Subscriber access provided by UNIV OF CONNECTICUT
Article
Droplet digital PCR outperforms real-time PCR in the detection of environmental DNA from an invasive fish species Hideyuki Doi, Teruhiko Takahara, Toshifumi Minamoto, Saeko Matsuhashi, Kimiko Uchii, and Hiroki Yamanaka Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.5b00253 • Publication Date (Web): 07 Apr 2015 Downloaded from http://pubs.acs.org on April 12, 2015
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 30
Environmental Science & Technology
1
Droplet digital PCR outperforms real-time PCR in the detection
2
of environmental DNA from an invasive fish species
3
Hideyuki Doi ‡1 *, Teruhiko Takahara‡2, Toshifumi Minamoto3, Saeko Matsuhashi1, Kimiko
4
Uchii4, and Hiroki Yamanaka5
5 6
1
7
739-8530 Japan 2 Graduate School of Integrated Arts and Sciences, Hiroshima University, 739-
8
8530 Higashi-Hiroshima, Japan 3 Graduate School of Human Development and Environment,
9
Kobe University, Kobe 657-0013 Japan 4 Faculty of Pharmacy, Osaka Ohtani University,
Institute for Sustainable Sciences and Development, Hiroshima University, Higashi-Hiroshima,
10
Tondabayashi 584-0066 Japan 5 Department of Environmental Solution Technology, Faculty of
11
Science and Technology, Ryukoku University, Otsu 520-2194 Japan
12 13
*Corresponding author: Hideyuki Doi
14
Institute for Sustainable Sciences and Development, Hiroshima University,
15
Hiroshima University, 1-3-1 Kagamiyama, Higashi-Hiroshima, 739-8530 Japan
16
e-mail:
[email protected] 17
Tel: +81-82-424-5732
Fax: +81-82-424-5732
18 19
ACS Paragon Plus Environment
1
Environmental Science & Technology
Page 2 of 30
20
21
ABSTRACT
22
Environmental DNA (eDNA) has been used to investigate species distributions in aquatic ecosystems.
23
Most of these studies use real-time PCR to detect eDNA in water; however, PCR amplification is often
24
inhibited by the presence of organic and inorganic matter. In droplet digital PCR (ddPCR), the sample is
25
partitioned into thousands of nano-liter droplets, and PCR inhibition may be reduced by the detection of
26
the end-point of PCR amplification in each droplet, independent of the amplification efficiency. In
27
addition, real-time PCR reagents can affect PCR amplification and consequently alter detection rates. We
28
compared the effectiveness of ddPCR and real-time PCR using two different PCR reagents for the
29
detection of the eDNA from invasive bluegill sunfish, Lepomis macrochirus, in ponds. We found that
30
ddPCR had higher detection rates of bluegill eDNA in pond water than real-time PCR with either of the
31
PCR reagents, especially at low DNA concentrations. Limits of DNA detection, which were tested by
32
spiking the bluegill DNA to DNA extracts from the ponds containing natural inhibitors, found that
33
ddPCR had higher detection rate than real-time PCR. Our results suggest that ddPCR is more resistant to
34
the presence of PCR inhibitors in field samples than real-time PCR. Thus, ddPCR outperforms real-time
35
PCR methods for detecting eDNA to document species distributions in natural habitats, especially in
36
habitats with high concentrations of PCR inhibitors.
37 38
ACS Paragon Plus Environment
2
Page 3 of 30
39
Environmental Science & Technology
Introduction
40 41
Documenting species distributions is fundamental to ecosystem management and is essential 1-3
42
for managing and conserving rare and endemic species,
as well as for controlling invasive
43
species. Recently, environmental DNA (eDNA) analyses have been applied to detect the
44
presence of species in aquatic and terrestrial ecosystems.4, 5 eDNA surveys have great potential
45
to contribute to species distribution data and reduce survey costs. eDNA methods have been
46
applied for the detection of many types of animals, especially in aquatic ecosystems; these
47
include common, 6-13 invasive, 14-19 and endangered species20-23 in various ecosystems, including
48
ponds and lakes, 10, 11,15,17,18 rivers and streams, 6, 7, 9, 12, 16, 19, 23 and oceans. 14
49
Most of these eDNA studies have used PCR techniques to detect eDNA originating from the
50
target species, thereby assessing the presence or absence of the species.6-9 eDNA detection was
51
first used in the field to assess the status of the bullfrog Rana catesbeiana (= Lithobates
52
catesbeianus) using electrophoresis on PCR products. 6 Recent eDNA studies generally use real-
53
time PCR platforms.9, 10, 20, 21 For example, Takahara et al. used real-time PCR to detect the
54
eDNA of invasive bluegill sunfish (Lepomis macrochirus) in ponds in Japan, obtaining better
55
detection rates than traditional observation methods.
56
crayfish, Procambarus clarkii, and New Zealand mud snails, Potamopyrgus antipodarum, have
57
been assessed using real-time PCR eDNA detection.16, 17
15
Also, the distributions of invasive
58
However, it is known that eDNA detection rates can be limited by very-low DNA sample
59
concentrations and the presence of substances which inhibit PCR.11-13 Increasing eDNA
60
concentrations in field samples by concentrating large volumes of water may also increase the
61
concentrations of inhibitory substances in the form of macromolecules present in the natural
ACS Paragon Plus Environment
3
Environmental Science & Technology
Page 4 of 30
62
environment. For example, humic and tannic acids, polymers, and many unknown materials are
63
commonly present in crude DNA extracts from soils and tissues, and even small quantities may
64
be sufficient to inhibit Taq polymerase.24,25,26 To extract and purify eDNA samples, thus
65
reducing concentrations of such PCR inhibitors, eDNA studies generally use DNA purification
66
kits, but these do not eliminate all macromolecules. Thus, we need to develop eDNA detection
67
methods with high sensitivities and tolerances to inhibitory substances.
68
Recently, droplet digital PCR (ddPCR), also known as a “third-generation PCR,” has been
69
developed.24 ddPCR is an emerging DNA detection method which generates thousands of nano-
70
liter droplets, some of which ideally contain only one or few copies of the target DNA. The PCR
71
reaction occurs in each droplet, and end-point PCR amplification is detected by the fluorescence
72
intensity of PCR probes.27 To date, ddPCR has been used as an alternative of real-time PCR to
73
quantify DNA concentrations.28-33 Recently, ddPCR has been applied to quantify fish eDNA
74
concentrations in mesocosm experiments.34-35 In ddPCR, PCR inhibitory substances should have
75
little effect on DNA quantification, because the end-point PCR amplification in each droplet can
76
be detected independent of the amplification efficiency.27,36 Thus, we may expect ddPCR to be
77
more suitable for the measurement of eDNA in field samples than real-time PCR. However, the
78
performance of ddPCR in detecting eDNA in field samples has never been evaluated.
79
In many real-time PCR eDNA studies, TaqMan Gene Expression Master Mix (GEMM, Life
80
Technologies) has been used because of its highly sensitive DNA detection. Recently, TaqMan
81
Environmental Master Mix 2.0 (EMM, Life Technologies) has been used for eDNA
82
detection14,17,19. EMM is tolerant to high concentrations of PCR inhibitors occurring in field
83
water samples, such as humic acid, and may improve eDNA detection. While a previous study11
84
showed that a real-time PCR with EMM detected target eDNA from several samples in which a
ACS Paragon Plus Environment
4
Page 5 of 30
Environmental Science & Technology
85
real-time PCR with GEMM detected no amplification, little is known so far about the relative
86
performance of the two for eDNA detection rate. In this study, we compared the accuracies of 1)
87
ddPCR, 2) real-time PCR with GEMM, and 3) real-time PCR with EMM in detecting eDNA of
88
invasive bluegill sunfish in field samples obtained from ponds, as studied by Takahara et al.15
89
We evaluated the differences in eDNA detection rates and limits of detection (LOD) using the
90
three PCR methods.
91 92
Materials and Methods
93 94
eDNA Sampling
95
We examined the presence of invasive bluegill sunfish (L. macrochirus) populations in ponds,
96
which were previously surveyed for eDNA by Takahara et al.15. These Japanese populations
97
were derived from a small number of individuals from a population in the Mississippi River,
98
USA.37 We used eDNA samples from 25 ponds which had been previously collected by
99
Takahara et al.15. In 19 out of the 25 ponds, bluegill eDNA had previously been detected in at
100
least in one of eight replicates in our previous study using real-time PCR with GEMM.15 The
101
other 6 ponds were randomly selected from the ponds in which eDNA was not detected and
102
bluegill was not found visually from the shore with walking of the whole shoreline for 10–20
103
min. Briefly, 1 L of water was collected from the surface of each pond from October to
104
December 2011. The water samples were stored in DNA-free bottles (Nalgene®) and
105
immediately transported in a cooling box on ice to the laboratory and stored at –30°C until
106
processed. After thawing, the water samples were filtered through a 3.0 µm membrane filter
107
(cellulose acetate, C300A142C; Advantec, Saijo, Japan). The filter discs were then placed in
ACS Paragon Plus Environment
5
Environmental Science & Technology
Page 6 of 30
108
autoclaved bottles and soaked in 10 mL autoclaved ultrapure water and agitated on a rotary
109
shaker at the maximum speed (250 rpm) for 10 min. Suspended matter in the bottles was
110
decanted into a centrifugal filter unit S10Y30 (30-kDa cutoff, 540642; Millipore, Billerica, MA)
111
and concentrated to 400 µL by spinning at 5000 × g. for 15 min. Then, eDNA was extracted
112
using a DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany) in a final volume of 100 µL. As
113
blanks, DNA-free pure water (1 L) were treated in the same manner as the samples during
114
filtration, eDNA extraction and PCR procedures, and bluegill eDNA was not detected in these in
115
subsequent ddPCR and real-time PCR assays. Full details of the methods for eDNA sampling
116
and(?) extraction have been described previously by Takahara et al.15
117 118
Real-time PCR assays
119
Real-time PCR for bluegill mitochondrial cytochrome b gene fragments was performed as
120
described by Takahara et al.15 eDNA was quantified using real-time TaqMan® PCR with a
121
StepOnePlus™ Real-Time PCR system (Life Technologies, Carlsbad, CA, USA). The
122
mitochondrial cytochrome b gene fragments (100 bp) were amplified and quantified with the
123
following
124
GCCTAGCAACCCAGATTTTAACA-3′),
125
ACGTCCCGGCAGATGTGT-3′),
126
CGACATCGCAACTGCCTTCTCTTCAGT-TAMRA-3′). 15 The specificity of the primers and
127
probe was tested with extracted DNA samples from other sunfish species present in Japan
128
(Micropterus salmoides and M. dolomieu) using real-time PCR, which generated no
primers
and
TaqMan
probe:
Bluegill_CytB_F
Bluegill_CytB_R
and
Bluegill_CytB_probe
ACS Paragon Plus Environment
(5′-
(5′-
(5′-FAM-
6
Page 7 of 30
Environmental Science & Technology
129
amplification signals. In addition, we directly sequenced the PCR amplicons to confirm the
130
specificity of the primer set. To confirm the specificity of the primer set in the field, real-time
131
PCR amplicons of all eDNA samples that were positive for the bluegill DNA were directly
132
sequenced after treatment with ExoSAP-IT (USB Corporation, Cleveland, OH, USA). Sequences
133
were determined by a commercial sequencing service (Takara Bio, Tokyo, Japan). All sequences
134
from each real-time PCR amplicon were confirmed as being from bluegill.
135
We used two master mix types (TaqMan Gene Expression Master Mix, GEMM, and TaqMan
136
Environmental Master Mix 2.0, EMM, Life Technologies) for real-time PCR. Each TaqMan
137
reaction contained 900 nM of each primer and 125 nM of TaqMan probe in 1× PCR master mix
138
kit (GEMM or EMM), along with 2 µL of DNA solution. Takahara et al.11 found that 2 µL of
139
DNA solution was optimal to detect fish eDNA using real-time PCR with minimum PCR
140
inhibitory effects for the pond samples. Eight replicates of each PCR reaction were performed in
141
a thermocycler (2 min at 50°C, 10 min at 95°C, and 55 cycles of 15 s at 95°C and 60 s at 60°C)
142
in a StepOnePlus Real Time PCR system (Life Technologies). We used 55 PCR cycles following
143
the standard protocol used in eDNA studies to increase eDNA detectability.11, 14, 15 Eight wells
144
were used as no-template negative controls for all real-time PCR assays; no amplification signal
145
was observed in these wells or the blank samples. To avoid contamination, we performed
146
the PCR set-up, including preparation and addition of the standards, in a separate room from that
147
where PCR cycling occurred.
148 149
ddPCR assays
150
We performed ddPCR analysis using the same DNA solutions and the same primers and probe
151
used for real-time PCR analysis. Each ddPCR reaction mixture (total 20 µL) contained 2 µL of
ACS Paragon Plus Environment
7
Environmental Science & Technology
Page 8 of 30
152
DNA solution, 900 nM of each primer, and 125 nM TaqMan probe in a 1× Bio-Rad Supermix kit
153
(Bio-Rad, Hercules, CA, USA), which was mixed with Bio-Rad droplet generator oil and
154
partitioned into nano-litter droplets using a Bio-Rad QX-100 droplet generator (Bio-Rad).
155
Droplets of each sample were separately applied to each well of a 96-well PCR plate. PCR was
156
performed with the 96-well plate sealed with pierceable sealing foil using a GeneAmp 9700
157
thermocycler (Applied Biosystems, Grand Island, NY, USA). PCR conditions were 10 min at
158
95°C, 45 cycles of denaturation for 30 s at 95°C and extension for 60 s at 60°C with ramp rate of
159
2.5°C s-1, followed by 10 min at 95°C and a hold at 4°C. Based on the results of a preliminary
160
experiment (Appendix S1) and the maximum Ct values of real-time PCR (40.8) in this and our
161
previous study,15 we decided to measure eDNA for all samples using PCR with 45 cycles. After
162
amplification, the PCR plate was transferred to a QX-100 droplet reader (Bio-Rad). The TaqMan
163
fluorescence of each droplet in each well was measured. Eight wells were used as no-template
164
negative controls for all ddPCR assays; droplets from the negative-control samples had
165
fluorescence (RFU) less than 800 after PCR amplification. We used Bio-Rad’s QuantaSoft
166
software version 1.3.2.0 to quantify the detection of target DNA. The threshold for a positive
167
signal was set at 1100 RFU, as determined by the negative control in LOD analysis, described
168
below. Any droplet with fluorescence above this threshold was counted as positive for bluegill
169
eDNA (Fig. 1, Fig. S1). The mean volume of droplet will be about 1 pL with 20-µL PCR
170
reaction. The mean detected droplets and analysis variance by ddPCR were 10,081 (± 358, 95%
171
CI) and 35.9 for all ddPCR analysis, respectively. We included all essential (E) information in
172
dMIQE required37 (Appendix).
173 174
Testing limit of detection (LOD)
ACS Paragon Plus Environment
8
Page 9 of 30
Environmental Science & Technology
175
To compare the minimum number of target DNA copies that could be detected among PCR
176
types (ddPCR, real-time PCR with GEMM, and real-time PCR with EMM), we performed two
177
limit of detection (LOD) analysis for each method. 1) DNA dilution experiment; we used the
178
following dilution series of bluegill DNA standard: 0, 1, 2, 3, and 10 copies per 2-µL PCR
179
template. The bluegill DNA used for LOD analysis was cloned with pGEM-T Easy Vector
180
(Promega, Tokyo, Japan), which concentration was quantified using a NanoDrop 1000
181
spectrophotometer (Thermo Fisher Scientific, Waltham, MA USA)15. For each PCR method, we
182
performed eight replicated LOD analyses for each number of DNA copies in the dilution series.
183
2) DNA spiking experiment; we mixed 10 pond samples (26 µL each), from which bluegill
184
eDNA were not detected, to prepare 260 µL of mixed DNA solution. The mixed DNA solution
185
was spiked with the bluegill DNA at concentrations of 0, 1, 2, 3, and 10 copies µL–1. Real-time
186
PCR and ddPCR were performed with 2 µL of the spiked sample as template in the same manner
187
as described above.
188 189
Statistical analyses
190
For the relationships between the counts of eDNA detection and eDNA concentrations, we fit
191
generalized linear models (GLMs) with binomial error distributions to eDNA counts from ponds
192
and LOD samples for each PCR method [eDNA count from eight replicates] = log10 [eDNA
193
copies]. To compare the differences in eDNA counts among PCR methods, we fit a GLM with
194
binomial error distribution, including an eDNA × PCR method interaction, [eDNA count from
195
eight replicates] = log10 [eDNA copies]*[PCR types].
196
To compare eDNA presence in ponds among PCR types, we also fit GLMs with binomial
197
error distributions in the same manner as above [eDNA presence of ponds (0 or 1)] = log10
ACS Paragon Plus Environment
9
Environmental Science & Technology
Page 10 of 30
198
[eDNA copies] and [eDNA presence of ponds (0 or 1)] = log10 [eDNA copies]*[PCR types]. All
199
statistical analyses were performed at a significance level of α = 0.05 using R ver. 3.1.0.39
200 201
Results
202 203
Detection of pond samples
204
There were significant positive relationships between DNA concentrations, which were
205
measured by real-time PCR with GEMM in Takahara et al.15, and the detection rate for all three
206
PCR methods (GLM, p < 0.001, Fig. 2). In pond samples, which had more than 150 copies L–1,
207
bluegill DNA was detected in all replicates by all PCR methods. ddPCR and real-time PCR with
208
EMM detected bluegill DNA in 3 and 1 pond sample in which the target DNA was not detected
209
by real-time PCR with GEMM, respectively (Fig. 3). In contrast, real-time PCR with EMM did
210
not detect the target DNA in one pond sample in which target DNA was detected by real-time
211
PCR with GEMM. The parameters of GLMs were significantly different among PCR types
212
(GLM, |z| > 3.094, p < 0.001 for all coefficients; DNA concentrations, PCR methods, and DNA
213
concentration × PCR method, Fig. 2), indicating that the slopes of GLMs differed significantly
214
among PCR methods, but the trends were associated with DNA concentration. ddPCR had
215
higher detection rates than real-time PCR with GEMM or EMM at low DNA concentrations.
216
However, the ddPCR detection rate was lower than that of real-time PCR at moderate DNA
217
concentrations (i.e., 3-5 DNA copies).
218
There were significant positive relationships between DNA concentration, as measured by
219
ddPCR, and detection probability (GLM, p < 0.001, Fig. 3). GLM parameters were not
ACS Paragon Plus Environment
10
Page 11 of 30
Environmental Science & Technology
220
significantly different among the PCR methods (GLM, |z| < 1.815, p = 0.064 for all coefficients:
221
DNA concentration, PCR method, and DNA concentration × PCR method, Fig. 3).
222 223
LOD analysis
224
1) DNA dilution experiment; The LOD trends (i.e., detection rates as a function of number of
225
DNA copies) did not differ significantly among PCR methods (GLM, z = 0.557, p = 0.578, Fig.
226
4). However, when using 2 and 3 DNA copies, real-time PCR with GEMM and EMM had
227
slightly higher detection rates than ddPCR. In all PCR types, the lower limit of detection was at 1
228
copy per reaction, and the detection rates tended to increase with increasing DNA concentration.
229
2) DNA spiking experiment; The LOD trends differed significantly among PCR methods
230
(GLM, z = 2.156, p = 0.0301, Fig. 5), indicating that ddPCR had higher detection rate for target
231
DNA than real-time PCR in field samples containing inhibitors. At target DNA concentrations of
232
5 and 10 copies per reaction, ddPCR had higher detection rates than real-time PCR. Only ddPCR
233
detected target DNA in all 8 replicates at the concentration of 10 DNA copies/reaction (Fig. S1).
234
In all PCR types, the lower limit of detection was 1 copy per reaction, and the detection rates
235
tended to increase along with increased DNA concentration.
236 237 238
Discussion
239 240
We found that eDNA detection rates using ddPCR were higher than those obtained using real-
241
time PCR with either GEMM or EMM, particularly in field samples with low DNA
242
concentrations. Real-time PCR with EMM had higher eDNA detection rates than real-time PCR
ACS Paragon Plus Environment
11
Environmental Science & Technology
Page 12 of 30
243
with GEMM. Thus, in general, detection rates for ddPCR were greater than those for real-time
244
PCR with EMM, which in turn were greater than those for real-time PCR with GEMM.
245
However, there was little difference in LOD results among the PCR methods.
246
One of the main reasons for higher eDNA detectability using ddPCR and real-time PCR with
247
EMM is likely to be that these PCR methods decrease PCR inhibition by compounds in the
248
natural environment. In ddPCR, each droplet contains 0 or 1 to few copies target DNA. This
249
distribution would be also true for inhibitors, which can reduce concentrations of PCR inhibitors
250
in ddPCR than present in real-time PCR.36 From our DNA spiking LOD test, ddPCR had higher
251
detection rate than real-time PCR and supported our explanation for PCR inhibitor effects on
252
detectability. Furthermore, ddPCR uses end-point PCR amplification in each nano-liter droplet to
253
independently detect amplified DNA amplification. Such methodological advantages of ddPCR
254
increase the detectability of DNA in field samples in the face sample-dependent limitations. In
255
real-time PCR, the tolerance of EMM to PCR inhibitors was greater than that of GEMM. Such
256
tolerance to PCR inhibitors should allow detection of DNA in samples with high concentrations
257
of PCR inhibitors. Thus, real-time PCR with EMM had higher detection rates than real-time PCR
258
with GEMM.
259
In an assessment of the effect of sample volume on PCR with GEMM, Takahara et al.11
260
suggested that 2 µL samples showed a smaller PCR inhibition effect on eDNA detection than 5
261
µL samples. However, using ddPCR or EMM may allow the PCR sample volume to be increased
262
to 5 µL or more (e.g., a previous study has used up to 13 µL eDNA sample in 20 µL PCR
263
reaction. 40). We were unable to test the effects of sample volume due to limited available sample
264
solution (100 µL), but future studies should consider estimating eDNA detection rates using
265
ddPCR and real-time PCR with EMM with larger PCR sample volumes.
ACS Paragon Plus Environment
12
Page 13 of 30
Environmental Science & Technology
266
The ddPCR detection rate was higher than that of the other PCR methods at low DNA
267
concentrations, but was slightly lower than that of real-time PCR at moderately low DNA
268
concentrations (e.g., 2-5 copies per PCR reaction). The relationship between detection rates and
269
DNA concentration was more linear for ddPCR than for real-time PCR. One possible
270
explanation would be that the detection mechanism of ddPCR is more resistant to external
271
factors, such as the effects of inhibitory substances, or low eDNA concentrations. Unlike in the
272
detection of end-point amplification in each droplet, as in ddPCR, real-time PCR can be easily
273
adversely affected by low amplification efficiency because it measures the fluorescence intensity
274
of amplicons in 20-µL PCR reactions. However, to further our understanding of the underlying
275
mechanisms, further studies would be necessary.
276
There has been increasing attentions to false positive and negative eDNA detection rates.41
277
Recently, a model for estimating false positive and negative rates of detection, while accounting
278
for eDNA sampling effort, has been proposed.42 ddPCR and the use of EMM in real-time PCR
279
assays have the potential to decrease false negative eDNA detection rates without increasing
280
sampling effort. ddPCR and EMM may contribute to increased certainty in not only eDNA
281
studies, but also in the development of species distribution and occupancy models, which are
282
widely applied to population management.42 In this study, we were unable to test false positive
283
DNA detection rates, although we did not find any DNA copies in the negative controls. More
284
intensive analysis using negative samples with/without spiking with target DNA were needed to
285
test false-positive detection rates for ddPCR and real-time PCR.
286
eDNA techniques have been applied to estimate species’ biomass and abundance.10, 20, 34, 35 To
287
accurately estimate eDNA concentrations in field samples, PCR platforms amenable to very low
288
template DNA concentrations are required. ddPCR was more accurate in quantifying fish eDNA
ACS Paragon Plus Environment
13
Environmental Science & Technology
Page 14 of 30
289
at low concentrations than real-time PCR, 35 and similar results were reported in a comparison of
290
ddPCR and real-time PCR using HIV DNA.30 It seems ddPCR can quantify even a few DNA
291
copies with high precision. Thus, ddPCR is a useful PCR platform due to its ability to for
292
estimate eDNA concentrations with higher precision than real-time PCR and detect eDNA
293
presence at low concentrations.
294
Taken together, our results suggest ddPCR is a better platform than real-time PCR for
295
detecting eDNA and quantifying its abundance in field samples. However, the disadvantages of
296
the three PCR methods should also be considered. 1) Cost and time: conducting ddPCR assays is
297
more expensive and time-consuming than conducting real-time PCR,35 and ddPCR facilities are
298
much more expensive than those for real-time PCR. The cost for real-time PCR is similar for
299
either GEMM or EMM because the prices of the two reagents are almost the same. 2) Carry-over
300
contamination risk: to minimize carry-over contamination of PCR products, GEMM contains a
301
Uracil-DNA Glycosylase, while ddPCR supermix and EMM do not. Thus, using GEMM in real-
302
time PCR can mitigate carry-over risks compared to ddPCR and real-time PCR with EMM. In
303
this study, we did not find any carry-over effects in any PCR platform (i.e. did not detect DNA in
304
any negative controls). However, future studies should test for carry-over effects in ddPCR and
305
real-time PCR with EMM assays. 3) Amplification and detection mechanisms: for ddPCR, an
306
appropriate threshold setting is essential for detection of eDNA. With a default fluorescence
307
threshold setting, such as >2000, we could not detect droplets containing DNA in some cases
308
(e.g., at lower DNA concentrations) and in such cases the detection rate of ddPCR was less than
309
that of real-time PCR. In this study, we manually set the threshold used in LOD analysis
310
according to outputs obtained from blank samples and those with only a few DNA copies. Some
311
of the samples in this study only had a few positive droplets in ddPCR. Such low number of
ACS Paragon Plus Environment
14
Page 15 of 30
Environmental Science & Technology
312
positive droplets can potentially increase false-positive detection rates. However, we did not find
313
any positive droplets from all negative controls (32 negative controls in total). Thus, the samples
314
with positive droplets using the threshold of 1100 RFU would have contained the target DNA.
315
Although further experiments are required to closely examine the optimum fluorescence
316
threshold to detect eDNA in ddPCR, the manual adjustment using negative control samples and
317
those with a few DNA copies would be one of the options to determine the threshold. In
318
addition, after considering preliminary tests and the maximum Ct value of real-time PCR (40.8),
319
we used 45 PCR cycles for ddPCR, which is less than the number used for real-time PCR (55
320
cycles). Despite this, the detection rates of ddPCR in field samples were higher than those of
321
real-time PCR. Because ddPCR measures the end-point of PCR amplification, we could not
322
check the amplification process (i.e., amplification plots and resulting amplicons). This is a
323
disadvantage of ddPCR; we could not identify non-target DNA amplification or other problems
324
occurring during PCR cycles. The number of PCR cycles for ddPCR should be carefully
325
determined with reference to the Ct values of real-time PCR or other information.
326
In conclusion, ddPCR is superior to real-time PCR for detecting eDNA in water from natural
327
habitats, such as ponds, especially at low DNA concentrations. In this study, we used aquatic
328
samples; DNA samples from soil and sediment generally contain more PCR inhibitory
329
substances.19 Thus, in other ecosystems, ddPCR would likely also have higher eDNA detection
330
rates than real-time PCR, which could help to better assess species distributions, but this needs
331
confirmation.
332 333
AUTHOR INFORMATION
334
Corresponding Author
ACS Paragon Plus Environment
15
Environmental Science & Technology
335
Institute for Sustainable Sciences and Development, Hiroshima University,
336
1-3-1 Kagamiyama, Higashi-Hiroshima, 739-8530 Japan
337
E-mail:
[email protected] 338
Tel: +81-82-424-5732
339
Fax: +81-82-424-5732
Page 16 of 30
340 341
Author Contributions
342
The manuscript was written through contributions of all authors. All authors have given approval
343
to the final version of the manuscript. ‡These authors contributed equally.
344 345
Funding Sources
346
Environment Research and Technology Development Fund (4RF-1302)
347
Industry-University Collaboration Grant for Young Scientists in Hiroshima University
348
JST-CREST
349 350
ACKNOWLEDGMENT
351
We sincerely thank Dr. Keiichi Fukaya, The Institute of Statistical Mathematics, who gave
352
advice on our experimental design and statistical analyses. We sincerely thank the Faculty of
353
Science and the Natural Science Center for Basic Research and Development, Hiroshima
354
University, for providing PCR facilities. This study was supported by the Environment Research
ACS Paragon Plus Environment
16
Page 17 of 30
Environmental Science & Technology
355
and Technology Development Fund (4RF-1302) of the Ministry of the Environment, Japan, JST-
356
CREST, and in part by the Industry-University Collaboration Grant for Young Scientists in
357
Hiroshima University to Teruhiko Takahara.
358 359
Supporting Information Available
360
Appendix S1 Testing for PCR cycle number for ddPCR
361
This information is available free of charge via the Internet at http://pubs.acs.org/ .
362 363 364 365 366 367 368 369 370 371
REFERENCES (1) Guisan, A.; Thuiller, W. Predicting species distribution: offering more than simple habitat
models. Ecol. Lett. 2005, 8 (9), 993-1009. (2) Elith, J.; Leathwick, J. R. Species distribution models: ecological explanation and prediction across space and time. Ann. Rev. Ecol. Evol. System. 2009, 40, 677-697. (3) Hutchinson, G. E. Homage to Santa Rosalia or why are there so many kinds of animals? Am. Nat. 1959, 93 (870), 145-159. (4) Rees, H. C.; Maddison, B. C.; Middleditch, D. J.; Patmore, J. R.; Gough, K. C. REVIEW:
372
The detection of aquatic animal species using environmental DNA–a
373
survey tool in ecology. J. Appl. Ecol. 2014, 51 (5), 1450-1459.
374
(5) Pedersen, W. P.; Overballe-Petersen, S.; Ermini, L.; Der
review of eDNA as a
Sarkissian,
C.;
Haile,
J.;
375
Hellstrom, M.; Spens, J.; Thomsen, P. F.; Bohmann, K.; Cappellini, E; Schnell, I. B.; Wales, N.
376
A.; Carøe, C.; Campos, P. F.; Schmidt, A. Z. M.; Gilbert, T. P.; Hansen, A. J.; Orlando, L.;
ACS Paragon Plus Environment
17
Environmental Science & Technology
Page 18 of 30
377
Willerslev, E. Ancient and modern environmental DNA. Philosoph. Trans. Royal Soc. B: Biol.
378
Sci. 2015, 370, 20130383.
379 380
(6) Ficetola, G. F.; Miaud, C.; Pompanon, F.; Taberlet, P. Species detection using environmental DNA from water samples. Biol. Lett. 2008, 4 (4), 423-425.
381
(7) Goldberg, C. S.; Pilliod, D. S.; Arkle, R. S.; Waits, L. P. Molecular detection of vertebrates
382
in stream water: a demonstration using rocky mountain tailed frogs and Idaho giant salamanders.
383
PLoS One. 2011, 6 (7), e22746.
384
(8) Dejean, T.; Valentini, A.; Miquel, C.; Taberlet, P.; Bellemain, E.; Miaud, C. Improved
385
detection of an alien invasive species through environmental DNA barcoding: the example of the
386
American bullfrog Lithobates catesbeianus. J. Appl. Ecol. 2012, 49 (4), 953-959.
387 388 389 390
(9) Minamoto, T.; Yamanaka, H.; Takahara, T.; Honjo, M. N.; Kawabata, Z. Surveillance of fish species composition using environmental DNA. Limnology 2012, 13 (2), 193-197. (10) Takahara, T.; Minamoto, T.; Yamanaka, H.; Doi, H.; Kawabata, Z. Estimation of fish biomass using environmental DNA. PLoS One. 2012, 7 (4). e35868.
391
(11) Takahara, T.; Minamoto, T.; Doi, H. Effects of sample processing on the detection rate of
392
environmental DNA from the Common Carp (Cyprinus carpio). Biol. Conserv. 2015, 183, 64-
393
49.
394
(12) Jane, S. F.; Wilcox, T. M.; McKelvey, K. S.; Young, M. K.; Schwartz, M. K.; Lowe, W.
395
H.; Letcher, B. H.; Whiteley, A. R. Distance, flow and PCR inhibition: eDNA dynamics in two
396
headwater streams. Molec. Ecol. Resourc. 2015, 15 (1), 216-227.
ACS Paragon Plus Environment
18
Page 19 of 30
Environmental Science & Technology
397
(13) McKee, A. M.; Spear, S. F.; Pierson, T. W. The effect of dilution and the use of a post-
398
extraction nucleic acid purification column on the accuracy, precision, and inhibition of
399
environmental DNA samples. Biol. Conserv. 2015, 183, 70-76.
400
(14) Thomsen, P. F.; Kielgast, J.; Iversen, L. L.; Møller, P. R.; Rasmussen, M. Willerslev, E.
401
Detection of a diverse marine fish fauna using environmental DNA from seawater samples. PLoS
402
One. 2012, 7 (8), e41732.
403 404
(15) Takahara, T.; Minamoto, T.; Doi, H. Using environmental DNA to estimate the distribution of an invasive fish species in ponds. PLoS One. 2013, 8 (2), e56584.
405
(16) Goldberg, C. S.; Sepulveda, A.; Ray, A.; Baumgardt, J.; Waits, L. P. Environmental DNA
406
as a new method for early detection of New Zealand mudsnails (Potamopyrgus antipodarum).
407
Freshw. Sci. 2013, 32 (3), 792-800.
408
(17) Tréguier, A.; Paillisson, J.-M.; Dejean, T.; Valentini, A.; Schlaepfer, M. A.; Roussel, J.-
409
M. Environmental DNA surveillance for invertebrate species: advantages and technical
410
limitations to detect invasive crayfish Procambarus clarkii in freshwater ponds. J. Appl. Ecol.
411
2014, 51 (4), 871-879.
412 413 414 415
(18) Maruyama, A.; Nakamura, K.; Yamanaka, H.; Kondoh, M.; Minamoto, T. The release rate of environmental DNA from juvenile and adult fish. PloS One. 2014, 9 (12), e114639. (19) Turner, C. R.; Uy, K. L.; Everhart, R. C. Fish environmental DNA is more concentrated in aquatic sediments than surface water. Biol. Conserv. 2014, 183, 93-102.
ACS Paragon Plus Environment
19
Environmental Science & Technology
Page 20 of 30
416
(20) Thomsen, P. F.; Kielgast, J.; Iversen, L. L.; Wiuf, C.; Rasmussen, M.; Gilbert, M. T.;
417
Orlando, L. ; Willerslev, E. Monitoring endangered freshwater biodiversity using environmental
418
DNA. Mol. Ecol. 2012, 21 (11), 2565-2573.
419 420 421 422
(21) Laramie, M. B.; Pilliod, D. S.; Goldberg, C. S. Characterizing the distribution of an endangered salmonid using environmental DNA analysis. Biol. Conserv. 2014, 183, 29-37. (22) Mächler, E.; Deiner, K.; Steinmann, P.; Altermatt, F. Utility of environmental DNA for monitoring rare and indicator macroinvertebrate species. Freshw. Sci. 2014, 33 (4), 1174-1183.
423
(23) Fukumoto, S.; Ushimaru, A.; Minamoto, T. A basin-scale application of environmental
424
DNA assessment for rare endemic species and closely related exotic species in rivers; A case
425
study of giant salamanders in Japan. J. Appl. Ecol. 2015, 52 (2), 358-365.
426
(24) Tebbe, C. C.; Vahjen, W. Interference of humic acids and DNA extracted directly from
427
soil in detection and transformation of recombinant DNA from bacteria and a yeast. Appl.
428
Environ. Microbiol. 1993, 59 (8), 2657-2665.
429 430 431 432
(25) Wilson, I. G. Inhabitation and facilitation of nucleic acid amplification. Appl. Environ. Microbiol. 1997, 63 (10), 3741-3751. (26) Opel, K. L.; Chung, D.; McCord, B. R. A study of PCR inhabitation mechanisms using real time PCR. J. Forensic Sci. 2010, 55 (1), 25-33.
433
(27) Hindson, B. J.; Ness, K. D.; Masquelier, D. A.; Belgrader, P.; Heredia, N. J.; Makarewicz,
434
A. J.; Bright, I. J.; Lucero, M. Y.; Hiddessen, A. L.; Legler, T. C.; Kitano, T. K.; Hodel, M. R.;
435
Petersen, J. F.; Wyatt, P. W.; Steenblock, E. R.; Shah, P. H.; Bousse, L. J.; Troup, C. B.; Mellen,
436
J. C.; Wittmann, D. K.; Erndt, N. G.; Cauley, T. H.; Koehler, R. T.; So, A. P.; Dube, S.; Rose, K.
ACS Paragon Plus Environment
20
Page 21 of 30
Environmental Science & Technology
437
A.; Montesclaros, L.; Wang, S.; Stumbo, D. P.; Hodges, S. P.; Romine, S.; Milanovich, F. P.;
438
White, H. E.; Regan, J. F.; Karlin-Neumann, G. A.; Hindson, C. M.; Saxonov, S.; Colston, B.W.
439
High-throughput droplet digital PCR system for absolute quantitation of DNA copy number.
440
Anal. Chem. 2011, 83 (22), 8604-8610.
441 442
(28) Hoshino, T.; Inagaki, F. Molecular quantification of environmental DNA using microfluidics and digital PCR. Syst. Appl. Microbiol. 2012, 35 (6), 390-395.
443
(29) Hindson, C. M.; Chevillet, J. R.; Briggs, H. A.; Gallichotte, E. N.; Ruf, I. K.; Hindson, B.
444
J.; Vessella, R. L.; Tewari, M. Absolute quantification by droplet digital PCR versus analog real-
445
time PCR. Nat. Meth. 2013, 10 (10), 1003-1005.
446
(30) Strain, M. C.; Lada, S. M.; Luong, T.; Rought, S. E.; Gianella, S.; Terry, V. H.; Spina, C.
447
A.; Woelk, C. H.; Richman, D. D. Highly precise measurement of HIV DNA by droplet digital
448
PCR. PLoS One. 2013, 8 (4), e55943.
449
(31) Hayden, R. T.; Gu, Z.; Ingersoll, J.; Abdul-Ali, D.; Shi, L.; Pounds, S.; Caliendo A. M.
450
Comparison of droplet digital PCR to real-time PCR for quantitative detection of
451
Cytomegalovirus. J. Clin. Microbiol. 2013, 51 (2), 540-546.
452
(32) Norton, S. E.; Lechner, J. M.; Williams, T.; Fernando, M. R. A stabilizing reagent
453
prevents cell-free DNA contamination by cellular DNA in plasma during blood sample storage
454
and shipping as determined by digital PCR. Clin. Biochem. 2013, 46 (15), 1561-1565.
455
(33) Kim, T. G.; Jeong, S. Y.; Cho, K. S. Comparison of droplet digital PCR and quantitative
456
real-time PCR for examining population dynamics of bacteria in soil. Appl. Microbiol.
457
Biotechnol. 2014, 98 (13), 6105-6113.
ACS Paragon Plus Environment
21
Environmental Science & Technology
Page 22 of 30
458
(34) Nathan, L. R.; Simmons, M. D.; Wegleitner, B.; Jerde, C. L.; Mahon, A. Quantifying
459
environmental DNA signals for aquatic invasive species across multiple detection platforms.
460
Env. Sci. Tech. 2014, 48 (21), 12800-12806.
461
(35) Doi, H.; Uchii, K.; Takahara, T.; Matsuhashi, S.; Yamanaka, H.; Minamoto, T. Use of
462
droplet digital PCR for estimation of fish abundance and biomass in environmental DNA
463
surveys. PLOS ONE 2015, 10 (3), e0122763.
464 465
(36) Dingle, T. C.; Sedlak, R. H.; Cook, L.; Jerome, K. R. Tolerance of droplet-digital PCR vs real-time quantitative PCR to inhibitory substances. Clin. Chem. 2013, 59 (11), 1670-1672.
466
(37) Kawamura, K.; Yonekura, R.; Katano, O.; Taniguchi, Y.; Saitoh, K. Origin and dispersal
467
of bluegill sunfish, Lepomis macrochirus, in Japan and Korea. Molec. Ecol. 2006, 15 (3), 613-
468
621.
469
(38) Huggett, J. F.; Foy, C. A.; Benes, V.; Emslie, K.; Garson, J. A.; Haynes, R.; Hellemans, J.;
470
Kubista, M.; Mueller, R. D., Nolan ,T.; Pfaffl, M. W., Shipley, G. L.; Vandesompele, J.; Wittwer,
471
C. T.; Bustin S. A. Guidelines for minimum information for publication of quantitative digital
472
PCR experiments. Clin. Chem. 2013, 59 (6), 892-902.
473 474
(39) R Core Team (2014) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria; http://www.R-project.org.
475
(40) Rees, H. C.; Bishop, K.; Middleditch, D. J.; Patmore, J. R.; Maddison, B. C.; Gough, K.
476
C. The application of eDNA for monitoring of the Great Crested Newt in the UK. Ecol. Evol.
477
2014, 4 (21), 4023-4032.
ACS Paragon Plus Environment
22
Page 23 of 30
478 479
Environmental Science & Technology
(41) Moyer, G. R.; Díaz-Ferguson, E.; Hill, J. E.; Shea, C. Assessing Environmental DNA detection in controlled lentic systems. PLoS One. 2014, 9 (7), e103767.
480
(42) Schmidt, B. R.; Kéry, M.; Ursenbacher, S.; Hyman, O. J.; Collins, J. P. Site occupancy
481
models in the analysis of environmental DNA presence/absence surveys: a case study of an
482
emerging amphibian pathogen. Meth. Ecol. Evol. 2013, 4 (7), 646-653.
483 484
FIGURES
485
Figure 1. Fluorescence amplitude (FAM) of droplets in samples of pond water (3/8 detected,
486
right plot) and negative control (DNA-free water, left plot), output from QuantaSoft software for
487
ddPCR. Pink line indicates the threshold to detect DNA at 1100 RFU. Blue and black dots mean
488
positive and negative droplets. Yellow dotted line separates each replicate.
489
Figure 2. Detection rate of bluegill eDNA in pond water for three PCR methods; real-time PCR
490
with GEMM, EMM and ddPCR. The blue lines indicate the binomial GLM results.
491
Figure 3. Bluegill eDNA presence in the ponds for three PCR methods; real-time PCR with
492
GEMM, EMM and ddPCR. The blue lines indicate the binomial GLM results.
493
Figure 4. Detection rates in limited of detection (LOD) analyses using DNA dilution experiment
494
for three PCR methods; real-time PCR with GEMM, EMM and ddPCR. The blue lines indicate
495
the binomial GLM results.
ACS Paragon Plus Environment
23
Environmental Science & Technology
Page 24 of 30
496
Figure 5. Detection rates in limited of detection (LOD) analyses using DNA spiking experiment
497
for three PCR methods; real-time PCR with GEMM, EMM and ddPCR. The blue lines indicate
498
the binomial GLM results.
499 500
ACS Paragon Plus Environment
24
Page 25 of 30
Environmental Science & Technology
501 502 503
Figure 1
504
ACS Paragon Plus Environment
25
Environmental Science & Technology
Page 26 of 30
505 506
Figure 2
507
ACS Paragon Plus Environment
26
Page 27 of 30
Environmental Science & Technology
508 509 510
Figure 3
511
ACS Paragon Plus Environment
27
Environmental Science & Technology
Page 28 of 30
512 513
Figure 4
514
ACS Paragon Plus Environment
28
Page 29 of 30
Environmental Science & Technology
515
516 517
Figure 5
ACS Paragon Plus Environment
29
Environmental Science & Technology Page 30 of 30
ACS Paragon Plus Environment