, 30-41.

121

Mita, S., Schmidt, B., Schon, E. A., DiMauro, S., and Bonilla, E. (1989). Detection of "deleted" mitochondrial

genomes in cytochrome-c oxidase-deficient muscle fibers of a patient with Kearns-Sayre syndrome. Proc

Natl Acad Sci U S A 86, 9509-9513.

Molloy, M. P. (2000). Two-dimensional electrophoresis of membrane proteins using immobilized pH gradients.

Anal Biochem 280, 1-10.

Molloy, M. P., Herbert, B. R., Walsh, B. J., Tyler, M. I., Traini, M., Sanchez, J. C., Hochstrasser, D. F.,

Williams, K. L., and Gooley, A. A. (1998). Extraction of membrane proteins by differential solubilization

for separation using two-dimensional gel electrophoresis. Electrophoresis 19, 837-844.

Molloy, M. P., Herbert, B. R., Williams, K. L., and Gooley, A. A. (1999). Extraction of Escherichia coli

proteins with organic solvents prior to two-dimensional electrophoresis. Electrophoresis 20, 701-704.

Molloy, M. P., Phadke, N. D., Maddock, J. R., and Andrews, P. C. (2001). Two-dimensional electrophoresis

and peptide mass fingerprinting of bacterial outer membrane proteins. Electrophoresis 22, 1686-1696.

Morin, D., Hauet, T., Spedding, M., and Tillement, J. (2001). Mitochondria as target for antiischemic drugs.

Adv Drug Deliv Rev 49, 151-174.

Motsenbocker, M. A. (1988). Sensitivity limitations encountered in enhanced horseradish peroxidase catalysed

chemiluminescence. J Biolumin Chemilumin 2, 9-16.

Munscher, C., Rieger, T., Muller-Hocker, J., and Kadenbach, B. (1993). The point mutation of mitochondrial

DNA characteristic for MERRF disease is found also in healthy people of different ages. FEBS Lett 317, 27-

30.

Murayama, K., Fujimura, T., Morita, M., and Shindo, N. (2001). One-step subcellular fractionation of rat liver

tissue using a Nycodenz density gradient prepared by freezing-thawing and two-dimensional sodium

dodecyl sulfate electrophoresis profiles of the main fraction of organelles. Electrophoresis 22, 2872-2880.

Nagley, P., and Wei, Y. H. (1998). Ageing and mammalian mitochondrial genetics. Trends Genet 14, 513-517.

Nekhaeva, E., Bodyak, N. D., Kraytsberg, Y., McGrath, S. B., Van Orsouw, N. J., Pluzhnikov, A., Wei, J. Y.,

Vijg, J., and Khrapko, K. (2002). Clonally expanded mtDNA point mutations are abundant in individual

cells of human tissues. Proc Natl Acad Sci U S A 99, 5521-5526.

Neubauer, G., King, A., Rappsilber, J., Calvio, C., Watson, M., Ajuh, P., Sleeman, J., Lamond, A., and Mann,

M. (1998). Mass spectrometry and EST-database searching allows characterization of the multi-protein

spliceosome complex. Nat Genet 20, 46-50.

Neubert, D., Bass, R., and Helge, H. (1966). [Conversion rate of DNA in the mitochondria of warm-blooded

animals]. Naturwissenschaften 53, 23-24.

O'Farrell, P. H. (1975). High resolution two-dimensional electrophoresis of proteins. J Biol Chem 250, 4007-

4021.

Osiewacz, H. D., and Hamann, A. (1997). DNA reorganization and biological aging. A review. Biochemistry

(Mosc) 62, 1275-1284.

Panfili, E., Sottocasa, G. L., Sandri, G., and Liut, G. (1980). The Ca2+-binding glycoprotein as the site of

metabolic regulation of mitochondrial Ca2+ movements. Eur J Biochem 105, 205-210.

Pang, C. Y., Lee, H. C., Yang, J. H., and Wei, Y. H. (1994). Human skin mitochondrial DNA deletions

associated with light exposure. Arch Biochem Biophys 312, 534-538.

Pappin, D. J. C., Hojrup, P., and Bleasby, A. J. (1993). Rapid identification of proteins by peptide-mass

fingerprinting. Curr Biol 3, 327-332.

Pascale, R. M., De Miglio, M. R., Muroni, M. R., Simile, M. M., Daino, L., Seddaiu, M. A., Pusceddu, S.,

Gaspa, L., Calvisi, D., Manenti, G., and Feo, F. (1998). Transferrin and transferrin receptor gene expression

and iron uptake in hepatocellular carcinoma in the rat. Hepatology 27, 452-461.

Perrin, R. J., Woods, W. S., Clayton, D. F., and George, J. M. (2001). Exposure to long chain polyunsaturated

fatty acids triggers rapid multimerization of synucleins. J Biol Chem 276, 41958-41962.

Pirt, S. J. (1991). Genome project. Nature 350, 104.

Rabilloud, T. (1990). Mechanisms of protein silver staining in polyacrylamide gels: a 10-year synthesis.

Electrophoresis 11, 785-794.

122

Rabilloud, T., Brodard, V., Peltre, G., Righetti, P. G., and Ettori, C. (1992). A comparison between low

background silver diammine and silver nitrate protein stains. Modified silver staining for immobilized pH

gradients. Electrophoresis 13, 429-439.

Rabilloud, T., Heller, M., Gasnier, F., Luche, S., Rey, C., Aebersold, R., Benahmed, M., Louisot, P., and

Lunardi, J. (2002). Proteomics analysis of cellular response to oxidative stress. Evidence for in vivo

overoxidation of peroxiredoxins at their active site. J Biol Chem 277, 19396-19401.

Rabilloud, T., Heller, M., Gasnier, F., Luche, S., Rey, C., Aebersold, R., Benahmed, M., Louisot, P., and

Lunardi, J. (2003). Membrane proteins ride shotgun. Nat Biotechnol 21, 508-510.

Rabilloud, T., Kieffer, S., Procaccio, V., Louwagie, M., Courchesne, P. L., Patterson, S. D., Martinez, P., Garin,

J., and Lunardi, J. (1998). Two-dimensional electrophoresis of human placental mitochondria and protein

identification by mass spectrometry: toward a human mitochondrial proteome. Electrophoresis 19, 1006-

1014.

Ramsby, M. L., Makowski, G. S., and Khairallah, E. A. (1994). Differential detergent fractionation of isolated

hepatocytes: biochemical, immunochemical and two-dimensional gel electrophoresis characterization of

cytoskeletal and noncytoskeletal compartments. Electrophoresis 15, 265-277.

Remacle, C., Gloire, G., Cardol, P., and Matagne, R. F. (2004). Impact of a mutation in the mitochondrial LSU

rRNA gene from Chlamydomonas reinhardtii on the activity and the assembly of respiratory-chain

complexes. Curr Genet 45, 323-330.

Richter, C. (1988). Do mitochondrial DNA fragments promote cancer and aging? FEBS Lett 241, 1-5.

Richter, C., Park, J. W., and Ames, B. N. (1988). Normal oxidative damage to mitochondrial and nuclear DNA

is extensive. Proc Natl Acad Sci U S A 85, 6465-6467.

Robinson, A. B., and Pauling, L. (1974). Evolution and the distribution of glutaminyl and asparaginyl residues

in proteins. Clin Chem 20, 961-965.

Rotilio, G., Bray, R. C., and Fielden, E. M. (1972). A pulse radiolysis study of superoxide dismutase. Biochim

Biophys Acta 268, 605-609.

Rowlatt, C., Chesterman, F. C., and Sheriff, M. U. (1976). Lifespan, age changes and tumour incidence in an

ageing C57BL mouse colony. Lab Anim 10, 419-442.

Rustin, P., von Kleist-Retzow, J. C., Vajo, Z., Rotig, A., and Munnich, A. (2000). For debate: defective

mitochondria, free radicals, cell death, aging-reality or myth-ochondria? Mech Ageing Dev 114, 201-206.

Santoni, V., Doumas, P., Rouquie, D., Mansion, M., Rabilloud, T., Rossignol, M., Kieffer, S., and Garin, J.

(1999). Large scale characterization of plant plasma membrane proteins. Towards the recovery of

hydrophobic proteins on two-dimensional electrophoresis gels. Biochimie 81, 655-661.

Santoni, V., Molloy, M., and Rabilloud, T. (2000). Membrane proteins and proteomics: un amour impossible?

Electrophoresis 21, 1054-1070.

Schagger, H., and von Jagow, G. (1991). Blue native electrophoresis for isolation of membrane protein

complexes in enzymatically active form. Anal Biochem 199, 223-231.

Seglen, P. O. (1976). Preparation of isolated rat liver cells. Methods Cell Biol 13, 29-83.

Seligman, A. M., Karnovsky, M. J., Wasserkrug, H. L., and Hanker, J. S. (1968). Nondroplet ultrastructural

demonstration of cytochrome oxidase activity with a polymerizing osmiophilic reagent, diaminobenzidine

(DAB). J Cell Biol 38, 1-14.

Shenkar, R., Navidi, W., Tavare, S., Dang, M. H., Chomyn, A., Attardi, G., Cortopassi, G., and Arnheim, N.

(1996). The mutation rate of the human mtDNA deletion mtDNA4977. Am J Hum Genet 59, 772-780.

Shevchenko, A., Wilm, M., Mann, M., Jensen, O. N., Podtelejnikov, A. V., Sagliocco, F., Vorm, O., Mortensen,

P., and Boucherie, H. (1997). Mass spectrometric sequencing of proteins silver-stained polyacrylamide gels.

J Protein Chem 16, 481-490.

Shibutani, S., Takeshita, M., and Grollman, A. P. (1991). Insertion of specific bases during DNA synthesis past

the oxidation-damaged base 8-oxodG. Nature 349, 431-434.

Shoubridge, E. A., Karpati, G., and Hastings, K. E. (1990). Deletion mutants are functionally dominant over

wild-type mitochondrial genomes in skeletal muscle fiber segments in mitochondrial disease. Cell 62, 43-49.

123

Sickmann, A., Reinders, J., Wagner, Y., Joppich, C., Zahedi, R., Meyer, H. E., Schonfisch, B., Perschil, I.,

Chacinska, A., Guiard, B., et al. (2003). The proteome of Saccharomyces cerevisiae mitochondria. Proc Natl

Acad Sci U S A 100, 13207-13212.

Sims, N. R. (1990). Rapid isolation of metabolically active mitochondria from rat brain and subregions using

Percoll density gradient centrifugation. J Neurochem 55, 698-707.

Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano, M. D., Fujimoto, E. K.,

Goeke, N. M., Olson, B. J., and Klenk, D. C. (1985). Measurement of protein using bicinchoninic acid. Anal

Biochem 150, 76-85.

Song, D. D., Shults, C. W., Sisk, A., Rockenstein, E., and Masliah, E. (2004). Enhanced substantia nigra

mitochondrial pathology in human alpha-synuclein transgenic mice after treatment with MPTP. Exp Neurol

186, 158-172.

Soong, N. W., Hinton, D. R., Cortopassi, G., and Arnheim, N. (1992). Mosaicism for a specific somatic

mitochondrial DNA mutation in adult human brain. Nat Genet 2, 318-323.

Stephenson, G., Marzuki, S., and Linnane, A. W. (1980). Biogenesis of mitochondria. Two-dimensional

electrophoretic analysis of mitochondrial translation products in yeast. Biochim Biophys Acta 609, 329-341.

Storm, T., Rath, S., Mohamed, S. A., Bruse, P., Kowald, A., Oehmichen, M., and Meissner, C. (2002). Mitotic

brain cells are just as prone to mitochondrial deletions as neurons: a large-scale single-cell PCR study of the

human caudate nucleus. Exp Gerontol 37, 1389-1400.

Stryer, L. (1995). Biochemistry, 4th edn (New York, W.H. Freeman and Company).

Sugiyama, S., Takasawa, M., Hayakawa, M., and Ozawa, T. (1993). Changes in skeletal muscle, heart and liver

mitochondrial electron transport activities in rats and dogs of various ages. Biochem Mol Biol Int 30, 937-

944.

Swain, M., and Ross, N. W. (1995). A silver stain protocol for proteins yielding high resolution and transparent

background in sodium dodecyl sulfate-polyacrylamide gels. Electrophoresis 16, 948-951.

Switzer, R. C., 3rd, Merril, C. R., Shifrin, S., Switzer, R. C., and Van Keuren, M. L. (1979). A highly sensitive

silver stain for detecting proteins and peptides in polyacrylamide gels. Anal Biochem 98, 231-237.

Takahashi, H., and Yamaguchi, M. (2000). Stimulatory effect of regucalcin on ATP-dependent Ca(2+) uptake

activity in rat liver mitochondria. J Cell Biochem 78, 121-130.

Taylor, R. W., Wardell, T. M., Connolly, B. A., Turnbull, D. M., and Lightowlers, R. N. (2001). Linked

oligodeoxynucleotides show binding cooperativity and can selectively impair replication of deleted

mitochondrial DNA templates. Nucleic Acids Res 29, 3404-3412.

Taylor, S. W., Warnock, D. E., Glenn, G. M., Zhang, B., Fahy, E., Gaucher, S. P., Capaldi, R. A., Gibson, B.

W., and Ghosh, S. S. (2002). An alternative strategy to determine the mitochondrial proteome using sucrose

gradient fractionation and 1D PAGE on highly purified human heart mitochondria. J Proteome Res 1, 451-

458.

Tengan, C. H., Gabbai, A. A., Shanske, S., Zeviani, M., and Moraes, C. T. (1997). Oxidative phosphorylation

dysfunction does not increase the rate of accumulation of age-related mtDNA deletions in skeletal muscle.

Mutat Res 379, 1-11.

Timmons, T. M., and Dunbar, B. S. (1990). Protein blotting and immunodetection. Methods Enzymol 182, 679-

688.

Towbin, H., and Gordon, J. (1984). Immunoblotting and dot immunobinding--current status and outlook. J

Immunol Methods 72, 313-340.

Trifunovic, A., Wredenberg, A., Falkenberg, M., Spelbrink, J. N., Rovio, A. T., Bruder, C. E., Bohlooly, Y. M.,

Gidlof, S., Oldfors, A., Wibom, R., et al. (2004). Premature ageing in mice expressing defective

mitochondrial DNA polymerase. Nature 429, 417-423.

Trounce, I., Byrne, E., and Marzuki, S. (1989). Decline in skeletal muscle mitochondrial respiratory chain

function: possible factor in ageing. Lancet 1, 637-639.

Ueda, K., Fukushima, H., Masliah, E., Xia, Y., Iwai, A., Yoshimoto, M., Otero, D. A., Kondo, J., Ihara, Y., and

Saitoh, T. (1993). Molecular cloning of cDNA encoding an unrecognized component of amyloid in

Alzheimer disease. Proc Natl Acad Sci U S A 90, 11282-11286.

124

van Os, C. H. (1987). Transcellular calcium transport in intestinal and renal epithelial cells. Biochim Biophys

Acta 906, 195-222.

Villa, A., Garcia-Simon, M. I., Blanco, P., Sese, B., Bogonez, E., and Satrustegui, J. (1998). Affinity

chromatography purification of mitochondrial inner membrane proteins with calcium transport activity.

Biochim Biophys Acta 1373, 347-359.

Vu, T. H., Tanji, K., Pallotti, F., Golzi, V., Hirano, M., DiMauro, S., and Bonilla, E. (2000). Analysis of

mtDNA deletions in muscle by in situ hybridization. Muscle Nerve 23, 80-85.

Wade, O. L., and Bishop, J. M. (1962). Cardiac output and Regional Blood Flow (Philadelphia, Oxford

Blackwell Scientific).

Wallace, D. C. (1992). Mitochondrial genetics: a paradigm for aging and degenerative diseases? Science 256,

628-632.

Wallace, D. C. (1997). Mitochondrial DNA in aging and disease. Sci Am 277, 40-47.

Wallace, D. C. (2001). Mitochondrial genes in degenerative disease and aging. ScientificWorldJournal 1, 83.

Wasinger, V. C., Cordwell, S. J., Cerpa-Poljak, A., Yan, J. X., Gooley, A. A., Wilkins, M. R., Duncan, M. W.,

Harris, R., Williams, K. L., and Humphery-Smith, I. (1995). Progress with gene-product mapping of the

Mollicutes: Mycoplasma genitalium. Electrophoresis 16, 1090-1094.

Wei, Y. H. (1998). Mitochondrial DNA mutations and oxidative damage in aging and diseases: an emerging

paradigm of gerontology and medicine. Proc Natl Sci Counc Repub China B 22, 55-67.

Weiss, W., Postel, W., and Gorg, A. (1992). Application of sequential extraction procedures and glycoprotein

blotting for the characterization of the 2-D polypeptide patterns of barley seed proteins. Electrophoresis 13,

770-773.

Wessel, D., and Flugge, U. I. (1984). A method for the quantitative recovery of protein in dilute solution in the

presence of detergents and lipids. Anal Biochem 138, 141-143.

Westermann, B., and Neupert, W. (2003). 'Omics' of the mitochondrion. Nat Biotechnol 21, 239-240.

Whitehead, T. P., Kricka, L. J., Carter, T. J., and Thorpe, G. H. (1979). Analytical luminescence: its potential in

the clinical laboratory. Clin Chem 25, 1531-1546.

Wilkins, M. R., Pasquali, C., Appel, R. D., Ou, K., Golaz, O., Sanchez, J. C., Yan, J. X., Gooley, A. A., Hughes,

G., Humphery-Smith, I., et al. (1996). From proteins to proteomes: large scale protein identification by two-

dimensional electrophoresis and amino acid analysis. Biotechnology (N Y) 14, 61-65.

Wilm, M., and Mann, M. (1996). Analytical properties of the nanoelectrospray ion source. Anal Chem 68, 1-8.

Xie, J. (2003) Establishment of a two-dimensional electrophoresis map of human mitochondrial proteins,

Dissertation, Humboldt-Universität zu Berlin, Berlin.

Xue, J. H., Takahashi, H., and Yamaguchi, M. (2000). Stimulatory effect of regucalcin on mitochondrial ATP-

dependent calcium uptake activity in rat kidney cortex. J Cell Biochem 80, 285-292.

Yamaguchi, M., and Isogai, M. (1993). Tissue concentration of calcium-binding protein regucalcin in rats by

enzyme-linked immunoadsorbent assay. Mol Cell Biochem 122, 65-68.

Yen, T. C., King, K. L., Lee, H. C., Yeh, S. H., and Wei, Y. H. (1994). Age-dependent increase of

mitochondrial DNA deletions together with lipid peroxides and superoxide dismutase in human liver

mitochondria. Free Radic Biol Med 16, 207-214.

Yen, T. C., Su, J. H., King, K. L., and Wei, Y. H. (1991). Ageing-associated 5 kb deletion in human liver

mitochondrial DNA. Biochem Biophys Res Commun 178, 124-131.

Yerushalmi, H., Lebendiker, M., and Schuldiner, S. (1995). EmrE, an Escherichia coli 12-kDa multidrug

transporter, exchanges toxic cations and H+ and is soluble in organic solvents. J Biol Chem 270, 6856-6863.

Zabel, C., Chamrad, D. C., Priller, J., Woodman, B., Meyer, H. E., Bates, G. P., and Klose, J. (2002).

Alterations in the mouse and human proteome caused by Huntington's disease. Mol Cell Proteomics 1, 366-

375.

Zeviani, M., Moraes, C. T., DiMauro, S., Nakase, H., Bonilla, E., Schon, E. A., Rowland, L. P., Tiranti, V., and

Piantadosi, C. (1998). Deletions of mitochondrial DNA in Kearns-Sayre syndrome. Neurology 51, 1525-

1533.

125

Zhang, C., Bills, M., Quigley, A., Maxwell, R. J., Linnane, A. W., and Nagley, P. (1997a). Varied prevalence of

age-associated mitochondrial DNA deletions in different species and tissues: a comparison between human

and rat. Biochem Biophys Res Commun 230, 630-635.

Zhang, C., Liu, V. W., Addessi, C. L., Sheffield, D. A., Linnane, A. W., and Nagley, P. (1998). Differential

occurrence of mutations in mitochondrial DNA of human skeletal muscle during aging. Hum Mutat 11, 360-

371.

Zhang, C., Liu, V. W., and Nagley, P. (1997b). Gross mosaic pattern of mitochondrial DNA deletions in skeletal

muscle tissues of an individual adult human subject. Biochem Biophys Res Commun 233, 56-60.

Zhao, W., Devamanoharan, P. S., and Varma, S. D. (2000). Fructose-mediated damage to lens alpha-crystallin:

prevention by pyruvate. Biochim Biophys Acta 1500, 161-168.

Zuo, X., Echan, L., Hembach, P., Tang, H. Y., Speicher, K. D., Santoli, D., and Speicher, D. W. (2001).

Towards global analysis of mammalian proteomes using sample prefractionation prior to narrow pH range

two-dimensional gels and using one-dimensional gels for insoluble and large proteins. Electrophoresis 22,

1603-1615.

126

Curriculum vitae

Mao Lei (Mrs.)

Seidenbau str. 6

D-12489 Berlin, Germany

Professional Experience

January 2002 – present:

Research assistant and doctoral student at the Institute of Human Genetics, University

hospital Campus Virchow-Klinikum Charité, worked on the dissertation project under

supervision of Prof. Dr. Dr. Klose.

January 1997 – December 2001:

Research assistant at the Department of Experimental Surgery, University hospital Campus

Virchow-Klinikum Charité, worked on the projects of “Development of hepatocyte

bioreactors as an alternative to animal experiments” and “Improvement of liver preservation

solutions for liver transplantation”.

August 1995 – December 1995:

Visiting scholar and interpreter at the Marmara Research Center, Gebze-Istanbul, Turkey.

Worked on a Sino-Turkish cooperation research project entitled “Trans-gene laboratory

animals by human gene microinjection”.

August 1991 – August 1995:

Research fellow of Biotechnology Research Center, Chinese Academy of Agricultural

Sciences, Beijing, China, focusing on the national project “Scale-up of monoclonal antibody

vaccine production for agricultural and veterinary usages”. Project manager in establishment

and management of a complete set of specific pathogen free (SPF) standard laboratory

animal facilities.

Education

October 2002 – July 2004:

Graduate study at University of Applied Sciences Berlin, Master of Computer Sciences in

Bioinformatics pursued. Graduate thesis titled “Mathematical simulation of the

accumulation of defective mitochondria during the aging process”, supervised by Dr.

Kowald and Prof. I. Koch.

October 1996 – December 2001:

Graduate study at the Department of biotechnology, Technical University Berlin. Dipom

Engineer of Biotechnology pursued. Graduate thesis titled “Simultaneous isolation of

hepatocytes and nonparenchymal cells from adult rat and human liver”, supervised by Uni.

Prof. Dipl. -Ing. Dr. U. Stahl and Dr. K. Zeilinger.

September 1987 – July 1991:

Undergraduate study at Beijing Agricultural University. Bachelor of Science in Agronomy

(majoring laboratory animal Sciences) awarded in July 1991. Undergraduate thesis titled

“The establishment of a diabetes animal using mini-pig”.

August 1991 – December 1991:

127

National certification courses in technology of germ-free laboratory animal husbandry and

National workshop on Specific-pathogen-free (SPF) animal facility management. Provided

by the Chinese Academy of Medical Sciences.

Publication:

1. Zeilinger K, Sauer IM, Pless G, Strobel C, Rudzitis J, Wang A, Nussler AK, Grebe A,

Mao L, Auth SH, Unger J, Neuhaus P, Gerlach JC.Three-dimensional co-culture of primary

human liver cells in bioreactors for in vitro drug studies: effects of the initial cell quality on

the long-term maintenance of hepatocyte-specific functions. Altern Lab Anim. 2002 Sep-

Oct; 30(5):525-38.

2. Zeilinger K, Auth SHG, Unger J, Grebe A, Mao L. Standardisierung eines Hepatozyten-

bioreaktorsystemes für in vitro-Metabolismusstudien als Alternative zum Tierversuch. In:

Schöffl H, Spielmann H, Tritthart HA, eds. Ersatz- und Ergänzungsmethoden zu

Tierversuchen. Springer-Verlag. 2000: 1-7.

3. Zeilinger K, Auth SHG, Unger J, Grebe A, Mao L, Petrik M, Holland G, Appel K,

Nüssler AK, Neuhaus P, Gerlach JC. Leberzellkultur in Bioreaktoren für in vitro Studien

zum Arzneimittelmetabolismus als Alternative zum Tierversuch. ALTEX 2000; 17:3-10.

4. Mao L, Zeilinger K, Roth S, Gerlach JC, Neuhaus P. Gleichzeitige Isolierung von

Hepatozyten, Itozellen und Simusendothelzellen der Leber aus dem selben Organ. Z.

Gastroenterol. 2000; 39:94.

5. Mao L, Zeilinger K, Auth S, Grebe A, Petrik M, Appel D, Schonoy N, Holland G,

Jennings G, Gerlach JC. Immortalisierte Hepatozyten: Zukunftsmodell zum Ersatz von

Metabolismusstudien am Tier? In: Schöffl H, Spielmann H, Tritthart HA, eds. Ersatz- und

Ergänzungsmethoden zu Tierversuchen. Springer-Verlag. 2000: 450.

6. Grebe A, Zeilinger K, Auth SHG, Mao L, Schonoy N, Holland G, Appel K, Jennings G,

Gerlach JC. Kultur primärer und immortalisierter humaner Hepatozyten in einem

Bioreaktor-Perfusionssystem. Z. Gastroenterol. 1999; 37: 71.

128

Acknowledgments

The work with this dissertation has been extensive and trying, but in the first place exciting,

instructive, and fun. Without help, support, and encouragement from some persons, I would never

have been able to finish this work.

First of all, I would like to express my deep gratitude to the unselfish effort of my external examiner,

Prof. Dipl.-Ing. Dr. Ulf Stahl and his untiring guidance and help throughout the years.

I would like to thank my supervisor Prof. Dr. Dr. Joachim Klose for making possible the conduct of

this research, for his inspiring and encouraging way to guide me to a deeper understanding of

knowledge, and for his invaluable comments during the whole work. His constructive comments and

invaluable suggestions have significantly improved the contents of this dissertation.

I would also like to acknowledge Professor Dr. Karl Sperling, Director General of our institute for

coming up with such a distinguished program.

I will also give a special thanks to Dr. Katrin Zeilinger, whose voices I always recall, especially

during the writing and proofreading of this work. It was she who led me grasp the spirit of discover.

I am very grateful to Prof. Sebastian Bachmann, Mrs. Petra Schrade and Dr. Kerim Mutig, who have

worked with the histopathological techniques of the biopsies. Thank you all for dedicating their

valuable time for my curiosity.

Priv.-Doz. Dr. Grosse-Siestrup and his co-workers, especially Melanie are appreciated for a fruitful

and agreeable collaboration.

I am also very grateful to everyone who has proofread parts of the manuscript, especially Dr. Claus

Zabel and Dr. Jin Xie, for their helpful advice and criticism throughout the course of this work. Dr.

Patrick Giavalisco is consistently appreciated for the exciting and endless discussions concerning the

work.

My thanks also go to all my colleagues at the Institute of Human Genetics for providing a superior

working atmosphere, especially Mrs. Grit Nebrich, Mr. Maik Wacker and Mrs. Silke Becker in the

core facility for their assistance of Mass spectrometry measurements. Mrs. Marion Herrmann is

acknowledged for allowing the use of relevant data for this study. Mr. Ingo Seefeldt is appreciated

for his assistance of ProteinWeaver software. Mrs. Bettina Esch, Mrs. Yvonne Kläre and Mrs. Janine

Stuwe are appreciated for their assistance of the laboratory works. Dr. Dijana Sagi is given thanks

for sharing her valuable experimental data relating to this study and fruitful discussion.

Support of mathematical modeling has been provided by Dr. Axel Kowald from Max Plank institute

for Molecular Genetics, and from Mr. Maciej Swat from the Institute of Theoretical Biology. I

appreciate their invaluable help and suggestions.

At last, I wish to express my heartiest gratitude and respect to a special person who is always there

for me: Mr. Rudi Beichel, to whom this work is dedicated.

129

Erklärung

Hiermit erkläre ich, Lei Mao, geboren am 29.04.1970 in Shanghai (China), an Eides Statt,

dass meine Dissertation mit dem Thema “The Mitochondrial Protein Profile Change during

the Aging Process” von mir selbst und ohne Hilfe Dritter verfasst wurde, auch in Teilen

keine Kopie andere Arbeiten darstellt und die benutzen Hilfsmittel sowie die Literatur

vollständig angegeben sind. Ich habe mich anderwärtig nicht um einen Doktorgrad

beworben und besitze einen entsprechenden Doktorgrad nicht. Ich erkläre die

Kenntnisnahme der dem Verfahren zugrunde liegenden Promotionsordnung der

Technischen Universität zu Berlin.

Berlin, den ____________________

_____________________________

Unterschrift

130

Appendix:

Tab.A1: Protein spots on the large-gel 2D-gels that have been identified in our study:

Spot Hit* gi-Nr.

Probability

based

Mowse

Score

Matched

Peptide

Sequence

coverage

[%]

[ kDa ] pI Protein name

9 gi|18655687 99 15 90 16 7..27 Chain B, Chimeric MOUSE CARBONMONOXY HEMOGLOBIN

gi|553919 59 5 63 13

6.78 alpha-1-globin

gi|13654245 77 10 65 21 5.02

major urinary protein 1 [Mus musculus]

gi|6677739 76 12 39 33 5.15

regucalcin [Mus musculus]

gi|6754976 132

15 50 22

8.26 peroxiredoxin 1; proliferation-associated gene A; osteoblast specific factor 3

37e

gi|8393343 84 11 68 14

8.59 fatty acid binding protein 1, liver; fatty acid binding protein liver [Mus musculus]

1.Hit gi|12842467 64 7 29 14 8.34

unnamed protein product [Mus musculus]

41 2.Hit gi|6753500 62 7 29 14

8.69 cytochrome c oxidase, subunit Vb [Mus musculus]

sgi|127134 63 9 48 17 4.62

Myosin light chain 3, skeletal muscle isoform (A2 catalytic) (Alkali) (MLC3F)

gi|7949078 164 20 74 19

4.82 myosin light chain 2, phosphorylatable, fast skeletal muscle [Mus musculus]

126

gi|127134 70 10 55 17

4.62 Myosin light chain 3, skeletal muscle isoform (A2 catalytic) (Alkali) (MLC3F)

128b

gi|7949078 149 13 63 19 4.82

myosin light chain 2, phosphorylatable, fast skeletal muscle [Mus musculus]

173b

gi|4506741 156 12 43 22 10.09 ribosomal protein S7; 40S ribosomal protein S7 [Homo sapiens]

173e

gi|31560385 98 13 46 19 10.49 ribosomal protein L21 [Mus musculus]

175b

gi|18079339 202 26 36 85 8.08

aconitase 2, mitochondrial [Mus musculus]

300

gi|20818892 96 9 41 22 8.25

similar to FLJ23469 protein [Mus musculus]

301

gi|18250284 132 17 36 40 6.27

isocitrate dehydrogenase 3 (NAD+) alpha [Mus musculus]

302

gi|18152793 122 22 60 39 6.41

pyruvate dehydrogenase (lipoamide) beta [Mus musculus]

303

gi|18152793 197 23 48 39 6.41

pyruvate dehydrogenase (lipoamide) beta [Mus musculus]

304

gi|18250284 141 19 40 40 6.27

isocitrate dehydrogenase 3 (NAD+) alpha [Mus musculus]

305

gi|18250284 74 11 31 40 6.27

isocitrate dehydrogenase 3 (NAD+) alpha [Mus musculus]

1.Hit gi|12846314 150 15 48 22 8.26

unnamed protein product [Mus musculus]

407(18) 2.Hit gi|6754976 150

15 48 22

8.26 peroxiredoxin 1; proliferation-associated gene A; osteoblast specific factor 3

Spot Hit* gi-Nr.

Probability

based

Mowse

Score

Matched

Peptide

Sequence

coverage

[%]

[ kDa ] pI Protein name

406

gi|8393343 143 22 91 14 8.59

fatty acid binding protein 1, liver; fatty acid binding protein liver [Mus musculus]

405(17) gi|6677739 155 22 63 33 5.15

regucalcin [Mus musculus]

1.Hit gi|226471 80 10 57 16 6.03 Cu/Zn superoxide dismutase

404_1 2.Hit gi|20896095 80 10 57 16

6.02 superoxide dismutase 1, soluble [Mus musculus]

1.Hit gi|226471 127 12 64 16 6.03 Cu/Zn superoxide dismutase

404_2 2.Hit gi|20896095 126 12 63 16

6.02 superoxide dismutase 1, soluble [Mus musculus]

1.Hit gi|26324826 96 20 27 73

8.47 unnamed protein product [Mus musculus]

403 2.Hit gi|6429156 86 19 24 75

8.64 peroxisomal acyl-CoA oxidase [Mus musculus]

402

gi|21313138 154 19 65 26 8.97

glutathione S-transferase class kappa [Mus musculus]

401

gi|31981724 161 27 64 25 8.76

glutathione S-transferase, alpha 3 [Mus musculus]

79 gi|7949005 92 6 54 12 9.36 ATP synthase, H+ transporting, mitochondrial F0 complex, subunit F;

mitochondrial ATP synthase coup

142 gi|18655687 69 5 43 16 7.27 Chain B, Chimeric HumanMOUSE CARBONMONOXY HEMOGLOBIN

175b gi|18079339 164 21 25 85 8.08

aconitase 2, mitochondrial [Mus musculus]

152

gi|6679299 81 12 47 30 5.57

prohibitin [Mus musculus]

1.Hit gi|6679583 98 14 51 24 5.64 RAB11B, member RAS oncogene family [Mus musculus]

154b 2.Hit gi|7108528 95 14 53 24

6.23 small GTPase [Mus musculus]

1.Hit gi|18390323 133 15 64 24 5.85

RAB14, member RAS oncogene family [Mus musculus]

2.Hit gi|6679583 86 10 38 24

5.64 RAB11B, member RAS oncogene family [Mus musculus]

3.Hit gi|34147513 58 8 45 23

6.4 RAB7, member RAS oncogene family; Ras-associated protein RAB7

154d

4.Hit gi|6679599 48 7 41 24 7.53

RAB7, member RAS oncogene family [Mus musculus]

gi|229552 123 35 54 66

5.76 albumin

1.Hit gi|547679 87 12 55 23 6.12 Heat shock 27 kDa protein (HSP 27)

160 2.Hit gi|91319 77 11 42 23

5.86 stress protein, 25K - mouse

180

gi|6680836 77 12 25 48 4.33

calreticulin [Mus musculus]

117a

gi|13385268 65 5 35 15 4.96

cytochrome b-5 [Mus musculus]

Spot Hit* gi-Nr.

Probability

based

Mowse

Score

Matched

Peptide

Sequence

coverage

[%]

[ kDa ] pI Protein name

126 gi|127134 184

Myosin light chain 3, skeletal muscle isoform (A2 catalytic) (Alkali) (MLC3F)

12 59 17 4.62

137a

gi|13385260 81 9 71 15 8.95

thioesterase superfamily member 2 [Mus musculus]

137b

gi|13385726 126 12 72 14 9.1

ubiquinol-cytochrome c reductase binding protein [Mus musculus]

144 gi|38075371 67 9 50 18 10.32 similar to NADH-ubiquinone oxidoreductase 13 kDa-A subunit, mitochondrial

(Complex I-13KD-A) (CI-13KD-A) [Mus musculus]

1.Hit gi|6755963 183 18 48 31

8.62 voltage-dependent anion channel 1 [Mus musculus]

153a 2.Hit gi|10720404 180 18 46 32 8.55

Voltage-dependent anion-selective channel protein 1 (VDAC-1)

(mVDAC1)

(Outer mitochondrial membrane protein porin 1)

1.Hit gi|13435636 101 10 30 25

6.67 2400003B06Rik protein [Mus musculus]

153c 2.Hit gi|26368202 97 10 27 27

8.49 unnamed protein product [Mus musculus]

155 gi|6679583 151 10 38 24

5.64 RAB11B, member RAS oncogene family [Mus musculus]

164 gi|38082750 67 9 37 27 7.63

NADH dehydrogenase (ubiquinone) flavoprotein 2 [Mus musculus]

1.Hit gi|27679110 109 19 58 21 10.2 similar to 60S RIBOSOMAL PROTEIN L17 (L23)

(ASI) [Rattus norvegicus]

173c

2.Hit gi|22001904 109 19 58 21

10.2 60S ribosomal protein L17 (L23)

173d

gi|20899100 170 19 82 21 8.19

NADH dehydrogenase (ubiquinone) 1 beta subcomplex, 10 [Mus musculus]

1.Hit gi|2501472 107 13 30 60 8.87

UDP-glucuronosyltransferase 1-1 precursor, microsomal (UDPGT)

174 2.Hit gi|31324690 107 13 30 60

8.94 UDP glycosyltransferase 1 family polypeptide A1 [Mus musculus]

1.Hit gi|14714615 80 15 23 92

4.74 Tumor rejection antigen gp96 [Mus musculus]

179 2.Hit gi|729425 89 16 24 92

4.78 Endoplasmin precursor (94 kDa glucose-regulated protein) (GRP94)

12c

gi|7440317 176 16 53 16 10.14 ribosomal protein S14 - mouse

13b gi|13384608 160 12 74 12 6.28 6-pyruvoyl-tetrahydropterin synthase (TCF1)

14 gi|22164792 109 9 35 22 9.34

mitochondrial ribosomal protein L12 [Mus musculus]

37a

gi|1708292 100 10 56 18 8.52

HEAT-RESPONSIVE PROTEIN 12

37c gi|38075371 90 8 61 18 10.32 similar to NADH-ubiquinone oxidoreductase 13 kDa-A subunit,

mitochondrial precursor (Complex I-13KD-A) (CI-13KD-A) [Mus musculus]

iii

Spot Hit* gi-Nr.

Probability

based

Mowse

Score

Matched

Peptide

Sequence

coverage

[%]

[ kDa ] pI Protein name

1.Hit gi|12852348 83 7 51 12

9.24 unnamed protein product [Mus musculus]

51b 2.Hit gi|6678047 79 7 45 14

4.74 synuclein, alpha; alpha SYN; alpha-synuclein [Mus musculus]

113

gi|31980806 108 8 35 22 4.57

progesterone receptor membrane component [Mus musculus]

114 gi|31980806 89 9 35 22 4.57

progesterone receptor membrane component [Mus musculus]

164a gi|6680674 85 12 21 56 5.98 thymoma viral proto-oncogene 2; RAC-beta serine/threonine protein kinase;

protein kinase B, beta [Mus musculus]

1.Hit gi|4506741 176 22 61 22

10.09 ribosomal protein S7; 40S ribosomal protein S7 [Homo sapiens]

172b 2.Hit gi|38081187 80 14 58 25

10.12 similar to 40S ribosomal protein S7 (S8) [Mus musculus]

gi|18655687 150 12 84 16 7.27

Chain B, Chimeric HumanMOUSE CARBONMONOXY HEMOGLOBIN

1.Hit gi|12833511 127 7 58 15 7.96

unnamed protein product [Mus musculus]

10 2.Hit gi|553919 110 6 58 13 6.78

alpha-1-globin

16 gi|13654245 179 18 80 21 5.02

major urinary protein 1 [Mus musculus]

17 gi|6677739 278 24 74 33 5.15

regucalcin [Mus musculus]

1.Hit gi|193761 79 6 89 6 6.82

alpha-globin

27 2.Hit gi|553919 65 6 42 13 6.78

alpha-1-globin

1.Hit gi|12859535 131 12 78 13 6.31

unnamed protein product [Mus musculus]

33 2.Hit gi|13386100 130 12 78 13 7.82

NADH dehydrogenase (ubiquinone) 1 alpha subcomplex, 5 [Mus musculus]

37e gi|8393343 148 15 76 14 8.59

fatty acid binding protein 1, liver; fatty acid binding protein liver [Mus musculus]

37f

gi|18655687 235 15 84 16 7.27

Chain B, Chimeric HumanMOUSE CARBONMONOXY HEMOGLOBIN

81 gi|127134 96 11 43 17 4.62

Myosin light chain 3, skeletal muscle isoform (A2 catalytic) (Alkali) (MLC3F)

1.Hit gi|117097 103 8 58 12 5.01

Cytochrome c oxidase polypeptide VA

85 2.Hit gi|6680986 76 7 34 16 6.08

cytochrome c oxidase, subunit Va [Mus musculus]

gi|127134 93 11 50 17 4.62

Myosin light chain 3, skeletal muscle isoform (A2 catalytic) (Alkali) (MLC3F)

gi|7949078 70 7 41 19 4.82

myosin light chain 2, phosphorylatable, fast skeletal muscle [Mus musculus]

gi|7949078 205 22 80 19 4.83

myosin light chain 2, phosphorylatable, fast skeletal muscle [Mus musculus]

128b

gi|19880190 102 14 17 69 9.31

DNA topoisomerase I [Mus musculus]

Spot Hit* gi-Nr.

Probability

based

Mowse

Score

Matched

Peptide

Sequence

coverage

[%]

[ kDa ] pI Protein name

177b gi|18250284 99 10 21 40 6.27

isocitrate dehydrogenase 3 (NAD+) alpha [Mus musculus]

175a

gi|28279468 158 25 30 81 7.52

leucine zipper-EF-hand containing transmembrane protein 1 [Mus musculus]

172d

gi|25051141 204 14 63 21 10.73 RIKEN cDNA 2510019J09 [Mus musculus]

1.Hit gi|12842467 66 8 29 14 8.34

unnamed protein product [Mus musculus]

122 2.Hit gi|6753500 64 8 29 14 8.69

cytochrome c oxidase, subunit Vb [Mus musculus]

128a gi|7949078 108 12 74 19 4.82

myosin light chain 2, phosphorylatable, fast skeletal muscle [Mus musculus]

139 gi|13385322 104 9 59 16 8.35

NADH dehydrogenase (ubiquinone) 1 beta subcomplex, 7 [Mus musculus]

145 gi|6680309 105 10 81 11 7.93

heat shock protein 1 (chaperonin 10); heat shock 10 kDa protein 1 (chaperonin

10);

mitochondrial chaperonin 10 [Mus musculus]

157b gi|30794280 85 17 25 69 5.82

albumin

159 gi|20913657 150 14 83 19 5.52

RIKEN cDNA 0610009D10 [Mus musculus]

1.Hit gi|26333821 88 10 18 75 8.24

unnamed protein product [Mus musculus]

172a 2.Hit gi|6429156 75 9 15 75 8.64

peroxisomal acyl-CoA oxidase [Mus musculus]

1.Hit gi|12805431 107 15 43 35 5.63

Pdhb protein [Mus musculus]

176b 2.Hit gi|18152793 101 15 40 39 6.41

pyruvate dehydrogenase (lipoamide) beta [Mus musculus]

181 gi|31980648 186 26 52 56 5.19

ATP synthase, H+ transporting mitochondrial F1 complex,

alpha subunit [Mus musculus]

gi|6754976 197 19 60 22 8.26 peroxiredoxin 1; proliferation-associated gene A; osteoblast specific factor 3;

macrophage 23 Kd stress protein; macrophase stress protein 23 kd; Trx dependent

19 gi|21703976 316 36 68 60 6.44

cDNA sequence BC021917 [Mus musculus]

1.Hit gi|18655687 147 15 84 16 7.27

Chain B, Chimeric HumanMOUSE CARBONMONOXY HEMOGLOBIN

26 2.Hit gi|16741459 79 10 61 19 5.52

RIKEN cDNA 0610009D10 [Mus musculus]

29 gi|18655687 192 16 84 16 7.27

Chain B, Chimeric HumanMOUSE CARBONMONOXY HEMOGLOBIN

1.Hit gi|12859535 103 8 73 13 6.31

unnamed protein product [Mus musculus]

37b 2.Hit gi|13386100 103 8 73 13 7.82

NADH dehydrogenase (ubiquinone) 1 alpha subcomplex, 5 [Mus musculus]

46 gi|1841387 62 7 34 28 4.72

14-3-3 zeta [Mus musculus]

Spot Hit* gi-Nr.

Probability

based

Mowse

Score

Matched

Peptide

Sequence

coverage

[%]

[ kDa ] pI Protein name

47 gi|1346412 78 13 27 74 6.54

Lamin A

49 gi|37589957 92 11 36 36 6.16

Malate dehydrogenase, soluble [Mus musculus]

61 gi|6753500 62 8 36 14 8.69

cytochrome c oxidase, subunit Vb [Mus musculus]

62 gi|6680986 66 8 28 16 6.08

cytochrome c oxidase, subunit Va [Mus musculus]

63 gi|6680986 86 14(Me18) 62 16 6.08

cytochrome c oxidase, subunit Va [Mus musculus]

75 gi|2253399 79 8 56 13 7.85

vesicle associated membrane protein 2 [Mus musculus]

78 gi|7949005 91 9 55 12 9.36

ATP synthase, H+ transporting, mitochondrial F0 complex, subunit F;

mitochondrial ATP synthase coupling factor 6 [Mus musculus]

1.Hit gi|12847456 66 5 22 18 5.03

unnamed protein product [Mus musculus]

80 2.Hit gi|20806153 66 5 22 18 5.16

ATP synthase, H+ transporting, mitochondrial F1 complex, delta subunit

88 gi|6680309 173 14 91 11 7.93

heat shock protein 1 (chaperonin 10); heat shock 10 kDa protein 1 (chaperonin

10);

mitochondrial chaperonin 10 [Mus musculus]

98 gi|6679299 224 18 61 30 5.57

prohibitin [Mus musculus]

gi|20913657 129 15 90 19 5.52

RIKEN cDNA 0610009D10 [Mus musculus]

100 gi|21759114 127 16 61 27 8.57

Electron transfer flavoprotein beta-subunit (Beta-ETF)

102

gi|21759114 131 16 52 27 8.57

Electron transfer flavoprotein beta-subunit (Beta-ETF)

1.Hit gi|21313618 112 13 62 26 8.56

RIKEN cDNA 0610041L09 [Mus musculus]

103 2.Hit gi|12805413 90 14 38 31 8.76

Echs1 protein [Mus musculus]

1.Hit gi|12850643 96 9 43 27 9.1

unnamed protein product [Mus musculus]

104 2.Hit gi|13182962 96 9 43 27 8.89

short chain L-3-hydroxyacyl-CoA dehydrogenase

111 gi|31980806 104 10 36 22 4.57

progesterone receptor membrane component [Mus musculus]

112 gi|31980806 132 14 51 22 4.57

progesterone receptor membrane component [Mus musculus]

115d gi|31542438 88 10 41 16 4.79

cytochrome b5 outer mitochondrial membrane precursor [Mus musculus]

117b gi|13385268 87 9 56 15 4.96

cytochrome b-5 [Mus musculus]

117c gi|13385268 154 11 58 15 4.96

cytochrome b-5 [Mus musculus]

Spot Hit* gi-Nr.

Probability

based

Mowse

Score

Matched

Peptide

Sequence

coverage

[%]

[ kDa ] pI Protein name

gi|15679953 140 14 51 33 6.75

Glycine N-methyltransferase [Mus musculus]

35 gi|38075371 93 11 61 18 10.32 similar to NADH-ubiquinone oxidoreductase 13 kDa-A subunit, mitochondrial

precursor (Complex I-13KD-A)

(CI-13KD-A) [Mus musculus]

40a gi|6996913 63 9 21 50 7.55 annexin A2; calpactin I heavy chain; annexin II; lipocortin II; chromobindin 8

41 gi|29789345 79 15 30 50 5.73

2410153K17 protein [Mus musculus]

1.Hit gi|20913929 264 32 58 57 4.77

prolyl 4-hydroxylase, beta polypeptide [Mus musculus]

43 2.Hit gi|129729 264 32 58 57 4.79 Protein disulfide isomerase precursor (PDI) (Prolyl 4-hydroxylase beta subunit)

(Cellular thyroid hormone binding protein) (P55) (ERP59)

2.Hit gi|21746161 161 30 56 50 4.78

tubulin, beta [Mus musculus]

45 gi|21311901 472 54 65 97 7.67

dimethylglycine dehydrogenase precursor [Mus musculus]

1.Hit gi|1174621 80 12 26 60 5.44

T-COMPLEX PROTEIN 1, THETA SUBUNIT (TCP-1-THETA) (CCT-THETA)

67 2.Hit gi|31560613 80 12 26 60 5.57

chaperonin subunit 8 (theta) [Mus musculus]

1.Hit gi|12859535 97 10 77 13 6.31

unnamed protein product [Mus musculus]

71a 2.Hit gi|13386100 96 10 77 13 7.82

NADH dehydrogenase (ubiquinone) 1 alpha subcomplex, 5 [Mus musculus]

71c gi|38075371 103 9 61 18 10.32 similar to NADH-ubiquinone oxidoreductase 13 kDa-A subunit, mitochondrial

precursor (Complex I-13KD-A)

(CI-13KD-A) [Mus musculus]

gi|5834958 8 9.88 ATP synthase F0 subunit 8

gi|12643945 31 8.84 Voltage-dependent anion-selective channel protein 3 (VDAC-3)

*Hit: Mascot search hit number

vii

Tab.A2: Protein spots on the Blue-native gels that have been identified in this study:

Spot

Mix* gi-Nr.

Probability

based

Mowse

Score

Matched

Peptide

Sequence

coverage

[%]

[ kDa

] pI Protein name

26 gi|15489120 205 42 34 116 6.36

Ogdh protein [Mus musculus]

gi|18079339 96 21 27 85 8.08

aconitase 2, mitochondrial [Mus musculus]

Mix1 gi|33859811 82 23 33 83 9.24 hydroxyacyl-Coenzyme A dehydrogenase

Mix2 gi|21704020 68 20 28 80

5.51 NADH dehydrogenase (ubiquinone) Fe-S protein 1 [Mus musculus]

Mix1 gi|27369581 73 26 41 75 8.43

solute carrier family 25 (mitochondrial carrier, Aralar), member 12 [Mus

musculus]

29 Mix2 gi|12597627 66 25 35 80

6.15 kinesin family member C3 [Mus musculus]

Mix1 gi|31980648 105 32 49 56 5.19

ATP synthase, H+ transporting mitochondrial F1 complex, beta subunit

Mix2 gi|6680748 103 24 49 60

9.22 ATP synthase, H+ transporting, mitochondrial F1 complex, alpha subunit [Mus

musculus]

Mix1 gi|2690302 129 36 66 47 9.05

aspartate aminotransferase precursor [Mus musculus]

31 Mix2 gi|38259206 74 31 52 47

8.64 creatine kinase, mitochondrial 2 [Mus musculus]

35 gi|21704020 94 27 37 80

5.51 NADH dehydrogenase (ubiquinone) Fe-S protein 1 [Mus musculus]

gi|23346461 66 21 46 53

6.52 NADH dehydrogenase (ubiquinone) Fe-S protein 2 [Mus musculus]

gi|20839603 104 20 50 30 6.67

NADH dehydrogenase (ubiquinone) Fe-S protein 3 [Mus musculus]

Mix1 gi|31980648 193 44 69 56 5.19

ATP synthase, H+ transporting mitochondrial F1 complex, beta subunit

41 Mix2 gi|6680748 175 35 56 60

9.22 ATP synthase, H+ transporting, mitochondrial F1 complex, alpha subunit [Mus

musculus]

gi|16741459 136 19 95 19

5.52 ATP synthase, H+ transporting, mitochondrial F0 complex, subunit d [Mus

musculus]

Mix1 gi|22267442 113 31 53 48 9.26

RIKubiquinol cytochrome c reductase core protein 2 [Mus musculus]

Mix2 gi|12846081 93 30 47 53 5.81

unnamed protein product [Mus musculus]

Mix3 gi|13384794 79 28 46 53

5.75 ubiquinol-cytochrome c reductase core protein 1 [Mus musculus]

49 Mix1 gi|18250284 88 18 41 40 6.27

isocitrate dehydrogenase 3 (NAD+) alpha [Mus musculus]

viii

Spot

Mix* gi-Nr.

Probability

based

Mowse

Score

Matched

Peptide

Sequence

coverage

[%]

[ kDa

] pI Protein name

Mix2 gi|126897 66 12 47 36

8.83 Malate dehydrogenase, mitochondrial precursor

gi|387422 76 6 26 36 8.93

malate dehydrogenase

gi|1372988 80 10 49 20 9.10

cytochrome c oxidase subunit IV

Mix1 gi|13096984 82 23 45 57 5.88

Glucose regulated protein [Mus musculus]

58 Mix2 gi|563510 68 25 48 63 5.85

epoxide hydrolase 2, cytoplasmic [Mus musculus]

Mix1 gi|6680836 208 40 61 48 4.33

calreticulin [Mus musculus]

60 Mix2 gi|38511616 62 19 32 49

9.16 Myef2 protein [Mus musculus]

gi|42415475 205 34 53 57

4.77 prolyl 4-hydroxylase, beta polypeptide; protein disulfide isomerase [Mus

musculus]

Mix1 gi|6680836 89 22 39 48 4.33

calreticulin [Mus musculus]

56 Mix2 gi|42415475 74 19 37 57

4.77 prolyl 4-hydroxylase, beta polypeptide; protein disulfide isomerase [Mus

musculus]

52 gi|6680027 104 29 47 61 8.05

glutamate dehydrogenase [Mus musculus]

Mix1 gi|31981562 135 31 54 58 7.18

pyruvate kinase 3 [Mus musculus]

50 Mix2 gi|6680027 123 31 55 61

8.05 glutamate dehydrogenase [Mus musculus]

Mix1 gi|18700024 127 25 51 42 8.76

isocitrate dehydrogenase 3, beta subunit; isocitrate dehydrogenase 3 beta

Mix2 gi|18250284 76 20 42 40 6.27

isocitrate dehydrogenase 3 (NAD+) alpha [Mus musculus]

Mix3 gi|6680345 67 16 31 43

9.17 isocitrate dehydrogenase 3 (NAD+), gamma [Mus musculus]

gi|53450 94 18 56 25 8.80

manganese superoxide dismutase [Mus musculus]

*Mix: Protein mixtures in the same protein spots.