Investigation of mature biofilm populations in the distribution of drinking water with attention to bacteria of hygienic relevance [original]

Technische Universität Berlin

Investigation of mature biofilm populations in the

distribution of drinking water with attention to bacteria of

hygienic relevance

vorgelegt von

Diplom-Biologin Bianca Conradi

aus Greetsiel

von der Fakultät III-Prozesswissenschaften

der Technischen Universität Berlin

zur Erlangung des akademischen Grades

Doktorin der Naturwissenschaften

-Dr. rer. nat.-

genehmigte Dissertation

Promotionsausschuss

Vorsitzender: Prof. Dr. rer. nat. Wolfgang Rotard, TU Berlin

Berichter: 1. Prof. Dr. rer. nat. Ulrich Szewzyk, TU Berlin

Berichterin: 2. Prof. Dr. rer. nat. Isolde Röske, TU Dresden

Tag der wissenschaftlichen Aussprache: 31.03.2011

Berlin 2011

D83

meiner Großmutter E. Hübel

1 INTRODUCTION....................................................................................................................................... 1

2 MATERIALAND METHODS................................................................................................................... 7

2.1 REACTOR SYSTEMS AT DIFFERENT LOCATIONS ...................................................................................... 7

2.1.1 Reactor system Berlin (Germany)................................................................................................... 7

2.1.1.1 Setup and function................................................................................................................................. 7

2.1.1.2 Sampling of material coupons............................................................................................................... 9

2.1.2 Reactor systems Duisburg (Germany)............................................................................................. 9

2.1.3 Reactor system Lundtofte (Denmark)............................................................................................ 10

2.2 PIPE SECTIONS FROM THE DRINKING WATER DISTRIBUTION SYSTEM.....................................................11

2.2.1 Free water samples of the distribution system .............................................................................. 12

2.2.2 Treatment of pipe deposits on the inner surface............................................................................ 12

2.2.2.1 Treatment of PVC pipes...................................................................................................................... 12

2.2.2.2 Treatment of metallic pipes................................................................................................................. 13

2.3 ANALYSIS OF BACTERIAL POPULATIONS BY CULTIVATION TECHNIQUES............................................... 13

2.3.1 Aerobic cultivation on modified R2A medium............................................................................... 13

2.3.2 Heterotrophic plate counts according to DIN EN ISO 6222......................................................... 13

2.3.3 E. coli and coliform bacteria according to DIN 38 411 K 6.......................................................... 14

2.3.4 Aerobic cultivation of P. aeruginosa according to DIN EN 12780................................................ 14

2.4 INVESTIGATION OF THE BACTERIAL POPULATION BY CULTURE INDEPENDENT METHODS ..................... 15

2.4.1 Total cell counts (TCC) determined by DAPI staining.................................................................. 15

2.4.1.1 Staining of biofilm suspensions........................................................................................................... 15

2.4.1.2 Staining on filter membrane ................................................................................................................ 15

2.4.1.3 Staining on coupons ............................................................................................................................ 16

2.4.1.4 Microscopic examination .................................................................................................................... 16

2.4.2 Fluorescence in situ hybridization (FISH).................................................................................... 16

2.4.2.1 Fixation of biofilm coupons, suspensions, and pure cultures .............................................................. 16

2.4.2.2 Hybridization procedure...................................................................................................................... 17

2.4.2.3 Development of a new oligonucleotide probe..................................................................................... 19

2.4.2.4 Oligonucleotide probes used in this study........................................................................................... 19

2.4.3 Extraction of total DNA from biofilm suspensions........................................................................ 20

2.4.3.1 Simple preparations of DNA from formaldehyde fixed and non-fixed biofilm suspensions............... 20

2.4.3.1.1 M IV (formaldehyde fixed) and BWB III (non-fixed) serial diluted.............................................. 20

2.4.3.1.2 Alkaline lysis of formaldehyde fixed M IV.................................................................................... 20

2.4.3.1.3 Alkaline lysis, enhanced sample volume, and ethanol precipitation .............................................. 21

2.4.3.1.4 Alkaline lysis, enhanced sample volume, and isopropanol precipitation ....................................... 21

2.4.3.1.5 Addition of bovine serum albumin (BSA) ..................................................................................... 22

2.4.3.2 FastDNASpin Sample Kit for soil...................................................................................................... 23

2.4.3.2.1 Extraction of DNAfrom biofilm suspensions................................................................................ 23

2.4.3.2.2 Evaluation of extraction efficiency ................................................................................................ 24

2.4.3.3 CTAB (hexadecyltrimethylammonium bromide) extraction ............................................................... 25

2.4.3.3.1 General extraction procedure......................................................................................................... 25

2.4.3.3.2 Variations of the protocol............................................................................................................... 27

2.4.3.4 DNAextraction with QIAamp DNAMini Kit..................................................................................... 27

2.4.3.5 Extraction and purification by Qiagen Genomic-tips 20 ..................................................................... 28

2.4.3.6 Verification of the extraction success .................................................................................................. 30

2.4.3.6.1 Control PCR................................................................................................................................... 30

2.4.3.6.2 Measurement of DNA.................................................................................................................... 32

2.4.4 Sequencing of bacterial 16S rDNA ............................................................................................... 33

2.4.4.1 Alkaline lysis....................................................................................................................................... 33

2.4.4.2 Sequencing reaction ............................................................................................................................ 33

2.4.5 RFLP analysis of the 16S rRNA gene of isolates........................................................................... 34

2.4.6 Phylogenetic analysis.................................................................................................................... 35

2.4.6.1 Pipe sample isolates............................................................................................................................. 35

2.4.6.2 Representative reactor sample isolates obtained from RFLP analysis................................................. 35

2.5 STATISTICALANALYSIS........................................................................................................................ 36

3 RESULTS................................................................................................................................................... 37

3.1 PIPE SAMPLES TAKEN FROM THE DISTRIBUTION SYSTEMS IN BERLIN AND THE RUHRGEBIET .............. 37

3.1.1 Macroscopic description of deposits on the inner pipe surface .................................................... 39

3.1.2 Total cell counts and culturable bacteria on the different pipe materials..................................... 39

3.1.3 Phylogenetic bacterial groups detected in the pipe samples......................................................... 45

3.2 RESULTS REACTOR SYSTEM BERLIN.................................................................................................... 56

3.2.1 Operation of the Berlin reactor system ......................................................................................... 56

3.2.2 P. aeruginosa in the bulk water phase of the reactor system in Berlin.......................................... 57

3.2.2.1 Specificity of the DIN EN 12780........................................................................................................ 58

3.2.2.2 Growth of isolated P. aeruginosa on modified R2Amedium.............................................................. 58

3.2.2.3 Detection of P. aeruginosa in the bulk water phase of the reactor system........................................... 59

3.2.3 Further microbiological investigations of the bulk water phase................................................... 61

3.2.3.1 Heterotrophic bacteria in the bulk water phase.................................................................................... 61

3.2.3.2 Statistical analysis of the data from the Berlin reactor system............................................................. 63

3.2.3.3 E. coli and coliform bacteria according to DIN 38 411 K 6 ................................................................ 65

3.2.4 Bacterial population on material coupons of the reactor systems in Germany............................. 66

3.2.4.1 Total cell counts (TTC) on material coupons ...................................................................................... 66

3.2.4.2 Culturable bacteria on six months exposed material coupons ............................................................. 70

3.2.4.3 Characterization of the biofilm isolates with specific oligonucleotid probes ...................................... 70

3.2.4.3.1 Verification of the new oligonucleotid probe................................................................................. 71

3.2.4.3.2 Identification of isolates grown on material coupons in the Berlin reactor system........................ 72

3.2.5 Culturable biofilm population of the two reactor systems in Germany and Denmark.................. 73

3.2.5.1 Pre-screening of drinking water isolates by RFLP analysis................................................................. 73

3.2.5.2 OTUs detected on biofilm and in bulk water....................................................................................... 75

3.2.5.3 Phylogenetic groups detected in the two reactor systems in Denmark and Germany.......................... 76

3.3 DNAEXTRACTION METHODS.............................................................................................................. 77

3.3.1 Inhibition and limitation of PCR by sample quality and preparation........................................... 77

3.3.1.1 Does sample composition, preparation or concentration effect PCR effectivity?................................ 77

3.3.1.2 Does alkaline lysis in formaldehyde fixed biofilm suspensions interfere with PCR amplification?.... 78

3.3.1.3 Do different DNAconcentration methods combined with enhanced sample volume effect DNA

amplification?........................................................................................................................................................... 80

3.3.1.4 Does bovine serum albumine (BSA) reduce the effect of inhibitory agents? ...................................... 81

3.3.2 Complex DNA extraction methods ................................................................................................ 82

3.3.2.1 FastDNASpin Sample Kit for soil...................................................................................................... 82

3.3.2.1.1 Concentration of DNAobtained..................................................................................................... 82

3.3.2.1.2 Dilution of DNA ............................................................................................................................ 83

3.3.2.1.3 Evaluation of extraction efficiency ................................................................................................ 83

3.3.2.2 CTAB based DNAextraction.............................................................................................................. 85

3.3.2.3 Extraction and purification by the QIAamp DNA Mini Kit ................................................................ 87

3.3.2.4 Extraction and purification by the Qiagen Genomic-tips 20................................................................ 88

4 DISCUSSION............................................................................................................................................. 91

4.1 BACTERIAL POPULATIONS IN MATURE DRINKING WATER BIOFILMS...................................................... 91

4.1.1 Investigated drinking water systems.............................................................................................. 91

4.1.2 Phylogenetic composition of the culturable population of the middle aged and old biofilms....... 91

4.1.2.1 Heterotrophic plate count (HPC) bacteria in the bulk water phase of the Berlin reactor system......... 96

4.1.3 Regrowth potential of the opportunistic pathogen P. aeruginosa in the Berlin reactor sytem.... 103

4.1.4 Total cell counts and heterotrophic plate count bacteria in middle aged and old biofilms..........110

4.2 LIMITATIONS OF MOLECULAR TECHNIQUES IN MATURE DRINKING WATER BIOFILMS ..........................117

4.2.1 Effect of sample quality and preparation on PCR........................................................................117

4.2.2 Complex DNA extraction methods .............................................................................................. 121

4.2.3 Methodological Perspective........................................................................................................ 124

4.3 IMPACT OF BACTERIAWITH PATHOGENIC POTENTIAL IN THE INVESTIGATED SYSTEMS....................... 125

5 OUTLOOK .............................................................................................................................................. 129

6 SUMMARY.............................................................................................................................................. 131

7 ZUSAMMENFASSUNG......................................................................................................................... 133

DANKSAGUNG................................................................................................................................................ 136

REFERENCES.................................................................................................................................................. 138

ABBREVIATIONS ........................................................................................................................................... 149

Introduction

1 Introduction

Nowadays, despite of the good drinking water quality in industrialized countries, dis-

eases related to drinking water have been reported. For example the “Centers for

Disease Control and Prevention” in the U. S. reported that in 1999 and 2000 twenty

outbreaks of waterborne disease could be associated with pathogens (Centers for

Disease Control and Prevention 2002). In addition, the World Health Organization

described an outbreak of E. coli O157 in Walkerton, Ontario, Canada in 2000 that

resulted in the death of five people and more than two dozen hospitalized people

(Leclerc et al. 2002; WHO 2000). This remains of the importance of high microbi-

ological quality of drinking water.

Despite the author is aware that in addition to the microbial item the chemical safety

of drinking water is the second point of major concern (Anonymous 2008; Heberer

2002), this study focussed on the impact of bacteria on quality of drinking water. Re-

garding the subject of microbiological drinking water quality the reviews of Szewzyk

et al. 2000, Leclerc et al. 2002 and the WHO guidelines for drinking water quality of

2008 give a comprehensive overview of pathogens belonging to bacteria, viruses,

protozoa, and Helminths with concern in drinking water (Anonymous 2008; Leclerc et

al. 2002; Szewzyk et al. 2000). Szewzyk et al. 2000 divided bacterial pathogens in

two groups, those with fecal origin and those that originated in water or soil. The first

group comprises bacteria like Campylobacter species, enterohemorrhagic Es-

cherichia coli,Salmonella species, Vibrio cholerae,Yersinia enterocolitica, or Helico-

bacter pylori. The second group of pathogens inhabit water or soil and are trans-

ported from these habitats into drinking water. These bacteria are able to grow if

parameters are getting appropriate and include representatives of Legionella species,

Pseudomonas aeruginosa,Aeromonas species, Acinetobacter species, and envi-

ronmental Mycobacteria. (Szewzyk et al. 2000) Furthermore, Mycobacterium avium,

Legionella pneumophila and Legionella species as well as Campylobacter species

and P. aeruginosa have been shown to survive in drinking water biofilms (Buswell et

al. 1998; Lehtola et al. 2007b; Moritz et al. 2010; Rogers et al. 1994).

In oligotrophic environments as drinking water with a low nutrient content of the free

water phase the formation of biofilms at solid-liquid interfaces has been described as

Introduction

typical (Fletcher and Marshall 1982). Moreover, it has been found for example by Van

der Wende et al. and Block et al. that biofilms in drinking water systems play an im-

portant role in contamination of the water phase (Block et al. 1993; van der Wende

and Characklis 1990). The rough calculation of Flemming that approximately 95 % of

the biomass in distribution systems is found in the biofilms that are not routinely ex-

amined and only 5 % in the water phase reflect the importance of the biofilms (Flem-

ming 2003).

In the last decades, a lot of work has been done in different natural and artificial

drinking water systems. These previous investigations on drinking water biofilms dif-

fered in four main aspects: The drinking water resource, systems in which the

biofilms were exposed, the substratum for development of microbial biofilms, the time

surfaces were exposed, and the methods applied to investigate the biofilm. For an

overview a selection of drinking water studies will be described.

The systems used for development or exposition of biofilms can be divided in three

categories. The smallest systems are devices of a portable dimension which allow a

simple exchange of biofilm coupons and can be installed easily in different locations.

Pedersen et al. used a rectangular box (polycarbonate and aluminium) with an o-ring

sealed lid, a test pile and two diffusors inside. They investigated biofilm development

on hydrophilic stainless steel and hydrophobic PVC surfaces for 4 to 5.6 months

(Pedersen 1990). Donlan et al. exposed test cylinders of cast iron in a device incor-

porated in water mains for up to 3.8 months (Donlan et al. 1994). Furthermore, Hal-

lam et al. fitted 45 cm pipe sections of MDPE, PVC, or cement between two PVC end

plates and exposed them for 21 days (Hallam et al. 2001). Niquette and co-workers

constructed a PVC cylinder for the exposition of coupons. They investigated different

materials that are used to different extent in distribution systems (PVC, PE, cemented

steel, asbestos-cement, cemented cast iron, tarred steel and grey iron) and exposed

them for 2 to 8 months. (Niquette et al. 2000) The materials PVC and cement were

also investigated by Camper et al.. In addition, ductile iron and an epoxy material

were exposed in an annular reactor for 3 to 8 months. (Camper et al. 2003) Zacheus

et al. constructed a combined system of a basin-like device with coupons and pipe

sections. They exposed the coupons of PVC and stainless steel from one week to 4.4

months and the PVC and PE pipe sections up to about 5.4 months. (Zacheus et al.

Introduction

2000) A cylinder-shaped stainless steal device named “Robbin`s device” or modifica-

tions of this device were used by several investigators. Kalmbach et al. exposed the

materials PE, PVC and as a non supporting material glass in this system. Incubation

was done 14 to 70 days. (Kalmbach et al. 1997b; Kalmbach 1998) An incubation time

of 8 to 15 days for drinking water biofilms is found in Schwartz et al. who investigated

steel and copper additionally to HDPE and PVC (Schwartz et al. 1997; Schwartz et

al. 1998). A modified Robbin`s device was also used by Kerr et al.. They exposed

cast iron, MDPE, and PVC up to 10 months. (Kerr et al. 1999)

Others used pipe systems of a greater dimension in which operating conditions were

expected to be more similar to the distribution system. Percival et al. 1998 studied

different grades of stainless steal that were exposed to drinking water for one and

two years. The stainless steel coupons were sorted horizontal to the flow direction in

the middle of eight centimetre stainless steel pipe sections. (Percival et al. 1998a;

Percival et al. 1998b) Martiny et al. also used a pipe system and investigated biofilms

on coupons installed in test plug modules from one day up to approximately three

years (Martiny et al. 2003). In a pilot-scale system with removable pipe sections of

PVC and iron pipes Norton et al. exposed the sections one to eight weeks before

scraping and washing the biofilms from the interior of PVC and iron pipes (Norton

and LeChevallier 2000). One month biofilms of a pilot plant with six pipe loops in

which removable coupons of PVC and cement were exposed have been investigated

by Block et al. (Block et al. 1993). Frias et al. investigated exposed PE surfaces up to

66 days in a pilot system of 200 m length with a diameter of 1.5 cm (Frias et al.

2001). Lethola et al. used a pilot scale system of PE and copper pipes for exposure

of biofilms for approximately five months (Lehtola et al. 2006). Deines et al. investi-

gated up to 11 days old biofilms on coupons developed in a pilot pipe loop system

constructed from actual distribution system PE pipes (Deines et al. 2010).

The last category describes pipe sections directly taken out of the drinking water dis-

tribution system. An early study was done by Olson and Ridgeway in 1981 who in-

vestigated approximately 40 years old biofilms developed on cement lined iron and a

galvanized iron pipe removed from the distribution system (Olson et al. 1981). Hallam

et al. removed HDPE pipes of 18 months and 10 years out of the distribution system

(Hallam et al. 2001). Coupon samples of 8 to 90 years old cast iron, cement lined

Introduction

cast iron, ductile iron, and asbestos cement pipes were sampled by LeChevallier et

al.. They also investigated deposits obtained from mechanical cleaning (pigging) or

scraping of pipe surfaces with a sterile spatula. (LeChevallier et al. 1987)

Traditionally, for the description of drinking water quality cultivation techniques have

been used and have been transferred to biofilm investigations. They described the

heterotrophic bacteria or focussed on indicator or pathogenic bacteria. Cultivation

has been used in several of the above described systems and further studies for in-

vestigations of drinking water biofilms (Block et al. 1993; Carter et al. 2000; Donlan et

al. 1994; Dutkiewicz and Fallowfield 1998; Percival et al. 1998b). Since scientists in

microbial ecology got aware of the limitation that cultivation only detects a small per-

centage of the total bacterial population, they looked for methods to circumvent this

bias for phylogenetic identification and quantification of bacteria (Amann and Kühl

1998; Ward et al. 1990). One favourite method for determination of total bacterial cell

counts is staining of double stranded DNA by fluorescent dyes like DAPI or acridine

orange and subsequent epi-fluorescent microscopy (Block et al. 1993; Schwartz et

al. 1998; Zacheus et al. 2000). For phylogenetic identification Kalmbach et al. and

Schwartz et al. used total cell counts in combination with rRNA targeted oligonucleo-

tide probes of different phylogenetic levels (Kalmbach 1998; Kalmbach et al. 2000;

Schwartz et al. 1998). The CARD(catalyzed reported deposition)-FISH method with

the advantage of enhanced signal intensity was successfully applied by Deines et al.

in 3 to 11 days old biofilms (Deines et al. 2010). With further development of molecu-

lar techniques in microbial ecology scientists investigated genomic DNA in drinking

water biofilms. Schwartz et al. used the polymerase chain reaction in combination

with southern blot hybridization for identification of facultative pathogens in a drinking

water biofilm (Schwartz et al. 1998). More recently Martiny and co-workers identified

bacterial DNA after extraction of total genomic DNA of drinking water biofilms grown

on stainless steel by cloning with subsequent sequencing (Martiny et al. 2003). In

addition, for comparison of population profiles of different biofilm samples on a

rougher phylogenetic level they used a fingerprinting technique. Therefore, after DNA

extraction and PCR Martiny et al. used 16S rDNA targeted terminal restriction frag-

ment length polymorphism (T-RFLP) and were able to detect correlations between

young (1 to 94 days) and old (571 to 1093 days) biofilms and the population profiles

(Martiny et al. 2003). Röder et al. used the fingerprinting technique DGGE (denatur-

Introduction

ing gradient gel electrophoresis) in disinfection experiments. Drinking water biofilms

were grown 10 to 38 months on bacterial growth supporting silicone rubber tubes

before disinfection. (Roeder et al. 2010) This technique was also used by Deines in 3

to 11 days old biofilms (Deines et al. 2010). Schmeisser et al. 2003 also analyzed the

bacterial population on a rubber-coated valve taken out of the distribution system by

a cloning sequencing approach but also did the next step from phyologenetic analy-

sis to the description of the metabolic potential by a metagenome analysis (Schmeis-

ser et al. 2003).

In addition to the above described selection of investigations furthermore work has

been done and provided a lot of invaluable insights about biofilms in drinking water.

In general different factors influencing the microbiological water quality direct or indi-

rect have been of interest including e. g. water source, concentration and kind of dis-

infectant, temperature, hydraulic conditions, assimilable organic carbon (AOC), age

of biofilms, material used for biofilm development and survival potential of selected

bacteria with focus on pathogen potential. Nevertheless, the findings of the investiga-

tion of the last decades left unanswered questions:

Is the bacterial composition of old biofilms comparable to the primarily de-

scribed young biofilms?

Are biofilms of the drinking water distribution system a reservoir for opportunis-

tic pathogenic, pathogenic or indicator bacteria and therefore a risk to human

health?

In the present study a biofilm reactor was constructed out of PE pipe sections with a

dimension that is routinely used in the Berlin distribution system. In each pipe sec-

tion, coupons of glass, copper, PE, stainless steel, and PVC were installed on the

inner surface in horizontal direction to the flow. The system was supplied directly by a

drinking water distribution system pipe. In this reactor system materials were ex-

posed 6 to 24 months and determined as “middle aged biofilms”. Furthermore, pipe

samples were taken from the actual distribution system in the Ruhrgebiet and Berlin.

The investigation focussed mainly on pipe materials of relevance in Germany like

cast iron, grey cast iron, and cement lined cast iron as well as PE and PVC. These

pipe samples have been exposed for 8 to 99 years in the distribution system and are

therefore named as “old biofilms” in this study. For investigation of the mature

Introduction

biofilms of these systems it was the aim to get a comprehensive insight by use of cul-

tivation in combination with molecular techniques.

This work was part of the cooperation project with focus on “Detection of growth and

the contamination risk of biofilms in distribution of drinking water” (Flemming 2003).

The dimension of the project and the need of specific methods and experience re-

sulted in the organization in a network of specialized scientists and drinking water

suppliers in Germany. The relevant organisms and project partners are summarized

in the subsequent table.

Investigated microorganisms Project partner

Pseudomonas sp., Legionella sp.,

E. coli, coliform bacteria H.-C. Flemming, J. Wingender

Universität Duisburg

Mycobacteria R. Schluze-Röbbecke, B. Ilg

Universität Düsseldorf

Aeromonas sp. R.H.W. Schubert

Universität Frankfurt

Campylobacter sp. and Yersinia sp. I. Feuerpfeil, A. Hummel

Federal Environmental Agency Germany,

Bad Elster

Helicobacter pylori M. Exner, A. Rechenburg

Universität Bonn

Cryptosporidium and Giardia M. Exner, C. Koch

Universität Bonn

Viruses K. Botzenhart

Universität Tübingen

Amoeba R. Michel, R. Hoffmann

Ernst-Rodenwald-Institute Koblenz

Fungi E. Göttlich, H.-C. Flemming

IWW, Mülheim a. d. Ruhr

Material and Methods

2 Material and Methods

2.1 Reactor systems at different locations

2.1.1 Reactor system Berlin (Germany)

2.1.1.1 Setup and function

The Berlin reactor system was designed in cooperation with the Berliner Wasserbetriebe

(BWB, the local water supplier in Berlin) and built up in a technical supply station of the Ber-

liner Wasserbertriebe in Berlin Lichterfelde. The drinking water influx in the system was sup-

plied via a 25 m (diameter: 8 cm) ductile cast pipe from the main pipe near the supply station.

After passage through the system the water was discarded. Construction and function had to

be as equivalent as possible to the real drinking water distribution system. Only material rou-

tinely used in drinking water distribution systems was utilized. In PE pipe sections with an

inner diameter of 10 cm and a height of each segment of 15.5 cm (hole column of 5 seg-

ments: 77.5 cm), the coupons (consisting of PE-HD: polyethylene high density, PVC: polyvi-

nylchloride, Cu: copper, V2A-steel, and glass) were fit to the inner surface of the segments in

flow direction (figure 1). PE, PVC, Cu and V2A-steel are materials which can be found in the

dinking water distribution system and in house installations. The coupons were treated by the

laboratory of the BWB with the purifier Extron (Merck, Germany) and rinsed with sterile des-

tilled water, incubated for some minutes in isopropanol (70 %) and air dried before being in-

serted in the segments. Each of the five reactors arranged in series included one hundred

coupons of the different materials that were allocated to the project partners. The valves at

the inlet and outlet of each reactor allowed sampling of one reactor without the need of drain-

ing all reactors. Each reactor was equipped with a fire resistant sampling valve at the outlet

for taking bulk water samples. The median water flow was 5 m³/h (range 0.5 to 45 m³/h) to

day 134 and 0.4 m³/h (range 0.05 to 3 m³/h) thereafter. During the first operation phase up to

day 110 flushing operations with enhanced flow rate (range 9 to 48 m³/h, lasting 2 to 19 h)

were performed 3 to 5 days every week. Furthermore, stagnation operations of 4 to 6.5 h and

from day 235 stagnation times of 16.5 to 19.5 h were done. The bulk water phase was regu-

larly investigated for heterotrophic plate counts (GDWR 1990, chapter 2.3.2) and P. aerugi-

nosa (DIN EN 12780, chapter 2.3.4). Coupons were exposed to drinking water for 6, 12, 18,

and 24 months.

Material and Methods

Fig. 1: Overview of the reactor system in Berlin (left). Two PE pipe sections with copper

coupons inside (right).

Fig. 2: Layout of the Berlin reactor system with five reactors arranged in series. typ a)

valve for regulation of flow direction typ b) valve used during coupon sampling typ c) fire

resistant valve used for water sampling

Material and Methods

2.1.1.2 Sampling of material coupons

To sample the coupons water flow in the reactor system was stopped but only the

sampled reactor was drained. After one reactor with coupons has been taken out an

equivalent sterile reactor without coupons was inserted and the water flow was

started again. The openings of the sampled reactor were covered with sterile plastic

bags to avoid contamination and drying and the reactor was transported to the labo-

ratory. In the laboratory the coupons were released of the segments and inserted into

a humid chamber (50 mL centrifugal tubes with some millilitres of drinking water on

cellulose paper were autoclaved before use) to avoid drying of the deposits and

stored at approx. 8°C until further preparations at the sampling day. To detach the

deposits from coupons the drinking water exposed surface (17.25 cm²) was scraped

with a metallic scraper. Afterwards the scraper and the surface were rinsed with a

defined volume of sterile drinking water in a sterile petri dish to avoid loss of bacteria.

At the end the biofilm suspension was filled in 1.5 mL reaction tubes (Eppendorf,

Germany) for further applications.

2.1.2 Reactor systems Duisburg (Germany)

The two reactor systems installed by the Duisburger project partner (working group of

Prof. H.-C. Flemming) had basically the same setup and function as the Berlin reac-

tor system. One reactor was located at the outlet of the treatment plant Dorsten-

Holsterhausen the other within the distribution system in Gladbeck, approx.. 15 km

from the treatment plant Dorsten-Holsterhausen. The treatment plant supplied non-

disinfected water form a groundwater source. The material coupons (PE, PVC, Cu,

V4A-steel) with a surface area of 16.5 cm² were fit to the inner surface of the reactors

(PE, height: 760 mm, inner diameter: 100 mm) and the 5 reactors were arranged in

series as in the Berlin reactor system. Coupons and reactors were cleaned with

methanol and sterile destilled water before installation and the whole system was

disinfected with H2O2for 24 h (400 mg/L) before operation. In Dorsten-Holsterhausen

flow rate was regulated to 4 m³/h (velocity: 0.12 m/s). In Gladbeck a flow rate of 2

m³/h (velocity: 0.06 m/s) was regulated from 6 a.m. to 18 p.m. and during night the

Material and Methods

flow was reduced to 0.5 m³/h (velocity: 0.016 m/s). (Flemming 2003; Wingender and

Flemming 2004)

2.1.3 Reactor system Lundtofte (Denmark)

Operation and sampling of the pilot plants

In cooperation with Adam C. Martiny of the University of Denmark a third reactor sys-

tem was investigated. Biofilm and bulk water samples were taken from the distribu-

tion system located in Lundtofte, Denmark (Boe-Hansen et al. 2002) and in Berlin,

Germany. In the two reactor systems neither disinfectant agents nor ozone or UV

irradiation was used. Both systems were supplied with groundwater. Physico-

chemical and microbiological parameters of both systems are shown in table 1. All

analyses were performed according to standard methods. The material coupons

were placed parallel to flow direction to avoid turbulence in the water flow and were

exposed to drinking water for more than 6 months. In the German reactor system the

materials PVC, PE, and steel were included in the comparative investigations,

whereas in the Danish system only steel coupons were positioned.

Bulk water from the inlet (Lundtofte) and the outlet (Berlin) was sampled in 100 mL

aliquots and stored on ice until further analysis (max. 2 h). Material coupons of the

Danish system were scraped off with a wetted sterile cotton tip and transferred to a

test tube containing 10 mL autoclaved, 0.22 m filtered drinking water. The tube was

vigorously shaken for 1 min. German coupons were treated as described in chapter

2.1.1.2. The Danish biofilm and bulk water samples were plated on standard R2A

medium (Reasoner and Geldreich 1985), whereas samples from Germany were

plated on modified R2A medium (starch replaced by Tween 80, see chapter 2.3.1). All

isolated strains from Berlin were tested positive for the ability to grow on standard

R2A medium. Plates were incubated for 7 (Berlin) respectively 10 days (Lundtofte) at

room temperature (approx. 23C) and transferred twice on the cultivation medium.

One hundred colonies were isolated from each sample, 400 in total. To minimize

possible biases introduced by selection, colonies were picked systematically from

one side of the plate towards the other (Lundtofte), while colonies in Berlin were

picked according to colony morphology.

Material and Methods

Tab. 1: Physico-chemical and microbiological parameters of bulk water and biofilm in the

pilot plants of Lundtofte (Denmark) and Berlin (Germany).

Parameters

Lundthof

Berlin

Biofilm

Exposure time of coupons [days] 319 197

Plate count

, steel [cfu/cm²] 7.7 x 10

1.0 x 10

Plate count

, PE [cfu/cm²] - 1.3 x 10

Plate count

, PVC [cfu/cm²] - 1.4 x 10

Bulk water

pH 7.6 7.6

Iron [mg/l] 0.04 0.02

Manganese [mg/l] <0.005 <0.02

Hardness [dH] 17.8 10.9

Calcium [mg/l] 93 95

Ammonia [mg/l] 0.1 0.052

Nitrite [mg/l] 0.03 <0.03

Nitrate [mg/l] 2.5 3.6

Phosphate [mg/l] <0.02 0.092

Chloride [mg/l] 109 55

Sulfate [mg/l] 11 78

NVOC [mg/l] 2.3 3.6

AOC [µg/l] 6.1 n.d.

Temperature [°C] 12.3 16.3

Flow rate [m³/h] 0.5 1 - 10

Plate count [cfu/mL] 680 265

a)determined by standard methods of the drinking water suppliers b)modified R2A

2.2 Pipe sections from the drinking water distribution system

In Berlin the pipes were sampled in cooperation with the Berliner Wasserbetriebe.

After excavation of the selected pipe section the pipe`s outer surface was cleansed

with a brush and sterile Milli-Q water, treated with 10 % H2O2for 5-10 min and rinsed

again with sterile Milli-Q water. The water supply was turned off and a first cut was

done in the pipe to slowly drain the water from inside the pipe. The pipes were cut

with different tools according to the pipe material. PVC pipes were cut with a profes-

sional hacksaw and the metallic pipes were cut with a professional pipe cutter which

avoided any heating of the cutting site. Only the first Berlin test sample BWB I was

cut with a circular saw. Cutting tools were cleaned with H2O2and Milli-Q water before

application as described above. Water was drained form the excavation to prevent

contamination by the re-entering water. After draining of the pipe the second cut was

carried out to completely excise the pipe section. Water which drained from the main

was pumped away before it could re-enter the pipe and cause contamination. The

openings of the pipe section were covered with sterile plastic bags to avoid contami-

nation and drying of the interior. Subsequent the pipe was directly transported to the

Material and Methods

laboratory (20 to 40 min) for further treatment. At warm days the pipes outer surface

was cooled with cool packs.

2.2.1 Free water samples of the distribution system

Before each pipe sampling water samples were collected at a standpipe which was

placed downstream the sampling site. The standpipe was flushed for about 20 min

prior to sample collection. Temperature and pH were determined on site. Water sam-

ples for microbiological analysis were taken in sterile glass bottles after fire steriliza-

tion and short flush of the tap to cool it down.

2.2.2 Treatment of pipe deposits on the inner surface

As described in chapter 3.1.1 there was a great difference in the macroscopic shape

of the inner surface. The PVC and cement pipes with less voluminous coatings and

the metallic pipes with voluminous irregular coatings. According to these differences

appropriate standard treatment protocols were developed with the working group of

the Universität Duisburg to ensure equal treatment of the pipe samples in both labo-

ratories. The first 4 to 5 cm at each opening were discarded to minimize contamina-

tion of the samples.

2.2.2.1 Treatment of PVC pipes

One opening of the pipe was closed with two layers of sterile plastic bags and fixed

strongly with tape (not in contact with the inside). One hundred gram sterile glass

pearls (diameter 0.5 cm) and 150 mL of sterile Milli-Q water were filled in. After the

second opening was closed with sterile plastic bags as described above the PVC

pipe with the glass pearls was shaken and rotated by hand for 10 min. The water

phase was decanted, fresh 100 mL sterile Milli-Q water filled in and shaken and ro-

tated again for 10 min. The procedure was repeated until the inner surface was clean

by optical inspection. Finally glass pearls were washed two times with 50 mL Milli-Q

water. The biofilm suspensions were pooled, homogenized by shaking and subdi-

vided into sterile glass bottles for the participating laboratories.

Material and Methods

2.2.2.2 Treatment of metallic pipes

For the metallic pipes a scraping method was used. One opening of the pipe was put

in a sterile bowl and the deposits were scraped from the other opening with different

kinds of metal scrapers. This was done from both sides of the pipe and in the end the

pipe was rinsed with sterile drinking water. If necessary, fragments were crushed with

a metal stick. Finally the biofilm suspension was homogenized by shaking and subdi-

vided as described before.

2.3 Analysis of bacterial populations by cultivation techniques

2.3.1 Aerobic cultivation on modified R2A medium

Cultivation of the biofilm suspension was performed on modified R2A medium

(Tween 80 instead of soluble starch 0.1 % v/v) as described by Kalmbach (Kalmbach

et al. 1999; Reasoner and Geldreich 1985). After scraping and homogenization

(chapter 2.1, 2.2) the samples were serially diluted in sterile Berlin drinking water (au-

toclaved 20 min at 121°C and filtered, 0.2 µm pore size Supor-200 membrane

WAT200539, Waters Corporation, Ann Arbor, Michigan, USA) and R2A medium in-

oculated with 100 µL by the spread-plate method in three to four replicates at 20±2°C

in the dark. Colony forming units (CFU) were determined after 7 and 14 days. For

further investigations two to three colonies of each morpho-type were selected and

transferred to a fresh R2A plate (Tween 80 modified). This was repeated until a pure

culture was obtained. Special focus was laid on Aquabacterium-like morphologies as

described by Kalmbach (Kalmbach et al. 1999). Because they often grow as small

plain colonies plates were scanned by a binocular loupe (Zeiss, magnification 8 - 10

times).

2.3.2 Heterotrophic plate counts according to DIN EN ISO 6222

To get an insight in the alteration of the heterotrophic bacteria in the Berlin reactor

system, bulk water samples were cultivated as described in the German Drinking

Material and Methods

Water Regulations 1990 (GDWR 1990)1according to the standard method DIN EN

ISO 6222. One mL sample volume was incubated 44 ±4 h at 20°C ±2°C and 36°C

±1°C on a nutrient rich, peptone and meat extract containing medium (DEV) with the

pour plate method. The guideline value is 100 CFU/mL. Regular investigations were

done by the laboratory of the Berliner Wasserbetriebe (BWB).

2.3.3 E. coli and coliform bacteria according to DIN 38 411 K 6

As it is described in the standard method (DIN EN 38 411 K 6 1991), bacteria of the

species E. coli are inhabitants of the gut of human and endotherm animals. The de-

tection of E. coli in water is assessed as indication of a fecal contamination. Coliform

bacteria may have a fecal source but are also able to multiply in sewage and surface

water. Primary incubation of the water sample was done in 1 % (w/v) lactose bouillon

at 36 ±1°C for 24 ±4 h, if negative up to 44 ±4 h. If gas and acid production is noticed

subcultures are prepared on Endo-agar or McConkey agar for 24 ±4 h to do further

physiological test for discrimination of E. coli and coliform bacteria. Investigated sam-

ple volume was 100 mL. The limit value for E. coli and coliform bacteria is non in 100

mL. Regular investigations were done by the laboratory of the BWB.

2.3.4 Aerobic cultivation of P. aeruginosa according to DIN EN 12780

The standard method DIN EN 127802(DIN EN 12780 2002) detects P. aeruginosa

after membrane filtration (0.45 µm) and incubation for 44±4 h at 36±2°C as colony

forming units on a selective cetrimide-containing medium. Typically blue-green colo-

nies (pyocyanin-production) are accepted as P. aeruginosa without further characteri-

zation. Additionally, colonies not featuring the typical colour on the selective medium

have to be further characterized. If they show fluorescence under UV light on the se-

lective medium and are able to produce ammonia out of acetamide they are also ac-

1GDWR 1990 – German drinking water regulation 1990 corresponds in German to Trinkwasserverord-

nung – TrinkwV vom 5. Dez. 1990 (Bundesgesetzblatt I S. 2600) established law from january the first

1991.

2In the beginning of the investigations the method was only described as draft standard. Because the

approach of this study is identical with the standard method 2002 this version is cited here.

Material and Methods

cepted as P. aeruginosa. If they are red-brown on the selective medium they have to

be oxidase positive, show fluorescence on King`s B medium under UV irradiation,

and have to be able to produce ammonia out of acetamide to be identified as P.

aeruginosa.

2.4 Investigation of the bacterial population by culture inde-

pendent methods

2.4.1 Total cell counts (TCC) determined by DAPI staining

2.4.1.1 Staining of biofilm suspensions

Biofilm suspensions were obtained from coupons of the reactors or pipe samples

(chapter 2.1, 2.2). Usually, 1 mL of biofilm suspension was mixed with 100 µL of

DAPI stock solution (conc. 100 µg/mL, Sigma, D-9542) in 1.5 mL reaction tubes (Ep-

pendorf, Germany). If the counts were above or below the limit of 40-100 counts per

ocular grid sample volume and staining solution were proportionally reduced. The

suspension was incubated for 5 min in the dark. During incubation, the vacuum filtra-

tion element (Sartorius, Germany) was prepared. It was flushed with 70 % ethanol

and a cellulose support membrane (0.45 µm pore size, Sartorius, Germany) was

placed under the polycarbonate membrane (0.2 µm pore size, 25 mm diameter,

GTBP 02500, Millipore, Germany). 10 mL of sterile drinking water was filled in the

filtration element and the stained biofilm suspension was transferred. After filtration

the filter was air dried and mounted with the anti fading reagent Citifluor AF87

(www.citifluor.co.uk, London, UK) on a microscopic slide.

2.4.1.2 Staining on filter membrane

For this procedure the biofilm sample was filtered on the 0.2 µm polycarbonate

membrane as described above and 10 to 15 µL of DAPI stock solution (10 µg/mL)

were directly dispersed on the filter and incubated for 20 min in the dark. After air dry-

ing the filter was mounted as described above. If necessary, to be in the range of 40

to 100 counts per ocular grid, biofilm suspension was diluted with autoclaved and

sterile filtered drinking water (0.2 µm pore size).

Material and Methods

2.4.1.3 Staining on coupons

Biofilm associated cells on the material coupons were stained with 50 to 100 µL of

DAPI solution (10 µg/mL) for 20 min in the dark. Afterwards the coupons were rinsed

with fresh Milli-Q water, air dried and mounted with anti fading reagent as described

above.

2.4.1.4 Microscopic examination

Immediately after staining fluorescent signals were counted with an Axioplan 2 (Carl

Zeiss, Germany) equipped with a HBO 100 lamp and the Zeiss filter no. 1 for DAPI

(excitation 365, dichroic mirror 395 nm, suppression 397 nm). A minimum of 10 ran-

domly chosen microscopic fields and 1000 cells were analysed. Results were docu-

mented with the camera Color View 12 (Soft Imaging System GmbH, Berlin, Ger-

many) or a Kodak EES 1600 color reversal film.

2.4.2 Fluorescence in situ hybridization (FISH)

2.4.2.1 Fixation of biofilm coupons, suspensions, and pure cultures

Biofilm on coupons

Fixation and washing of the coupons was done in 50 mL centrifugal tubes which were

filled with sterile 3.7 % formaldehyde and 1 x PBS (Sambrook and Russel 2001) so-

lution respectively. After the coupons have been taken out of the reactor and trans-

ported in the laboratory they were first fixed with formaldehyde at 4°C for two hours

and then washed twice in 1 x PBS. At the end the coupons were air dried and stored

at room temperature in the dark. (Kalmbach 1998)

Biofilm suspension

Formaldehyde fixation:

Seven hundred µL of the biofilm suspension were filled in a 1.5 mL reaction tube

(Eppendorf, Germany) and centrifuged at 13,000 rpm for 5 min. The supernatant was

discarded. After addition of 700 µL 3.7 % sterile formaldehyde solution (0.2 µm pore

Material and Methods

size filtered, GTBP, Millipore, Germany) the pellet was resuspended by Vortex mixing

or with a manual micro-homogenisator. The resuspended pellet was incubated for 2 h

at 4°C in the dark. Afterwards the pellet was washed twice with sterile 1 x PBS (Sam-

brook and Russel 2001) and centrifuged as described above. Finally the pellet was

resuspended in 50 to 500 µL PBS (1x)/ethanol (96 %) mixture (1:2) and stored at -

20°C.

Ethanol fixation:

After centrifugation as described above the biofilm suspension was resuspended in

50 to 500 µL PBS (1x)/ethanol (96 %) mixture (1:2), incubated for 2 h at 4°C in the

dark and stored at -20°C.

Bacterial pure cultures

Formaldehyde fixation:

Bacterial pure cultures were incubated in the appropriate medium to the exponential

phase. Afterwards 700 to 1500 µL were centrifuged for 5 min at 13,000 rpm to get a 3

to 4 mm diameter pellet. The supernatant was discarded, 700 µL 3.7 % sterile for-

maldehyde added, the pellet resuspended, and incubated for 2 h at 4°C. After cen-

trifugation (13,000 rpm, 5 min) the pellet was washed two times with 1 x PBS and

finally resuspended in PBS (1x)/ethanol (96 %) mixture (1:2) solution and stored at -

20°C.

Ethanol fixation:

Bacterial cultures were incubated and pelleted as described for the formaldehyde

fixation and resuspended in an appropriate volume (depending on the suspension

density, 50 to 500 µL) of PBS (1x)/ethanol (96 %) mixture (1:2), incubated at 4°C for

2 h and stored at -20°C.

2.4.2.2 Hybridization procedure

Oligonucleotide probes were diluted in sterile Milli-Q water to a working solution of 50

ng/µL and stored at -20°C. To obtain a concentration of 5 ng/µL in the hybridization

solution the working solution was diluted 1:10 with hybridization buffer containing 0.9

M NaCl, 20 mM Tris/HCl (pH 7.2), 0.03 % SDS, and an appropriate volume of for-

mamide (molecular biology grade, Merck, Germany). The hybridization was per-

Material and Methods

formed for 1.5 to 4h in a humid chamber at 46°C. To remove unbound oligonucleo-

tides the samples were washed 18 minutes in the pre-warmed washing buffer con-

sisting of 20 mM Tris/HCl, 0.01 % SDS, and an appropriate volume of NaCl corre-

sponding to the used formamide stringency. Finally, the hybridized object was rinsed

with Milli-Q water, air dried and mounted with antifading reagent Citifluor AF2

(Citifluor Ltd., London, UK). (Manz et al. 1998; Manz et al. 1993)

Counterstaining was done by overlaying the hybridized area with a 1 µg/mL (bacterial

cultures) or 10 µg/mL (biofilm suspensions and coupons) DAPI (Sigma, D-9542) solu-

tion for 15 to 20 min. After gently rinsing off the staining with Milli-Q water and air dry-

ing the fields were mounted with antifading reagent.

Fluorescent signals were counted with an Axioplan 2 (Carl Zeiss, Germany) equipped

with an HBO 100 lamp and the Zeiss filter no. 1 for DAPI (excitation 365, dichroic

mirror 395 nm, suppression 397 nm), no. 9 oregon green (excitation 450-490, di-

chromatic mirror 510 nm, suppression 520 nm) and HQ light filter 41007 (excitation

535-550 nm, dichroic mirror 565 nm, suppression 610-675 nm, AF Analysentechnik,

Tübingen, Germany) for detection of Cy3 labelled probes. A minimum of ten ran-

domly chosen microscopic fields of the ocular grid and 1000 cells were analysed.

Results were documented with the camera Color View 12 (Soft Imaging System

GmbH, Berlin, Germany) or a Kodak EES 1600 color reversal film.

Bacterial pure cultures

About 10 µL of fixed bacterial pure culture was placed on the cavities of a teflon-

coated microscopic slide (Marienfeld, Bad Mergentheim, Germany) and dried at 46°C

in the hybridization oven. The slides were dehydrated with ethanol (50, 80 and 96 %,

3 min each), 10 µL hybridization solution added to each cavity and incubated,

washed and mounted with antifading agent as described above.

Material and Methods

2.4.2.3 Development of a new oligonucleotide probe

A new probe for P. aeruginosa was developed with the ARB software package and

the probe design and match tool (Strunk et al. 1999). Specificity was checked by

comparative sequence database analysis and hybridization with target and non-

target bacteria. For the probes optimization of hybridization stringency was performed

as described by Manz et al. (Manz et al. 1998). Hybridization stringency was ad-

justed by addition of varying formamide concentrations to the hybridization buffer and

sodium chloride to the washing buffer. Hybridization was prepared as described

above. (chapter 2.4.2.2)

2.4.2.4 Oligonucleotide probes used in this study

Table 2 shows the oligonucleotids used in this study with their specificity and strin-

gency. Oligonucleotids were labeled with the indocarbocyanine dye Cy3 or oregon

green by the company Metabion (Martinsried, Germany).

Tab. 2: Oligonucleotide probes used in this study.

Common

name

Probe Squence 5` - 3` Target-organisms FAa)

Reference

EUB 338 GCT GCC TCC CGTAGGAGT domain Bacteria 20 (Amann et al. 1990)

non-EUB 338 ACT CCTACG GGA GGC AGC negative control 20 “

PsearB TCT CGG CCT TGA AAC CCC P. aeruginosa 40 (Hogardt et al. 2000)

PsearE CCC ACC CGA GGT GCT GG P. aeruginosa 50 present study

Ps GCT GGC CTA GCC TTC most true pseudomonads 35-50 (Schleifer et al. 1992)

a)FA: formamide

Material and Methods

2.4.3 Extraction of total DNA from biofilm suspensions

2.4.3.1 Simple preparations of DNA from formaldehyde fixed and non-fixed

biofilm suspensions

To test the influence of formaldehyde fixation on extraction and amplification proce-

dure, both fixed sample material and untreated samples were tested:

a) Formaldehyde fixed sample of M IV (chapter 2.4.2.1). The fixed biofilm suspen-

sion was washed with 1 x PBS and resuspended in 1 x PBS.

b) Untreated biofilm suspension of BWB III which has been stored in a closed glass

bottle in the fridge for some weeks. Because the influence of the chemical com-

pounds was the aim of this approach, changes caused by fridge storage were as-

sessed to be low.

As positive control the DNA of the isolate DK 79 (drinking water isolate of the reactor

system in Denmark) or E. coli (DSM 5695) was added to the PCR mixture. DNA con-

centration of both controls was optimized to give a well-defined band on a 1.7 % aga-

rose gel after universal PCR (chapter 2.4.3.6, 0.5 µL of AE diluted alkaline lysis,

both). Addition of control DNA to biofilm suspensions after DNA preparation allowed

assessment of inhibition and/or adsorption. Amplification of control DNA without

biofilm suspension was the confirmation that the approach configuration was suitable

in general. The following different approaches were made:

2.4.3.1.1 M IV (formaldehyde fixed) and BWB III (non-fixed) serial diluted

The formaldehyde fixed and the non-fixed biofilm suspensions M IV and BWB III

were applied serial diluted (to 10-6) to the PCR test series a) and b).

a) addition of DK 79 DNA

b) without addition of DK 79 DNA

2.4.3.1.2 Alkaline lysis of formaldehyde fixed M IV

Two hundred µL of the biofilm suspension were centrifuged at 13,000 rpm for 5

min (Biofuge 13, Heraeus Instruments, Newtown, CT, USA) and the supernatant

was discarded. The general lysis procedure is described in chapter 2.4.4.1. Be-

cause of the higher pellet volume the volumina of the solutions were adjusted.

Material and Methods

The pellet (75 µL) was resuspended in 100 µL lysis solution and heated at 95°C

(Thermomixer, 5436, Eppendorf, Germany) for 30 min. Every 5 min the suspen-

sion was shaken by hand and vented if necessary. Finally the lysed suspension

was diluted with 900 µL AE buffer (chapter 2.4.4). As described above two PCR

test series were accomplished.

a) addition of DK 79 DNA

b) without addition of DK 79 DNA

2.4.3.1.3 Alkaline lysis, enhanced sample volume, and ethanol precipitation

Alkaline lysis was prepared with 2 mL biofilm suspension of formaldehyde fixed M IV

and addition of 250 µL lysis solution as described above. After lysis the suspension

was transferred to a 50 mL centrifugal tube and 2.5 mL ammonium acetate (end-

conc. 2 M, NH4Ac, Merck, Germany) and 225 µL dextran blue (conc. 1 mg/mL, Fluka,

Germany) were added, and cooled for 10 min on ice. After centrifugation (15 min,

12,000 rpm, 15°C, Sorvall RC-5B refrigerated Superspeed Centrifuge DuPont, Wil-

mington, DE, USA) the double volume of ice cold ethanol (96 %, p. a. Fluka, Ger-

many) was added. The suspension was vortexed and centrifugated again (15 min,

12,000 rpm, 15°C). The supernatant was discarded and the pellet washed with 70 %

ethanol and transferred to a 1.5 mL centrifugal tube before centrifugation. After drying

at 70°C for 3 min the pellet was resuspended in 30 µL AE buffer (Qiagen, Germany).

Afterwards universal PCR was prepared as described in chapter 2.4.3.6) with differ-

ent dilutions of the DNA suspension and:

a) addition of DK 79 DNA

b) without addition of DK 79 DNA

2.4.3.1.4 Alkaline lysis, enhanced sample volume, and isopropanol precipita-

tion

Five caps of formaldehyde fixed biofilm suspension of M IV were washed with 1 x

PBS (in 50 mL centrifugal tube) which resulted in a pellet of 5 mL biofilm suspension.

This pellet was resuspended in 5 mL alkaline lysis solution. The lysis was prepared at

95°C in a waterbath for 30 min. Every 5 min the tube was shaken by hand. After lysis,

500 µL 3 M sodium acetate solution and 250 µL dextran blue (conc. 1 mg/mL, Fluka,

Material and Methods

Germany) were added and carefully mixed. After centrifugation (6 min, 12,000 rpm,

15°C, Sorvall RC-5B Refrigerated Superspeed Centrifuge DuPont, Wilmington, DE,

USA) the supernatant was transferred to a clean centrifugal tube, 5 mL isopropanol

(99.9 %, Fluka, Germany) added, and mixed by hand. Once again the solution was

centrifugated (6 min, 12,000 rpm, 15°C) and the supernatant discarded. The pellet

was washed in 0.8 mL 70 % ethanol (10 min, 15,000 rpm, 15°C) and transferred to a

1.5 mL centrifugal tube before centrifugation. Finally the pellet was dryed at 70°C to

remove remaining ethanol and resuspended in 30 µL AE buffer (Qiagen, Germany).

PCR was prepared as described in chapter 2.4.3.6 and again the two test series ac-

complished.

a) addition of DK 79 DNA

b) without addition of DK 79 DNA

The isopropanol precipitation was repeated with the following modifications: E. coli

(DSM 5695) was used parallel to the biofilm suspension. Furthermore isopropanol

precipitation was extended to 15 min.

2.4.3.1.5 Addition of bovine serum albumin (BSA)

PCR preparations were done as described in chapter 2.4.3.6. Bovine serum albumin

(BSA, acetylated, Promega, Germany), biofilm suspension (BWB III) or control DNA

(E. coli DSM 5695) was added to PCR preparations as follows:

a) untreated sample of BWB III serial diluted to 10-6 +E. coli DSM 5695 + 0.1

µg/µL BSA (conc. in 25 µL PCR volume)

b) untreated sample of BWB III serial diluted to 10-6 + 0.1 µg/µL BSA (conc. in 25

µL PCR volume)

c) E. coli DSM 5695 + 0.1 µL/µL BSA (conc. in 25 µL PCR volume)

d) E. coli DSM 5695

After this first approach the concentration dependent effect of BSA was tested:

a) serial dilution of untreated BWB III sample up to 10-6 + 0.1 µg/µL BSA

b) serial dilution of untreated BWB III sample up to 10-6 + 0.3 µg/µL BSA

c) serial dilution of untreated BWB III sample up to 10-6 + 0.6 µg/µL BSA

Material and Methods

2.4.3.2 FastDNA Spin Sample Kit for soil

After sampling total nucleic acids were directly extracted from the biofilm suspensions

or frozen at –20°C until extraction procedure. DNA was prepared with the method

provided by Bio 101 (1070 Joshua, Vista, CA 92083, USA) as FastDNA Spin Sample

Kit for soil according to the manufacturer`s instructions. The extraction method in-

cludes a mechanical-chemical lysis with ceramic and silica particles, a protein pre-

cipitation and several washing and purification steps. Two hundred fifty µL of the

biofilm suspension of different samples were added to the Multimix tube and the DNA

was finally eluted with 50 µL DES (kit supplied pure water).

2.4.3.2.1 Extraction of DNA from biofilm suspensions

The extraction of total DNA according to the standard protocol did not succeed for

various samples (pipes and material coupons of the reactor, data not shown). To

concentrate DNA the DNA extracts of different biofilm suspensions of pipe samples

were pooled as described in the following table:

Tab. 3: Pooled biofilm suspensions of pipe samples before DNAextraction.

Biofilm suspension Tubes pooled Resulting µL

BWB V 4 tubes 150

BWB VII 2 tubes 80

BWB VII 3 tubes 120

M X 3 tubes 130

Material and Methods

To precipitate the extracted DNA with isopropanol chemicals were added as follows:

Tab. 4: Additon of chemicals before isopropanol precipitation.

Biofilm Resulting Dextran blue Sodium acetate PCR H2O Isopropanol

susp

nsion

(conc. 1 mg/mL)

µL

BWB V 150 10 20 20 200

BWB VII 80 5 10 5 100

BWB VII 120 7,5 15 7,5 150

M X 130 10 20 40 200

After the addition of isopropanol (99.9 %) the suspension was mixed by hand and

precipitated 15 min at room temperature. The centrifugation was done for 15 min at

15.000 rpm and 15°C and the supernatant discarded. The blue pellet (dextran blue)

was dried in the Thermomixer at 70°C for 3 min and finally diluted in 30 µL AE buffer

(Qiagen, Germany). The solution was stored at -20°C until PCR preparation.

To test whether the concentration of DNA in the biofilm DNA preparations was to high

to amplify it in the PCR, dilutions of the DNA preparations of BWB VI, M VII and RB I

PE were made (1:10, 1:100, 1:1000) and different volumina appointed in the PCR.

2.4.3.2.2 Evaluation of extraction efficiency

Defined amounts of DNA of Aquabacterium citratiphilum DSM 11900 were appointed

to the FastSpin extraction to assess the extraction efficiency. At first A. citratiphilum

was cultivated in liquid modified R2A medium at room temperature and harvested in

the exponential phase. The bacterial cells were lysed under alkaline conditions as

described before and diluted in AE buffer (1:10). The concentration of the DNA was

determined with the PicoGreen dsDNA Quantitation kit (chapter 2.4.3.6.2). Subse-

quent the defined DNA amounts were applied to the FastDNA Spin Kit method and

concentration of DNA was again measured at the end of the procedure to determine

the loss of DNA. To further assess the loss of DNA by the mechanical-chemical lysis

the extraction was repeated without the kit recommended lysis.

Material and Methods

2.4.3.3 CTAB (hexadecyltrimethylammonium bromide) extraction

The CTAB extration is described more detailed than the other DNA preparations be-

cause the classical method allows for more biases than an optimized kit method. The

general extraction method was described by Ogram in 1998 and was modified as

described in the following chapters (Ogram 1998).

2.4.3.3.1 General extraction procedure

Solutions and technical equipment

0.12 M sodium phosphate buffer: 0.11 M Na2HPO4(Sigma, Germany), 0.01 M

NaH2PO4(Sigma, pH 8.0)

alkaline lysis solution: 0.25 % SDS (sodium dodecylsufate, Sigma), 50 mM

NaOH (Merck, Germany), sterilised by a 0.2 µm Minisart-filter (Satorius, Ger-

many), Qiagen AE-buffer (Qiagen, Germany)

5 M NaCl (Merck, Germany)

10 % (w/v) CTAB (cetyltrimethylammonium bromide, Sigma) in 0.7 M NaCl

(Merck)

Chloroform-isoamylalkohol (24:1): chloroform (Merck), isoamylalcohol (Merck)

13 % (w/v) PEG-(8000) (polyethylene glycol, Sigma) in 0.7 M NaCl

dextran blue (1 mg/mL, Fluka)

99.8 % ethanol (Roth, Germany)

70 % ethanol (99.8 % ethanol diluted with sterile Milli-Q water)

1 x TE (Tris-EDTA): 10 mM Tris-Cl (tris(hydroxymethly)- aminomethane hydro-

chloride (Riedel-de-Haen, Germany), 1 mM EDTA (Sigma), pH 8

6 M NH4Ac (ammonium acetate, Merck): sterilised by a 0.2 µm Minisart-filter

(Satorius)

chloroform-isoamyl alcohol (24:1) (both Merck)

Sterile glas pipettes and pipettboy (Pipetus-Standard, Hirschmann, Eberstadt,

Germany)

Biofuge 13 (Heraeus Instruments, Newtown, CT, USA) for 1.5 mL reaction

tubes

Material and Methods

Sorvall RC-5B Refrigerated Superspeed Centrifuge (DuPont, Wilmington, DE,

USA) for 50 mL centrifugal tubes

Thermomixer Eppendorf 5436

Preparation of biofilm suspenions

1. The sample (1 to 5 g) was prepared into sterile 50 mL centrifugal tubes

(Kisker, Steinfurt, Germany), 10 mL sodium phosphate buffer added, vortexed

for 1 min and allowed to settle for 10 min with occasional mixing.

2. Suspension was centrifugated at 6,500 rpm for 10 min and the supernatant

discarded. Step 1 and 2 were repeated.

3. 16 mL alkaline lysis solution were added (0.25 % SDS, 50 mM NaOH), mixed

by hand, incubated in a waterbath for 20 min at 95°C and shaken with hand

every 5 min.

4. Suspension again centrifugated at 6,500 rpm for 10 min and supernatant

transfered in a clean centrifugal tube. The transferred volume was listed and

the pellet discarded.

5. 2.7 mL 5 M NaCl (final conc. 0.7 M NaCl) and 2.1 mL of 10 % CTAB (final

conc. 1 % CTAB) were added to the supernatant, mixed by hand and incu-

bated at 65°C in the waterbath.

6. An equal volume of chloroform-isoamyl alcohol (24:1) was added and mixed

briefly to achieve emulsion.

7. Suspension was centrifugated at 4,200 rpm for 5 min and the upper aqueous

phase transfered to a clean centrifugal tube, the lower phase was discarded.

8. An equal volume of 13 % polyethylene glycol (PEG 8000) in 0.7 M NaCl was

added, mixed by hand and keeped on ice for 10 min.

9. Suspension was again centrifugated at 8,500 rpm for 15 min and the super-

natant discarded. 500 µL of 70 % ethanol and 30 µL dextran blue solution

(1 mg/mL, Fluka) were added, mix and centrifugated for a short time (max.

2,000 rpm, to collect liquid at the bottom). The liquid was transfered to a 1.5

mL reaction tube and centrifugated for 15 min at 12,000 rpm. The supernatant

was discarted and the pellet dried at 70°C in the Thermomixer for 3 min. Fi-

nally, nucleic acids were resuspended in 750 µL TE.

Material and Methods

10.390 µL 6 M NH4Ac (final conc. 2 M) were added and the suspension keep on

ice for 10 min.

11.Again a centrifugation step was done at 12,000 rpm for 15 min, the super-

natant transfered in a clean reaction tube, the transferred volume listed, and

the supernatant discarded.

12.Afterwards 2 volumes (listed in step 11) of 99.8 % ethanol were added and

centrifugated for 15 min at 8,500 rpm and the supernatant discarded. The pel-

let was washed with 500 µL 70 % ethanol and shortly centrifugated (max.

2,000 rpm) to collect liquid at the bottom and the liquid was transfered into a

clean reaction tube. The suspension was centrifugated at 12,000 rpm for 15

min, the supernatant discarded and, the pellet dryed at 70°C for about 3 min.

Finally the DNApellet was resuspended in 100 µL QiagenAE buffer and store

at -20°C until further applications.

2.4.3.3.2 Variations of the protocol

Some of the above described protocol steps were varied as follows:

Step 5: If the protocol was prepared without the lysis steps, the missing liquid volume

was replaced by 10 mL or 16 mL double autoclaved Milli-Q water, when 10 mL water

was added the volumes of NaCl and CTAB were adapted.

Step 10: In the case of leaving out the PEG precipitation the added volume of 6 M

ammonium acetate was adapted to achieve a final conc. of 1.5 to 2 M NH4Ac.

Step 12: In the validation process of the method after addition of ethanol in step 12

the solution was frozen over night at -20°C and the next day the procedure was con-

tinued with centrifugation.

2.4.3.4 DNA extraction with QIAamp DNA Mini Kit

In this kit the DNA is lysed, bound to a silica-gel membrane, two times washed with

different buffers and eluted with buffer AE. The extraction was in principle performed

as described in the manual (02/2003, QIAamp DNA Mini Kit, Qiagen, Germany) ex-

cept the modifications described in the following.

Material and Methods

Two different lysis procedures were tested:

a) Kit recommended lysozyme lysis in protocol D (lysozyme 20 mg/mL, from chicken

egg white, Sigma, Germany; 20 mM Tris HCl (Riedel de Haen, Germany) pH 8.0;

2 mM EDTA(Sigma); 1.2 % Triton (100 %,Fluka), Qiagen proteinase K

b) Alkaline lysis as described in chapter 2.4.4.1.

The following samples were tested:

a) E. coli DSM 5695 liquid culture in nutrient agar (DSMZ media, incubated for 7

days) as control of the extraction method was extracted in parallel. The culture

was centrifuged at 12,000 rpm for 15 min (Sorvall, RC-5B) and the supernatant

discarded.

b) Formaldehyde fixed biofilm suspension of BWB III. For the larger sample volume

the volume of the added solutions was adapted. Six caps (approx. 4.2 mL) of for-

maldeyd fixed samples were pooled, alkaline lysed (3 mL alkaline lysis solution,

30 min at 95°C, shaken by hand every 5 min) and diluted in 27 mLAE buffer. The

mixture was centrifuged again (12,000 rpm, 6 min) to separate the sediment from

the DNA in the liquid phase, transferred to a clean tube and an adapted volume of

15 mL 96 % ethanol was added (step 4 Qiagen protocol) before the solution was

applied to the spin column in 600 µL portions.

2.4.3.5 Extraction and purification by Qiagen Genomic-tips 20

The extraction was accomplished as described in the manufacturer`s manual

(08/2001, Qiagen, Germany) except some variations that will be described below.

The manufacturer describes that the procedure is based on an optimized buffer sys-

tems for careful lysis of cells followed by binding of genomic DNA to anion-exchange

resin under appropriate low salt and pH conditions. Impurities are removed by a me-

dium-salt wash and the genomic DNA is eluted in a high-salt buffer and concentrated

and desalted by isopropanol precipitation.

The extraction procedure was divided in two parts:

a) sample preparation and lysis protocol for bacteria

Material and Methods

1. The lysozyme solution (from chicken egg white, Sigma; stock solution: 100

mg/mL) was prepared as described in the manual and protease K stock solu-

tion used from Qiagen.

2. The bacterial culture was pelleted at 12,000 rpm (Sorvall, RC-5B) instead of

5,000 g because in the environmental sample very small bacteria were ex-

pected which might not pellet at low g forces.

3. As a control a 24 to 48 h E. coli DSM 5695 liquid culture in nutrient agar

(DSMZ media) was used. 800 µL of the culture were centrifuged.

4. As an alternative lysis procedure the alkaline lysis as described in chapter

2.4.4.1 was used. To the pellet of 1 mL biofilm suspension 1 mL lysis solution

was added, but the pellet was not washed with 1 x PBS. 30 µL lysis solution

were added and after lysis diluted with 270 µLAE (Qiagen, Germany). To

avoid a negative influence of SDS included in the lysis solution, in some ap-

proaches the anionic detergent was precipitated by 1 M potassium acetate for

15 min in the fridge. This was prepared directly after lysis without dilution with

AE buffer.

b) Genomic-tips 20 protocol for isolation of genomic DNA from bacteria

This part of the protocol was to a large extent performed as described in the man-

ual.

1. Genomic tips 20/G were used that were recommended for up to 4.5 x 109cells

in the sample extracted.

2. Eluted DNA was collected in sterile 15 mL polypropylen centrifugal tubes

(Kisker, Germany).

3. The isopropanol precipitated DNAwas centrifuged at 4°C and 5,000 rpm for

15 min to concentrate the DNA at the bottom and the supernatant removed be-

fore the DNA pellet was washed with 70 % ethanol. Isopropanol and 70 %

ethanol were filtered (0.2 µm pore size single use syringe filter, Satorius, Ger-

many).

4. The final drying of the DNA pellet, after washing with 70 % ethanol, was done

at 46°C in a hybridization oven for up to 30 min.

5. The DNApellet was finally resuspended in 100 or 300 µLAE buffer.

Material and Methods

2.4.3.6 Verification of the extraction success

The extraction success was evaluated by two methods, polymerase chain reaction

with universal bacterial primers and/or measurement of double-stranded DNA with

the sensitive fluorescent nucleic acid stain PicoGreen.

2.4.3.6.1 Control PCR

The above extracted bacterial DNA was amplified by PCR in 25 µL reaction mixtures

as described in table 5. To test extracted DNA a volume PCR-H2O was replaced by a

volume of the DNA extract, analogous controls were prepared. The Taq polymerase

kit contained the appropriate buffer and MgCl2(5 U/µL Taq no. 1647687, Roche Di-

agnostics GmbH, Germany). The universal 16S rDNA primers 616 forward (5`-AGA

GTT TGA TYM TGG CTC AG-3`) and 1525 reverse (5`-AAG GAG GTG WTC CAR

CC-3`) were used (Lane 1991). Reaction mixtures were incubated in a gradient

thermal cycler (Whatman, T-Gradient Thermoblock, Biometra, Germany) with one of

the three following cycling conditions.

1. Initial denaturation one cycle at 96°C for 2 min followed by 35 cycles of 94°C for

30 sec, 60°C for 2 min, 72°C for 3 min, and a final extension cycle at 72°C for 15

min. program 1

2. Initial denaturation one cycle at 96°C for 2 min followed by 35 cycles of 94°C for

30 sec, 57°C for 2 min, 72°C for 3 min, and a final extension cycle at 72°C for 15

min. program 2

3. Initial denaturation one cycle at 96°C for 2 min followed by 40 cycles of 94°C for

30 sec, 52°C for 2 min, 72°C for 3 min, and a final extension cycle at 72°C for 15

min. program 3

Material and Methods

Tab. 5: Composition of PCR reaction mixture per 25 µL total volume. To test extracted

DNA the appropriate PCR-H2O was replaced by the sample volume.

Solution Conc. stock solution Volume in 25 µL total

PCR reaction mixture

Buffer 10 x 2.5

MgCl225 mM 1.5

dNTP 10 mM (2.5 mM each base) 0.5

Primer 616 F 10 pmol/µL 0.25

Primer 1525 R 10 pmol/µL 0.25

Taq 1 U/µL 1.9

PCR-H2O 18,1

total volume 25

PCR products were visualized on a 1.7 % or 1 % agarose gel (SeaKem LE agarose,

FMC BioProductss, Rockland, Maine, USA) casted in a biozym chamber (Biozym

midi chamber, Biozym Scientific GmbH, Hess. Oldendorf, Germany). The investi-

gated DNA solution was mixed with the loading buffer on a parafilm M strip. As load-

ing buffer a Ficoll based buffer (0.25 % bromophenol blue, 0.25 % xylene cyanol FF,

15 % Ficoll Type 400, (Sambrook and Russel 2001) or the buffer offered with the lad-

der (GeneRuler 100bp DNA Ladder Plus, MBI Fermentas, Leon-Rot, Germany) was

used. The running conditions are shown in table 6.

Tab. 6: Running conditions of agarose gels.

Agarose%

w/v

TAEa)

Agarose

Voltage

Running time

h (approx.)

1.7 75 1.28 100 1.2

1.0 75 0.75 80 2

a)1 x TAE buffer: 40 mM Tis-acetate, 1 mM EDTA, (Sambrook and Russel 2001)

Table 7 illustrates which PCR amplification conditions were applied to which DNA

extraction approach. Due to the limited volume of the biofilm suspension and as a

Material and Methods

consequence the limited volume of DNA extracts, it was not possible to test every

extract with every PCR conditions.

Tab. 7: PCR conditions applied to the DNA extracts of the different preparation attempts.

In every PCR the primers 616 F and 1525 R were used. fix.: formaldehyde fixed biofilm

suspension

DNA preparation PCR-program Applied sample

inhibition and adsorption

chapter 3.3.1.1 prog. 2

57°C annealing, 35 cycles M IV (fix.): undiluted 1+5 µL, dilution

1:10 to 1:1061 µL each

BWB III: 1 µL of undiluted and each

dilution

with and without control DK 79

alkaline lysis

chapter 3.3.1.2 prog. 2

57°C annealing, 35 cycles M IV (fix.): undiluted 1+5 µL, dilution

1:10 to 1:1061 µL each

ethanol precipitation

chapter 3.3.1.3 prog. 2

57°C annealing, 35 cycles M IV (fix.): undiluted. 1+5 µL, dilution

1:10 to 1:1061 µL each

isopropanol precipitation

chapter 3.3.1.3 prog. 2

57°C annealing, 35 cycles M IV: undiluted. 1+5 µL, dilution 1:10 to

1:1061 µL each

BSA

chapter 3.3.1.4 prog. 2

57°C annealing, 35 cycles BWB III: undiluted and 1: 10 to 1:106,

1 µL each

with BSAin different concentrations

with and without spiked E. coli

Fast Prep

chapter 3.3.2.1 prog. 2

57°C annealing, 35 cycles BWB V, BWB VII, M X: 0.5, 1, 5,17.5 µL

undiluted. extract

Fast Prep diluted

chapter 3.3.2.1.2 prog. 3

52°C annealing, 40 cycles M VIII, RBI PE, BWB VI:

dilution 1:10 to 1:1000, MVIII=10µL

each, RBI PE=10 µL each

BWB VI=1, 10, 17.5 µL each

CTAB

chapter 3.3.2.2 prog. 3

52°C annealing, 40 cycles

prog. 2

57°C annealing, 35 cycles

A. citratiphilum spiked: 1, 8, 17.5µL

A. citratiphilum spiked: 4 parallels, 1,

10 µL

QIAamp DNA Mini Kit

chapter 3.3.2.3 prog. 1

60°C annealing, 35 cycles BWB III fix., 1. eluat, 1, 10, 17.5 µL

Genomic tips 20

chapter 3.3.2.4 prog. 2

57°C annealing, 35 cycles BWB VI, E.coli (fix., not fix.), lysozym

lysis, alk. lysis, alk. lyis combined with

precipitation, all undiluted, 1 + 10 µL

2.4.3.6.2 Measurement of DNA

To determine small amounts of double-stranded DNA the sensitive fluorescent nu-

cleic acid stain PicoGreen was used (PicoGreen dsDNA Quantitation kit P-7589, Mo-

lecular Probes Europe BV, Leiden, Netherlands). With this fluorescent dye it is possi-

ble to determine as little as 25 pg/mL of dsDNA. The preparation of standards and

samples is described in detail in the product information sheet. It is possible to pre-

pare two standard curves depending on the expected DNA concentration of the in-

Material and Methods

vestigated sample. One curve ranges from 1 ng/mL to 1 µg/mL the other from 25

pg/mL to 25 ng/mL. Measurement was done with a Hitachi Fluorescence Spectrome-

ter F-4500 equipped with a Xenon lamp.

2.4.4 Sequencing of bacterial 16S rDNA

2.4.4.1 Alkaline lysis

Bacterial isolates incubated on agar plates or in liquid medium were transferred to a

1.5 mL centrifugal tube (pelleted and washed with 1 x PBS). Bacterial cells were re-

suspended in 20 to 30 µL alkaline lysis solution (0.25 % SDS, Sigma; 50 mM NaOH,

Merck, filtered through a 0.2 µm Minisart-Filter, Satorius) and heated at 95°C (Ther-

momixer 5436, Eppendorf) for 15 min under permanent agitation. Afterwards the DNA

was diluted 1:10 in AE buffer (Qiagen, Germany) and usually the concentration and

the DNA/Protein ratio determined at 260 and 280 nm (1 OD260= 50 µg DNA/mL). The

DNA concentration was usually adjusted to 10 to 100 ng in the PCR reaction.

2.4.4.2 Sequencing reaction

For phylogenetic identification selected isolates obtained from different drinking water

habitats were sequenced as follows. One hundred µL total PCR reaction mixture con-

tained 70 µL diluted DNA extract (see above) and 30 µL master mix. The 30 µL mas-

ter mix contained 1x reaction buffer, 1.5 mM MgCl2, 200 mM dNTP (Promega, Ger-

many), 1µM each primer, and 2.5 U Taq (no. 1647687, Roche Diagnostics GmbH,

Germany). Samples were initial denaturated at 96°C for 1 min 30 sec followed by 28

cycles of 96°C for 30 sec, 57°C for 2 min 30 sec, 72°C for 3 min 30 sec, and a final

extension cycle at 72°C for 10 min in a Biometra cycler (Biometra, Germany). The

16S rRNA genes were amplified with the universal primer pairs 616F (5`-AGA GTT

TGA TYM TGG CTC AG 3`) and primer 1492R (5`-CGG YTA CCT TGT TAC GAC-3`)

or 63F (5`-CAG GCC TAA CAC ATG CAA GTC-3`) and 1387R (5`-GGG CGG WGT

GTA CAA GGC-3`) (Lane 1991; Marchesi et al. 1998). Sequence analysis was per-

formed with an ABI Prism 310 sequencer (Perkin-Elmer Applied Biosystems Deutsch-

land GmbH, Weiterstadt, Germany) using an Applied Biosystems Big Dye Terminator

Material and Methods

Ready Reaction Mix Kit according to manufacturer’s instruction. Cycle sequencing

was done in a Gene Amp PCR System 9700 (Perkin-Elmer Applied Biosystems,

Germany) and additionally to the above mentioned primers the following were used

to amplify the hole double stranded 16S rRNA gene: 699R (5`-RGG GTT GCG CTC

GTT-3`); 610R (5`-ACC GCG GCT GCT GGC AC-3`), 610F (5`-GTG CCA GCA GCC

GCG GT-3`), nonEUB (5`-ACT CCTACG GGA GGC AGC-3`).

2.4.5 RFLP analysis of the 16S rRNA gene of isolates

Pure cultures of the selected isolates were harvested from plates (Sly et al. 1999),

washed with sterile 1 x PBS (Sambrook and Russel 2001), resuspended in 20 l 50

mM NaOH and 0.25 % SDS, and heated for 15 min at 94C. PCR amplification of the

16S rRNA gene fragment was done with the primers 9F (5'-GAG TTT GAT CCT GGC

TCA G-3') and 1512R (5'-ACG GCT ACC TTG TTA CGA CTT-3') targeting most bac-

teria (Martiny et al. 2003). The following thermal cycling program was applied: 94C

for 5 min, followed by 25 cycles of 94C for 45 sec, 52C for 45 sec and 72C for 2

min and a final extension step at 72C for 10 min. The enzymes RsaI and MspI were

used separately to restrict the amplified 16S rDNA fragments to increase resolution

power. Six U of RsaI and 8 U of MspI (New England BioLabs, Beverly, Mass., USA)

were employed and the mixtures were incubated at 37C for 2 h. A third enzyme

BstUI was applied to check the resolution power of the technique. Six U of BstUI was

used and the mixture was incubated at 60C for 2 h. The resulting restricted product

was separated on a 1.7 % agarose gel. The software package containing GeneScan

and GeneTools (Syngene, Cambridge, U.K.) was used to obtain and compare the

resulting band patterns. The band pattern was checked manually to control for re-

striction fragments at the same position and short fragments resulting in bands of low

intensity. Isolates with the same band pattern were considered an operational taxo-

nomic unit (OTU) and the abundance was scored. Novel OTUs were identified when

fragment sizes differed more than 5 % and bands below 50 bp were not scored due

to the low fluorescence intensity in the gel following the recommendations by

Vaneechoutte and Heyndrickx (Vaneechoutte and Heyndrickx 2001).

Material and Methods

2.4.6 Phylogenetic analysis

2.4.6.1 Pipe sample isolates

Sequences of the 112 isolates of the pipe samples were aligned by using ClustalW

(http://www.ebi.ac.uk/Tools/clustalw/). The aligned sequences were admitted to the

MEGA 4 software package to group the pipe isolates of this study and to detect iden-

tical isolates (sequence difference <1 %). The distance matrix was calculated

(neighbour-joining, pairwise deletion, p-distance, Transitions+Transversions) and

OTUs determined. Representatives of each OTU were chosen and the closest 16S

rRNA sequence (clone or isolate) and the closest described relative (only published

isolates selected) matching the sequence in the Genbank database were identified

by the BLASTN tool of the NCBI (http://www.ncbi.nlm.nih.gov/blast/Blast.cgi). The

closest sequence and described relative has to show the maximum number of com-

pared bases and lowest gaps.

2.4.6.2 Representative reactor sample isolates obtained from RFLP analysis

Sequence analysis of the OTU representatives obtained from the RFLP anaylsis was

performed with an ABI Prism 310 sequencer as described in chapter 2.4.4. 16S rDNA

sequences were aligned with the ARB software package (Strunk et al. 1999) and

manually corrected for errors. The new sequence was analyzed against the phyloge-

netic tree containing all sequences in the ARB database (6spring2001) using the

maximum parsimony “quick add” tool to get a first estimate of the affiliation of new

bacterial species. The family of each strain was identified and a new tree was recon-

structed using sequences from all species within the corresponding family as de-

scribed in Bergey’s Manual (Bergey`s Manual Trust 2001). Strains were identified

using a consensus based on the neighbour-joining, maximum parsimony and fastdna

maximum likelihood algorithms and the bacterial nomenclature described in the latest

edition of Bergey’s Manual. A base frequency filter was generated based on the se-

lected sequences excluding all positions different in more than 70 % of the strains to

enable a comparison of homologous positions.

Material and Methods

2.5 Statistical analysis

The statistical analysis was performed either with the statistic tools of the Excel 2000

program for simple descriptive statistics or the program STATISTIKA 7.1 (StatSoft,

Inc. 2005, www.statsoft.com). The samples were tested for normal distribution by the

Kolmogoroff-Smirnoff test with Lilliefors correction and the Shapiro-Wilk test. If the

values showed no normal distribution they were log transformed and 1 was added to

all values to avoid negative log values. If not further noted, non-transformed data

were used. As a consequence of the results parametric or non parametric test statis-

tic for dependent or independent samples was chosen. The chosen tests are de-

scribed in the result chapters. The error probability alpha was 0.05 for all tests.

Results

3 Results

3.1 Pipe samples taken from the distribution systems in Berlin

and the Ruhrgebiet

As shown in table 8 in total 18 pipe samples were taken in the drinking water distribution sys-

tems of the Ruhrgebiet and in Berlin by the working groups of Prof. Dr. H.-K. Flemming of the

Universität Duisburg or the working group of Prof. Dr. Szewzyk at the TU Berlin in the years

1999, 2000 and 2001. The pipe samples were assigned with the internal code M II to XII for

the samples in the Ruhrgebiet and BWB I to VII for pipes taken from the distribution system

of the Berliner Wasserbetriebe. Most of the pipes consisted of the materials PVC, cement

and the metallic materials grey cast, cast iron and Tyton. Two exceptional materials are a

cast iron pipe with an inline material and a tin coated steel pipe. Thirteen of the sampled

pipes had a diameter of DN 100, in single cases it was DN 200, 150, 125 and 50. The time of

exposition in the drinking water system varied from 12 to 34 years for the PVC pipes and

from 8 to 20 years for cement pipes. The metallic materials showed a greater range of expo-

sition time from 24 to 99 years. Another parameter of interest for the pipe samples is the

scraped inner pipe surface. For 15 of the 18 pipe samples more than 2500 cm² were scraped

(min. 2540 cm², max. 5511 cm²). For the 10 year old cement pipe the minimum inner pipe

surface scraped was 298 cm². The pH of the scraped biofilm suspensions ranged between

7.9 and 8.2 for the PVC pipes and between 5.7 and 7.7 for the metallic materials. Values

above pH 9 were measured for M XI and M XII due to the material cement. Two further inter-

esting parameters for the description of the pipe samples are disinfection and temperature of

the supplied drinking water. These parameters were measured in the bulk water after it had

flown through the pipe and before the pipe section was cut out of the distribution system. In

the drinking water distribution system of the Ruhrgebiet the water was generally disinfected

this resulted in a free chlorine content of the bulk water phase between ≤ 0.01 mg/mL and

0.13 mg/mL3(Flemming 2003). The Berlin drinking water was normally not disinfected except

the tin coated steel pipe where a locally limited chlorination was carried out. Temperature of

the free water at the sampling day ranged between a minimum of 6°C for a 24 years old grey

cast iron pipe and a maximum value of 16.4°C measured in a 20 years old cement pipe.

3The free chlorine conc. was determined by the N,N-diethyl-p-phenylenediamine colorimetric method on site by

the working group in Duisburg.

Results

Tab. 8: Pipe samples taken from the drinking water distribution systems in Berlin and the Ruhrgebiet.

Sample/

date Pipe material Sampling site Age in

years Scraped inner

pipe surface

cm²

pH Free water phase

free chlo-

rine

mg/mL

Temperature

°C

M VI

19/10/99 PVC

DN 100 Dorsten-

Holsterhausen

Rhade 12 3142 7.9 n. dis. n.d.

BWB II

02/02/00 PVC

DN 200 Berlin

Lichtenberg 24 5511 7.9 n. dis. 8.2

BWB IV

17/04/00 PVC

DN 150 Berlin

Lichtenberg 24 2779 8.0 n. dis. 10.9

M II

08/06/99 PVC

DN 100 Duisburg

Rheinhausen 28 3110 8.0 <0.01 n.d.

M VII

09/11/99 PVC

DN 100 Duisburg

Friemersheim 34 3456 8.2 <0.01 n.d.

M XI

11/09/01 Cement

DN 100 Duisburg

Ungelsheim 8 3272 9.2 0.05 12.7

BWB I

23/06/99 Cement

n.d. Berlin

Lichterfelde approx.

10 298 n.d. n. dis. n.d.

M XII

09/10/01 Cement

DN 100 Duisburg

Wanheim 20 3343 9.1 0.13 16.4

BWB III

29/02/00 grey cast iron

DN 100 Berlin

Lichtenrade 24 1860 7.2 n. dis. 6

M V

14/09/99 grey cast iron

DN 100 Duisburg

Neudorf 30 3340 6.7 <0.01 n.d.

M III

13/07/99 grey cast iron,

cement

DN 125

Mülheim (Ruhr)

Raadt 37 3244 7.7/7.3 <0.01 n.d.

M IV

10/08/99 grey cast iron

DN 100 Oberhausen 99 3202 5.7 0.01 n.d.

M X

06/03/01 cast iron

DN 100 Duisburg

Buchholz 62 2540 6.4 0.02 6.6

BWB V

05/06/00 cast iron

DN 100 Brandenburg approx.

73 2942 7.3 n. dis. 14.8

M IX

17/10/00 Tyton

(spheroidal

graphite)

DN 100

Duisburg

Wanheimerort 26 3189 6.7 0.01 13.9

M VIII

17/10/00 Tyton

(spheroidal

graphite)

DN 100

Duisburg

Bissingheim 27 3810 6.4 <0.01 13.8

BWB VI

10/10/00 tin coated

steel

DN 50

Berlin

Schönerlinde approx.

10 976 7.2 0.13 15

BWB VII

23/04/01 cast iron with

inline material

DN 100

Berlin

Grunewald 2 3009 7.5 n. dis. 9.9

n. dis.: no disinfection n.d.: not determined