An integrative approach to identify novel target genes
for reduction of diacetyl production in lager yeast
vorgelegt von
B.Sc-Cam Thuy Duong
aus Hanoi
Von der Fakultät III für Prozesswissenschaften
der Technischen Universität Berlin
zur Erlangung des akademischen Grades
Doktor der Naturwissenschaften
- Dr.rer.nat. -
genehmigte Dissertation
Promotionsausschuss:
Vorsitzender: Prof. Dr. D. Knorr
Berichter: Prof. Dr.U.Stahl
Berichter: Prof. Dr. H.N.Truong
Tag der wissenschaftliche Aussprache: 26.02.2009
Berlin 2009
D 83
ACKNOWLEDGMENTS
First and foremost, I am deeply grateful to Prof. Dr. Ulf Stahl for giving me the opportunity to
work at the Department of Microbiology and Genetics, for his financial support and for his
warm care from the very first days when I was in Berlin.
I would like to particularly express my gratitude to PD Dr. Elke Nevoigt for her great
supervision, helpful discussion and encouragement. I would like to thank her for introducing
me into this interesting topic. Throughout my thesis-writing period, she provided valuable
advices, good teaching and every bit of her precious time to read the manuscript critically.
Without her devoted guidance, this thesis could have not been completed.
I would like to acknowledge Prof. Dr. Hai Nam Truong, Institute of Biotechnology, Hanoi,
Vietnam for his dedicated guidance during the initial step of my scientific career, for giving
me the chance to work abroad and for his continuous support and encouragement.
My special thank go to Dr. Huyen Nguyen Thi Thanh for her enthusiasm, helpful advices and
encouragement to my work and personal life when they were most needed.
Many thanks go to all colleagues in the Institute of Microbiology and Genetics who
contributed to bring this work to completion. I am especially thankful to the colleagues and
student in the Laboratory I: Lysann Strack, Almut Popp, Dörte Müller, Maria Krain and
Dr. Huyen Nguyen Thi Thanh, Isil Bakil, Georg Hubmann for such nice working atmosphere
and technical support.
For finance support, I am very grateful to Das Bundesministerium für Bildung und Forschung
(BMBF) for the Dissertation Scholarship.
I am thankful to Dr. Yukiko Kodama for giving me the opportunity to work at Tokyo Institute of
Technology and Suntory Ltd. and for her enthusiastic cooperation throughout the project. In
addition, I would like to thank our collaboration partners Erich Schuster, Yoshihiro Nakao,
Dr. Yuki Katou, Dr. Matthias E. Futschik and Prof. Dr. Frank-Juergen Methner for their great
contribution to the success of the project.
I would like to thank Dr. Olaf Kniemeyer, Dr. Huyen Nguyen Thi Thanh, Alastair Warren and
Jochen Hoffmann for patient proofreading and English corrections of my thesis. I am thankful
to Nam Dzung Hoang for his help to submit this thesis on time.
Finally, I would like to express my profound gratitude to my parents and my sister for their
unconditional love and support. Much love and thanks go to my husband Tuan Tran and my
little son Tom who give me a lot of strength and encouragement to make this story goes to
the end.
iii
TABLE OF CONTENTS
I
Literature review...............................................................................1
Brewers’ yeast: Targets and strategies for strain improvement ................1
1
Introduction....................................................................................................1
1.1 Overview about brewers’ yeast: history, taxonomy and genetic features....1
1.1.1 History.........................................................................................................1
1.1.2 Taxonomy....................................................................................................2
1.1.3 Genetic features..........................................................................................4
1.2 Brewing process and role of yeast in beer production.................................5
2
Targets and strategies for optimisation of brewers’ yeast strains.............8
2.1 Improvement of carbohydrate consumption ................................................8
2.1.1 Dextrin.........................................................................................................8
2.1.2 Maltose and maltotriose ..............................................................................9
2.2 Improvement of by-product profile.............................................................12
2.2.1 Reduction of diacetyl production................................................................12
2.2.2 Increased production of acetate esters......................................................16
2.2.3 Increase of sulphite production..................................................................17
2.2.4 Elimination of sulphide compounds...........................................................20
2.3 Alteration of flocculation behaviour............................................................22
3
“Omics” technologies in studies regarding brewers’ yeast.....................25
3.1 Genomics..................................................................................................26
3.2 Transcriptomics.........................................................................................27
3.3 Proteomics ................................................................................................28
4
Lager brewers’ yeast genome sequence and its perspective in
brewers’ yeast global studies.....................................................................31
5
Conclusions..................................................................................................33
iv
II
Experimental part............................................................................34
An integrative approach to identify novel target genes for reduction of
diacetyl production in lager yeast.................................................34
1
Introduction..................................................................................................34
1.1 The need to optimise brewers’ yeast.........................................................34
1.2 Former attempts to improve brewers’ yeast ..............................................34
1.2.1 Classical genetic manipulations.................................................................35
1.2.2 Rational metabolic engineering.................................................................37
1.3 Inverse metabolic engineering as an alternative approach for improving
brewers’ yeast...........................................................................................39
2
Aim of work...................................................................................................41
3
Materials and methods ................................................................................43
3.1 Materials....................................................................................................43
3.1.1 Equipment.................................................................................................43
3.1.2 Enzymes, chemicals and kits ....................................................................44
3.1.3 Strains.......................................................................................................44
3.1.4 Media and culture conditions.....................................................................46
3.1.5 Plasmids....................................................................................................47
3.1.6 Oligonucleotides........................................................................................47
3.2 Methods.....................................................................................................48
3.2.1 DNA methods............................................................................................48
3.2.1.1 Isolation of yeast genomic DNA.................................................................48
3.2.1.2 Isolation of plasmid DNA from E. coli (minipreparation) ............................48
3.2.1.3 Agarose DNA gel electrophoresis..............................................................48
3.2.1.4 PCR...........................................................................................................49
3.2.1.5 Transformation of lager brewers’ yeast .....................................................50
3.2.1.6 Transformation of E. coli............................................................................51
3.2.1.7 Microarray-based comparative genomic hybridisation...............................51
3.2.2 RNA method: Microarray-based comparative transcriptome analysis.......53
3.2.2.1 Isolation of brewers’ yeast total RNA.........................................................53
3.2.2.2 Sample preparation and array hybridization..............................................53
3.2.2.3 Data acquisition and analysis....................................................................54
3.2.3 Protein methods........................................................................................55
v
3.2.3.1 Comparative proteome analysis................................................................55
3.2.3.1.1 Isolation of total protein from bottom fermenting yeast..............................55
3.2.3.1.2 Two-dimensional gel electrophoresis ........................................................55
3.2.3.1.3 MALDI-TOF mass spectrometry................................................................57
3.2.3.2 Determination of protein concentration......................................................57
3.2.3.3 Determination of enzyme activity...............................................................57
3.2.3.3.1 Preparation of permeabilized cell proteins.................................................57
3.2.3.3.2 In vitro acetohydroxyacid synthase (AHAS) assay....................................58
3.2.4 Fermentations............................................................................................59
3.2.5 Analytical methods ....................................................................................61
4
Results..........................................................................................................62
4.1 Phenotypes of the three selected lager brewers’ yeast strains producing
different levels of diacetyl ..........................................................................62
4.1.1 Wort sugar consumption during the main fermentation of the three
selected lager brewers’ yeast strains.........................................................62
4.1.2 Diacetyl production....................................................................................63
4.1.3 Flocculation behaviour...............................................................................64
4.1.4 Other fermentation by-products.................................................................65
4.2 Global molecular analyses of the three selected brewers’ yeast strains
possessing variations in diacetyl production..............................................67
4.2.1 Genome level: Microarray-based comparative genome hybridization
using bottom fermenting yeast DNA microarray........................................67
4.2.1.1 Differences of the studied strains in their chromosome patterns...............70
4.2.1.1.1 Strain A vs strain C....................................................................................70
4.2.1.1.2 Strain B vs strain C....................................................................................72
4.2.1.2 Identification of differences in copy number of known genes relevant to
diacetyl formation and flocculation.............................................................75
4.2.2 Transcriptome level: Microarray-based comparative transcriptome
analysis .....................................................................................................78
4.2.2.1 Transcriptome analysis using bottom fermenting yeast DNA microarray ..78
4.2.2.2 Identification of differences at transcriptional level of genes relevant to
diacetyl and flocculation ............................................................................80
4.2.3 Proteome level: Comparative proteome analysis using two-dimensional
gel electrophoresis ....................................................................................83
vi
4.2.3.1 Identification of protein spots which showed significant different
intensities among the three studied lager brewers’ yeast strains ..............83
4.2.3.2 Mass spectrometry identification of differentially expressed proteins........84
4.3 Sc-ILV6, a potential novel target gene for reducing diacetyl production in
brewers’ yeast...........................................................................................88
4.4 In vitro acetohydroxyacid synthase activity in the studied strains..............88
4.5 Disruption of Sc-ILV6 in strain C for reduced diacetyl production..............89
4.5.1 Deletion of the first Sc-ILV6 gene copy in strain C: generation of a
Sc-ilv6
∆
single deletion strain....................................................................89
4.5.2 Deletion of the second Sc-ILV6 gene copy: generation of a
Sc-ilv6
∆
/Sc-ilv6
∆
double deletion strain.....................................................92
4.5.3 In vitro acetohydroxyacid synthase activity in strains Sc-ilv6
∆
and
Sc-ilv6
∆
/Sc-ilv6
∆
........................................................................................97
4.5.4 Fermentation characteristics and vicinal diketone production of strains
Sc-ilv6
∆
and Sc-ilv6
∆
/Sc-ilv6
∆
under the laboratory scale fermentation...98
4.5.5 Fermentation characteristics and vicinal diketone production of strain
Sc-ilv6
∆
/Sc-ilv6
∆
under industrially relevant brewery fermentation...........99
4.5.6 Determination of flavour-relevant products in green beer produced by
strain Sc-ilv6
∆
/Sc-ilv6
∆
...........................................................................101
5
Discussion..................................................................................................103
5.1 Global genetic analyses of the three lager brewers’ yeast strains
producing different levels of diacetyl........................................................104
5.2 Identification of potential novel target genes for reduction of diacetyl
production: Sc-ILV6, Sc-BAT1, non-Sc-BAT1, non-Sc-BAT2..................108
5.3 Impact of Sc-ILV6 disruption on in vitro AHAS activity and vicinal
diketone reduction...................................................................................114
5.4 Fermentation performance of the Sc-ilv6 double deletion mutant............116
5.5 Slight change of by-product profile in the green beer produced by strain
Sc-ilv6
∆
/ Sc-ilv6
∆
....................................................................................117
5.6 Concluding remarks and outlook.............................................................118
6
Summary.....................................................................................................121
7
Zusammenfassung ....................................................................................123
8
References..................................................................................................125
vii
Abbreviations
2D Two-dimensional
AHAS acetohydroxyacid synthase
BCAA Branched-chain amino acid
bp base pair(s)
BPB Bromophenol blue
BSA Bovine serum albumin
cDNA complementary deoxyribonucleic acid
cRNA complementary ribonucleic acid
CHAPS 3-[(3-Cholamidopropyl)-dimethylammonio]-1-propane sulfonate
CGH comparative genomic hybridisation
DEPC Diethyl Pyrocarbonate
DMSO Dimethylsulfoxide
DNA Deoxyribonucleic acid
DTT Dithiothreitol
E. coli Escherichia coli
EDTA Ethylendiaminetetraacetic acid
EtBr Ethidium bromide
FAD Flavin adenin dinucleotide
Fig. Figure
G418 Geneticine
GC-ECD Gas Chromatography - Electron Capture Detector
GCOS GeneChip Operating System
GMOs Genetically modified organisms
h hour
hl hectorlitre
IAA Idoacetamid
IVT In vitro transcription
kb kilobase
Km
R
Kanamycin resistance
LB Luria-Bertani
M, mM Molar, Millimolar
MALDI-TOF Matrix-Assisted Laser Desorption/Ionization Time of Flight
MEBAK Mitteleuropäische Brautechnische Analysenkommission
min minute
nm nanometre
viii
nt nucleotide
OD optical density
ORF open reading frame
PCR polymerase chain reaction
PMSF phenylmethylsulfonyl flourid
RNA Ribonucleic acid
rpm Rotations per minute
RT room temperature
S. bayanus Saccharomyces bayanus
S. carlsbergensis Saccharomyces carlsbergensis
S. cerevisiae Saccharomyces cerevisiae
Sc-type Saccharomyces cerevisiae type
Non-Sc-type non-Saccharomyces cerevisiae type
S. monacensis Saccharomyces monacensis
S. pastorianus Saccharomyces pastorianus
s second
SGD Saccharomyces Genome Database
SDS Sodium dodecyl sulphate
Tab. Table
TEMED Tetramethylethylenediamine
ThDP Thiamin pyrophosphate
Tris Tris (hydroxymethyl) aminomethane
V Voltage
VLB Versuchs-und Lehranstalt fuer Brauerei in Berlin
Vol Volume
w/v weight/volume
wt wild type
YED Yeast Glucose medium
YEPD Yeast Peptone Glucose medium
ix
Index of figures and tables
Figures Page
Figure 1 Previous modifications of the valine biosynthetic pathway to reduce diacetyl production
in brewers’ yeast 13
Figure 2 Modifications of sulphate assimilation and sulphur-containing amino acid biosynthetic
pathway to control sulphite and sulphide production 18
Figure 3 Time courses of apparent extract during the fermentation by the three selected lager
brewers’ yeast strains 63
Figure 4 Different diacetyl productions of the three studied lager brewers’ yeast strains 64
Figure 5 Difference in flocculation behaviour between the studied lager brewers’ yeast strains 64
Figure 6 Putative chromosomal structure of lager brewers’ yeast strain Weihenstephan Nr.4
(34/70) (Kodama et al., 2006) 68
Figure 7 Translocation between the non-Sc-type chromosomes II and IV and between the
non-Sc-type chromosomes VIII and XV in lager brewers’ yeast (Y.Nakao, Genomic
analysis of lager brewing yeast and its application to brewing)
69
Figure 8 Array-based genomic comparison of the studied lager brewers’ yeast strains by means
of bottom fermenting yeast DNA microarray 71
Figure 9 Chromosomal differences and possible compositions of some chromosomes in strains
B and C 73
Figure 10 Microarray DNA hybridisation signals of genes encoding enzymes of the valine
biosynthetic pathway in the studied lager brewers’ yeast strains 76
Figure 11 Microarray-based DNA and microarray-based normalised RNA hybridisation signals of
genes encoding enzymes involved in the valine biosynthetic pathway in the studied
strains
81
Figure 12 Two-dimensional gel image of a lager brewers’ yeast strain (strain C) at apparent
extract of 8% 84
Figure 13 Significant differentially expressed protein spots among the three studied lager brewers’
yeast strains at apparent extract of 8% detected in 2D gels 86
Figure 14 In vitro activity of acetohydroxyacid synthase (AHAS or Ilv2p) in strains B and C 88
Figure 15 Disruption of the first copy of Sc-ILV6 in strain C 90
Figure 16 Diagnostic PCR to check the correct single deletion mutant Sc-ilv6
∆
in the 8 selected
clones 91
Figure 17 Disruption of the second copy of Sc-ILV6 in the Sc-ilv6
∆
single deletion strain 93
Figure 18 Diagnostic PCR to confirm the correct Sc-ilv6
∆
/Sc-ilv6
∆
double deletion in the selected
clones 94
Figure 19 Diagnostic PCR to check the correct disruption of Sc-ILV6 instead of the wrong
disruption of non-Sc-ILV6 ORF 96
Figure 20 In vitro activity of acetohydroxyacid synthase (AHAS or Ilv2p) in strains B, C,
Sc-ilv6
∆
and Sc-ilv6
∆
/Sc-ilv6
∆
97
Figure 21 Vicinal diketone production by strains B, C and Sc-ilv6
∆
and Sc-ilv6
∆
/Sc-ilv6
∆
98
Figure 22 Fermentation performance of strain Sc-ilv6
∆
/Sc-ilv6
∆
in comparison to the wild type
strain C 99
Figure 23 Vicinal diketone production by strain Sc-ilv6
∆
/Sc-ilv6
∆
in comparison to the wild type
strain C 100
Figure 24 Branched-chain amino acid biosynthesis and degradation in yeast and some related
metabolites 113
x
Tables Page
Table 1 Strains used in this study 45
Table 2 Plasmids used in this study 47
Table 3 Analysis of green beers produced by the three studied brewers’ lager yeast strains
(harvested at apparent extract of 3%) 66
Table 4 Differences of genes belonged to the flocculation gene family identified at genome level
in the studied lager yeast strains 77
Table 5 Number of significant differences identified at transcriptome level among the studied
lager yeast strains via analysis using bottom fermenting yeast DNA microaray 79
Table 6 Differences in flocculation genes identified at level of genome and transcriptome in the
studies strains 82
Table 7 MALDI-TOF mass spectrometry identification of significantly different protein spots
detected in the proteome comparisons of the three studied lager yeast strains 85
Table 8 Analysis of green beers produced by strain Sc-ilv6
∆
/Sc-ilv6
∆
and the reference strain C
(harvested at apparent extract of 3%) 102
1
I Literature review
Brewers’ yeast: Targets and strategies for strain improvement
1 Introduction
1.1 Overview about brewers’ yeast: history, taxonomy and genetic features
1.1.1
History
Beer brewing is one of the oldest technologies in the world and its history can be
traced for several millenniums. For most of the time, beer brewing was considered as
a spontaneous event. It was originally performed based on the experience that cereal
grains used for brewing would potentially result in alcohol production when they had
been stored under wet condition (Corran, 1975). Although beer brewing has an
ancient history, the role of yeast in beer fermentation has been only known from 19
th
century. In 1680, Antonie Van Leeuwenhook observed “yeast flocs” in fermenting
wort through a microscope. Nevertheless, no comment about the role of yeast in
fermentation was stated (Briggs et al., 2004). The presence of microorganisms in
fermentation was only recognised between 1836 to 1838 as the result of independent
works of Theodore Schwann, Friedrich Traugott Kuetzig and Charles Cagniard-
Latour (Briggs et al., 2004). Based on observation of “yeast” cells under
miscroscopes, Kuetzig and Cagniard assumed they were living organisms and were
necessary in the brewing process. Schwann also observed the growth of yeast cells
through a microscope and designated them as ’Zuckerpilz’’. The theory of living
organism being responsible for the alcoholic fermentation process encountered a
strong opposition by some eminent chemists for a long time (Barnett, 2004). Only in
1861s was the importance of yeast in fermentation generally accepted, thank to the
work of Louis Pasteur. Another milestone in the history of brewing was the work of
Emile Hansen (1883). By developing the technique of generating pure cultures in
solid media invented by Robert Koch (1881), Hansen isolated the first pure culture
2
brewing yeast named “Carlsberg Yeast Number 1”. From that time, the use of pure
cultured yeast became popular in beer brewing.
In general, there are two main kinds of beer i.e. ale and lager beer, these are
dependent on the yeast and conditions used for fermentation. For ale beer brewing,
the fermentation is carried out at room temperature using ale yeast strains (from
about 20 to 25
ο
C). The fermentation of lager beer is performed under lower
temperatures (8-14
ο
C) using the lager brewers’ yeast strains. After the main
fermentation, ale beer production is subjected to a short period of aging whilst the
lager beer production has to undergo a long maturation period lasting from one to
three weeks at low temperature (around 0
o
C). Ale beer has a fruity aroma whilst
lager beer is paler, drier and usually has a lower alcohol content (Polaina, 2002). At
the end of fermentation, ale yeast rises to the top of the fermentation vessels whilst
lager yeast settles down to the bottom. They are therefore called top-fermenting and
bottom-fermenting yeast, respectively. Whilst ale beer is believed to be produced in
3000 BC in Mesopotamia (Corran, 1975), history of lager beer is much shorter, only
being recorded from the 19
th
century. Bottom-fermenting yeast was secretly used by
Bavarian brewers’ until the 1840s when it was illegally transported to Czechoslovakia
and Denmark (Boulton and Quain, 2001). The lager yeasts were then spread
throughout other parts of Europe and North America. Currently, lager beers are
brewed worldwide and comprise of 90% beer production of the world while ale beers
are mostly produced on the British Isles (Kodama et al., 2006).
1.1.2
Taxonomy
Recent descriptions about brewers’ yeast taxonomy were given by Boulton and
Quain (2001) and by Briggs and colleagues (2004). Whilst ale yeast is classified as
Saccharomyces cerevisiae, lager brewers’ yeast is taxonomically much more
complicated and has been renamed several times. Barnett (2004), in his review
about yeast taxonomy study pointed out factors for the instability in yeast
3
nomenclature including: i) criteria used for classification, ii) development of lab
techniques, iii) discovery of new kinds of yeast and iv) nomenclature correction of
one taxon which is unintentionally named several times. The change of lager yeast’s
nomenclature is predominantly consequence of the first two factors listed.
The aforementioned first pure brewers’ yeast strain in the world, propagated by
Emile Hansen, was a lager brewers’ yeast strain. In 1908, Hansen named this
bottom-fermenting yeast as S. carlbergensis in recognition of its difference from ale
yeast which has been used in the traditional beer production of Belgium, Germany
and Britain. This strain is suggested to be closely related to most current lager
brewers’ yeast strains (Hansen and Kielland-Brandt, 2003). Lager brewers’ yeast
was then consolidated in S. uvarum since it was recognised to be almost
undistinguishable from this kind of wine yeast (Campbell, 2000). Later, based on the
criteria of nurtrient consumption, cell morphology and mode of reproduction, Yarrow
(1984) assigned lager brewers’ yeast assigned to the species S. cerevisiae.
From the beginning to Yarrow’s classification, taxonomy of brewers’ yeast was
mostly based on its ability to assimilate certain substrates, its colony and cell
morphology, mode of reproduction and based on microscopic experience of
scientists. With the development of recombinant DNA technology, from 1985, DNA
characteristic criteria have been applied and provided a more precise classification of
brewers’ yeast. By using DNA re-association, Vaughan-Martini and Kurztman (1985)
demonstrated that the DNA of the original S. carlbergensis showed high homology to
both S. cerevisiae (53%) and S. bayanus (72%). Since the genomes of S. bayanus
and S. cerevisiae showed little similarity (20%), Vaughan-Martini (1985) suggested
that lager brewers’ yeast was the hybrid of S. cerevisiae and S. bayanus. Later, lager
brewers’ yeast was grouped into S. pastorianus based on the fact that they were
93% homologous in genome constitution (Vaughan-Martini and Martini, 1987). Until
recently, it has been generally accepted that ale yeast is S. cerevisiae and lager
yeast is S. pastorianus. Compared to Yarrow’s classification, this taxonomy seems to
4
be more “comfort” to brewers in its clear reflection of the physiological differences
between ale and lager brewers’ yeast.
1.1.3
Genetic features
According to their genetic constution, ale and lager yeast are different. In addition,
ale yeast strains are much more diverse than lager yeast strains. A chromosomal
fingerprinting study proved that most lager yeast in the brewing world have one or
two basic fingerprints namely “Turborg” or “Carlsberg” while ale yeasts failed to show
any common fingerprint (Casey, 1996). Ale yeast strains revealed to be polyploid and
closely related to laboratory strains of S. cerevisiae. In contrast, lager yeast strains
are allopolyploid hybrids of S. cerevisiae and another Saccharomyces yeast
(see I.1.1.2). In comparison to laboratory yeast strains, the amplified fragment length
polymorphism (AFLP) pattern of ale yeasts showed 93.7% commonality while it was
only about 74.6 % in the case of lager yeasts (Azumi and Goto-Yamamoto, 2001).
Two-dimensional gel electrophoresis of the proteomes also showed that ale yeast
strains were much closely related to lab yeast S288c than the lager yeast strains
(Kobi et al., 2004).
Using DNA re-association experiment, Vaughan Martini (1985) was the first
author to verify the hybrid nature of lager brewers’ yeast (see I.1.1.2). Since then, the
existence of diverged genomes in lager yeast was confirmed in a series of studies
using different methods such as Southern analysis of several genes, kar-mediated
single chromosome transfer and hybridisation of radioactive probes to chromosome-
sized DNA separated pulse-field electrophoresis (review by Kodama et al., 2006).
Several attempts have been aimed in elucidating the origins of lager brewers’ yeast.
Even though most studies agree that S. cerevisiae is the first parent of lager brewers’
yeast, there were different ideas about its second ancestor. By comparing DNA
homology, Vaughan Martini et al. (1985) recognized the second parent of lager
brewing yeast as a S. bayanus type strain (CBS 380). This hypothesis was supported
5
by the investigation of homology in the the non-S. cerevisiae sequences between
lager brewers’ yeast and S. bayanus (Tamai et al., 1998; Yamagishi and Ogata,
1999; Casaregola et al., 2001; Kodama et al., 2001a). In contrast, several studies
based on Southern analysis and molecular cloning suggested that an S. monacensis
type strain (CBS 1503) could be the other contributor of the lager brewing genome
(Pedersen, 1986a; Pedersen, 1986b; Hansen and Kielland-Brandt, 1994). However,
it was revealed that S. bayanus CBS 380 (Hansen and Kielland-Brandt, 1994;
Pedersen, 1986a) and S. monacensis CBS 1503 (Andersen et al., 2000; Casaregola
et al., 2001) themselves are hybrids containing divergent versions of many genes.
The species S. bayanus contains two varieties: S. bayanus var. bayanus and
S. bayanus var. uvarum. It was proved that between these two varieties, only
S. bayanus var. bayanus contained strains which contributed to lager brewers’ yeast
genomes (Naumova et al., 2005). This variety contains a collection of hybrid strains
which are similar to S. bayanus CBS380. Lager brewers’ yeast therefore seems to
represent one among many hybridisation events occurred between S. cerevisiae and
S. bayanus. In a laborious work using sequencing and restriction analysis of 48 gene
fragments chosen randomly within the S. bayanus genome, Rainieri and colleagues
(2006) identified the third group of S. bayanus which only contains two pure genetic
lines: IF0539 and IFO1948 without sequence from S. cerevisiae. These two isolates
are supposed to represent the pure non-S. cerevisiae genomic content of lager
brewers’ yeast (Kodama et al., 2006; Rainieri et al., 2006).
1.2 Brewing process and role of yeast in beer production
The aim of the brewing process is the conversion of grain starch and proteins to
fermentable sugars and amino acids, subsequently to extract these nutrients with
water and to ferment them with yeast to produce beer, an alcoholic, carbonated and
aromatic beverage. The brewing process involves five main stages: i) malting,
6
ii) milling, mashing and wort production, iii) wort boiling, iv) fermentation and v) post
fermentation treatment.
In the malting step, barley is germinated by being steeped into water. The
germination process is about two weeks long and results in the biosynthesis and
activation of amylotic and proteolytic enzymes to convert barley starches and
proteins into fermentable sugars and free amino nitrogen, respectively. This
germination is stopped by heating. In the second step, the dry malt is milled and
mixed with water. The temperature of this mixture is then increased in several steps
to provide optimal conditions for the activity of amylotic and proteolytic enzymes
which facilitate sugar and protein degradation. After that, sweet wort is produced by
separating the aqueous phase from the residual grains. Sweet wort is then boiled
with hops to extract aroma and bitter hop compounds. The product is called brewers’
(hopped) wort ready for use in fermentation. Following this, freshly propagated yeast
is inoculated into wort and the fermentation begins. In this process, yeast utilises
fermentable sugars and nutrients in the wort for growth and maintenance and in turn
releases ethanol, carbon dioxide and various by-products to form “green beer”. The
green beer becomes “drinkable beer” after the maturation, filtration and
pasteurisation. The maturation course is also called “secondary fermentation” and is
needed for the improvement of beer flavour and aroma.
In ale brewing, the fermentation lasts about two or three days at room
temperature, whilst lager brewing fermentation takes from 5 to 10 days at lower
temperatures of between 8-15
ο
C. At the end of fermentation, yeast is harvested and
often used in subsequent fermentations. Depending on the type of fermentation,
yeast cells are collected either from the surface (ale yeast) or from the bottom
(lager yeast) of the fermentation vessels. During the maturation period, many
undesirable organoleptic compounds are reduced to the acceptable levels. Among
these undesirable substances, diacetyl is of the most concern to brewers, especially
in lager beer brewing. Even present at low concentrations, it has a strong impact by
causing a butter-like flavour in beer (Virkajarvi, 2006). Diacetyl may be a part of the
7
flavour in some ale beers; however, it is an off-flavour in lager beers. Removing
diacetyl in lager beer production requires a long maturation process which may last
from one to three weeks.
Brewers’ yeast only participates in two phases, i.e the fermentation and
maturation of beer brewing process. The role of yeast in the fermentation, however,
is active and decisive by the fact that most of beer components such as alcohol,
carbon dioxide and flavours compounds are brewing yeast’s metabolites and this
metabolite pattern depends strongly on brewing yeast genotypes. Improvement of
brewers’ yeast strain has therefore received a great attention in the optimisation of
the fermentation process. In the following part, I will give an overview of targets and
strategies for the genetic improvement of brewers’ yeast strain. The “omics” studies
of brewers’ yeast are also mentioned. The application of “omics” technologies in
brewers’ yeast studies has led to an increased knowledge about cellular activities of
brewers’ yeast during the main fermentation. The lager brewers’ yeast genome
sequence is emphasized in its perspective in global studies of brewers’ yeast. The
accessibility of brewing yeast genome database can provide useful tools for global
studies; thus endowing an insight into the nature of brewers’ yeast. The enlargement
of knowledge about yeast nature will be a valuable basis for the strain improvement
of brewers’ yeast.
8
2 Targets and strategies for optimisation of brewers’ yeast strains
2.1 Improvement of carbohydrate consumption
2.1.1
Dextrin
During the malting and mashing process, barley starch is degraded to simple
sugars which are fermentable by brewers’ yeast. These fermentable sugars are
comprised of 75% of wort carbohydrates, including glucose, fructose, maltose,
galactose and maltotriose. Another product of barley starch degradation is
polysaccharides (dextrins) of varying length. Dextrins constitute at least 20% of
brewers’ wort carbohydrates; however, they are not utilisable by brewers’ yeast. The
creation of brewers’ yeast capable of fermenting dextrin has become a target of
industrial brewing in the production of low calorie beer and the production of higher
alcoholic amounts from the same amount of malt (Campbell, 2000).
Dextrins are mixtures of D-glucose polymers which have linear glycosidic α-1,4
and branched glycosidic α-1,6 linkages. Among Saccharomyces yeasts,
S. diastaticus is known to have the capability of hydrolysing and fermenting dextrin
by producing extracellular glucoamylases. These enzymes are encoded by three
unlinked polymeric genes i.e. STA1, STA2 and STA3. Several attemptes were made
to confer this feature of S. diastaticus to brewers’ yeast. The first approach was the
hybridisation of brewers’ yeast strain with the wild yeast S. diastaticus. The resulting
progeny was able to utilise dextrin, however, it also governs other genetic make-up
from S. diastaticus, notably the inheritance of the POF1 gene. The presence of this
gene in the hybrid genome confers the ability of ferulic acid decarboxylation, resulting
in a phenolic off-flavour in beer (Hansen and Kielland-Brandt, 1997).
Other attempts to improve dextrin utilisation in brewers’ yeast involved the direct
transfer of either S. diastaticus STA1 or STA2 gene of to brewers’ yeast using
plasmid expression or integration approaches (Meaden and Tubb, 1985; Perry and
Meaden, 1988; Sakai et al., 1989; Vakeria and Hinchliffe, 1989; Park et al., 1990).
These studies more or less gained certain success especially those which involved
9
the integrated approaches and thus conferred a better genetic stability to the
transformants. Nonetheless, a drawback of glucoamylases derived from
S. diastaticus is the lack of α-1,6 debranching activity leading to a high amount of
unfermented dextrin. To gain higher dextrin-fermenting efficiency, genes endowing
both α-1,6 and α-1,4 glucoamylase activity from other fungi were introduced into
brewers’ yeast (Yocum, 1986a; Cole et al., 1988; Gopal and Hammond, 1992).
Integration of the glucoseamylase gene from Aspergillus niger to brewers’ yeast
genome was highly successful. This was demonstrated by the fact that 50% of wort
dextrin was utilised resulting in a 20% increase in ethanol concentration (Gopal and
Hammond, 1992). This enzyme from Aspergillus niger, however, is heat stable and
therefore not being denatured after pasteurisation of beer. The presence of active
glucoamylase made the beer became sweet during the storage. To overcome this
problem, the GAM1 gene from Swanniomyces occidentalis was introduced into
brewers’ yeast (Lancashire et al., 1989). The resulting glucoamylase is both
heat-labil and possesses debranching activity and the transformant can ferment
dextrin efficiently.
2.1.2
Maltose and maltotriose
Maltose is the most abundant sugar in brewing wort accounting for ca. 60% of
total fermentable sugars (Vidgren et al., 2005). The fermentation of maltose only
starts when 50% of wort glucose is consumed (Stewart et al., 1983). This
phenomenon results from the glucose repression of genes which are responsible for
the uptake and hydrolysis of these sugars in the cell. The improvement of maltose
utilisation of brewers’ yeast is important in brewing fermentation, especially in the
high gravity fermentation in which glucose is present at high amounts and in
accelerating the rate of fermentation.
Maltose assimilation in yeast requires the presence of at least one among five
unlinked MAL loci namely MAL1-4 and MAL6. Each MAL locus consists of three
10
genes MALxT, MALxR and MALxS (with x referring to the number of the locus)
encoding for a maltose permease, a positive regulator and a maltase, respectively.
These three genes are repressed by Mig1p in the presence of glucose (Hu et al.,
1995). In addition, MALT and MALS expression is induced by maltose. The
repression involves MIG1 while induction involves MALR (Klein et al., 1996). Apart
from transcriptional regulation, maltose assimilation also involves post-transcriptional
regulation and post-translational control in which the presence of glucose leads to the
increase in the lability of the MALS transcript and to the inactivation of the maltose
permease (Go¨rts, 1969; Siro and Lo¨vgren, 1979; Federoff et al., 1983; Peinado and
Loureiro-Dias, 1986; Hu et al., 1995; Lucero et al., 1993). Besides maltose
permease, maltose is also being taken up via AGT1-encoded transporter which is a
broad specificity α-glucosidase transporter (Han et al., 1995). AGT1 is allelic but is
only 57% identical to MAL1T.
To improve maltose fermentation efficiency, Kodama and colleagues (1994)
overexpressed MAL genes by using a constitutive promoter which is not repressed
by glucose in one brewers’ yeast strain. In high gravity fermentation, the constitutive
expression of MALT gene was effective to improve maltose fermentation efficiency,
whilst the expression of MALS or MALR had no impact on maltose consumption.
Another attempt to accelerate maltose fermentation in yeast was based on the
removal of the repression factor which regulated the transcription of MAL genes
(Klein et al., 1996). Disruption of the MIG1 gene resulted in a decrease in maltose
repression only in a haploid laboratory strain while it led to a stricter glucose control
on maltose metabolism in an industrial yeast strain. That effect on the industrial strain
is supposed to be caused by the increased glucose control on the maltose permease
resulting in the alteration in the uptake of maltose (Klein et al., 1998).
Some recent studies have focused on the clarification of MALT gene combination
and on functionality of maltose transporters in brewers’ yeast. In an examination
involving 25 lager and 5 ale yeast strains, Jesperson and colleagues (1999), by using
hybridisation genes probes to seperate chromosomes, showed that different brewers’
11
yeast strains had diverse combinations of MAL genes. In fact, all 30 studied brewers’
strains contained MAL1T, MAL3T and AGT1 and only one of those strains lacked the
MAL4 gene. MAL2T was not detected in 12 lager yeast strains, nor in any of the
tested ale strains. MAL6T was not found in any of all 30 tested brewers’ yeast
strains. Through the use chromosome blot and hybridisation, another study on
different maltose transporters showed that maltose was mostly taken up via the
MALxT transporters in lager strains while in ale strains it was predominantly carried
out by AGT1-encoded transporter (Vidgren et al., 2005). This study also indicated
that some apparent multiple maltose transporter genes in several brewers’ yeast
strains did not encode functional transporters.
Besides maltose, the fermentation of maltotriose is also of concern in brewing
fermentation. Maltotriose is the second most abundant fermentable sugar in brewing
wort (comprising 15-20 %); however, it is least preferred to be taken up by yeast cells
compared to glucose and maltose (Sergio L. Alves et al., 2008). The consumption of
maltotriose in ale yeast is significantly slower than in lager yeast and is therefore
more problematic in the ale brewing fermentation. Hydrolysis of maltose and
maltotriose requires the same maltase, however, it was unclear whether there exists
a specific transporter for maltotriose or if maltotriose is co-transported with maltose
via maltose transporters (Salema-Oom et al., 2005). Recently, a novel gene MTY1
encoding an α-glucoside transporter was identified in lager brewers’ yeast (Salema-
Oom et al., 2005). This new gene is 90% and 45% identical to MAL3T and AGT1
genes respectively. Overexpression of MTY1 conferred the capability of fermenting
maltose and maltotriose in an S. cerevisiae Mal
-
strain. Interestingly, the Mytp is
distinct from other α-glucoside transporters as it has higher affinity for maltotriose
than maltose.
12
2.2 Improvement of by-product profile
2.2.1
Reduction of diacetyl production
Vicinal diketones (diacetyl and 2,3-pentanedione) impart undesirable butter-like
flavour to beer. Among these two substances, diacetyl is of more concern to brewers
since it has a much lower taste threshold than 2,3-pentanedione. Diacetyl is a
by-product of the valine biosynthetic pathway which is formed from the
non-enzymatic oxidative decarboxylation of α-acetolactate. The latter compound
leaks out from the cells during the main fermentation (Fig. 1). Diacetyl was then
reabsorbed into yeast cells and there it was reduced to acetoin and subsequently to
2,3-butanediol, a compound which has much higher taste threshold in beer. Diacetyl
is reduced to acceptable levels during the maturation. The main purpose of the
maturation process in lager beer brewing is indeed the diacetyl removal and the
completion of this process may last from one to three weeks. Prevention of diacetyl
production would therefore help to shorten the maturation process thus accelerating
lager beer production.
In general, diacetyl production can be reduced by different strategies:
i) elimination of diacetyl formation from its precursor α-acletolactate, ii) reduction of
α-acetolactate production and iii) increase of the conversion of α-acetolactate
towards the valine biosynthetic pathway.
In the first approach, to prevent the formation of diacetyl from its precursor
α-acetolactate, heterogeneous α-acetolactate decarboxylase was either introduced
into green beer or expressed in brewers’ yeast. This enzyme catalyzes the direct
conversion of α-acetolactate to acetoin, thereby eliminating diacetyl formation
(Fig. 1). The addition of α-acetolactate decarboxylase isolated from Enterobacter
aerogenes to green beer led to a decrease in vicinal diaketones levels under the
taste-threshold after 24 h at 10
o
C (Godtfredsen et al., 1987). The use of acetolactate
decarboxylases in brewing was approved in 2001 by US Food and Drug
Administration in USA (Hannemann, 2002); however, the addition of this enzyme is
13
incompatible with the German beer purity law (Donalies et al., 2008). Alternatively,
ALDC genes encoding α-acetolactate decarboxylase from different bacteria
i.e. Enterobacter aerogenes, Klebsiella terrigena, Lactococcus lactis and Acetobacter
aceti were expressed in yeast using either episomal plasmids, genomic or rDNA
integration (Sone et al., 1987; Goelling and Stahl, 1988; Sone et al., 1988; Fujii et al.,
1990; Blomqvist et al., 1991; Yamano et al., 1994). By using this strategy, diacetyl
formation was reduced efficiently and in some cases, the maturation period could be
ignored. The ALDC genes from Lactococcus lactis and Acetobacter aceti are
considered to be more acceptable for food application since these organisms have
been already used in food production (Hammond, 1995). In the current opinion, the
“self-cloned” yeast strains which do not contain any additional heterogeneous DNA
are assumed to be more accepted in food and beverage approval (Akada, 2002).
Pyruvate
α
αα
α-Acetolactate
α
αα
α, β
ββ
β- Dihydroxy
-isovalerate
α
αα
α- Ketoiso-valerate
Valine
Acetoin
ILV2
ILV5
2,3-Butanediol
α
αα
α-Acetolactate
Diacetyl
ALDC
ILV3
Diacetyl
ILV6
Acetoin
ALDC
BAT1, 2
Pyruvate
α
αα
α-Acetolactate
α
αα
α, β
ββ
β- Dihydroxy
-isovalerate
α
αα
α- Ketoiso-valerate
Valine
Acetoin
ILV2
ILV5
2,3-Butanediol
α
αα
α-Acetolactate
Diacetyl
ALDC
ILV3
Diacetyl
ILV6
Acetoin
ALDC
BAT1, 2
Fig. 1 Previous modifications of the valine biosynthetic pathway to reduce diacetyl
production in brewers’ yeast. The big and disrupted arrows indicate overexpression and
prevention of corresponding enzyme activity respectively. The dash arrow indicates the introduction
of heterogenouos enzyme. ILV2, ILV6: acetohydroxyacid synthase, ILV5: reductoisomerase, ILV3:
dihydroxyacid dehydratase; BAT, BAT2: branched-chain amino acid transaminase.
14
The second genetic strategy to reduce diacetyl has been based on blocking the
formation of its precursor α-acetolactate. As acetohydroxyacid synthase (also called
acetolactate synthase) is the enzyme responsible for the formation of α-acetolactate,
there have been different attempts in eliminating the activity of this enzyme.
Gjermansen et al. (1988) completely remove acetolactate synthase activity in one
lager brewers’ yeast strain by introduction of in vitro constructed ilv2 deletion. The
resulting deletion strain no longer produced α-acetolactate but encountered nutrient
deficiency since valine uptake from extracellular medium was not sufficient for growth
(Kiellandt-Brandt et al., 1995). It was discovered that sulfometuron methyl (SM) is an
inhibitor of acetolactate synthase and mutation of the ILV2 gene leads to the
insensitivity of this enzyme to SM (Falco and Dumas, 1985). Based on this fact, a
partial block of acetolactate synthase was obtained. In this approach, at first
spontaneous primary allodiploid mutants which were resistant to SM and prototrophic
for valine and isoleucine were isolated (Kiellandt-Brandt et al., 1989). These mutants
were then treated with UV radiation and the secondary mutants reversely sensitive
for SM were screened. In S. cerevisiae, this procedure would result in the strain
carrying two copies of wildtype ILV2 gene dues to the high frequency of mitotic
recombination. In brewers’ yeast, since these mutants were alloploid, the frequency
of mitotic recombination was low. Thus, these secondary mutants were expected to
carry one copy of wild-type ILV2 gene while the SM resistant gene was inactivated.
Secondary mutants were selected by screening for slow growing colonies on medium
lacking of valine and isoleucine. Mating of these secondary mutants resulted in
allotetraploid brewing yeast strains which had a lower diacetyl production and
acceptable brewing characteristics (Kiellandt-Brandt et al., 1995). Another method to
reduce acetolactate synthase activity involved the exploitation of ILV2 anti-sense
mRNA (Vakeria et al., 1991). In this case, lower diacetyl production was obtained but
the resulting transformant was not able to ferment wort well. Recently, some authors
combined the disruption of the ILV2 gene with integration of either AMY, LSD1,
FLONS genes into the ILV2 locus (Liu et al., 2004; Wang et al., 2008; Zhang et al.,
15
2008). Apart from producing less diacetyl, the resulting brewing yeast mutants
conferred other beneficial brewing phenotypes: capability to utilise starch (AMY),
capability to utilise dextran T-70 (LSD1) and controllable NewFlo flocculation property
(FLONS), respectively.
Diacetyl reduction can also be achieved by increasing the flux towards the
formation of valine. To this end, overexpression of ILV3 encoding dihydroxyacid
reductase and ILV5 encoding reductoisomerase was performed (Goossens et al.,
1987; Villanueba et al., 1990; Goossens et al., 1993; Mithieux and Weiss, 1995).
Both approacheas led to the enhancement in activity of their corresponding enzymes,
however, only an increase in activity of ILV5 encoding enzyme resulted in a reduction
of diacetyl level. A decrease in vicinal diketone concentration of up to 60% was
obtained in brewers’ yeast strains with overexpression of ILV5 encoding
enzyme (Villanueba et al., 1990; Goossens et al., 1993; Mithieux and Weiss, 1995).
It is reported that Ilv5p has a high turnover in mitochondria. As Ilv5p is responsible
for the valine biosynthesis and mitochondrial DNA maintenance, the overexpression
of Ilv5p might therefore cause abnormal situation relating to stoichiometry of
mitochondrial DNA and nucleoid which is might be undesirable from a view point of
the quality of brewers’ yeast strain (MacAlpine et al., 2000). Besides that,
manipulation of a certain metabolic pathway can result in unwanted change in the
organoleptic properties of beer. To avoid this problem, Omura (2008) aimed to
overexpress a functional cytosolic Ilv5p enzyme. For this purpose, Ilv5p mutants with
different N-terminal truncations were generated. Among those, the mutant which had
46 residues deleted (Ilv5p-∆46), was found to stably function solely in the cytosol but
was not present in the mitochondria. Overexpression of the Ilv5p-∆46 in a lager yeast
strain resulted in the same reduction of VDK production as the overexpression of
wild-type Ilv5p using a constitutive promoter. Moreover, cytosolic Ilv5p-∆46
overexpression did not alter the production of aromatic compounds and organic acids
important for organoleptic properties of beer. In contrast, there existed an alteration
of production of some organic acids (pyruvate, acetate), fusel alcohols (amyl
16
alcohols, isobutyl alcohol) and acetate ester (isoamyl acetate) in the case of wild-type
Ilv5p overexpression.
2.2.2
Increased production of acetate esters
Besides ethanol and carbon dioxide, during the fermentation, brewers’ yeast
produces other organoleptic compounds which define beer flavour and aroma. The
largest group of those compounds is fusel alcohols and their acetate esters. They are
intermediates of branched-chain amino acid pathways i.e. valine, isoleucine and
leucine. Among those, isoamyl acetate is the most important component which
imparts distinct banana and pear flavour to beer. It is a by-product of leucine
biosynthesis and is formed from the esterification of acetyl-coA and isoamyl alcohol
by catalysis of the ATF1 and ATF2 encoded alcohol acetyltransferases. Isoamyl
alcohol is formed from a-ketoisocaproate in two enzymatic steps.
Alpha-ketoisocaproate is the intermediate of the leucine biosynthetic pathway which
is formed from a-ketoisovalerate in three enzymatic steps. The enzyme responsible
for the first of the three steps is α-isopropylmalate synthase which catalyses the
conversion of a-ketoisovalerate to α-isopropylmalate. Thus, one strategy to increase
isoamyl acetate production was based on the alteration of the activity of
α-isopropylmalate synthase. This enzyme is encoded by the LEU4 gene and is
feedback-inhibited by leucine. Overexpression of the LEU4 gene in a sake
S. cerevisiae yeast resulted in a slight increase in the amount of isoamyl alcohol and
its acetate esters (Hirata et al., 1992). In addition, it was revealed that the LEU4
encoding enzyme was strikingly insensitive to leucine inhibition in the mutant which is
resistant to one toxic analogue of leucine (Santyanarayana et al., 1968). This
strategy was applied to achieve a bottom fermenting strain which produced a higher
amount of isoamyl alcohol and its corresponding acetate ester (Lee et al., 1995). In
addition, α-ketoisocaproate is formed by the degradation of leucine via Ehrlich
pathway. Thus, the increase in leucin uptake could result in an increase in
α-ketoisocaproate level and isoamyl acetate levels. It was shown that the constitutive
17
expression of BAP2, a gene encoding branched-chain amino acid permease in
brewers’ yeast showed an increase in α-ketoisocaproate and in isoamyl alcohol
levels (Kodama et al., 2001b).
Overexpression of either ATF1 or ATF2 genes in brewers’ yeast resulted in a
great enhancement of isoamyl acetate production (Fujii et al., 1994; Nagasawa et al.,
1998; Verstrepen et al., 2003a; Verstrepen et al., 2003b). In addition, increases of
other acetate esters such as ethyl acetate, isobutyl acetate, pentyl acetate, hexyl
acetate and octyl acetate were also obsevered. As lager brewers’ yeast is the hybrid
of S. cerevisiae and another non-Saccharomyces yeast, it contains two diverged
versions of many genes in its genome, namely S. cerevisiae type gene (Sc-type) and
non-Saccharomyces type gene (non-Sc-type). It has been demonstrated that
overexpression of different alleles of ATF genes e.g. Sc-ATF1, Sc-ATF2,
non-Sc-ATF1 led to different impacts on rate of ester production (Fujii et al., 1994;
Verstrepen et al., 2003b). Based on this knowledge, Verstrepen (2003b) suggested
that different aroma patterns produced by different brewers’ yeast strains might result
from various mutations of their ATF genes.
2.2.3
Increase of sulphite production
Sulphite plays an important role in beer flavour stabilization. As an antioxidant,
sulphite prevents oxidative reactions that may occur during post-fermentation
processes, thereby helping to increase beer’s shelf life. In addition, it stabilizes beer
flavour by trapping undesirable carbonyl compounds. These complexes of carbonyl-
sulphite have much higher taste threshold compared to free carbonyls. Sulphite is an
intermediate of the reductive sulphate assimilation which is significant for the
biosynthesis of the sulfur-containing amino acids i.e. methionine and cystein (Fig. 2).
Sulphite, however, is usually produced in yeast at low levels. For the improvement of
beer flavour stability, several efforts have concentrated on increasing sulphite
production in brewers’ yeast.
18
One approach was based on prevention of sulphite (S0
2
) reduction to hydrogen
sulphide (H
2
S). In yeast, sulphite is reduced to sulphide by the activity of sulphite
reductase. This enzyme is a heterogeneous tetramer which is composed of two
subunits α and β, encoded by the genes MET10 and MET5 respectively (Fig. 2).
Hansen and Kielland-Brandt (1996a) eliminated sulphite reductase activity in a lager
brewers’ yeast strain by disrupting all MET10 alleles. The resulting mutant showed a
striking enhancement of sulphite production. Moreover, this study also succeeded in
eliminating hydrogen sulphide, an unwanted by-product causing a rotten-egg flavour
to beer.
Sulphate
APS
MET3
PAPS
MET14
Sulphite (SO
2
)
MET16
Sulphide (H
2
S)
MET5,10
SUL1,2
Aspartate
Homoserine O-acetyl
Homoserine (OAH)
MET25
MET2
Homocysteine Methionine
SAM
SAH
Cysteine
Gluthaonine
Threonine
Isoleucine
Sulphate
Sulphite
STR4 STR3
STR1 STR2
SUU1
HOM3
Sulphate
APS
MET3
PAPS
MET14
Sulphite (SO
2
)
MET16
Sulphide (H
2
S)
MET5,10
SUL1,2
Aspartate
Homoserine O-acetyl
Homoserine (OAH)
MET25
MET2
Homocysteine Methionine
SAM
SAH
Cysteine
Gluthaonine
Threonine
Isoleucine
Sulphate
Sulphite
STR4 STR3
STR1 STR2
SUU1
HOM3
Fig. 2 Modifications of sulphate assimilation and sulphur-containing amino acid biosynthetic
pathway to control sulphite and sulphide production. APS: adenosyl phosphosulphate, PAPS:
phosphoadenosyl phosphosulphate. SAM: S-adenosyl homomethionine, SAH: S-adenosyl
homocystein. Big arrows indicate overexpression of correspondent enzymes. Interrupted arrows
signify decrease of enzyme activity.
19
Another approach to increase sulphite production in brewers’ yeast was based on
enhancing the flux towards from sulphate to sulphite (Fig. 2). To this end, MET3 and
MET14 encoding ATP sufurylase and APS kinase respectively were overexpressed
in brewers’ yeast from multi-copy plasmids. It was reported that overexpression of
MET14 had the highest impact on sulphite production and it even led to an increase
in sulphite production in a met5 mutant (Korch et al., 1991). Expression of MET14
under the control of the strong promoter TIP1 in one sulphite reductase deficient
S. cerevisiae strain also resulted in an increase in sulphite production. Moreover,
Donalies et al. increased the production of sulphite 10-fold in a S. cerevisiae strain by
combining the enhancement MET14 encoded enzyme activity with the
overexpression of SSU1, the gene encoding a sulphite efflux pump (Donalies and
Stahl, 2002).
Repression of the transcription of MET genes is mediated by cysteine. MET2
encodes L-homoserine-O-acetyltransferase which catalyzes the conversion of
homoserine to O-acetyl homoserine (OAH) (Fig. 2). Disruption of MET2 in brewers’
yeast led to the shortage of OAH and consequently to the prevention of cysteine
formation. In this way, genes participating in sulphate assimilation were depressed,
leading to an increase in sulphite production. An enhancement in hydrogen sulphide
production, however, was also observed (Hansen and Kielland-Brandt, 1996b).
It has been reported that sometimes a brewers’ yeast strain with low sulphite
production is desirable for beer brewing (Hansen and Kielland-Brandt, 2003).
It derives from the fact that due to being early accumulated; sulphite will form
complexes with carbonyl compounds and therefore prevent them from being reduced
to their corresponding alcohols. Consequently, the flavour will be negatively affected
(Dufour, 1991). By inactivating 4 copies of MET14 gene in one brewers’ yeast strain,
Johanesen et al. proved that sulphite production during main fermentation resulted in
an increase of acetaldehyde in beer (Johannesen et al., 1999). A brewers’ yeast
strain with a late formation of sulphite is therefore necessary for beer brewing in
preventing the accumulation of acetaldehyde. This demand can be afforded by the
20
utilisation of HSP26 and HPS30 promoters which allowed gene expression at the end
of the exponential phase or stationary phase, respectively. It was shown that
overexpression of the MET14 gene under control of HPS26 promoter led to a
delayed increase in sulphite production (Donalies and Stahl, 2002).
Compared to baker’s yeast, lager brewers’ yeast produces higher amounts of
sulphite (SO
2
) and hydrogen sulphide (H
2
S). In a recent study, Yoshida et al. (2008),
via using integrated metabolome and transcriptome analyses, found out the genetic
basis for these differences. The higher amounts of SO
2
and H
2
S produced by lager
brewers’ yeast than baker’s yeast were due to the limiting amount of OAH. The study
also revealed that the flux from aspartate to OAH had a greater effect on production
of H
2
S than sulphite (SO
2
) (Fig. 2). In contrast, the flux from sulphate to SO
2
had a
greater effect on SO
2
production than H
2
S production. With the aim of increasing
sulphite production, Yoshida created a prototype brewers’ yeast strain by
simultaneously increasing the flux from aspartate to OAH (Sc-HOM3 overexpression)
and the flux from sulphate to sulphite (MET14 overexpression). The resulting mutant
showed a higher level of sulphite and lower level of hydrogen sulphide than the
parental strain.
2.2.4
Elimination of sulphide compounds
Hydrogen sulphide is an unwanted by-product causing a rotten-egg flavour to
beer. It is generated via sulphate assimilation or degradation of sulphur containing
amino acids when nitrogen is depleted. As its taste threshold flavour is low, small
amounts of hydrogen sulphide cause an organoleptic problem in beer. Different
strategies to reduce hydrogen sulphide formation in brewers’ yeast were developed.
As previously mentioned, the deletion of all MET10 genes in brewers’ yeast led to the
inactivation of sulphite reductase. This resulted in a strong abolishment of H
2
S
formation and an accumulation of sulphite in the mutant strain (see I.2.2.3) (Hansen
and Kielland-Brandt, 1996a). Moreover, hydrogen sulphide can also be reduced by
21
the overexpression of both Sc-HOM3 and MET14 (see I.2.2.3) (Yoshida et al., 2008).
Besides the reduction in H
2
S production, the resulting mutant also showed a higher
sulphite production than the wild-type strain.
Other approaches to eliminate H
2
S involved the orientation of H
2
S into the flux of
sulphur containing amino acid biosynthesis. The expression of MET25 under the
control of a constitutive promoter in brewers’ yeast gave rise to a several fold
enhancement of homocystein synthase (Met25p) activity. In the pilot-scale
fermentation, the resulting mutant showed approximately 10-fold decrease in
sulphide production (Omura and Shibano, 1995). In another study, overexpression of
STR4 encoding cysthaonine β-synthase in bottom fermenting yeasts resulted in the
suppression of sulphide formation (Tezuka et al., 1992). The suppression of H
2
S
production was partly due to an increased requirement for homocysteine when STR4
was overexpressed. In addition, it was also explained by the authors that STR4
overexpression led to an increased amount of intracellular cysteine, thus causing to
an increased repression of sulphur assimilation genes, comprising those responsible
for H
2
S formation (Hansen and Kielland-Brandt, 1994). Moreover, hydrogen sulphide
can also be reduced by the overexpression of both Sc-HOM3 and MET14 as
indicated in section I.2.2.3 (Yoshida et al., 2008). Besides the reduction in H
2
S, the
resulting mutant also showed a higher sulphite production than the wildtype strain.
Dimethyl sulphide (DMS) is also a compound affecting organoleptic beer
characteristics, especially in the case of lager beer. The presence of DMS causes an
unwanted corn-like smell and flavour in beer. It is formed both in the wort boiling
stage by thermal degradation of S-methyl methionine and during fermentation by
reduction of dimethyl sulfoxide (DMSO). Hansen et al. (1999) revealed that disruption
of MXR1, a gene encoding methionine sulfoxide reductase led to the incapability of
DMSO reduction in a laboratory yeast strain. Based on that fact, a brewers’ yeast
strain producing lower level of DMS was obtained by disrupting the MXR1 gene
(Hansen et al., 2002).
22
2.3 Alteration of flocculation behaviour
Yeast flocculation is a common phenomenon in beer brewing which involves the
spontaneous asexual aggregation of yeast cells into flocs and their subsequent
removal from fermentation medium by sedimentation (lager yeast) or floating to the
surface (ale yeast) of the fermentation tank. Flocculation is beneficial to beer brewing
since it provides an easy and effective way to separate yeast cells from green beer.
Yeast strains with good flocculation behaviour are of great concern to the brewers.
Besides being strongly aggregated, an ideal yeast strain must flocculate at the proper
time of fermentation (Dequin, 2001). An early flocculation leads to unfinished
fermentation and subsequently to abnormal flavour and aroma of beer. On the other
hand, late flocculation results in cloudy beer due to incomplete separation of yeast
cells. In general, yeast strain with strong flocculation behaviour towards the end of
the main fermentation is necessary for the production of aromatic sufficient and
apparent beer.
Even though the exact mechanism of flocculation is still unclear, it is generally
accepted that flocculation results from the interaction between lectin-like proteins
(flocculins) of a cell with the sugar residues of adjacent cells (Stratford, 1989;
Stratford, 1992). Flocculation is inhibited by the presence of free saccharide
molecules in the medium, seemingly because these free sugars competitively
interact with flocculins, thereby preventing them from interacting with the sugar
residuals on cell surface. Depending on the types of sugar affecting inhibition,
different flocculation phenotypes have been described i.e.: i) the Flo1 phenotype
where flocculation is inhibited by mannose but not by other sugars like glucose,
maltose, sucrose, galactose (Stratford and Assinder, 1991), ii) the NewFlo phenotype
in which flocculation is inhibited by mannose and other sugars like glucose or
maltose (Stratford and Assinder, 1991) and iii) the third flocculation phenotype which
is not inhibited by mannose (Masy et al., 1992; Dengis and Rouxhet, 1997).
It is also reported that flocculation is controlled by genetic factors. Flocculins are
encoded by flocculation genes. Sequence analysis revealed that flocculation genes
23
belonged to a multi-gene family which localized to the telomeric regions (Teunissen
and Steensma, 1995). Among those, two dominant flocculation genes conferring the
Flo1 phenotype in yeast are FLO1 and FLO5 ( Johnston and Reader, 1982; Johnston
and H. P, 1983). Besides that, the flocculation gene family includes the genes FLO9
and FLO10 which are 94% and 58% homologous to FLO1, respectively (Teunissen
and Steensma, 1995; Sieiro et al., 1997). In lager brewers’ yeast, another homologue
of FLO1 gene named Lg-FLO1 was identified. This gene is not present in ale yeast
strains and is responsible for the NewFlo phenotype of most lager brewers’ yeast
strains (Kobayashi et al., 1995; Kobayashi et al., 1998; Sato et al., 2002). In addition
to these structural genes, flocculation gene family also includes FLO8, which
encodes a transcription activator required for regulation of flocculation as well as
other phenotypes such as diploid filamentous growth, and haploid invasive
growth (Liu et al., 1996; Kobayashi et al., 1996; Kobayashi et al., 1999; Pan and
Heitman, 1999). Another dominant flocculation gene is FLO11 which encodes a
Flo1-type flocculin (Lo and Dranginis, 1996; Bayly et al., 2005). FLO11 is distinct
from other flocculation genes in that it locates near the centromere instead of the
telomere (Lo and Dranginis, 1996). Besides being regulated by Flo8p, the expression
of FLO11 is also controlled by other factors such as mating type.
Several attempts have been made in altering the flocculation behaviour of
brewers’ yeast. The protoplast fusion of one non-flocculent brewers’ yeast strain with
a flocculent S. cerevisiae strain resulted in the formation of a flocculent yeast strain
which could be used for beer brewing (Urano et al., 1993). Other efforts to control the
flocculation in brewers’ yeast involved the manipulation of flocculation genes. The
constitutive expression of the FLO1 gene using the ADH1 promoter resulted in a
strong flocculation phenotype in one non-flocculent brewers’ yeast strain (Watari et
al., 1994). Nevertheless, this strain was not suitable for beer brewing since the onset
of flocculation occurred too early, thus leading to the incomplete fermentation.
Identification of promoters which can precisely control the gene expression under
24
industrial fermentation conditions is therefore necessary for the improvement of the
flocculation behaviour in yeast.
The promoter HSP30 is induced during the entry of yeast cells into stationary
phase of the fermentation (Riou et al., 1997). In the S. cerevisiae FY32 strain, a
mutation of FLO8 gene led to the inactivation of transcription of the FLO genes and
thus to the non-flocculent phenotype in this strain (Winston et al., 1995; Liu et al.,
1996; Verstrepen et al., 2005). For the alteration of flocculation phenotype in this
strain, the wild-type FLO1 promoter was replaced by HSP30 promoter (Verstrepen et
al., 2001). The resulting transformant showed a strong flocculation toward the end of
the laboratory fermentation. Besides that, ADH2 promoter, which is repressed during
the growth on glucose (Price et al., 1990; Gancedo, 1998) and is derepressed with
transition on the growth on ethanol (Noronha et al., 1998), is a possible system to
alter the flocculation in yeast. Govender et al. (2008) placed the dominant flocculation
genes FLO1, FLO5 and FLO11 of the S. cerevisiae FY32 strain under the
transcriptional control of either ADH2 or HSP30 promoters. It was shown that the six
gene-promoter combinations resulted in specific flocculation phenotypes in terms of
timing and intensity. The results suggested that the flocculation behaviour of brewers’
yeast could be improved by fine-tuning the expression of dominant flocculation
genes.
25
3 “Omics” technologies in studies regarding brewers’ yeast
The progress in DNA recombinant technology has enabled the improvement of
yeast strains via metabolic engineering. In metabolic engineering, one crucial factor
for the determination of targets and approaches for strain improvement is the
understanding of how a phenotype is determined by a genotype (Attfield and Bell,
2003). This knowledge has been accumulated mostly via reductionistic approaches
in which the linkage between the genotype and the phenotype has been identified via
the modification of a particular gene (Bro and Nielsen, 2004). However, the fact that
a phenotype can be defined by multiple genes or the modification of a single gene
may lead to pleiotrophic effects can make it difficult to discover this
genotype-phenotype relationship.
The accessibility of complete genome sequences of several organisms has
facilitated the development of the so-called “omics” technologies mainly genomics,
transcriptomic and proteomics. The development of “omics” technologies in turn has
allowed the studies of cellular activities on a global scale and thus has provided an
insight into the cellular responses to genetic alterations or environmental changes.
The availability of the complete genome sequence of S. cerevisiae has laid the
foundation for the utilisation of “omics” technologies in yeast studies i.e. the
construction of DNA microarrays for global genome and transcriptome analyses and
the protein database for proteome studies and thus has led to an accumulation of the
huge knowledge about cellular activities. The accumulation of the knowledge is
highly advantegous for strain improvement.
As previously mentioned, lager brewers’ yeast strains are aneuploid hybrids
between S. cerevisiae and another Saccharomyces yeast. As the genome sequence
of brewers’ yeast has not been publicly accessible, the application of “omics”
technologies in the brewers’ yeast studies has been performed mostly by exploiting
the current knowledge of S. cerevisiae genome sequence. In this section, I review
26
some recent applications of genomics, transcriptomics and proteomics technologies
in studies regarding brewers’ yeast.
3.1 Genomics
Microarray-based comparative genomic hybridisation (CGH) is a useful tool for
global scale genome analysis. Microarray-based CGH has been used successfully in
detecting gene deletions, quantification of gene copy numbers and giving information
on the chromosomal aneuploidies as well as the translocations at genomic scale
(Winzeler et al., 1999; Daran-Lapujade et al., 2003). Identification of genomic
differences between strains showing different degrees of a certain phenotype can
directly reveal relevant target genes for strain improvement as long as they can linke
to this phenotype. In brewers’ yeast, DNA microarray-based hybridisation was used
to study the complexity of lager brewers’ yeast genome. By hybridising total genomic
DNA of two lager brewers’ yeast strains to the S. cerevsiae array, Bond et al. (2004)
detected conserved and discrete translocation regions in the genomes of the studied
lager brewers’ yeast strains. In addition, large regions of S. cerevisiae genome were
found to be absent in lager brewers’ yeast. The study provided more evidence about
the aneuploid nature of lager brewers’ yeast as well as the diversity of genome
composition between different lager brewers’ yeast strains.
Pope et al. (2007) used different genomic fingerprinting approaches to
discriminate different lager, ale and S. cerevisiae strains. Among the genomic
fingerprinting methods, amplified fragment length polymorphism providing a snapshot
of DNA sequences across the whole genome resulted in relatively good
discrimination of the studied strains. In contrast, the array-based GCH by means of
S. cerevisiae microarray failed in providing meaningful differentiation of studied
strains. The unsatifying result obtained by using microarray-based CGH had
supposedly been caused by the lack of the non-S. cerevisiae component in the
analysis.
27
Recently, a two-species microarray has been developed based on the genome
sequence of the strain S. cerevisiae S.288C and contig sequences of one
S. bayanus var. uvarum strain (CBS 7001 type strain) (Dunn and Sherlock, 2008).
The analysis of 17 different lager brewers’ yeast strains using this two-species
microarray revealed the presence of two genomically distinct groups of lager brewers’
yeast strains which correlated to specific breweries and geographical regions. The
first identified group lack a significant portion of the S. cerevisiae genome but retains
all of the S. bayanus genome. The second group retains nearly all of the genomic
content of the both genomes. The 1
st
group represents Saarz type beer and
Carlsberg brewery strains while the 2
nd
group contains strains from the Netherlands,
non-Carlsberg Danish breweries and two North American breweries. The analysis
also revealed some consistent break points or regions of amplifications or deletions
in the studied strains and thus presumably included genes of selective importance in
brewing conditions
3.2 Transcriptomics
DNA microarray is also a powerful tool for the global-scale transcriptome analysis.
Microarray-based transcriptome analyis enables the examination of abundance of all
transcripts in the cell at a given state or condition and thus allows the identification of
genes which are co-regulated as well as the analysis of global responses to genomic
mutations. Furthermore, by comparing the transcriptional profiles of one strain in
different conditions or between various strains showing different phenotypes, genetic
basis relevant to these differences can be revealed (Pandey et al., 2007). Based on
that, target genes for strain improvement can be identified.
DNA microaray has been applied to study transcriptional profiles of brewers’ yeast
during fermentation. So far, these studies have been performed by the means of the
S. cerevisiae array. Olesen et al. (2002) studied the dynamics of one lager brewers’
yeast transcriptome at different points of time during a pilot-scale brewery
28
fermentation. The analysis revealed that the average gene expression increased
rapidly and reached a maximum value after two days of the main fermentation. The
average expression was then declined as sugar was consumed. Those genes with
high average expression value were mostly genes involved in protein synthesis,
glycolysis and lipid synthesis while a large number of genes with unknown biological
functions showed a low average expression. Another study using the S. cerevisiae
microarray to study the transcriptomes of two lager brewers’ yeast strains at different
points of time during a small-scale brewery fermentation revealed a high level of
expression of ORFs involved in fatty acid and ergosterol biosynthesis early in
fermentation (James et al., 2003). Genes involved in respiration and mitochondrial
protein synthesis also showed a high level of expression early in the fermentation.
Furthermore, a near complete repression of many stress response genes and gene
involved in protein biosynthesis was observed at the end of fermentation compared
to that at the start of fermentation.
3.3 Proteomics
Like transcriptomics, proteomics reveals the global response of gene expression
to environmental and genetic changes. Nevertheless, compared to transcriptome
analysis, proteomics brings us one level closer to the phenotype (Bro and Nielsen,
2004) by disclosing the questions related to gene functions such as the mRNA
translation efficiency, protein translation modification or protein stability. The standard
method for quantitative protein analysis involves the protein separation by
two-dimensional gel electrophoresis (2D gel electrophoresis) and mass spectrometry
(MS) or tandem MS (MS/MS) identification of protein spots (Gygi et al., 2000).
Despite its potential in study of gene function, the disadvantage in proteome analysis
is caused by the standard method used for proteomic study. 2D gel electrophoresis is
considered more laborious, less sensitive and less reproducible than DNA
microarray, the common method used in global-scale transcriptome study. However,
29
the optimisation of the 2D gel electrophoresis and the development of new methods
for global quantification of proteins based on mass spectrometry will bring about
more perspectives for global-scale proteome studies (Aebersold and Mann, 2003).
Proteome analysis using 2D gel electrophoresis has been used to define the
relatedness between different kinds of brewers’ yeast as well as between brewers’
yeast and other Saccharomyces yeasts (Joubert et al., 2000; Kobi et al., 2004). The
first proteome maps of lager brewers’ yeast and ale brewers’ yeast were respectively
presented by Joubert et al. in 2000 and Kobi et al. in 2004. Comparison of the
proteomes of one ale yeast strain, one lager yeast strain and the S. cerevisiae strain
S288C confirmed that the ale yeast strain is much more closely related to
S. cerevisiae than the lager yeast strain at proteomic level (Kobi et al., 2004). In
agreement with the hypothesis that lager brewers’ yeast is a hybrid of at least two
different Saccharomyces yeasts, the proteome of lager brewers’ yeast appeared to
be the superimposition of two elementary patterns, one corresponded to
S. cerevisiae proteins and the other was best represented by one S. pastorianus
strain (Joubert et al., 2000).
In addition, through the use of 2D gel electrophoresis and differential gel exposure
in which the proteomes of S. cerevisiae strain S288C and one lager brewers’ yeast
strain were labelled with different isotopes and then being separated in one 2D gel,
Joubert et al. (2000) discovered that a large percentage of S. cerevisiae proteins
(83%) was co-migrated with lager brewers’ yeast proteins. In contrast, the
co-migration of proteins of either S. bayanus or S. uvarum type strains with this lager
yeast strain was markedly lower, only 35% and 37%, respectively. Furthermore, the
proteome analysis using 2D gel electrophoresis, MS, MS/MS and S. cerevisiae
database searching allowed the identification of many novel non-S. cerevisiae
proteins of lager brewers’ yeast (Joubert et al., 2001). These newly identified proteins
of lager brewers’ yeast corresponded to the protein spots that did not co-migrate with
known proteins of S. cerevisiae separated on the 2D gels.
30
In addition to the study of strain relatedness, proteomics was applied in studying
brewers’ yeast gene expression at different stages of fermentation and as well as at
different generations of successive fermentations (Kobi et al., 2004). The proteome
of one ale yeast strain during production-scale fermentation was studied at the
beginning and at the end of the first and the third usage of the yeast (the 1
st
and 3
rd
generation of successive fermentations). It was shown that the most pronounced
changes in protein expression occurred in the 1
st
generation, during the switch from
aerobic propagation to anaerobic fermentation. Even though yeast propagation was
performed in saccharose medium before inoculated in brewing wort, no drastic
protein change directly related to the change in sugar source (from saccharose
medium to wort) was observed. The variation in protein expression in the 3
rd
generation was much lower in comparison to the 1
st
generation. Unsurprisingly, no
difference in protein expression related to the switch from aerobic to anaerobic
condition in the 3
rd
generation was observed. However, certain stress response
proteins i.e. Hsp26p, Ssa4p, Pnc1p were induced during first generation and
constitutively expressed in the subsequent generations. These are stress-response
proteins induced by variety of treatments. The induction of these stress-response
proteins during the first fermentation suggested that the switch from oxidative to
fermentative condition was an environmental stress to yeast cells (Kobi et al., 2004).
In addition, the authors explained that the constituve expression of these stress
response proteins in subsequent fermentations was probably important for the
maintance of viability of the yeast cells which encountered stressful fermentative
conditions.
There has not yet been any study regarding the dynamics of lager brewers’ yeast
proteome during fermentation. However, proteome analysis was used to identified
the proteins which are induced during the lag and early exponential phase
(early-induced proteins) in glucose-containing medium (Brejning et al., 2005). After
5 h of inoculation, several proteins were identified as early-induced including Ade17p,
Eno2p, Ilv5gp, Sam1p, Rsp21 and Ssa2p. The induction of most of these proteins did
31
not match the transcriptional expression of genes in the glucose-containing medium.
Nevertheless, under brewing conditions, the transcriptional expression of these
genes, except for ENO2 and SSA2, were strongly induced early in the lag phase
(Brejning et al., 2005). The monitor of these early-induced genes and proteins could
be useful in creating physiological markers for optimisation and control of growth
initiation during brewing fermentation.
4 Lager brewers’ yeast genome sequence and its perspective in
brewers’ yeast global studies
The application of “omics” technologies based on current knowledge of
S. cerevisie genome sequence in brewers’ yeast studies has certain limitations,
especially regarding the lager brewers’ yeast studies. As aforementioned, the
microarray-based global genome analysis failed to discriminate different lager
brewers’ yeast strains due to the lack of non-S. cerevisiae sequences in the
microarray (Pope et al., 2007). The usage of the two-species array composed of
S. cerevisiae and S. bayanus var. uvarum sequences in global genome analysis
provided a better opportunity to precisely differentiate lager brewers’ yeast strains.
However, it was estimated that the S. bayanus sequence which contributed to the
lager brewers’ yeast genome was about 10% divergent to the sequence of the
S. bayanus var. uvarum strain (Dunn and Sherlock, 2008), thus the exploitation of
this two-species microarray could not fully evaluate the genotype of lager brewers’
yeast. In addition, the transcriptome analysis of lager brewers’ yeast using
S. cerevisiae arrays only revealed the expression pattern of half of the genome while
the other half is uncovered. The proteome studies have also been obstructed due to
the lack of the non-S. cerevisiae sequences. Non-S. cerevisiae proteins cannot be
unambiguously identified by using the common method of peptide fingerpringting
(MALDI-TOF MS) and database searching with S. cerevisiae sequence. The
identification of non-S. cerevisiae proteins thus requires more time-consuming
32
methods such as tandem MS or nano-electrospray tandem MS/MS (Joubert et al.,
2001; Brejning et al., 2005) and has to be based on sequence homologies.
Recently, the genome of one lager brewers’ yeast strain i.e WH 34/70 was
sequenced (Kodama et al., 2006). The total size of the lager brewers’ yeast genome
was 23.2 million bp, approximately twice the size of the S. cerevisiae laboratory yeast
genome. The contig sequences of the lager brewers’ yeast genome are divided in
two groups: i) Sc-type with more than 98% to homomology to S. cerevisiae
sequences and ii) non-Sc-type with identity around 85% identical to S. cerevisiae
sequences. The sequencing project also confirmed the hybrid nature of lager
brewers’ yeast with the presence of three kinds of chromosomes: Sc-type,
non-Sc-type and various chimeral types.
Although the sequence of the lager brewers’ yeast genome has not been publicly
available, production of bottom fermenting yeast microarray based on this sequence
has been announced (Nakao et al., 2008). In general, the bottom fermenting yeast
DNA microarray contains 22,977 probesets representing 22,483 regions from the
whole genome sequence information of the lager brewing yeast strain 34/70,
403 S. cerevisiae ORFs which are not identified in the WH 34/70 strain,
64 control genes and 27 S. pastorianus ORFs submitted in Genbank
(http://www.ncbi.nlm.nih.gov/Genbank/). The 22,483 regions from strain WH 34/70
composed of 7640 Sc-type ORFs, 6307 non-Sc-type ORFs, 28 mitochondrial ORFs,
7955 intergenic regions and 553 other sequences which showed a similarity to
S. cerevisiae proteins by NCBIblastX homology searching (Nakao et al., 2008). The
availability of the bottom fermenting yeast DNA microarray will allow more reliable
global transcriptome and genome analyses of lager brewers’ yeast strains while the
accessibility of the bottom fermenting yeast DNA sequence will bring about more
convenience for proteomic studies. These advancements will give rise to an outburst
knowlege about brewers’ yeast physiology in the near future and is thus highly
advantageous for the improvement of brewers’ yeast strain.
33
5 Conclusions
The progress of genetic engineering has led to the creation of numerous novel
brewers’ yeast strains highly beneficial for brewing industry. Many of these strains
were constructed according to requirements for commercial production, e.g the
generation of “self-cloned” strains which contains no heterologous gene and no
addition of selectable marker. However, the use of recombinant strains in beer
production has not been worldwide approved. So far, there is only one brewers’ yeast
strain received official approval for commercial use from British government. Even so,
it has not yet been used commercially. The limited public acceptance for genetically
modified organisms (GMOs) is clearly the obstacle for the development of novel
brewers’ yeast strains. This phenomenon is derived from the concern of consumers
about the danger of genetically modified (GM) foods and beverages. In the future, a
better communication between scientists and consumers as well as between
scientists and legislators should be established to improve the community’s
knowledge about the low risk and high benefit that certain GMOs can bring about.
This will be the important requirement for application of biotechnology in food and
beverage industry in general and in brewing industry in particular.
The application of global molecular methods i.e. genomic, transcriptomic and
proteomics has already enabled the accumulation of some knowledge about cellular
activities of brewers’ yeast during fermentation. So far, these global analyses have
been mostly performed based on current knowledge about S. cerevisiae genome
sequence and thus have certain limitations in studying lager brewers’ yeast, the
hybrid between S. cerevisiae and another Saccharomyces yeast. Nevertheless,
recent achievements regarding the complete sequence of bottom fermenting yeast
sequence and the bottom fermenting yeast DNA microarray will strongly facilitate the
enlargement of the basic knowledge and the improvement of brewers’ yeast strain.
34
II Experimental part
An integrative approach to identify novel target genes
for reduction of diacetyl production in lager yeast
1 Introduction
1.1 The need to optimise brewers’ yeast
Although beer brewing is a well-established traditional process, the current
brewers’ yeast strains are far from optimised for beer production as stated by
Hammond (1995). For example, beer has to undergo a secondary fermentation
process which lasts about two weeks to remove diacetyl, an off-flavour in beer. As
this process requires a lot of time and capacity, it would be a great benefit to have a
yeast strain with a low diacetyl production. For beer brewing, an “ideal” yeast strain
should comprise a number of good features including the ability to consume a wide
range of substrates, the fast fermentation of wort sugars under low temperature, a
good flocculation towards the end of main fermentation, a balanced pattern of
by-products and a low level of diacetyl production. There are brewers’ strains with
ideal features; however, these features are not included in one single strain.
Therefore, a complete understanding of the relationship between phenotype and
genotype is needed to allow the transfer of a good trait from one strain to another
strain.
1.2 Former attempts to improve brewers’ yeast
Due to the demand for brewers’ yeast strains with improved properties, there has
been much research focusing on engineering brewers’ yeast. As lager beer
constitutes 90% of world beer production (Kodama et al., 2006), the object for most
of this research has been the bottom fermenting yeast. In general, genetic
35
improvement of brewers’ yeast strains has been achieved through traditional genetic
manipulations (breeding, traditional mutagenesis and cytocyduction) and rational
metabolic engineering. Application of these approaches in brewers’ yeast
improvement has advantages and drawbacks which will be discussed as follows.
1.2.1
Classical genetic manipulations
The first approach for genetic improvement of yeast strains was mating (also
denoted as breeding or cross hybridisation) of parents owning favourable traits and
selection for progenies with combined desirable phenotypes. The approach produces
high genetic diversity and can be used to combine optimal genotypes and traits in
one strain (Attfield and Bell, 2003). Moreover, it does not relate to the matter of
“Genetically modified organism” (GMO). In general, this approach is applicable to any
yeast strain which can produce a number of viable spores (Attfield and Bell, 2003).
Among industrial yeasts, lager brewers’ yeast is striking in its genetic constitution
of being an aneuploid hybrid between S. cerevisiae and another Saccharomyces
yeast, probably S. bayanus. Besides this, lager brewers’ yeast sporulates poorly and
even if it sporulates, a low number of spores can survive (Hammond, 1995). Due to
this limitation, early studies encountered problem when breeding brewers’ yeasts
(Andersen et al., 2000). Nevertheless, matable spores from brewers’ yeast were
isolated and some novel brewers’ yeast strains with optimised features were created
by cross hybridisation. Gjemansen isolated segregants of one lager brewers’ yeast
strain and obtained viable spores of the two mating types (Gjermansen and
Sigsgaard, 1981). Pairwise cross was performed between spores of opposite mating
types resulting in a number of hybrids. Investigation of these hybrids under brewery
conditions revealed one strain as good as the parental strain. Besides that, the
hybridisation of maters from brewers’ yeast with maters from other yeasts such as
S. cerevisiae led to the formation brewers’ yeast strains which harboured good traits
from the non-brewing parent (Bilinski et al., 1987; Bilinski and Casey, 1989).
36
Still, breeding of brewers’ yeast, in particular lager brewers’ yeast is troublesome
since it is quite laborious and the results cannot be anticipated.
Classical genetic approaches also involve mutagenesis by inducing UV radiation
or alkylating reagents. The approach itself is indirect and the frequency of obtaining a
desirable phenotype is very low. In addition, it always has a high frequency of
generating harmful gene alterations. As it was previously mentioned, bottom
fermenting yeast strains are polyploid or aneuploid. Therefore, if the mutation was
not dominant, it must occur in all alleles of a gene for the alteration of the phenotype
(Attfield and Bell, 2003). Thus, even though this method is possible for creation of
desirable haploid laboratory yeast strains, it is almost not accessible for polyploid
organisms such as brewers’ yeast.
Another strategy of traditional genetic manipulations of brewers’ yeast is
protoplast fusion. This approach allows the hybridisation of individuals without
considering the mating types (Hansen and Kielland-Brandt, 2003). Early studies
using this method led to the creation of hybrid brewers’ yeast strains producing
off-flavours in beer (Attfield and Bell, 2003). Nonetheless, protoplast fusion was
successfully used for improving the flocculation behaviour of one brewers’ yeast
strain. The fusion of protoplast of one flocculent S. cerevisiae strain with a
non flocculent brewers’ yeast led to the formation of a flocculent yeast strain which
could be used for beer brewing (Urano et al., 1993). However, this approach is
unpredictable since the characteristics of the two parents are not averaged. The
protoplast fusion tends to lead to chromosomal dominance of one nucleus from one
parent while the other nucleus is the subordinate which containes the loss of most
nuclear genome (Attfield and Bell, 2003). The desirable trait from parents thus can be
absent in the offspring.
To sum up, application of traditional genetic approaches has gained several
successes for strain improvement of brewers’ yeast. However, they are not direct and
the probability of having the correct combination of desirable traits is low (Attfield and
Bell, 2003).
37
1.2.2
Rational metabolic engineering
The development of recombinant DNA technology has allowed the improvement
of yeast strains by rational metabolic engineering. The approach was defined as “the
improvement of cellular activities by manipulation of enzymatic, transport and
regulatory functions with the use of recombinant DNA technology” (Bailey, 1991).
In general, rational metabolic engineering consists of two important parts:
i) identification of target genes for genetic manipulation and ii) genetic engineering of
the cell for construction of recombinant strain (Ostergaard et al., 2000). The rational
metabolic engineering is distinguished from other classical genetic approaches as it
allows the direct transfer of genetic information via genetic manipulation of the target
genes.
In rational metabolic engineering, the identification of target genes is based on
knowledge about enzymes, pathways and regulatory factors relevant to the
phenotype. From that, a rate-controlling step is identified or a model system is
constructed. The problem is then solved by genetic manipulation of the identified
target genes. Application of this approach has led to a number of successes in strain
improvement of Saccharomyces yeast in general and brewers’ yeast in particular.
Rational metabolic engineering was effectively employed to improve numerous
phenotypes of brewers’ yeast. Concretely, it has been used in the alteration of
by-product formation i.e. the prevention of unwanted substances diacetyl (Sone et
al., 1987; Villanueba et al., 1990) and sulphide compounds (Omura and Shibano,
1995) as well as the enhancement of production of desirable by-products like acetate
esters (Lee et al., 1995) and sulphite (Korch et al., 1991; Hansen and Kielland-
Brandt, 1996a). In addition, rational metabolic engineering has been used to improve
brewers’ yeast utilisation of carbohydrates such as dextrin (Yocum, 1986b; Park et
al., 1990) and maltose (Kodama et al., 1994). Optimisation of flocculation phenotype
in brewers’ yeast was also achieved by rational metabolic engineering (Watari et al.,
1991). Thus, the application of rational metabolic genetic engineering has actually
covered most targets for brewers’ yeast strain improvement.
38
Even though numerous achievements have been made via employing rational
metabolic engineering, many studies failed or were not as successful as expected. In
some cases, the genetically modified strains governed other unfavourable
phenotypes in addition to the desirable traits. For example, the reduction of diacetyl
formation in brewers’ yeast by the complete elimination of acetolactate synthase
activity led to the inability of the mutant strain to produce valine. This led to a nutrient
deficiency in the mutant strain since valine uptake from the extracellular medium was
not sufficient for cellular activities (Gjermansen et al., 1988; Kiellandt-Brandt et al.,
1995). The reason for these failures is likely due to an incomplete understanding of
the complex global metabolic network and its response to genetic alterations (Bailey
et al., 2002; Nevoigt, 2008).
Identification of the target for rational metabolic engineering is based on the
available knowledge regarding the relationship between phenotype and genotype.
Therefore, the more knowledge about the targets and the relevant factors for genetic
manipulation is accumulated, the more successful is the approach altogether.
Regarding this aspect, rational metabolic engineering is less accessible to brewers’
yeast compared to laboratory yeast. As the genome sequence of lager brewers’
yeast has not yet been published, the knowledge used for genetic engineering of
brewers’ yeast so far has been mostly based on studies of S. cerevisiae. However,
the genome of lager brewers’ yeast is strikingly different from the laboratory yeast. As
previously mentioned, the genome of lager brewers’ yeast is comprised of sequences
from both S. cerevisiae and another Saccharomyces yeast, probably S. bayanus
(Vaughan-Martini and Kurztman, 1985; Naumova et al., 2005; Kodama et al., 2006;
Rainieri et al., 2006). In fact, there have been several examples that genetic
engineering of certain target genes led to different results in laboratory and brewers’
yeast backgrounds (Klein et al., 1996). Rational metabolic engineering to improve
brewers’ yeast is therefore limited since the effects of gene homologs from the
non S. cerevisiae origin on the phenotype of lager brewers’ yeast cannot be correctly
predicted.
39
1.3 Inverse metabolic engineering as an alternative approach for improving
brewers’ yeast
To overcome the limitations of rational metabolic engineering, inverse metabolic
engineering can be used an alternative for the improvement of brewers’ yeast. The
concept of inverse metabolic engineering was codified by Bailey et al. in 2002 and
was described as “The elucidation of a metabolic engineering strategy by first,
identifying, constructing, or calculating a desired phenotype; second, determining the
genetic or particular environmental process factors conferring the phenotypes; and
third, endowing that phenotype on another strains or organism by directed
engineering environmental manipulation”.
The selection of the desired phenotype is the first step of inverse metabolic
engineering. This desired phenotype can arise naturally or can be obtained via
appropriate evolutionary engineering (Sauer and Schlattner, 2004). The second step
is the identification of the target genes for genetic modification. It has been carried
out by analysing the molecular basis for the differences between the strains with
desirable phenotypes and the host/production strain, i.e the strain to be modified.
This identification of target genes is considered the most challenging step in inverse
metabolic engineering. However, the availability of methods for genome-wide and
global functional analyses has enabled the screening of differences at various
molecular levels (i.e. genome, transcriptome, proteome, metabolome) and thus has
facilitated the identification of crucial genetic information relevant for a certain
phenotypic trait (Nevoigt, 2008). In the last step, the desirable trait is conferred to the
host strain by genetic engineering.
In contrast to rational metabolic engineering which tries to genetically engineer
the cellular activities based on available knowledge or on a “human-deduced model
system”, inverse metabolic engineering is more advantageous as it is based on a
“living model system” i.e. a desirable phenotype that exists in nature (Sauer and
Schlattner, 2004). Besides that, inverse metabolic engineering comprises many other
benefits compared to the rational metabolic engineering (Nevoigt, 2008). Firstly, it
40
requires no preliminary knowledge about cellular subsystems e.g. enzymes,
pathways and regulation factors relevant to the favourable trait. This is actually
helpful for the improvement of brewers’ yeast strain since the accumulated
knowledge about the cellular activities of brewers’ yeast is still limited (see above).
Another advantage of inverse metabolic engineering is that one can directly use the
industrial strains and industrially relevant conditions to identify the target genes for
genetic modification. Moreover, inverse metabolic engineering can be applied to
combine various valuable traits in one strain. In this model of analysis, one strain with
the 1
st
desirable trait will play the role of the host while the other strain with the 2
nd
favourable trait plays the role of the model strain. By comparison the genotypes and
gene expression patterns of the two studied strains, the genetic basis for the 2
nd
valuable trait will be identified and genetic manipulation of the explored target genes
in the host strain will lead to the creation of a yeast strain with combined favourable
traits. In an approach where the host and the model organism are taxonomically
closely related, e.g. they both belong to one species, genetic modification will then be
based on homologous genes. The modified strains therefore can be considered as
“self-cloned” and can be better accepted in food and beverage industry. Lastly, by
studying naturally diverse strains, inverse metabolic engineering approach provides
good chances to explore novel target genes for strain improvement i.e. those which
have never been found in the rational approach.
Inverse metabolic engineering was effectively employed for yeast strain
optimisation in an increasing number of studies (Nevoigt, 2008). For example, it was
applied to improve the galactose utilisation in S. cerevisiae (Bro et al., 2005). Another
application was to construct a yeast strain with the improvement in xylose uptake and
ethanol production (Jin et al., 2005). Recently, inverse metabolic engineering was
successfully employed to increase sulphite production in a bottom fermenting yeast
strain (Yoshida et al., 2008). It can be said that inverse metabolic engineering, with
its outstanding advantages, is clearly a powerful approach for yeast strain
improvement.
41
2 Aim of work
Diacetyl is a by-product of the valine biosynthetic pathway which causes an
unwanted butter flavour in beer. The reduction of diacetyl production is one of the
most relevant targets for brewer’ yeast strain improvement since it helps to shorten
the maturation process and thereby enabling an acceleration of lager beer production
(see also section I.2.2.1).
The aim of this thesis is to reduce diacetyl production in lager brewers’ yeast by
means of inverse metabolic engineering. In inverse metabolic engineering, the most
important step is the identification of target genes for genetic modification. This task
herein is performed by means of an integrative approach using global analyses at
different genetic molecular levels.
A step by step scenario reflected by the following key questions was designed to
find out the most suitable solution:
1) What is the strategy for brewers’ yeast improvement using inverse metabolic
engineering?
Firstly, lager brewers’ yeast strains producing various levels of diacetyl will be
selected. Secondly, the genetic basis responsible for the phenotypic differences in
the selected strains will be identified. The last step will be the improvement of a
production brewers’ yeast strain via genetic engineering of the target genes.
2) How novel target genes for reducing diacetyl production in lager brewers’ yeast
are identified?
To identify the genetic basis for the strain’s phenotypic differences relevant to
brewing, an integrative approach is chosen using global analyses at different
molecular levels of genome (microarray-based comparative genomic hybridisation),
transcriptome (microarray-based transcriptome analysis) and proteome
(two-dimensional gel electrophoresis, MALDI-TOF MS). For the microarray-based
genome and transcriptome analyses, bottom fermenting yeast DNA microarrays will
42
be used. Incorporating data of these analyses will lead to the identification of novel
target genes for strain improvement.
3) How the roles of the relevant target genes for diacetyl production are verified?
Genetic manipulations of the relevant target genes will be performed in one
production lager yeast strain. Fermentations of the resulting recombinant strains will
be carried out under laboratory and industrially relevant brewery conditions. The
roles of target genes will be verified by measuring diacetyl productions of the wildtype
and the recombinant strains during the main fermentation.
4) How do the genetic modifications affect the fermentation performance of the lager
brewer’ strain?
Fermentation performances of the recombinant strains will be investigated under
industrially relevant brewery conditions. During the fermentation, the time courses of
apparent extract and time couses of pH will be recorded. In addition, the
concentrations of non-sedimented cells in the wort medium will be measured. At the
end of the main fermentations, the green beers produced by the recombinant strains
will be taken for by-product analyses.
43
3 Materials and methods
3.1 Materials
3.1.1
Equipment
Autoclave Varioklav 500 EV (H+P Labortechnik, Oberschleissheim)
Array Bottom fermenting yeast DNA microarray (Affymetrix
customer array)
Array washing system Fluidics Station 450
Balances Type 1907 (Sartorius, Göttingen)
Centrifuges Sorvall RC-5B (for SS34- and GSA-Rotors) (Sorvall DuPont,
Bad Homburg); Microrapid/K (Hettich, Tuttlingen); DW41
(Qualitron, Korea)
Clean bench Uniflow UVUB1200 (UniEquip, Martinsried)
Electrophoresis
chambers
Mini-Sub and Wide Mini-Sub DNA-Cell (Biorad); Ettan
TM
IPGphor
TM
Isoelectric Focusing System; Ettan
TM
Dalt six
Electrophoresis Unit (Amersham Pharmacia Biotech)
Homogenizer Micro-Dismembrator; Braun, Melsungen
Hybridization oven Compact Line OV4 (Biometra, Göttingen)
Incubators Typ B6420 & FunctionLine Typ B20 (Heraeus Instruments,
Hanau); MultiTemp III (Amersham Pharmacia Biotech)
Microscope Labophot (Nikon, Japan)
PCR equipment GeneAmp 9600 (PerkinElmer, Norwalk, USA);
MyCycler
TMthermal cycler
(Biorad)
pH-electrode IntLab 412 (Mettler-Toledo, Urdorf, Schweiz)
pH meter Digital-pH-Meter (Knick)
Pipetting equipment Gilson Pipetman P10, P20, P200, and P1000 (Abimed,
Langenfeld)
Shaker Certomat U (Braun Biotech, Melsungen); Polymax 1040T
(Heidolph, Kelheim)
Scanner Affymetrix Array Scanner 3000, Image Scanner (Amersham)
44
Spectrophotometer UV-160A (Shimadzu, Japan)
Vacuum equipment Laborota 4000 (Heidolph, Kelheim)
Water bath Lauda A100 (Wobser, Lauda-Königshofen)
Other materials Immobiline
TM
DryStrip Gels pH 3-7
3.1.2
Enzymes, chemicals and kits
Agarose Seakem LE, GTG & Gold, Incert
R
Agarose (FMC
Bioproducts, Denmark)
Biochemicals/chemicals Acetylated BSA, Control Oligo B2, DMSO, Eukaryotic
Hybridisation Buffer (Affymetrix); Pharmalyte pH 3-
7,
PlusOne APS, PlusOne CHAPS, PlusOne Dithiothreitol,
PlusOne Drytrip Cover Fluid, PhastGel
TM
Blue R, PlusOne
Glycin, PlusOne SDS, PlusOne Urea, (Amersham
Pharmacia Biotech, Uppsala Sweden); Ampiciline (BioMol,
Hamburg); PAA 29:1, UltraPure Urea (ICN Biomedicals,
US); BSA, Geneticin, UltraPure™ Tris, (Invitrogen,
Karlsruhe); Phleomycin (Invivogen, Toulouse, France);
Agar-Agar, Yeast extract, Peptone, Triptone, Glucose,
Maltose, Isoamyl Alcohol (Merck, Darmstadt; Serva,
Heidelberg; Difco, MI, USA); DEPC, Sodium Acetate
(Fluka, Switzland); BPB, Chloroform, Isoamyl Alcohol,
Hydrochloric acid (MERK, Darmstadt); Biotin-N6-ddATP
(NEL); 6x SSPE Buffer (NIPPON GENE); One-Phor-All
Buffer (Pharmacia); Herring Sperm (Promega); Acetic acid,
Ethanol, Glycerol, Isopropanol, (Roth, Karlsruhe); EDTA,
IAA, Phenol, PMSF, TEMED, Acetoin, Thiamin
pyrophosphate, FAD, Creatine, Alpha-napthol, Pyruvate,
Triton-X (Sigma, St. Louis, USA);
Enzymes Restriction enzymes (New England Biolabs, Schwalbach);
DNAseI (Amersham); Zymolase, Protease, RNAse (Gibco
BLR)
Kits BioRad Protein-Assay (BioRad, München); AccuPre® PCR
45
Purification Kit (Bioneer, Seoul), Genomic extraction Kit
(Quiagen); GeneChip IVT Labelling Kits (Affymetrix)
Nucleic acids λ-DNA (Roche Diagnostic, Mannheim); GeneRuler TM
DNA Ladder Mix (MBI Fermentas, Litauen), 100bp Plus
DNA ladder (Bioneer, Seoul)
Oligonucleotides Metabion, Berlin
PCR-reagents Deoxynicleoside-Triosephosphate Set, Top-polymerase,
Pfu-polymerase (Roche Diagnostics, Mannheim; Bioneer,
Seoul)
All chemicals not listed above were obtained from the following companies: Fluka,
Merck, Roche Diagnostic, Sigma, Serva, Pharmacia and were of analytical grade or
better quality.
3.1.3
Strains
Table 1. Strains used in this study
Strain Lab name Genotype Source or reference
S. cerevisiae
BY4741 MATa his
3
∆1 leu2∆0 met15∆0 ura3∆0 Euroscarf
Brewers’ yeast
Sc-06136 A Bottom fermenting yeast Institut für Gaerungsgewerbe Berlin
Sc-06168 or H06 B Bottom fermenting yeast Institut für Gaerungsgewerbe Berlin
Sc-06165 C Bottom fermenting yeast, arose from
strain A via single cell isolation Institut für Gaerungsgewerbe Berlin
Sc-06165-ilv6
∆
Sc-ilv6
∆
Sc-ILV6/
Sc-ilv6∆::loxP-kanMX-loxP This study
Sc-06165- ilv6
∆∆
Sc-ilv6
∆
/ Sc-ilv6
∆
Sc-ilv6∆::loxP-kanMX-loxP/
Sc-ilv6∆::loxP-ble
r
-loxP
This study
E. coli
SupE44, ∆lacU169(
φ
80lacZ
∆
M15),
hsdR17, recA1, endA1, gyrA96, thi-1,
relA1
Melson M. and Yuan R. (1968)
46
3.1.4
Media and culture conditions
E. coli
E. coli strains were cultivated in LB medium (1% bactotryptone, 0.5% yeast
extract, 1% NaCl, pH 7.5) at 37
ο
C. Recombinant E. coli strains were selected in LB
medium with 150 µg/ml of Ampicilline. Solid LB medium suppelementing with
150 µg/ml of Ampicilline was obtained by adding 1.5% agar. For stock culture
storage, 1 ml of liquid culture was mixed with 0.3 ml glycerol 100% and then being
kept at -70
ο
C.
Yeasts
In general, S. cerevisiae and wild-type brewers’ yeast strains were grown in
Erlenmeyer flasks on a rotary shaker at 170 rpm in YEPD medium (2% peptone,
1% yeast extract, and 2% glucose) or on YEPD plates (1% yeast extract,
2% peptone, 2% glucose, 1.5% agar) at 30
ο
C.
The mutant Sc-ilv6
∆
where one copy of Sc-ILV6 was replaced by the
loxP-kanMX-loxP cassette was first propagated at 30
ο
C in YEPD-containing shaking
flask. Cells were then grown on selective YED plates (1% yeast extract, 2% glucose;
pH 6.3) supplemented with 17.5 µg/ml G418. YEPD plate was not being used since
the presence of peptone increased the threshold of selective concentration of this
antibiotic to yeast.
The mutant Sc-ilv6
∆/
Sc-ilv6
∆
which carries deletions in both copies of Sc-ILV6
ORFs and thus habouring the kanMX and ble
r
markers was first generated on YEPD
medium (1% yeast extract, 2% peptone, 2% glucose) at 30
ο
C. After that, cells were
replica-selected on YEPD plates supplemented with 17.5 µg/ml phleomycin and
afterwards on YED plates plus 50 µg/ml G418.
For storage, 1 ml stock of yeast culture was mixed with 0.3 ml glycerol and kept
at -70
ο
C. For preparation of any stock cultures, yeast strains were grown in selective
media.
47
3.1.5
Plasmids
Table 2. Plasmids used in this study
Plasmid Description Reference
pUG6 E. coli/ S. cerevisiae
shuttle vector, containing
Amp
+
, loxP-kanMX-loxP disruption cassette
Güldner et al., 1996
;
Güldner et al., 2002
pUG66 E. coli/ S. cerevisiae shuttle vector, containing
Amp
+
, loxP-ble
r
-loxP disruption cassette
Güldner et al., 2002
3.1.6
Oligonucleotides
The oligonucleotides used in this work were synthesized by Metabion (Berlin)
1) Primers used for disruption of Sc-ILV6 gene in lager brewers’ yeast
P1 (Forward primer):
5’- TACAGAATCTTTAGAACATCTGAGCTCACTAACCCAGTCTTTCTAccgccagctgaagcttcg - 3´
P2 (Reverse primer):
5’ - ATTTCGGCGACCAATTCTTGGGAGTCAGCGGCGCCAGCATTGGTGgcataggccactagtggatc - 3´
2) Primers used to verify the Sc-ILV6 deletion:
P3 (Forward primer): 5´- ATATGGAAGTACATAGTTCG - 3´
P4 (Reverse primer): 5´- TTCGGCGACCAATTCTTG - 3´
3) Primers used to verify the existence of non-Sc-ILV6
P5 (Forward primer): 5´- TAAGTCACATACGTAGTTTG - 3´
P6 (Reverse primer): 5´- TCGGCAACTAACTCGTTG - 3´
48
3.2 Methods
3.2.1
DNA methods
3.2.1.1 Isolation of yeast genomic DNA
Genomic DNA was isolated from S. cerevisiae and brewers’ yeast following the
method of Hoffman and Winston (1987). Yeast cells were grown overnight in 5 ml
YEPD medium. Next, 1.5 ml of the culture was harvested by centrifuging for
10 minutes at 12000 rpm at 4
ο
C and washed with 500 µl sterile distilled water. After
centrifugation, cell pellet was resuspended in residual water. For the cell disruption,
200 µl of lysis buffer (2% Triton X-100; 1% SDS; 0.1 M NaCl; 10 mM Tris-HCl,
pH 8.0; 1 mM EDTA), 200 µl of phenol:chloroform:isoamyl alcohol (25:24:1) and
300 mg of glass beads (0.45-0.5 mm) were added to the cell suspension. The tube
was vortexed vigorously for 4 minutes and then centrifuged for 5 minutes at
12000 rpm at 4
ο
C and the supernatant was collected. For the removal of protein
residuals, one volume of chroloform was added to the supernatant. The mixture was
vortexed for 30 s and centrifuged again for 5 minutes at 12000 rpm at 4
ο
C. The upper
phase was collected and 0.7 volume of isopropanol was added for DNA precipitation.
The precipitated DNA was obtained by centrifuging for 15 minutes at 12000 rpm
at 4
ο
C. DNA pellet was washed with 70% ethanol and then dissolved in 50 µl of
sterile distilled water.
3.2.1.2 Isolation of plasmid DNA from E. coli (minipreparation)
For analysis of E. coli transformants, the plasmid DNA was isolated from 1.5 ml
culture using the alkaline method of Birnboim and Doly (1979). Lager amounts of
plasmid DNA were isolated using GeneJET
TM
Plasmid Miniprep Kit (Fermentas).
3.2.1.3 Agarose DNA gel electrophoresis
Agarose gels of 0.8% (w/v) were used for DNA electrophoresis. The gel was
prepared by boiling agarose in 1x TBE buffer until agarose was totally dissolved.
49
Gels were casted at about 60
ο
C. DNA samples were mixed with 1/4 volume of
loading buffer (20 mM EDTA, pH 8.0, 0.025% bromophenol blue, 60% saccharose)
and loaded into the lanes of agarose gel. The electrophoresis was run in 1x TBE
buffer (89 mM Tris; 89 mM Boric acid; 2 mM EDTA, pH 8.0) at about 80 mA. For size
determination of DNA fragments, either the DNA size marker “GeneRuler
TM
”
(Fermentas) or “100bp Plus DNA leader” (Bioneer) was used. The gels were stained
in 0.5 mg/ml EtBr solution for 20 minutes and the DNA fragments were visualized
under UV light (λ = 254 nm).
3.2.1.4 PCR
PCR were performed using either GeneAmp9600 (Perkin Elmer) or Mycycler
(BioRad). The reactions were performed in 25 µl or 50 µl volume in 0.2 ml reaction
tubes (GeneAmp9600). The reaction mix contained 0.25 mM of each dNTP; 1 µM of
each primer; 1 ng/µl of template DNA; 1x reaction buffer, 1-2.5 units of DNA
polymerase and H
2
O up to the final volume. For generation of disruption cassettes,
Pfu DNA polymerase (Bioneer) was used. For PCR diagnostic of Sc-ILV6 disruption,
Top DNA polymerase (Bioneer) was used.
In general, PCR programe set up included following steps: i) pre-heat treatment
at 95
ο
C for 2 minutes; ii) 25 cycles as followed: denaturation for 45s at 94
ο
C, 45s at
annealing temperature and elongation at 72
ο
C and iii) end-elongation for 10 min
at 72
ο
C.
The annealing temperature was calculated based on melting temperatures of the
forward and reverse primers. The melting temperature of the primers was calculated
using pDRAW32 software. The annealing temperature was adjusted to two degree
lower than melting temperatures of the primers. The elongation time was calculated
depending on the length of the expected PCR. On average, 60 s elongation times
was used for the amplification of DNA fragment of 1 kb length.
50
3.2.1.5 Transformation of lager brewers’ yeast
Transformation of disruption cassettes into brewers’ yeast was performed using
the Lithiumacetate/PEG method. Yeast strains were precultured overnight in 20 ml
YEPD medium at 30
ο
C. For main culture preparation, 200 ml YEPD was inoculated
with the preculture by adjusting to an OD of 0.2. The main cultures were grown at
30
ο
C till an OD of 0.7 and cells were harvested by centrifugation for 2 minutes at
6000 rpm. The supernatant was decanted and cells were washed with 25 ml distilled
water. The cells were then centrifuged for 2 minutes at 6000 rpm and resuspended in
1 ml distilled water. After that, 400 µl of the cell suspension was washed and
resuspended again in 400 µl water. Next, this cell suspension was dispatched into
aliquots of 100 µl. The aliquots were kept on ice until being added with transformation
mixture.
The transformation mixture (without the DNA cassette) contained 240 µl PEG
4000 (50 % w/v), 36 µl 1 M lithium acetate, 50 µl single stranded carrier DNA (herring
sperm DNA 2 mg/ml). The carrier DNA was boiled and cooled on ice for generation of
single stranded DNA before being added to mixture.
Up to 1 µg of PCR product (dissolved in 34 µl water) was added to the yeast cell
suspension (100µl). The tube was vortexed vigorously to allow good contact between
cells and DNA. Next, the previously prepared transformation mix was added to the
tube. The complete mixture was vortexed vigorously and incubated for 40 minutes at
42
ο
C. Afterwards, the mixture was transferred into an Erlenmeyer flask containing
5 ml YEPD. The flask was shaked overnight at 170 rpm and 30
ο
C for the expression
of enzymes which conferred antibiotic resistance to the transformants. Cells were
harvested by centrifugation (13000 rpm, 30 sec) and then washed with 5 ml of
0.85% NaCl. After that, cell pellet was resuspended in 1ml of 0.85% NaCl. Cell
suspension was spread onto selective agar plates. The plates were incubated for 2-4
days at 30
ο
C for the appearance of colonies.
51
3.2.1.6 Transformation of E. coli
Transformation of E. coli was carried out using the heat shock method according
to Inoue et al. (1990). E. coli cells were made competent by CaCl
2
treatment. For
competent cells preparation, E. coli cells were precultured in 20 ml LB-medium from
a frozen stock culture (-70
ο
C) for 16 hours at 37
ο
C. Next, 1 ml of preculture was
transferred into an Erlenmeyer flask containing 100 ml of LB-medium. Cells were
grown until an OD of 0.4 and the main culture was kept on ice for 1 hour. Cells were
then collected by centrifuging for 5 minutes at 4
ο
C at 4000 rpm. Cells were washed
with 100 ml of ice-cold solution I (0.1 M MgCl
2
; 0.01 M Tris-HCl, pH 7.6). After that,
cells were dissolved in 50 ml of ice-cold solution II (0.1 M CaCl
2
; 0.01 M Tris-HCl,
pH 7.6) by gently mixing. The cell suspension was then centrifuged again as in the
previous step and the supernatant was decanted. Cell pellet was then resuspended
in 4 ml of ice-cold solution II and the suspension was kept on ice for 30 minutes. The
competent cells were either directly used for transformation or stored at -70
ο
C after
the addition of glycerol (15% final concentration).
For transformation, approximately 100 ng of plasmid DNA were mixed gently with
200 µl of competent cells. The mixture was kept on ice for 30 minutes and then
subjected to a heat shock at 42
ο
C for 30 sec. The cells were then put immediately on
ice for 1 minute and then mixed with 1 ml of LB medium. The tube was incubated
with agitation at 37
ο
C for 1 h. After that, various dilutions of the sample were spread
onto LB plates containing ampiciline (150 µg/ml).
3.2.1.7 Microarray-based comparative genomic hybridisation
Sample preparation and array hybridisation
Yeast cells were grown in YEPD medium and genomic DNA was extracted from
50 ml of yeast culture at an OD of 1.5 using genomic extraction kit from QIAGEN.
Protein and RNA were removed using zymolase, protease K and RNAse treatment
for 45 minutes. Sample preparation was carried out as following method described by
52
Winzeler (2003) with slight modifications. Genomic DNA (10 µg) was fragmented with
0.15 units DNAseI (Gibco BLR, PCR grade) in 1x One-Phor-All buffer (Pharmacia)
supplemented with 1.5 mM CoCl
2
for 3 minutes at 37
ο
C. DNAse I reaction was
inactivated by heating the sample at 95
ο
C for 15 minutes. Fragmented DNA was
labelled with
1 nmol Biotin-N6-ddATP (NEL) using 25 units termininal transferase
(Roche) at 37
ο
C for 1 hr. Labelled DNA fragments were dissolved in 200 µl
hybridisation solution containing 50 pM Control Oligonucleotide B2 (Affymetrix),
1x Eukaryotic Hybridisation Controls (Affymetrix), 20 µg herring sperm (Promega),
6x SSPE (0.9 M NaCl, 60 mM NaH
2
P0
4
, 6 mM EDTA) (NIPPON-GENE) and
0.005% Triton-X (SIGMA). After 10 minutes incubation at 100
ο
C, the hybridisation
solution was transferred on ice for a few minutes and afterwards hybridised to DNA
microarray.
Hybridisation was carried out for 16 h at 42
ο
C in hybridisation oven with
permanent mixing at 60 rpm. Genomic DNA of each lager yeast strain was hybridized
to one single array. Washing, staining and scanning of arrayss were carried out as
described in Affymetrix Technical manual (Affymetrix, 2004).
Data acquisition and analysis
Data analyses were performed in collobration with of Yoshihiro Nakao (Suntory
Ltd). Detection of signal intensities of microarrays was carried out using Affymetrix
Gene Chip Analysis Basic System and Affymetrix GeneChip® Operating
Software (GCOS) v 1.0. A probeset was detected as absent “A” or present “P” based
on detection p-value calculated by detection algorithm with default parameter in
GCOS. In each pairwise comparison, signal log
2
ratio and change p-value of every
probe set was calculated. A probeset had a change call of decrease “D”, increase “I”,
medium increase “MI”, medium decrease “MD” or not change “NC” based on the
change p-value calculated by Change Algorithm with default parameter in GCOS.
53
3.2.2
RNA method: Microarray-based comparative transcriptome analysis
3.2.2.1 Isolation of brewers’ yeast total RNA
Brewers’ yeast strains were grown in 30 litre fermenters under relevant brewing
conditions (11.5
ο
P brewers’ wort, 11
ο
C). For total RNA isolation, brewers’ yeast cells
were collected at apparent extract of 8%. Cell sampling was performed as described
by Piper (2002) with slight changes. Roughly 20 ml of culture corresponding to
240 mg yeast wet weight was harvested in triplicate for each strain from the 30 liter
scale fermentations. The broth was frozen instantly by pouring it directly into a
beaker containing liquid nitrogen. The frozen sample was then broken into small
fragments and transferred to two 50 ml centrifuge tubes. The sample was then
thawed at 0
ο
C by subsequent vigorous vortexing and keeping on ice. Next, the
samples were centrifuged at 5000 rpm at 0
ο
C for 4 minutes and re-suspended in
1.8 ml AE-cold buffer (50 mM sodium acetate, 10 mM EDTA, pH 5.0). The content of
the two tubes was pooled and afterwards aliquoted into 10 eppendorf tubes, 400 µl
each. RNA extraction was carried out using the hot phenol method (Schmitt et al.,
1990). The resulted RNA was treated with DNAseI (Amersham) (approximately
0.1 units per 1 µg RNA) to remove DNA. RNA sample was then purified again with
phenol:chloroform:isoamyl alcohol (25:24:1) for DNAseI removal. After that, RNA
sample was precipitated by adding 0.7 volumes of isopropanol and 0.1 volumes of
3 M sodium acetate, pH 5.2. The RNA pellet was washed with 80% ethanol and
resuspended in DEPC-treated water. The RNA sample was then kept at -80
o
C until
being used for hybridization.
3.2.2.2 Sample preparation and array hybridization
cDNA synthesis, biotin-labeled cRNA synthesis and fragmentation, hybridization
and washing, scanning of the array were performed following Affymetrix user’s
manual (Affymetrix, 2004). In short, single strand cDNAs was synthesed from total
RNA (15 µg) by incorporating T7 RNA-polymerase promoter. Subsequently, double
54
strand cDNAs was synthesized and was then used as the template for the synthesis
of biotin-labeled cRNA using GeneChip IVT Labeling Kit (Affymetrix). Biotin-labelled
cRNA was then fragmented. Hybridization cocktail (300 µl) containing 15 µg
fragmented Biotin-Labeled cRNA, 50 pM Control Oligonucleotide B2, 1x Eykaryotic
Hybridization Controls, 0.1 mg/ml Herring Sperm DNA, 0.5 mg/ml Acetylated BSA,
1x hybridisation buffer and 10% DMSO (Affymetrix) was applied to bottom fermenting
yeast DNA microarray. The hybridization was carried out for 16 hour at 45
ο
C and at
60 rpm rotation. Array hybridizations were carried out in technical triplicate, i.e. the
three independent RNA isolations for each strain, respectively. The arrays were
washed and stained as described in Affymetrix user’s manual (Affymetrix, 2004)
using Fluidics Station 450. Arrays were then scanned with Affymetrix GeneChip
Scanner 3000.
3.2.2.3 Data acquisition and analysis
Data analyses were performed in collaboration with Dr. Matthias E. Fuschik
(Humbolt University) and Yoshihiro Nakao (Suntory Ltd.). Detection of signal
intensities of micoarrays was carried out using Affymetrix Gene Chip Analysis Basic
System and Affymetrix GeneChip® Operating Software (GCOS) v 1.0. Every
transcript was flagged as absent “A”, present “P” or marginal present “M” based on
the detection p-value calculated by Detection Algorithm with default parameter in
GCOS. Subsequently, the data was adjusted by quantile normalization (Bolstad et
al., 2003). For every transcript in each pairwise comparison, we calculated the
logged average fold-change. The significance of differential expression was
assessed using the CyperT approach, which is based on a Bayesian t-test (Baldi and
Long, 2001). As significant threshold for the pairwise comparisons, a false discovery
rate of 0.001 was chosen. Pathway analysis of significant differences was performed
using SGD database and Microsoft Access program.
55
3.2.3
Protein methods
3.2.3.1 Comparative proteome analysis
3.2.3.1.1 Isolation of total protein from bottom fermenting yeast
Brewers’ yeast strains were grown in 3 litre fermenters under relevant brewing
conditions (11.38
ο
P brewers’ wort, 12
ο
C). For protein isolation, brewers’ yeast cells
were collected at an apparent extract of 8% by centrifugation at 6000 rpm for
5 minutes. Subsequently, 1 g wet weight of yeast cells was washed twice with 100 ml
and then with 50 ml of ice-cold 10 mM Tris-HCl, pH 8.0 buffer and centrifuged at
4000 rpm for 3 minutes in each step. Cells were then resuspended in the same buffer
and dropped into liquid nitrogen in the form of small bits. Frozen cells were ground
into powder using Braun Dismembrator at 2500 rpm for 2 minutes. Frozen cell
powder was transferred into an eppendorf tube and thawed to 4
ο
C. Next, 10 µl of
100mM PMSF was added to the cell samples and the mixtures were centrifuged for
8 minutes at 8000 g at 4
ο
C. Supernatant was then collected and centrifuged again at
16000 rpm for 30 minutes at 4
ο
C. Protein concentration of the supernatant was
determined using Bradford assay. Protein sample was dispatched into eppendorf
tubes in small aliquots of 70 µl and stored at -80
ο
C until being used for
electrophoresis.
3.2.3.1.2 Two-dimensional gel electrophoresis
Protein sample (800 µg) was solublized in 600 µl reswelling solution (8 M urea,
1% CHAPS, 0.4% DTT, 0.5 v/v Pharmalyte 3-10, 0.002% bromophenol blue) by
shaking for 30 minutes at 600 rpm. The samples were then centrifuged for 5 minutes
at 12000 rpm and the supernatants were applied to a 24 cm, nonlinear Immobiline
Drystrip, pH 3-7 (Amersham Biosciences). The strips were covered with silicon oil.
Protein focusing was performed on IPGphor at 20
ο
C as follow: 30 V for 15 h, 200 V
for 1 h, 500 V for 1 h, 1000 V for 1 h, gradient 8000 V for 30 minutes, and 4 h at
8000 V. The strips were equilibrated twice for 15 minutes firstly in 10 ml buffer
56
(6 M urea, 2% SDS, 50 mM Tris-HCl pH 8.0, 30% glycerol) supplemented with
1% (w/v) DTT and then in 10 ml of the same buffer supplemented with 2.5% (w/v)
idoacetamid.
After equilibration, IPG strips were load onto 12.5% acrylamide/bis-acrylamide
(29:1) gels. The two gels were sealed with 0.5% (w/v) agarose solution containing
0.002% BPB. The second dimension was perfomed using Ettan
TM
Dalt six
Electrophoresis Unit (Amersham Biosciences). The gels were run at 20
ο
C at 3W per
gel for 30 minutes and then at 20 W per gel for 4.5 h in the electrophoresis buffer
containing 25 mM Tris-base, 192 mM glycine, 1% (w/v) SDS. Gels were stained
overnight with Comassie Brillant Blue and washed several times with destaining
solution (25% Ethanol, 5% acetic acid). After that, gels were washed again with
sterile distilled water for 30 minutes and scanned with the Image Scanner
(Amersham).
Protein samples of the three strains were isolated in triplicate from two
independent fermentations. Concretely, protein samples of three strains were
isolated twice from the first fermentation and once from the second fermentation. For
each strain, 2D gel of protein extract was run in triplicate. Protein samples of the
three selected strains were always run concurrently.
Scanned images of analytical gels were analysed using Delta 2D Software v 4.0
(Decodon, Greifswald, GmbH). A master gel for each strain was created from all
replicates using all-to-one warping strategy. The average volume (in percentage to
the total volume) and standard deviation of each spot in replicates were calculated.
The master gels were used for pairwise comparisons between selected lager yeast
strains. Differentially expressed proteins were selected with a fold change of 2 of
average volumes. Statistical analysis was performed allowing a standard deviation
≤ 30% for each spot from the three replicates and a p-value of 0.05 or below in the
Student’s t-test at pairwise comparison.
57
3.2.3.1.3 MALDI-TOF mass spectrometry
Protein spots were excised from 2D gels; trypsin digested and identified using
MALDI-TOF mass spectrometry (Kohler et al., 2005). The analysis was performed
using the service of Greifswald University, Institut für Marine Biotechnologie
(Prof. Thomas Schweder group). MS data were investigated using Mascot search
engine (Matrix Science Ltd) against the protein database of S. cerevisiae. Peptides
that yield a protein score of at least 100 and a sequence coverage ≥ 30% for
duplicate identifications were regarded as positively detected.
3.2.3.2 Determination of protein concentration
The protein concentration in cell extracts was determined using Bradford assay
(1976). The dye reagent (Bio-Rad) was diluted 1:4 with distilled water. Cell protein
extract (200 µl) was mix with 800 µl of the diluted dye solution and incubated for 5-10
minutes at room temperature. The extinction was measured at a wavelength of
595 nm against a control made up of 800 µl of the prepared dye solution and 200 µl
of the buffer used to dissolve cell protein extract. For protein determination, a
calibration curve of BSA with concentrations of 20-70 µg/ml was prepared.
3.2.3.3 Determination of enzyme activity
3.2.3.3.1 Preparation of permeabilized cell proteins
Yeast strains were pre-cultivated in 20ml wort by shaking at 24
ο
C for one day.
Yeast strains were then inoculated at the concentration of 1 x 10
7
cells/ml in 90 ml
brewers’ wort (11.38
ο
P). Fermentations were carried out in 100 ml bottles closed with
airlocks at 12
ο
C until the apparent extract was reduced to a value varying from 8.3 to
8.7. Cells were harvested by centrifuging for 5 minutes at 5000 rpm at 4
ο
C and
washed twice with sterile distilled water. Next, cells were resuspended in 1 ml
ice-cold buffer (0.1 M Tris-HCl pH7.5, 0.1 M NaCl and 0.1 M EDTA) supplemented
with 2 mM PMSF. Cells were permeabilized by adding 100 µl chloroform and
58
vortexing for 30 s. Cell samples were then centrifuged at 6000 rpm for 5 minutes 4
ο
C
and the supernatant was removed. Permeated cell were resuspended in the same
buffer and were placed at 4
ο
C for being used for AHAS assay within 2 h.
3.2.3.3.2 In vitro acetohydroxyacid synthase (AHAS) assay
AHAS was assayed based on its activity to convert pyruvate into α-acetolactate.
The enzyme activity was assayed using the method described by Byrne and
Meacock (2001) with some modifications. The assay was performed in a volume of
100 µl. For each sample, two tubes, the control and sample itself were set up. A 90 µl
mixture containing 65 µl cell suspension, 5 µl ThDP 20 mM, 5 µl 0.2 M MgCl
2,
5 µl of
4 mM FAD and 10 µl of 1 M K
3
PO
4
(pH7.5) was added to each tube and incubated
for 10 minutes at 30
ο
C. After that, 10 µl 1 M pyruvate was added to the mixture and
the reactions were carried out for 20 minutes at 30
ο
C. The reactions were then
stopped by the addition of 11.3 µl 9.9 M H
2
SO
4
to the sample and 150
µl 6M NaOH to
the control tube. The sample tubes were then incubated at 60
ο
C for 30 minutes to
allow the efficient conversion of α-acetolactate to acetoin. After that, 140 µl 8 M
NaOH was added to each sample tube to stop the reaction. At this point, the sample
and control had the same volume (250 µl) and the same pH. The yield of α-
acetolactate was determined by the amount of acetoin produced in the
decarboxylation reaction which was taken place in the acidic condition at high
temperature. In the control tubes, α-acetolactate was not decarboxylated to acetoin,
therefore the amount of acetoin produced in the decarboxylation reaction was
measured by subtracting the amount acetoin in the background (control tube) from
the total amount of acetoin (sample tube).
The concentration of acetoin in the control and sample tubes was determined
using colorimetric method (Westerfeld, 1945). Each tube was filled with 750 µl water,
200 µl 0.5% creatine and 200 µl 5% α-napthol freshly prepared in 2.5 M NaOH. The
tubes were vortexed for 2 s and kept for 1 h at RT to allow colour development. The
59
reaction mixtures were centrifuged for 2 minutes for clarification. Absorbances of the
supernatants at 525 nm were measured against the blank made up of 1 ml water,
200 µl 0.5% creatine and 200 µl 5% α-napthol. AHAS activity was calculated as
acetoin produced per mg of permeabilized cell protein per h.
3.2.4
Fermentations
Yeast cells and green beers used for different experiments (e.g. protein isolation,
RNA isolation and green beer analysis) were harvested from different fermentations
carried out under sightly different conditions.
For determination of characteristics of green beers produced by brewers’ yeast
strains and the fermentation performance of brewers’ yeast strains, yeast strains
were cultivated and fermented using brewers’ wort with original gravity of 11.38
ο
P
kindly provided by a German brewery. Yeast strains were cultivated in 200 ml wort by
shaking at room temperature. After three days, 600 ml wort was added and the
cultures was cultivated under the same condition for another two days. The cultures
were then filled with 1000 ml wort and incubated without shaking at 12
ο
C for 24 h.
The cells were harvested by fermentation and inoculated into fresh wort at the
density of 1 x 10
7
cells/ml for primary fermentation. The fermentations were carried
out in 3 litre glass fermenters with stirring at 50 rmp at 10
ο
C. The fermenters were
closed with airlocks for elimination of oxygen but allowing the release of gases.
During the main fermentation, cell density, decrease of wort apparent extract and the
diacetyl concentration were recorded. The fermentations were finished when wort
apparent extract decreased to a value between 2.8% to 3%. After the main
fermentation, green beers produced by yeast strains were taken for analysis. Besides
ethanol and glycerol, other by-products were analysed included higher alcohol,
organic acids, vicinal diketones, esters and fatty acids.
60
For protein isolation, yeast cells were cultivated and fermented in 3 litre-scale
galss fermenters as described above. Yeast cells were harvested at an apparent
extract of 8% for total protein isolation.
For isolation of brewers’ yeast total RNA, yeast strains were cultivated and
fermented using brewers’ wort of 11.5
ο
P produced by our collaboration partner from
VLB (Prof. Frank Jürgen Methner group). Yeast strains were pre-cultivated in 30ml
wort by shaking at 24
ο
C for one day and then transferred into a bottle containing
500 ml wort. The secondary pre-cultures were cultivated under the same condition
until they reached a cell density of 1 x 10
8
cells/ml. Next, 5 litre wort was inoculated
with each secondary preculture and yeast strains were cultivated under the same
condition. Main fermentations were carried out in 30 litre scale-tanks at 11
ο
C. Wort
was inoculated at a density of approximately 1.6 x 10
7
cells/ml for the main
fermentation. For RNA isolation, cells were harvested at an apparent extract of 8%.
For determination of fermentation performance and vicinal diketone production by
the engineered strain Sc-ilv6
∆/
Sc-ilv6
∆
, yeast strains (the reference strain C and
strain Sc-ilv6∆/Sc-ilv6∆) were cultivated and fermented under relevant industrial
brewery conditions using brewers’ wort with original gravity of 11.38
ο
P. Yeast strains
were innoculated in 5 litre bottles at the cell density of about 1.1 x 10
7
cells/ml. The
cultures were incubated at RT (20-25
ο
C) until they reached the cell density of
ca. 1.3 x 10
8
cells/ml. Wort was inoculated with the precultures at the cell density of
approximately 1 x 10
7
cells/ml for the main fermentation. Main fermentations were
carried out in 30-litre fementation tank filled with 24 litre brewers’ wort at 10.5
ο
C.
During the fermentation, cell density, decrease of wort apparent extract and the
diacetyl concentration in wort medium were recorded. The fermentations were
finished until wort apparent extract was decreased to a value from 2.8% to 3%. After
the main fermentation, green beers produced by yeast strains were taken for
analysis. The measured compounds included ethanol, glycerol and other
flavour-relevant products such as vicinal diketones, fusel alcohols, esters and fatty
acids.
61
For in vitro AHAS assay and determination of diacetyl production levels of strains
B, C and Sc-ilv6 mutants, yeast strains were cultivated and fermented using brewers’
wort with original gravity of 11.38
ο
P kindly provided by a German brewery. Yeast
strains were precultured in 20 ml wort by shaking at room temperature. The second
preculture was performed in 100 ml wort until it reached to a cell density varying from
5 x 10
7
to 1 x 10
8
cells/ml. Cells were harvested by centrifugation. For fementation,
wort was inoculated with the preculture at concentration of 1 x 10
7
cells/ml. The
fermentations were carried out in 100 ml bottles containing 90 ml wort. The strains
were fermented in 100 ml bottles closed with airlocks at 12
ο
C until an apparent
extract varying from 8.3 to 8.8 was reached. Cells were then harvested for
preparation of permeabilized cell proteins and the culture supernatant was collected
for diacetyl measurement.
3.2.5
Analytical methods
Vicinal diketone concentration of the wort medium harvested from the laboratory
scale fermentations was determined using GC-ECD method derived from MEBAK
(Band II, 1.2.1, 1996). Vicinal diketone measurement was carried out by our
collobration partner at “Institut für Versuchs- und Lehranstalt für Brauerei in Berlin”
(VLB) (Prof. Frank-JürgenMethner group).
Determination of apparent extract of wort medium harvested from the laboratory
scale fermentations was performed using the method according to MEBAK (Band III,
1.1.1, 1.1.4, 1996).
Determination of apparent extract of wort medium and components of the green
beer derived from the relevan industrial brewery fermentations were performed
according to international standardized methods edited by European Brewery
Convention (1998). These analyses were carried out by our collobration partner in a
German beer factory.
62
4 Results
4.1 Phenotypes of the three selected lager brewers’ yeast strains
producing different levels of diacetyl
This work focuses on applying the strategy of inverse metabolic engineering to
reduce diacetyl production in lager brewers’ yeast. To that aim, we at first selected
lager brewers’ yeast strains which showed different levels of diacetyl production.
These strains are derived from the collection of “Institut für Versuchs- und Lehranstalt
für Brauerei in Berlin” (VLB) with the lab names A, B and C (see materials and
methods). Strain A produces the highest level of diacetyl. Strain B was previously
selected for a very low level of diacetyl production. Strain C arose from strain A from
single cell isolation and produces a slightly higher level of diacetyl than strain A.
Strain C is a common production strain for lager beer brewing.
The performances of these brewers’ yeast strains were investigated under
industrially relevant brewing conditions (11.38
ο
P wort, 12
ο
C). For each strain, the
time courses of apparent extract were recorded. The time courses of apparent
extract are the readouts for the consumption of wort sugar during the fermentation of
brewers’ yeast strains. At the end of the main fermentation, green beer produced by
each strain was taken for product analysis. The measured compounds included
ethanol, glycerol and other flavour-relevant products such as vicinal diketones, fusel
alcohols, esters and fatty acids.
4.1.1
Wort sugar consumption during the main fermentation of the three
selected lager brewers’ yeast strains
The decreases of wort gravity in strains A and C were almost similar. However, a
difference in fermentation rate between strain B and the other strains was observed
after day 3 of the main fermentation. Thus, compared to strains A and C, strain B
showed a slower rate of wort sugar consumption. Strain B needed more time,
63
i.e. 60hrs and 40 hours, respectively than strains A and C to reach the wort
attenuation (apparent extract of 3%) (Fig. 3).
0
2
4
6
8
10
12
14
0 20 40 60 80 100 120 140 160 180 200 220
A
B
C
Fermentation time (h)
Apparent extract (%)
0
2
4
6
8
10
12
14
0 20 40 60 80 100 120 140 160 180 200 220
A
B
C
Fermentation time (h)
Apparent extract (%)
Fig. 3 Time courses of apparent extract during the fermentation by the three selected lager
brewers’ yeast strains. Fermentation was carried out in 3 litre glass fermenters under conditions
relevant to industrial brewery fermentations (11.38
ο
P brewers’ wort, 11
ο
C).
4.1.2
Diacetyl production
The relative difference in diacetyl production of the three selected lager brewers’
yeast strains was investigated under industrially relevant brewing conditions (11.38
ο
P
wort, 12
ο
C) in 3 litre glass fermenters. The fermentations were carried out in
duplicate. For each strain, diacetyl was investigated at apparent extract of 8%, 6%
and 3% corresponding to the beginning, the middle and the end stages of the main
fermentation, respectively (Fig. 4). It is obvious from Fig. 4 that from all investigated
points of apparent extract; strain A produced the highest level of diacetyl while the
diacetyl production of strain B was the lowest. By the end of fermentation
(ca. apparent extract of 3%), the diacetyl concentration of beer produced by strain B
was of about 17% and 34% compared to those of strains A and C, respectively.
Interestingly, the diacetyl production of strain B by the end the main fermentation was
below the beer diacetyl taste-threshold of 0.1 mg/ml (Fig. 4).
64
0
0.1
0.2
0.3
0.4
0.5
234
5
6
78
9
Apparent extract (%)
Diacetyl in medium (mg/l)
A
B
C
Diacetyl taste threshold in beer
0
0.1
0.2
0.3
0.4
0.5
234
5
6
78
9
Apparent extract (%)
Diacetyl in medium (mg/l)
A
B
C
A
B
C
Diacetyl taste threshold in beer
Fig. 4 Different diacetyl productions of the three studied lager brewers’ yeast strains.
Fermentation was carried out in 3 litre glass fermenters under conditions relevant to industrial brewery
fermentations (11.38
ο
P brewers’ wort, 11
ο
C). Values at apparent extract of 6% and 3% are the
average from two independent experiments, including standard deviations. Values at apparent extract
of 8% were measured from a single fermentation.
4.1.3
Flocculation behaviour
Fig. 5 Difference in flocculation behaviour between the studied lager brewers’ yeast strains.
Fermentation was carried out in Lietz device using 11.38
ο
P brewers’ wort at 12
ο
C. The picture was
taken after 7 days fermentation. Strain B flocculated much earlier than strains A and B.
Strain A Strain B Strain C
Strain A Strain B Strain C
65
Despite the very low level of diacetyl production (Fig. 4), strain B is not useful for
brewing industry since it flocculates very early. The flocculation behaviour of three
selected lager brewers’ yeast strains was easily detectable using Lietz fermentation
devices. It can be seen from Fig. 5 that the medium of strain B was much clearer
than that of strains A or C after 7 days of the main fermentation. The low number of
non-sedimented cells left in the medium explaines the fact that strain B needed more
time than strains A and C to reach the wort attenuation (II.4.1.1).
4.1.4
Other fermentation by-products
Green beer analysis showed that the patterns of by-products of the three selected
lager brewers’ yeast strains were almost similar except for some slight differences in
isobutyl acetate, acetaldehyde and 2,3-pentanedione concentrations (Table 3). The
difference in 2,3-pentanedione production in the studied strains corresponds to their
differences in diacetyl productions. This result matches the fact that the precursors of
these two by-products are both formed by the activity of acetohydroxyacid synthase.
66
Table 3. Analysis of green beers produced by the three studied brewers’ lager yeast strains
(harvested at apparent extract of 3%). Fermentation was carried out as indicated in Fig. 3. Except
as noted, the results shown are mean values of two independent experiments including standard
deviations. The bold lines indicate difference of by-product production in the studied strains.
Strains A B C
Ethanol (g l
-1
) 37.65 ± 0.25 36.30 ± 0.80 37.30 ± 0.07
Glycerol (g l
-1
) * 1.5 1.5 1.5
pH 4.155 ± 0.015 4.325 ± 0.065 4.185 ± 0.020
Acetaldehyde (mg l
-1
)* 1.76 4.70 2.65
Fusel alcohols
n-propyl alcohol (mg l
-1
)* 19.3 16.3 20.9
Isoamyl alcohol (mg l
-1
)* 39.2 38.1 47.1
Iso-butyl alcohol (mg l
-1
)* 8.9 6.5 10.6
2-Phenylethyl alcohol (mg l-
1
) 12.10 ± 0.00 13.60 ± 1.60 14.50 ± 1.55
Vicinal diketones
2,3-pentandione (mg l
-1
) 0.355 ± 0.005 0.090 ± 0.000 0.260 ± 0.060
Diacetyl (mg l
-1
) 0.425 ± 0.005 0.075 ± 0.005 0.255 ± 0.110
VDK (mg l
-1
) 0.780 ± 0.010 0.165 ± 0.005 0.515 ± 0.180
Esters
Ethyl acetate (mg l-1) * 11.62 11.52 10.6
Butyl acetate (mg l
-1
) 0 ± 0 0 ± 0 0 ± 0
Isoamyl acetate (mg l
-1
) 0.495 ± 0.035 0.408 ± 0.120 0.475 ± 0.030
Isobutyl acetate (mg l
-1
) 0.017 ± 0.008 0.040 ± 0.012 0.030 ± 0.014
2-Phenylethyl acetate (mg l
-1
) 0.150 ± 0.000 0.175 ± 0.025 0.170 ± 0.010
Ethyl butyrate (mg l
-1
) 0.040 ± 0.000 0.045 ± 0.005 0.045 ± 0.000
Ethyl caproate (mg l
-1
) 0.110 ± 0.010 0.085 ± 0.015 0.090 ± 0.010
Ethyl caprate (mg l
-1
) 0.01 ± 0.00 0.01 ± 0.00 0.01 ± 0.00
Ethyl caprylate (mg l
-1
) 0.170 ± 0.010 0.135 ± 0.015 0.150 ± 0.000
Fatty acids
Isovalerate (mg l
-1
) 1.58 ± 0.06 1.36 ± 0.11 1.42 ± 0.12
Caproate (mg l
-1
) 1.385 ± 0.005 1.245 ± 0.015 1.265 ± 0.030
2-Ethyl caproate (mg l
-1
) 0.120 ± 0.051 0.150 ± 0.093 0.110 ± 0.065
Caprate (mg l
-1
) 0.23± 0.00 0.26 ± 0.00 0.235 ± 0.02
Caprylate (mg l
-1
) 3.145 ± 0.035 3.035 ± 0.075 2.920 ± 0.010
*
)
Values were recorded from a single experiment
67
4.2 Global molecular analyses of the three selected brewers’ yeast strains
possessing variations in diacetyl production
To understand the genetic basis leading to the difference in diacetyl production in
the selected lager brewers’ yeast strains, molecular global analyses of the strains
were carried out at the level of genome, transcriptome and proteome. The cell
samples used for transcriptome and proteome analyses were harvested at apparent
extract of 8% of the main fermentation. At this point of time, all strains were in the
logarithmic growth and had the same cell densities (Fig. 3).
4.2.1
Genome level: Microarray-based comparative genome hybridization
using bottom fermenting yeast DNA microarray
It is well known that lager brewers’ yeast is an aneuploid hybrid between
S. cerevisiae and another Sacchromyces yeast, probably S. bayanus (Kodama et al.,
2006; Rainieri et al., 2006). The DNA component from S. cerevisiae of lager brewers’
yeast is often referred to Sc-like type (Sc-, cer-, -CE) whilst its other DNA component
is differently denoted as S. pastorianus (-Sp), lager (-Lg), non-S. cerevisiae (non-Sc)
or S. carlsbergensis (-CA) (Kodama et al., 2006). Here the terms
Sc-
type and
non-Sc-type are consistently used.
Before having a closer look at the results obtained at genome level, it is important
to review some recent knowledge regarding the chromosomal structure of lager
brewers’ yeast. Different studies have shown that lager brewers’ yeast genome
contains three kinds of chromosomes: Sc-type, non-Sc-type and various chimeral
types (Fig. 6) (De Barros Lopes et al., 2002; Kodama et al., 2006). By applying
comparative genomic hybridisation (CGH) with the use of S. cerevisiae array,
Kodama et al. (2006) revealed that different lager brewers’ yeast strains could have
variable compositions of chromosomes. Concretely, they can contain different types
and copy number of some certain chromosomes in their genetic set-up. For example,
68
most lager yeast strains possess chromosome XVI in two different Sc/non-Sc hybrid
types while only few strains contain Sc-type and non-Sc-type ones.
Fig. 6 Putative chromosomal structure of lager brewers’ yeast strain Weihenstephan Nr.4
(34/70) (Kodama et al., 2006). The break points between Sc-type and non-Sc-type DNA in
chromosomes are shown as constrictions.
Furthermore, one notable point about chromosomal structure of lager brewers’
yeast is that the pure non-Sc-type of chromosomes III, VII and XVI are mostly not
observed. Instead of that, these chromosomes are mostly found in Sc-type and
non-Sc/Sc hybrid type (Fig. 6) (Y. Nakao, pers.comm.). In addition, lager brewers’
yeast genome is remarkable by the translocations between non-Sc-types of
chromosomes II and IV; chromosomes VIII and XV which resulted in the presence of
hybrid non-Sc-type between these chromosomes (Fig. 7) (Kodama et al., 2006),
(Y. Nakao, pers.comm.).
To investigate the genetic basis for the differences in diacetyl production of the
threeselected lager brewers’ yeast strains at genomic level, we employed microarray-
based CGH by means of bottom fermenting yeast DNA microarray to study our
strains. In the pairwise comparison, a sequence was designated as “increased” if the
69
calculated change p-value was ≤ 0.002 and as “decreased” if the change p-value was
≥ 0.998. In general, the genome analysis revealed thousands of significant changes
in the studies strains i.e. A vs B (5730); B vs C (6003) and A vs C (2094). These
include differences regarding the coding and intergenic regions. This result is
consistent to the fact that strain A and C are genetically related and different from
strain B.
Sc/Non-Sc type
ChrII
ChrIV
ChrVIII
ChrXV
Non-Sc typeSc type
non-Sc-IVnon-Sc-II
non-Sc-IV
non-Sc-VIII
non-Sc-XV
non-Sc-XV
non-Sc-VIII
non-Sc-II
Sc/Non-Sc type
ChrII
ChrIV
ChrVIII
ChrXV
Non-Sc typeSc type
non-Sc-IVnon-Sc-II
non-Sc-IV
non-Sc-VIII
non-Sc-XV
non-Sc-XV
non-Sc-VIII
non-Sc-II
Fig. 7 Translocation between the non-Sc-type chromosomes II and IV and between the
non-Sc-type chromosomes VIII and XV in lager brewers’ yeast (Y.Nakao, Genomic analysis of
lager brewing yeast and its application to brewing). The break points between Sc-type DNA and
non-Sc-type DNA in chromosomes are shown as constrictions.
In each pairwise strain comparison, log
2
hybridisation ratio for every single gene
was calculated. If the DNA components of two strains were identical, a log
2
hybridisation ratio of 0 was expected while deviations of this value indicated
variations between the two samples. To estimate the differences in the chromosome
constitutions between the compared strains, the log
2
hybridisation ratio of each single
ORF was plotted versus its position on the chromosomes, using the gene order of
S. cerevisiae. Since strains C and A are genetically closely related, we expected that
70
the genome patterns of strains A and C would display high similarity and be different
from that of strain B. The chromosomal comparison, therefore, was focused on two
pairs: strain A vs C and strain B vs C
.
4.2.1.1 Differences of the studied strains in their chromosome patterns
4.2.1.1.1 Strain A vs strain C
In accordance to the fact that strains A and C are genetically related, microarray-
based CGH revealed that the signal intensities of genes on most of chromosomes of
these two strains were in fact very similar (Fig. 8B). Comparison of these two strains
only showed changes in relative log
2
hybridisation ratios for part of chromosomes VIII
and XV (Fig. 8B). The relative log
2
hybridisation ratios of Sc-type and non-Sc-type
ORFs signals on other chromosomes were equal to 0 indicating that the two
compared strains possess a similar constitution of these chromosomes in their
genomes.
In the case of chromosome VIII, log
2
hybridisation ratios of Sc-type ORFs
between strains A and C were about 0.45 (Fig. 8B). These log
2
hybridisation ratios
corresponded to a hybridisation ratio of 1.36, suggesting that strain A contains more
copy numbers of Sc-type chromosome VIII than strain C. For example, strain A may
contain 4 copies while strain C may possess 3 copies of this chromosome.
Log
2
hybridisation analysis showed one region of no difference in Sc-type
chromosome XV while in the other parts, the relative log
2
hybridisation ratio was
equal to 0.42 corresponding to the relative hybridisation of 1.34 (Fig. 8B). The region
of no signal difference spread from the gene y0r343w to yor065c, similar to the
previously described “jump region” or the translocation points on chromosome XV in
brewers’ yeast (Bond et al., 2004). As lager brewer’ yeast has three types of
chromosome XV: Sc-type, non-Sc and hybrid Sc/non-Sc types (Fig. 7) I deduced that
genome of strain A might have more copies of the hybrid type of chromosome XV
than strain C.
71
Fig. 8 Array-based genomic comparison of the studied lager brewers’ yeast strains by means
of bottom fermenting yeast DNA microarray. A) Strain B vs C. B) Strain A vs C. In each pairwise
comparison, the log
2
hybridisation ratio of each ORF was plotted versus its position on chromosome.
The red and blue colours indicate the log
2
hybridisation ratios of Sc-type and non-Sc-type ORFs,
respectively. The points where signal show abrupt changes are considered sites of recombination
which gave rise to chimeral chromosomes and are simply denoted as translocation points.
chr.1
Sc
Non-Sc
0
1
-1
0
1
-1
A)
B)
Log
2
hybridisation ratio
chr.1
Sc
Non-Sc
0
1
-1
0
1
-1
A)
B)
chr.1
Sc
Non-Sc
0
1
-1
0
1
-1
A)
B)
Log
2
hybridisation ratio
72
Non-Sc type chromosome VIII and non-Sc-type XV in A vs C comparison showed
regions of relative log
2
hybridisation of -1 respectively on the left side and right side
(Fig. 8B). That means the hybridisation signals of the ORFs located on these regions
in this chromosome of strain C were two times higher compared to strain A. As
aforementioned, the non-Sc-type chromosomes VIII and XV lager brewers’ yeast are
present as the heterogenic hybrid types in comparison to S. cerevisiae chromosomes
(Fig. 7). Based on this knowledge, I deduce that strain C contains a higher copy
number of non-Sc type chromosomes XV compared to strain A.
4.2.1.1.2 Strain B vs strain C
A number of slight differences in relative hybridisation ratios were observed in
chromosomal comparison between strains B and C. These included differences in
chromosomes II, IV, V, VII, XI, XII, XIII, XIV and XVI (Fig. 8A). As a high percentage
of ORFs on these chromosomes have been detected as “not changed” in the
comparison of the two strains, we concluded that these differences were not due to
the difference in chromosomes copy numbers. It is more likely that these differences
resulted from the varying specificities of the probesets of the bottom fermenting yeast
DNA microarray to the genomes sequences of the two compared lager brewers’
strains.
The most striking differences in strains B and C comparison were found on
chromosomes I, III, VI, VIII, IX, X and XV (Fig. 8A). Based on the relative
hybridisation signals and the chromosome structure of lager brewers’ yeast, we
suppose some differences in chromosomes constitutions between strains B and C
(Fig. 9).
73
Fig. 9 Chromosomal differences and possible compositions of some chromosomes in strains B
and C. The red and blue colours indicate the log
2
hybridisation signal ratios of Sc-type ORFs and non-
Sc-type ORFs, respectively. The points where signal show abrupt changes are considered sites of
recombination which gave rise to chimeral choromosomes and are simply denoted as translocation
points.
Regarding chromosome VIII, both Sc-type and non-Sc-type showed regions of no
change in its left side in B vs C comparison (Fig. 9). In contrast, there was one region
on the right side of Sc-type chromosome VIII which exhibit a relative hybridisation
ratio of 2 times lower in strain B compared to strain C. In addition, another region on
XV
X
IX
VIII
VI
III
I
Log
2
hyb. signal ratios vs.
chromosome composition
Chromosome
structure of lager
brewing yeast
Chr.
Non-Sc
Sc
Non-Sc
Sc
Non-Sc
Sc
Non-Sc
Sc
Non-Sc
Sc
Non-Sc
Sc
Non-Sc
Sc
Hybrid type: B > C
Sc type: B = C
Non-Sc-type: B > C
Sc-type: B < C
Sc type: B = C
Non-Sc-type: B < C
No explanation
Sc-type: B > C
Non-Sc-type: B = C
No explanation
No explanation
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
Possible chromosomal composition in
strains B and C derived from results
B C
: Centromere: Region of no synteny to S. cerevisiae, > 20kb
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
: Sc-type chromosome : Hybrid type chromosome : non-Sc-type chromosome
XV
X
IX
VIII
VI
III
I
Log
2
hyb. signal ratios vs.
chromosome composition
Chromosome
structure of lager
brewing yeast
Chr.
Non-Sc
Sc
Non-Sc
Sc
Non-Sc
Sc
Non-Sc
Sc
Non-Sc
Sc
Non-Sc
Sc
Non-Sc
Sc
Non-Sc
Sc
Non-Sc
Sc
Non-Sc
Sc
Non-Sc
Sc
Non-Sc
Sc
Non-Sc
Sc
Non-Sc
Sc
Hybrid type: B > C
Sc type: B = C
Non-Sc-type: B > C
Sc-type: B < C
Sc type: B = C
Non-Sc-type: B < C
No explanation
Sc-type: B > C
Non-Sc-type: B = C
No explanation
No explanation
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
Possible chromosomal composition in
strains B and C derived from results
B C
: Centromere: Region of no synteny to S. cerevisiae, > 20kb
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
1
- 1
0
: Sc-type chromosome : Hybrid type chromosome : non-Sc-type chromosome
74
the right side of non-Sc-type chromosome VIII with the relative ratio of about 1.5
times higher in strain B compared to strain C was observed (Fig. 9). Based on the
fact that the chromosome VIII of lager brewers’ existed in three types: Sc, non-Sc,
Sc/non-Sc hybrid types (Fig. 6) (Rainieri et al., 2006), we assumed that strains B and
C contained the same copy numbers of non-Sc-type chromosome VIII. In addition,
strain B contained more copies of hybrid type while strain C possessed a higher copy
number of Sc-type chromosome VIII (Fig. 9).
Comparison of chromosomes I, VI, IX of two strains B and C showed that the
relative hybridisation signal ratios of one type of these chromosomes (either Sc-type
or non-Sc-type) was the same while the other type was different. Since there was no
hybrid type in these chromosomes, we assumed that the two compared strains
possessed the same copy numbers in either Sc-type or non-Sc type but different
copy numbers in the other type of these chromosomes. For example, relative Sc-type
hybridisation ratio on chromosomes IX was equal to 0 while of relative non-Sc-type
hybridisation signal ratio was higher in strain B. Thus, we presumed that regarding
chromosome IX, strain B had the same copy numbers of Sc-type and more copies of
non-Sc-type compared to strain C. As the relative non-Sc-type hybridisation signal
ratio was about 1.5 times higher in strain B, it was supposed strain B contained 3
copies while strain C possessed 2 copies of the non-Sc-type chromosome IX (Fig. 9).
Chromosomes III, X, XV displayed the same relative hybridisation ratio patterns.
Relative hybridisation signal of either Sc-type or non-Sc type were the same in both
strains while that of other type (either Sc or non-Sc) was only similar in one region of
these chromosomes (Fig. 9). So far, we do not have a reasonable explanation for
these differences on chromosomal level.
75
4.2.1.2 Identification of differences in copy number of known genes relevant to
diacetyl formation and flocculation
The genes identified as different in the studied strains at genomic level were
incorporated to yeast pathways listed by Saccharomyces Genome Database (SGD)
using the Microsoft Access software. As diacetyl is the by-product of the valine
biosynthetic pathway, we at first focused on comparing the hybridisation ratios of
genes encoding enzymes participated in this pathway in the selected strains (Fig.
10). Based on the results obtained in microarray-based CGH, differences related to
the valine biosynthetic pathway were detected including Sc-IVL6, Sc-BAT1 and
non-Sc-BAT1 (Fig. 10). All of these ORFs were found to be different in both B vs C
and B vs A pairwise comparisons (Fig. 10). This result is consistent to the fact that
the diacetyl production of strain B strikingly differences from those of strains A and
C (Fig. 4).
Among the studied strains, strain A produced the slightly higher of level of diacetyl
than strain C while diacetyl production of strain B was much lower than those of
strains A and C. In accordance to this fact, the level of Sc-BAT1 hybridisation signal
was highest in strain A and lowest in strain B (Fig. 10). Since the difference in
hybridisation signal of each ORF is directly related to the difference in gene copy
numbers, from the hybridisation ratios, I deduced that strains A, C and B might
contain three, two and one copies of Sc-BAT1 ORF, respectively. The pairwise
comparison of hybridisation signals of non-Sc-BAT1 and Sc-ILV6 genes also
suggested that the three studied strains possessed different copy numbers of these
genes. Concretely, strain B might contain one copy whilst strains A and C might
contain two or three copies of Sc-ILV6 gene. These differences in gene copy number
might be the reason for the different expression levels of these genes and thus for
difference in diacetyl phenotype of the studied strains. However, this assumption has
to be confirmed by incorporating with results obtained from other analyses at
transcriptome and proteome levels.
76
Pyruvate
Dihydroxyisovalerate
ILV2
α-acetolactate
ILV6
α-ketoisovalerate
Valine
ILV5
ILV3
BAT1
BAT2
DIACETYL
0
250
500
750
0
200
400
600
A B C A B C
Hybridisation signal
Hybridisation signal
0
150
300
450
A B C A B C
0
400
800
600
200
Hybridisation signal
Hybridisation signal
0
250
500
750
1000
A B C
Hybridisation signal
0
200
400
600
0
200
400
600
A B C A B C
Hybridisation signal
Hybridisation signalv
0
200
400
600
800
0
250
500
750
A B C A B C
Hybridisation signal
Hybridisation signalv
0
150
300
450
0
150
300
450
600
A B C A B C
Hybridisation signalv
Hybridisation signalv
Sc-type genes Non-Sc-type genes
)
))
Pyruvate
Dihydroxyisovalerate
ILV2
α-acetolactate
ILV6
α-ketoisovalerate
Valine
ILV5
ILV3
BAT1
BAT2
DIACETYL
0
250
500
750
0
250
500
750
0
250
500
750
0
200
400
600
0
200
400
600
0
200
400
600
A B CA B C A B CA B C
Hybridisation signal
Hybridisation signal
0
150
300
450
0
150
300
450
0
150
300
450
A B CA B C A B CA B C
0
400
800
600
200
0
400
800
600
200
0
400
800
600
200
Hybridisation signal
Hybridisation signal
0
250
500
750
1000
0
250
500
750
1000
0
250
500
750
1000
A B CA B C
Hybridisation signal
0
200
400
600
0
200
400
600
0
200
400
600
0
200
400
600
0
200
400
600
0
200
400
600
A B CA B C A B CA B C
Hybridisation signal
Hybridisation signalv
0
200
400
600
800
0
200
400
600
800
0
200
400
600
800
0
250
500
750
0
250
500
750
0
250
500
750
A B CA B C A B CA B C
Hybridisation signal
Hybridisation signalv
0
150
300
450
0
150
300
450
0
150
300
450
0
150
300
450
600
0
150
300
450
600
0
150
300
450
600
A B CA B C A B CA B C
Hybridisation signalv
Hybridisation signalv
Sc-type genes Non-Sc-type genes
))
))))
Fig. 10 Microarray DNA hybridisation signals of genes encoding enzymes of the valine
biosynthetic pathway in the studied lager brewers’ yeast strains. The microoarray-based CGH
among the studied strains was performed using bottom fermenting yeast DNA microarray. The stars
indicate significant differences in the studied strains. Signinicant differences were selected with
change p-value ≤ 0.002 or ≥ 0.998. ILV2, ILV6: acetohydroxyacid synthase, ILV5: reductoisomerase,
ILV3: dihydroxyacid dehydratase; BAT1, BAT2: branched-chain amino acid transaminase.
77
Besides the differences directly related to diacetyl formation, global analysis at
genome level also revealed differences in copy numbers of flocculation genes in the
studied strains (Table 4). The comparison between the highly flocculent strain B and
other strains revealed many differences including Sc-FLO1, non-Sc-FLO1, Sc-FLO8,
non-Sc-FLO9, non-Sc-FLO10, Sc-FLO11, non-Sc-FLO11 (Table 4). Among those,
the sequence of non-Sc-FLO10 (id-fix Lg_4227_1) was found to be present only in
strain B. In addition, except for Sc-FLO1 (id-fix Sc_2439_1) other genes showed
higher hybridisation ratios in strain B than in both strains C and A, suggesting that
strain B contained more copies number of flocculation genes than the two other
strains. No difference was detected in the comparison between strains A and C. This
result was reasonable since flocculation phenotype of the strains A and C was
similar.
Table 4. Differences of genes belonged to the flocculation gene family identified at genome
level in the studied lager yeast strains. The DNA analysis was performed using microarray-based
CGH by means of bottom fermenting yeast DNA microarray. Black sheets indicate changes in the
pairwise strain comparison. Changes were selected based on a change p-value ≤ 0.002 or ≥ 0.998.
Some sequences, e.g. FLO1 appeared several times as the results being divided in two halves or
more because of frame shift and stop codon mutation. Gene was detected as present based on a
detection p-value of 0.05.
Gene detection Hybridisation ratios
Gene name Gene type id_fix sys.gene A B C A/B A/C
C/B
FLO1 Sc Sc_2439_1
YAR050W
P P P -1.5 1.1 -1.7
FLO1 Sc Lg_6958_1 YAR050W
P P P 1.9 -1.1
1.9
FLO1 non-Sc Lg_1617_1 YAR050W
P P P 1.6 1.1 1.6
FLO1 non-Sc Lg_3309_1 YAR050W
P P P 1.2 1.0 1.2
FLO1 non-Sc Lg_3309_1 YAR050W
P P P 1.3 -1.1
1.3
FLO1 non-Sc Lg_4229_1 YAR050W
P P P 1.3 1.0 1.4
FLO8 Sc Sc_4622_1
YER109C P P P 1.1 1.0 1.2
FLO9 Sc Lg_427_1 YAL063C P P P -1.4 1.1 -1.4
FLO9 non-Sc Lg_934_2 YAL063C P P P 1.2 1.0 1.2
FLO10 non-Sc Lg_4227_1 YKR102W
A P A
FLO10 non-Sc Lg_4227_2 YKR102W
P P P 1.2 1.0 1.2
FLO11 Sc Sc_1180_1
YIR019C P P P 1.3 1.0 1.3
FLO11 Sc Sc_1180_2
YIR019C P P P 1.3 1.0 1.2
FLO11 Sc Lg_2876_1 YIR019C P P P -1.4 1.1 -1.5
FLO11 non-Sc Lg_2876_2 YIR019C P P P -1.3 1.0 -1.4
*
)
Gene detection: A means absent, P means present
78
To conclude, by using microarray-based CGH, we identified differences regarding
the type and copy number of some chromosomes between the studied strains.
Differences directly related to diacetyl and flocculation phenotypes were determined.
The differences relating to diacetyl phenotype will be integrated with the results
obtained at the level of transcriptome and proteome for identifying target genes for
reduction of diacetyl production in lager yeast.
4.2.2
Transcriptome level: Microarray-based comparative transcriptome
analysis
4.2.2.1 Transcriptome analysis using bottom fermenting yeast DNA microarray
To identify the genetic basis for the phenotypic differences at transcriptome level,
we carried out comparative microarray-based transcriptome analysis on the selected
lager brewers’ yeast strains. The transcriptome profiles of lager brewers’ yeast
strains were examined using yeast cells harvested at apparent extract of 8% of the
main fermentation. Customized bottom fermenting yeast DNA microarrays were used
in order to differentiate the expressional levels of Sc-type and non-Sc-type genes.
Transcriptome of each strain was studied in technical triplicates, i.e. total RNA of
each strain was isolated in three replicates from cells harvested from a single
fermentation. Mean values and statistical analysis for triplicate microarray
hybridisations were calculated for each strain. In each pairwise comparison, the
logged average fold-changes were calculated and statistical analysis was performed
using Cyper-T approach with a false discovery rate of 0.001.
Using this criterion, 1851 significant differences were identified at transcriptome
level among the three studied strains. The number and categories of significant
differences in each pairwise comparison are shown in Table 5. In general, the
comparison of strain B to either strain A or C revealed more than 1000 significant
differences while the number of significant differences in the A vs C comparison was
much lower (rougly 300). More than 600 common significant differences were found
79
in both A vs B and B vs C comparison (data not shown). The results were consistent
to the fact that strains A and C are genetically related. Transcriptional profiling also
revealed differences regarding SGD-type genes i.e. Sc-type ORFs from SGD
database that are not present in WH34/70, in the studied strains (Table 5).
Furthermore, differences reagrding several S. patorianus sequences published in the
Genbank, intergenic regions and other sequences, which showed similarity to
S. cerevisiae proteins by NCBIblastX homology searching, were also detected at
transcriptomic level. There were no differences regarding mitochondrial genes
among the studied strains.
Table 5. Number of significant differences identified at transcriptome level among the studied
lager yeast strains via analysis using bottom fermenting yeast DNA microaray. Significant
differences were chosen with a false discovery rate of 0.001. Sc: S. cerevisiae type; non-Sc:
non-S. cerevisiae type; SGD-type: Sc-type sequences from SGD database which are not detected in
the genome sequence of the lager brewers’ yeast strain WH34/70. Others: sequences from
S. pastorianus published in Genbank, intergenic sequences, sequences which showed similarity to
S. cerevisiae proteins by NCBIblastX homology searching
Pairs Type genes No of significant differences
Sc-type 488
Non-Sc-type 471
SGD-type 20
A vs B
Others 197
1176
Sc-type 86
Non-Sc-type 181
SGD-type 3
A vs C
Others 68
338
Sc-type 560
Non-Sc-type 436
SGD-type 16
C vs B
Others 224
1236
80
4.2.2.2 Identification of differences at transcriptional level of genes relevant to
diacetyl and flocculation
The identified differentially expressed ORFs between the studied strains have
been further analysed using the Microsoft Access software and Saccharomyces
Genome Database (SGD). The differentially expressed ORFs in the studied strains
were incorporated into 90 of the total 156 biological pathways listed in SGDs. Valine
biosynthetic pathway (from which diacetyl is formed as a by-product) was one of the
pathways which showed many differences in the studied strains. These included
Sc-ILV6, Sc-BAT1, non-Sc-BAT1 and non-Sc-BAT2 (Fig. 11). For data evaluation,
the result of genomic analysis of genes participated in valine biosynthesis were
incorporated with result of transcriptome level (Fig. 11).
It is shown from (Fig. 11) that Sc-BAT1 gene and Sc-ILV6 transcript were both
less abundant in strain B in comparison to strains A and C. In addition, the
abundance of non-Sc-BAT1 ORF and transcript were both lowest A and highest
strain B. This positive correlation between genome and transcriptome analyses
suggested that the differences regarding abundance of Sc-ILV6, Sc-BAT1 and
non-Sc-BAT1 transcripts among the studied strains might result from the differences
in the gene copy numbers.
Besides that, the result at transcriptional level regarding non-Sc-BAT2 ORF did
not match that of genome analysis. In the genome analysis, non-Sc-BAT2 gene was
identified as “not change” in the studied strains. However, at mRNA level it was about
two-fold higher in strain B compared to strains A and C. Thus, the difference of
non-Sc-BAT2 gene cannot be due to the difference in gene copy numbers but it
rather resulted from other factors such as differences in transcriptional regulation,
promoter strength or mRNA stability.
81
Fig. 11 Microarray-based DNA and microarray-based normalised RNA hybridisation signals of
genes encoding enzymes involved in the valine biosynthetic pathway in the studied strains.
The experiments were performed using bottom fermenting yeast DNA microarrays. In each strain
analysis, the hybridisation signals at DNA level were obtained from a single array hybridisation whilst
the normalized hybridisation signals at RNA level were the average of normalized signals from
triplicate array hybridisations with standard deviations. The stars indicate significant differences in the
studied strains. Significant differences at DNA level were chosen with change p-value ≤ 0.002 or
≥ 0.998. Significant differences at RNA level were selected based on a false discovery rate of 0.001.
Nor.hyb.signal: normalised hybridisation signal. For each ORF, the Sc-type and non-Sc-type signals
were generated. ILV5 non-Sc-type ORF and transcript are absent since the bottom fermenting yeast
DNA microarray does not contain probeset for this ORF. ILV2, ILV6: acetohydroxyacid synthase, ILV5:
reductoisomerase, ILV3: dihydroxyacid dehydratase; BAT1, BAT2: branched-chain amino acid
transaminase
Pyruvate
Dihydroxy
isovalerate
ILV2
α-acetolactate
ILV6
α-keto
isovalerate
Valine
ILV5
ILV3
BAT1
BAT2
DIACETYL
0
200
400
600
Hybridisation signal
0
200
400
600
0
200
400
600
Hybridisation signal
0
400
800
600
200
0
400
800
600
200
Hybridisation signal
0
150
300
450
600
0
150
300
450
600
Hybridisation signalv
Sc-type Non-Sc-type
0
200
400
600
0
200
400
600
Hybridisation signalv
0
250
500
750
0
250
500
750
Hybridisation signalv
0
250
500
750
0
250
500
750
Hybridisation signal
0
150
300
450
0
150
300
450
Hybridisation signal
0
250
500
750
1000
0
250
500
750
1000
0
250
500
750
1000
Hybridisation signal
0
150
300
450
0
150
300
450
Hybridisation signalv
0
200
400
600
0
200
400
600
Hybridisation signal
0
200
400
600
800
0
200
400
600
800
Hybridisation signal
DNA level
0.0
0.3
0.6
0.9
1.2
0.3
0.6
0.9
1.2
Nor.hyb.signal
0
0.3
0.6
0.9
1.2
Nor.hyb.signal
0.00
0.25
0.50
0.75
1.00
1.25
0.00
0.25
0.50
0.75
1.00
1.25
0.00
0.25
0.50
0.75
1.00
1.25
Nor.hyb.signal
0.00
0.25
0.50
0.75
1.00
1.25
0.00
0.25
0.50
0.75
1.00
1.25
Nor.hyb.signal
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
Nor.hyb.signal
0.0
0.3
0.6
0.9
1.2
0.0
0.3
0.6
0.9
1.2
0.0
0.3
0.6
0.9
1.2
Nor.hyb.signal
0.0
0.3
0.6
0.9
1.2
0.0
0.3
0.6
0.9
1.2
0.0
0.3
0.6
0.9
1.2
Nor.hyb.signal
0.0
0.3
0.6
0.9
1.2
0.0
0.3
0.6
0.9
1.2
0.0
0.3
0.6
0.9
1.2
A B C
Nor.hyb.signal
0.0
0.5
1.0
1.5
0.0
0.5
1.0
1.5
Nor.hyb.signal
A B CA B C
A B CA B C A B CA B C
0.0
0.3
0.6
0.9
1.2
0.0
0.3
0.6
0.9
1.2
0.0
0.3
0.6
0.9
1.2
Nor.hyb.signal
0.00
0.25
0.50
0.75
1.00
1.25
0.00
0.25
0.50
0.75
1.00
1.25
Nor.hyb.signal
A B CA B CA B CA B C A B CA B CA B CA B C
A B CA B C A B CA B CA B CA B C A B CA B C
A B CA B CA B CA B C A B CA B CA B CA B C
A B C A B C
A B CA B C A B CA B CA B CA B C A B CA B C
RNA level
Sc-type Non-Sc-type
)) )
)))) ))
)
82
Table 6. Differences in flocculation genes identified at level of genome and transcriptome in the
studies strains. The genome and transcriptome analyses were performed using comparative
microarray analysis by means of bottom fermenting yeast DNA microarray. Bold sheets indicate
significant changes or differences in a pairwise comparison. Changes at DNA level were chosen with
a change p-value ≤ 0.002 or ≥ 0.998. Significant differences at RNA level were selected based on a
false discovery rate of 0.001. Some sequences appeared more than once as the result of being
divided in two halves or more because of frame shift or stop codon mutation. For gene detection and
transcript flag, P means present; A means absent; A, P: ambiguously detected (either present or
absent).
DNA level RNA level
Gene detection Hybridisation ratios Transcript flags Hybridisation ratios
Gene
name Gene
type id_fix A B C A/B A/C C/B A B C A/B A/C C/B
FLO1 Sc Sc_2439_1 P P P -1.5 1.1 -1.7 P P P -6.3 -1.8 -3.5
FLO1 Sc Lg_6958_1 P P P 1.9 -1.1 1.9 P P P 1.6 1.3 1.2
FLO1 non-Sc Lg_1617_1 P P P 1.6 1.1 1.6 P,A P,A P 1.1 -1.2 1.4
FLO1 non-Sc Lg_3309_1 P P P 1.2 1.0 1.2 P P P 1.2 1.2 1.0
FLO1 non-Sc Lg_3309_1 P P P 1.3 -1.1 1.3 P P P 1.0 1.2 -1.3
FLO1 non-Sc Lg_4229_1 P P P 1.3 1.0 1.4 P P P 1.4 1.3 1.1
FLO8 Sc Sc_4622_1 P P P 1.1 1.0 1.2 P P P 1.2 1.0 1.2
FLO9 Sc Sc_2692_1 P P P 1.0 1.0 -1.1 P P P -91.3 1.2 -111.0
FLO9 Sc Lg_427_1 P P P -1.4 1.1 -1.4 P P P -37.3 1.8 -68.5
FLO9 non-Sc Lg_934_2 P P P 1.2 1.0 1.2 P P P -1.2 1.3 -1.6
FLO10 non-Sc Lg_4227_1 A P A P,A P P,A
FLO10 non-Sc Lg_4227_2 P P P 1.2 1.0 1.2 P P P 1.4 1.3 1.1
FLO11 Sc Sc_1180_1 P P P 1.3 1.0 1.3 P P P -2.4 1.4 -3.3
FLO11 Sc Sc_1180_1 P P P 1.1 -1.1 1.1 P P P -2.5 1.3 -3.3
FLO11 Sc Sc_1180_2 P P P 1.3 1.0 1.2 P,A P P,A -2.8 1.8 -4.9
FLO11 Sc Lg_2876_1 P P P -1.4 1.1 -1.5 P P P -2.1 1.0 -2.2
FLO11 non-Sc Lg_2876_2 P P P -1.3 1.0 -1.4 A P,A A -2.2 1.2 -2.6
Regarding the flocculation phenotype, no difference was observed at transcription
level in the comparison between strains A and C. However, many flocculation genes
were found to be up-regulated in strain B compared to strains A and C (Table 6).
These included Sc-FLO1, Sc-FLO9, non-Sc-FLO9, non-Sc-FLO10, Sc-FLO11,
non-Sc-FLO11. Among those, Sc-FLO9 (id-fix Sc_2692_1) was more than 90-fold up
regulated in strain B in compared to strains A and C. The result was consistent with
the fact that strain B flocculates much earlier than strains A and C during the main
fermentation (see II.4.1.3). In this data set, the detected difference of non-Sc-FLO10
(id-fix Lg_4227_1) in the studied strains was due to the absence of this sequence in
83
genome of strains A and C. Interestingly, in A vs B and B vs C comparisons, some
genes which were not more than 2-fold changed at genomic level showed big
differences at transcriptional level, i.e. Sc-FLO1, id_fix Sc_2349_1 (6.3-fold lower in
A vs B, 4-fold lower in C vs B); Sc-FLO9, id-fix Sc_2692_1 (91-fold lower in A vs B,
111-fold lower in C vs B), non-Sc-FLO9, id-fix Lg_427_1 (37-fold lower in A vs B,
69-fold lower in C vs B comparisons). This result suggested that the differences in
the expression levels of these genes should result from factors such as differences in
transcriptional regulation, promoter strength or mRNA stability.
To sum up, the microarray-based comparative transcriptome analysis revealed
many significant differences directly relevant to diacetyl and flocculation behavious in
the studied strain. The identified differences directly related diacetyl formation
included Sc-ILV6, Sc-BAT1, non-Sc-BAT1 and non-Sc-BAT1. The differences
regarding the abundance of these transcripts might be responsible for the difference
in diacetyl production of the studied strains. Nevertheles, integration of the result at
transcriptome level with the result obtained at proteome levels is needed for the
identification of potential target genes for reducing diacetyl production in lager
brewers’ yeast.
4.2.3
Proteome level: Comparative proteome analysis using two-dimensional
gel electrophoresis
4.2.3.1 Identification of protein spots which showed significant different
intensities among the three studied lager brewers’ yeast strains
To identify genetic basis for differences in diacetyl production of the studied lager
brewers’ yeast strains at proteome level, two-dimensional (2D) gel analysis was
performed. Protein samples were isolated from the cells harvested at apparent
extract of 8% from two independent fermentations. The 2D gel electrophoresis was
carried out in triplicate as described in the materials and methods section. The 2D
gels of brewers’ yeast strains were scanned and analysed using the Delta 2D
84
software version 4.0. Separation of total protein by 2D gel electrophoresis resulted in
the detection of 520 spots in total in the 2D gel of each studied strain (Fig. 12).
Statistical analysis was performed allowing a standard deviation ≤ 30% for each spot
from the three replicates and a p-value of 0.05 or below in the Student’s t-test at
pairwise comparison. Using these criteria, 9 spots were identified as more than 2-fold
significantly different among the three studied lager yeast strains. Among these
9 spots, 4 spots were identified in A vs B comparison, 4 spots were identified as
significantly different in B vs C comparison and 1 spot was identified as significantly
different in the comparison between B vs A/C (Fig. 13).
26
170
130
95
72
55
43
34
17
11
Mr (kDa)
3
pI
7
26
170
130
95
72
55
43
34
17
11
Mr (kDa)
26
170
130
95
72
55
43
34
17
11
26
170
130
95
72
55
43
34
17
11
Mr (kDa)
3
pI
7
Fig. 12 Two-dimensional gel image of a lager brewers’ yeast strain (strain C) at apparent extract
of 8%. Gel was stained with Comassie Brillant Blue.
4.2.3.2 Mass spectrometry identification of differentially expressed proteins
Protein spots identified as different in studied strains were excised from the gels
and characterised using MALDI-TOF MS. Among the 9 spots analysed by
MALDI-TOF MS, spots 1-8 were positively identified with a protein score of at least
100 and a sequence coverage ≥ 30% for duplicate identification (Table 7). Spot 9
85
was identified as Eno2p with protein sequence coverage of about 25%. As this
sequence coverage was quite low, it was considered as being ambiguously
identified. In addition, it is well known that lager brewers’ yeast is a hybrid of two
Saccharomyces yeast, thus it proteome may contain two version of certain proteins,
Sc-type and non-Sc-type. As a protein database of lager brewers’ yeast has not been
available, Mascot search engine of peptides generated in the MALDI-TOF
experiment was only performed with the protein database of S. cerevisiae. Due to
that fact, a protein of non-Sc may not be detectable as in the case of spot 9.
Table 7. MALDI-TOF mass spectrometry identification of significantly different protein spots
detected in the proteome comparisons of the three studied lager yeast strains
Spots
Accession No
Name
pI
MW
(kD) Mowse
Score Protein
coverage (%) Identification
1 gi|6321968 ENO2 5.67 46.89 576 61.09 Eno2p [S. cerevisiae]
2 gi|6322790 FBA1 5.51 39.60 942 68.8 Fructose 1,6-bisphosphate aldolase
3 gi|10383781 PGK1 7.11 44.71 514 66.35 3-phosphoglycerate kinase
4 gi|151942494 HSP31 5.26 25.62 500 82.7 heat-shock protein
5 gi|10383781 PGK1 7.11 44.71 266 52.59 3-phosphoglycerate kinase
6 gi|151944335 SSB2 5.32 66.55 546 53.51 stress-seventy subfamily B protein
7 gi|151941387 SSA1 5 69.60 506 46.88 stress-seventy subfamily A protein
8 gi|6321968 ENO2 5.67 46.89 422 38.92 Eno2p [S. cerevisiae]
9 gi|157830958 ENO2 6.04 46.60 447 25% Ambiguously identified
The eight spots identified by MALDI-TOF included two stress-seventy subfamily
proteins (Ssa1p, Ssb2p), one heat shock protein (Hsp31p) and three proteins
participating in the glycolytic pathway (Fba1p, Eno2p, Pgk1p). Some spots at
different positions on the 2D gels were identified as one protein, i.e. spots 1 and 8
were identified as Eno2p, spots 3 and 5 were detected as Pgk1p. Spot 5 and spot 8
appeared respectively to be fragments of Pgk1p and Eno2p since their molecular
weights were smaller than those of corresponding proteins.
86
Fig. 13 Significant differentially expressed protein spots among the three studied lager
brewers’ yeast strains at apparent extract of 8% detected in 2D gels. Significant differences were
identified with p-value of 0.05, a spot standard deviation ≥ 30% and 2 fold-change regulated.
Spot 1-4: significant differences between strain A and B. Spot 6-9: Significant differences between
strain C and B. Spot 5 was the significant difference identified between strain B vs strains A and C.
0.0
0.1
0.2
0.3
0.4
0.5
Average %
volume
0
0.1
0.2
0.3
0.4
0.5
Average %
volume
0.0
0.2
0.4
0.6
Average %
volume
0.00
0.03
0.06
0.09
0.12
0.15
Average %
volume
0.00
0.03
0.06
0.09
0.12
Average %
volume
0.0
0.2
0.4
0.6
Average %
volume
0.00
0.03
0.06
0.09
Average %
volume
0.00
0.02
0.04
0.06
0.08
Average %
volume
0.00
0.03
0.06
0.09
0.12
0.15
Average %
volume
Spot 1
Spot 2
Spot 3
Spot 8
Spot 4
Spot 5
Spot 9
Spot 6
Spot 7
Strain A Strain B
Strain C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
87
All of these 9 spots were shown to be different in the comparison of strain B
versus strains A and C. Spots 1, 2, 3, 4, 5 and 8 were shown to be at the highest
level while spots 6, 7, 9 were least abundant in strain B. No significant difference was
observed in proteome comparison between strains A and C. In addition to the global
genetic analyses at transcriptomic and genomic levels, the result obtained at
proteome level once again fitted well to the fact that strains A and C are closely
related and are phenotypically different from strain B.
To conclude, global analysis at proteome level only revealed differences
regarding glycolytic and stress proteins. None of these differences directly related to
the diacetyl and flocculation phenotypes.
88
4.3 Sc-ILV6, a potential novel target gene for reducing diacetyl production
in brewers’ yeast
Global molecular analyses only revealed differences directly relevant to diacetyl
and flocculation phenotypes at level of genome and transcriptome. Among those,
Sc-ILV6 is one of the most promising targets for the reduction of diacetyl production.
Significant differences regarding this ORF were identified at both genome and
transcriptome levels in the studied strains. Sc-ILV6 is proposed to encode a
regulatory subunit of acetohydroxyacid synthase (AHAS, Ilv2p). AHAS catalyzes the
conversion of pyruvate to α-acetolactate which is the precursor of diacetyl.
Compared to strains A and C, strain B contains a lower copy number of Sc-ILV6 ORF
and a lower level of Sc-ILV6 transcript. The lower concentration of Sc-ILV6 mRNA in
strain B might be responsible for a lower activity of AHAS and thus for the lower
levels of α-acetolactate and diacetyl in this strain.
4.4 In vitro acetohydroxyacid synthase activity in the studied strains
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Strain B Strain C
AHAS activity (µmol/mg.h)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Strain B Strain C
AHAS activity (µmol/mg.h)
Fig. 14 In vitro activity of acetohydroxyacid synthase (AHAS or Ilv2p) in strains B and C. AHAS
activity was measured using permeabilized cell proteins. For this experiment, fermentation was carried
out in 100 ml bottle using 11.38
ο
P brewers’ wort at 12
ο
C. Permeabilized cell proteins were prepared
from cells harvested at apparent extract of 8% of the fermentation. Results shown are mean values of
two independent experiments including standard deviations.
89
To demonstrate the hypothesis that the lower level of Sc-ILV6 transcript in strain
B compared to strain A and C could lead to the lower AHAS activity in this strain
(section II.4.3), in vitro AHAS activity in two strains B and C was measured. In vitro
AHAS activity in strain B was only slightly different from that in strain C (about 87%)
(Fig. 14). In spite of this result, we decided to delete the Sc-ILV6 in strain C to verify
its role in the formation of diacetyl in brewers’ yeast.
4.5 Disruption of Sc-ILV6 in strain C for reduced diacetyl production
4.5.1
Deletion of the first Sc-ILV6 gene copy in strain C: generation of a
Sc-ilv6
∆
∆∆
∆
single deletion strain
Disruption of Sc-ILV6 in strain C was mediated via homologous recombination.
The first copy of Sc-ILV6 in strain C was deleted by using a loxP-kanMX-loxP
disruption cassette amplified from the plasmid pUG6 (Gueldener et al., 2002).
The disruption cassette consisted of the kanMX module which was flanked by two
direct repeated 34 bp loxP sequence. The kanMX module itself consisted of kan
r
gene flanked by the TEF promoter and terminator originated from filamentous fungus
Ashybya gossypii. The kan
r
gene of the kanMX module originated from E.coli
transposon Tn 903. Insertion of the loxP-kanMX-loxP module into the genome of
brewers’ yeast would confer the resistance to the antibiotic Geneticin 418 (G418) to
this strain. Once inserted to the yeast genome, the kanMX marker can be rescued by
transformation of a plasmid expressing Cre recombinase under control of GAL
promoter. Upon growth on galactose, Cre recombinase action at the repeated loxP
sites would excise the kanMX marker, leaving behind one loxP sequence at the site
of the Sc-ILV6 disruption cassette.
90
Fig. 15 Disruption of the first copy of Sc-ILV6 in strain C. Gray bars indicate 45 bp homolougous
regions needed for mediating the homologous recombination. The integration was verified by
diagnostic PCR using primers P3 and P4.
Sc-ilv6
∆
∆∆
∆
single deletion mutant
-69
0 930
0 930
ORF end
-45
596551
ATG
loxP-kanMX-loxP
Sc-ILV6 (930bp)
P3
loxP-kanMX-loxP
1.73 kb
P4
0930
ORF endATG 594
P3 P4
0.63 kp
3. Diagnostic PCR
P2
loxP-kanMX-loxP
pUG6
P1
loxP-kanMX-loxP
Disruption cassette (1707bp)
1. Amplification of the Sc-ILV6
disruption cassette
2. Homologous recombination
Wild-type
-69 ORF endATG 594
P3 P4
0.63 kb
-69
0 930
ORF endATG 594
P3 P4
0.63 kb
1
st
copy of Sc-ILV6
2
nd
copy of Sc-ILV6
1
st
copy of Sc-ILV6
is deleted
2
nd
copy of Sc-ILV6
Sc-ilv6
∆
∆∆
∆
single deletion mutant
-69
0 930
0 930
ORF end
-45
596551
ATG
loxP-kanMX-loxP
Sc-ILV6 (930bp)
P3
loxP-kanMX-loxP
1.73 kb
P4
0930
ORF endATG 594
P3 P4
0.63 kp
3. Diagnostic PCR
P2
loxP-kanMX-loxP
pUG6
loxP-kanMX-loxP
pUG6
P1
loxP-kanMX-loxPloxP-kanMX-loxP
Disruption cassette (1707bp)
1. Amplification of the Sc-ILV6
disruption cassette
2. Homologous recombination
Wild-type
-69 ORF endATG 594
P3 P4
0.63 kb
-69
0 930
ORF endATG 594
P3 P4
0.63 kb
-69 ORF endATG 594
P3 P4
0.63 kb
-69
0 930
ORF endATG 594
P3 P4
0.63 kb
1
st
copy of Sc-ILV6
2
nd
copy of Sc-ILV6
1
st
copy of Sc-ILV6
is deleted
2
nd
copy of Sc-ILV6
91
The 1.7 kb loxP-kanMX-loxP cassette was generated via PCR using
recombinogenic primers P1 and P2 (see materials and methods). The forward P1
and reverse P2 primers respectively contained 18 bp and 20 bp homologous to
plasmid pUG6 at 3’ end which were necessary for PCR amplification of the disruption
cassette. At the 5’ ends, the two primers contained 45 bp homologous to brewers’
yeast sequences. The 45 bp at the 5’end of P1 primer was designed as the
sequence flanking the left side of Sc-ILV6 gene while the 45 bp at 5’end of P2 primer
was designed as the sequence at position -551 till -596 of Sc-ILV6 gene (Fig. 15).
These homologous sequences were selected as specific for Sc-ILV6 replacement to
prevent the disruption of non-Sc-ILV6 ORF. The sequence of non-Sc-ILV6 ORF as
well as its flanking sequences was kindly provided by Dr. Kodama from Suntory Ltd.
Homologous recombination would result in the replacement of 596 bp of the Sc-ILV6
coding region, starting from the ATG start codon, by the disruption cassette.
Fig. 16 Diagnostic PCR to check the correct single deletion mutant Sc-ilv6
∆
∆∆
∆
in the 8 selected
clones. NC: negative control, strain C, which contains two copies of Sc-ILV6
1.73 kb
0.66 kb
NC
Clone 1
Clone 2
Clone 3
Clone 4
Clone 5
Clone 6
Clone 7
Clone 8
92
The PCR product was introduced into strain C using the PEG/Lithium acetate
transformation method. The transformants carrying loxP-kanMX-loxP cassette were
selected on YED plate supplemented with 17.5 µg/ml of G418. After that, the
transformants were transferred to a new YED plate with a higher concentration of
G418 (50 µg/ml). Eight randomly chosen transformants which were able to grow on
the latter medium were then selected for further investigation. The disruption of
Sc-ILV6 in these transformants was confirmed by diagnostic PCR using primers P3
and P4 (see materials and methods). The primer P3 located 49 bp upstream and the
primer P4 spread from position 551 to 596 of the Sc-ILV6 (Fig. 15). Microarray CGH
revealed that strain C contained two copies of Sc-ILV6. Thus, diagnostic PCR using
these two primers in the mutant strain where one copy of Sc-ILV6 was deleted would
result in two fragments of 1.7 kb (loxP-kanMX-loxP band) and 660 bp (the remaining
Sc-ILV6 band) (Fig. 15). In contrast, the untransformed strain C led to the
amplification of only one fragment of 660 bp (control band). The results showed that
all of the eight selected transformants carried the correct single deletion of
Sc-ILV6 (Fig. 16).
4.5.2
Deletion of the second Sc-ILV6 gene copy: generation of a
Sc-ilv6
∆
∆∆
∆
/Sc-ilv6
∆
∆∆
∆
double deletion strain
One of the Sc-ilv6∆ single deletion mutants was used for the generation of the
Sc-ilv6∆/Sc-ilv6∆ double deletion strain. The disruption of the second copy of Sc-ILV6
in strain C was also mediated via homologous recombination. The plasmid pUG66
was used as the template for PCR amplification of the second disruption cassette
loxP-ble
r
-loxP (Gueldener et al., 2002). The reaction was carried out with the same
primers P1 and P2 previously used to amplify the loxP-kanMX-loxP cassette. The ble
r
module of the second disruption cassette was composed of the ble
r
gene which was
flanked by the TEF promoter and terminator of Ashybya gossypii. The ble
r
gene of
the plasmid pUG66 originated from transposon Tn5. The transformant harbouring the
ble
r
module in the genome would render resistance to the antibiotic phleomycine.
93
Fig. 17 Disruption of the second copy of Sc-ILV6 in the Sc-ilv6
∆
∆∆
∆
single deletion strain. Gray bars indicate 45 bp of homolougous regions for mediating
homologous recombination. The integration of disruption cassette was diagnosed by PCR using primers P3 and P4. In the positive transformant, the
loxP-ble
r
-loxP cassette replaced the second copy Sc-ILV6, diagnostic PCR resulted in two bands of 1.3 kb (loxP-bler-loxP containing band) and 1.73 kb
(loxP-kanMX-loxP containing band). In the negative transformant, loxP-bler-loxP cassette replaced loxP-kanMX-loxP cassette, diagnostic PCR resulted in two
bands of 660 bp (Sc-ILV6 containing band) and 1.3 kb (loxP-ble
r
-loxP containing band)
ORF end
P1
P2
loxP-bler-loxP
-45 596 551
ATG
Sc-ILV6
Positive transformant
Diagnostic PCR
loxP-kanMX-loxP
0
loxP-bler-loxP
loxP-kanMX-loxP
P4
P3
P4
P3
1.3 kb
1.73 kb
Negative transformant
Diagnostic PCR
-45 596 551
Sc-ILV6
P1
P2
loxP-kanMX-loxP
loxP-bler-loxP
loxP-bler-loxP
Sc-ILV6
-69
P3 P4
0
594
P3 P4
0
930
660 bp
1.3 kb
ATG
930
930
ORF end
94
Fig. 18 Diagnostic PCR to confirm the correct Sc-ilv6
∆/
∆/∆/
∆/
Sc-ilv6
∆
∆∆
∆
double deletion in the selected
clones
The second disruption cassette loxP-ble
r
-loxP was transformed into the strain
Sc-ilv6∆ using the PEG/Lithium acetate method. As the two disruption cassettes
loxP-kanMX-loxP and loxP-ble
r
-loxP were amplified by using the same primers,
integration of the second disruption cassette loxP-ble
r
-loxP into the genome of the
strain Sc-ilv6∆ would appear in two possibilities: i) loxP-ble
r
-loxP replaced
loxP-kanMX-loxP cassette and the 2
nd
copy of Sc-ILV6 remained (negative
transformant) and ii) the loxP-kanMX-loxP cassette substituted the 2
nd
copy of
Sc-ILV6, generating the double Sc-ilv6∆/Sc-ilv6∆ strain (positive transformant)
(Fig. 17). The Sc-ilv6∆/Sc-ilv6∆ double deletion mutants harbouring both
loxP-ble
r
-loxP and loxP-kanMX-loxP cassettes rendered resistance to both G418 and
phleomycine. Thus, for the selection of the double deletion mutants, transformants
were first selected on a YEPD plate containing 17.5 µg/ml phleomycine and then
replica selected on a YED plate containing 50 µg/ml of G418. On the first selective
plates (YEPD plus 17.5 µg/ml phleomycine), 130 transformants were obtained. The
transfer of 80 of these transformants onto the second selective plates (YED plus
Clone 1
Clone 2
Clone 3
Clone 4
Clone 5
Wildtype-C
Sc
-
ilv6
∆
1.7 kb
1.3 kb
0.66 kb
95
50µg/ml of G418) led to the growth of 9 transformants. Five of these transformants
were used for diagnostic PCR using primers P3 and P4 (Fig. 17).
In the original Sc-ilv6∆ single deletion mutant containing one copy of Sc-ILV6 and
the loxP-kanMX-loxP cassette, diagnostic PCR would result in two bands of 1.7 kb
and 660 bp (Fig. 17). In the negative transformant which the loxP-kanMX-loxP
cassette was replaced by loxP-ble
r
-loxP cassette, it resulted in two bands of 1.3 kb
and 660 bp. The desired double deletion transformant carried no copy of Sc-ILV6
and thus, led to an amplification of two fragments of 1.3 kb and 1.7 kb (Fig. 17)
Diagnostic PCR revealed that only one of these five tested transformants was the
correct Sc-ilv6
∆
/Sc-ilv6
∆
double deletion mutant (Fig. 18) (clone 5). In addition, PCR
of one transformant (clone 1) resulted in three bands of 1.7 kb, 1.3 kb and 660 bp.
Thus, this clone appeared to be the mixture of the correct double deletion strain
either with the single deletion strain or with the negative transformant. Diagnostic
PCR of the other three transformants only resulted in one band of 660 bp. The
appearance of only one band of 660 bp indicates that these three clones contained
no disruption cassette. However, why these clones were able to grow on G418 and
phleomycin selective media remains questionable to us.
As aforementioned, brewers’ yeast containes two versions of many genes
(Sc-type and non-Sc-type). The microarray-based CGH analysis revealed that all the
studied brewers’ yeast strains in this work contained both Sc-ILV6 and non-Sc-ILV6
ORFs. Sequence analysis showed that these two ORFs had the same length and
were about 86% identical (Dr. Yukiko Kodama, personal information). To verify that
the disruptions were specific for Sc-ILV6 ORF, diagnostic PCR using non-Sc primers
P5 and P6 (see material and methods) was set up for the generated Sc-ilv6∆ single
and Sc-ilv6∆/Sc-ilv6∆ double deletion strains (Fig. 19). The primer P5 located 46 bp
upstream while primer P6 was designed as the sequence located from position 575
to 593 of the non-Sc-ILV6 ORF. PCR amplification using non-Sc primers P5 and P6
gave no product in S.cereviae strain BY4741. In contrast, it resulted in one band of
660 bp (non-Sc-type ILV6 band) in the wild-type strain C, the single deletion mutant
96
1.73 kb
1.30 kb
0.66 kb
0.66 kb
BY4741
Strain C
Sc
-
ilv6
∆
Sc
-
ilv6
∆
/
Sc
-
ilv6
∆
Sc-ilv6∆ and the double deletion mutant Sc-ilv6∆/Sc-ilv6∆. The result confirmed the
correct disruption of Sc-ILV6 ORF in the mutant strains.
Fig. 19 Diagnostic PCR to check the correct disruption of Sc-ILV6 instead of the wrong
disruption of non-Sc-ILV6 ORF. Upper gel: Diagnostic PCR using the Sc-type primers P3 and P4.
Lower gel: Diagnostic PCR using non-Sc-type primers P5 and P6.
97
4.5.3
In vitro acetohydroxyacid synthase activity in strains Sc-ilv6
∆
∆ ∆
∆
and
Sc-ilv6
∆
∆∆
∆
/Sc-ilv6
∆
∆∆
∆
As previsouly mentioned, Ilv6p has been proposed as an enhancer of
acetohydroxyacid synthase (Ilvp2, AHAS). Thus, the activity of AHAS in the Sc-ilv6
single (Sc-ilv6
∆
) and in the double deletion strain (Sc-ilv6
∆/
Sc-ilv6
∆
) was measured
(Fig. 20). It was showed that the AHAS activity in the two mutants was slightly
different in comparison to that in the reference strain C. Concretely, the AHAS activity
in the Sc-ilv6 single and double deletion strains was about 97% and 90% compared
to that in the reference strain C. The AHAS activity of strain B which contains one
copy of Sc-ILV6 ORF was about 82% compared to strain C.
Fig. 20 In vitro activity of acetohydroxyacid synthase (AHAS or Ilv2p) in strains B, C,
Sc-ilv6
∆
∆ ∆
∆
and Sc-ilv6
∆
∆∆
∆
/Sc-ilv6
∆
∆∆
∆
. AHAS activity was measured using permeabilized cell proteins. For
this experiment, fermentation was carried out in 100 ml bottle using 11.38
o
P brewers’ wort at 12
o
C.
Permeabilized cell proteins were prepared from cells harvested at apparent extract of 8% of the
fermentation. Results shown are mean values of two independent experiments including standard
deviations.
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
B C Sc-ilv6
∆
∆∆
∆
Sc-ilv6
∆
∆∆
∆
/Sc-ilv6
∆
∆∆
∆
AHAS activity (
µ
µ
µ
µ
mol/mg.h
)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
B C Sc-ilv6
∆
∆∆
∆
Sc-ilv6
∆
∆∆
∆
/Sc-ilv6
∆
∆∆
∆
AHAS activity (
µ
µ
µ
µ
mol/mg.h
)
98
4.5.4
Fermentation characteristics and vicinal diketone production of strains
Sc-ilv6
∆
∆ ∆
∆
and Sc-ilv6
∆
∆∆
∆
/Sc-ilv6
∆
∆ ∆
∆
under the laboratory scale fermentation
The diacetyl productions of Sc-ilv6 deletion mutant strains were investigated
under laboratory scale fermentation (see materials and methods). For each strain,
the fermentations were done in duplicate. Diacetyl was always measured when the
apparent extract varying from 8.3 to 8.8% was reached. The results showed that the
production of 2,3-pentanedione of the two mutant strains was similar to that of the
reference strain C. In contrast, a strong decrease in diacetyl formation was observed
in the two mutant strains (Fig. 21). The diacetyl production of strains Sc-ilv6
∆
and
Sc-ilv6
∆
/Sc-ilv6
∆
were reduced to about 87% and 60% compared to that of the
wildtype strain C. Nevertheless, vicinal diketone production of strain B which contains
one copy of Sc-ILV6 was still much lower than those of strains Sc-ilv6
∆
and
Sc-ilv6
∆
/Sc-ilv6
∆
. Diacetyl and 2,3-pentanedione productions of strain B at apparent
extract of 8% was about 9% and 25% compared to those of strain C, respectively.
The diacetyl concentration of these strains had a correlation to the AHAS activity
even though the difference in AHAS activity was low.
Fig. 21 Vicinal diketone production by strains B, C and Sc-ilv6
∆
∆ ∆
∆
and Sc-ilv6
∆
∆∆
∆
/Sc-ilv6
∆
∆∆
∆
. Diacetyl
and 2,3-pentanedione were measured at apparent extract varying from 8.3 to 8.8%. Yeast strains
were fermented in 100 ml bottles using 11.38
o
P brewers’ wort at 12
o
C. Total diacetyl: the sum diacetyl
and its precursor α-acetolacte in the wort medium; total 2,3-pentanedione: the sum of
2,3-pentanedione and its precursor α-aceto-α-hydroxybutyrate in the medium.
0.0
0.5
1.0
1.5
2.0
2.5
B C Sc-ilv6
∆
∆∆
∆
Sc-ilv6
∆
∆∆
∆
/Sc-ilv6
∆
∆∆
∆
Total diacetyl
concentration (mg/l)
Total 2,3-pentanedione
concentration (mg/l)
0.0
0.5
1.0
1.5
2.0
2.5
B C Sc-ilv6
∆
∆∆
∆
Sc-ilv6
∆
∆∆
∆
/Sc-ilv6
∆
∆∆
∆
Total diacetyl
concentration (mg/l)
Total 2,3-pentanedione
concentration (mg/l)
99
4.5.5
Fermentation characteristics and vicinal diketone production of strain
Sc-ilv6
∆
∆∆
∆
/Sc-ilv6
∆
∆∆
∆
under industrially relevant brewery fermentation
Fig. 22 Fermentation performance of strain Sc-ilv6
∆
∆∆
∆
/Sc-ilv6
∆
∆∆
∆
in comparison to the wild type
strain C. Fermentation was carried out in 30-litre tank under industrially relevant brewing conditions
(10.5
o
C, 11.38
o
P brewers’ wort). A) Concentration of non-sedimented cells, B) Time courses of
apparent extract, C) Time courses of pH
The growth of the strain Sc-ilv6∆/Sc-ilv6∆ was virtually similar to that of the
reference strain C during the main fermentation (Fig. 22A). At the end of
fermentation, the strain Sc-ilv6∆/Sc-ilv6∆ showed a slower sedimentation compared
to the reference strain C (Fig. 22A).
0
1
2
3
4
5
0 2 4 6 8 10
Fermentation time (d)
Conc. of non
-
sedimented
cells
(x 10
7
cells/ml)
Wildtype, strain C
Sc-ilv6∆/Sc-ilv6∆
A)
Apparent extract (%)
0
2
4
6
8
10
12
0 1 2 3 4 5 6 7 8
Fermentation time (d)
4.0
4.2
4.4
4.6
4.8
5.0
5.2
5.4
0 2 4 6 8 10
Fermentation time (d)
pH
B) C)
0
1
2
3
4
5
0 2 4 6 8 10
Fermentation time (d)
Conc. of non
-
sedimented
cells
(x 10
7
cells/ml)
Wildtype, strain C
Sc-ilv6∆/Sc-ilv6∆
Wildtype, strain C
Sc-ilv6∆/Sc-ilv6∆
A)
Apparent extract (%)
0
2
4
6
8
10
12
0 1 2 3 4 5 6 7 8
Fermentation time (d)
4.0
4.2
4.4
4.6
4.8
5.0
5.2
5.4
0 2 4 6 8 10
Fermentation time (d)
pH
B) C)
100
By the end of the main fermentation, the consumption of wort sugars in the strain
Sc-ilv6∆/Sc-ilv6∆ was slightly slower in comparison to the wildtype. The reference
strain C needed seven days while it took the strain Sc-ilv6∆/Sc-ilv6∆ eight days to
reach wort attenuation (Fig. 22B).
The time courses of the pH values were not influenced by the disruption of
Sc-ILV6 ORFs (Fig. 22C).
2,3 pentanedione conc. (mg/l)
0.0
0.1
0.2
0.3
0.4
024681012
Apparent extract (%)
0.0
0.1
0.2
0.3
0.4
0.5
024681012
Apparent extract (%)
Diacetyl con. (mg/l)
A) B)
Wild type, strain C Sc-ilv6∆/Sc-ilv6∆
2,3 pentanedione conc. (mg/l)
0.0
0.1
0.2
0.3
0.4
024681012
Apparent extract (%)
0.0
0.1
0.2
0.3
0.4
0.5
024681012
Apparent extract (%)
Diacetyl con. (mg/l)
A) B)
Wild type, strain C Sc-ilv6∆/Sc-ilv6∆
2,3 pentanedione conc. (mg/l)
0.0
0.1
0.2
0.3
0.4
024681012
Apparent extract (%)
0.0
0.1
0.2
0.3
0.4
0.5
024681012
Apparent extract (%)
Diacetyl con. (mg/l)
A) B)
Wild type, strain C Sc-ilv6∆/Sc-ilv6∆
Wild type, strain C Sc-ilv6∆/Sc-ilv6∆
Fig. 23 Vicinal diketone production by strain Sc-ilv6
∆
∆∆
∆
/Sc-ilv6
∆
∆ ∆
∆
in comparison to the wild type
strain C. Fermentation was carried out in 30-litre tank under industrially relevant brewing conditions
(10.5
o
C, 11.38
o
P brewers’ wort). A) Diacetyl concentration during fermentation. B) 2,3-pentanedione
concentration during fermentation
The diacetyl content of the Sc-ilv6
∆
/Sc-ilv6
∆
double deletion strain was
investigated under brewing condition. Diacetyl contents were measured at different
apparent extract during the main fermentation. By the end of fermentation (at
apparent extract of 2.8%), the diacetyl production of strain Sc-ilv6∆/Sc-ilv6∆ was
reduced by 65% in comparison to that of the wild type.
However, the complete disruption of Sc-ILV6 gene in strain C only resulted in a
slighter change in the final production of 2,3-pentanedione. Compared to the wild
type, the strain Sc-ilv6∆/Sc-ilv6∆ showed a 26% reduction in 2,3-pentanedione
concentration by the end of the fermentation (Fig. 22).
101
4.5.6
Determination of flavour-relevant products in green beer produced by
strain Sc-ilv6
∆
∆∆
∆
/Sc-ilv6
∆
∆ ∆
∆
Compared to the reference beer, the green beer produced by the mutant
Sc-ilv6∆/Sc-ilv6∆ showed a 25% reduction in acetaldehyde concentration (Table 8).
Moreover, an increase in production of some acetate esters (ethyl acetate,
2-phenylethyl acetate, isoamyl acetate) and ethyl esters (ethyl caprate, ethyl
formiate) was observed. A decrease in production of some fusel alcohols (iso-butyl
alcohol, isoamyl alcohol, 2-phenylethyl alcohol) and fatty acid (capric acid) was also
observed. However, the concentrations of these by-products were in the beer normal
range.
102
Table 8. Analysis of green beers produced by strain Sc-ilv6
∆
∆∆
∆
/Sc-ilv6
∆
∆ ∆
∆
and the reference
strain C (harvested at apparent extract of 3%). Fermentation was carried out in 3 litre glass
fermenters under conditions relevant to industrial brewing fermentation (11.38
o
P, 12
o
C). The results
obtained for the strain Sc-ilv6
∆
/Sc-ilv6
∆
are mean values of two dependent experiments including
standard deviations. Bold lines indicate alteration in production of some by-products
Strains C Sc-ilv6∆
∆∆
∆/Sc-ilv6∆
∆∆
∆
Ethanol (g l
-1
) 38.4 36.95 ± 0.35
pH 4.29 4.27 ± 0.06
Acetaldehyde (mg l
-1
) 5.25 3.93 ± 0.04
Organic acids
Acetic acid (mg l
-1
) 187.11 176.12 ± 49.12
Butyric acid (mg l
-1
) 1.47 1.77 ± 0.11
Fusel alcohols
1-propanol (mg l
-1
) 21.2 23.19 ± 0.49
Active amyl alcohol (mg l
-1
) 15.6 9.9 ± 0.4
Isoamyl alcohol (mg l
-1
) 49.3 31.45 ± 0.95
Iso-butyl alcohol (mg l
-1
) 13.4 8.2 ± 1.2
2-phenylethyl alcohol (mg l
-1
) 48.1 36.1 ± 6.8
Esters
Ethyl acetate (mg l
-1
) 3.34 7.91 ± 2.09
Isoamyl acetate (mg l
-1
) 0.35 0.46 ± 0.09
Isobutyl acetate (mg l
-1
) 0 0 ± 0
2-Phenylethyl acetate (mg l
-1
) 1.05 1.47 ± 0.09
Ethyl formiate (mg l
-1
) 0.82 0.65 ± 0.03
Ethyl butyrate (mg l
-1
) 0.04 0.050 ± 0.005
Ethyl caproate (mg l
-1
) 0.07 0.080 ± 0.005
Ethyl caprate (mg l
-1
) 1.45 2.81 ± 0.12
Fatty acids
Isovalerate (mg l
-1
) 6.48 6.59 ± 0.56
Caproate (mg l
-1
) 5.17 4.75 ± 0.79
2-Ethyl capronate (mg l
-1
) 4.7 4.26 ± 0.17
Caprate (mg l
-1
) 4.61 3.47± 0.02
Caprylate (mg l
-1
) 9.28 7.86 ± 0.31
Phenylacetic acid (mg.l
-1
) 0.86 0.95 ± 0.14
103
5 Discussion
So far, the improvement of brewers’ yeast has been mostly attempted by classical
genetic manipulation or rational metabolic engineering. In this study, we used the
inverse metabolic engineering approach to identify novel target genes for
optimisation of lager brewers’ yeast strains. The predominant target addressed in the
current work is the reduction of diacetyl production. Diacetyl causes an unwanted
butter-like flavour in beer. The reduction of diacetyl to an acceptable level in beer
during maturation requires a lot of time. A lager brewers’ strain producing low level of
dicetyl would be of great advantage for industry since it helps to accelerate the beer
brewing process, i.e. it would save alot of time and storage capacity.
Comparative global molecular analyses of different lager brewers’ yeast strains
producing various levels of diacetyl revealed that Sc-ILV6 was one of the potential
novel target genes for reduction of diacetyl production. Sc-Ilv6p is proposed to be the
regulatory subunit of Ilv2p. The latter enzyme is directly involved in diacetyl formation
since it catalyses the reaction to convert pyruvate into α-acetolacte which is the
precursor of diacetyl. The significant differences regarding Sc-ILV6 were found at
both level of genome (gene copy number) and transcriptome (mRNA concentration)
in the studied strains. The resulting difference regarding Sc-ILV6 expression level in
the studied strains might be the reason for the difference in activity of Ilv2p (AHAS)
and thus might have led to the difference in diacetyl production in the studies strains.
The in vitro activity of Ilv2p was shown to be slightly different in the studied strains.
Despite of this fact, I subsequently deleted two copies of Sc-ILV6 in one production
industrial brewers’ yeast strain to verify its role in diacetyl formation.
The disruption of the Sc-ILV6 in brewers’ yeast only led to a slight reduction in
AHAS activity and 2,3-pentanedione concentration. However, a significant decrease
in diacetyl production was achieved. Growth and wort sugar consumption of strain
Sc-ilv6
∆
/Sc-ilv6
∆
were slightly slower than those of the reference strain. Analysis of
green beer produced by the Sc-ilv6
∆
/Sc-ilv6
∆
mutant revealed a slight change in
concentrations of acetaldehyde, some fusel alcohols, esters and fatty acids.
104
5.1 Global genetic analyses of the three lager brewers’ yeast strains
producing different levels of diacetyl
To identify the crucial differences determining the various levels of diacetyl in the
three selected lager yeast strains, comparative global analyses were performed at
the level of genome, transcriptome and proteome. The genome and transcriptome
analyses were carried out by means of comparative microarray analysis using
customized bottom fermenting yeast DNA microarrays. The proteomes of the
selected strains were studied using 2D gel electrophoresis and mass spectrometry
(MALDI-TOF MS). To obtain an insight into the transcriptome and proteome at the
same point in time, mRNA and protein samples were always isolated from cells
harvested at apparent extract of 8% which corresponds to the early phase of the
main fermentation.
Global analysis at genomic level using bottom fermenting yeast DNA probes
(22,977 probesets including coding and intergenic regions) revealed thousands of
sequences which showed to be significantly differently abundant in the there studied
strains. In each pairwise comparison at the genome level, a sequence was
designated as “increased” if the calculated change p-value was ≤ 0.002 and as
“decreased” if the change p-value was ≥ 0.998. Global genome analysis revealed
that the chromosome patterns of the two strains A and C were in fact quite similar
and were different from that of strain B (Fig. 8). Expression analysis at transcriptome
level using the same bottom fermenting yeast DNA microarray resulted in the
identification of total 1851 transcripts whose levels were significant different in the
studied strains (false discovery rate of 0.001). Among those, 338 transcripts were
found to be differentially abundant between strains A and C (Table 5). In contrast, the
number of significant differences identified in comparison between strain B and the
two other strains are quite high i.e 1176 (B vs A) and 1236 (B vs C). In addition, the
transcriptome analysis releaved many common differences (more than 600) in the B
vs A and B vs C comparisons. The results obtained at genome and transcriptome
105
levels were consistent to the fact that strains A and C are genetically closely related
and phenotypic different from strain B.
The differences identified at both genome and transcriptome levels included those
which have an obvious link to the diacetyl and flocculation phenotypes of the
investigated strains. In addition to Sc-ILV6, non-Sc-BAT1 which can be linked to the
diacetyl production pathway, many differences directly related to flocculation were
obtained (e.g. those regarding FLO1, FLO8 and FLO9). In a number of cases, the
identified genes showed significantly different abundances at both levels, i.e. gene
copy number and mRNA level. The results indicate that genome and transcriptome
analyses with the use of bottom fermenting yeast DNA microarray is a useful tool to
study the genetic basis of brewing relevant phenotypic differences of lager brewers’
yeast strains.
Microarray-based genome and transcriptome analyses of lager brewers’ yeast in
previous studies were solely performed with S. cerevisiae arrays (Olesen et al., 2002;
James et al., 2003; Bond et al., 2004; Pope et al., 2007). Thus, these analyses only
allowed the identification of the differences regarding Sc-type genes and transcripts
while the non-Sc-type ones were not accessible. The use of bottom fermenting yeast
DNA microarrays (kindly provided by Suntory Ltd.) in the current study allowed a
more profound study of significant lager brewer’s yeast strain’s differences and
revealed whether such a difference was based on Sc-type or non-Sc-type gene or
transcript. Therefore, it led to more reliable transcriptome and genome data which
are important for the identification of real target genes for strain improvement.
Furthermore, the integration of results obtained for a certain gene at both genome
and transcriptome levels allowed us to draw a hypothesis about the factors affecting
its differential expression in the studied strains. For example, both Sc-ILV6 copy
number and Sc-ILV6 transcript (mRNA) showed a low abundance in strain B (low
diacetyl producer) compared to the other two strains. This positive correlation
between gene copy number and transcript abundant level suggested that the higher
Sc-ILV6 transcript level in the strain B might result from a higher gene copy number.
106
In contrast, non-Sc-BAT2 showed the same gene copy number among all strains but
a higher mRNA level in strain B (Fig. 11). One can conclude that the higher Sc-BAT2
transcript level in strain B must result from other factors than copy number such as
transcriptional activity and regulation or mRNA stability. The combination of global
analyses at different genetic molecular levels is therefore crucial for a better
understanding of the differences in gene expression and for deducing rational
strategies for the strain improvement.
The expression patterns of genes directly related to diacetyl and flocculation
behaviour at transcriptome level did not match at all the result at proteome level.
Indeed, no differences directly related to diacetyl and flocculation phenotype was
identified when comparing the proteomes of three studied strains. In contrast to
genome and transcriptome analyses, the number of differences identified in
proteome comparisons of the studied strains was very low. It only revealed 6 proteins
which were at least 2-fold differentially expressed (up- or down-regulated) in the
studied strains with a t-test p-value ≤ 0.05 and spot standard deviation of
30% (Table 7). These differentially abundant proteins included 3 glycolytic enzymes
(Fba1p, Eno2p and Pgk1p) and 3 stress proteins (Ssa1p, Ssb2p and Hsp31p)
(Table 7, Fig. 13). Among those, 4 proteins were found to be differentially expressed
in A vs B comparison including Eno2p, Fba1p, Pgk1p and Hsp31p. In addition,
4 proteins were identified as differentially expressed in B vs C comparison including
Ssb2p, Ssa1p, Eno2p and Pgk1p. No difference was identified when proteome of
strain A was compared to that of strain C. Similar to the analyses at genome and
transcriptome levels, the comparison at peoteome level fitted well to the fact that
strains A and C are closely related and phenotypically different from strain B. As
diacetyl and flocculation phenotypes of strain B was considerably different from those
of strains A and C, the significant differences identified in both A vs B and B vs C
comparison were considered to be most important for the phenotype, i.e. Eno2p and
Pgk1p. Interestingly, no differences regarding ENO2 and PGK1 mRNA abundances
were found in transcriptome analysis. The result suggested that the differences in
107
protein concentration must have resulted from differences in translational efficiency
or protein stability.
Different levels of the two glycolytic enzymes Eno2p and Pgk1p might even be
related to the differences in diacetyl production in the studied lager yeast strains. It
was found that strain B has a lower amount of two glycolytic enzyme Eno2p and
Pgk1p in comparison to strains C and A. The lower abundance of these enzymes in
strain B (compared to strains A and C) could have led to a lower level of pyruvate
intermediate concentration in strain B. As pyruvate is the substrate for the reaction
catalysed by AHAS to form the precursors for vicinal diketones (diacetyl and
2,3-pentanedione), the lower level of pyruvate could result in a lower level of vicinal
diketone production in strain B.
The negative correlation between the identified differentially expressed mRNA
species and proteins can be due to a number of reasons e.g. the sensitivity and
reproducibility of the used methods, post-translational modifications of protein or time
incompatible between levels of expression of a transcript and a protein. In this study,
the relative low number of differences identified at proteome level might result from
the common limitations of the protein analysis method used in this study. The
detection limit of a protein in a 2D gel is estimated to be at least 1000 protein copies
per cell (Mijalski et al., 2005). In contrast, the DNA microarray allows detecting less
than one copy of mRNA per 20 yeast cells (Lockhart et al., 1996; Wodicka et al.,
1997). Thus, low abundant proteins might not have been detected in our 2D gels.
Since the 2D gel electrophoresis requires many stages of manual work, its
reproducibility is lower than that of the microarray method. In this study, protein
analysis was carried out in biological triplicates (cell samples for protein isolation
were taken in triplicate from two independent fermentations) while the transcriptome
analysis was performed in technical triplicates (total RNA samples were isolated in
triplicate from cells harvested from one single fermentation). The difference in the
triplicate models used in the transcriptome and proteome analyses could account for
the lower reproducibility of our proteome analysis compared to that of transcriptome
108
analysis. Due to these facts, only a low number of significant differences (regarding
high abundant proteins) were identified at proteom level. In addition, flocculation
proteins are integral cell wall proteins and thus might have been lost during the
protein extraction. Due to these reasons, the strain specific differences directly
related to diacetyl production and flocculation (found at transcriptome level) might not
be detectable at proteome level.
5.2 Identification of potential novel target genes for reduction of diacetyl
production: Sc-ILV6, Sc-BAT1, non-Sc-BAT1, non-Sc-BAT2
As aforementioned, diacetyl is a by-product of the valine biosynthetic pathway,
which is formed from the non-enzymatic decarboxylation of α-acetolactate. When
transcriptomes of the three studied lager brewers’ yeast strains were compared,
several significant differences were identified for genes whose products are involved
in the valine biosynthesis pathway (Sc-ILV6, Sc-BAT1, non-Sc-BAT1 and
non-Sc-BAT2) and differences in their expression could thus be directly relevant for
diacetyl formation (Fig. 11). Among those, the levels of Sc-BAT1 and Sc-ILV6
transcripts were respectively 3-fold and 4-fold lower in strain B compared to strains A
and C. Strain B produced the lowest level of diacetyl while strain A has a slightly
higher diacetyl production than strain C (Fig. 4). The lower levels of Sc-ILV6 and
Sc-BAT1 mRNA in strain B might therefore be responsible for the lower level of
diacetyl production in strain B compared to A and C. In addition, the level of
non-Sc-BAT1 and non-Sc-BAT2 transcripts were about 2-fold lower in strain A
compared to strains B and C. These differences might thus be related to the highest
diacetyl production in strain A.
ILV6 which is one of the differentially expressed genes in the studied strains, is
assumed to encode a regulatory subunit of Ilv2p. This protein confers
acetohydroxyacid synthase (AHAS) activity and is responsible for the formation of
α-acetolactate from pyruvate. The different studies that have resulted in the
109
assumption that Ilv6p is a regulatory subunit of Ilv2p are as follows. First, three active
AHAS isozymes (AHASI, AHASII, and AHASIII) exist in most bacteria, e.g. E. coli
and Samonella typhimurium. All three isozymes have a tetrameric structure (α2β2)
and each consists of two subunits of different molecular weights (small subunits:
≈ 10-17 kD, large subunits ≈ 60 kD) (Pang and Duggleby, 1999). It was demonstrated
that AHAS activity was conferred by the large subunits while the small subunits act
as a regulator by enhancing the activity of the catalytic subunits and conferring
sensitivity to feedback inhibition by valine of the enzyme. In S. cerevisiae, only one
isozyme of AHAS has been characterized which is also composed of a catalytic
subunit (Ilv2p) and a regulatory subunit (Ilv6p) (McCourt and Duggleby, 2006). The
disruption of ILV6 in S. cerevisiae led to the insensitivity of AHAS to feedback
inhibition by valine (Cullin et al., 1996). Although the role of yeast Ilv6p in feedback
regulation of the enzyme is obvious, its role as an enhancer for catalytic subunit Ilv2p
is still unclear. It was showed that an ilv6 null mutant did not show any change in
AHAS in vitro activity (Cullin et al., 1996). Nevertheless, Pang et al. (1999)
overexpressed yeast Ilv2p and Ilv6p in E. coli and carried out in vitro reconstitution of
these two proteins. In fact, the activity of the reconstituted enzyme in high salt
concentration buffer was 7-fold higher than that of the catalytic subunit (Ilv2p) alone.
Global genetic analysis at mRNA level revealed that Sc-ILV6 transcript was about
4-fold less abundant in strain B compared to strains A and C. As Ilv6p probably acts
as an enhancer for Ilv2p in yeast, the lower expression level of Sc-ILV6 in strain B
may lead to lower the AHAS activity and accordingly to the lower level of diacetyl
production in this strain. Regarding this fact, Sc-ILV6 is a potential target gene for the
reduction of diacetyl production in lager brewers’ yeast. The influence of genetic
modification of Sc-ILV6 on diacetyl production and fermentation performances of
lager brewers’ yeast will be discussed in details in the latter parts.
Apart from the detected difference regarding Sc-ILV6 transcript abundance, strain
B also contained a lower level of Sc-BAT1 transcript (approximately 2-fold) compared
to strains A and C (Fig. 11). BAT1 encodes a mitochondrial branched-chain amino
110
acid (BCAA) transaminase which is required both for the BCAA biosynthetic pathway
and the BCAA degradation via the Ehrlich pathway (Fig. 24). In yeast, this enzyme is
highly expressed during logarithmic phase and repressed during stationary phase. In
the first period of the fermentation which can be considered as the “logarithmic
phase”, brewers’ yeast cells do not need to produce BCAAs since they can be taken
up from the medium. Thus, the higher level of Sc-BAT1 expression in strains A and C
could lead to a higher valine consumption and to a higher level of α-keto-isovalerate
formed from the valine degradation in these strains. The higher level of
α-keto-isovalerate could result in a lower metabolic flux through the valine
biosynthetic pathway. This could lead to a higher level of intermediate α-acetolactate
and accordingly a higher level of diacetyl in strains A and C compared to strain B.
Strain A produced the highest level of diacetyl among the studied strains. It was
shown that non-Sc-BAT1 and non-Sc-BAT2 transcripts were least abundant in strain
A compared to strains B and C (Fig. 11). The difference regarding non-Sc-BAT2
expression level might be responsible for difference in diacetyl production between
strain A and C or in other words, for the highest diacetyl production in strain A
compared to strains B and C. BAT2 encodes a cytosolic BCAA transaminase which
is, like BAT1, involves in the BCAA biosynthetic pathway and the BCAA degradation
via the Ehrlich pathway (Fig. 24). In contrast to BAT1, this enzyme is repressed
during logarithmic phase and highly expressed during stationary phase. Starting from
the middle of the fermentation which corresponds to the “balance phase”, brewers’
yeast has to synthesize BCAAs needed for the cellular activities via the BCAA
biosynthetic pathway as the BAACs in wort medium is depleted. Thus, the lower level
of non-Sc-BAT2 expression in this period in strain A could lead to the lower level of
valine formation in this strain. As AHAS is feedback inhibited by valine, the lower
level of valine formation might result in the lesser extent of valine inhibition to AHAS
and thus in the higher activity of AHAS in strain A. The higher activity of this enzyme
in turn might lead to a higher level of diacetyl production in strains A and C. In this
study, the transcriptomes of the studied strains were analysed during the early stage
111
of fermentation, which corresponded to the late logarithmic growth of the brewers’
yeast. Thus, the lower level of non-Sc-BAT2 in strain A during the early stage might
not be necessarily reflect the lower level of this transcript in strain A at the middle
stage (the balance phase) of the fermentation. To confirm this hypothesis, the
investigation of the expression levels of non-Sc-BAT2 in the studied strains during
the middle stage of fermentation is required.
In addition to the low level of non-Sc-BAT2 transcript, strain A also showed a
lower level of non-Sc-BAT1 transcript compared to strain C. As discussed above, the
lower level of BAT1 expression may lead to the higher level of diacetyl production.
Although strain A has the lower level of non-Sc-BAT1 transcript, it still produced
higher level of diacetyl than strain C. One possible explanation for this result is the
low level of non-Sc-BAT1, which might result in the lower level of diacetyl production,
could not compensate the higher expression level of non-Sc-BAT2, which might
result in the higher level of diacetyl production in strain A. Thus, in total strain A
produced higher level of diacetyl than strain C. It was demonstrated that Sc-type and
non-Sc-type ORFs are about 85% homologous (Kodama et al., 2006). Therefore, the
negative correlation between non-Sc-BAT1 expression level and diacetyl production
in strain A could be also explained by the possible difference in the function of
Sc-Bat1p and non-Sc-Bat1p. Due to that reason, the lower level of non-Sc-BAT1
might not be related to higher level of diactyl production in strain A.
In brewers’ yeast, so far there has not been any study of ILV6, BAT1 and BAT2
expression level in relation to diacetyl formation. Regarding these above analyses,
these Sc-ILV6, Sc-BAT1, and non-Sc-BAT2 ORFs can be considered as the novel
potential target genes for reducing diacetyl production in yeast. The roles of these
genes in diacetyl formation can be verified via the deletion of Sc-BAT1 and Sc-ILV6
in strains A or C or via overexpression of non-Sc-BAT2 genes in strain A. In addition,
the role of non-Sc-BAT1 in diacetyl formation can also be verified via the
overexpression of this gene in any selected lager yeast strain. Besides the genetic
modification of every single gene, the simultaneous manipulation of both Sc-BAT1
112
and Sc-ILV6 or both non-Sc-BAT1 and non-Sc-BAT2 or of even all of these genes
might be crucial to verify their role in diacetyl production.
In this thesis, I performed the genetic modification of one of these promising
target genes, Sc-ILV6 to reduce diacetyl production in lager yeast. For that purpose,
Sc-ILV6 gene was disrupted in strain C which is the production strain of an industrial
German brewery. The success in reducing diacetyl production in this strain may lead
to the creation of a brewers’ strain producing low level of diacetyl which is useful for
beer brewing. In addition, the BAT1 and BAT2 genes were addressed in a parallel
work carried out by Lysann Strack.
113
Fig. 24 Branched-chain amino acid biosynthesis and degradation in yeast and some related
metabolites. ILV1: threonin deaminase; ILV2, ILV6: acetohydroxyacid synthase, ILV3: dihydroxyacid
dehydratase; ILV5: reductoisomerase; BAT1, BAT2: branched-chain amino acids transaminase;
ATF1, ATF2: alcohol acetyltransferese; IAH1: isoamyl acetate-hydrolyzing esterase
Pyruvate
α-acetotolactate
Dihdroxy-
isovalerate
α-keto-
isovalerate
Diacetyl
Valine
Isobutyl
acetate
Isobutyl
alcohol
α-keto-isocaproate
Leucine
a-ketobutyrate
α-aceto-α
hydroxybutyrate
Dihydroxy-β-
methyl valerate
α-keto-β-methyl
valerate
Active amyl
alcohol
D-amyl acetate
pyruvate pyruvate
CO
2
CO
2
2,3 pentanedione
Isoamyl alcohol
Isoamyl acetate
Isoleucine
ILV2
ILV6
ILV5
ILV3
BAT1
BAT2
ILV2
ILV6
ILV5
ILV3
BAT1
BAT2
Threonine
ILV1
BAT1
BAT2
Isoamyl alcohol
Active amyl alcohol
Isobutyl
alcohol
ATF1
ATF2
ATF1
ATF2
IAH1
IAH1
ATF1ATF2
IAH1
Pyruvate
α-acetotolactate
Dihdroxy-
isovalerate
α-keto-
isovalerate
Diacetyl
Valine
Isobutyl
acetate
Isobutyl
alcohol
α-keto-isocaproate
Leucine
a-ketobutyrate
α-aceto-α
hydroxybutyrate
Dihydroxy-β-
methyl valerate
α-keto-β-methyl
valerate
Active amyl
alcohol
D-amyl acetate
pyruvate pyruvate
CO
2
CO
2
2,3 pentanedione
Isoamyl alcohol
Isoamyl acetate
Isoleucine
ILV2
ILV6
ILV5
ILV3
BAT1
BAT2
ILV2
ILV6
ILV5
ILV3
BAT1
BAT2
Threonine
ILV1
BAT1
BAT2
Isoamyl alcohol
Active amyl alcohol
Isobutyl
alcohol
ATF1
ATF2
ATF1
ATF2
IAH1
IAH1
ATF1ATF2
IAH1
114
5.3 Impact of Sc-ILV6 disruption on in vitro AHAS activity and vicinal
diketone reduction
The disruption of one and two copies of Sc-ILV6 gene in the reference strain C
only led to insignificant changes in in vitro AHAS activity (Fig. 20). The AHAS activity
of the Sc-ilv6 single deletion mutant was about 97% compared to that of the
reference strain C (Fig. 20). Here, only the mean values were taken into account.
Roughly, 10% reduction in AHAS activity was obtained in the Sc-ILV6 double
deletion strain. These results are consistent with the work of Cullin (1996) in which
the disruption of ILV6 ORF also did not result in the alteration of AHAS activity in a
laboratory S. cerevisiae strain. In contrast to the insignificant differences regarding
AHAS activity, a remarkable reduction in diacetyl production in Sc-ilv6 mutants was
observed (Fig. 21, Fig. 23). The analysis of wort at apparent extract of 8% in
laboratory scale fermentations showed that the diacetyl production of strains Sc-ilv6
∆
and Sc-ilv6
∆
/S-ilv6
∆
was reduced by 13% and 40%, respectively compared to that of
the wildtype. In addition, under conditions relevant to industrial brewery
fermentations, the diacetyl production of strain Sc-ilv6
∆
/Sc-ilv6
∆
showed a decrease
of about 65% compared to the reference strain C at the end of the primary
fermentation (Fig. 23).
As previously mentioned, diacetyl is formed from the non-enzymatic
decarboxylation of α-acetolactate. As AHAS catalyses the reaction to convert
pyruvate into α-acetolactate, the reduction of dicetyl production in the Sc-ilv6 mutants
can be considered as the readout for the reduction of in vivo AHAS activity in the
Sc-ilv6 mutant strains. It is a common fact that the in vitro activtity does not
necessarily reflect the in vivo activity of an enzyme due to the lack of allosteric
regulation or of some unknown crucial regulation factors or simply due to the fact that
the in vitro conditions are not optimal for the stability and activity of the enzyme.
Compared to previous studies, our study was the first study which showed a
reduction of diacetyl production by disrupting ILV6 in yeast, and thus was successful
in providing data which supports the role of Ilv6p as the enhancer of AHAS in vivo.
115
Comparative genome analysis revealed that strain B contained one copy of
Sc-ILV6 while strain C contained two copies of Sc-ILV6. The strains Sc-ilv6
∆
and
Sc-ilv6
∆
/Sc-ilv6
∆
generated from strain C contains one and zero copy of Sc-ILV6,
respectively. It was shown that the Sc-ilv6
∆
and Sc-ilv6
∆
/Sc-ilv6
∆
respectively
showed a diacetyl reduction of 13% and 40% compared to the reference strain C.
Nevertheless, these two Sc-ilv6 mutants produced higher levels of diacetyl compared
to strain B (Fig. 21). The results implied that the difference in diacetyl production
between strain B and C not solely resulted from the difference in Sc-ILV6 copy
numbers. Other possible factors causing the different levels of diacetyl production in
strains B and C could involve the differential expression levels of Sc-BAT1 or
glycolytic enzymes as previously discussed in section II.5.2 and II.5.1, respectively.
It was shown that the disruption of Sc-ILV6 gene in the reference strain C resulted
in the stronger impact on the reduction of diacetyl than of 2,3-pentanedione (Fig. 21,
Fig. 23). Diacetyl and 2,3-pentanedione are formed from the non-enzymatic oxidative
decarboxylation of their precursors α-acetolactate and α-aceto-α-hydroxybutyrate,
respectively (Fig. 24). Alpha-acetolactate and α-aceto-α-hydroxybutyrate are formed
by the condensation of one pyruvate molecule respectively with another pyruvate
molecule and one molecule α-ketobutyrate, respectively (Fig. 24). These two
reactions are both catalyzed by AHAS. One possible elucidation for the different
impacts of Sc-ILV6 deletion on diacetyl and 2,3-pentanedione reduction could be that
the deletion had a much stronger influence on reducing the activity
of AHAS in the reaction to form α-acetolactate than in the reaction to form
α-aceto-α-hydroxybutyrate. For example, the reaction to form
α-aceto-α-hydroxybutyrate might be catalyzed by only AHAS large subunit (Ilv2p)
while the reaction to form α-acetolactate might be catalysed by both catalytic subunit
(Ilv2p) and the holoenzyme (Ilv2p+Ilv6p). The deletion of Sc-ILV6 could lead to the
absence of holoenzyme and thus to the stronger reduction in diacetyl than
2,3-pentanedione in strain Sc-ilv6
∆
/Sc-ilv6
∆
.
116
5.4 Fermentation performance of the Sc-ilv6 double deletion mutant
Investigation of fermentation performance of strain Sc-ilv6
∆/
Sc-ilv6
∆
under
industrially relevant brewery conditions showed a slightly slower decrease in wort
gravity in comparison to that of the reference strain C (Fig. 22B). It seemed that the
decrease in wort sugar consumption started from middle of the main fermentation
(approximately from day 3 or day 4). Due to the slower wort consumption, it took
strain Sc-ilv6
∆/
Sc-ilv6
∆
about 20 hrs more than the reference strain to reach wort
attenuation. Growth of strain Sc-ilv6
∆/
Sc-ilv6
∆
was virtually similar to that of the
reference strain C during the fermentation. Nonetheless, by the end of the main
fermentation, the strain Sc-ilv6
∆/
Sc-ilv6
∆
showed a slower sedimentation rate than
the reference strain C (Fig. 22A). Integration of these results suggested that the
slower sedimentation in the mutant strain might result from the slower consumption of
wort sugars compared to the reference strain.
In the beginning of the main fermentation, brewers’ yeast takes up valine as well
as other branched-chain amino acids (BCAAs) from the extracellular wort medium for
growth and maintenance (Petersen et al., 2004). As the internal valine concentration
increases during the uptake of valine, AHAS is inhibited (Inoue and Kashihara,
1995). When the valine in wort is depleted, valine and other BCAAs are produced via
the BCAA biosynthetic pathway. Based on this knowledge, we supposed that the
slower wort sugar consumption starting from the middle of the main fermentation in
strain Sc-ilv6/Sc-ilv6
∆
could result from the lower level of BCAAs formed during this
period. In the beginning of the fermentation, growth of the reference strain C and
strain Sc-ilv6
∆/
Sc-ilv6
∆
were similar as there was no difference in the level of valine
taken up from the extracellular wort medium. As valine from the wort was consumed,
the yeast cells had to synthesize BCAAs necessary for the cellular activities. In the
strain Sc-ilv6
∆/
Sc-ilv6
∆
, AHAS activity was lower than that in the reference strain C
due to the absence of Sc-Ilv6p. The lower level in AHAS activity in strain
Sc-ilv6
∆/
Sc-ilv6
∆
could result in lower level of BCAAs and thus to the slower
consumption of pyruvate substrate. Due to that reason, strain Sc-ilv6
∆/
Sc-ilv6
∆
117
needed more time to consume wort sugars and thus sedimented slower than the
reference strain C.
5.5 Slight change of by-product profile in the green beer produced by strain
Sc-ilv6
∆
∆∆
∆
/ Sc-ilv6
∆
∆ ∆
∆
The green beer produced by strain Sc-ilv6
∆/
Sc-ilv6
∆
showed differences in
concentrations of some acetate esters (isoamyl acetate, 2-phenylethyl acetate, ethyl
acetate) and ethyl esters (ethyl formiate, ethyl butyrate, ethyl caprate) (Table 8).
Besides that, some fusel alcohols (isoamyl alcohol, iso-butyl alcohol, active amyl
alcohol, 2-phenylethyl alcohol) and some fatty acids (capric acid, caprylic acid) were
reduced. A decrease in acetaldehyde level in the green beer produced by strain
Sc-ilv6
∆/
Sc-ilv6
∆
was also observed. However, the concentrations of these
by-products were in the normal range for lager beer and no significant difference in
the taste of beer produced by the strain Sc-ilv6
∆/
Sc-ilv6
∆
was detected in
comparison to that of the reference beer.
Isoamyl alcohol, active amyl alcohol and iso-butyl alcohol are intermediates of the
BCAA biosynthetic pathway (Fig. 24). The concentrations of these fusel alcohols
were lower in green beer produced by strain Sc-ilv6
∆/
Sc-ilv6
∆
in comparison to
those of the reference beer (Table 8). In addition to the decreaded isoamyl alcohol,
an increase in its corresponding acetate ester (isoamyl acetate) was observerd.
Besides that, a decrease in the production of 2-phenylethyl alcohol and an increase
in the production of its acetate ester (2-phenylethyl acetate) were also observed
(Table 8). Therefore, the lower level of these fusel alcohols might be caused by the
increase of their corresponding acetate esters.
In addition to the enhancement in concentrations of acetate esters, green beer
produced by strain Sc-ilv6
∆/
Sc-ilv6
∆
also showed an increased level of some ethyl
esters (ethyl formiate, ethyl butyrate and ethyl caprate). Ethyl esters are formed from
the esterification of ethanol with the fatty acids under activity of esterase. In addition
118
to the increase of ethyl caprate, a decrease in its corresponding fatty acid (capric
acid) concentration was observed.
The decrease of fusel alcohols in correlation to the increase of their acetate esters
and the decrease of fatty acids in correlation to the increase of the corresponding
ethyl esters could be explained by the alteration of esterases that are responsible for
the esterification reactions between acetyl-coA and fusel alcohols and the
esterification reations between ethanol and fatty acids. The disruption of Sc-ILV6
might have unknown impact resuting in an increase in activity of these esterases.
Due to that reason, the reactions between acetyl-coA and fusel alcohols as well as
reactions between ethanol and fatty acids could be accelerated, thereby resulting in
the decreased fusel alcohols and fatty acids as well as the increase of the
corresponding acetate esters and ethyl ester production. The question raised here is
why the production of some esters was altered when the concentrations of other
esters were unchanged. One possible explanation could be that the disruption of
Sc-ILV6 could have resulted in the increase of activity of some certain esterases,
which catalyzed specific esterification reactions between ethanol and some certain
fatty acids as well as between acetyl-coA and some certain fusel alcohols.
5.6 Concluding remarks and outlook
The global analyses are powerful tools for the identification of potential novel
target genes for diacetyl reduction i.e. Sc-ILV6, Sc-BAT1, non-Sc-BAT1 and
non-Sc-BAT2 in lager brewers’ yeast in particular when the results of different
molecular level were integrated. A striking decrease in diacetyl production was
obtained by the disruption of one and two copies Sc-ILV6 in an industrial lager
brewers’ yeast strain. The strain Sc-ilv6
∆
/Sc-ilv6
∆
which containes no copy of
Sc-ILV6 ORF showed a reduction in diacetyl production by 65% under the relevant
industrial brewery fermentation. The result supports the role Sc-ilv6p as an enhancer
119
of Ilv2p. In addition, the results confirm that inverse metabolic engineering is a useful
tool for identifying novel target genes for the improvement of brewers’ strain.
Besides the reduction of diacetyl production, small changes in concentrations of
alcetaldehyde, esters, fusel alcohols, fatty acids in green beer produced by the strain
Sc-ilv6
∆
/Sc-ilv6
∆
were observed. The concentrations of these by-products, however,
are in the normal range for lager beer. The taste of the beer produced by the strain
Sc-ilv6
∆
/Sc-ilv6
∆
showed no significant difference compared to that of the reference
beer. The characterisation of the beer produced by strain Sc-ilv6
∆
/Sc-ilv6
∆
suggested that such a modification strain would be useful for the beer brewing with a
shortening lagering period.
In the genome of strain Sc-ilv6
∆
/Sc-ilv6
∆
, the two copies of Sc-ILV6 ORF were
replaced by the loxP-kanMX-loxP and loxP-ble
r
-loxP disruption cassettes. One of the
aims of the future work will be the removal of these disruption cassettes for conferring
a better acceptance in commercial use to this strain
.
This task can be carried out by
introducing a plasmid expressing Cre recombinase under the control of GAL
promoter. In the presence of galactose, Cre recombinase action at the repeated loxP
sites will excise the kanMX and ble
r
markers, leaving behind one loxP sequence at
the site of the each disruption cassette. After that, the Cre recombinase plasmid can
be removed by subcultivations under non-selective condition. To this end, apart from
the two loxP sequences, strain Sc-ilv6
∆
/Sc-ilv6
∆
will contain no other hetorologous
DNA. In this case, the engineered strain can be considered as “self-cloned” and can
be better accepted in food and beverage industry.
It is very likely that diacetyl can be reduced even more. The expression levels of
Sc-BAT1, non-Sc-BAT2 and non-Sc-BAT1 as well as the abundant level of Eno2p
and Pgk1p could influence the level of diacetyl production of lager brewers’ yeast.
Thus, the outlook of this thesis involves the additional genetic manipulation of other
target genes directly relevant to diacetyl formation i.e Sc-BAT1, non-Sc-BAT2 and
non-Sc-BAT1 or even of target genes which may be relate to diacetyl formation i.e.
ENO2 and PGK1. The verification of the hypothesis that the expression levels of
120
these genes could have influenced the diacetyl production can provide more
knowledge about the genotype-phenotype relationship in brewers’ yeast and is
crucial for the construction of new brewers’ yeast strain with desired diacetyl
phenotype.
121
6 Summary
This thesis aimed at improving lager brewers’ yeast by means of inverse
metabolic engineering. The primary target is the reduction of diacetyl production. To
that aim, three lager brewers’ yeast strains producing different levels of diacetyl were
selected: i) strain A which produces the highest level of diacetyl, ii) strain B which
shows a very low level of diacetyl production and iii) strain C, a currently used
industrial production strain, whose diacetyl level is slightly lower than that produced
by strain A but much higher than strain B. Although strain B seems to be an ideal
strain due to its low diacetyl production, it is not useful for beer brewing because of a
very strong and early flocculation which results in uncomplete wort attenuation.
In order to identify the genetic basis for the strain’s phenotypic differences
relevant to brewing, an integrated approach was chosen using global analyses at
different molecular levels which influence protein expression, i.e. gene copy numbers
(analysed by microarray-based comparative genome hybridisation), mRNA
concentrations (via microarray-based comparative transcriptome analysis) and
protein abundance (two-dimensional gel electrophoresis plus mass spectrometry).
Based on the result, an industrial production strain was modified and its diacetyl
production could be significantly reduced without negatively affecting any important
property of the strain relevant in brewing.
The main results obtained were as follows:
1) Genome and transcriptome analyses revealed numerous significant
differences regarding the abundance of gene copies and transcripts in the studied
strains. Among those, several differences obviously related to flocculation and
diacetyl phenotype were identified. In contrast, the number of significant differences
at proteome level was very low and none of these few was directly related to the
difference in diacetyl and flocculation phenotypes.
2) The comparative transcriptome analysis of the three studied strains revealed
several differences in mRNA concentrations of genes, i.e Sc-ILV6, Sc-BAT1,
non-Sc-BAT1 and non-Sc-BAT2, whose products are directly involved in the valine
biosynthesis pathway and could thus have affected the production of diacetyl, a
by-product of valine biosynthesis, simply by strain-dependent differences in their
protein activities. Thus, these genes were considered promising targets for the
reduction of diacetyl production in brewers’ yeast.
122
3) Among the potential target genes, Sc-ILV6 was chosen for further
investigations. In fact, Ilv6p has previously been proposed to be a regulatory subunit
of Ilv2p (AHAS), the enzyme which is responsible for the formation of α-acetolactate,
the precursor of diacetyl. To verify the role of Sc-ILV6 gene in diacetyl formation, two
copies of Sc-ILV6 were subsequently disrupted in the industrial production strain C,
leading to the generation of a single mutant (Sc-ilv6
∆
) and a double mutant
(Sc-ilv6
∆
/Sc-ilv6
∆
), respectively.
4) The disruption of two copies of Sc-ILV6 in the production strain only led to an
insignificant decrease in in vitro AHAS activity. However, an industrially relevant
brewery fermentation revealed that the diacetyl production of strain Sc-ilv6
∆
/Sc-ilv6
∆
was reduced by 65% at the end of the main fermentation and this concentration is
0.08 mg/ml and below the taste-threshold of 0.1 mg/ml for diacetyl in beer. The
concentration of 2,3-pentanedione was only slightly reduced by Sc-ILV6 deletion
5) Comparative genome analysis revealed that strain B contained one copy of
Sc-ILV6 while strain C contained two copies of Sc-ILV6 ORF. However, compared to
strain B, the diacetyl production of the Sc-ilv6 double deletion mutant (derived from
strain C) was still much higher. The results imply that the lower level of diacetyl
production in strain B compared to strain C could not be solely caused by the higher
copy number of Sc-ILV6 in strain C.
6) Examination of the Sc-ilv6 double deletion strain under conditions relevant in
industrial brewery fermentations revealed that the green beer produced by strain
Sc-ilv6
∆
/Sc-ilv6
∆
showed only small alterations in the concentrations of some fusel
alcohols, esters, fatty acids and acetaldehyde. Nevertheless, the levels of these
by-products are in the normal range for lager beer. Sensory investigation revealed no
significant differences in the taste of the beer produced by the Sc-ilv6
∆
/Sc-ilv6
∆
compared to that of the reference beer.
123
7 Zusammenfassung
Die vorliegende Arbeit beschäftigt sich mit der Optimierung von untergärigen
Brauhefen mittels Inversem Metabolic Engineering. Das vorrangige Ziel war die
Reduzierung der Diacetylproduktion. Zu diesem Zweck wurden drei untergärige
Brauhefen mit unterschiedlicher Diacetylproduktion selektiert: i) Stamm A, der am
meisten Diacetyl bildet, ii) Stamm B, der das niedrigste Diacetyl-Niveau zeigt und
iii) Stamm C, ein zur Zeit eingesetzter industrieller Produktionsstamm, der etwas
weniger Diacetyl bildet als Stamm A, jedoch wesentlich mehr als Stamm B. Trotz
seiner geringen Diacetylbildung ist Stamm B nicht für Brauereien geeignet, da dieser
Stamm stark und sehr früh flockuliert, was zu einer unvollständigen Würzevergärung
führt.
Um die genetische Basis für die brauerei-relevanten stammspezifischen
phenotypischen Unterschiede zu identifizieren wurde ein integrierter Ansatz gewählt,
der globale Analysenmethoden beeinhaltete um unterschiedlichen molekulare
Ebenen zu analysieren, die einen Einfluß auf die Proteinexpression haben, d.h.
Genkopiezahl (Microarray-basierte vergleichende Genomhybridisierung), mRNA
Konzentrationen (genomweite Transkriptom-Analyse mittels Hefe-Microarrays) und
Proteinkonzentrationen (zweidimensionale Gelelektrophorese und Massen-
spektrometrie). Basierend auf den Ergebnissen wurde der Produktionsstamm C so
verändert, dass die Diacetylproduktion signifikant reduziert und keine der anderen
brauerei-relevante Eigenschaften negativ beeinflußt war.
Im folgenden sind die wesentlichen Ergebnisse der Arbeit zusammengefaßt:
1) Die Genom- und Transkriptomanalyse ergab eine Vielzahl von signifikanten
stammspezifischen Unterschieden. Unter den identifizierten Genen waren einige, die
offensichtlich einen direkten Bezug zum Diacetyl- bzw. Flockulationsphenotyp hatten.
Dagegen war die Anzahl der stammspezifischen Unterschiede auf der
Proteomebene nur sehr gering und die wenigen detektierten Gene hatten keinen
direkten Bezug zum Diacetyl- bzw. Flockulationsphenotyp.
2) Die vergleichende Transkriptomanalyse der drei Stämme ergab Unterschiede
in den Konzentrationen von verschiedenen mRNAs, d. h. Sc-ILV6, Sc-BAT1,
non-Sc-BAT1, non-Sc-BAT2, dessen Genprodukte direkt in die Valinbiosynthese
involviert sind. Stammspezifische Unterschiede bezüglich der entsprechenden
124
Proteinaktivitäten könnten die Produktion von Diacetyl beeinflußt haben, welches ein
Nebenprodukt der Valinbiosynthese ist. Daher wurden diese Gene als potentielle
Targets für eine Reduzierung der Diacetylproduktion in Brauhefen betrachtet.
3) Aus den potentiellen Targetgenen, wurde Sc-ILV6 für weitere Untersuchungen
ausgewählt. Arbeiten anderer Autoren ließen vermuten, dass Ilv6p eine
regulatorische Untereinheit von Ilv2p (AHAS) ist, welches für die Bildung von
α-Acetolaktat verantwortlich ist, dem Precursor von Diacetyl. Um die Rolle von
Sc-ILV6 in der Diacetylbildung zu verifizieren, wurden beide Kopien dieses Gens im
industriellen Produktionsstamm C nacheinander deletiert; d. h. es wurde sowohl eine
Einfachmutante (Sc-ilv6
∆
) und eine Doppelmutante (Sc-ilv6
∆
/Sc-ilv6
∆
) generiert.
4) Die Deletion der beiden Sc-ILV6 Kopien in Produktionsstamm C führte zu
einer nicht-signifikanten Verminderung der in vitro AHAS-Aktivität. Allerdings zeigten
Fermentationen unter industriell relevanten Bedingungen, dass die
Diacetylkonzentration der Doppelmutante Sc-ilv6
∆
/Sc-ilv6
∆
am Ende der
Fermentation im Vergleich zur Kontrolle um 65% reduziert war. Die erreichte
Konzentration von 0.08 mg/ml lag sogar unter dem Geschmacksschwellenwert von
Diacetyl im Bier (0.1 mg/l)
5) Die vergleichende Genomanalyse zeigte, dass Stamm B eine Kopie von
Sc-ILV6 enthielt, während Stamm C zwei Kopien aufwies. Allerdings produzierte die
Sc-ilv6
∆
/Sc-ilv6
∆
Doppelmutante (die von Stamm C abgeleitet wurde) immer noch
wesentlich mehr Diacetyl als Stamm B. Diese Ergebnisse lassen folgern, dass das
niedrige Diacetyl-Niveau in Stamm B nicht ausschließlich auf der niedrigeren
Kopiezahl von Sc-ILV6 in diesem Stamm beruhen kann.
6) Untersuchungen der Sc-ilv6
∆
/Sc-ilv6
∆
Doppelmutante unter brauerei-
relevanten Fermentationsbedingungen ergaben, dass das produzierte Jungbier nur
geringfüge Veränderungen in Bezug auf die Konzentrationen einiger höherer
Alkohole, Ester, Fettsäuren und Acetaldehyd aufwies. Insgesamt lagen die
Konznetrationen jedoch alle in einem Bereich, der für untergäriges Bier als normal
angesehen wird. Sensorische Tests ergaben keine erkennbaren Unterschiede im
Geschmack im Vergleich zum Referenzbier.
125
8 References
Aebersold, R. and Mann, M. (2003). Mass spectrometry-based proteomics. Nature 422, 198-
207.
Affymetrix (2004). GeneChip Expression Analysis: Technical Manual 701021 Rev. 5. Santa
Clara, CA, Affymetrix.
Akada, R. (2002). Genetically modified industrial yeast ready for application. J Biosci Bioeng
94, 536-44.
Andersen, T., Hoffmann, L., Grifone, R., Nilsson-Tillgren T and Kielland-Brandt, M. (2000).
Brewing yeast genetics, EBC Monograph Fachverlag Hans Carl, pp. 140-147.
Attfield, A. V. and Bell, P. J. L. (2003). Genetics and Classical Genetic Manipulations of
Industrial Yeasts in Winde, J. H. d. (Ed), Functional genetics of industrial yeasts, pp.
17-55.
Azumi, M. and Goto-Yamamoto, N. (2001). AFLP analysis of type strains and laboratory and
industrial strains of Saccharomyces sensu stricto and its application to phenetic
clustering. Yeast 18, 1145-54.
Bailey, J. E. (1991). Toward a science of metabolic engineering. Science 252, 1668-75.
Bailey, J. E., Sburlati, A., Hatzimanikatis, V., Lee, K., Renner, W. A. and Tsai, P. S. (2002).
Inverse metabolic engineering: a strategy for directed genetic engineering of useful
phenotypes. Biotechnol Bioeng 79, 568-79.
Baldi, P. and Long, A. D. (2001). A Bayesian framework for the analysis of microarray
expression data: regularized t -test and statistical inferences of gene changes.
Bioinformatics 17, 509-19.
Barnett, J. A. (2004). A history of research on yeasts 8: taxonomy. Yeast 21, 1141-93.
Bayly, J. C., Douglas, L. M., Pretorius, I. S., Bauer, F. F. and Dranginis, A. M. (2005).
Characteristics of Flo11-dependent flocculation in Saccharomyces cerevisiae. FEMS
Yeast Res 5, 1151-6.
Bilinski, C. A. and Casey, J. (1989). Developments in sporulation and breeding of brewer's
yeast. Yeast 5, 429-438.
Bilinski, C. A., I., R. and Stewart , G. (1987). Cross breeding of Saccharomyces cerevisiae
and Saccharomyces uvarum (carlsbergensis) by mating of meiotic segregants:
isolation and characterization of species hybrids, Proc 21st Congr Eur Brew Conv.
Blomqvist, K., Suihko, M. L., Knowles, J. and Penttila, M. (1991). Chromosomal Integration
and Expression of Two Bacterial alpha-Acetolactate Decarboxylase Genes in Brewer's
Yeast. Appl Environ Microbiol 57, 2796-2803.
Bolstad, B. M., Irizarry, R. A., Astrand, M. and Speed, T. P. (2003). A comparison of
normalization methods for high density oligonucleotide array data based on variance
and bias. Bioinformatics 19, 185-93.
Bond, U., Neal, C., Donnelly, D. and James, T. C. (2004). Aneuploidy and copy number
breakpoints in the genome of lager yeasts mapped by microarray hybridisation. Curr
Genet 45, 360-70.
Boulton, C. A. and Quain, D. (2001). Brewing Yeast and Fermentation. Blackwell Science, 7.
Brejning, J., Arneborg, N. and Jespersen, L. (2005). Identification of genes and proteins
induced during the lag and early exponential phase of lager brewing yeasts. J Appl
Microbiol 98, 261-71.
Briggs, D. E., Boulton, C. A., Brookes, P. A. and Stevens, R. (2004). Brewing Science and
Practice. Blackwell Science, Oxford.
Bro, C., Knudsen, S., Regenberg, B., Olsson, L. and Nielsen, J. (2005). Improvement of
galactose uptake in Saccharomyces cerevisiae through overexpression of
phosphoglucomutase: example of transcript analysis as a tool in inverse metabolic
engineering. Appl Environ Microbiol 71, 6465-72.
126
Bro, C. and Nielsen, J. (2004). Impact of 'ome' analyses on inverse metabolic engineering.
Metab Eng 6, 204-11.
Byrne, K. L. and Meacock, P. A. (2001). Thiamin auxotrophy in yeast through altered
cofactor dependence of the enzyme acetohydroxyacid synthase. Microbiology 147,
2389-98.
Campbell, I. (2000). Brewing yeast selection. Microbiology today 27, 122-24.
Casaregola, S., Nguyen, H. V., Lapathitis, G., Kotyk, A. and Gaillardin, C. (2001). Analysis
of the constitution of the beer yeast genome by PCR, sequencing and subtelomeric
sequence hybridization. Int J Syst Evol Microbiol 51, 1607-18.
Casey, J. P. (1996). Practical application of pulsed field electrophoresis and yeast
chromosome fingerprinting in brewing QA and R&D. Technical Quarterly of the
Master Brewers of Association of the Americas 33, 1-10.
Cole, G. E., McCabe, P. C., Inlow, D., Gelfand, D. H., Ben-Bassat, A. and Innis, M. A.
(1988). Stable expression of Aspergillus awamori glucoamylase in distiller’s yeast.
Biotechnology 6, 417-421.
Corran, H. (1975). A history of brewing.
Cullin, C., Baudin-Baillieu, A., Guillemet, E. and Ozier-Kalogeropoulos, O. (1996).
Functional analysis of YCL09C: evidence for a role as the regulatory subunit of
acetolactate synthase. Yeast 12, 1511-8.
Daran-Lapujade, P., Daran, J. M., Kotter, P., Petit, T., Piper, M. D. and Pronk, J. T. (2003).
Comparative genotyping of the Saccharomyces cerevisiae laboratory strains S288C
and CEN.PK113-7D using oligonucleotide microarrays. FEMS Yeast Res 4, 259-69.
De Barros Lopes, M., Bellon, J. R., Shirley, N. J. and Ganter, P. F. (2002). Evidence for
multiple interspecific hybridization in Saccharomyces sensu stricto species. FEMS
Yeast Res 1, 323-31.
Dengis, P. B. and Rouxhet, P. G. (1997). Surface properties of top- and bottom-fermenting
yeast. Yeast 13, 931-43.
Dequin, S. (2001). The potential of genetic engineering for improving brewing, wine-making
and baking yeasts. Appl Microbiol Biotechnol 56, 577-88.
Donalies, U. E., Nguyen, H. T., Stahl, U. and Nevoigt, E. (2008). Improvement of
Saccharomyces Yeast Strains Used in Brewing, Wine Making and Baking. Adv
Biochem Eng Biotechnol.
Donalies, U. E. and Stahl, U. (2002). Increasing sulphite formation in Saccharomyces
cerevisiae by overexpression of MET14 and SSU1. Yeast 19, 475-84.
Dufour, J.-P. (1991). Influence of industrial brewing and fermentation working conditions on
beer SO2 level and flavour stability. Proceedings of the 23rd Congress Lisbon, pp.
209.
Dunn, B. and Sherlock, G. (2008). Reconstruction of the genome origins and evolution of the
hybrid lager yeast Saccharomyces pastorianus. Genome Res 18, 1610-23.
European Brewery Convention (1998) ANALYTICA-EBC. Verlag Hans Carl, Getraenke-
Fach-verlag.
Falco, S. C. and Dumas, K. S. (1985). Genetic analysis of mutants of Saccharomyces
cerevisiae resistant to the herbicide sulfometuron methyl. Genetics 109, 21-35.
Federoff, H. J., Eccleshall, T. R. and Marmur, J. (1983). Carbon catabolite repression of
maltase synthesis in Saccharomyces carlsbergensis. J Bacteriol 156, 301-7.
Fujii, T., Kondo, K., Shimizu, F., Sone, H., Tanaka, J.-I. and Inoue, T. (1990). Application of
a ribosomal DNA integration vector in the construction of a brewer´s yeast having -
acetolactate decarboxlase activity. Appl Env Microbiol 56, 997.
Fujii, T., Nagasawa, N., Iwamatsu, A., Bogaki, T., Tamai, Y. and Hamachi, M. (1994).
Molecular cloning, sequence analysis, and expression of the yeast alcohol
acetyltransferase gene. Appl Environ Microbiol 60, 2786-92.
127
Gancedo, J. M. (1998). Yeast carbon catabolite repression. Microbiol Mol Biol Rev 62, 334-
61.
Gjermansen, C., Nilsson-Tillgren, T., Petersen, J. G., Kielland-Brandt, M. C., Sigsgaard, P.
and Holmberg, S. (1988). Towards diacetyl-less brewers' yeast. Influence of ilv2 and
ilv5 mutations. J Basic Microbiol 28, 175-83.
Gjermansen, C. and Sigsgaard, P. (1981). Construction of a hybrid brewing strain of
Saccharomyces carlsbergensis by mating meiotic segregants. Carlsberg Res Comm 46,
1-11.
Go¨rts, C. P. M. (1969). Effect of glucose on the activity and the kinetics of the maltose-
uptake system and of a-glucosidase in Saccharomyces cerevisiae. Biochim. Biophys.
Acta 184, 299–305.
Godtfredsen, S. E., Ottensen, M., Sigsgaard, P., Erdal, K., Mathiasen, T. and B. Ahrenst-
Larsen, B. (1987). Use of alpha-acetolactate decarboxylase for accelerated maturation
of beer, Proc. 21
st
EBC Congress, IRL Press, Oxford, UK, 161-168.
Goelling, D. and Stahl, U. (1988). Cloning and expression of an alpha-acetolactate
decarboxylase gene from Streptococcus lactis subsp. diacetylactis in Escherichia coli.
Appl Environ Microbiol 54, 1889-91.
Goossens, E., Debourg, A., Villanueba, K. and Masschelein, C. (1993). Decreased diacetyl
production in lager brewing yeast by integration of the ILV5 gene XIII of
Saccharomyces cerevisiae. Proceedings of the 24th Congress Oslo, pp. 251.
Goossens, E., Dillemans, M., Debourg, A. and Masschelein, C. (1987). Control of diacetyl
formation by the intensification of the anabolic flux of acetohydroxyacid
intermediates. Proceedings of the 21st Congress. Madrid, pp. 553.
Gopal, C. and Hammond, J. (1992). Use of genetically-modified yeasts for beer production.
Proc 24th Int Brew Tech Conf. Harrogate, pp. 297.
Govender, P., Domingo, J. L., Bester, M. C., Pretorius, I. S. and Bauer, F. F. (2008).
Controlled expression of the dominant flocculation genes FLO1, FLO5, and FLO11 in
Saccharomyces cerevisiae. Appl Environ Microbiol 74, 6041-52.
Gueldener, U., Heinisch, J., Koehler, G. J., Voss, D. and Hegemann, J. H. (2002). A second
set of loxP marker cassettes for Cre-mediated multiple gene knockouts in budding
yeast. Nucleic Acids Res 30, e23.
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, 9390-5.
Hammond, J. R. (1995). Genetically-modified brewing yeasts for the 21st century. Progress to
date. Yeast 11, 1613-27.
Han, E. K., Cotty, F., Sottas, C., Jiang, H. and Michels, C. A. (1995). Characterization of
AGT1 encoding a general alpha-glucoside transporter from Saccharomyces. Mol
Microbiol 17, 1093-107.
Hannemann, W. (2002). Reducing beer maturation time and retaining quality. Tech. Q.
Master Brew. Assoc. Am 39(3), 149-155.
Hansen, J. (1999). Inactivation of MXR1 abolishes formation of dimethyl sulfide from
dimethyl sulfoxide in Saccharomyces cerevisiae. Appl Environ Microbiol 65, 3915-9.
Hansen, J., Bruun, S. V., Bech, L. M. and Gjermansen, C. (2002). The level of MXR1 gene
expression in brewing yeast during beer fermentation is a major determinant for the
concentration of dimethyl sulfide in beer. FEMS Yeast Res 2, 137-49.
Hansen, J. and Kielland-Brandt, M. (1996a). Inactivation of MET10 in brewer´s yeast
specifically increases SO2 formation during beer production. Nature Biotechnol 14,
1587.
Hansen, J. and Kielland-Brandt, M. (1997). Brewer´s yeast. Technomic Publishing
128
Hansen, J. and Kielland-Brandt, M. C. (1994). Saccharomyces carlsbergensis contains two
functional MET2 alleles similar to homologues from S. cerevisiae and S. monacensis.
Gene 140, 33-40.
Hansen, J. and Kielland-Brandt, M. C. (1996b). Inactivation of MET2 in brewer's yeast
increases the level of sulfite in beer. J Biotechnol 50, 75-87.
Hansen, J. and Kielland-Brandt, M. C. (2003). Brewer's yeast: genetic structure and targets
for improvement in Winde, J. H. d. (Ed), Functional genetics of industrial yeasts, pp.
143-203.
Hirata, D., Aoki, S., Watanabe, K., Tsukioka, M. and Suzuki, T. (1992). Stable
Overproduction of Isoamyl Alcohol by Saccharomyces cerevisiae with Chromosome-
integrated Multicopy LEU4 Genes. Bioscience, Biotechnology and Biochemistry 56,
1682-1683.
Hu, Z., Nehlin, J. O., Ronne, H. and Michels, C. A. (1995). MIG1-dependent and MIG1-
independent glucose regulation of MAL gene expression in Saccharomyces cerevisiae.
Curr Genet 28, 258-66.
Inoue, T. and Kashihara, T. (1995). The importance of indices related to nitrogen metabolism
in fermentation control. Tech. Q. Master Brew. Assoc. Am 32, 109-113.
James, T. C., Campbell, S., Donnelly, D. and Bond, U. (2003). Transcription profile of
brewery yeast under fermentation conditions. J Appl Microbiol 94, 432-48.
Jespersen, L., Cesar, L. B., Meaden, P. G. and Jakobsen, M. (1999). Multiple alpha-glucoside
transporter genes in brewer's yeast. Appl Environ Microbiol 65, 450-6.
Jin, Y. S., Alper, H., Yang, Y. T. and Stephanopoulos, G. (2005). Improvement of xylose
uptake and ethanol production in recombinant Saccharomyces cerevisiae through an
inverse metabolic engineering approach. Appl Environ Microbiol 71, 8249-56.
Johannesen, P., Nyborg, M. and Hansen, J. (1999). Construction of S. carlsbergensis brewer´s
yeast without production of sulfite. Proceedings of the 27th Congress. Cannes, pp.
655.
Johnston, J. R. and H. P, R. (1983). Genetic control of flocculation in J. F. T. Spencer, D. M.
S., and A. R. W. Sait (Ed), Yeast genetics, fundamental and applied aspects, pp. 205–
224.
Johnston, J. R. and Reader, H. P. (1982). Genes conferring and suppressing focculation in
brewing and laboratory strains of Saccharomyces. Proceedings of the 11th
International Conference on Yeast Genetics and Molecular Biology. Montpelier, pp.
p.124.
Joubert, R., Brignon, P., Lehmann, C., Monribot, C., Gendre, F. and Boucherie, H. (2000).
Two-dimensional gel analysis of the proteome of lager brewing yeasts. Yeast 16, 511-
22.
Joubert, R., Strub, J. M., Zugmeyer, S., Kobi, D., Carte, N., Van Dorsselaer, A., Boucherie,
H. and Jaquet-Guffreund, L. (2001). Identification by mass spectrometry of two-
dimensional gel electrophoresis-separated proteins extracted from lager brewing yeast.
Electrophoresis 22, 2969-82.
Kiellandt-Brandt, M., Gjermansen, C., Nilsson-Tillgren, T. and Holmberg, S. (1989). Yeast
breeding. Proceedings of the 22nd Congress. Zürich pp. 37.
Kiellandt-Brandt, M., Nilsson-Tillgren, T., Gjermansen, C., Holmberg, S. and Petersen, M.
(1995). Genetics of brewing yeasts. Academic Press, London.
Klein, C. J., Olsson, L. and Nielsen, J. (1998). Glucose control in Saccharomyces cerevisiae:
the role of Mig1 in metabolic functions. Microbiology 144 ( Pt 1), 13-24.
Klein, C. J., Olsson, L., Ronnow, B., Mikkelsen, J. D. and Nielsen, J. (1996). Alleviation of
glucose repression of maltose metabolism by MIG1 disruption in Saccharomyces
cerevisiae. Appl Environ Microbiol 62, 4441-9.
129
Kobayashi, O., Hayashi, N., Kuroki, R. and Sone, H. (1998). Region of FLO1 proteins
responsible for sugar recognition. J Bacteriol 180, 6503-10.
Kobayashi, O., Hayashi, N. and Sone, H. (1995). The FLO1 genes determine two flocculation
phenotypes distinguished by sugar inhibition, Proc Congr Eur Brew Conv 16, pp.
361–367.
Kobayashi, O., Suda, H., Ohtani, T. and Sone, H. (1996). Molecular cloning and analysis of
the dominant flocculation gene FLO8 from Saccharomyces cerevisiae. Mol Gen Genet
251, 707-15.
Kobayashi, O., Yoshimoto, H. and Sone, H. (1999). Analysis of the genes activated by the
FLO8 gene in Saccharomyces cerevisiae. Curr Genet 36, 256-61.
Kobi, D., Zugmeyer, S., Potier, S. and Jaquet-Gutfreund, L. (2004). Two-dimensional protein
map of an "ale"-brewing yeast strain: proteome dynamics during fermentation. FEMS
Yeast Res 5, 213-30.
Kodama, Y., Fukui, N., Ashikari, T. and Shibano, Y. (1994). Improvement of Maltose
Fermentation Efficiency: Constitutive Expression of MAL Genes in Brewing Yeasts. .
Am Soc Brew Chem 59, 24-29.
Kodama, Y., Kielland-Brandt, M. C. and Hansen, J. (2006). Lager brewing yeast, Topics in
Current Genetics.
Kodama, Y., Omura, F. and Ashikari, T. (2001a). Isolation and characterization of a gene
specific to lager brewing yeast that encodes a branched-chain amino acid permease.
Appl Environ Microbiol 67, 3455-62.
Kodama, Y., Omura, F., Miyajima, K. a. and Ashikari, T. (2001b). Control of Higher Alcohol
Production by Manipulation of the BAP2 Gene in Brewing Yeast. J Am Soc Brew
Chem 59(4), 157-162.
Kohler, C., Wolff, S., Albrecht, D., Fuchs, S., Becher, D., Buttner, K., Engelmann, S. and
Hecker, M. (2005). Proteome analyses of Staphylococcus aureus in growing and non-
growing cells: a physiological approach. Int J Med Microbiol 295, 547-65.
Korch, C., Mountain, H., Gyllang, H., Winge, H. and Brehmer, P. (1991). A mechanism for
sulfite production in beer and how to increase sulfite levels by recombinant genetics.
Proceedings of the 23rd Congress. Lisbon, pp. 201.
Lancashire, W., Carter, A., Howard, J. and Wilde, R. (1989). Superatenuating brewing yeast.
Proceedings of the 22nd Congress. Zürich, pp. 491.
Lee, S., Villa, K. and Patino, H. (1995). Yeast strain development for enhanced production of
desirable alcohol/esters in beer. J. Am. Soc. Brew. Chem 53, 153-156.
Liu, H., Styles, C. A. and Fink, G. R. (1996). Saccharomyces cerevisiae S288C has a
mutation in FLO8, a gene required for filamentous growth. Genetics 144 967–978.
Liu, Z., Zhang, G. and Liu, S. (2004). Constructing an amylolytic brewing yeast
Saccharomyces pastorianus suitable for accelerated brewing. J Biosci Bioeng 98, 414-
9.
Lo, W. S. and Dranginis, A. M. (1996). FLO11, a yeast gene related to the STA genes,
encodes a novel cell surface flocculin. J Bacteriol 178, 7144-51.
Lockhart, D. J., Dong, H., Byrne, M. C., Follettie, M. T., Gallo, M. V., Chee, M. S.,
Mittmann, M., Wang, C., Kobayashi, M., Horton, H. and Brown, E. L. (1996).
Expression monitoring by hybridization to high-density oligonucleotide arrays. Nat
Biotechnol 14, 1675-80.
Lucero, P., Herweijer, M. and Lagunas, R. (1993). Catabolite inactivation of the yeast maltose
transporter is due to proteolysis. FEBS Lett 333, 165-8.
MacAlpine, D. M., Perlman, P. S. and Butow, R. A. (2000). The numbers of individual
mitochondrial DNA molecules and mitochondrial DNA nucleoids in yeast are co-
regulated by the general amino acid control pathway. EMBO J 19, 767-75.
130
Masy, C. L., Henquinet, A. and Mestdagh, M. M. (1992). Flocculation of Saccharomyces
cerevisiae: inhibition by sugars. Can J Microbiol 38, 1298-306.
McCourt, J. A. and Duggleby, R. G. (2006). Acetohydroxyacid synthase and its role in the
biosynthetic pathway for branched-chain amino acids. Amino Acids 31, 173-210.
Meaden, P. G. and Tubb, R. S. (1985). A plasmid vector system for the genetic manipulation
of brewing strains. Proceedings of the European Brewery Convention Congress.
Helinski, pp. 219- 226.
Mijalski, T., Harder, A., Halder, T., Kersten, M., Horsch, M., Strom, T. M., Liebscher, H. V.,
Lottspeich, F., de Angelis, M. H. and Beckers, J. (2005). Identification of coexpressed
gene clusters in a comparative analysis of transcriptome and proteome in mouse
tissues. Proc Natl Acad Sci U S A 102, 8621-6.
Mithieux, S. M. and Weiss, A. S. (1995). Tandem integration of multiple ILV5 copies and
elevated transcription in polyploid yeast. Yeast 11, 311-6.
Nagasawa, N., Bogaki, T., Iwamatsu, A., Hamachi, M. and Kumagai, C. (1998). Cloning and
nucleotide sequence of the alcohol acetyltransferase II gene (ATF2) from
Saccharomyces cerevisiae. Biosci. Biotechnol. Biochem. 62, 1852-1857.
Nakao, Y., Nakamura, N., Kodama, Y., FUJIMURA, T. and Ashikari, T. (2008). Screening
method for genes of brewing yeast
Naumova, E. S., Naumov, G. I., Masneuf-Pomarede, I., Aigle, M. and Dubourdieu, D. (2005).
Molecular genetic study of introgression between Saccharomyces bayanus and S.
cerevisiae. Yeast 22, 1099-115.
Nevoigt, E. (2008). Progress in metabolic engineering of Saccharomyces cerevisiae.
Microbiol Mol Biol Rev 72, 379-412.
Noronha, S., Kaslow, D. and J., S. (1998). Transition phase in the production of recombinant
proteins in yeast under the ADH2 promoter: an important step for reproducible
manufacturing of a malaria transmission blocking vaccine candidate. J. Ind.
Microbiol. 20, 192–199.
Olesen, K., Felding, T., Gjermansen, C. and Hansen, J. (2002). The dynamics of the
Saccharomyces carlsbergensis brewing yeast transcriptome during a production-scale
lager beer fermentation. FEMS Yeast Res 2, 563-73.
Omura, F. (2008). Targeting of mitochondrial Saccharomyces cerevisiae Ilv5p to the cytosol
and its effect on vicinal diketone formation in brewing. Appl Microbiol Biotechnol 78,
503-13.
Omura, F. and Shibano, Y. (1995). Reduction of hydrogen sulfide production in brewing
yeast by constitutive expression of MET25 gene. J Am Soc Brew Chem 53, 58.
Ostergaard, S., Olsson, L. and Nielsen, J. (2000). Metabolic engineering of Saccharomyces
cerevisiae. Microbiol Mol Biol Rev 64, 34-50.
Pan, X. and Heitman, J. (1999). Cyclic AMP-dependent protein kinase regulates
pseudohyphal differentiation in Saccharomyces cerevisiae. Mol Cell Biol 19, 4874-87.
Pandey, G., Yoshikawa, K., Hirasawa, T., Nagahisa, K., Katakura, Y., Furusawa, C., Shimizu,
H. and Shioya, S. (2007). Extracting the hidden features in saline osmotic tolerance in
Saccharomyces cerevisiae from DNA microarray data using the self-organizing map:
biosynthesis of amino acids. Appl Microbiol Biotechnol 75, 415-26.
Pang, S. S. and Duggleby, R. G. (1999). Expression, purification, characterization, and
reconstitution of the large and small subunits of yeast acetohydroxyacid synthase.
Biochemistry 38, 5222-31.
Park, C. S., Park, Y. J., Lee, Y. H., Park, K. J., Kang, H. S. and Pek, U. H. (1990). The novel
genetic manipulation to improve the plasmid stability and enzyme activity in the
recombinant brewing yeast, Technical Quarterly of the Master Brewers Association of
the Americas, pp. 112-116.
131
Pedersen, M. (1986a). DNA sequence polymorphism in the genus Saccharomyces. III.
Homologous chromosome III of Saccharomyces bayanus, S. carlsbergensis and S.
uvarum. Carlsberg Res Commun 51, 163.
Pedersen, M. (1986b). DNA sequence polymorphism in the genus Saccharomyces. IV.
Restriction endonuclease fragment patterns of chromosomal regions in brewing and
other yeast strains. Carlsberg Res Commun 51, 185.
Peinado, J. M. and Loureiro-Dias, M. C. (1986). Reversible loss of affinity induced by
glucose in the maltose-H+ symport of Saccharomyces cerevisiae. Biochim Biophys
Acta 856, 189-92.
Perry, C. and Meaden, P. (1988). Properties of a genetically engineered dextrin-fermenting
strain of brewer´s yeast. J Inst Brew 94, 64.
Petersen, E. E., Mensour, N. A., Margaritis, A., Stewart, R. J. and Pilkington, P. H. (2004).
The effects of wort valine concentration on the total diacetyl profile and levels late in
batch fermentations with brewing yeast Saccharomyces carlsbergensis. American
Society of Brewing Chemists 62, 131-139
Piper, M. D., Daran-Lapujade, P., Bro, C., Regenberg, B., Knudsen, S., Nielsen, J. and Pronk,
J. T. (2002). Reproducibility of oligonucleotide microarray transcriptome analyses. An
interlaboratory comparison using chemostat cultures of Saccharomyces cerevisiae. J
Biol Chem 277, 37001-8.
Polaina, J. (2002). Brewer's yeast: Genetics and Biotechnology. Applied Mycology and
Biotechnology 2.
Pope, G. A., MacKenzie, D. A., Defernez, M., Aroso, M. A., Fuller, L. J., Mellon, F. A.,
Dunn, W. B., Brown, M., Goodacre, R., Kell, D. B., Marvin, M. E., Louis, E. J. and
Roberts, I. N. (2007). Metabolic footprinting as a tool for discriminating between
brewing yeasts. Yeast 24, 667-79.
Price, V., Taylor, W., Clevenger, W., Worthington, M. and Young, E. (1990). Expression of
heterologous proteins in Saccharomyces cerevisiae using the ADH2 promoter.
Methods Enzymol. 185, 308–318.
Rainieri, S., Kodama, Y., Kaneko, Y., Mikata, K., Nakao, Y. and Ashikari, T. (2006). Pure
and mixed genetic lines of Saccharomyces bayanus and Saccharomyces pastorianus
and their contribution to the lager brewing strain genome. Appl Environ Microbiol 72,
3968-74.
Riou, C., Nicaud, J. M., Barre, P. and Gaillardin, C. (1997). Stationary-phase gene expression
in Saccharomyces cerevisiae during wine fermentation. Yeast 13, 903-15.
Sakai, K., Fukui, S., Yabuuchi, S., Aoyagi, S. and Tsumura, Y. (1989). Expression of the
Saccharomyces diastaticus STAI gene in brewing yeasts. Journal of the American
Society of Brewing Chemists 47, 87-91.
Salema-Oom, M., Valadao Pinto, V., Goncalves, P. and Spencer-Martins, I. (2005).
Maltotriose utilization by industrial Saccharomyces strains: characterization of a new
member of the alpha-glucoside transporter family. Appl Environ Microbiol 71, 5044-
9.
Santyanarayana, T., Umbarger, H. E. and Lindengren, G. (1968). Biosynthesis of branched-
chain amino acids in yeast: regulation ofleucine biosynthesis in prototrophic and
leucine auxotrophic strains. J.Bacteriol 96, 2018-2024.
Sato, M., Maeba, H., Watari, J. and Takashio, M. (2002). Analysis of an inactivated Lg-FLO1
gene present in bottom-fermenting yeast. J Biosci Bioeng 93, 395-8.
Sauer, U. and Schlattner, U. (2004). Inverse metabolic engineering with phosphagen kinase
systems improves the cellular energy state. Metab Eng 6, 220-8.
Schmitt, M. E., Brown, T. A. and Trumpower, B. L. (1990). A rapid and simple method for
preparation of RNA from Saccharomyces cerevisiae. Nucleic Acids Res 18, 3091-2.
132
Sergio L. Alves, J., Herberts, R. A., Hollatz, C., Trichez, D., Miletti, L. C., de Araujo, P. S.
and Stambuk, B. U. (2008). Molecular Analysis of Maltotriose Active Transport and
Fermentation by Saccharomyces cerevisiae Reveals a Determinant Role for the AGT1
Permease. Appl Environ Microbiol 74, 1494-1501.
Sieiro, C., Reboredo, N. M., Blanco, P. and Villa, T. G. (1997). Cloning of a new FLO gene
from the flocculating Saccharomyces cerevisiae IM1-8b strain. FEMS Microbiol Lett
146, 109-15.
Siro, M.-R. and Lo¨vgren, T. (1979). Influence of glucose on the a-glucoside permease
activity of yeast. Eur. J. Appl. Microbiol. 7:59–66.
Sone, H., Fujii, T., Kondo, K., Shimuzu, F., Tanaka, J. and Inoue, T. (1988). Nucleotide
sequence and expression of the Enterobacter aerogenes alpha-acetolactate
decarboxlase gene in brewer´s yeast. Appl Env Microbiol 54, 38.
Sone, H., Kondo, K., Fujii, T., Shimuzu, F., Tanaka, J. and Inoue, T. (1987). Fermentation
properties of brewer´s yeast having -acetolactate decarboxlase gene Proceedings of
the 21st Congress. Madrid, pp. 545.
Stewart, G. G., Russell, I. and Sills, A. M. (1983). Factors that control the utilisation of wort
carbohydrates by yeast. Tech. Q. Master Brew. Assoc. Am 20, 1-8.
Stratford, M. (1989). Yeast flocculation: calcium specifity. Yeast 5, 487.
Stratford, M. (1992). Yeast flocculation: reconcilation of physiology and genetic viewpoints.
Yeast 8, 25.
Stratford, M. and Assinder, S. (1991). Yeast flocculation: Flo1 and NewFlo phenotypes and
receptor structure. Yeast 7, 559.
Tamai, Y., Momma, T., Yoshimoto, H. and Kaneko, Y. (1998). Co-existence of two types of
chromosome in the bottom fermenting yeast, Saccharomyces pastorianus. Yeast 14,
923-33.
Teunissen, A. W. and Steensma, H. Y. (1995). Review: the dominant flocculation genes of
Saccharomyces cerevisiae constitute a new subtelomeric gene family. Yeast 11, 1001-
13.
Tezuka, H., Mori, T., Okumura, Y., Kitabatake, K. and Tsumara, Y. (1992). Cloning of a gene
supressing hydrogen sulfide production by Saccharomyces cerevisiae and its
expression in a brewing yeast. J Am Soc Brew Chem 50 130.
Urano, N., Sahara, H. and Koshino, S. (1993). Conversion of a non-flocculent brewer's yeast
to flocculent ones by electrofusion. I: Identification and characterization of the fusants
by pulsed field gel electrophoresis. Journal of biotechnology 28, 237-247.
Vakeria, D., Box, W. and Hinchliffe, E. (1991). The control of gene expression in brewing
yeast with anti-sense RNA. Proceedings of the 23rd Congress Lisbon pp. 305.
Vakeria, D. and Hinchliffe, E. (1989). Amylolytic brewing yeast: their commercial and
legislative acceptability. Proceeding of the 22nd Congress Zürich, pp. 475.
Vaughan-Martini, A. and Kurztman, C. (1985). Deoxyribonucleic Acid Relatedness among
Species of Saccharomyces sensu stricto. Int J Syst Bacteriol 35, 508-511.
Vaughan-Martini, A. V. and Martini, A. (1987). Three newly delimited species of
Saccharomyces sensu stricto. Antonie Van Leeuwenhoek 53, 77-84.
Verstrepen, K. J., Derdelinckx, G., Delvaux, F. R., Winderickx, J., Thevelein, J. M., Bauer, F.
F. and Pretorius, I. S. (2001). Late fermentation expression of FLO1 in
Saccharomyces cerevisiae. Am. Soc. Brew. Chem. 59:, 69–76.
Verstrepen, K. J., Jansen, A., Lewitter, F. and Fink, G. R. (2005). Intragenic tandem repeats
generate functional variability. Nat Genet 37, 986-90.
Verstrepen, K. J., N. Moonjai, G. Derdelinckx, J.-P. Dufour, J. Winderickx, J. M. Thevelein,
I. S. Pretorius and Delvaux, F. R. (2003a). Genetic regulation of ester synthesis in
brewer's yeast: new facts, insights and implications for the brewer, Brewing yeast
fermentation performance, 2nd ed., Blackwell Science.
133
Verstrepen, K. J., Van Laere, S. D., Vanderhaegen, B. M., Derdelinckx, G., Dufour, J. P.,
Pretorius, I. S., Winderickx, J., Thevelein, J. M. and Delvaux, F. R. (2003b).
Expression levels of the yeast alcohol acetyltransferase genes ATF1, Lg-ATF1, and
ATF2 control the formation of a broad range of volatile esters. Appl Environ
Microbiol 69, 5228-37.
Vidgren, V., Ruohonen, L. and Londesborough, J. (2005). Characterization and functional
analysis of the MAL and MPH Loci for maltose utilization in some ale and lager yeast
strains. Appl Environ Microbiol 71, 7846-57.
Villanueba, K., Gossens, E. and Masschelein, C. (1990). Sub-threshold vicinal diketone levels
in lager brewing yeast fermentations by means of ILV5 gene amplification. J Am Soc
Brew Chem 48, 111.
Virkajarvi, I. (2006). Accelerated processing of beer in Bamforth, C. (Ed), Brewing: New
Technology, Woodhead Publishing Limited.
Wang, D., Wang, Z., Liu, N., He, X. and Zhang, B. (2008). Genetic modification of industrial
yeast strains to obtain controllable NewFlo flocculation property and lower diacetyl
production. Biotechnol Lett.
Watari, J., Takata, Y., Murakami, J. and Koshino, S. (1991). Breeding of flocculent brewer´s
yeast by genetic engineering Proceedings of the 23rd Congress Lisbon pp. 297.
Watari, J., Takata, Y., Ogawa, M., Sahara, H., Koshino, S., Onnela, M. L., Airaksinen, U.,
Jaatinen, R., Penttila, M. and Keranen, S. (1994). Molecular cloning and analysis of
the yeast flocculation gene FLO1. Yeast 10, 211-25.
Westerfeld, W. W. (1945). A colorimetric determination of blood acetoin. J Bio Chem 161,
495-502.
Winston, F., Dollard, C. and Ricupero-Hovasse, S. L. (1995). Construction of a set of
convenient Saccharomyces cerevisiae strains that are isogenic to S288C. Yeast 11, 53-
5.
Winzeler, E. A., Castillo-Davis, C. I., Oshiro, G., Liang, D., Richards, D. R., Zhou, Y. and
Hartl, D. L. (2003). Genetic diversity in yeast assessed with whole-genome
oligonucleotide arrays. Genetics 163, 79-89.
Winzeler, E. A., Lee, B., McCusker, J. H. and Davis, R. W. (1999). Whole genome genetic-
typing in yeast using high-density oligonucleotide arrays. Parasitology 118 Suppl,
S73-80.
Wodicka, L., Dong, H., Mittmann, M., Ho, M. H. and Lockhart, D. J. (1997). Genome-wide
expression monitoring in Saccharomyces cerevisiae. Nat Biotechnol 15, 1359-67.
Yamagishi, H. and Ogata, T. (1999). Chromosomal structures of bottom fermenting yeasts.
Syst Appl Microbiol 22, 341-53.
Yamano, S., Tanaka, J. and Inoue, T. (1994). Cloning and expression of the gene encoding
alpha-acetolactate decarboxylase from Acetobacter aceti ssp. xylinum in brewer's
yeast. J Biotechnol 32, 165-71.
Yarrow, D. (1984). Saccharomyces Meyen ex Reess, The Yeasts, A Taxonomic Study, 3rd
edn, Elsevier, pp. 379-395.
Yocum, R. (1986a). Genetic engineering of industrial yeasts. Proc Bio Expo 86. Boston pp.
171.
Yocum, R. (1986b). Genetic engineering of industrial yeasts. Proc Bio Expo 86 Boston pp.
171.
Yoshida, S., Imoto, J., Minato, T., Oouchi, R., Sugihara, M., Imai, T., Ishiguro, T., Mizutani,
S., Tomita, M., Soga, T. and Yoshimoto, H. (2008). Development of bottom-
fermenting saccharomyces strains that produce high SO2 levels, using integrated
metabolome and transcriptome analysis. Appl Environ Microbiol 74, 2787-96.
134
Zhang, Y., Wang, Z. Y., He, X. P., Liu, N. and Zhang, B. R. (2008). New industrial brewing
yeast strains with ILV2 disruption and LSD1 expression. Int J Food Microbiol 123,
18-24.
