w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 4 8 5 0 – 4 8 5 9

Avai lab le at www.sc iencedi rect .com

journa l homepage : www.e lsev i er . com/ loca te /wat res

Rapid decay of host-specific fecal Bacteroidales cellsin seawater as measured by quantitative PCRwith propidium monoazide

Sungwoo Bae, Stefan Wuertz*

Department of Civil and Environmental Engineering, 2001 EU III, University of California, Davis, One Shields Avenue, Davis, CA 95616, USA

a r t i c l e i n f o

Article history:

Received 20 March 2009

Received in revised form

22 June 2009

Accepted 28 June 2009

Published online 3 July 2009

Keywords:

Persistence and survival

Host-specific Bacteroidales

Microbial source tracking

Fecal pollution

Propidium monoazide

* Corresponding author. Tel.: þ1 530 754 640E-mail address: swuertz@ucdavis.edu (S.

0043-1354/$ – see front matter ª 2009 Elsevidoi:10.1016/j.watres.2009.06.053

a b s t r a c t

We investigated the persistence of feces-derived Bacteroidales cells and their DNA in

seawater under natural conditions using an optimized chemical method based on co-ex-

traction of nucleic acids with propidium monoazide (PMA), which interferes with PCR

amplification of molecular markers from extracellular DNA and dead bacterial cells. The

previously validated Bacteroidales assays BacUni-UCD, BacHum-UCD, BacCow-UCD, and

BacCan-UCD were utilized to determine concentrations of Bacteroidales genetic markers

targeting all warm-blooded animals, humans, cows and dogs, specifically, over a period of

24 d. Microcosms containing mixed feces in dialysis tubing were exposed to seawater

under flow-through conditions at ambient temperature in the presence and absence of

sunlight. Using a two-stage plus linear decay model, the average T99 (two-log reduction) of

host-specific Bacteroidales cells was 28 h, whereas that of host-specific Bacteroidales DNA

was 177 h. Natural sunlight did not affect the survival of uncultivable Bacteroidales cells and

their DNA with the exception of the BacCow-UCD marker. Bacteroidales DNA, as measured

by quantitative PCR (qPCR) without PMA, persisted for as long as 24 d at concentrations

close to the limit of detection. Culturable Enterococcus cells were detected for only 70 h,

whereas Enterococcus cells measured by qPCR with and without PMA persisted for 450 h. In

conclusion, measuring Bacteroidales DNA without differentiating between intact and dead

cells or extracellular DNA may misinform about the extent of recent fecal pollution events,

particularly in the case of multiple sources of contamination with variable temporal and

spatial scales due to the relatively long persistence of DNA in the environment. In contrast,

applying qPCR with and without PMA can provide data on the fate and transport of fecal

Bacteroidales in water, and help implement management practices to protect recreational

water quality.

ª 2009 Elsevier Ltd. All rights reserved.

1. Introduction inadequate to identify the source of fecal contamination

For more than 100 years, fecal indicator bacteria (FIB) such as

total and fecal coliforms, E. coli and Enterococcus, have been

used to assess recreational water quality. However, FIB are

7; fax: þ1 530 752 7872.Wuertz).er Ltd. All rights reserved

because they exist in the feces of a variety of warm-blooded

animals; they are far less prevalent in cold-blooded animals

(Field and Samadpour, 2007). Microbial source tracking (MST)

is a rapidly emerging approach to discriminate between

.

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 4 8 5 0 – 4 8 5 9 4851

human and non-human fecal contamination from various

livestock and wildlife sources in water using microbiological

traits of source identifiers (Field and Samadpour, 2007; Santo

Domingo et al., 2007). Among MST techniques, cultivation-

and library-independent methods such as qPCR assays tar-

geting host-specific molecular makers have been highlighted

because of their simplicity, sensitivity, specificity, and ability

to generate quantitative data (Santo Domingo et al., 2007).

However, without information regarding the persistence of

source identifiers in the environment, it is difficult to develop

effective management and implementation plans for recrea-

tional waters.

Members of the order of Bacteroidales have been proposed

both as an alternative to the fecal indicator bacteria Entero-

coccus and E. coli and as an MST identifier because of their

abundance in the gastrointestinal tract with host-specific

distributions (Dick et al., 2005; Seurinck et al., 2005; Layton

et al., 2006; Reischer et al., 2006; Kildare et al., 2007). Previous

decay studies have been laboratory-based and the DNA of

fecal Bacteroides was detected by PCR for days and even weeks

in freshwater incubating in a batch reactor, whereas cultur-

able Bacteroides fragilis cells could survive only a few hours

(Kreader, 1998; Savichtcheva and Okabe, 2006). More recently,

the persistence of host-specific Bacteroides genetic markers

was examined at different incubation temperatures and

salinities (Seurinck et al., 2005; Okabe and Shimazu, 2007; Bell

et al., 2009). However, the fate of MST identifiers in water was

influenced by a complex mix of physical, chemical, and bio-

logical factors that are difficult to simulate in the laboratory

(Anderson et al., 2005). Consequently, even though the

persistence of some Bacteroidales genetic markers was inves-

tigated in laboratory experiments, little is known regarding

the survival and persistence of fecal Bacteroidales cells and

their DNA in the natural environment.

Enumeration of non-cultivable host-specific Bacteroidales

cells in feces was previously attempted using bromodeox-

yuridine (BrdU) DNA labeling and immunocapture followed by

PCR, but the technique was not suitable to evaluate the

persistence of Bacteroidales cells because of high detection

limits (Walters and Field, 2006). In recent times, an optimized

quantitative PCR method with propidium monoazide (PMA-

qPCR) successfully discriminated between viable (or intact)

and dead Bacteroidales cells (Bae and Wuertz, 2009). Propidium

monoazide, a DNA-intercalating dye, will enter only dead cells

with a compromised cell envelope and then bind to DNA, or

attach to extracellular DNA (Nocker et al., 2006). Upon expo-

sure to bright visible light, the photoactive azide group on the

dye is converted to a nitrene radical, which is readily attached

to a carbon atom of the DNA through a C–H insertion, resulting

in an inhibition of amplification of DNA from dead cells and

free DNA. Residual unbound PMA is simultaneously trans-

formed to a hydroxylamine, which is incapable of covalently

binding to DNA (Nocker et al., 2006, 2007a; Bae and Wuertz,

2009).

The objective of the present study was to determine the

survival and persistence in seawater of host-specific fecal

Bacteroidales cells and their DNA using PMA-qPCR, because

fecal Bacteroidales cells cannot be cultivated at present. Repli-

cate microcosms were constructed in flowing seawater and

incubated outdoors to simulate environmental conditions by

allowing exposure to diurnal cycles of natural sunlight and

ambient temperature. We also tested the hypothesis that

sunlight was the single most important contributor to survival

of Bacteroidales cells and their DNA because natural sunlight is

considered to greatly affect bacterial die-off in marine waters.

Decay rate models were developed for both host-specific

Bacteroidales cells and their DNA in seawater.

2. Materials and methods

2.1. Fecal sample collection and preparation of inocula

Individual fecal samples were collected with a sterile utensil

and placed in a sterile 50-mL tube. Human fecal samples were

donated from six healthy adults and one infant. A total of 15

fecal samples were obtained from a cattle farm (4 samples;

Winters, CA) and a dog park (11 samples; Davis, CA). Each fecal

sample was immediately transported on ice to the laboratory.

Ten milligrams of feces was diluted with 1� PBS solution, and

serially diluted samples were analyzed by qPCR with host-

specific Bacteroidales assays (Kildare et al., 2007) as described

below to verify the specificity of each assay. Once the speci-

ficity had been confirmed, 500 mg of each fecal sample from

different hosts was mixed with 30 ml of 1� sterile PBS solu-

tion. The mixed samples were immediately shipped on ice to

the Bodega Bay Marine Laboratory and then combined with

2 l of seawater before distribution into dialysis tubing.

2.2. Microcosm study design

Microcosms were designed to simulate environmental condi-

tions by allowing exposure to ambient insolation, seawater

temperature, salinity and other factors characteristic of marine

environments. Experiments were conducted at outdoor facilities

at the Bodega Bay Marine Laboratory (38�19.1100 N 123�04.2940

W). All microcosms were placed in a rectangular tank with

dimensions of 80 cm by 45 cm by 30 cm. The rectangular tank

was designed to be exposed to diurnal cycles of insolation with

all dialysis tubes receiving equivalent amounts of sunlight.

Tubes were flatly spread across the bottom of the tank and tubes

were randomly selected during sampling events.

Fresh seawater was continuously introduced into the tank

at a flow rate of a 1.9 l/min. The seawater exiting from the tank

was disinfected by UV irradiation before discharging into the

Pacific Ocean. Temperature, salinity, conductivity, light

intensity, meteorological, and oceanographic data of the

seawater were obtained from the Bodega Ocean Observing

Node for experimental periods from 10 October to 15

December 2007. (http://www.bml.ucdavis.edu/boon/index.

html). In order to check for the presence of host-specific Bac-

teroidales prior to the microcosm experiments, a 100-ml

seawater sample was subjected to direct DNA extraction using

the Ultraclean� Water DNA Isolation kit (Mo Bio Laboratories

Inc., Carlsbad, CA).

Microcosms were employed to monitor the decay rates of

culturable fecal indicator bacteria and host-specific Bacter-

oidales cells and their DNA. A ready-to-use Spectra/Por

7 membrane dialysis tubing (MWCO 50000, 28 mm flat width;

Spectrum Laboratories, Inc., Rancho Dominguez, CA) was cut

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 4 8 5 0 – 4 8 5 94852

into pieces of 8-cm length. The transparent dialysis tubes

were permeable to UV and visible light and contained the

same diluted mixture of feces from three hosts as described in

Section 2.1. One half of the 66 individual dialysis tubes were

placed inside a 6.4 cm long PVC pipe shielding them from

sunlight. The remaining dialysis tubes were exposed to

sunlight at a diurnal cycle. Triplicates of the dialysis tubes for

both light and dark conditions were randomly sampled for

24 days and immediately shipped on ice to the University of

California, Davis, for further analysis.

2.3. Enumeration of fecal indicator bacteria

Procedures for enumeration of Enterococcus were carried out

using membrane filtration techniques (USEPA, 2002).

Preliminary experiments were conducted to determine the

initial concentrations of Enterococcus using 100 ml of

seawater obtained at the Bodega Bay Marine Laboratory. For

the microcosms, between 5 ml and 10 ml of fecal sample

from dialysis tubes were diluted with a sterile buffer solu-

tion and filtered through 47-mm diameter sterile membrane

filter with a 0.45 mm pore size (Millipore Corp., Bedford,

Mass). The filters were subsequently placed on mEI agar and

MI agar plates. Triplicate plates were then incubated at

41 �C for 24 h. The TaqMan� qPCR assay for Enterococcus

targeting the 23S rRNA gene was used as previously

described in (Haugland et al., 2005) after optimizing primer

and probe concentrations at 600 nM and 200 nM, respec-

tively. The assay limit of detection (ALOD) was 10 gene copies

per reaction and calculated as previously described (Kildare

et al., 2007). Otherwise, qPCR procedures were performed as

explained in Section 2.7.

2.4. PMA treatment

PMA (Biotium Inc., Hayward, CA, USA) was dissolved in 20%

DMSO to create a stock concentration of 2 mM and stored in the

dark at �20 �C. Exposure time and PMA concentration were in

accordance with previous optimization experiments for fecal

Bacteroidales (Bae and Wuertz, 2009). PMA was added to

samples at a final concentration of 100 mM. Following a 5-min

incubation in the dark, samples were exposed for 10 min with

a 600-W halogen light source at a distance of 20 cm from the

light source. Microcentrifuge tubes were placed on ice prior to

light exposure to avoid excessive heating.

2.5. Nucleic acid extraction

A 400-ml aliquot from a microcosm was extracted using the

FastDNA� spin kit for soil (Biomedicals, Solon, OH, USA).

Cell lysis was achieved by bead beating using a bead mill

Minibead beater� (Biospec Products Inc., Bartlesville, OK,

USA) at 2400 rpm for 20 s. Cell debris was removed by

centrifugation for 5 min at 14,000 � g before adding 250 ml of

PPS solution (provided with the kit). DNA was eluted twice

with a volume of 80 ml DES buffer (provided in the kit).

Otherwise, DNA extraction was performed according to the

manufacturer’s instructions. For fresh seawater samples,

the UltraClean� Water DNA Isolation kit (Mo Bio Laborato-

ries Inc., Carlsbad, CA) was used with a 0.22-mm sterile filter

(provided with the kit) according to the manufacturer’s

protocol.

2.6. PCR amplification of host-specific Bacteroidales

The real-time 50-nuclease PCR assays for Universal

(BacUni-UCD), mixed human-specific (BacHum-UCD), dog-

specific (BacCan-UCD), and cow-specific (BacCow-UCD)

Bacteroidales were used to quantify host-specific Bacter-

oidales markers in microcosms and seawater samples.

Each 25-ml PCR reaction contained 12.5 ml of qPCR Master-

Mix Plus (Eurogentec, San Diego, CA) with 400 nM forward

and reverse primers and 80 nM probe. Ten microliters of

DNA extracts was assayed in 96-well plates in a ABI PRISM

7700 Sequence Detection system (Applied Biosystems,

Foster City, CA). The thermal protocol for Bacteroidales

consisted of 2 min at 50 �C, followed by 10 min at 95 �C,

and 40 cycles of 15 s at 95 �C and 60 s at 60 �C. Each sample

was analyzed in duplicate and a serial dilution approach

was employed for each duplicate to alleviate the effects of

potentially present PCR inhibitors.

The limit of detection (LOD) was calculated using the

following equation.

LOD ¼�

ALOD

VT

��Vel

Vex

�(1)

Here, V (ml) refers to the volume of sample that was extracted

(Vex), the volume of eluate from the nucleic acid extraction

(Vel), and the volume of nucleic acid template added to the PCR

reaction (VT). The assay limits of detection (ALOD) were previ-

ously reported as 2, 3.5, 3.7 and 1 gene copies per reaction for

the universal, human-specific, cow-specific and dog-specific

Bacteroidales genetic markers, respectively (Kildare et al.,

2007).

2.7. Decay rate calculations

The decay rates of culturable Enterococcus cells were calculated

by a firstorder decay model (N ¼ N0 � exp�k1t). A two-stage

exponential-plus-linear model was used to fit the experi-

mental data and to calculate the decay rates as well as T99

(time for 2-log reduction) for Bacteroidales.

The model equation is

N ¼ N1 � exp�k1t þN2 � exp�k2t � k3 � tþ y0 (2)

where N is the number of gene copies per milliliter of host-

specific Bacteroidales at time t; the sum of N1 and N2 is the

initial concentration of host-specific Bacteroidales; y0 is

a constant; and k1, k2, and k3 are the decay rate constants.

2.8. Statistical analysis

Statistical tools provided in MINITAB� version 15 were

used to perform ANOVA tests at each sampling point. All

standard deviations were calculated from three indepen-

dent replicates. The effects of light/dark and PMA/non-

PMA of the variables tested were analyzed by two-way

ANOVA. The non-linear regression analysis and fitting of

persistence data was performed using Origin Pro 8�. To

Fig. 1 – Decay curves for culturable Enterococcus cells and the Enterococcus 23S rRNA gene measured by qPCR with and

without PMA from mixed stool samples in seawater. All standard deviations were calculated from three independent

replicate microcosms. (A) The number of copies of the Enterococcus 23S rRNA gene measured by qPCR and (B) culturable

Enterococcus cells under natural sunlight exposure. (C) The number of copies of the Enterococcus 23S rRNA gene measured by

qPCR and (D) culturable Enterococcus cells under dark conditions. The white circles represent Enterococcus cells measured by

PMA-qPCR, whereas dark circles represent total DNA including intact cells, dead cells and extracellular DNA.

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 4 8 5 0 – 4 8 5 9 4853

statistically analyze the two-stage exponential-plus-linear

model, we compared data sets by F-test using Origin

Pro 8�.

3. Results

A microcosm study was conducted to estimate the persistence

of fecal indicator bacteria and host-specific Bacteroidales as

identifiers of microbial source tracking (MST) in seawater.

Characteristic meteorological and seawater conditions were

simulated by conducting experiments outdoors with contin-

uously flowing seawater, natural sunlight, ambient tempera-

ture variance, predators, etc,. The average salinity, pH and

dissolved oxygen values throughout the study were 31.5 ppt,

7.5, and 6.5 mg/l, respectively. The temperature decreased

from 14.8 �C to 13.5 �C during the study, which was not

a significant change. Solar radiation intensity of natural

sunlight varied from 0 to 700 W/m2 depending on the presence

and absence of natural sunlight. The background concentra-

tion of the fecal indicator Enterococcus in seawater flowing

through the water tank was less than 2 CFU/100 ml. The

universal fecal Bacteroidales gene marker (BacUni-UCD) was

detected at concentrations of 30 gene copies/ml in flowing

seawater and human-, cow-, and dog-specific Bacteroidales

were not found.

3.1. Persistence of Enterococcus measured by PMA-qPCR, PCR and culture techniques in seawater

Culturable cells of enterococci from dialysis tubes in the

microcosms were enumerated to assess the persistence of the

traditional indicator organism. They were only cultured

during the first 72 h because of the limited volume available in

dialysis tubes (Fig. 1). Data for enumerated enterococci in the

seawater were fitted with an exponential decay model.

Enterococci decreased with decay constants of 0.17 h�1 and

0.10 h�1 under natural sunlight and dark conditions with R2

values of 0.70 and 0.56, respectively. The decay curves for

Enterococcus under natural sunlight and dark conditions were

statistically different using the F-test ( p-value <0.001). Hence,

the culturable Enterococcus cells were susceptible to natural

sunlight, as would be expected from previous studies.

However, based on both PMA-qPCR and qPCR, intact cells and

DNA were more persistent than were culturable Enterococcus

cells. The number of copies of the Enterococcus 23S rRNA gene

Ta

ble

1–

Kin

eti

cp

ara

mete

rso

fex

po

nen

tia

l-p

lus-

lin

ea

rm

od

els

ap

pli

ed

toh

ost

-sp

eci

fic

Ba

ctero

ida

les

DN

Aa

nd

cell

deca

y.

Ty

pe

Ho

st-s

peci

fic

Ba

ctero

ida

les

DN

AC

ell

N0

(gc/

ml)

k 1(/

h)

N1

(gc/

ml)

k 2(/

h)

k 3(/

h)

y 0(g

c/m

l)R

2v

alu

eN

0(g

c/m

l)k 1

(/h

)N

1(g

c/m

l)k 2

(/h

)k 3

(/h

)y 0

(gc/

ml)

R2

va

lue

Lig

ht

Un

ivers

al

3.5

9�

10

70.1

16

5.3

7�

10

60.0

15

0.5

856

0.7

61.9

2�

10

60.1

92

1.9

2�

10

60.1

92

0.4

5613

0.6

2

Hu

ma

n7.0

4�

10

60.0

81

1.1

8�

10

60.0

13

1609

0.7

11.4

5�

10

60.1

75

5.8

0�

10

40.2

12

0.0

4745

0.6

4

Co

w5.2

2�

10

50.1

11

5.2

2�

10

46.0

927.9

1.0

2�

10

40.9

32.9

5�

10

40.1

53

00

034

0.9

9

Do

g3.5

5�

10

50.0

98

9.1

0�

10

40.0

15

00

0.5

76.8

0�

10

40.2

01

538

0.2

15

073.5

0.5

4

Da

rkU

niv

ers

al

3.7

1�

10

72.0

55

8.1

3�

10

60.0

2250

1.9�

10

50.9

25.1

7�

10

60.1

24

9.0

3�

10

30.0

12

0.4

87

379

0.8

5

Hu

ma

n4.6

2�

10

61.0

83.2

4�

10

60.0

23

0.8

6062

0.8

21.4

4�

10

60.5

93

2.3

1�

10

40.0

16

0.0

8211

0.7

9

Co

w4.5

7�

10

54.7

6.4

0�

10

40.0

14

0.2

9183

0.8

62.9

5�

10

40.1

84

00

0.1

01

52.8

0.9

9

Do

g4.4�

10

50.1

01

5.9�

10

40.0

14

00

0.9

26.8

1�

10

40.1

89

273

0.0

09

00

0.8

9

gc/

ml,

gen

eco

pie

sn

um

ber/

ml.

N¼

N1�

ex

p�

k 1tþ

N2�

ex

p�

k 2t�

k 3�

tþ

y 0w

here

Nis

the

gen

eco

pie

sn

um

ber/

mlo

fh

ost

-sp

eci

fic

Ba

ctero

ida

les

at

tim

et,

N1

an

dN

2a

reth

ein

itia

lco

nce

ntr

ati

on

of

ho

st-

speci

fic

Ba

ctero

ida

les,

y 0is

con

sta

nt

an

dk 1

,k 2

,a

nd

k 3a

reth

ed

eca

yra

teco

nst

an

ts.

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 4 8 5 0 – 4 8 5 94854

rapidly decreased by one log and then persisted for the

remainder of the experiment. Natural sunlight did not affect

the survival of Enterococcus cells and their DNA (Fig. 1A and C).

3.2. Determination of the decay rates and T99 values ofhost-specific Bacteroidales

We used PMA-qPCR to discriminate between intact cells and

total DNA including viable cells, dead cells, and extracellular

DNA in dialysis tubes submerged in seawater. The dialysis

tubes were placed outdoors in a rectangular tank to simulate

natural conditions including the presence of dissolved

oxygen, natural light, temperature variance and others. The

exponential-plus-linear model (Eq. (2)) was used to fit the

decay function of host-specific Bacteroidales when comparing

kinetic parameters for cell-bound and total DNA molecular

markers (Table 1). The decay functions for each host-specific

Bacteroidales marker were nearly identical under natural

sunlight and dark conditions and exposure to natural

sunlight had no effect based on the F-test ( p-value >0.05).

The only exception was the persistence of the cow-specific

Bacteroidales DNA in natural sunlight and dark conditions,

which was significantly different based on the F-test ( p-value

<0.01). The cow-specific Bacteroidales DNA persisted longer in

dark conditions than in natural sunlight conditions.

T99 (2-log reduction) values of host-specific Bacteroidales

were calculated based on the exponential-plus-linear model.

T99 values for total DNA were similar under natural sunlight

and dark conditions for each host-specific Bacteroidales

assay: for natural sunlight exposure the T99 values were

170 h (universal), 180 h (human), 160 h (cow), and 193 h

(dog), and for dark conditions they were 167 h (universal),

172 h (human), 182 h (cow), and 183 h (dog). As expected, the

DNA was more persistent over the study period than were

intact cells. Statistically significant differences between

intact cell detection and DNA detection were observed

( p-value <0.001) in both sunlight exposure and dark condi-

tion. The T99 values for intact cells of Bacteroidales were

much shorter for both light and dark conditions, ranging

from 24 h to 31 h in natural sunlight and from 26 h to 38 h in

the dark (Figs. 2 and 3).

Both universal and human-specific Bacteroidales cells and

their DNA were detectable during the entire experimental

period. In contrast, under light and dark conditions cow- and

dog-specific Bacteroidales cells were not evident after 140 h and

260 h, respectively (Figs. 2 and 3).

3.3. Estimation of fecal pollution ages

Knowledge about the approximate age of fecal pollution

detected by qPCR would be of great benefit to water quality

regulators and coastal managers alike. We describe a simple

equation based on the ratio of threshold values obtained from

qPCR runs (Ct values in samples with and without PMA

treatment) to estimate the age of contamination by different

fecal sources (Fig. 4).

Age of fecal pollution ¼ CtPMA

Ctcycle � Ctwithout PMA(3)

Fig. 2 – Decay curves for host-specific Bacteroidales cells measured by PMA-PCR and for total DNA including intact/dead cells

and extracellular DNA under natural sunlight exposure in seawater. All standard deviations were calculated from three

independent replicate microcosms. White squares represent intact cells and black squares the total Bacteroidales DNA

including intact cells, dead cells and extracellular DNA. The dotted line and the dash-dotted line were plotted by the two-

stage exponential-plus-linear model as shown in Table 1.

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 4 8 5 0 – 4 8 5 9 4855

where CtPMA ¼ Ct number in PMA treatment, Ctcycle ¼ total

number of cycles of PCR amplification (e.g., 40 cycles), and

Ctwithout PMA ¼ Ct number without PMA treatment. The initial

ratios for universal, human-, dog-, and cow-specific Bacter-

oidales at the beginning of the decay experiment were 1.00,

1.49, 1.51 and 1.97, respectively. Over time, the ratios

increased. At the end of the microcosm experiment, the ratios

for host-specific Bacteroidales assays ranged from 3.66 to 27.55.

Under dark conditions, a linear relationship was observed for

ratios and time during the entire microcosm exposure except

for cow-specific Bacteroidales, which had a linear ratio only for

350 h. The ratios for each host-specific Bacteroidales assay

under both natural light and dark conditions were similar for

300 h after which time the ratios for microcosms exposed to

natural sunlight increased rapidly (Fig. 4).

4. Discussion

4.1. Persistence of Enterococcus

The factors affecting the survival of enteric bacteria in marine

and freshwater environments are well studied. Both abiotic and

biotic factors can influence inactivation of enteric bacteria.

Abiotic factors include sunlight, salinity, temperature, pH,

nutrient deprivation, temperature, hydrostatic pressure, sedi-

mentation and others (Anderson et al., 1979, 1983; Rhodes and

Kator, 1988; Sinton et al., 1999; Rozen and Belkin, 2001; Craig

et al., 2004; Whitman et al., 2004; Anderson et al., 2005); biotic

factors comprise grazing and competition, bacteriophages,

antibiotics and toxins (Rozen and Belkin, 2001; Whitman et al.,

2004). Among the abiotic factors sunlight and temperature

appear to be critical for survival of enteric bacteria in seawater

and freshwater. The variance in temperature during the entire

experimental period in our study was insignificant. The effect of

sunlight was in accordance with previous reports that natural

sunlight adversely affects the cultivability of fecal indicator

bacteria. The survival of Enterococcus cells in our study was

significantly different under light and dark conditions ( p-value

<0.001) and rapid decreases were observed in the presence of

sunlight (Fig. 1). Unfortunately, we could investigate the

persistence of culturable fecal indicators only for a few days due

to the relatively small sample volume in dialysis tubes

(approximately 10 ml). Our finding that light had no effect on

the decay of the Enterococcus 23S rRNA gene as measured by

qPCR is consistent with recent work reporting no statistical

difference in the decay rates of this marker in light- and dark-

exposed sewage microcosms (Walters et al., 2009).

Fig. 3 – Decay curves for host-specific Bacteroidales cells measured by PMA-PCR and for total DNA including intact/dead cells

and extracellular DNA under dark conditions in seawater. White squares represent intact cells and black squares the total

Bacteroidales DNA including viable cells, dead cells and extracellular DNA. The dotted line and the dash-dotted line were

plotted by the two-stage exponential-plus-linear model as shown in Table 1.

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 4 8 5 0 – 4 8 5 94856

A rapid decrease in intact cells of Enterococcus and their

DNA was observed by qPCR at the beginning of the experiment

(Fig. 1) but both later persisted during the course of the

experiment. Interestingly, the decay of the Enterococcus 23S

rRNA gene based on qPCR alone and qPCR with PMA was very

similar, indicating that either the PMA-qPCR method was not

working well in the case of Enterococcus or cell envelope

integrity was largely maintained, and hence most of the rRNA

gene copies were present in intact cells. Cells can remain

metabolically active in the environment without being able to

divide in or on nutritional media, and this was demonstrated

for many enteric bacterial species during seawater incubation

in the dark and in the light (Rozen and Belkin, 2001).

4.2. Persistence of Bacteroidales

To date the host-specific Bacteroidales cells that are targeted by

the genetic markers have not been isolated and cultivated. To

our knowledge, this is the first study to differentiate between

fecal Bacteroidales cells and DNA present in dead cells or

extracellular DNA to assess the decay rates of host-specific

genetic markers. The PMA-qPCR method provides a relatively

fast, simple and inexpensive means to discriminate intact/

dead cells in mixed populations (Nocker et al., 2007a; Bae and

Wuertz,2009). One of the primary defining features of our study

was the use of a continuous flow-through system situated

outdoors. In situ experiments were believed to better simulate

environmental conditions compared to a laboratory batch

setup where nutrients become depleted and waste products

accumulate during the course of the study.

The quantitative nature of host-specific Bacteroidales data

obtained by qPCR or PMA-qPCR can in principle lead to an

estimation of the relative contribution of each source.

Because qPCR assays are capable of detecting a few copies of

a gene per reaction the associated limit of detection (LOD)

values provide a measure of significance of detects. If

concentrations are low and close to the LOD they may be

discounted for analytical reasons or represent background

levels found in a particular watershed. For example, in the

microcosm study, cow- and dog-specific Bacteroidales cells

were not detected by PMA-qPCR after 300 h, while universal

and human-specific Bacteroidales cells remained present, but

hovered just above the LOD, throughout the experiment

(24 d). Host-specific Bacteroidales DNA was also detectable

during the course of the experiment, except in the case of

cow-specific Bacteroidales DNA.

The effect of natural sunlight, which has long been recog-

nized as the most lethal factor in the survival of enteric bacteria,

on the persistence of host-specific Bacteroidales cells and their

DNA has not been studied before. In this study, exposure to

Fig. 4 – Ratio of Ct values with and without PMA treatment for host-specific Bacteroidales calculated by the equation

(ratio[CtPMA=CtcycleLCtwithout PMA) as a function of time during the decay experiment. White circles and black squares

represent natural sunlight exposure and dark conditions, respectively. The ratio can be used to estimate the time of a fecal

pollution event.

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 4 8 5 0 – 4 8 5 9 4857

natural sunlight did not significantly influence the persistence

of Bacteroidales cells and their DNA ( p-value>0.05), except in the

case of cow-specific Bacteroidales DNA, which was no longer

detected after 350 h of sunlight exposure. This result contrasts

with a concurrent study where the Bacteroidales human-specific

marker BacHum-UCD as measured without PMA decayed faster

in light microcosms (Walters et al., 2009). The disagreement

between the two studies could be attributed to the experimental

setup of microcosms (continuous flow system vs. batch system)

and the type of inoculum (feces vs. sewage). Short-wave solar

radiation (UV-B) damage to the bacterial cell primarily results

from direct photochemical mechanisms, leading to the forma-

tion of DNA photoproducts of which cyclobutane pyrimidine

dimer and pyrimidine (6–4) pyrimidinone are the most common

(Sinton, 2006). Therefore, the PMA-qPCR method seems to be

able to detect even Bacteroidales cells that have been damaged by

UV-B exposure, because UV-B did not directly lead to a break-

down of membrane integrity allowing PMA uptake. Similarly,

others reported that loss of viability after short UV exposure

could not be monitored with PMA as UV light affects viability by

inducing DNA damage without directly affect membrane

permeability (Nocker et al., 2007b). In the same vein, natural

sunlight did not affect the decay of Enterococcus cells measured

by PMA-qPCR in our study. The most significant parameter

affecting the persistence of anaerobic Bacteroidales cells in

microcosms was most likely the presence of dissolved oxygen in

water.

A recent study using a non-quantitative PCR assay to detect

Bacteroides suggested that Bacteroides markers were a reliable

source identifier only on high flow samples with higher

concentrations of culturable fecal indicators and could not

describe precise discrimination between livestock- and

human-derived feces in the larger land use pattern (Stapleton

et al., 2007). It is plausible that a genetic marker would be less

predictive in identifying the source of fecal pollution when

extracellular DNA and dead cells are also detected. DNA can

persist in marine water, freshwater, and sediment for up to

55 days, up to 21 days, and 40 days, respectively (Nielsen et al.,

2007). Even though dissolved DNA serving as nutrient sources

can be decomposed by either chemical or enzymatic reaction

such as DNA hydrolysis, DNA oxidation, and nucleases (Pin-

chuk et al., 2008), significant concentrations of dissolved DNA

have been found in marine water, freshwater, and sediments

at concentrations varying from 1 mg to above 80 mg dissolved

DNA per liter (Nielsen et al., 2007). Host-specific Bacteroidales

assays combined with PMA-PCR can resolve this issue and

simultaneously identify the sources of different fecal loadings

and estimate recent fecal pollution in watersheds. The ratio of

Ct values from qPCR with and without PMA treatment

provides a useful estimate of ages of fecal contamination from

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 4 8 5 0 – 4 8 5 94858

different animal sources based on the persistence of Bacter-

oidales genetic markers.

4.3. Modeling Bacteroidales decay

Understanding host-specific Bacteroidales decay rates is vital for

predicting their environmental persistence. We employed

exponential-plus-linear models to fit the experimental data and

calculated the decay rates of host-specific Bacteroidales. The

survival of culturable indicators such as E. coli and Enterococcus is

generally represented as a first-order decay function, asgiven by

Chick’s Law in water (Gonzalez, 1995; Pachepsky et al., 2006).

However, this exponential model may not fit well with experi-

mental data possessing random variations of field experiments.

Therefore, it was suggested that bacterial inactivation curves

are usually characterized by two or more phases, showing

‘‘shoulder’’ behavior (slow removal/die-off or perhaps growth

early in the experiment followed by apparent first-order decay

(Whitman et al., 2004). The third phase is an apparent period of

‘‘recovery’’ or ‘‘stability’’, which may or may not be quantified.

Usually, the decay rates are calculated within only certain

exponential periods and the shoulder phase or the third phase

in the experiment is ignored. Others have proposed that the

modelshouldcontain at least threeparameters, which is known

as the two-stage decay model. A realistic die-off model should

account for multiple stressors from abiotic or biotic factors and

even regrowth in the early stage; hence the linear portion is

explained by the uncertainty of the model due to experimental

errors or other marginal factors (Crane and Moore, 1986;

Pachepsky et al., 2006). At the molecular level microorganisms

do not follow a simple decay model, because various lethal

factors simultaneously affect their viability and the population

may separate into two subpopulations, namely, a rapid and

a slow decrease population. Therefore, the two-stage expo-

nential-plus-linear model provides more mechanistic insights

into the decay of Bacteroidales in water. The ‘‘shoulder’’ behavior

observed was not due to regrowth given that the prevailing

oxygen concentration of 6.5 mg l�1 would inhibit cell division,

but probably reflects the combined effects of predator grazing,

release of free DNA by cell lysis, desorption of DNA from sus-

pended solids, and other environmental variables.

Several questions have yet to be addressed. Sediment

resuspension represents a significant source of bacteria to the

water column (Fries et al., 2008). In the same way, Bacteroidales

might survive in anaerobic sediments for prolonged periods of

time or the DNA of Bacteroidales may persist while sorbed to

sediment. If true, the resuspension of sediments would

interfere in the analysis of microbial source tracking data.

Further, it is essential to determine the relationship between

the detection of a source identifier like Bacteroidales and the

presence of fecal pathogens as well as the incidence of illness

associated with recreational activities in impacted waters.

5. Conclusions

� The PMA-qPCR method to quantify host-specific Bacter-

oidales cells can detect a recent fecal contamination event

and significantly enhances the usefulness of source tracking

data obtained from watersheds experiencing inputs from

multiple fecal sources. The information is highly relevant

for the development of water quality criteria and traditional

FIB methods are unsuited to this purpose.

� Host-specific Bacteroidales cells had an average two-log

reduction time in seawater of 28 h whereas that of host-

specific Bacteroidales DNA was 177 h.

� Host-specific Bacteroidales showed similar decay rates in

seawater. Natural sunlight did not affect the survival of

uncultivated Bacteroidales cells and their DNA, with the

exception of cow-specific Bacteroidales DNA, which persisted

longer in the dark. A likely explanation in that the presence of

dissolved oxygen acts as principal factor in the decay of cells.

� The ratio of gene copies or Ct values calculated using qPCR

with and without PMA treatment can be used to estimate

the age of fecal pollution originating from different sources.

In an absolute sense, only one discrete source of host-

specific fecal contamination ought to be present when

applying the ratio. In many scenarios it is reasonable to

assume that there has been one type of major input of fecal

contamination, such as in form of runoff after a storm

event. The ratio can also serve to pinpoint human waste-

water inputs from sources not representing publicly owned

treatment works, based on the fact that treated and dis-

infected wastewater would have hardly any intact Bacter-

oidales cells. Hence, if a low ratio is detected for the human-

specific assay, the implication is that there are either septic

tank leakages or other recent human inputs.

� Decay rate constants for each host-specific Bacteroidales

genetic markerwerecalculatedfromatwo-stageexponential-

plus-linear model. These constants can provide valuable

information to establish fate and transport models of host-

specific Bacteroidales and to understand the different survival

stages in seawater.

� A rapid decrease in intact cells of Enterococcus and their DNA

was observed by qPCR at the beginning of the microscosm

decay experiment. However, the 23S rRNA gene remained

detectable throughout the study by both qPCR and qPCR with

PMA, necessitating further studies before this assay can be

used to inform and enforce new water quality criteria.

r e f e r e n c e s

Anderson, I.C., Rhodes, M., Kator, H., 1979. Sublethal stress inEscherichia coli: a function of salinity. Appl. Environ. Microbiol.38 (6), 1147–1152.

Anderson, I.C., Rhodes, M.W., Kator, H.I., 1983. Seasonal variationin survival of Escherichia coli exposed in situ in membranediffusion chambers containing filtered and nonfilteredestuarine water. Appl. Environ. Microbiol. 45 (6), 1877–1883.

Anderson, K.L., Whitlock, J.E., Harwood, V.J., 2005. Persistenceand differential survival of fecal indicator bacteria insubtropical waters and sediments. Appl. Environ. Microbiol. 71(6), 3041–3048.

Bae, S., Wuertz, S., 2009. Discrimination of viable and dead fecalBacteroidales bacteria by quantitative PCR with propidiummonoazide. Appl. Environ. Microbiol. 75, 2940–2944.

Bell, A., Layton, A.C., McKay, L., Williams, D., Gentry, D., Sayler, D.,2009. Factors influencing the persistence of fecal Bacteroides instream water. J. Environ. Qual. 38, 1224–1232.

Craig, D.L., Fallowfield, H.J., Cromar, N.J., 2004. Use of microcosmsto determine persistence of Escherichia coli in recreational

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 4 8 5 0 – 4 8 5 9 4859

coastal water and sediment and validation with in situmeasurements. J. Appl. Microbiol. 96 (5), 922–930.

Crane, S.R., Moore, J.A., 1986. Modeling enteric bacterial die-off –a review. Water, Air, Soil Pollut. 27 (3–4), 411–439.

Dick, L.K., Bernhard, A.E., Brodeur, T.J., Santo Domingo, J.W.,Simpson, J.M., Walters, S.P., Field, K.G., 2005. Hostdistributions of uncultivated fecal Bacteroidales bacteriareveal genetic markers for fecal source identification. Appl.Environ. Microbiol. 71 (6), 3184–3191.

Field, K.G., Samadpour, M., 2007. Fecal source tracking, theindicator paradigm, and managing water quality. Water Res.41 (16), 3517–3538.

Fries, J.S., Characklis, G.W., Noble, R.T., 2008. Sediment-waterexchange of Vibrio sp., fecal indicator bacteria: implicationsfor persistence and transport in the Neuse River Estuary,North Carolina, USA. Water Res. 42 (4–5), 941–950.

Gonzalez, J.M., 1995. Modelling enteric bacteria survival in aquaticsystems. Hydrobiologia 316 (2), 109–116.

Haugland, R.A., Siefring, S.C., Wymer, L.J., Brenner, K.P.,Dufour, A.P., 2005. Comparison of Enterococcus measurementsin freshwater at two recreational beaches by quantitativepolymerase chain reaction and membrane filter cultureanalysis. Water Res. 39 (4), 559–568.

Kildare, B.J., Leutenegger, C.M., McSwain, B.S., Bambic, D.G.,Rajal, V.B., Wuertz, S., 2007. 16S rRNA-based assays forquantitative detection of universal, human-, cow-, and dog-specific fecal Bacteroidales: a Bayesian approach. Water Res.41 (16), 3701–3715.

Kreader, C.A., 1998. Persistence of PCR-detectable Bacteroidesdistasonis from human feces in river water. Appl. Environ.Microbiol. 64 (10), 4103–4105.

Layton, A., McKay, L., Williams, D., Garrett, V., Gentry, R.,Sayler, G., 2006. Development of Bacteroides 16S rRNA geneTaqMan-based real-time PCR assays for estimation of total,human, and bovine fecal pollution in water. Appl. Environ.Microbiol. 72 (6), 4214–4224.

Nielsen, K.M., Johnsen, P.J., Bensasson, D., Daffonchio, D., 2007.Release and persistence of extracellular DNA in theenvironment. Environ. Biosafety Res. 6 (1–2), 37–53.

Nocker, A., Cheung, C.Y., Camper, A.K., 2006. Comparison ofpropidium monoazide with ethidium monoazide fordifferentiation of live vs. dead bacteria by selective removal ofDNA from dead cells. J. Microbiol. Methods 67 (2), 310–320.

Nocker, A., Sossa-Fernandez, P., Burr, M.D., Camper, A.K., 2007a.Use of propidium monoazide for live/dead distinction inmicrobial ecology. Appl. Environ. Microbiol. 73 (16), 5111–5117.

Nocker, A., Sossa, K.E., Camper, A.K., 2007b. Molecular monitoring ofdisinfectionefficacy using propidium monoazide incombinationwith quantitative PCR. J. Microbiol. Methods 70 (2), 252–260.

Okabe, S., Shimazu, Y., 2007. Persistence of host-specificBacteroides-Prevotella 16S rRNA genetic markers inenvironmental waters: effects of temperature and salinity.Appl. Microbiol. Biotechnol. 76 (4), 935–944.

Pachepsky, Y.A., Sadeghi, A.M., Bradford, S.A., Shelton, D.R.,Guber, A.K., Dao, T., 2006. Transport and fate of manure-bornepathogens: modeling perspective. Agric. Water Manage 86(1–2), 81–92.

Pinchuk, G.E., Ammons, C., Culley, D.E., Li, S.M.W., McLean, J.S.,Romine, M.F., Nealson, K.H., Fredrickson, J.K., Beliaev, A.S.,2008. Utilization of DNA as a sole source of phosphorus,carbon, and energy by Shewanella spp.: ecological andphysiological implications for dissimilatory metal reduction.Appl. Environ. Microbiol. 74 (4), 1198–1208.

Reischer, G.H., Kasper, D.C., Steinborn, R., Mach, R.L.,Farnleitner, A.H., 2006. Quantitative PCR method for sensitivedetection of ruminant fecal pollution in freshwater andevaluation of this method in alpine karstic regions. Appl.Environ. Microbiol. 72 (8), 5610–5614.

Rhodes, M.W., Kator, H., 1988. Survival of Escherichia coli andSalmonella spp. in estuarine environments. Appl. Environ.Microbiol. 54 (12), 2902–2907.

Rozen, Y., Belkin, S., 2001. Survival of enteric bacteria in seawater.FEMS Microbiol. Rev. 25 (5), 513–529.

Santo Domingo, J.W., Bambic, D.G., Edge, T.A., Wuertz, S., 2007.Quo vadis source tracking? Towards a strategic framework forenvironmental monitoring of fecal pollution. Water Res. 41(16), 3539–3552.

Savichtcheva, O., Okabe, S., 2006. Alternative indicators of fecalpollution: relations with pathogens and conventionalindicators, current methodologies for direct pathogenmonitoring and future application perspectives. Water Res. 40(13), 2463–2476.

Seurinck, S., Defoirdt, T., Verstraete, W., Siciliano, S.D., 2005.Detection and quantification of the human-specific HF183Bacteroides 16S rRNA genetic marker with real-time PCR forassessment of human faecal pollution in freshwater. Environ.Microbiol. 7 (2), 249–259.

Sinton, L.W., 2006. Biotic and abiotic effects. In: Belkin, S.,Colwell, R.R. (Eds.), Oceans and Health: Pathogen in the MarineEnvironment. Springer ScienceþBusiness Media, New York,pp. xiv, 464 pp.

Sinton, L.W., Finlay, R.K., Lynch, P.A., 1999. Sunlight inactivationof fecal bacteriophages and bacteria in sewage-pollutedseawater. Appl. Environ. Microbiol. 65 (8), 3605–3613.

Stapleton, C.M., Wyer, M.D., Kay, D., Crowther, J., McDonald, A.T.,Walters, M., Gawler, A., Hindle, T., 2007. Microbial sourcetracking: a forensic technique for microbial sourceidentification? J. Environ. Monit. 9 (5), 427–439.

USEPA, 2002. Method 1600: Enterococci in Water by MembraneFiltration Using Membrane- Enterococcus Indoxyl-b-D-Glucoside Agar (mEI). Office of Water, Washington, DC (EPA821-R-02-022).

Walters, S.P., Field, K.G., 2006. Persistence and growth of fecalBacteroidales assessed by bromodeoxyuridineimmunocapture. Appl. Environ. Microbiol. 72 (7), 4532–4539.

Walters, S.P., Yamahara, K.M., Boehm, A.B., 2009. Persistenceof nucleic acid markers of health-relevant organisms inseawater microcosms: implication for their use inassessing risk in recreational waters. Water Res. 43 (19),4929–4939.

Whitman, R.L., Nevers, M.B., Korinek, G.C., Byappanahalli, M.N.,2004. Solar and temporal effects on Escherichia coliconcentration at a Lake Michigan swimming beach. Appl.Environ. Microbiol. 70 (7), 4276–4285.