Environ. Sci. Technol. 2003, 37, 343-351

Real-Time PCR Quantification of
Nitrifying Bacteria in a Municipal
Wastewater Treatment Plant
G E R D A H A R M S , †,§ A L I C E C . L A Y T O N , †,§
H E B E M . D I O N I S I , †,§
I G R I D R . G R E G O R Y , †,§
V I C T O R I A M . G A R R E T T , †,§
S H A W N A . H A W K I N S , ‡,§
K E V I N G . R O B I N S O N , ‡,§ A N D
G A R Y S . S A Y L E R * ,†,§
Department of Microbiology, Department of Civil and
Environmental Engineering, and Center for Environmental
Biotechnology, The University of Tennessee,
Knoxville, Tennessee, 37996

Real-time PCR assays using TaqMan or Molecular
Beacon probes were developed and optimized for the
quantification of total bacteria, the nitrite-oxidizing bacteria
Nitrospira, and Nitrosomonas oligotropha-like ammonia
oxidizing bacteria (AOB) in mixed liquor suspended solids
(MLSS) from a municipal wastewater treatment plant
(WWTP) using a single-sludge nitrification process. The
targets for the real-time PCR assays were the 16S rRNA
genes (16S rDNA) for bacteria and Nitrospira spp. and
the amoA gene for N. oligotropha. A previously reported
assay for AOB 16S rDNA was also tested for its application
to activated sludge. The Nitrospira 16S rDNA, AOB 16S
rDNA, and N. oligotropha-like amoA assays were loglinear over 6 orders of magnitude and the bacterial 16S
rDNA real-time PCR assay was log-linear over 4 orders
of magnitude with DNA standards. When these real-time
PCR assays were applied to DNA extracted from MLSS,
dilution of the DNA extracts was necessary to prevent
PCR inhibition. The optimal DNA dilution range was
broad for the bacterial 16S rDNA (1000-fold) and
Nitrospira 16S rDNA assays (2500-fold) but narrow for
the AOB 16S rDNA assay (10-fold) and N. oligotrophalike amoA real-time PCR assay (5-fold). In twelve MLSS
samples collected over one year, mean cell per L
values were 4.3 ( 2.0 × 1011 for bacteria, 3.7 ( 3.2 ×
1010 for Nitrospira, 1.2 ( 0.9 × 1010 for all AOB, and
7.5 ( 6.0 × 109 for N. oligotropha-like AOB. The percent
of the nitrifying population was 1.7% N. oligotropha-like
AOB based on the N. oligotropha amoA assay, 2.9% total
AOB based on the AOB 16S rDNA assay, and 8.6% nitriteoxidizing bacteria based on the Nitrospira 16S rDNA assay.
Ammonia-oxidizing bacteria in the wastewater treatment
plant were estimated to oxidize 7.7 ( 6.8 fmol/hr/cell based
on the AOB 16S rDNA assay and 12.4 ( 7.3 fmol/hr/cell
based on the N. oligotropha amoA assay.
* Corresponding author phone: (865)974-8080; fax: (865)974-8086;
e-mail: sayler@utk.edu.
† Department of Microbiology.
‡ Department of Civil and Environmental Engineering.
§ Center for Environmental Biotechnology.
10.1021/es0257164 CCC: $25.00
Published on Web 12/04/2002

© 2003 American Chemical Society

1. Introduction
Chemolithotrophic nitrification is a two-step process involving two groups of bacteria: ammonia-oxidizing bacteria
(AOB) oxidize NH3 to NO2- and nitrite-oxidizing bacteria
(NOB) oxidize NO2- to NO3- (1). Nitrification is an important
process in biogeochemical nitrogen cycling and in controlling
effluent toxicity in wastewater treatment. The physiological
activity and abundance of these organisms in wastewater
processing is critical in the design and operation of waste
treatment systems, particularly since these organisms display
low growth rate and high sensitivity to environmental
disturbances and inhibitors (2). An important aspect of
activity relates to reactor design. Single sludge wastewater
treatment designs accomplish nitrification concurrent with
removal of carbonaceous oxygen demand in one reactor;
activated sludge in these reactors contains both heterotrophs
and nitrifiers that necessarily compete for resources such as
oxygen (3). Single sludge reactors are a dominant design in
the United States (4). Alternatively, nitrification can be
accomplished using a series of reactors, the first dedicated
to carbonaceous oxygen demand removal and the second to
nitrification. To date, molecular quantification of nitrifying
populations in activated sludge from full-scale wastewater
treatment plants (WWTPs) has been performed on samples
obtained from facilities with high nitrogen loads (5) or twostage reators (6, 7). Given that single-sludge nitrification
processes may contain fewer nitrifiers and are more susceptible to plant upsets, attempts at quantification of nitrifiers
in these plants is warranted.
In recent studies, a competitive Polymerase Chain Reaction (cPCR) procedure was developed to quantify nitrifying
bacteria by PCR amplification of ammonia monooxygenase
(amoA) and Nitrospira spp. 16S rRNA genes (16S rDNA),
respectively (8, 9). The cPCR procedure is a well established
quantitative method that has been used to enumerate both
culturable and nonculturable organisms, including nitrifying
bacteria (10-13). This method relies on the measurement of
PCR products at the endpoint, after gel electrophoresis, and
it has a log-linear detection range of only 2 to 3 orders of
magnitude (14, 15). The cPCR technique is robust due to the
presence of a stringent internal control. However, cPCR is
difficult to use in routine process monitoring of populations
in wastewater treatment processes because it is labor- and
cost-intensive with low throughput.
In the current study, real-time PCR was investigated for
applications in monitoring nitrifying populations in wastewater treatment because it combines high throughput with
high analytical sensitivity and precision, offering a dynamic
detection range of 6 orders of magnitude or more (15, 16).
Although real-time PCR has been widely applied in medical
research, its application to environmental research has been
slower (17-21). In real-time PCR, amplicons are detected by
measurement of a fluorescence signal without post-PCR
sample processing such as gel electrophoresis (Figure 1).
Several different fluorescent probes can be used in real-time
PCR including TaqMan (22) or Molecular Beacon (23, 24). A
TaqMan probe is a linear oligonucleotide complementary to
a target nucleic acid sequence, with a fluorophore attached
to the 5′end and a quencher to the 3′end (22) (Figure 1). The
TaqMan probe is cleaved by the 5′ exonuclease activity of
Taq DNA polymerase as the primer is extended, resulting in
the separation of the reporter dye from quencher dye and
an increase in fluorescence signal emitted by the reporter.
In each cycle, additional reporter dye molecules are cleaved,
resulting in an increase in fluorescence intensity proportional
to the amount of amplicon produced (22). In contrast, a
VOL. 37, NO. 2, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

343

FIGURE 1. Schematic representation of the quantification of specific targets from mixed-liquor suspended solids (MLSS) samples. After
DNA extraction, the DNA template is used in real-time PCR assays. (A) TaqManprobe assay. A TaqMan probe is a linear oligonucleotide
probe complementary to the target sequence, with a fluorescent dye (green circle, F) attached to the 5′end and a quencher (blue triangle,
Q) to the 3′end. The proximity of these two dyes quenches the signal. Steps: (1) Denaturation of the DNA at 95 °C separates the template.
(2) Annealing of primers (usually 18-22 nucleotides) and probe to target DNA at temperatures of 50 °C to 60 °C. (3) Extension and synthesis
of the DNA strand results in the 5′ exonuclease activity by Taq DNA polymerase (red oval, P) separating the fluorescent dye from the
quencher resulting in an increase in fluorescence. Steps 2 and 3 are often performed simultaneously. The spectrofluorimetric thermal
cycler measures the relative fluorescence at the end of step 3 (eye). Steps 1-3 are repeated 45 to 50 times. In each cycle, additional
reporter dye molecules are cleaved, resulting in an increase in fluorescence intensity proportional to the amount of amplicon produced
(22, 43). (B) Molecular beacon assay. A Molecular beacon probe has a stem and loop structure. The loop section of the probe is
complementary to the target and the stem results from the annealing of artificially designed arm sequences (23, 24, 43). A fluorescent
dye (green circle, F) and a quencher (blue triangle, Q) are attached to the arm sequences. Step: (1) Denaturation at 95 °C results in the
separation of the template DNA and separation of the stem on the molecular beacon probe, separating the quencher from the dye allowing
fluorescence. (2) Annealing at 60 °C allows the probe to hybridize to the template resulting in fluorescence. Alternatively, if the probe
does not hybridize with the template the stem-and-loop structure is reformed and fluorescence is quenched. The relative fluorescence
is measured at the beginning of this step (eye) and is proportional to the number of target sequences. (3) Extension and synthesis of DNA
at 72 °C results in dissociation of the probe from the target and fluorescence is quenched. Steps 1-3 are repeated 45 to 50 times. The
final output for both assays shows relative fluorescence as a function of cycle number. Amplification curves for template concentrations
between 107 and 10 copies per PCR reaction are shown (left to right). The threshold is calculated as 10 times the standard deviation of
the background fluorescence (dashed line). The point where the fluorescence signal crosses the threshold is the threshold cycle (CT) (dotted
line) and is lower when more copies of the template are present at the beginning of the reaction.
344

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 2, 2003

TABLE 1. Primers and Probes Used in This Study
assay

target

primer/probea

sequence (5′-3′)b

TM (°C)c

reference

1055f
1392r
16STaq1115
amoNo550D2f
amoNo754r
amoNoTaq729

5′-ATGGCTGTCGTCAGCT-3′
5′-ACGGGCGGTGTGTAC-3′
5′-(6-FAM)-CAACGAGCGCAACCC-(TAMRA)-3′
5′-TCAGTAGCYGACTACACMGG-3′
5′-CTTTAACATAGTAGAAAGCGG-3′
5′-(6-FAM)-CCAAAGTACCACCATACGCAG(TAMRA)-3′
5′-GGAGRAAAGCAGGGGATCG-3′
5′-GGAGGAAAGTAGGGGATCG-3′
5′-CGTCCTCTCAGACCARCTACTG-3′
5′-(6-FAM)-CAACTAGCTAATCAGRCATCRGCCGCTC-(TAMRA)-3′
5′-CCTGCTTTCAGTTGCTACCG-3′
5′-GTTTGCAGCGCTTTGTACCG-3′
5′-(6-FAM)-AGCACTCTGAAAGGACTGCCCAGG(TAMRA)-3′
5′-(6-FAM)-GCTGCACC|AGCACTCTGAAAGGACTGCCCAGG|GGTGCAGC-(DABCYL)-3′

57.7
58.9
62.9
55.4
55.0
64.1

(25)
(26)
this study
this study
this study
this study

63.5
62.3
62.8
71.0

(20)
(20)
(20)
(20)

64.7
67.8
73.9

( 9)
( 9)
this study

bacterial
16S rDNA

bacterial
16S rDNA

N. oligotrophalike amoA

N. oligotropha
amoA gene

AOB

ammonia-oxidizing
bacterial
16S rDNA

CTO 189fA/B
CTO 189fC
RT1r
TMP1

Nitrospira
16S rDNA

Nitrospira spp.
16S rDNA

NSR1113f
NSR1264r
NSR1143Taq
NSR1143Beac

this study

a Primer/probe abbreviations: f ) forward primer, r ) reverse primer, Taq ) TaqMan probe, Beac ) beacon probe. b 5′-Fluorophore-probequencher-3′ in case of TaqMan probe; 5′-fluorophore-arm|probe|arm-quencher-3′ in case of Molecular Beacon. 6-FAM ) 6-carboxyfluorescein;
TAMRA ) carboxytetramethylrhodamine; DABCYL ) 4-(4-dimethylaminophenyl)azo)benzoic acid. c Melting temperatures were calculated using
the oligo calculator from Sigma Genosys (http://www.genosys.com/cgi-win/oligo_calconly.exe; Sigma Genosys, The Woodlands, TX).

Molecular Beacon probe has a stem and loop structure, with
the loop section of the probe complementary to the target
and the stem formed by the annealing of an artificially
designed arm sequence (24). A fluorescent dye and a
quencher are attached to both ends of the molecule (Figure
1). When free in solution the Molecular Beacon adopts a
hairpin structure, which results in fluorophore quenching.
In the presence of a complementary target, the hairpin
structure unfolds and the separation of the fluorophore and
the quencher leads to emission of fluorescence. Using either
a TaqMan or a Molecular Beacon probe the threshold cycle
of a sample is inversely proportional to the logarithm of the
amount of template DNA initially added to the PCR reaction
for both types of probe (22-24). Although the design,
chemistry and portion of the PCR cycle in which fluorescence
is detected differ for these two probes, similar results should
be obtained in the enumeration of molecules.
The primary objectives of the current study were to
develop methods and to quantify ammonia-oxidizing bacteria
and nitrite-oxidizing bacteria (Nitrospira) in a single-stage
type municipal WWTP. Real-time PCR assays for Nitrospira
16S rDNA and Nitrosomonas oligotropha amoA genes were
developed from existing cPCR assays for monitoring nitrifying
bacteria, and a real-time PCR assay developed to enumerate
AOB 16S rDNA in soil (20) was applied to mixed-liquor
suspended solids (MLSS). A bacterial 16S rDNA TaqMan assay
was also developed to monitor total biomass in the MLSS
samples. Secondary objectives of this study were to compare
real-time PCR assays using different probe designs (molecular
beacon versus TaqMan), compare real-time PCR assays
directed toward ribosomal RNA genes versus catabolic genes
(amoA), and validate real-time PCR assays by comparison
with previous data in copies per liter obtained by dot-blot
hybridization and competitive PCR. The results of this study
indicate that real-time PCR can be implemented as a tool to
facilitate molecular monitoring and quantification of critical
sub-populations, such as nitrifying bacteria, in wastewater
treatment processes.

(NH4+-N) removal designed with average and peak flow
capacities of 40 million gallons per day (MGD) and 70 MGD,
respectively. In the year 2000, this single stage reactor system
was operated at an average solids retention time of 12 days.
The reactor system consists of six identical basins arranged
in parallel, each basin being 183 feet long, 32 feet wide, and
unusually deep at 33 feet. The basins are in turn sub-divided
into five compartments in series along the flow path. Each
compartment is aerated with ceramic fine bubble diffusers
that provide for complete mixing. The basins are fed via a
common influent channel. Gravity overflow from the basins
first combines and then splits to flow to a series of circular
clarifiers. MLSS samples were collected from the upwell at
the center of one of the clarifiers prior to the effluent flow
entering the clarifier quiescent zone. Reactor specific flow
and influent ammonia data along with plant effluent data
was obtained. The mean influent BOD5 was 302 (( 47) mg/L,
and mean mixed-liquor volatile suspended solids (MLVSS)
was 1971 ( 178 mg/L over the one-year period. The mean
monthly influent temperature for the year was 17 °C, with
a low of 9 °C and a high of 25 °C. Genomic DNA was extracted
in triplicate from 2 mL of MLSS samples using a FastDNA kit
(BIO 101, Vista, CA) with minor modifications as described
by Dionisi et al. (9).

2. Methodology

Real-Time PCR Assays. Real-time PCR assays were
developed for the quantification of bacterial 16S rDNA, Nitrospira spp. 16S rDNA, and N. oligotropha-like amoA. Three
TaqMan probes, 16sTaq1115, amoNoTaq729, and NSR1143Taq, were designed (Table 1) using the guidelines provided
by Applied Biosystems (http://home.appliedbiosystems.com;
Applied Biosystems, Foster City, CA). The primers and probes
were synthesized by Sigma Genosys (Sigma Genosys, The
Woodlands, TX). A Molecular Beacon probe, NSR1143Beac,
was designed using the guidelines provided at http://
www.molecular-beacons.org/protocol.html and synthesized
by Stratagene (Stratagene, La Jolla, CA). The optimal artificial
arm sequences were determined by using the Zuker DNA
folding program (mfold) (http://bioinfo.math.rpi.edu/∼mfold/
dna/form1/cgi).

Samples. MLSS samples were collected monthly for one year
from a local municipal WWTP treating mainly municipal
wastewater. The WWTP employs single stage reactors for
carbon (biological oxygen demand (BOD5)) and nitrogen

The real-time PCR assay for AOB used two forward primers
CTO 189A/B and CTO189C, one reverse primer RT1r and the
TaqMan probe TMP1 (Table 1) as described by Hermansson
and Lindgren (20).
VOL. 37, NO. 2, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

345

Real-Time PCR for Quantification of Bacterial 16S rDNA.
Bacterial 16S rDNA was amplified using primers 1055f (25)
and 1392r (26) (Table 1). The TaqMan probe 16STaq1115
was modified from the 1114f primer (26). The PCR mix with
a total volume of 25 µL contained Platinum Quantitative PCR
SuperMix-UDG (Life Technologies, Inc., Gaithersburg, MD)
with 5 mM MgCl2, 15 pmol primers 1055f and 1392r, 6.25
pmol TaqMan probe 16STaq1115, 3.2 to 7.0 ng of sample
DNA or dilutions of plasmid pCR2.1 vector (Invitrogen,
Carlsbad, CA) carrying a 16S rRNA gene for Nitrospira
(GenBank accession number AF420301) (9) as standard (from
4.5 × 103 to 4.5 × 108 copies of the 16S rDNA gene). The PCR
program was 3 min at 50 °C, 10 min at 95 °C, 45 cycles at 95
°C for 30 s, 50 °C for 60 s, and 72 °C for 20 s.
Real-Time PCR for Quantification of Nitrospira 16S
rDNA. The Nitrospira 16S rDNA primers NSR1113f and
NSR1264r (Table 1) were previously designed and tested using
genomic DNA extracted from municipal and industrial MLSS
as templates (9). The TaqMan probe NSR1143Taq (Table 1)
was derived from a conserved sequence region between the
primers NSR1113f and NSR1264r. The probe portion of the
Molecular Beacon NSR1143Beac (Table 1) is identical to that
one used in the TaqMan probe. The probe region of
NSR1143Beac has a TM of 68.0 °C, and the stem region has
a TM of 70.2 °C (with 5.0 mM MgCl2).
Real-time PCR assays using NSR1143Taq were performed
in a total volume of 25 µL with 5 mM MgCl2, Platinum
SuperMix, 15 pmol of primers NSR1113f and NSR1264r, 6.25
pmol TaqMan probe NSR1143Taq, and 3.2 to 7.0 ng of sample
DNA or 30 to 3 × 107 copies of the standard (151 bp fragment
of Nitrospira 16S rDNA from AF420301) (9). PCR amplification
consisted of 2 min at 50 °C, 10 min at 95 °C, 55 cycles at 95
°C for 30 s, 63 °C for 60 s.
The Molecular Beacon assay contained Platinum SuperMix with 5 mM MgCl2, 25 pmol of primers NSR1113f and
NSR1264r, 8.5 pmol Molecular Beacon probe NSR1143Beac,
3.2 to 7.0 ng of sample DNA or 30 to 3 × 106 copies of standard
DNA in 25 µL. PCR amplification consisted of 5 min at 95 °C,
55 cycles at 95 °C for 30 s, 60 °C for 60 s, and 72 °C for 10 s.
Real-Time PCR for Quantification of N. oligotrophalike amoA Gene. The primers amoNo550D2f and amoNo754r
(Table 1) were designed to target the amoA gene of ammoniaoxidizing bacteria found in the full-scale municipal WWTP
(9) based on alignment of amoA gene sequences using the
CLUSTAL W program (27). Alignments consisted of amoA
sequences from clonal libraries obtained from the WWTP
(9), four bench-scale municipal wastewater treatment systems (8), and amoA sequences available in GenBank (28).
The forward primer amoNo550D2f contained two degenerate
bases in order to amplify all amoA clones from the libraries,
as well as N. urea (AJ388585) and N. oligotropha (AF272406)
amoA genes. The TaqMan probe amoNoTaq729 was derived
from a conserved sequence region within the primer pair
amoNo550D2f and amoNo754r (Table 1). The 25 µL PCR
mix contained TaqMan Universal PCR Master Mix (PE
Applied Biosystems, Foster City, CA) with 7.5 pmol primers
amoNo550D2f and amoNo754r, 6.25 pmol TaqMan probe
amoNoTaq729, 0.3 to 0.7 ng of sample DNA. Standards
consisted of the plasmid pCR2.1 carrying the M-20 amoA
gene (GenBank accession number AF420299) (9) adjusted to
10 to 1.0 × 107 copies per PCR. PCR amplification consisted
of 3 min at 50 °C, 10 min at 95 °C, 55 cycles at 95 °C for 30 s,
56 °C for 60 s.
Real-Time PCR for Quantification of AOB 16S rRNA
Gene. Real-time PCR assays were performed as described by
Hermansson and Lindgren (20) in a total volume of 25 µL
with Universal PCR Master Mix (PE Applied Biosystems), 7.5
pmol of a 2:1 ratio of primers CTO 189fA/B and CTO 189fC,
7.5 pmol of the reverse primer RT1r, 3.125 pmol TaqMan
probe TMP1, and 0.3 to 0.7 ng of sample DNA or 60 to 6 ×
346

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 2, 2003

107 copies of the standard (Nitrosomonas europea 16S rDNA
cloned into pCR2.1). PCR amplification consisted of 2 min
at 50 °C, 10 min at 95 °C, 40 cycles at 95 °C for 30 s, 60 °C
for 60 s.
Acquisition and Data Analysis. Real-time PCR assays for
bacterial 16S rDNA and Nitrospira 16S rDNA were run on a
Bio-Rad iCycler with the iCycler iQ fluorescence detector
and iCycler software version 2.3 (Bio-Rad, Hercules, CA).
Plate well factors were determined prior to each PCR run to
normalize background fluorescence intensities from each
single well. amoA and AOB 16S rDNA real-time PCR assays
were run using a DNA Engine Opticon Continuous fluorescence Detection System (MJ Research, Waltham, MA). The
threshold was determined by the computer software as 10
times the standard deviation of the background fluorescence
averaged over at least 5 cycles at the start of the run. The
threshold cycle (CT) of each PCR reaction was automatically
determined by detecting the cycle at which the fluorescence
exceeded the calculated threshold. For the Molecular Beacon
probe, the data window was adjusted to capture roughly
10% of all data, chosen among the data points in the beginning
area of the annealing step. During each PCR run, the CT
values obtained from the DNA standards were used for the
construction of standard curves.
Data for bacterial 16S rDNA, Nitrospira 16S rDNA, and N.
oligotropha-like amoA was previously collected for the same
set of samples used in this study using dot-blots and
hybridization with a P32 labeled universal 16S rDNA probe
and competitive PCR assays for the Nitrospira 16S rDNA and
amoA genes (9). These data were compared to the data
obtained by real-time PCR for each gene using paired samples
t-tests. Paired samples t-tests compute the differences
between the values of the two variables for each case and
tests whether the average differs from zero (SPSS version
11.01, SPSS Inc., Chicago, IL). The null hypothesis was that
there were no significant differences between the copies/L
on each sample date obtained by the real-time PCR assay
and the analogous competitive PCR assay or dot-blot
hybridization.
Application of Real-Time PCR Assays to MLSS Samples.
All real-time PCR assays were performed using three replicates per sample, and all PCR runs included control reactions
without template. The effect of sample concentration on PCR
performance was determined using dilutions of sample DNA
(initial concentration of 50 ng/µL) containing 5 pg to 50 ng
in sterile water followed by real-time PCR analysis for 16S
rDNA, Nitrospira 16S rDNA, AOB 16S rDNA, and N. oligotropha-like amoA as described above.
Gene copies were initially calculated by comparison of
threshold cycles obtained in each PCR run from known
standard DNA concentrations. To reduce variability between
PCR runs, data were recalculated using a second standard
curve generated from 11 standard curves for bacterial 16S
rDNA (r2 ) 0.94), 22 standard curves for Nitrospira 16S rDNA
(r2 ) 0.84), 3 standard curves for AOB 16S rDNA (r2 ) 0.98),
and 5 standard curves for N. oligotropha amoA (r2 ) 0.90).
In the case of the Nitrospira 16S rDNA and bacterial 16S
rDNA assays, one universal standard curve was applied for
calculations, because both assays were shown to function
alike with standard plasmid AF420301 as template (Figure 2).

3. Results
Development and Optimization of Real-Time PCR Assays.
The real-time PCR assays were validated using known
concentrations of standard DNA. The linear range of detection
for the real-time PCR assay for bacterial 16S rDNA was 4
orders of magnitude, from 4.5 × 104 to 4.5 × 108 copies per
PCR and the detection limit for this assay was 4.5 × 103 target
DNA copies. The linear range of detection for the real-time
PCR assays for Nitrospira 16S rDNA, N. oligotropha amoA,

FIGURE 2. Comparison of bacterial 16S rDNA and Nitrospira 16S
rDNA real-time PCR TaqMan assays. The plasmid containing
Nitrospira 16S rDNA was used as the target for both assays. The
dashed line represents the 95% confidence interval. Error bars
indicate the standard deviation of 3 PCRs and are smaller than the
symbols.
and AOB 16S rDNA were at least 6 orders of magnitude, from
30 to 3.0 × 107, 10 to 1.0 × 107, and 60 to 6 × 107 copies per
PCR, respectively. The regression coefficient (r2) values for
standard curves for all real-time PCR assays in each run were
always above 0.90. In addition, copies per PCR calculated for
the standard plasmid AF420301 by the bacterial 16S rDNA
and Nitrospira 16S rDNA assays were highly correlated and
demonstrated a linear relationship (r2 ) 0.97) with a slope
of 1.14, indicating that the assays for bacterial 16S rDNA and
Nitrospira 16S rDNA were functioning alike (Figure 2).
Application of Real-Time TaqMan PCR Assays to WWTP
Samples. In environmental samples, PCR techniques can be
biased by the presence of inhibitory compounds that copurify with the DNA or low target concentrations in a high
background of heterologous DNA (15, 29). Inhibition effects
by the environmental samples on the real-time PCR assays
were tested using serial dilutions of genomic DNA extracted
from MLSS. PCR amplification was completely inhibited in
the undiluted sample at 50 ng per PCR assay for all four
real-time PCR assays and also in diluted samples containing
10 ng per PCR assay for the bacterial 16S rDNA and the N.
oligotropha-like amoA assays. CT values were proportional
to DNA at concentrations ranging from 5 pg to 5 ng for the
bacterial 16S rDNA (r2 ) 0.98, slope ) -3.87), at concentrations ranging from 5 pg to 5 ng for the Nitrospira 16S rDNA
(r2 ) 0.98, slope ) -3.45), at concentrations from 250 pg to
2.5 ng for the AOB 16S rDNA (r2 ) 0.98, slope ) -3.15), and
at concentrations ranging from 200 pg to 1 ng for the N.
oligotropha-like amoA assay (r2 ) 0.97, slope ) - 3.01) (Figure
3A). In the no-template control reactions no CT values were
obtained for the Nitrospira 16S rDNA, AOB 16S rDNA, and
N. oligotropha-like amoA assays. In the no template control
reactions using the 16S rDNA assay a CT value of 29.7 ( 0.43
was obtained. Conversion of the CT value to copies per PCR
based on the standard curve indicated that there were 3.8 (
0.9 × 103 copies bacterial 16S rDNA in the control reaction
without sample DNA. This background value may result from
bacterial DNA contamination of the Taq enzyme or other
reagents in the PCR mix. The MLSS sample containing greater
than 5 pg DNA per PCR assay was 20-fold higher than the
detection limit (Figure 3B) so the effect of background
contamination on the calculated value would be less than
5% in the most dilute samples. Conversion of the CT values
to copies per PCR for the other 3 assays indicated that the
detection limit for quantifying these targets in MLSS samples
were 180 copies for Nitrospira 16S rDNA, 1.2 × 103 for N.
oligotropha-like amoA, and 2.6 × 103 AOB 16S rDNA assays
(Figure 3 B).

FIGURE 3. (A) Threshold cycle (CT) measurements in diluted DNA
from a MLSS sample and (B) calculated copies of target genes in
diluted DNA from a MLSS sample. TaqMan real-time PCR assays
shown are Bacterial 16S rDNA (b), N. oligotropha-like amoA gene
(1), AOB 16S rDNA (3), and Nitrospira 16S rDNA (9). Error bars
indicate the standard deviation of 3 PCRs.
Comparison of TaqMan and Molecular Beacon Probes
in the Nitrospira 16S rDNA Real-Time PCR Assay. Fluorescence obtained with NSR1143Taq was twice as high as
the fluorescence obtained with NSR1143Beac (Figure 4a),
indicating a higher signal-to-noise ratio using the TaqMan
probe. The r2 values of standard curves obtained using
NSR1143Beac and NSR1143Taq were similar, ranging from
0.97 to 0.99 and 0.92 to 0.99, respectively. When MLSS samples
were analyzed using both probes, the copies obtained with
the Molecular Beacon probe were up to 12.5 times lower
than those obtained with the TaqMan probe, except for the
last 3 samples in which the values were almost equivalent
(Figure 4b). Nitrospira 16S rDNA copies per liter in the
municipal WWTP ranged from 1.7 × 1010 to 1.2 × 1011 using
the TaqMan probe and from 2 × 109 to 2.1 × 1011 using the
Molecular Beacon probe. In a paired t-test the differences
in results obtained by the Molecular Beacon probe and the
TaqMan probe assays were not significant (t ) -0.125, p )
0.903) (Table 2).
Validation of N. oligotropha amoA, Nitrospira 16S
rDNA, and Bacterial 16S rDNA Real-Time PCR Assays. To
verify the specificity of the amoA primers, PCR product
obtained with genomic DNA extracted from MLSS of the
municipal WWTP as template was cloned and sequenced as
described previously (9). Amplification of genomic DNA from
suspended solids using the primers amoNo550D2f and
amoNo754r produced a product with the expected size of
approximately 205 bp (data not shown). In a clonal library,
VOL. 37, NO. 2, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

347

FIGURE 4. (A) Measurements of Nitrospira 16S rDNA copies during
real-time PCR using TaqMan probe NSR1143Taq (9) or molecular
beacon probe NSR1143Beac (b) in DNA extracted from MLSS. The
fluorescence intensity was measured during the annealing step of
each temperature cycle. No-template controls are shown as (0)
and (O) for the TaqMan and molecular beacon probes, respectively.
Error bars indicate the standard deviation of 3 reactions. The dashed
line represents the threshold value (42 RFU) calculated for the
molecular beacon assay. The threshold value for the TaqMan assay
was 20 RFU. (B) Nitrospira 16S rDNA copies per liter as determined
by real-time PCR using TaqMan probe NSR1143Taq (9) or Molecular
Beacon probe NSR1143Beac (b) in DNA extracted from MLSS.

TABLE 2. Comparison of Real-Time PCR Assays with Other
Molecular Methods Using Paired Samples t-Test

compared groups
microbial 16S dot-blota
vs bacterial 16S TMb
amoAc TM vs amoA cPCRd
Ntspaf TM vs Ntspa cPCR
Ntspa MBg vs Ntspa cPCR

mean
paired
differences

t

df

significance
(2-tailed)

2.2 × 1013 10.79h 11

0.000

1.5 × 1010 4.324h 11
1.4 × 1010 1.456 11
1.5 × 1010 0.813 11

0.001
0.173
0.433

a dot-blot ) dot-blot hybridization. b TM ) TaqMan based real-time
PCR assay. c amoA ) N. oligotropha-like amoA. d cPCR ) competitive
quantitative PCR. e AOB ) ammonia-oxidizing bacteria 16S rDNA.
f Ntspa ) Nitrospira spp. 16S rDNA. g MB ) Molecular Beacon based
real-time PCR assay. h Significant at the prescribed R.

23 randomly selected clones were identified as amoA
sequences. These sequences were 89-94% similar to clone
M20 and 92-94% similar to clone M379, which were amoA
clone sequences previously isolated from this WWTP (9). N.
oligotropha-like amoA gene copies ranged from 2.6 × 109 to
4.3 × 1010 per L MLSS of the municipal WWTP. In contrast,
348

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 2, 2003

no amplification was observed when DNA isolated from
an industrial WWTP, where no N. oligotropha-like amoA
sequences were detected (9), was used as template (data not
shown).
Paired samples t-tests were used to determine whether
gene copies obtained in the real-time PCR assays were
equivalent to the values previously obtained using dot-blot
hybridization and competitive PCR assays for Nitrospira 16S
rDNA and N. oligotropha-like amoA (Table 2). The mean
bacterial 16S rDNA copies per liter previously obtained by
dot-blot hybridization was 2.3 ( 0.7 × 1013 (9). In this study,
the average 16S rDNA copies per liter obtained by real-time
PCR was 1 order of magnitude lower at 1.6 ( 0.7 × 1012 and
the differences in values obtained by these two methods
were statistically significant (Table 2). The mean Nitrospira
16S rDNA copies per liter previously obtained by competitive
PCR was 2.4 ( 1.3 × 1010 compared to 3.7 ( 3.2 × 1010 and
3.9 ( 6.4 × 1010 copies per liter obtained using the Nitrospira
TaqMan and Molecular Beacon assays, respectively. The
differences in the values obtained by the three assays were
not statistically significant (Table 2). The mean values
obtained using the N. oligotropha-like amoA competitive PCR
and the real-time PCR were 3.4 ( 2.3 × 108 and 1.5 ( 1.2 ×
1010 copies per liter, respectively. The differences in the mean
values obtained by the two assays were statistically significant
and may reflect the differences in the primers used for the
two assays.
Calculation of Cells/Liter from Copies/Liter and Ammonia-Oxidizing Activity in MLSS Samples. The number of
total bacteria, AOB, N. oligotropha-like AOB and Nitrospira
cells per liter MLSS were calculated from copies per liter
using several assumptions regarding gene copies per cell
(Table 3, Figure 5). First, the average 16S rDNA copies per
genome in bacterial cells were assumed to be 3.6 copies based
on the average 16S rDNA copies found in cultured bacteria
(30). Second, one cell of N. oligotropha was assumed to
contain 2 copies amoA based on the copies reported for N.
europaea (31). Third, both AOB and Nitrospira were assumed
to contain 1 copy 16S rDNA per cell based on copies 16S
rDNA found in Nitrobacter and the AOB Nitrosomonas and
Nitrosospira (32, 33).
In the 12 monthly samples, total bacteria ranged from
2.3 × 1011 to 8.4 × 1011 and Nitrospira spp. ranged from
1.7 × 1010 to 1.2 × 1011 cells per liter MLSS and were relatively
constant through the year (Figure 5). The number of
ammonia-oxidizing bacteria as determined using the AOB
16S rDNA and N. oligotropha amoA assays were not as
constant with a drop in AOB in the August, September and
October samples (Figure 5). Although the number of AOB
calculated using the 16S rDNA assay were about 2-fold the
values calculated using the amoA assays, the assays were
significantly correlated with a Pearsons coefficient of 0.901.
Ammonia-oxidizing activity per cell‚hour was calculated
from the AOB cell number and WWTP plant data using the
following formula:

activity )

+
(NH+ - NH+ - NH4OUT) × Q
4IN
4UP
AOB

NH4+IN is the average ammonia concentration (measured
in mg nitrogen/L) in the secondary influent into the WWTP
reactor. NH4+OUT is the average ammonia concentration in
the plant effluent. NH4+UP is used to account for ammonia
assimilation, because ammonia is removed by both assimilation into cells for cell growth and by autotrophic
ammonia oxidation by AOB, and equals 0.20×NH4+SI. Ammonia assimilation values of 0.1 and 0.2 were previously
used by Daims et al. (6, 34). In this WWTP an ammonia
assimilation value of 0.3 was estimated based on a Monte
Carlo analysis (35) using the steady-state ASM1 model (36).

TABLE 3. Conversion of Copies/L to Cells/L, Cells/g, and Percent of Biomass
target cellsa

copies/L

cells/Lb

cells/gc

% biomassd

bacteria 16S rDNA
N. oligotropha amoA
AOB 16S rDNA
Nitrospira 16S rDNA

1.6 ( 0.7 × 1012
1.5 ( 1.2 × 1010
1.2 ( 0.9 × 1010
3.7 ( 3.2 × 1010

4.3 ( 2.0 × 1011
7.5 ( 6.0 × 109
1.2 ( 0.9 × 1010
3.7 ( 3.2 × 1010

2.2 ( 0.97 × 1011
3.8 ( 3.0 × 109
6.1 ( 4.7 × 109
1.9 ( 1.6 ×1010

100
1.7
2.9
8.6

a Values are averaged for all 12 samples. b Cells/L ) copies/L ÷ gene copy number/cell. Assumed gene copy number/cell is 3.6 for bacterial
16S rDNA, 1 for Nitrospira 16S rDNA, and 2 for amoA gene. c Cells/g ) cells/L ÷ 1.97 g/L (mean mixed liquor volatile suspended solids (MLVSS)).
d Bacterial 16S rDNA provides normalization for comparison to the subpopulations and is set at 100%.

FIGURE 5. Total bacteria (b), N. oligotropha (1), AOB (3), Nitrospira
(9) cells per L MLSS in a municipal WWTP determined using the
bacterial 16S rDNA, N. oligotropha amoA, AOB 16S rDNA and
Nitrospira real-time PCR Taqman assays. Error bars indicate the
standard deviation of 3 reactions. Quantitative detection limits for
each assay were as follows: bacterial 16S rDNA, 1.0 ×1010, N.
oligotropha-like amoA, 1.5 × 109, AOB 16S rDNA 2.6 × 109 and
Nitrospira 16S rDNA, 9 × 107 cells per L.
An intermediate value of 0.2 was used for activity calculations.
Q is the average influent flowrate to the WWTP reactor (L/
hr). AOB is the total AOB cell number in the basin as
determined by real-time PCR.
Mean monthly ammonia-oxidizing activity based on the
N. oligotropha amoA cell calculations ranged from 3.5 to
56.2 fmol/hr/cell. The highest activity value, which was
greater than twice the standard deviation of the data set, was
excluded resulting in a mean activity of 12.4 ( 7.3 fmol/
hr/cell. The mean ammonia-oxidizing activity based on the
AOB 16S rDNA cell values was 7.7 ( 6.8 fmol/hr/cell.

4. Discussion
Real-time PCR using a fluorescent internal probe combines
high throughput with high analytical sensitivity for the
detection of specific genes present in low concentrations in
complex and variable DNA mixtures, such as those extracted
from environmental samples. The log-linear detection range
using standards in the real-time PCR assays was very broad,
with up to 6 orders of magnitude and detection limits of 10
to 60 copies for the Nitrospira 16S rDNA, 16S AOB rDNA, and
N. oligotropha amoA assays. These values were comparable
to other real-time PCR assays (15-20). The bacterial 16S rDNA
assay showed a log-linear detection range across 4 orders of
magnitude. The higher detection limit (4.5 × 103 copies per
PCR) in the bacterial assay was comparable to the detection
limit for other bacterial real-time PCR assays (21, 37). These
high detection limits in bacteria real-time PCR assays are
generally attributed to DNA contamination of the Taq DNA
polymerase enzyme or other sources such as water and
plasticware (37). The high detection limit in the bacterial
real-time PCR assay did not affect the use of this assay in
DNA extracted from MLSS because these samples contained
approximately 108 bacterial 16S rDNA copies per µg of DNA.

The bacterial 16S rDNA real-time PCR assay was designed
to be a broad-based assay for the estimation of total bacterial
biomass in the MLSS. This assay may also serve as an internal
standard, since bacterial 16S rDNA in the MLSS remained
relatively constant. Although the PCR primers and probes
used in this study have zero base pair mismatches with more
than 9000 bacterial 16S rDNA sequences found in GenBank,
the number of bacterial cells calculated needs to be considered an estimate because some bacterial strains/and or
species may not hybridize to the primers or probes and thus
may not be detected. The cell/L values obtained using the
real-time PCR assay were 10-fold less than those obtained
by dot-blots using a single probe. This may result from the
reduction of background due to the combined specificity
levels of the primers and probe used for the 16S rDNA assay.
In addition the probe 1392r used for the dot-blot hybridizations (9) is a universal probe that hybridizes to eukaryotic
DNA found in the MLSS and measures total microbial
population including both eukaryotic and bacterial cells. The
bacterial real-time PCR assay used in this study has not been
compared to other bacterial 16S rDNA real-time PCR assays
(21, 37, 38).
Application of PCR assays to environmental samples is
complicated by several factors such as low concentration of
targets and the possible presence of PCR inhibitors. The
inhibitory effects of complex genomic DNA at concentrations
of 10 to 50 ng/ul and other PCR inhibitors can be minimized
by dilution of the DNA extracts. However, over-dilution may
result in low total DNA and target concentrations resulting
in high variability and potential over- or under-estimation
of the target. In this study, the acceptable DNA per PCR range
varied between the assays with the bacterial 16S rDNA and
Nitrospira 16S rDNA assays producing similar results over a
broad DNA concentration range (1000 to 2500-fold). In
contrast, the N. oligotropha-like amoA and AOB 16S rDNA
assay produced consistent results over a narrow DNA
concentration range (5 to 10-fold). These results suggest that
in DNA extracts from environmental samples real-time PCR
assays may need to be performed using several dilutions.
Other factors that may affect real-time PCR assays are a loss
in fluorescence signal due to a large excess of complex DNA
or an overestimation of target DNA due to limited probe
specificity (15).
The design of specific PCR assays for Nitrospira 16S rDNA
was straightforward because regions of the Nitrospira 16S
rDNA are well conserved between Nitrospira species and
distinct from nearest phylogenetic groups (39, 40). The
primers used for the Nitrospira real-time PCR assays in this
study were previously demonstrated to be specific for
Nitrospira sp. in the cPCR assay (9). The Molecular Beacon
and TaqMan probes were designed using the same target
sequence used in the cPCR assay. The differences in values
obtained previously in individual MLSS samples using cPCR
and more recently using the Molecular Beacon and TaqMan
real-time PCR assays were statistically insignificant. This
suggests that quantitative PCR methods are robust and that
real-time PCR assays can be adapted from cPCR assays. Slight
differences in numbers obtained with cPCR and real-time
VOL. 37, NO. 2, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

349

PCR using the TaqMan probes have also been reported for
the quantification of Mycobacterium tuberculosis DNA in
sputum (41).
In contrast to the Nitrospira assay, designing a real-time
PCR assay to detect AOB is more complicated for two reasons.
First, wastewater treatment plants may contain multiple
species of AOB (28). Second, most AOB are phylogenetically
closely related to other activated sludge bacteria within the
beta-subdivision class of Proteobacteria.
Both the 16S rDNA and amoA gene provide well-studied
genetic markers for the characterization of AOB (28). DNA
probes and primers targeting AOB 16S rDNA have been used
to detect essentially all known beta-subdivision AOB species
(2, 5-7, 11, 12, 28). The 16S rDNA assay described by
Hermansson and Lindgren (20) is broad based and can
quantify a range of AOB species including N. oligotropha,
the presumptive AOB species at this WWTP. One potential
problem with probes and primers designed to target AOB
16S rDNA is that they may hybridize with closely related
non-AOB species resulting in false positives. Indeed, BLAST
analysis of the primers and probe used in the AOB 16S rDNA
assay indicated that one of the degenerate permutations of
each of the RT1r and TMP1 sequences had zero mismatches
to Ralstonia eutropha, a non-AOB species. However, in realtime PCR assays, 3 oligonucleotides (2 primers and 1 probe)
must bind for efficient amplification and detection, so the
effect of nonspecific hybridization by one primer or probe
is reduced.
The amoA gene can be used as an alternative phylogenetic
marker or target molecule for the detection of AOB (28). The
amoA gene is found only in AOB and thus serves as a more
specific marker than the 16S rDNA. However, the amoA gene
has higher sequence variability between AOB than the 16S
rDNA, thus making it more difficult to design a single assay
to detect all ammonia oxidizing bacteria. In this study, the
real-time PCR assay was designed to detect N. oligotrophalike bacteria because it was the only AOB previously found
within the municipal WWTP used in this study (9). The
primers amoNo550D2f and amoNo754r were redesigned from
the previously published cPCR primers based on additional
sequence information obtained from amoA clonal libraries
made to MLSS obtained from a bioreactor system (8). The
specificity of the primer pair was confirmed by additional
sequence analysis of clonal libraries obtained after PCR
amplification and cloning of the product. When compared
to the cPCR assay, the real-time PCR with the redesigned
primers detected about 50-fold more amoA copies in the
municipal MLSS samples. The specificity of the N. oligotropha-like amoA real-time PCR assay was confirmed by
the lack of amplification found in industrial MLSS samples
which lack N. oligotropha but contain N. nitrosa AOB (9).
Because the AOB 16S rDNA assay has the potential to
produce false positives and the N. oligotropha amoA assay
has the potential to produce false negatives, the use of the
two assays in these samples provides complimentary data
for the detection of AOB. The 16S rDNA assay and the amoA
assay were highly correlated and both assays indicated a
drop in AOB during August, September, and October
suggesting that both assays were useful for AOB quantification. The reason for the drop in AOB is unclear as the percent
nitrification remained constant and there was no apparent
link to the basin temperature. The AOB cell numbers
calculated by the 16S rDNA assay were approximately 2-fold
higher than the N. oligotropha cell numbers using the amoA
assay. Although AOB have not been cultured from this plant,
these results suggest that N. oligotropha-like AOB comprise
at least 50% of the AOB population in this WWTP.
The percent of the AOB population determined using
either the AOB 16S rDNA assay (2.9%) or the N. oligotropha
amoA assay (1.7%) were 3-4-fold lower than reported for an
350

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 2, 2003

activated sludge sample obtained from the second stage of
a 2-stage WWTP measured by fluorescence in situ hybridization (FISH) (8.4%) (6). Additionally, the AOB percent was
approximately 2-3-fold lower than reported for a sludge
sample from an industrial plant connected to a rendering
factory (7%) (7). Differences in the percent of ammoniaoxidixing bacteria between these studies may reflect the
differences in the operation and design of the WWTPs, e.g.
single-sludge nitrification versus two-stage rectors or sludge
with a high nitrogen load. Alternatively, the differences may
reflect differences in the methods, FISH versus real-time PCR.
Interestingly, the percent Nitrospira (8.6%) calculated in this
study is consistent with values obtained by FISH in the sludge
from an industrial plant connected to a rendering factory at
9% to 12% Nitrospira population (7, 40).
The ammonia-oxidizing activity per cell per hour was
calculated based on the estimated number of ammoniaoxidizing cells in the basin and the estimated amount of
ammonia oxidized per hour. The estimated ammoniaoxidizing rate of 7.7 fmol/hr/cell based on AOB 16S rDNA,
or 12.4 fmol/hr/cell based on the N. oligotropha amoA, were
in the range of values reported using FISH (2.3 ( 0.4 fmol/
hr/cell) (6), a cPCR assay (16 to 43 fmol/hr/cell) (34) and for
pure cultures (4 to 23 fmol/hr/cell) (42). Given that the
estimated ammonia-oxidizing rates are in the expected range,
it is likely that the real-time PCR assays used in this study
detects one of, if not, the major organisms mediating
ammonia oxidation in the WWTP under study.
The results of this study indicate that real-time PCR
technology is a valuable tool for quantification of uncultivable
or difficult to culture microbes in environmental samples,
offering high throughput, analytical sensitivity, and precision.
The bacterial 16S rDNA assay and the Nitrospira 16S rDNA
assays described in the study and the previously published
AOB 16S rDNA may have broad utility to other wastewater
treatment plants and environmental samples. The N. oligotropha amoA assay will be more useful in wastewater systems
where the N. oligotropha is a known member of the AOB
population. The amoA assay may also prove useful in
developing mRNA based reverse transcriptase real-time PCR
assays to measure physiological responses of N. oligotropha
to changes in environmental conditions.

Acknowledgments
This work was funded by a Water Environment Research
Foundation research grant (WERF project #98-CTS-2) and
by the University of Tennessee, Waste Management Research
and Education Institute. H.D. is a recipient of a postdoctoral
fellowship from CONICET. We thank Arthur Meyers of
Eastman Chemical Company (Kingsport, TN) for technical
advice and Neil Quigley at the Molecular Biology Resource
Facility (University of Tennessee, Knoxville, TN) for DNA
sequencing. We thank Knoxville Utilities Board (KUB,
Knoxville, TN) for providing samples.

Literature Cited
(1) Bock, E.; Koops, H.-P.; Ahlers, B.; Harms, H. In The prokaryotes,
2nd ed.; Balows, A., Truper, H. G., Dworkin, M., Harder, W.,
¨
Schleifer, K.-H., Eds.; Springer-Verlag: New York, 1992; Vol. I,
Chapter 17, pp 414-430.
(2) Okabe, S.; Satoh, H.; Watanabe, Y. Appl. Environ. Microbiol.
1999, 65, 3182-3191.
(3) Rittman, B. E.; McCarty, P. L. In Environmental Biotechnology:
Principles and Applications; McGraw-Hill Companies, Inc.: New
York, 2001; Chapter 9, pp 470-496.
(4) U.S. Environmental Protection Agency. 1993. EPA/625/R-93/
010.
(5) Wagner, M.; Rath, G.; Amann, R.; Koops, H. P.; Schleifer, K. H.
Sys. Appl. Microbiol. 1995, 18, 251-264.
(6) Daims, H.; Ramsing, N. B.; Schleifer, K.-H.; Wagner, M. Appl.
Environ. Microbiol. 2001, 67, 5810-5818.

(7) Juretschko, S.; Loy, A.; Lehner, A.; Wagner, M. Sys. Appl.
Microbiol. 2002, 25, 84-99.
(8) Dionisi, H. M.; Layton, A. C.; Robinson, K. G.; Brown, J. R.;
Gregory, I. R.; Parker, J. J.; Sayler, G. S. Water Environ. Res. 2002,
74, 462-469.
(9) Dionisi, H. M.; Layton, A. C.; Harms, G.; Gregory, I. R.; Robinson,
K. G.; Sayler, G. S. Appl. Environ. Microbiol. 2002, 68, 245-253.
(10) Johnsen, K.; Enger, Ø.; Jacobsen, C. S.; Thirup, L.; Torsvik, V.
Appl. Environ. Microbiol. 1999, 65, 1786-1789.
(11) Kowalchuk, G. A.; Naoumenko, Z. S.; Derikx, P. J. L.; Felske, A.;
Stephen, J. R.; Arkhipchenko, I. A. Appl. Environ. Microbiol.
1999, 65, 396-403.
(12) Mendum, T. A.; Sockett, R. E.; Hirsch, P. R. Appl. Environ.
Microbiol. 1999, 65, 4155-4162.
(13) Stephen, J. R.; Chang, Y.-J.; Macnaughton, S. J.; Kowalchuk, G.
A.; Leung, K. T.; Flemming, C. A.; White, D. C. Appl. Environ.
Microbiol. 1999, 65, 95-101.
(14) Zimmermann, K.; Mannhalter, J. W. BioTechniques 1996, 21,
268-279.
(15) Becker, S.; Boger, P.; Oehlmann, R.; Ernst, A. Appl. Environ.
¨
Microbiol. 2000, 66, 4945-4953.
(16) Heid, C. A.; Stevens, J.; Livak, K. J.; Williams, P. M. Genome Res.
1996, 6, 986-994.
(17) Becker, S.; Fahrbach; M., Boger; P.; Ernst, A. Appl. Environ.
¨
Microbiol. 2002, 68, 4486-4494.
(18) Bowers, H. A.; Tengs, T.; Glasgow, H. B., Jr.; Burkholder, J. M.;
Rublee, P. A.; Oldach, D. W. Appl. Environ. Microbiol. 2000, 66,
4641-4648.
(19) Gruntzig, V.; Nold, S. C.; Zhou, J.; Tiedje, J. M. Appl. Environ.
¨
Microbiol. 2001, 67, 760-768.
(20) Hermansson, A.; Lindgren, P.-E. Appl. Environ. Microbiol. 2001,
67, 972-976.
(21) Suzuki, M. T.; Taylor, L. T.; DeLong, E. F. Appl. Environ. Microbiol.
2000, 66, 4605-4614.
(22) Livak, K. J.; Flood, S. J. A.; Marmaro, J.; Giusti, W.; Deetz, K. PCR
Methods Applications 1995, 4, 357-362.
(23) Tyagi, S.; Kramer, F. R. Nat. Biotechnol. 1996, 14, 303-308.
(24) Broude, N. E. Trends Biotechnol. 2002, 6, 249-256.
(25) Ferris, M. J.; Muyzer, G.; Ward, D. M. Appl. Environ. Microbiol.
1996, 62, 340-346.
(26) Lane, D. J. In Nucleic Acid Techniques in Bacterial Systematics;
Stackebrandt, E., Goodfellow, M., Eds.; John Wiley & Sons:
Chichester, West Sussex, 1991; pp 115-148.

(27) Thompson, J. D.; Higgens, D. G.; Gibson, T. J. Nucleic Acids Res.
1994, 22, 4673-4680.
(28) Purkhold, U.; Pommerening- Roser, A.; Juretschko, S.; Schmid,
¨
M. C.; Koops, H.-P.; Wagner, M. Appl. Environ. Microbiol. 2000,
66, 5368-5382.
(29) Chandler, D. P. J. Ind. Microbiol. Biotechnol. 1998, 21, 128-140.
(30) Klappenbach, J. A.; Saxman, P. R.; Cole, J. R.; Schmidt, T. M.
Nucleic Acids Res. 2001, 29, 181-184.
(31) McTavish, H.; Fuchs, J. A.; Hooper, A. B. J. Bacteriol. 1993, 175,
2436-2444.
(32) Navarro, E.; Fernandez, M. P.; Grimont, F.; Clays-Josserand, A.;
Bardin, R. Int. J. Syst. Bacteriol. 1992, 42, 554-560.
(33) Aakra, A.; Utaker, J. B.; Nes, I. F. Int. J. Sys. Bacteriol. 1999, 49,
123-130.
(34) Daims, H.; Purkhold, U.; Bjerrum, L.; Arnold, E.; Wilderer, P. A.;
Wagner, M. Water Sci. Technol. 2000, 43, 9-18.
(35) Cox, C. D. Integration of uncertainty and water quality
considerations in the design of activated sludge plants. WEFTEC
2000 Annual Conference Proceeedings, Session 55, CD-ROM,
Water Environmental Federation, 2000.
(36) Henze, M.; Grady, C. P. L.; Gujer, W.; Marais, G. v. R.; Matsuo,
T. Activated Sludge Model No. 1. IAWPRC Scientific and Technical Report, No. 1, IAWPRC, London, 1987; ISSN: 1010-707X.
(37) Corless, C. E.; Guiver, M.; Borrow, R.; Edwards-Jones, V.;
Kaczmarski, E. B.; Fox, A. J. J. Clin. Microbiol. 2000, 38, 17471752.
(38) Lyons, S. R.; Griffen, A. L.; Leys, E. J. J. Clin. Microbiol. 2000, 38,
2362-2365.
(39) Burrell, P. C.; Keller, J.; Blackall, L. L. Appl. Environ. Microbiol.
1998, 64, 1878-1883.
(40) Juretschko, S.; Timmermann, G.; Schmid, M. C.; Schleifer, K.H.; Pommerening-Roser, A.; Koops, H.-P.; Wagner, M. Appl.
¨
Environ. Microbiol. 1998, 64, 3042-3051.
(41) Desjardin, L. E.; Chen, Y.; Perkins, M. D.; Teixeira, L.; Cave, M.
D.; Eisenach, K. D. J. Clin. Microbiol. 1998, 36, 1964-1968.
(42) Belser, L. W. Annu. Rev. Microbiol. 1979, 33, 309-333.
(43) Bustin, S. A. J. Mol Endocrinol. 2000, 25, 169-193.

Received for review April 12, 2002. Revised manuscript received October 1, 2002. Accepted October 18, 2002.
ES0257164

VOL. 37, NO. 2, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

351

