Similaire à Domestic-wastewater-influent-profiling-using-mitochondrial-real-time-PCR-for-source-tracking-animal-contamination_2008_Journal-of-Microbiological-Methods
Similaire à Domestic-wastewater-influent-profiling-using-mitochondrial-real-time-PCR-for-source-tracking-animal-contamination_2008_Journal-of-Microbiological-Methods (20)
2. contamination (Siewicki et al., 2007; Somarelli et al., 2007; Ram et al.,
2007; Yan et al., 2007).
2. Materials and methods
2.1. Sample collection and concentration
Composite samples (500 ml pooled from 24 hourly samples) of
domestic wastewater influent were collected weekly at the same day
and time for twelve weeks from two local wastewater treatment
facilities. The Town of Holly Springs Wastewater Treatment Plant
processes 1.1 million gallons per day of mostly domestic wastewater
from a population of approximately 20,000 people. Holly Springs
WWTF discharges into Utley Creek in southwestern Wake County, NC.
Besides domestic sources, Holly Springs also processes wastewater
from two light industries: a corrugated container plant that produces
plastic and cardboard waste and a microbrewery, which produces beer
locally. The South Cary Water Reclamation Plant is a biological
nutrient removal system with filtration. The plant has a treatment
capacity of 12.8 million gallons per day and services 55,000 residents.
Located on Middle Creek in southern Wake County, it processes
wastewater from homes and small, non-industrial businesses such as
offices and restaurants.
Influent samples were collected in 500 ml sterile centrifuge bottles
and concentrated by centrifugation using a Sorvall RC-5B plus
superspeed centrifuge (Thermo Electron Corp., Asheville, NC) at
9000 g for 15 min. Pellets were resuspended in remaining liquid
(~5–25 ml) after supernatant aspiration. Concentrations obtained
were approximately 6-fold as determined by DNA measurements
(260 nm) taken with NanoDrop spectrophotometer (NanoDrop
Technologies, Wilmington, DE) before and after centrifugation.
2.2. Chemical parameters
Holly Springs WWTF assayed influent for fecal coliforms (CFU/
100 ml), BOD (biological oxygen demand, mg/l), TSS (total
suspended solids, mg/l), ammonia (mg/l), total P (phosphorus,
mg/l), and total N (nitrogen, mg/l) according to Standard Methods
18th edition (AWWA et al., 1992). South Cary WWTF assayed for
alkalinity (mg/l), TSS, TVSS (total volatile suspended solids, mg/l),
ammonia, BOD, TKN (total Kjeldahl nitrogen, mg/l), NOx (nitrate+
nitrite nitrogen, mg/l), total N, total P and pH (SU) according to
Standard Methods 20th edition (AWWA et al., 1999). Both facilities
test chemical parameters from composite influent samples on a
daily basis to monitor process controls for the National Pollutant
Discharge Elimination System (NPDES, 2008), (http://cfpub.epa.
gov/npdes/home.cfm?program_id=13).
2.3. Spectrophotometric parameters
A NanoDrop spectrophotometer (NanoDrop Technologies,
Wilmington, DE) was used to assess influent samples before and
after concentration by centrifugation (initial and final readings,
respectively) (Table 3). Wavelengths of 260, 600 and 340 nm were
recorded, corresponding to DNA concentration, bacterial culture
optical density (OD600) and humic acid concentration (Lakay et al.,
2007), respectively.
2.4. DNA extraction
Influent samples were frozen at −20 °C overnight then thawed at
55 °C. One ml aliquots were heated at 99 °C, 300 rpm for 5 min. This
formed a crude DNA preparation suitable for real-time PCR.
2.5. Primer and probe design
Three species-specific primer and dual-labeled probe sets (human,
bovine, and swine) specific for amplification of mitochondrial gene
NADH dehydrogenase subunit 5 (ND5) for a multiplex real-time PCR
assay were previously published (Caldwell et al., 2007): human
forward primer (5′-CAG CAG CCA TTC AAG CAA TGC-3′), human
reverse primer (5′-GGT GGA GAC CTA ATT GGG CTG ATT AG-3′),
human probe (5′-TAT CGG CGA TAT CGG TTT CAT CCT CG-3′), bovine
forward primer (5′-CAG CAG CCC TAC AAG CAA TGT-3′), bovine
reverse primer (5′-GAG GCC AAA TTG GGC GGA TTA T-3′), bovine
probe (5′-CAT CGG CGA CAT TGG TTT CAT TTT AG-3′), swine forward
primer (5′-ACA GCT GCA CTA CAA GCA ATG C-3′), swine reverse
primer (5′-GGA TGT AGT CCG AAT TGA GCT GAT TAT-3′), and swine
probe (5′-CAT CGG AGA CAT TGG ATT TGT CCT AT-3′). The dual-labeled
probes were conjugated with Quasar 570, Cal Red, and FAM at the 5′
ends for human, bovine, and swine probe, respectively. The probe 3′
ends utilized Black Hole quenchers (BioSearch Technologies; Novato,
CA).
New species-specific primers and probes for real-time PCR
(singleplex) were developed using Primer Quest software (http://
scitools.idtdna.com/Primerquest/) for amplification of mitochon-
drial genes NADH dehydrogenase subunit 5 (ND5) for dog and cat,
ND2 for Canada goose, and cytochrome b (cytb) for white-tailed
deer (Table 1). All primers, including those previously published,
were adjusted for mismatch amplification mutation assay (Cebula
et al., 1995) utilizing penultimate primer mismatch kinetics to
increase their specificity. Primers were purchased from IDT (http://
www.idtdna.com). Dual-labeled probes were purchased from
BioSearch (www.biosearch.com) with 3′ black hole quenchers,
(BHQ), and 5′fluorophores Quasar 570, Cal Red, and Quasar 670
corresponding to Cy3, Texas Red and Cy5, respectively (Table 1). All
oligonucleotides were reconstituted in TE buffer (pH 7.5) and stored
at −20 °C prior to use.
2.6. Standard curves and assay specificity
Standard curves were generated using 10-fold serial dilutions
(107
–100
) of mtDNA copies produced from double PCR amplifications
Table 1
Mitochondrial real-time PCR primers and probes
Primer or
probea
Nucleotide sequence (5 to 3′) Tm
(°C)
Location
within
targetb
Amplicon
size (bp)
Dog forward GGCATGCCTTTCCTTACAGGATTC 58.2 1144–1167 102
Dog reverse GGGATGTGGCAACGAGTGTAATTATG 57.9 1227–1245
Dog probe TCATCGAGTCCGCTAACACGTCGAAT 61.2 1184–1209
Cat forward AACTATTCATCGGCTGAGAGGCA 58.1 416–438 143
Cat reverse GCTATGATGAAGCCTACGTCTCCAAAG 58.9 537–560
Cat probe ATGCAAACACTGCCGCCCTACAAGCAAT 64.5 488–515
Canada goose
for
CTAACATCCAAATCCCTCGACCCA 58.5 430–453 77
Canada goose
rev
TCCTATTCAGCCTCCTAGTGCTCT 58.8 484–507
Canada goose
probe
TACTCACCGCCATAGCCCTAGCCT 63.1 458–482
Deer forward TAACCCGATTCTTCGCCTTCCTCT 59.7 524–547 122
Deer reverse GTCTGCGTCTGATGGAATTCCTGAT 58.8 624–648
Deer probe CCTCCCATTTATCATCGCAGCACTTGCT 62.3 592–579
a
Primers and probes were designed using IDT Primer Questsm
program (http://scitools.
idtdna.com/Primerquest/) and adjusted for mismatch amplification mutation assay
(MAMA) (Cebula et al., 1995) in primers and species-specificity in probes. The dual-
labeled probes were conjugated at the 5′ ends with Quasar 670 for cat and deer and FAM
for dog and C. goose. The probe 3′ ends utilized Black Hole quenchers (BioSearch
Technologies; Novato, CA).
b
Positions of the oligonucleotides are listed relative to the numbering of gene ND5
(dog and cat), ND2 (C. goose) and cytb (deer) in VectorNTI version 10.1 (2005, Invitrogen
Corp). Nucleotide sequences were retrieved from GenBank™
(http://www3.ncbi.nlm.nih.
gov) under accession numbers AY656739 (dog), NC_1700 (cat), NC_007011 (C. goose), and
DQ379370 (deer).
2 J.M. Caldwell, J.F. Levine / Journal of Microbiological Methods xxx (2009) xxx–xxx
ARTICLE IN PRESS
Please cite this article as: Caldwell, J.M., Levine, J.F., Domestic wastewater influent profiling using mitochondrial real-time PCR for source
tracking animal contamination, J. Microbiol. Methods (2009), doi:10.1016/j.mimet.2008.11.007
3. of dog (ND5), cat (ND5), Canada goose (ND2) and white-tailed deer
(cytb) mitochondrial clones. Using the real-time PCR assay (Section
2.7), a composite of two identical experiments (performed on separate
days) is shown (Fig. 1). Amplicons were cloned using the TOPO TA kit
(Invitrogen, Carlsbad, CA), sequenced (Section 2.8) and NCBI BLAST
searches were performed to verify sequence identity. Standard curves
and PCR amplification efficiencies for multiplex real-time PCR of
human, bovine and swine mtDNA were published previously
(Caldwell et al., 2007).
Pet and wildlife primers and probes were tested for real-time PCR
cross-reactions with other avian, mammalian and piscine feces and
found to be species-specific (data not shown). Annealing tempera-
tures, thermal cycler background and cycle threshold levels were
optimized for each primer/probe set.
2.7. Real-time PCR
Multiplex and singleplex real-time PCR was run in 25 µl volume
reaction tubes (Cepheid; Sunnyvale, CA) with OmniMix Bead (TaKaRa
Bio Inc.; Madison, WI), (1.5U TaKaRa hot start Taq polymerase, 200 µM
dNTP, 4 mM MgCl2, 25 mM HEPES pH 8.0), forward and reverse
primers (300 nM each) and probe (320 nM) (Table 1), additional
1.5 mM MgCl2 (5.5 mM final MgCl2), 5 µl of crude DNA from influent or
known mitochondrial copies for the standard curve, and RT-PCR water
(Ambion, Austin, TX) to final volume.
Amplifications were performed in a Cepheid Smart Cycler II
thermal cycler (Cepheid, Sunnyvale, CA) with the following condi-
tions: 95 °C for 120 s, 40 cycles of 94 °C for 10 s, 60 °C for 12 s, and 72 °C
for 10 s. Four fluorophore channel optics were in use during
annealing: FAM, Cy3, Texas Red, and Cy5. This rapid real-time PCR
assay runs ca. 40 min.
No template (NTC) and positive controls (103
–105
ND5, ND2, or
cytb amplicon copies) were used for all assays. For a sample to be
considered positive, its CT value had to be 5 CT values less than all
negative control reactions and its corresponding amplification curve
had to exhibit the three distinct phases of real-time PCR: lag, linear,
and plateau. Positive signals below 100 copies per reaction for the
multiplex (human, bovine and swine) and 10 copies per reaction for
Fig. 1. Standard curves for mitochondrial real-time PCR assays. The composite of two identical experiments (performed on separate days) is shown. Mitochondrial copies of DNA
from dog (A), cat (B), Canada goose (C) and white-tailed deer (D) were diluted 10-fold in RT-PCR water (Ambion, Austin, TX) to construct the standard curves. Threshold values (CT)
were plotted against the corresponding mitochondrial copy numbers, and the linear range, the slope (m) and goodness-of-fit of the linear regression coefficient (r2
) were
determined. The PCR amplification efficiencies were 93.5, 95.5,104.4, and 105.8% for dog, cat, Canada goose and white-tailed deer, respectively. PCR efficiencies were calculated by
the formula: E=(10(− 1/slope)
)−1.
3J.M. Caldwell, J.F. Levine / Journal of Microbiological Methods xxx (2009) xxx–xxx
ARTICLE IN PRESS
Please cite this article as: Caldwell, J.M., Levine, J.F., Domestic wastewater influent profiling using mitochondrial real-time PCR for source
tracking animal contamination, J. Microbiol. Methods (2009), doi:10.1016/j.mimet.2008.11.007
4. singleplex assays (dog, cat, Canada goose, white-tailed deer) were
discarded as noise. Internal amplification controls (IAC) were
employed to check for PCR inhibitors: 103
copies of human ND5
amplicon were added to a sample aliquot and compared to the human
mitochondrial copy number standard curve or another IAC sample
with only water and master mix added.
2.8. Amplicon sequencing
Real-time PCR amplicons and mitochondrial clones of dog, cat,
Canada goose and white-tailed deer were sequenced following the
recommended protocol with the ABI BigDye v. 3.1 sequencing kit
(Applied Biosystems; Foster City, CA). Sequencing reactions were
purified using an ethanol/ammonium acetate precipitation protocol
(Irwin et al., 2003) and visualized using an ABI 3130XL Automated
Sequencer (Applied Biosystems, Foster City, CA). Sequences were
compiled in Sequencher version 4.5 (Gene Codes Corp., Ann Arbor, MI)
and NCBI BLASTsearches were performed to verify sequence identities.
2.9. Statistical analyses
Amplification efficiency (E) of the PCR assay was determined for
pet and wildlife primer/probe sets using the slope of the standard
curve: E=(10−1/slope
)−1. Data analysis of the real-time PCR standard
curves was performed using Origin software version 7.5 (OriginLab
Corp., Northampton, MA). Least squares linear regression coefficient of
determination (r2
, goodness-of-fit) and slope were used to assess the
quality of each real-time primer and probe set (Fig. 1).
Pearson correlation coefficient analysis using SAS 9.1 software was
used to compare bacterial, chemical and spectrophotometric para-
meters with human, bovine, and dog mtDNA concentrations by site
(Table 3).
3. Results
3.1. Linear range and amplification efficiency of real-time PCR assays
Standard curves were generated using serial dilutions of known
mitochondrial ND5, ND2 or cytb copies to determine the linear range
and amplification efficiencies of the real-time PCR assay for dog, cat,
Canada goose and white-tailed deer (Sections 2.5 and 2.6). Linear
ranges between 101
and 107
copies were noted for dog and cat,100
and
107
for Canada goose and white-tailed deer (Fig. 1). These are
comparable to ranges in clinical real-time PCR literature. Polymerase
chain reaction amplification efficiencies for dog, cat, Canada goose and
white-tailed deer were 94, 96, 104 and 106%, respectively (Fig. 1).
Linear regression coefficients (r2
) were 0.99 for all standard curves
except for cat (r2
=0.97). Linear range and efficiencies for multiplex
real-time PCR (human, bovine and swine) were determined pre-
viously (Caldwell et al., 2007).
3.2. PCR assay specificity for dog, cat, Canada goose and white-tailed deer
Mitochondrial clones created for standard curves of dog, cat,
Canada goose and white-tailed deer (NCBI accession numbers
EU078704–EU078707) exhibited 99, 97, 98, and 97% sequence identity,
respectively, to their species of origin when subjected to NCBI BLAST
analysis (data not shown). Amplicons (77–143 bp) were found to have
100% identity to their designated species when subjected to NCBI
BLAST analysis (data not shown). Detection of mitochondrial DNA was
species-specific for new primer/probe sets in fecal samples with no
cross-reactions with other species tested nor human, bovine, swine,
horse, sheep, goat, turkey, duck, chicken, or tilapia (data not shown).
White-tailed deer was also tested against llama with no cross-reaction.
3.3. Mitochondrial DNA detection in influents
A freeze/thaw method followed by a brief heat treatment (99 °C,
300 rpm for 5 min) was used to extract DNA from wastewater
influents. The heat treatment was found to improve detection results
in complex effluent samples. Multiplex (human, bovine and swine
together) and singleplex (dog, cat, Canada goose, and white-tailed
deer separately) real-time PCR using species-specific primers and
probes were used to detect, quantify and characterize mtDNA in
wastewater influents from two local municipal plants.
Human and dog mtDNA signals were detected in all 24 influent
samples taken over a period of approximately 6 months (Tables 2A
and 2B), although dog signals under 10 copies/reaction were not
included due to previously determined lower limits (Section 2.7),
(Caldwell et al., 2007). Also, sample 1 at the South Cary facility
exhibited only 76 human copies/reaction (Table 2B) which was below
the previously set lower limit of 100 copies for human mtDNA
detection in the multiplex assay. Mean human mtDNA copies/ml were
1.2×105
and 7.3×104
for Holly Springs (HS) and South Cary (SC)
WWTF, respectively. Mean dog mtDNA copies/ml were approximately
Table 2A
Holly Springs WWTF mitochondrial copies/ml influenta
Sample Human Bovine Swine Dog Cat Deer Canada goose
1 2.3×104
2.5×104
0 ⁎ ⁎ 0 0
2 1.3×105
5.5×104
1.5×104
⁎ 0 0 0
3 1.1×105
4.3×103
0 ⁎ 0 ⁎ 0
4 1.4×105
⁎ ⁎ 8.9×102
0 0 0
5 2.5×105
1.9×105
0 3.7×102
0 0 0
6 2.1×104
1.3×105
0 8.3×102
0 0 0
7 2.0×105
3.6×104
0 4.4×102
0 0 0
8 3.8×104
3.9×104
7.6×103
7.8×102
0 0 ⁎
9 1.7×105
8.0×104
0 7.7×102
0 ⁎ ⁎
10 1.9×105
2.5×104
0 ⁎ 0 0 0
11 8.9×104
0 0 3.1×103
0 7.9×102
0
12 3.7×104
0 0 4.3×102
0 0 3.8×102
Mean 1.2×105
4.8×104
1.9×103
6.4×102
0 6.6×101
3.1×101
SD 0.8×105
5.8×104
4.6×103
8.6×102
0 2.3×102
1.1×102
Italics = Only 1 out of 3 reps had CT N0. ⁎Positive reading below predetermined lower
limits (Section 2.7).
a
Detection of mtDNA in crude preparations of domestic and light industrial
wastewater influents taken weekly from 1/9/07 to 6/26/07. DNA was extracted using
freeze/thaw method followed by heating at 99 °C, 300 rpm for 5 min and assayed by
multiplex real-time PCR (human, bovine, swine) or singleplex real-time PCR (dog, cat,
deer, Canada goose) using species-specific primer and probe sets. All numbers corrected
by concentration factors.
Table 2B
South Cary WWTF mitochondrial copies/ml influenta
Sample Human Bovine Swine Dog Cat Deer Canada goose
1 2.2×103b
0 0 ⁎ 0 0 0
2 5.5×104
0 0 3.3×102
0 0 ⁎
3 3.4×104
6.2×103
0 5.2×102
0 0 0
4 1.4×105
0 0 4.9×102
0 0 0
5 4.5×104
0 0 4.9×102
0 0 0
6 1.7×105
1.0×105
⁎ 6.6×102
0 0 0
7 1.5×105
0 0 8.4×102
0 0 0
8 4.6×104
⁎ 0 3.4×102
0 ⁎ ⁎
9 9.1×104
⁎ 0 3.6×102
⁎ ⁎ 0
10 5.9×104
1.7×104
0 2.9×102
1.3×103
⁎ 0
11 5.5×104
2.4×103
0 5.0×102
⁎ ⁎ ⁎
12 3.4×104
1.2×103
0 1.5×102
0 ⁎ ⁎
Mean 7.3×104
1.1×104
0 4.2×102
1.1×102
0 0
SD 5.3×104
2.9×104
0 2.2×102
3.8×102
0 0
Italics = Only 1 out of 3 reps had CT N0. ⁎Positive reading below predetermined lower
limits (Section 2.7).
a
Detection of mtDNA in crude preparations of domestic and light industrial
wastewater influents taken weekly from 1/9/07 to 6/26/07. DNA was extracted using
freeze/thaw method followed by heating at 99 °C, 300 rpm for 5 min and assayed by
multiplex real-time PCR (human, bovine, swine) or singleplex real-time PCR (dog, cat,
deer, Canada goose) using species-specific primer and probe sets. All numbers corrected
by concentration factors.
b
76 copies/reaction.
4 J.M. Caldwell, J.F. Levine / Journal of Microbiological Methods xxx (2009) xxx–xxx
ARTICLE IN PRESS
Please cite this article as: Caldwell, J.M., Levine, J.F., Domestic wastewater influent profiling using mitochondrial real-time PCR for source
tracking animal contamination, J. Microbiol. Methods (2009), doi:10.1016/j.mimet.2008.11.007
5. 200 times lower than human signal at 6.4×102
and 4.2×102
,
respectively. With one exception, all human and dog replicates
(three for each sample) had CT values N0. Bovine mtDNA signal
(mean of both plants=3.0×104
copies/ml) was less consistent having 9
out of 12 and 5 out of 12 positive samples for HS and SC, respectively.
Swine mtDNA (mean=1.9×103
copies/ml for HS) was intermittent,
found in only 2 positives out of 24 total samples and in only 1/3
replicates in one of the 2 positive samples. The bovine signal was
approximately 1 and the swine approximately 2 orders of magnitude
lower than the human mtDNA signal. Cat mtDNA (1.1×102
copies/ml)
was found only once in SC and at lower levels than dog. Deer and
Canada goose signals (6.6×101
and 3.1×101
copies/ml, respectively)
each had 1 positive out of 24 samples (Tables 2A and 2B).
3.4. Correlations between mtDNA concentrations and other parameters
Human mtDNA concentration (copies/ml) was positively corre-
lated with ammonia concentration (P=0.01) and initial OD600 reading
(P=0.02) at Holly Springs WWTF only (Table 3). Bovine mtDNA was
positively correlated with biological oxygen demand (BOD) (P=0.02),
final DNA concentration (P=0.03), initial and final humic acid (P=0.01,
P=0.01), and final OD600 (P=0.03) at Holly Springs WWTF and total
suspended solids (TSS) (P=0.04, P=0.09) at HS and SC plant,
respectively. Fecal coliforms were not correlated with mtDNA of any
species assayed. No other significant (Pb0.05) correlations were
apparent in influents from the South Cary (SC) facility.
4. Discussion
Wastewater treatment systems use a variety of vital biotechnolo-
gical and natural processes to treat fecal waste and render the liquids
safe for release into local streams or land applications. If a wastewater
system fails due to environmental catastrophes, mechanical malfunc-
tions or mismanagement, methods are required to trace the location
and extent of the breach. However, when enteric organisms are
documented in surface waters, their origin is sometimes not readily
apparent or difficult to pinpoint with bacterial methods (Yan et al.,
2007). Amplification of eukaryotic mitochondrial DNA (mtDNA)
identifies the source directly and can be used after culture and
chemical methods to characterize contaminants. We used real-time
PCR to characterize and quantify mtDNA in domestic wastewater
influent (untreated sewage) using primer/probe sets for human,
bovine, swine, dog, cat, Canada goose and white-tailed deer. These
species mtDNA values provide a baseline for identifying unintentional
discharges of influent from domestic wastewater facilities and
differentiating among other mammalian effluents such as hog lagoon
wastes or dairy barn run-off. Mitochondrial DNA shed from the
epithelial cells of eukaryotic hosts can be used in conjunction with
more traditional bacterial source tracking methods to directly identify
the fecal source contaminating environmental surface waters.
The triplex mitochondrial real-time PCR assay (human, bovine,
swine) was used previously to quantify animal waste effluents
(Caldwell et al., 2007). Human signal was diluted 115-fold in domestic
wastewater: 1.1×107
copies/g feces (Caldwell et al., 2007) compared
to 9.6×104
copies/ml influent, (mean value of two WWTF influents,
Tables 2A and 2B). In the same article, carry-over mtDNA signal from
beef consumed was found in human feces at least two orders of
magnitude less than the signal for human mtDNA (2×104
and
3×105
copies/g feces). Bovine signal is similar in domestic wastewater
(mean of both plants=3.0×104
copies/ml, Tables 2A and 2B) as
compared to human carry-over mtDNA in feces. Bovine mtDNA
concentration shows no dilution in influent, as is the case with human
mtDNA. Swine mtDNA exhibited no carry-over signal in human feces
(Caldwell et al., 2007) but averaged 1.9×103
copies/ml influent in two
out of 24 samples. Therefore, real-time PCR signals from bovine and
swine mtDNA in domestic wastewater are probably primarily from
food waste flushed down the sink and not carryover in feces.
The most surprising result of these effluent profiles was the con-
sistent signal from dog mtDNA in domestic wastewater. The
Humane Society of the United States (http://www.hsus.org/pets/
issues_affecting_our_pets/pet_overpopulation_and_owership_statis-
tics/us_pet_ownership_statistics.html) cites 73 million owned dogs in
the US. The American Pet Products Manufacturers Association states
that 45% of dogs in the 2000 census were large dogs of 40 lb or more
(http://www.usatoday.com/news/science/2002-06-07-dog-usat.
htm). Therefore, the potential for large volumes of dog waste is
indisputable; but apparently many dog owners flush dog waste
down the sanitary drain. In many suburban areas, dog owners are
required to pick up their pet's feces. Toilet disposal of this waste
may account for this consistent dog mtDNA signal in domestic
influent. Cat mtDNA signal only occurred once in 24 samples.
Therefore, cat owners are not flushing cat waste as often as dog
owners, the total volume of cat feces is considerably less than
canine feces, or total cat mtDNA is below detectable levels for real-
time PCR.
Canada goose and white-tailed deer mtDNA were intended to be
negative controls for these domestic wastewater influent studies.
However, each wild species had one positive sample in HS plant
(Table 2A). Although, the goose mtDNA signal was positive in only 1
out of 3 replicates of the same sample (Table 2A). Chimeric ampli-
fication of complex DNA mixtures could give spurious results such as
these. Low, intermittent, and inconsistent (only one positive out of
three replicates) real-time PCR signals should be evaluated conser-
vatively. Due to small sampling size and complex biological back-
grounds, it can be challenging to assess environmental samples with
quantitative molecular techniques.
Human mtDNA can be attributed to human feces and possibly
bath or wash water from homes and small businesses. Bovine and
swine signal might come from human feces carry-over (Caldwell
et al., 2007) but more likely from meats, grease and animal products
washed down kitchen drains. Dog signal is not as strong as human,
bovine or swine signal, but consistent at ca. 102
copies/ml influent
and is probably a result of flushing of dog feces in household toilets
after daily walks and “accidents”. Cat signal comes from flushing
feces from the cat box; the lower amounts for cat might reflect lower
feces weights of the smaller mammals. Deer and Canada goose signal
were not expected and could be caused by chimeric PCR amplifica-
tion. For source tracking purposes, a combination of human
(105
copies/ml) and dog mtDNA signal (102
copies/ml) could be
indicative of municipal domestic wastewater contamination of
environmental waters.
Table 3
Comparison of influent parameters at Holly Springs WWTF
Mitochondrial copies/ml influent
Human Bovine
r P r P
Fecal coliforms (CFU/ml) 0.31 0.32 −0.05 0.88
BOD (mg/l) 0.09 0.79 0.66 0.02
TSS (mg/l) 0.19 0.55 0.60 0.04
Ammonia (mg/l) 0.70 0.01 0.43 0.16
Final DNA (260 nm) 0.23 0.47 0.62 0.03
Initial humic acid (340 nm) 0.43 0.17 0.68 0.01
Final humic acid (340 nm) 0.37 0.24 0.69 0.01
Initial OD600 0.65 0.02 0.15 0.64
Final OD600 0.56 0.06 0.64 0.03
r = Pearson correlation coefficient; Pb0.05 in bold type (N=12). Dog mitochondrial
copies/ml influent showed no significant positive or negative correlations. Total P, total
N and initial DNA concentration showed no significant positive or negative correlations
with either human or bovine mitochondrial copies/ml influent. Initial and final refer to
before and after centrifugation, respectively. South Cary WWTF had no correlations at
Pb0.05. BOD = Biological oxygen demand; TSS = total suspended solids.
5J.M. Caldwell, J.F. Levine / Journal of Microbiological Methods xxx (2009) xxx–xxx
ARTICLE IN PRESS
Please cite this article as: Caldwell, J.M., Levine, J.F., Domestic wastewater influent profiling using mitochondrial real-time PCR for source
tracking animal contamination, J. Microbiol. Methods (2009), doi:10.1016/j.mimet.2008.11.007
6. Other researchers (Plummer and Long, 2007) have demonstrated
statistically that one measure each of particulate matter (turbidity,
particle counts), organic matter (total organic carbon, dissolved organic
carbon, UV254 absorbance), and indicator organisms (fecal coliforms,
enterococci) was adequate for characterizing source water quality. In
this study, ammonia concentration at one WWTF exhibited a strong
positive correlation with human mtDNA concentration. Neither human
nor bovine mtDNA concentration exhibited a significant correlation
with fecal coliforms (Table 3). Yet, human mtDNA showed a strong
positive correlation to initial OD600 reading (P=0.02) and bovine mtDNA
likewise to final OD600 reading (P=0.03). We interpret this as a strong
correlation to total bacteria in the influent, since OD600 is commonly
used to quantify bacterial cell concentrations in cultures. However, it is
possible that TSS (total suspended solids) could confound the OD600
reading. Bovine mtDNA also had strong positive correlations with TSS
(P=0.04), final DNA (P=0.03), initial and final humic acid (P=0.01 for
both) while human mtDNA did not. Correlations with total suspended
solids and both humic acid readings point to discarded food products as
the primary source for bovine mtDNA. Strong ammonia and initial
OD600 correlations suggest human waste, urine and feces, respectively,
as the primary indicators or components of human mtDNA.
Humic acid concentration can be calculated from standard curves
and absorbancy readings at 340 nm (Lakay et al., 2007). We used
340 nm as a quick indicator of humic acid, an organic contaminant and
potential PCR-inhibitory compound that can co-purify with DNA. We
found no PCR inhibition as tested by internal amplification controls
and therefore could not relate inhibition to humic acid concentration.
No significant correlations between mtDNA concentrations and
other influent parameters were noted at the South Cary WWTF. This
could reflect variations in laboratory techniques or personnel, or the
influent compositions from each site.
This study provides a baseline profile of domestic wastewater
influent for mtDNA-based differentiation of sources of fecal contam-
ination. Further studies are needed to create mtDNA profiles of
residential septic systems, agricultural and wildlife sources of fecal
contamination. Comparison of these profiles with other molecular,
bacterial, chemical and spectrophotometric parameters will refine our
ability to source track fecal contamination in surface waters.
Acknowledgments
Funds supporting these studies were provided by the United States
Department of Agriculture Cooperative State Research, Education and
Extension Service (USDA CSREES); National Research Initiative
Epidemiological Approaches to Food Safety Program and the USDA
CSREES supported Food Safety Research and Response Network: a
USDA Cooperative Agricultural Project. We thank Leisha Collins and
Tony Szempruch for their assistance in the collection and processing
of influent; Amy Moore and the staff of the Town of Holly Springs
Department of Water Quality, Cecil Martin and Kelly Spainhour of the
South Cary Water Reclamation Facility for influent samples and
influent chemical data.
References
Andreasson, H., Gyllensten, U., Allen, M., 2002. Real-time DNA quantification of nuclear
and mitochondrial DNA in forensic analysis. BioTechniques 33, 407–411.
AWWA, APHA, WEF, 1992. Standard Methods for the Examination of Water and
Wastewater, 18th ed.
AWWA, APHA, WEF, 1999. Standard Methods for the Examination of Water and
Wastewater, 20th ed.
Caldwell, J.M., Raley, M.E., Levine, J.F., 2007. Mitochondrial multiplex real-time PCR as a
source tracking method in fecal-contaminated effluents. Environ. Sci. Technol. 41,
3277–3283.
Cebula, T.A., Payne, W.L., Feng, P., 1995. Simultaneous identification of strains of
Escherichia coli serotype O157:H7 and their shiga-like toxin type by mismatch
amplification mutation assay-multiplex PCR. J. Clin. Microbiol. 33, 248–250.
Gerber, A.S., Loggins, R., Kumar, S., Dowling, T.E., 2001. Does nonneutral evolutions
shape observed patterns of DNA variation in animal mitochondrial genomes? Annu.
Rev. Genet. 35, 539–566.
Irwin, D.L., Mitchelson, K.R., Findlay, I., 2003. PCR product cleanup methods for capillary
electrophoresis. BioTechniques 34, 932–936.
Iyengar, V., Albaugh, G.P., Lohani, A., Nair, P.P., 1991. Human stools as a source of viable
colonic epithelial cells. FASEB J. 5, 2856–2859.
Lakay, F.M., Botha, A., Prior, B.A., 2007. Comparative analysis of environmental DNA
extraction and purification methods from different humic acid-rich soils. J. Appl.
Microbiol. 102, 1364–5072.
Martellini, A., Payment, P., Villemur, R., 2005. Use of eukaryotic mitochondrial DNA to
differentiate human, bovine, porcine and ovine sources in fecally contaminated
surface water. Water Res. 39, 541–548.
National Pollutant Discharge Elimination System. 2008. U.S. EPA.
Plummer, J.D., Long, S.C., 2007. Monitoring source water for microbial contamination:
evaluation of water quality measures. Water Res., doi:10.1016/j.watres.2007.05.004.
Ram, J.L., Thompson, B., Turner, C., Nechvatal, J.M., Sheehan, H., Bobrin, J., 2007.
Identification of pets and raccoons as sources of bacterial contamination of urban
storm sewers using a sequence-based bacterial source tracking method. Water Res.,
doi:10.1016/j.watres.2007.04.013.
Siewicki, T.C., Pullaro, T., Pan, W., McDaniel, S., Glenn, R., Stewart, J., 2007. Models of total
and presumed wildlife sources of fecal coliforms bacteria in coastal ponds. J.
Environ. Manag. 82, 120–132.
Somarelli, J.A., Makarewicz, J.C., Sia, R., Simon, R., 2007. Wildlife identified as major
source of Escherichia coli in agriculturally dominated watersheds by BOX A1R-
derived genetic fingerprints. J. Environ. Manag. 82, 60–65.
Yan, T., Hamilton, M.J., Sadowsky, M.J., 2007. High-throughput and quantitative
procedure for determining sources of Escherichia coli in waterways by using host-
specific DNA marker genes. Appl. Environ. Microbiol. 73, 890–896.
6 J.M. Caldwell, J.F. Levine / Journal of Microbiological Methods xxx (2009) xxx–xxx
ARTICLE IN PRESS
Please cite this article as: Caldwell, J.M., Levine, J.F., Domestic wastewater influent profiling using mitochondrial real-time PCR for source
tracking animal contamination, J. Microbiol. Methods (2009), doi:10.1016/j.mimet.2008.11.007