APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Aug. 2003, p. 4901–4909
Copyright © 2003, American Society for Microbiology. All Rights Reserved.
Vol. 69, No. 8
Strain-Specific Ureolytic Microbial Calcium Carbonate Precipitation
Frederik Hammes,1† Nico Boon,1Johan de Villiers,2Willy Verstraete,1*
and Steven Douglas Siciliano1‡
Laboratory of Microbial Ecology and Technology (LabMET), Ghent University, B-9000 Ghent, Belgium,1
and Department of Earth Sciences, University of Pretoria, 0001 Pretoria, South Africa2
Received 5 December 2002/Accepted 2 May 2003
During a study of ureolytic microbial calcium carbonate (CaCO3) precipitation by bacterial isolates collected
from different environmental samples, morphological differences were observed in the large CaCO3crystal
aggregates precipitated within bacterial colonies grown on agar. Based on these differences, 12 isolates were
selected for further study. We hypothesized that the striking differences in crystal morphology were the result
of different microbial species or, alternatively, differences in the functional attributes of the isolates selected.
Sequencing of 16S rRNA genes showed that all of the isolates were phylogenetically closely related to the
Bacillus sphaericus group. Urease gene diversity among the isolates was examined by using a novel application
of PCR-denaturing gradient gel electrophoresis (DGGE). This approach revealed significant differences be-
tween the isolates. Moreover, for several isolates, multiple bands appeared on the DGGE gels, suggesting the
apparent presence of different urease genes in these isolates. The substrate affinities (Km) and maximum
hydrolysis rates (Vmax) of crude enzyme extracts differed considerably for the different strains. For certain
isolates, the urease activity increased up to 10-fold in the presence of 30 mM calcium, and apparently this
contributed to the characteristic crystal formation by these isolates. We show that strain-specific calcification
occurred during ureolytic microbial carbonate precipitation. The specificity was mainly due to differences in
urease expression and the response to calcium.
Numerous diverse microbial species participate in the pre-
cipitation of mineral carbonates in various natural environ-
ments, including soils, geological formations, freshwater bio-
films, oceans, and saline lakes (9, 28, 30, 31). In some instances
the resulting natural precipitates represent the fossil records or
fingerprints of bacterial activity (28, 38, 40). For example, this
process was at the heart of the original debate about whether
mineral sediments on the ALH84001 Martian meteorite rep-
resent proof of life in outer space (24, 39, 40). In addition to its
natural historical importance, microbial carbonate precipita-
tion (MCP) is also the basis of several innovative applications
in both the building sector and the wastewater industry (13, 15,
The precise role of microbes in the carbonate precipitation
process is still not clear. Boquet et al. (5) suggested that almost
all bacteria are capable of CaCO3precipitation. Knorre and
Krumbein (20) concluded that MCP occurs as a by-product of
common microbial metabolic processes, such as photosyn-
thesis, urea hydrolysis, and sulfate reduction. These meta-
bolic processes increase the alkalinity (increase the pH and
dissolved inorganic carbon content) of the environment and
thereby favor CaCO3precipitation (8, 9). Alternatively, it is
possible that there are specific attributes of certain bacteria
that promote and affect CaCO3precipitation. The negatively
charged nature and specific functional groups of microbial cell
walls favor the binding of divalent cations (Ca2?, Mg2?), there-
by making microorganisms ideal crystal nucleation sites (30,
34). Microbial extracellular polymeric substances are also an
important factor in precipitation, either through trapping and
concentration of calcium ions or as a result of specific proteins
that influence precipitation (19). Kawaguchi and Decho (19)
suggested that specific proteins present in biological extracel-
lular polymeric substances cause the formation of different
CaCO3polymorphs. A third hypothesis combines the common
metabolism and strain specificity hypotheses to suggest that
CaCO3precipitation, possibly influenced by intracellular cal-
cium metabolism (8, 17, 23), might play a role in the ecology of
the precipitating organism (2, 23).
Enzymatic hydrolysis of urea presents a straightforward
model for studying microbial CaCO3precipitation. The urease
enzyme (urea amidohydrolase; EC 188.8.131.52) is common in a
wide variety of microorganisms, can be readily induced by
adding an inexpensive substrate, and is involved in several
biotechnological applications (12, 16, 37). One mole of urea is
hydrolyzed intracellularly to 1 mol of ammonia and 1 mol of
carbamate (equation 1), which spontaneously hydrolyzes to
form an additional 1 mol of ammonia and carbonic acid (equa-
tion 2) (7). These products subsequently equilibrate in water to
form bicarbonate and 2 mol of ammonium and hydroxide ions
(equations 3 and 4). The latter give rise to a pH increase, which
in turn can shift the bicarbonate equilibrium, resulting in the
formation of carbonate ions (equation 5), which in the pres-
ence of soluble calcium ions precipitate as CaCO3(equation 6)
CO(NH2)2? H2O 3 NH2COOH ? NH3
NH2COOH ? H2O 3 NH3? H2CO3
* Corresponding author. Mailing address: Laboratory of Microbial
Ecology and Technology (LabMET), Ghent University, Coupure
Links 653, B-9000 Ghent, Belgium. Phone: 32-(0)9 264 59 76. Fax: 32-
(0)9 264 62 48. E-mail: Willy.Verstraete@rug.ac.be.
† Present address: Department of Microbiology and Molecular Eco-
toxicology, Swiss Federal Institute for Environmental Science and
Technology, CH-8600 Duebendorf, Switzerland.
‡ Present address: Department of Soil Science, University of Sas-
katchewan, Saskatoon, Saskatchewan S7N 5A8, Canada.
?? H?(pKa2? 6.37)(3)
2NH3? 2H2O 7 2NH4
?? H?? 2NH4
?? 2OH?7 CO3
2?? Ca2?7 CaCO3(Kso? 3.8 ? 10?9) (6)
During laboratory work over the past few years, we found
that different bacterial isolates precipitated visually different
crystal aggregates in a repeatable manner when they were
cultivated on semisolid media under similar growth condi-
tions. The aim of this work was to characterize some of the
primary mechanisms responsible for this diversity in crystal
MATERIALS AND METHODS
Isolation. CaCO3-precipitating strains were isolated from two soils (garden
soil and landfill soil from Ghent, Belgium), from a freshly cut concrete surface
(Portland, type CEM I 42.5 R), and from calcareous sludge from a biocatalytic
calcification reactor. The latter treats calcium-rich industrial wastewater through
ureolytic microbial carbonate precipitation (15, 16). The natural samples were
suspended in a sterile physiological solution (8.5 g of NaCl liter?1), diluted
appropriately, and plated on precipitation agar containing (per liter) 3 g of
nutrient broth (Oxoid), 20 g of urea (Riedel-de Hae ¨n), 2.12 g of NaHCO3
(Sigma), 10 g of NH4Cl (Sigma), and 30 mM CaCl2? 2H2O (Sigma) (12, 37).
Incubation was at 28°C. Colonies were assessed every 5 days with a stereomi-
croscope (Wild, Heerbrugg, Belgium), and positive colonies were selected based
on visual crystal formation within 10 days. Positive isolates were purified by
repetitive dilution and plating (as described above). The purified strains were
deposited in the BCCM/LMG Culture Collection (Ghent, Belgium) as strains
LMG21776 to LMG21787.
Microscopy and phase identification. Crystal-precipitating colonies were ex-
amined after 5 and 10 days of cultivation by light microscopy with an Axioskop
II Plus microscope (Zeiss) and by stereomicroscopy (Wild). Digital images were
captured with a 1-CCD camera (Hamamatsu Photonics GmbH, Herssching,
Germany). A total of 12 isolates were selected based on visual differences in
precipitate morphology. Large crystal aggregates that precipitated within single
colonies of the isolates were subsequently harvested from the agar surface,
washed in sterile water, and dried (28°C, 3 days). The dried aggregates were
ground to the appropriate particle size for X-ray diffraction (XRD) analysis
(diameter, ?10 ?m) with a McCrone micronizing mill and then analyzed by
using a Siemens D-501 diffractometer equipped with an Ni filter and a CuK?
radiation source. Phase identification was done after background subtraction by
using the Bruker EVA software.
Urease activity and location of urease activity. All the isolates were tested for
urease activity, as well as the location of urease activity. This was done by
streaking the purified cultures on urease test agar (BBL, Becton Dickinson and
Company, Sparks, Md.) and inoculating urease test broth (as described above)
with viable liquid cultures, as well as filtrates (pore size, 0.22 ?m; Millipore) of
the liquid cultures. A change in color following incubation for 5 days at 28°C was
recorded as a urease-positive reaction.
PCR amplification of 16S rRNA genes. A DNA template for PCR amplifica-
tion from pure cultures was obtained by extracting total genomic DNA according
to the manufacturer’s instructions (Wizard genomic DNA purification kit; Pro-
mega, Leiden, The Netherlands). The PCR master mixture contained each
primer at a concentration of 500 nM, each deoxynucleoside triphosphate at a
concentration of 200 ?M, 1.5 mM MgCl2, 10 ?l of thermophilic DNA polymer-
ase 10? reaction buffer (MgCl2free), 2.5 U of Taq DNA polymerase (Promega,
Madison, Wis.), 400 ng of bovine serum albumin (Boehringer) per ?l, and
enough DNase- and RNase-free filter-sterilized water (Sigma-Aldrich Chemie,
Steinheim, Germany) so that the final volume was 100 ?l. One microliter of
DNA template was added to 24 ?l of the master mixture. The 16S rRNA gene
fragments were obtained by amplifying the 16S rRNA gene with primers P63f
(5?-CAGGCCTAACACATGCAAGTC-3?) and P1378r (5?-CGGTGTGTACA
AGGCCCGGGAACG-3?). PCR was performed in a 9600 thermal cycler (Per-
kin-Elmer, Norwalk, Conn.) with a program consisting of 10 min at 95°C, fol-
lowed by 30 cycles of 1 min at 94°C, 1 min at 53°C, and 2 min at 72°C and a final
elongation step for 10 min at 72°C.
PCR-DGGE of the ureC gene. For PCR-denaturing gradient gel electrophore-
sis (DGGE) analysis of the ureC gene the DNA extraction procedure and the
PCR master mixture composition were identical to those described above. Ba-
cillus pasteurii ATCC 6453 was included as a positive control for ureolytic mi-
crobial CaCO3precipitation (12, 37). To amplify the genes coding for the ureC
subunit of the urease enzyme, PCR was performed with primers UreC-F (5?-T
GGGCCTTAAAATHCAYGARGAYTGGG-3?) and UreC-R (5?-GGTGGTG
GCACACCATNANCATRTC-3?) as previously described by Reed (29). The
length of the expected amplified fragment was 382 bp. To examine the diversity
of the partial ureC DNA fragments by DGGE, a 40-bp GC clamp (5?-CGCCC
GCCGCGCGCGGCGGGCGGGGCGGGGGCACGGGGGG-3?) (27) was at-
tached to the 5? end of the UreC-F primer. The PCR conditions were as follows:
94°C for 5 min, followed by 35 cycles of 92°C for 1 min, 50°C for 1 min, and 72°C
for 2 min. A final extension was carried out at 72°C for 10 min. DGGE was
performed with the Bio-Rad D gene system (Bio-Rad, Hercules, Calif.). PCR
samples were loaded onto 7% (wt/vol) polyacrylamide gels in 1? TAE (20 mM
Tris, 10 mM acetate, 0.5 mM EDTA; pH 7.4). The polyacrylamide gels were
made with a 40 to 60% denaturing agent gradient; 7 M urea and 40% formamide
were defined as 100% denaturing agent. Electrophoresis was performed for 16 h
at 60°C and 45 V. The resulting gels were stained with SYBR Green I nucleic
acid gel stain (1:10,000 dilution; FMC BioProducts, Rockland, Maine) and pho-
tographed (3). Processing of the DGGE gels was done with the Bionumerics 2.0
software (Applied Maths, Kortrijk, Belgium). Calculation of similarities was
based on the Dice correlation coefficient and the results in a distance matrix. The
unweighted pair group with mathematical average clustering algorithm was used
to calculate dendrograms for the DGGE gels.
DNA sequencing and sequence analysis. DNA sequencing of the PCR frag-
ments was carried out by ITT Biotech-Bioservice (Bielefeld, Germany). Analysis
of DNA sequences and homology searches were completed by using standard
DNA sequencing programs and the BLAST server of the National Center for
Biotechnology Information (www.ncbi.nlm.nih.gov) with the BLAST algorithm
(1) and specifically with the blastn program for comparison of a nucleotide query
sequence with a nucleotide sequence database. A phylogenetic tree was con-
structed by using the RDP Phylip 3.5c interface (22). Distance matrix analyses
were done with the Jukes-Cantor (18) correction, and tree construction was done
by the neighbor-joining method (33). Putative ureC gene fragments were excised
from a DGGE gel, reamplified, and sent out for sequencing (see above). The
resulting nucleotide sequences were translated into protein sequences by using
the blastx software (BLAST) (National Center for Biotechnology Information),
and these sequences were compared with a protein sequence database.
Protein extraction. Isolates were cultivated in 1 liter of nutrient broth (Oxoid)
containing 20 g of urea liter?1for 48 h at 30°C with shaking (100 rpm). Bacterial
cells were harvested by centrifugation (8,000 ? g, 30 min) and washing of the
pellet in extraction buffer (100 mM NaH2PO4, 1 mM EDTA; pH 7.0), followed
by centrifugation (15,000 ? g, 10 min) and resuspension in extraction buffer (see
above). Crude enzyme was extracted by bead beating (three 90-s pulses with 10-s
pauses between pulses). The protein concentration was determined by the pro-
cedure of Bradford (6). This included adding 200 ?l of a commercial Bio-Rad
protein assay solution to 800 ?l of an appropriately diluted sample and measur-
ing the color development spectrophotometrically at 595 nm. Bovine serum
albumin was used as the standard.
Urease activity assay. Urease activity was assayed in 1 ml of buffer containing
100 mM NaH2PO4and 1 mM EDTA (pH 8.0). Ten different substrate (urea)
concentrations between 5 and 100 mM were used. Both the crude enzyme and
the reaction mixtures were incubated for 5 min at 25°C prior to the urease assay.
The reaction was initiated by addition of the crude enzyme, and urease activity
was determined by measuring the amount of total ammonium nitrogen released
after 10 min spectrophotometrically (425 nm) by using the Nessler assay method
(14). One unit of urease activity was defined as the amount of enzyme that
hydrolyzed 1 ?mol of urea per min. Michaelis-Menten kinetic constants (Kmand
Vmax) were estimated by graphing the data in a Lineweaver-Burk plot. In another
assay to determine the effect of pH on the relative urease activity, the procedure
described above was repeated with 50 mM urea solutions at pH 7 and 8 (in 100
mM phosphate buffer) and pH 9 (in 100 mM Tris-HCl buffer). In both assays,
commercial urease from Jack beans (type III from Jack beans; Sigma) was used
as a positive control.
Effect of calcium on urease activity. The isolates were cultivated and concen-
trated as described above. Sterile solutions containing urea (100 mM), nutrient
broth (1 g ? liter?1), and NaHCO3(10 mM) at pH 7 (1 N HNO3) were prepared
and inoculated along with the isolates to obtain final concentrations of 107to 108
CFU ? ml?1. Similar solutions that also contained 30 mM CaCl2were also inoc-
ulated. These solutions were all incubated at 25°C with stirring and were sampled
every 30 min. Total ammonium nitrogen was measured as described above, and
4902HAMMES ET AL.APPL. ENVIRON. MICROBIOL.
the results were expressed as the urease activity in the presence of calcium
divided by the urease activity in the absence of calcium. All experiments were
done in triplicate, and all measurements were obtained in duplicate.
PCA, multivariate analysis of variance, and discriminant analysis. Principal-
component analysis (PCA) was carried out by using Bionumerics software and
incorporating as parameters for each isolate (i) the number of urease bands from
the DGGE, (ii) the Ca2?/urea ratio, (iii) the Kmdata, (iv) the Vmaxdata, and (v)
the calcium-urease activity data. This was done to establish a statistical correla-
tion between the various parameters and the morphological differences observed
in the crystal aggregates. Multivariate analysis of variance and discriminant
analysis were also used to identify the parameter(s) that was the primary con-
tributing parameter(s) to these morphological differences. The following three
different groups were defined based on the PCA: (i) strains CPB 1 to CPB 4, (ii)
strains CPB 5 and CPB 6, and (iii) strains CPB 7 to CPB 12. Accounting for
covariance structure and relative character importance were both used in the
Nucleotide sequence accession numbers. The almost complete sequences (800
to 1,300 kb) of 16S rRNA genes of all isolates have been deposited in the
GenBank database under accession no. AF548874 to AF548885. Six nucleotide
sequences (?300 kb) of urease genes have also been deposited in the GenBank
database under accession no. AY178982 to AY178987.
Isolation, crystallization, and initial urease analysis. About
10% of all bacterial colonies isolated from various sources
induced crystallization on the precipitation agar. Precipitation
started with a darkening in the center of the bacterial colony,
which was attributed to amorphous CaCO3formation (after
between 20 h and 5 days, depending on the isolate), and this
was followed by crystallization and crystal maturation with
time (Fig. 1). The precipitate always formed within the bacte-
rial colony on the agar surface, which also captured bacteria
within the crystal structure. Twelve isolates were selected for
further study based on morphological differences observed in
the large crystal aggregates which were precipitated in single
colonies on precipitation agar (Fig. 2). The isolates were
termed calcium-precipitating bacteria (strains CPB 1 to CPB
12). Four basic morphologically distinct groups of crystal ag-
gregates were initially distinguished. CPB 1 to CPB 4 all pro-
duced large, light-brown structures which formed rapidly (20 to
48 h for crystallization), and aggregates accounted for up to
98% of the total colony surface (Fig. 2). CPB 5 and CPB 6
precipitated at similar rates, but in this case the result was
distinctly sharp, whitish and transparent crystal aggregates.
CPB 7 to CPB 10 precipitated at noticeably lower rates than
the first strains (about 3 to 5 days for crystallization). The
resulting precipitates were thick, brown aggregates, with an
almost sponge-like appearance. CPB 11 and CPB 12 required
between 5 and 10 days for crystallization and were character-
ized by initial formation of separate, small, spherical crystals
that coalesced only after a couple of days. Note that the dif-
ferent crystal groups reflected neither the initial origin of the
isolates (indicated in Fig. 2) nor the cell morphology, since all
isolates formed rod-shaped cells that were between 1 and 3 ?m
long. XRD analysis of the crystals revealed that the primary
crystal component of all the classes was rhombohedral calcite,
although hexagonal vaterite was detected in some cases as well.
For all isolates, urease activity was cell associated. All 12 iso-
lates produced a urease-positive color when they were grown
on urease test agar or when they were inoculated into urease
test broth. No filtrate of a liquid culture gave a positive reac-
Identification and sequencing. All of the isolates were
closely related to one another and to cultured bacteria belong-
ing to the Bacillus sphaericus group. BLAST results suggested
that the closest relatives of this group are Bacillus pasteurii
(CPB 1 to CPB 4), Bacillus psychrophilus (CPB 5 to CPB 7),
Planococcus okeanokoites (CPB 8), and Bacillus globisporus
(CPB 9 to CPB 12), while another species that is closely re-
lated to all of the isolates is Filibacter limicola. A phylogenetic
(neighbor-joining) tree was constructed by using Escherichia
coli and B. subtilis as outgroups (Fig. 3). Although all of the
CPB isolates were related to the B. sphaericus group, there was
still sufficient diversity in the 16S rRNA genes to place the
isolates into different clusters. The largest cluster contained
CPB 1 to CPB 4, CPB 7, and CPB 8 and was closely related to
B. pasteurii. The second cluster contained CPB 11 and CPB 12,
and the third cluster contained CPB 10. CPB 9 was situated
between the second and third clusters. The fourth cluster con-
tained CPB 5 and CPB 6 and was noticeably distantly related
to the other clusters.
Urease gene diversity. PCR-DGGE, performed with UreC
primers, revealed that the positions and numbers of bands
were different for different isolates (Fig. 4), and a cluster anal-
ysis, based on band positions, showed that there were two large
groups, as well as four isolates that were not in any group (CPB
7, CPB 8, CPB 10, and CPB 12). The first group included CPB
1 and CPB 2 (100% similarity), CPB 11, CPB 3, and CPB 4.
The second group included CPB 5 and CPB 6 (100% similar-
ity) and CPB 9, as well as B. pasteurii. All of the bands were
excised from the DGGE gel, but only six bands (as indicated in
Fig. 3) were successfully purified, amplified, and sequenced. In
all cases, amino acid sequences were related to the alpha sub-
unit (UreC) of previously described ureases, thus indicating
that the PCR amplification was indeed specific. The highest
levels of similarity were found with sequences from Bacillus
FIG. 1. Typical ureolytic CaCO3precipitation sequence, starting with the formation of amorphous CaCO3, followed by crystallization and
VOL. 69, 2003MICROBIAL CaCO3PRECIPITATION4903
FIG. 2. Morphological differences in calcite precipitates within bacterial colonies of ureolytic calcium-precipitating bacteria grown on semisolid
media. The origins of isolates are indicated in parentheses.
4904 HAMMES ET AL.APPL. ENVIRON. MICROBIOL.
(NP_599339.1), B. pasteurii (S_47104), and Haemophilus influ-
enzae (NP_438697.1). The levels of similarity between the iso-
late sequences and these corresponding sequences ranged
from 66 to 90%, and the levels of similarity between the isolate
sequences ranged from 81 to 95%.
Urease activity. There were some distinct differences among
the isolates when both the substrate affinity (Km) and rate
(Vmax) values were compared (Fig. 5). Isolates CPB 1 to CPB
6 showed high urea affinities, while the affinities of CPB 7 to
CPB 12 were about three times lower. Isolates CPB 5 and CPB
6 displayed noticeably high Vmaxvalues, while the Vmaxvalues
of CPB 7 and CPB 11 were distinctly lower than the other Vmax
values. For all the isolates, the relative activity of the crude
enzyme extract was between 40 and 60% lower at pH 7 than at
pH 8, while pH 9 had no significant effect at all (results not
Effect of calcium on urease activity. Figure 6 shows a com-
parison between the urease activities of the isolates with and
without calcium added to the medium. CPB 1 to CPB 4 dis-
played large increases in urease activity (4- to 10-fold) in the
presence of calcium ions compared to the activity in the ab-
sence of calcium. For the other isolates there were either no
differences or calcium had slightly adverse effects on the urease
PCA. PCA confirmed that at least three crystal groups could
be explained by the various urease data, thereby suggesting
that the differences in precipitation could largely be ascribed to
diversity in the urease gene fragments of the isolates, with
resulting biochemical differences in the urease activities of the
isolates (Fig. 7). The first three principal components (PC)
explained 99.7% of the total variation (PC 1, 79.4%; PC 2,
18.9%; PC 3, 1.4%). The PCA, however, did not separate the
different crystal aggregate groups completely; most notably,
CPB 11 and CPB 12 clustered together with CPB 7 to CPB 10.
Based on the PCA, three groups were included in the discrimi-
nant analysis. The results showed that for the first discriminant
(accounting for 91% of the discrimination) in descending order
of importance, the substrate affinity (Km) made a positive con-
tribution, while the maximum hydrolysis rate (Vmax) and the
number of bands both made negative contributions (P ?
The aim of this study was to investigate the primary mech-
anisms that cause apparent differences in crystal aggregate
formation during ureolytic MCP. Such information should be
of particular interest to workers investigating carbonate pre-
cipitation as a bacterial fossil record and to workers investigat-
ing carbonate precipitation as a potential biotechnological
technique. MCP is an established tool for the in situ restora-
tion of buildings and monuments (8, 37) and is an emerging
tool in the treatment of calcium-rich industrial wastewater (15,
16). For both these applications, improved knowledge of the
precise role of bacteria in the precipitation process, specifically
the mechanisms governing precipitation rates and types, is
vitally important (8). For example, recent biotechnology de-
signed to remove Ca2?from industrial wastewater specifically
stimulates and accelerates urease activity in activated sludge,
FIG. 3. Neighbor-joining tree, based upon partial 16S rRNA gene
sequences of the 12 isolates (CPB 1 to CPB 12) and their closest
relatives. E. coli was used as an outgroup. A sequence analysis was
done as described in the text, and 402 base positions were included in
FIG. 4. DGGE separation of selectively amplified urease PCR products from the different isolates, with B. pasteurii as a positive control. The
tree was constructed by calculating the Dice coefficient and clustering with an unweighted pair group with mathematical average algorithm. The
white vertical lines represent bands, and individual bands indicated by arrows were excised and sequenced, as described in the text.
VOL. 69, 2003 MICROBIAL CaCO3PRECIPITATION 4905
resulting in removal of 99% of the calcium. Interestingly, dur-
ing adaptation of this reactor community, the bacterial com-
munity changed, but urease activity was apparently dominated
by only two ureases (15).
Precipitation within microbial colonies on agar presents a
unique opportunity to study MCP within a specific localized
microenvironment created by the microorganisms (Fig. 1). In
previous studies workers have used XRD data to suggest that
species-specific microbial precipitation occurs (19, 31), but
other workers contend that species-specific differences are due
primarily to environmental rather than microbial factors (20).
Although morphological differences in crystallization were ev-
ident (Fig. 2), XRD analysis showed that in all cases rhombo-
hedral calcite was the primary component, with vaterite de-
tected only in some cases. The latter finding is interesting, as
vaterite is metastable at the normal temperature and atmo-
spheric pressure, and it has been suggested that the metastable
polymorphs form initially and subsequently convert to a stable
polymorph (e.g., calcite) (39). Enzymatic precipitation exper-
iments performed with liquid solutions of urea and CaCl2at
room temperature revealed that the sequence of precipitation
is as follows: first amorphous CaCO3, then vaterite, and finally
calcite (35). Although calcite is commonly precipitated during
ureolytic carbonate precipitation (37), microbes precipitate
other polymorphs, such as aragonite (19, 31). The fact that all
the samples produced very similar XRD results, while clear
morphological differences were evident, suggests that the dif-
ferences were a result of variations in crystal growth rates
along different planes of the crystal structure. This could have
been a result of the colony growth rate and/or actual urease
activity, which thus influenced the rate of supply of chemical
species required for precipitation (35). Alternatively, crystal
growth can be inhibited or altered by the adsorption of pro-
teins, organic matter, or inorganic components to specific crys-
tallographic planes of the growing crystal (19, 30, 34).
Initially, we hypothesized that the differences in crystal mor-
phology arose because the bacteria belonged to different gen-
era. Thus, it was surprising that for such a widespread function
as urease the isolates obtained were all so closely related to the
B. sphaericus group. Several ureolytic CaCO3-precipitating
FIG. 5. Michaelis-Menten kinetic values (Kmand Vmax) determined with crude enzyme extracts from the various ureolytic calcium-precipitating
isolates. The error bars indicate standard deviations.
FIG. 6. Effect of calcium ions (30 mM) on the specific urease activities (100 mM urea assay) of the various isolates, expressed as the urease
activity with calcium (UAcalcium) divided by the urease activity without calcium present (UA). The error bars indicate standard deviations.
4906 HAMMES ET AL.APPL. ENVIRON. MICROBIOL.
species have been characterized previously, including B. pas-
teurii (12, 37), Pseudomonas spp. and Variovorax spp. (13), and
Leuconostoc mesenteroides (12). We therefore emphasize that
the isolation and identification results presented here do not
necessarily represent a unique group of calcium-precipitating
bacteria but rather represent organisms that proliferate and
express the urease gene under the cultivation conditions used.
However, the close relationship of the various isolates ob-
tained in this study within the B. sphaericus group is remark-
able. To some extent, the predominance of a specific phyloge-
netic group can be attributed to the environments from which
the isolates were obtained. Both the soils and the concrete
surface represent dry conditions in which spore-forming organ-
isms, such as Bacillus species, proliferate. In this regard Felske
et al. (10), for instance, showed with fluorescent in situ hybrid-
ization that at least 40% of all bacteria in a Dutch soil belonged
to Bacillus species and some other low-G?C-content organ-
isms. The isolation and cultivation conditions could also have
had a selective influence; nutrient broth with urea added is a
preferred growth medium of ureolytic Bacillus species (12, 37).
Confirming this, when nonculturing methods are used, urease
biotechnology reactors are dominated by a variety of bacterial
species (15). Thus, the significance of the B. sphaericus group
is unknown, but the observation that marked differences in
crystal morphology occurred among closely related species
suggests that further research is needed to determine how bac-
teria precipitate CaCO3in order to optimize biotechnological
applications of this process.
A new approach to study the diversity of functional genes is
analysis of PCR products of these genes with DGGE (4, 32).
To our knowledge, this is the first study in which DGGE was
used to examine the diversity of a urease gene fragment. The
first noticeable result of the urease DGGE analysis is the ap-
parent presence of isozymes. Although the presence of urease
isoforms in a single organism has been described previously
(7), the notion of multiple isozymes has been largely rejected
(25). DGGE examination of PCR products amplified with de-
generated primers often reveals more than one band from a
single unique starting template (21). The generation of multi-
ple bands was, however, not entirely consistent, since not all
CPB DNA templates resulted in multiple bands (Fig. 4). As a
consequence, in some of the CPB strains isolated, the possi-
bility that there are different isozymes cannot be excluded.
DGGE also revealed distinct differences as well as evident
similarities between isolates, which were highlighted by the
cluster analysis (Fig. 4). Interestingly, some correlation be-
tween the clustering results and the crystal groups could be
detected. For example, the 100% similarity between CPB 1 and
CPB 2 and the 100% similarity between CPB 5 and CPB 6
coincide with the corresponding distinct and almost identical
crystallization patterns (Fig. 2). On the other hand, B. pasteurii
clustered with CPB 5, CPB 6, and CPB 9, in contrast to the
phylogenetic clustering (Fig. 3). These results confirm previous
reports that there is a definite degree of divergence in the
genetic make-up of microbial ureases (25, 26). DGGE makes it
possible to reveal this urease diversity, since this technique
reveals a 1-bp difference between two sequences (11). This
DGGE approach, applied to total DNA extracted from various
environmental habitats, could be especially useful for further
investigation of the diversity of urease genes in microbial com-
munities without prior cultivation of the urease-positive organ-
The results reported here showed that urease activity was
present in all the isolates and that the urease enzymes were not
extracellular in any of the isolates, which was consistent with
most of the previously described data concerning this point (7,
25, 37). Kmvalues for urease enzymes ranging from 0.1 to 100
mM urea have been reported for bacteria (24), suggesting that
the values obtained for CPB 7 to CPB 12 (35 to 55 mM) reflect
rather low substrate affinities. On the other hand, the Vmax
values for most isolates were rather high. Stocks-Fisher et al.
(37) reported a Vmaxvalue for B. pasteurii urease of 1.72 to 3.55
mM ? min?1? mg of protein?1(depending on the pH). The
standard reported values range from 1 to 5.5 mM ? min?1? mg
of protein?1for purified microbial enzymes, while Ureoplasma
ureolyticum was shown to have values between 33 and 180
mM ? min?1? mg of protein?1(25). Crude enzyme extracts
should be used comparatively rather than as preparations that
are reflective of the actual urease activity of an organism, as
the actual urease activity is significantly affected by transmem-
brane transport of urea (via urea permease enzymes), ammo-
nia, and protons. Nonetheless, the activity results in Fig. 5 do
explain to some extent the differences seen in crystallization.
Isolates CPB 1 to CPB 6 all exhibited moderate to high sub-
FIG. 7. PCA incorporating biochemical and molecular urease data and revealing at least three major clusters.
VOL. 69, 2003MICROBIAL CaCO3PRECIPITATION4907
strate affinities and high specific rates, which resulted in rapid
crystallization of CaCO3aggregates that were crystalline in
appearance and had a clear to light-brown color. This was
especially the case for isolates CPB 5 and CPB 6. Isolates CPB
7 to CPB 12, representing the initial classes 5 and 6, displayed
low substrate affinities and low specific rates (with the excep-
tion of isolate CPB 12, which displayed a rather high specific
rate). This could explain the slower crystal formation, which
allowed more colony growth, and thus the interference by the
sorption of organic matter in the crystal structure. The latter
would have caused both the less crystalline appearance of the
aggregates and the darker color.
The presence of calcium modulated urease activity. South-
am (36) suggested that surface-associated mineralization could
result in limitations of nutrient transport and eventual disrup-
tion of the proton motive force, which suggested that precipi-
tation resulting from urea hydrolysis might be detrimental to
bacterial cells and thus to further urease activity. Figure 6
shows that for most of the isolates no difference was detected
between urease activity in the presence of calcium and urease
activity in the absence of calcium. However, isolates CPB 1 to
CPB 4 exhibited remarkable increases in urease activity in the
presence of soluble calcium. This coincides with characteristic
similar macrocrystallization by these isolates (Fig. 2) and
strengthens the conclusion described above that high urease
affinities and high specific rates were the basis of the differ-
ences in macrocrystallization. To our knowledge, calcium ions
have not previously been associated with increased urease ac-
tivity. While Ca2?could theoretically facilitate better trans-
membrane transport or improve intracellular signaling pro-
cesses, it should be expected that these processes occur in all
related organisms, such as the case described here. An alter-
native possibility is a detoxification response of the bacteria to
high calcium concentrations (17). Active calcium metabolism
requires energy (ATP) (23), and several microorganisms (e.g.,
B. pasteurii) have previously been shown to produce ATP
through urea hydrolysis (25). The fact that some strains dis-
played such a pronounced effect warrants further research on
the possibilities of an ATP-driven calcification process.
In conclusion, in this paper we clearly show that strain-
specific precipitation occurs during ureolytic microbial CaCO3
precipitation in a closely related group of bacteria. The differ-
ences in precipitation could largely be ascribed to diversity
in the urease enzymes of the various isolates, coupled to the
pronounced effect of calcium on urease activity in some strains.
This research was sponsored in part by an equipment grant from the
Flemish Fund for Scientific Research (FWO-Vlaanderen).
We thank Sabine Verryn for performing the XRD analysis and
phase identification, R. Hausinger for providing comments on the
urease molecular results, and Vanessa Vermeirsen, Hanne Lievens,
and Sofie Dobelaere for critically reading the manuscript.
1. Altschul, S. F., T. L. Madden, A. A. Schaffer, J. Zhang, Z. Zhang, W. Miller,
and D. J. Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generation
of protein database search programs. Nucleic Acids Res. 25:3389–3402.
2. Anderson, S., V. D. Appanna, J. Huang, and T. Viswanatha. 1992. A novel
role for calcite in calcium homeostasis. FEBS Lett. 308:94–96.
3. Boon, N., J. Goris, P. De Vos, W. Verstraete, and E. M. Top. 2000. Bioaug-
mentation of activated sludge by an indigenous 3-chloroaniline-degrading
Comamonas testosteroni strain, I2gfp. Appl. Environ. Microbiol. 66:2906–
4. Boon, N., J. Goris, P. De Vos, W. Verstraete, and E. M. Top. 2001. Genetic
diversity among 3-chloroaniline- and aniline-degrading strains of the Co-
mamonadaceae. Appl. Environ. Microbiol. 67:1107–1115.
5. Boquet, E., A. Boronat, and A. Ramos-Cormenzana. 1973. Production of
calcite (calcium carbonate) crystals by soil bacteria is a general phenomenon.
Nature (London) 246:527–529.
6. Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein-dye bind-
ing. Anal. Biochem. 72:248–254.
7. Burne, R. A., and Y. Y. Chen. 2000. Bacterial ureases in infectious diseases.
Microbes Infect. 2:533–542.
8. Castanier, S., G. Le Me ´tayer-Levrel, and J. P. Perthuisot. 1999. Ca-carbon-
ates precipitation and limestone genesis—the microbiologist point of view.
Sediment. Geol. 126:9–23.
9. Douglas, S., and T. J. Beveridge. 1998. Mineral formation by bacteria in
natural microbial communities. FEMS Microbiol. Ecol. 26:79–88.
10. Felske, A., A. D. L. Akkermans, and W. M. De Vos. 1998. In situ detection of
an uncultured predominant Bacillus in Dutch grassland soils. Appl. Environ.
11. Felske, A., A. Wolterink, R. Van Lis, W. M. De Vos, and A. D. L. Akkermans.
1999. Searching for predominant soil bacteria: 16S rDNA cloning versus
strain cultivation. FEMS Microbiol. Ecol. 30:137–145.
12. Ferris, F. G., and L. G. Stehmeier. January 1993. Bacteriogenic mineral
plugging. U.S. patent 5,143,155
13. Fujita, Y., F. G. Ferris, R. D. Lawson, F. S. Colwell, and R. W. Smith. 2000.
Calcium carbonate precipitation by ureolytic subsurface bacteria. Geomicro-
biol. J. 17:305–318.
14. Greenberg, A. E., L. S. Clesceri, and A. D. Eaton (ed.). 1992. Standard
methods for the examination of water and wastewater, 18th ed. American
Public Health Association, Washington, D.C.
15. Hammes, F., N. Boon, G. Clement, J. de Villiers, S. D. Siciliano, and W.
Verstraete. Molecular, biochemical and ecological characterisation of a bio-
catalytic calcification reactor. Appl. Microbiol. Biotechnol., in press.
16. Hammes, F., A. Seka, S. De Knijf, and W. Verstraete. 2003. A novel ap-
proach to calcium removal from calcium-rich industrial wastewater. Water
17. Hammes, F., and W. Verstraete. 2002. Key roles of pH and calcium metab-
olism in microbial carbonate precipitation. Re. Environ. Sci. Bio/Technol.
18. Jukes, T. H., and C. R. Cantor. 1969. Evolution of protein molecules, p.
21–132. In H. N. Munro (ed.), Mammalian protein metabolism. Academic
Press, Inc., New York, N.Y.
19. Kawaguchi, T., and A. W. Decho. 2002. A laboratory investigation of cya-
nobacterial extracellular polymeric secretions (EPS) in influencing CaCO3
polymorphism. J. Cryst. Growth 240:230–235.
20. Knorre, H., and W. Krumbein. 2000. Bacterial calcification, p. 25–31. In
R. E. Riding and S. M. Awramik (ed.), Microbial sediments. Springer-
Verlag, Berlin, Germany.
21. Kowalchuk, G., J. Stephen, W. De Boer, J. Prosser, T. Embley, and J.
Woldendorp. 1997. Analysis of ammonia-oxidizing bacteria of the beta sub-
division of the class Proteobacteria in coastal sand dunes by denaturing
gradient gel electrophoresis and sequencing of PCR-amplified 16S ribosomal
DNA fragments. Appl. Environ. Microbiol. 63:1489–1497.
22. Maidak, B. L., J. R., Cole, T. G. Lilburn, C. T. Parker, Jr., P. R. Saxman,
R. J. Farris, G. M. Garrity, G. J. Olsen, T. M. Schmidt, and J. M. Tiedje.
2001. The RDP-II (Ribosomal Database Project). Nucleic Acids Res. 29:
23. McConnaughey, T. A., and F. F. Whelan. 1997. Calcification generates pro-
tons for nutrient and bicarbonate uptake. Earth Sci. Rev. 42:95–117.
24. McKay, D. S., E. K. Gibson, Jr., K. Thomas-Keprta, H. Vali, C. S. Romanek,
S. J. Clemett, X. D. F. Chillier, C. R. Maechling, and R. N. Zare. 1996.
Search for past life on Mars: possible relic biogenic activity in Martian
meteorite ALH84001. Science 273:924–930.
25. Mobley, H. L. T., and R. P. Hausinger. 1989. Microbial ureases: significance,
regulation and molecular characterization. Microbiol. Rev. 53:85–108.
26. Mobley, H. L. T., M. D. Island, and R. P. Hausinger. 1995. Molecular biology
of microbial ureases. Microbiol. Rev. 59:451–480.
27. Muyzer, G., E. C. De Waal, and A. Uitterlinden. 1993. Profiling of complex
microbial populations using denaturing gradient gel electrophoresis analysis
of polymerase chain reaction-amplified genes coding for 16S rRNA. Appl.
Environ. Microbiol. 59:695–700.
28. Peckman, J., J. Paul, and V. Thiel. 1999. Bacterially mediated formation of
diagenetic aragonite and native sulphur in Zechstein carbonates (Upper
Permian, central Germany). Sediment. Geol. 126:205–222.
29. Reed, K. E. 2001. Restriction enzyme mapping of bacterial urease genes:
using degenerate primers to expand experimental outcomes. Biochem. Mol.
Biol. Edu. 29:239–244.
30. Rivadeneyra, M. A., G. Delgado, A. Ramos-Cormenzana, and R. Delgado.
1998. Biomineralisation of carbonates by Halomonas eurihalina in solid and
liquid media with different salinities: crystal formation sequence. Res. Mi-
31. Rivadeneyra, M. A., G. Delgado, M. Soriano, A. Ramos-Cormenzana, and R.
4908HAMMES ET AL.APPL. ENVIRON. MICROBIOL.
Delgado. 2000. Precipitation of carbonates by Nesterenkonia halobia in liquid Download full-text
media. Chemosphere 41:617–624.
32. Rosado, A. S., G. F. Duarte, L. Seldin, and J. D. Van Elsas. 1998. Genetic
diversity of nifH gene sequences in Paenibacillus azotofixans strains and soil
samples analyzed by denaturating gradient gel electrophoresis of PCR-am-
plified gene fragments. Appl. Environ. Microbiol. 64:2770–2779.
33. Saitou, N., and M. Nei. 1987. The neighbor-joining method: a new method
for reconstructing phylogenetic trees. Mol. Biol. Evol. 4:406–425.
34. Schultze-Lam, S., D. Fortin, B. S. Davis, and T. J. Beveridge. 1996. Miner-
alisation of bacterial surfaces. Chem. Geol. 132:171–181.
35. Sondi, I., and E. Matijevic. 2001. Homogeneous precipitation of calcium
carbonates by enzyme catalyzed reaction. J. Colloid Interface Sci. 238:208–
36. Southam, G. 2000. Bacterial surface-mediated mineral formation, p. 257–
276. In D. R. Lovley (ed.), Environmental microbe-metal interactions. ASM
Press, Washington D.C.
37. Stocks-Fisher, S., J. K. Galinat, and S. S. Bang. 1999. Microbiological
precipitation of CaCO3. Soil Biol. Biochem. 31:1563–1571.
38. Trewin, N. H., and A. H. Knoll. 1999. Preservation of Devonian chemotro-
phic filamentous bacteria in calcite veins. Palaios 14:288–294.
39. Vecht, A., and T. G. Ireland. 2000. The role of vaterite and aragonite in the
formation of pseudo-biogenic carbonate structures: implications for Martian
exobiology. Geochim. Cosmochim. Acta 66:2719–2725.
40. Westfall, F. 1999. The nature of fossil bacteria: a guide to the search for
extraterrestrial life. J. Geophys. Res. E Planets 104:16437–16451.
VOL. 69, 2003MICROBIAL CaCO3PRECIPITATION4909