Appl Microbiol Biotechnol (2001) 57:776–785
DOI 10.1007/s002530100826
O R I G I N A L PA P E R
E. Vincke · N. Boon · W. Verstraete
Analysis of the microbial communities
on corroded concrete sewer pipes – a case study
Received: 26 July 2001 / Accepted: 27 August 2001 / Published online: 18 October 2001
© Springer-Verlag 2001
Abstract Conventional as well as molecular techniques
have been used to determine the microbial communities
present on the concrete walls of sewer pipes. The genetic
fingerprint of the microbiota on corroded concrete sewer
pipes was obtained by means of denaturing gradient gel
electrophoresis (DGGE) of 16S rRNA gene fragments.
The DGGE profiles of the bacterial communities present
on the concrete surface changed as observed by shifts
occurring at the level of the dominance of bands from
non-corroded places to the most severely corroded plac-
es. By means of statistical tools, it was possible to distin-
guish two different groups, corresponding to the micro-
bial communities on corroded and non-corroded surfac-
es, respectively. Characterization of the microbial com-
munities indicated that the sequences of typical bands
showed the highest level of identity to sequences from
the bacterial strains Thiobacillus thiooxidans, Acidithio-
bacillus sp., Mycobacterium sp. and different hetero-
trophs belonging to the α-, β- and γ-Proteobacteria, Ac-
idobacteria and Actinobacteria. In addition, the presence
of N-acyl-homoserine lactone signal molecules was
shown by two bio-assays of the biofilm on the concrete
under the water level and at the most severely corroded
places on the concrete surface of the sewer pipe.
Introduction
Biogenic sulfuric acid attack of concrete sewer pipes,
wastewater collection systems and treatment plants can
cause severe damage and has been described in the Unit-
ed States (Islander et al. 1991; Padival et al. 1995; Davis
et al. 1998; Peccia et al. 2000), Germany (Sand et al.
1987; Schmidt et al. 1997), Japan (Mori et al. 1991;
E. Vincke · N. Boon · W. Verstraete (✉)
Laboratory of Microbial Ecology and Technology (LabMET),
Faculty of Agricultural and Applied Biological Sciences,
Ghent University, Coupure Links 653, 9000 Ghent, Belgium
e-mail: Willy.Verstraete@rug.ac.be
Tel.: +32-9-2645976, Fax: +32-9-2646248
URL: http://welcome.to/labmet
Tazawa et al. 1996) and many other places in the world as
well (Vincke et al. 2000). Mansfeld et al. (1990) suggest-
ed that the emergence of the problem in the USA could
be related to the National Pollution Discharge Elimina-
tion Systems (NPDES) in 1972, which regulates the loads
of toxic chemicals discharged into sewers and water bod-
ies. As a result, microbial activity increased significantly
in wastewater systems. In Belgium, a general concern is
the uncoupling of sewage and surface water leading to
higher concentrations of nutrients and a longer residence
time of the sewage in the pipes. The sulfuric acid causing
corrosion of sewer crowns is generated by a complex mi-
crobial ecosystem (Islander et al. 1991). Anaerobic con-
ditions in sewage support microorganisms that convert
oxidized sulfur compounds. The formed hydrogen sulfide
volatilizes to the sewer atmosphere and dissolves in the
condensate on the sewer crown. Sulfur-oxidizing bacteria
are then capable of oxidation of elemental sulfur (Sand
and Bock 1984; Kelly 1982; Trüper 1984; Hazeu et al.
1988). The effects of the corrosive attack on concrete in
sewers can be of the order of several mm per year (Mori
et al. 1991). Proper design in new collection systems re-
duces anaerobic microbial activity and consecutively sul-
fide generation. However, control should also be possible
at the point of the aggressive sulfuric acid production due
to microbiological sulfide oxidation. A very important
factor in this process is microbial colonization. Efforts to
restore concrete sewer pipes will be ineffective without a
better understanding of the fundamental processes of bio-
genic sulfuric acid corrosion.
The pH of uncorroded concrete is approximately 12, and
sulfur-oxidizing bacteria cannot grow when directly ex-
posed to such a high pH. After neutralization of the con-
crete due to carbonation and acidic H2S, sulfate-produc-
ing bacteria are able to grow on the surface. The estab-
lishment of Thiobacillus thiooxidans, which especially
dominates at very low pH, in corroded concrete pipes is
presumed to be preceded by moderately acidophilic
Thiobacillus species, for example, T. intermedius, T.
neapolitanus and T. thioparus (Milde et al. 1983). Cul-
ture studies have shown that growth of T. thiooxidans re-

quires a relationship with an acidophilic heterotroph
(Sand and Bock 1984; Cho and Mori 1995). The degra-
dation of the excreted organics of the sulfur-oxidizing
bacteria by an acidophilic heterotroph is required for a
continued growth of the acid-producing bacteria. Cho
and Mori (1995) found the coexistence with T. thiooxi-
dans of an acid-resistant and H2S-oxidizing fungus on
severely corroded sewer pipes. This fungus oxidizes H2S
to thiosulfate, probably as a sulfide detoxification pro-
cess. Thiosulfate can then be used by T. thiooxidans as
energy source. Sand and Bock (1984) found heterotrophs
on concrete surfaces from pH 6.2 to pH 1.5.
The present study describes the analysis of microbial
communities on corroded sewer pipes. To obtain a good
representation of the actual populations at the site, con-
ventional as well as DNA-based identification techniques
were used. Traditional cultivation techniques for the en-
richment and isolation of microbial cells yield only a lim-
ited fraction of all microorganisms present (Amann et al.
1995). Molecular techniques, more specifically polymer-
ase chain reaction (PCR) and subsequently denaturing
gradient gel electrophoresis (DGGE) and cloning, have
been proven to be useful in describing more accurately
the microbial communities in environmental samples and
overcome some of the limitations and biases of culture-
based techniques (Muyzer 1999; Boon et al. 2000). How-
ever, the sensitivity of the molecular techniques is also
limited (Muyzer et al. 1993) and environmental samples
can give some difficulties for molecular analysis, as, for
example, calcium and heavy metals inhibit nucleic acid
amplification during PCR (Wilson 1997). Therefore, the
two techniques were used in parallel.
To get a more complete image of the biofilm compo-
sition, the presence of N-acyl homoserine lactone (AHL)
activity in corroded concrete samples was studied by
means of two bio-assays. Recently, there has been an in-
creasing appreciation and knowledge that bacteria can
perceive and respond to other bacteria using signal mole-
cules (Greenberg 1997). These small molecules are ex-
creted by cells and accumulate in cultures as a function
of cell density. Typical signal molecules excreted by
gram-negative bacteria are AHL derivatives (McLean et
al. 1997). At a threshold population density, the accumu-
lated AHL molecules can interact with receptors on
the bacterial cell surface that control gene expression
(Stickler 1999). This regulatory mechanism is called
autoinduction or quorum-sensing. Different AHL mole-
cules exist and bacteria are able to produce some specific
molecules with different properties. Signal specificity is
conferred by the length and the nature of the substitution
at C-3 of the acyl side chain (Cha et al. 1998). Neverthe-
less, many genes can be activated by heterologous AHL
autoinducers, implying that bacteria can sense and react
to population densities of their own species and popula-
tion densities in general. As biofilms typically contain
high concentrations of cells, AHL molecules were pro-
posed as essential components of biofilm physiology by
McLean et al. (1997). These authors showed for the first
time AHL activity in naturally occurring biofilms.
777
Material and methods
Sampling
Samples were taken at different places on a concrete sewer pipe
exhibiting moderate to heavy corrosion (Fig. 1). The corroded
pipe is part of the sewerage network of Oostende (Belgium). The
first series of samples (I.1-I.5, II) was taken on different parts of
the sewer pipe at the most severely corroded point (A), situated
immediately after connection of two pressure lines into the gravity
sewer pipe. The second series of samples was taken at the crown,
just above the water line and at the bottom of the pipe on different
points starting at the connection of the pressure line (A), and
100 m (B) and 200 m (C) downstream, respectively.
Biofilm samples were taken with a sterile spatula and trans-
ferred into sterile Falcon tubes. Samples of biofilm and corrosion
products were collected and 2 g of each were weighed for DNA
extraction and further molecular analysis. Samples for DNA ex-
traction were stored at –20 °C before analysis. For conventional
counting techniques and signal molecule analysis, 1 g of the cor-
roded samples, consisting of biofilm and corrosion products, was
suspended in 9 ml of sterile physiological solution (8.5 g NaCl
l–1). In addition to the biofilm samples, water was sampled after
the pressure line. The same procedure was followed as for the bio-
film samples.
Conventional counting techniques
Total heterotrophic colony-forming units were determined with
nutrient agar (Difco) using the spread plate technique. For the de-
termination of fungi, malt extract agar (2%, Difco) was used
(Atlas 1993). The plates were incubated in the dark at 28 °C for 3
and 7 days respectively (n=3).
The cell numbers of thiobacilli were determined by a most
probable number (MPN) technique using three selective nutrient
solutions (Vincke et al. 1999).
DNA-based identification techniques
DNA extraction and purification
Total DNA from the corrosion samples was obtained by a modi-
fied extraction method described previously (El Fantroussi et al.
1999). To obtain DNA, 2 g of the samples or 2 ml of the sewage
water was transferred to 4 ml of 10 mM Tris/HCl (pH 9) and 3 g
of beads (0.10–0.11 mm). This mixture was beaten three times for
90 s using a bead beater at 2,000 rpm (Braun Biotech Internation-
al, Melsungen, Germany). After this, 160 µl of 50 g l–1 lysozyme
was added, and the suspension was incubated at 28 °C for 15 min
on a shaker (200 rpm). Chemical lysis of the bacterial cells was
achieved by adding 300 µl of 20% SDS and slowly mixing the
suspension for 5–10 min. Subsequently, 1 ml of 8 M ammonium
acetate was added. DNA was obtained from the lysates using stan-
dard chloroform-isoamyl alcohol (24:1) extraction and isopropa-
nol precipitation procedures (Boon et al. 2000). The total amount
of nucleic acids extracted from the corrosion samples and the wa-
ter sample was resuspended in 500 and 250 µl DNase-free water,
respectively. Aliquots of, respectively, 250 (corrosion samples)
and 150 (water sample) µl of the crude extract were further puri-
fied using Wizard PCR preps (Promega, Madison, Wis.).
Amplification of 16S rRNA genes
The extracted prokaryotic DNA was amplified by a nested PCR in
two subsequent rounds using a 9600 thermal cycler (Perkin-Elmer,
Norwalk, Conn., USA). PCR amplification was carried out in 25 µl-
reactions and 2 µl of extracted total DNA was added. The 16S rRNA
genes from the microbial community in the water samples were am-
plified by PCR (Weidner et al. 1996; Boon et al. 2000) using primers
778
Fig. 1 A Overview of sewer
system, samples were taken
near the connection (A), and
100 m (B) and 200 m(C) down-
stream. B Detailed scheme of
corroded concrete pipe near the
connection and profile of cor-
rosion measured during the
first series of sampling

complementary to conserved regions of the 16S rDNA. Using the
forward primer R1n (5′GCTCAGATTGAACGCTGGCG3′-forward)
and the reverse primer U2 (5′ACATTTCACAACACGAGCTG3′-re-
verse) DNA fragments were amplified (Weidner et al. 1996). After
the first amplification round, the amplified product was used to start
a new PCR round. In the second PCR round, the forward primer
P63f (5′CAGGCCTAACACATGCAAGTC3′-forward), which was
prolonged at its 5′ end with a 40-base GC clamp, and the reverse
primer P518r (5′ATTACCGCGGCTGCTGG3′-reverse) were used.
The PCR master mix contained 0.5 µM of each primer, 200 µM
of each deoxynucleoside triphosphate, 1.5 mM MgCl2, 10 µl of
thermophilic DNA polymerase 10× reaction buffer (MgCl2-free),
2.5 U of Taq DNA polymerase (Promega, Madison, Wis., USA),
400 ng µl–1 of bovine serum albumin (Boehringer) and sterile wa-
ter to a final volume of 100 µl. PCR was performed in a 9600 ther-
mal cycler as follows: 94 °C for 5 min, followed by 30 cycles of
92 °C for 1 min, 53 °C for 1 min, and 72 °C for 2 min. All PCR
products (10-µl volumes) were analyzed by electrophoresis in 1%
(w/v) agarose gels before DGGE was performed.
Analysis of PCR products by DGGE
DGGE was performed as previously described by Boon et al.
(2000) using the Bio-Rad D Gene System (Hercules, Calif., USA)
with 6% (w/v) polyacrylamide gels in 1× TAE (20 mM Tris,
10 mM acetate, 0.5 mM EDTA, pH 7.4) containing a linear chemi-
cal gradient ranging from 45 to 60% denaturant (100% denaturant
contains 7 M urea and 40% formamide). PCR products (8 µl) ob-
tained from the second PCR round of the corrosion and water
samples were used for separation in denaturing gradient gels. The
electrophoresis was run for 17 h at 60 °C and 45 V. After comple-
tion of electrophoresis, the gels were soaked for 5 min in fixation
buffer (10% ethanol, 0.5% acetic acid) and subsequently for
10 min in SYBR green I nucleic-acid gel stain (1:10.000 dilution;
FMC BioProducts, Rockland, Me., USA). The stained gel was im-
mediately photographed on an UV transillumination table with a
video camera module (Vilbert Lourmat, Marne-la vallé, France).
DGGE pattern analysis
Statistical comparison of different DGGE banding patterns was
done with the GelCompar software 4.1 (Applied Maths, Kortrijk,
Belgium). The clustering algorithm of Ward (1963) was used to
calculate dendrograms.
DNA cloning and sequencing
Amplified 16S rDNA fragments, obtained after the first PCR
round, were cloned by using the TOPO TA cloning kit (Invitrogen,
Carlsbad, Calif., USA) according to the manufacturer's instruc-
tions. DNA sequencing was performed by Genetik (Bielefeld,
Germany) on 22 clones. Analysis of DNA sequences and homolo-
gy searches were carried out with the BLAST server of the
National Center for Biotechnology Information (NCBI) using
the BLAST algorithm for the comparison of a nucleotide query
sequence against a nucleotide sequence database (Altschul et al.
1997).

Nucleotide sequence accession numbers
Nucleotide sequences for fragments 1–22 were deposited in Gen-
Bank database and are given in Table 2.
N-acyl homoserine lactone detection
Bacterial strains
The strains Agrobacterium tumefaciens A136 (Ti–) (pCF218)
(pCF372), A. tumefaciens KYC6 (traM::Tn5-gusA harboring
pCF218) and Chromobacterium violaceum CV026 were provided
by C. Fuqua (Fuqua and Winans 1996). For long-term storage, all
cultures were suspended in a mixture of LB broth and glycerol and
frozen at –70 °C. Prior to use, frozen cultures were removed from
storage and incubated on AT medium supplemented with 50 µg
spectinomycin ml–1 and 4.5 µg tetracycline ml–1 (A. tumefaciens
A136), AT medium with 100 µg kanamycin ml–1 and 4.5 µg tetra-
cycline ml–1 (A. tumefaciens KYC6), and LB medium (C. violaceum
CV026) respectively.
AHL cross-feeding assay
A cross-feeding assay (Fuqua and Winans 1996) was used for
AHL detection. This assay consisted of streaking the AHL report-
er strain, A. tumefaciens A136, on AT medium containing X-gal
(40 µg ml–1) and then placing the sample (synthetic AHLs or cor-
rosion product) approximately 1 cm away. If AHL activity was
present, it would diffuse through the agar to the reporter strain and
be detected due to activation of the traI-lacZ fusion by TraR. Posi-
tive and negative controls consisted of culturing the reporter strain
with A. tumefaciens KYC6 (AHL overproducer) and A. tumefaci-
ens A136 (AHL reporter strain).
AHL reporter plate bioassay
Chromobacterium violaceum CV026 was grown overnight, shak-
ing in LB broth at 30 °C. From this culture, 50 µl was added to
5 ml of molten semi-solid LB-agar 0.6% (w/v). Immediately after
inoculation, it was poured over the surface of prewarmed LB agar
plates. After solidifying of the overlaid agar, wells were punched
in the agar with a sterile cork-borer (diameter 6 mm). The wells
were filled with the samples to be assayed (i.e. synthetic AHLs,
corrosion products). The Petri dishes were incubated in the upright
position for 24 h at 30 °C. The induction of violacein synthesis by
AHL signal molecules is indicated by blue/purple pigmentation of
the bacterial lawn around the wells. Positive (N-hexanoyl DL-ho-
moserine lactone) and negative (sterile LB broth) controls were in-
cluded in each assay plate (McClean et al. 1997).
Table 1 Determination of the bacterial/fungal densities and sulfur-oxidizing bacteria at different sections of a corroded concrete sewer pipe
Bacteriaa Fungia Thiobacillus
thiooxidansb
I.1 5.17±0.68 5.98±0.43 6.15
I.2 6.23±0.45 5.20±0.58 5.65
I.3 8.90±0.40 6.30±0.50 4.98
I.4 8.04±0.55 5.00±0.60 4.40
I.5 8.71±0.70 5.60±0.68 3.97
II 6.04±0.95 4.48±0.70 3.40
a log CFU (g corrosion material)–1 (I.1–I.5) or log CFU ml–1 (II); data
10 10
b log MPN (g corrosion material)–1 (I.1–I.5) or log MPN ml–1 (II)
10 10
779
Results
A first series of samples was taken immediately after the
connection of the pressure lines into the gravity sewer
pipe (A). At that point, different samples were taken
from the concrete surface, starting at the crown of the
pipe (I.1) to the bottom of the pipe (I.5) and the sewage
water (II) (Fig. 1). After 1 month, new samples were tak-
en on the respective places and the same procedures
were followed. The samples were labeled a and b, re-
spectively.
Together with an increasing corrosion level from be-
neath the water level to the crown of the sewer pipe, a
pH drop of the concrete surface was seen. The pH varied
from 7.2 in the biofilm covering the bottom to 2.9 at the
crown of the sewer pipe. Conventional plate-counting
techniques on general media revealed the smallest
amount of bacteria in the crown samples of the sewer
pipe (Table 1). The crown (I.1), characterized by a low
pH and a highly corroded surface, was colonized by ap-
proximately 105 and 106 CFU (g corrosion product or
biofilm)–1 of bacteria and fungi, respectively. Above the
water level (I.3), characterized by a moderately corroded
surface, higher bacterial and fungal densities were de-
tected, i.e. 108 and 106 CFU (g corrosion product)–1 for
bacteria and fungi, respectively. The largest amount of
bacteria but fewer fungi were found under the water lev-
el, 5×108 CFU (g biofilm)–1 and 4×105 CFU (g bio-
film)–1. MPN counts of the most important sulfur-oxidiz-
ing bacteria showed the largest number of acidophilic
sulfur-oxidizing bacteria, more specifically T. thiooxi-
dans, at the crown of the sewer pipe (Table 1); 106 MPN
(g corrosion product)–1 were found in specific media for
T. thiooxidans. The use of specific media for moderately
acidophilic thiobacilli showed the largest abundance of
these bacteria in samples I.2–I.4 i.e. at moderately cor-
roded places.
In order to monitor and compare the microbial com-
munities on different places of a corroded sewer pipe,
the diversity of 16S rRNA gene fragments was exam-
ined. The DGGE profile showed a decrease in the micro-
bial diversity from the sewage water to the more severe-
ly corroded concrete surface in the crown of the sewer
pipe (Fig. 2A). A limited amount of bands can be ob-
served in sample I.1. Gradually, an increasing amount of
Thiobacillus Thiobacillus
novellus/intermediusb neapolitanusb
4.81 4.18
6.15 4.40
5.04 4.88
6.15 4.40
4.30 3.98
4.60 3.00
are means±SD (n=3)
780
Fig. 2 A Denaturing gradient gel electrophoresis (DGGE) profile
of the microbial diversity of the sampled biofilms and water
phase. Samples b were taken 1 month later at the respective places
as samples a. B Cluster analysis of the DGGE pattern; C densio-
gram of DGGE profile for samples I.5b (bottom), I.3b (above the
water level) and I.1b (crown)
bands appeared in the DGGE profile. A larger spreading
of the bands, as well as a larger quantity of bands, is ob-
served in sample I.4–I.5. To determine the information
content of the banding patterns in terms of differences in
microbial community structure, they were analyzed by
clustering. Two groups were obtained (Fig. 2B). It is in-
teresting to notice the clear difference between the mi-
crobial communities on the corroded places in compari-
son with the microbial communities under the water lev-
el (water, as well as biofilm). The fingerprints of the
samples showed differences in both band position and
intensity. This was visualized by means of a densiogram
based on the DGGE profile of samples I.1, I.3 and I.5
(Fig. 2C). An important alteration of the microbial com-
munities on the concrete pipe from the sewer crown to
the bottom of the sewer pipe was seen. The most domi-
nant species in the corroded samples I.1, I.2 and I.3 were
identified (Table 2). At the most severely corroded place,
not only acidophilic bacteria (Thiobacillus and Acidi-
thiobacillus sp.) were found, but also Mycobacterium sp.
In sample I.2, which represents a corroded surface sur-
rounding the steel bar, a larger variety of microorgan-
isms was identified, belonging to the α-Proteobacteria,
β-Proteobacteria, γ-Proteobacteria, Actinobacteria, Ac-
idobacteria, Flavobacteria and the Bacillus group.
Again a smaller diversity was seen in sample I.3; the
identified bacteria belonged to the Actinobacteria, α-
Proteobacteria and γ-Proteobacteria.
Two bio-assays showed the presence of AHL signal
molecules in the biofilm under the water level and at the
most severely corroded places (Table 3). A strong signal
was seen for the biofilm samples under the water level.
A deep blue coloration of the indicator strain A. tumefa-
ciens A136 occurred in the cross-streaking assay. Posi-
tive results were confirmed with the AHL reporter plate
bioassay by purple staining of the reporter strain in the
surrounding of the biofilm samples I.4 and I.5. No signal
molecules could be detected in water sample II, sample
I.2 and sample I.3. A small AHL signal could be ob-
served in sample I.1; a weak blue coloration in the A. tu-
mefaciens cross-streaking test and a slight coloration in
the C. violaceum test occurred.
A second series of samples was taken at the connec-
tion (A), and 100 m (B) and 200 m (C) beyond the con-
nection of the pressure lines downstream into the gravity
pipe. At each distance, three samples were taken, namely
at the crown (1), the water level (2) and the bottom (3) of
the sewer pipe, respectively. In the first series of samples
a large difference could be observed in the pH profile of
the sewer pipe. This was confirmed in the second series
(Table 4). By moving further from the connection, a
weaker corrosive action was observed as well as smaller
differences in the pH profile. At a distance of 100 m, the
pH varied from 4.2 to 7.4. The pH variation of the bio-
film on the concrete surface was even smaller 200 m
downstream of the connection; it varied from 5.5 to 7.4.
Conventional counting techniques showed only small
differences between the bacterial densities for the differ-
ent profiles. At each sampling point, an increasing
amount of bacteria were present from the crown to the
bottom of the sewer pipe, varying in the order of 105–106
to 108–109 per g biofilm. MPN counting revealed some
differences in the number of acidophilic and moderately
acidophilic sulfur-oxidizing bacteria. In the samples tak-
en of the silt or the biofilm in the bottom of the pipe, the
amounts of T. thiooxidans, T. novellus/intermedius and T.
neapolitanus remained more or less constant. Some larg-
er differences could be detected in samples taken at the
crown or above the water level. Comparing samples tak-
en at the crown immediately after the connection and
100 m downstream revealed an alteration of the sulfur-
 781
Table 2 Sequence similarities after cloning and sequencing of ter level (I.3b). Sequences were aligned to their closest relatives in
amplified 16S rDNA fragments for samples taken at the crown public databases by using the BLASTN 2.0.13 program
(I.1b), in the surrounding of the steel bar (I.2b) and above the wa-

Sample Similarity (%) Speciesa Accession no. Taxonomic description

I.1
1 99% (533/537) Acidithiobacillus thiooxidans (B-S3) (X75269.1) AF379017 γ-Proteobacteria
2 99% (512/516) Thiobacillus thiooxidans (Y11596.1) AF379018 γ-Proteobacteria
3 98% (530/537) Mycobacterium sp. IMVS B76676 (AF016407.1) AF379016 Actinobacteria
4 99% (526/528) Mycobacterium IWGMT 90174 (X88908.1) AF379019 Actinobacteria
5 99% (443/444) Thiobacillus thiooxidans (Y11596.1) AF379020 γ-Proteobacteria
I.2
6 97% (421/430) Methylobacterium sp. (Z23158.1) AF379021 α-Proteobacteria
7 94% (437/462) Aeromicrobium erythreum (AF005021.1) AF379022 Actinobacteria
8 89% (379/422) Metal-contaminated soil clone K20–60 (AF145853.1) AF379037 α-Proteobacteria
9 99% (434/436) Methylobacterium sp. (Z23158.1) AF379023 α-Proteobacteria
10 96% (451/469) Flavobacterium aff. xylanivorum (AJ297440.1) AF379024 Flavobacteria
11 98% (458/467) Unidentified β-Proteobacterium (D84603.1) AF379025 β-Proteobacteria
12 93% (405/431) Agrobacterium sp. (AB006037.1) AF379026 α-Proteobacteria
13 95% (465/487) Cellvibrio sp. R4079 (AJ289164.1) AF379027 γ-Proteobacteria
14 98% (315/320) Bacillus psychrodurans(AJ277984.1) AF379028 Bacillus group
15 96% (476/493) Stenothrophomonas maltophilia (AJ131907.1) AF379029 γ-Proteobacteria
16 98% (395/399) Methylobacterium sp. (Z23160.1) AF379030 α-Proteobacteria
17 96% (377/391) Pseudomonas pseudoalcaligenes (Z76666.1) AF379031 γ-Proteobacteria
I.3
18 98% (305/311) Uncultured bacterium DA008 (Y12597.1) AF379032 Acidobacteria
19 97% (307/314) Mycobacterium sp. IMVS B76676 (AF016407.1) AF379033 Actinobacteria
20 98% (262/266) Propionibacterium acnes (AF145256.1) AF379034 Actinobacteria
21 97% (302/310) Mycobacterium sydneyiensis (AF101243.2) AF379035 Actinobacteria
22 95% (383/402) Sphingomonas sp. MK331 (D84519.1) AF379036 α-Proteobacteria

Table 3 Detection of N-acyl-

homoserine lactone production
in the sampled biofilm and
corrosion products and in the
water phase
Reporter: Reporter:
A. tumefaciens NTL4 C. violaceum CV026
Overproducer: A. tumefaciens KYC6 +++ –
Negative control – –
I.1: Heavily corroded surface + +
I.2: Corroded surface surrounding steel bar – –
I.3: Moderately corroded surface – –
I.4: Not corroded surface under the water level ++ +
I.5: Silt, bottom of sewer pipe ++ +
II: Sewage water – –

Table 4 Overview on pH, detection of N-acyl-homoserine lactone (AHL) molecules and bacterial density at different locations of a cor-
roded concrete sewerage system

pH AHLa Bacteriab T. thiooxidansc T. novellus/ T. neapolitanusc

intermediusc

Connection pressure A.1 3.1 + 5.60±0.78 6.15 4.81 3.40

lines-gravity pipe A.2 3.4 – 7.91±0.45 5.15 6.15 4.98
A.3 7.2 ++ 8.48±0.60 3.18 5.40 4.18
100 m beyond B.1 4.2 + 6.11±0.63 5.65 5.81 4.98
connection B.2 4.9 – 8.60±0.71 4.40 6.04 4.88
B.3 7.4 ++ 9.30±0.43 3.97 4.30 3.00
200 m beyond C.1 5.5 – 6.95±0.55 4.40 4.40 4.18
connection C.2 6.0 – 7.84±0.53 4.40 6.04 4.40
C.3 7.4 ++ 8.90±0.76 3.40 4.30 3.00
Water D 7.8 – 8.85±0.49 4.40 5.60 4.00
aCross-feeding bioassay (reporter A. tumefaciens NTL4)
blog CFU (g corrosion material)–1 (A–C) or log CFU ml–1 (D); data are means±SD (n=3)
10 10
clog MPN (g corrosion material)–1 (A–C) or log MPN ml–1 (D)
10 10
782
Fig. 3 DGGE band pattern of the second series of samples. Sam-
ples were taken at the crown (1), at the water level (2) and under
the water level (3), respectively, at the connection (A), and 100 m
(B) and 200 m (C) downstream. Sample D is the water sample
oxidizing bacteria from acidophilic to moderately acido-
philic sulfur-oxidizing bacteria. In sample A.1 approxi-
mately 106, 105 and 103 MPN (g corrosion product)–1 for
T. thiooxidans, T. novellus/intermedius and T. neapolit-
anus, respectively, were detected. For sample B.1 ap-
proximately 105, 106 and 105 MPN (g corrosion prod-
uct)–1 for T. thiooxidans, T. novellus/intermedius and T.
neapolitanus, respectively, were counted.
The microbial communities on the different places
were further analyzed by examining the diversity of 16S
rRNA fragments (Fig. 3). As described earlier, a large
difference occurred in the pattern of DGGE bands near
the connection (A1–3). Also the water sample was char-
acterized by a large diversity, illustrated by the large va-
riety (D). The global profile of the three samples taken
100 m downstream (B1–3) corresponds to a large extent
with the profile obtained near the connection (A1–3).
The number of bands is more or less the same, although
the bands differ in intensity. Moving further from the
connection, a larger difference in the profiles, both in the
number of bands and in intensity, was observed (C1–3).
The in-between difference of the samples taken 200 m
downstream is small in comparison with the in-between
difference of the samples taken near the connection and
100 m downstream.
The confirmation of the presence of signal molecules,
detected during the first series, on the concrete surface
was obtained during the second series of samples (Ta-
ble 4). No signal molecules were detected in the sewage
water. A clear signal was obtained for each sample of the
biofilm taken at the concrete surface under the water lev-
el, regardless of the distance from the connection or the
corrosion front. A faint signal was detected for the sam-
ples taken at the most corroded places, i.e. the crown, at
point A and B. However, the moderately corroded sam-
ples showed no evidence of signal molecule production
in the formation of AHL by using the two bio-assays.
Discussion
Reports in the literature show that severe corrosion of
concrete sewer pipes often occurs at points of high tur-
bulence and consecutively H2S release into the sewer at-
mosphere (DeHollander 1998). This is confirmed in our
case study. Severe damage is located near the connection
of two pressure lines into a gravity pipe, which is accom-
panied by a high-turbulence flow.
The corrosion front, with the most severe corrosion
damage at the crown, is related to the surface pH and
characterized by the appearance of specific microbial
communities. Conventional counting techniques were
used to get an idea of the microbial communities present
on the corroded surface of concrete sewer pipes. From
the study of Gu et al. (1998), it was apparent that thioba-
cilli were not the sole microorganisms causing degrada-
tion of concrete. The work of Davis et al. (1998) showed
the absence of nitrate-reducing bacteria, sulfate-reducing
bacteria and anaerobic heterotrophs in corrosion sam-
ples. Therefore, this work focused on aerobic bacterial
and fungal counts on general media and MPN counts in
specific media for sulfur-oxidizing bacteria. To get a bet-
ter and more reliable profile of the genetic diversity of
the complex microbial communities, molecular tools
were used. With the above-described procedure, reliable
and reproducible DGGE band patterns could be obtained
for corroded concrete samples.
Conventional MPN techniques showed the presence of
T. thiooxidans, T. novellus/intermedius and T. neapolit-
anus. These results correspond with data obtained in liter-
ature (Davis et al. 1998). Bacterial counts of the same or-
der of magnitude as reported by these authors were ob-
tained in the outer part of the corrosion layer. The species
of the genus Thiobacillus fall into the α-, β- and γ-sub-
classes of the Proteobacteria. These species exhibit al-
most as much diversity in DNA composition and physiol-
ogy as is found in all other proteobacterial groups. Re-
cently, T. thiooxidans and some other Thiobacillus spe-
cies were reassigned to three newly designated genera
within the γ-Proteobacteria, namely Acidithiobacillus
(e.g. Thiobacillus thiooxidans, Thiobacillus ferrooxi-
dans), Halothiobacillus (e.g. Thiobacillus neapolitanus)

and Thermithiobacillus (Kelly and Wood 2000). Howev-
er, by DGGE of the obtained fragments, no moderately
acidophilic sulfur-oxidizing bacteria were found, in con-
trast to the results obtained by conventional MPN tech-
niques with specific media (Table 1). This can possibly
be explained by: (1) the sampling procedure and/or (2) by
the limitations of the used technique. It has been demon-
strated in the literature that results may be complicated by
variation in community composition with depth of the
corrosion layer. The lowest pH values are found at the
surface, while deep in the concrete pH values are high. It
is notable that Sand and Bock (1984) found a substantial
abundance of T. intermedius/novellus and T. neapolitanus
along with the higher abundance of T. thiooxidans on
concrete at pH 1.5. But these authors used whole pieces
of concrete rather than collecting organisms from the sur-
face. It is possible that T. thiooxidans was growing in the
low-pH environment at the surface, while the other spe-
cies were growing at higher pH values within the con-
crete. In the described procedure, samples were taken of
the outer layer of the corroded concrete without structural
strength, i.e. composed of softened concrete, corrosion
products and micro-organisms. In addition to these con-
siderations, it must be stressed that also the sensitivity of
the detection of 16S rDNA sequence variants is limited.
Experiments (Muyzer et al. 1993) showed the ability to
distinguish a specific band in the mixture in which the
target DNA of the species comprised 1% of the total mix-
ture. Therefore species in minority in the microbial popu-
lations are not detectable or difficult to detect. This may
explain why after cloning of the extract of sample I.1, I.2
and I.3 no or a limited amount of moderately acidophilic
sulfur-oxidizing bacteria were retrieved. To resolve this
problem, it may be interesting to make use of more spe-
cific primers for PCR amplification of the obtained DNA
fractions. Such basic information can be used further in
the development or application of new techniques such as
microscopy with fluorescent in situ hybridization (FISH)
using probes based on variable regions of 16S rRNA of
specific acidophilic bacteria. For example, such an oli-
gonucleotide probe for T. thiooxidans was developed
(Peccia et al. 2000). However, to get a complete image of
the microbial community in an environmental sample of a
corroded concrete pipe with FISH probes, foregoing fun-
damental studies and identification of the present bacteria
is necessary.
With conventional counting techniques, Davis et al.
(1998) detected large numbers of aerobic heterotrophs in
the outer corrosion layer as well as the inner corrosion
layer of concrete sewer pipes. Results of sequence com-
parison of the obtained clones with the Genbank dat-
abase showed the presence of different heterotrophs. The
presence of these bacteria can be explained by: (1) the
providence of vital nutrients for the sulfur-oxidizing bac-
teria, or (2) the removal capacity of toxic substances as
pyruvate and oxalate, as T. thiooxidans excretes pyruvate
and oxalate, which are self-inhibitory at concentrations of
2×10–5 to 7×10–5 M (Cho and Mori 1995; Borichewsky
1967; Peccia et al. 2000), or (3) the production of organ-
783
ic acids that with calcium may form soluble calcium
complexes. Our results clearly show the presence of non-
sulfur-oxidizing heterotropic bacteria. These bacteria ob-
viously grow together with the sulfur-oxidizing bacteria.
It is clearly seen that a limited amount of some specific
bacteria occur at the heavily corroded concrete surfaces
characterized by a low pH. Most of the bacteria found in
the presence of the sulfur-oxidizing bacteria are chemo-
organotrophs and are characterized by polysaccharide
slime production, acid-fastness or a special cell wall
composition. The resistance of Mycobacteria sp. to acids
is considered to be due to the high levels of mycolic acid
in the cell wall, which make the cells of mycobacteria
wax-like and strongly hydrophobic (Schlegel 1993).
Data of Gu et al. (1998) indicated that a Fusarium sp.
is capable of degrading the concrete material. Possible
explanation may be the formation of soluble calcium
complexes with the concrete, resulting in dissolution (Gu
et al. 1998). Our results showed also the presence of fun-
gi; however, they were not further analyzed or identified.
Near the steel bar, more bacterial species were pres-
ent. This could be explained by: (1) higher pH values
and/or (2) the presence of steel bar and subsequently the
presence of micro-organisms involved in corrosion of
steel.
The first series of samples showed that a large differ-
ence in community composition exists between the cor-
roded samples and the biofilm under the water level. The
microbial analysis of the second series of samples con-
firmed these results. Additionally, it was interesting to
see that this difference in microbial diversity and pH of
the concrete surface gradually declined as the distance
from the connection increases and the corrosion damage
decreased. The DGGE patterns of the three samples tak-
en 200 m after the connection showed a considerable
similarity and indicate that no major differences in bacte-
rial species occur under and above the water level. How-
ever bands of different intensity appear, and more de-
tailed analysis is necessary to get a complete image of
the microbial communities present in sewer pipes.
In addition, AHL production was detected in the se-
verely corroded concrete samples of the crown of the
sewer pipe. Results obtained in this work suggest the in-
tensity of the AHL production; however, for quantitative
results, the different molecules should be separated, for
example by thin-layer chromatography (Shaw et al.
1997). The detection of AHL molecules on severely cor-
roded concrete surfaces in sewer pipes offers a potential
opportunity for corrosion control. Different compounds
were described that interfere not only with intraspecies
cell-cell communication but also with interspecies com-
munication in the process of colonization of bacterial
communities. For example, halogenated furanones struc-
turally resemble AHL molecules and show inhibitory ac-
tivity at ecologically realistic concentrations in AHL bio-
assays (Manefield et al. 1999).
AHL activity was also detected in the biofilm samples
taken under the water level. These results can lead to fur-
ther insight into the occurrence of signal molecules and
784
an understanding of the fundamental mechanisms of bio-
film formation in environmental samples. The presence
of naturally occurring AHL production has already been
shown in aquatic biofilms growing on submerged stones
(McLean et al. 1997). This is, to our knowledge, the first
report showing AHL production in sewage and microbi-
ologically corroded concrete.
In conclusion, molecular tools (PCR, DGGE and
cloning of 16S rRNA gene fragments), applied directly
to environmental samples containing large amounts of
calcium and corrosion products characterized by a low
pH, were successfully used. The results suggest that es-
pecially the acidophilic sulfur-oxidizing bacteria (Acidi-
thiobacillus thiooxidans) are active at low pH on heavily
corroded concrete surfaces, in association with some het-
erotrophs (e.g. Mycobacterium sp.). Interestingly, AHL
molecules were detected in these samples; however, their
function in the corrosion process is still unclear.
Acknowledgements This research was supported by the Fund
for Scientific Research Flanders (FWO-Vlaanderen, project no.
G.0279.98N). The authors thank Eva Top for the useful ideas and
Siska Maertens for her practical help with the molecular tech-
niques. We also thank Aquafin NV for the technical support dur-
ing sampling and Geert Rombaut, Frederik Hammes, Sarah Philips
and Kris Van Hege for useful comments on the manuscript.
References
Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller
W, Lipman DJ (1997) Gapped BLAST and PSI-BLAST: a new
generation of protein database search programs. Nucleic Acids
Res 25: 3389–3402
Amann RI, Ludwig W, Schleifer K-H (1995) Phylogenetic identi-
fication and in situ detection of individual microbial cells
without cultivation. Microbiol Rev 59: 143–169
Atlas RM (1993) Malt extract agar. In: Parks LC (ed) Handbook
of microbiological media. CRC, London, p. 538
Boon N, Goris J, De Vos P, Verstraete W, Top E (2000) Bioaug-
mentation of activated sludge by an indigenous 3-chloroani-
line degrading Comamonas testosteroni strain I2gfp. Appl En-
viron Microbiol 66: 2906–2913
Borichewsky RM (1967) Keto acids as growth limiting factors in
autotrophic bacteria that share their habitat. Annu Rev Micro-
biol 38: 265–292
Cha C, Gao P, Chen Y-C, Shaw PD, Farrand SK (1998) Produc-
tion of N-acyl-homoserine lactone quorum-sensing signals by
gram-negative plant-associated bacteria. Mol Plant-Microbe
Interact 11(11): 1119–1129
Cho K-S, Mori T (1995) A newly isolated fungus participates in
the corrosion of concrete sewer pipes. Water Sci Technol 31:
263–271
Davis JL, Nica D, Shields K, Roberts DJ (1998) Analysis of con-
crete from corroded sewer pipe. Int Biodeter Biodegr 42(1):
75–84
DeHollander GR (1998) Gaseous emissions from wastewater fa-
cilities. Water Environ. Res 70: 580–585
El Fantroussi SL, Verschuere L, Verstraete W, Top EM (1999) Ef-
fect of phenylurea herbicides on soil microbial communities
estimated by analysis of 16S rRNA gene fingerprints and com-
munity-level physiological profiles. Appl Environ Microbiol
65: 982–988
Fuqua C, Winans SC (1996) Conserved cis-acting promoter ele-
ments are required for density-dependent transcription of
Agrobacterium tumefaciens conjugal transfer genes. J Bacte-
riol 178(2): 435–440
Greenberg EP (1997) Quorum sensing in gram-negative bacteria.
ASM News 63: 371–377
Gu J-D, Ford TE, Berke NS, Mitchell R (1998) Biodeterioration
of concrete by the fungus Fusarium. Int Biodeter Biodegr 41:
101–109
Hazeu W, Battenburg WH, Bos P, van der Pas RK, Kuenen JG
(1988) The production and utilization of intermediary elemen-
tal sulphur during the oxidation of reduced sulphur com-
pounds by Thiobacillus ferrooxidans. Arch Microbiol 150:
574–579
Islander RL, Devinny JS, Mansfeld F, Postyn A, Shih H (1991)
Microbial ecology of crown corrosion in sewers. J Environ
Eng 117: 751–770
Kelly DP (1982) Biochemistry of the chemolithotrophic oxidation
of inorganic sulfur. Phil Trans R Soc Lond Ser B 298:
499–528
Kelly DP, Wood AP (2000) Reclassification of some species of
Thiobacillus to the newly designated genera Acidithiobacillus
gen. nov. Halothiobacillus gen. nov. and Thermithiobacillus
gen.nov. Int J Syst Evol Micr 50: 511–516
Manefield M, de Nys R, Kumar N, Read R, Givskov M, Steinberg
P, Kjelleberg S (1999) Evidence that halogenated furanones
from Delisea pulchra inhibit acylated homoserine lactone
(AHL)-mediated gene expression by displacing the AHL sig-
nal from its receptor protein. Microbiology 145: 283–291
Mansfeld F, Shih H, Postyn A, Devinny J, Islander R, Chen CL
(1990) Corrosion monitoring and control in concrete sewer
pipes. Corrosion 90, Paper No 113, National Association of
Corrosion Engineers, Houston, Texas, USA
McClean KH, Winson MK, Fish L, Taylor A, Chabra SR, Camara
M, Daykin M, Lamb JH, Swift S, Bycroft BW, Stewart GSAB,
Williams P (1997) Quorum sensing and Chromobacterium vio-
laceum: exploitation of violacein production and inhibition for
the detection of N-acylhomoserine lactones. Microbiology
143: 3703–3711
McLean RJC, Whiteley M, Stickler DJ, Fuqua WC (1997) Evi-
dence of autoinducer activity in naturally occurring biofilms.
FEMS Microbiol Lett 154: 259–263
Milde K, Sand W, Wolff W, Bock E (1983) Thiobacilli of the cor-
roded concrete walls of the Hamburg sewer system. J Gen Mi-
crobiol 129: 1327–1333
Mori T, Koga M, Hikosaka Y, Nonaka T, Mishina F, Sakai Y,
Koizumi J (1991) Microbial corrosion of concrete sewer
pipes, H2S production form sediments and determination of
corrosion rate. Water Sci Technol 23(7–9): 1275–1282
Muyzer G (1999) DGGE/TGGE a method for identifying genes
from natural ecosystems. Curr Opin Microbiol 2: 317–322
Muyzer G, de Waal EC, Uitterlinden A (1993) Profiling of complex
microbial populations using denaturing gradient gel electropho-
resis analysis of polymerase chain reaction-amplified genes
coding for 16S rRNA. Appl Environ Microbiol 59: 695–700
Padival NA, Weiss JS, Arnold RG (1995) Control of Thiobacillus
by means of microbial competition: implications for corrosion
of concrete sewers. Water Environ Res 67: 201–205
Peccia J, Marchand EA, Silverstein J, Hernandez M (2000) Devel-
opment and application of small-subunit rRNA probes for as-
sessment of selected Thiobacillus species and members of the
genus Acidophilium. Appl Environ Microbiol 66: 3065–3072
Sand W, Bock E (1984) Concrete corrosion in the Hamburg sewer
system. Envir Tech Lett 5: 517–528
Sand W, Bock E, White DC (1987) Biotest system for rapid evalu-
ation of concrete resistance to sulfur-oxidizing bacteria. Mater
Performance 26: 14–17
Schlegel HG (1993) General microbiology, 2nd edn. Cambridge
University Press, Cambridge, pp 102–104
Schmidt M, Hormann K, Hofmann F-J, Wagner E (1997) Beton
mit erhöhtem Widerstand gegen Säure und Biogene Schwefel-
säurekorrosion. Concrete Precasting Plant Technol 4: 64–70
Shaw PD, Ping G, Daly SL, Cha C, Cronan JE, Rinehart KL
(1997) Detecting and characterizing N-acyl-homoserine lac-
tone signal molecules by thin-layer chromatography. Proc Natl
Acad Sci USA 94: 6036–6041

Stickler D (1999) Biofilms. Curr Opin Microbiol 2(3): 270–275
Tazawa E-I, Morinaga T, Kawai K (1996) The deterioration of
concrete in sewerworks caused by metabolites of aerobic mi-
croorganisms, and preventive measures. L'Industria Italiana
del Cemento 11: 792–780
Trüper HG (1984) Microorganisms and the sulphur cycle. In:
Müller A, Krebs B (eds) Sulphur, its Significance for Chemis-
try, for the Geo-, Bio-, and Cosmosphere and Technology.
Studies in inorganic chemistry. Elsevier, Amsterdam, pp
351–365
Vincke E, Verstichel S, Monteny J, Verstraete W (1999) A new
test procedure for biogenic sulfuric acid corrosion of concrete.
Biodegradation 10: 421–428
785
Vincke E, Monteny J, Beeldens A, De Belie N, Taerwe L, Van
Gemert D, Verstraete W (2000) Recent developments in the
research on biogenic sulfuric acid attack of concrete. In: Lens
P, Hulshoff Pull L (eds) Biotechnological treatment of sulphur
pollution. IWA, London, pp 515–541
Ward JH (1963) Hierarchical grouping to optimize an objective
function. J Am Stat Assoc 58: 236–244
Weidner S, Arnold W, Pühler A (1996) Diversity of uncultured
microorganisms associated with the seagrass Halophila
stipulacea estimated by restriction fragment length polymor-
phism analysis of PCR-amplified 16S rRNA genes. Appl En-
viron Microbiol 62: 766–771
Wilson IG (1997) Inhibition and facilitation of nucleic acid ampli-
fication. Appl Environ Microbiol 63: 3741–3751