Analytical validation of a reverse transcriptase droplet digital PCR (RT-ddPCR) for quantitative detection of infectious hematopoietic necrosis virus

Accepted Manuscript Title: Analytical validation of a reverse transcriptase droplet digital PCR (RT-ddPCR) for quantitative detection of infectious hematopoietic necrosis virus Authors: Peng Jia, Maureen K. Purcell, Guang Pan, Jinjin Wang, Shifu Kan, Yin Liu, Xiaocong Zheng, Xiujie Shi, Junqiang He, Li Yu, Qunyi Hua, Tikang Lu, Wensheng Lan, James R. Winton, Ningyi Jin, Hong Liu PII: DOI: Reference:

S0166-0934(16)30581-X http://dx.doi.org/doi:10.1016/j.jviromet.2017.03.010 VIRMET 13225

To appear in:

Journal of Virological Methods

Received date: Revised date: Accepted date:

19-10-2016 23-3-2017 23-3-2017

Please cite this article as: Jia, Peng, Purcell, Maureen K., Pan, Guang, Wang, Jinjin, Kan, Shifu, Liu, Yin, Zheng, Xiaocong, Shi, Xiujie, He, Junqiang, Yu, Li, Hua, Qunyi, Lu, Tikang, Lan, Wensheng, Winton, James R., Jin, Ningyi, Liu, Hong, Analytical validation of a reverse transcriptase droplet digital PCR (RT-ddPCR) for quantitative detection of infectious hematopoietic necrosis virus.Journal of Virological Methods http://dx.doi.org/10.1016/j.jviromet.2017.03.010 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Analytical validation of a reverse transcriptase droplet digital PCR (RT-ddPCR) for quantitative detection of infectious hematopoietic necrosis virus

1 2 3 4

Peng Jia a, c, e, Maureen K. Purcell b, Guang Pan a, c, Jinjin Wang a, c, Shifu Kan d, Yin Liu a, c, Xiaocong Zheng a, c, Xiujie Shi a, c, Junqiang He a, c, Li Yu a, c

, Qunyi Hua a, c, Tikang Lu a, c, Wensheng Lan a, c, James R. Winton b, Ningyi Jin, e, Hong Liu *, a, c

5 6 7

a

Shenzhen Entry-exit Inspection and Quarantine Bureau, Shenzhen, People’s Republic of China, 518045;

8

b

US Geological Survey, Western Fisheries Research Center, 6505 Northeast 65th Street, Seattle, WA 98115, USA;

9

c

Shenzhen Academy of Inspection and Quarantine Sciences, Shenzhen, People’s Republic of China, 518045;

10

d

Shenzhen Supervision and Testing Center for Quality and Safety of Agri-products, Shenzhen, People’s Republic of China, 518005;

11

e

Institute of Military Veterinary Medicine, Academy of Military Medical Sciences of PLA, Jilin, People’s Republic of China, 130117;

12 13 14

Highlights

1 2 3



Droplet digital PCR (ddPCR) has been used for precise and sensitive quantification of virus loads

4



We used ddPCR to measure IHNV loads in fish tissue samples without the need of standard curves

5



We compared the results of ddPCR with those from a published real-time PCR assay

6



The study presented here is the first report of a RT-ddPCR assay for the detection of IHNV RNA

7 8 9 10 11

ABSTRACT:

12

Infectious hematopoietic necrosis virus (IHNV) is an important pathogen of salmonid fishes. A validated universal reverse transcriptase quantitative

13

PCR (RT-qPCR) assay that can quantify levels of IHNV in fish tissues has been previously reported. In the present study, we adapted the published set

14

of IHNV primers and probe for use in a reverse-transcriptase droplet digital PCR (RT-ddPCR) assay for quantification of the virus in fish tissue

15

samples. The RT-ddPCR and RT-qPCR assays detected 13 phylogenetically diverse IHNV strains, but neither assay produced detectable amplification

16

when RNA from other fish viruses was used. The RT-ddPCR assay had a limit of detection (LOD) equating to 2.2 plaque forming units (PFU) / µl

17

while the LOD for the RT-qPCR was 0.2 PFU / µl. Good agreement (69.4 – 100%) between assays was observed when used to detect IHNV RNA in

18

cell culture supernatant and tissues from IHNV infected rainbow trout (Oncorhynchus mykiss) and arctic char (Salvelinus alpinus). Estimates of RNA

19

copy number produced by the two assays were significantly correlated but the RT-qPCR consistently produced higher estimates than the RT-ddPCR.

20

The analytical properties of the N gene RT-ddPCR test indicated that this method may be useful to assess IHNV RNA copy number for research and

21

diagnostic purposes. Future work is needed to establish the within and between laboratory diagnostic performance of the RT-ddPCR assay.

1 2 3

Keywords: IHNV; analytical sensitivity; analytical specificity; RT-ddPCR; RT-qPCR

1

1. Introduction

2

Infectious hematopoietic necrosis virus (IHNV) is a major pathogen of rainbow and steelhead trout (Oncorhynchus mykiss), Chinook salmon (O.

3

tshawytscha), sockeye and kokanee salmon (O. nerka) and other salmonid fish species (Wolf, 1988). IHNV has been detected in North America, Asia

4

and Europe (OIE, 2015). In July 2016, the first occurrence of IHNV was reported in Nairobi, Kenya, in the Southern Hemisphere (OIE, 2016). Like all

5

members of the family Rhabdoviridae, IHNV is a linear, non-segmented, single-stranded, negative-sense RNA virus. Phylogenetic analyses of IHNV

6

isolates from many areas of the world where the virus is endemic or has been introduced have identified the existence of five major genogroups,

7

namely, U, M, L, E, and J (Enzmann et al., 2005; Jia et al., 2014; Kurath et al., 2003; Nishizawa et al., 2006). The J genogroup is widespread in the

8

salmon and trout farming regions of Asia and includes two subgroups, namely, J Nagano and J Shizuoka (Jia et al., 2014; Nishizawa et al., 2006).

9

A number of methods have been established for quantitative detection of IHNV. Plaque assay is the most established method for quantifying IHNV

10

(Batts et al., 1989; Fendrick et al., 1982). While having the advantage of estimating titers of infectious virus, the plaque assay is prone to limitations

11

such as variation in cell line sensitivity and assay turnaround time (Te et al., 2015), making the assay impractical for certain purposes (Purcell et al.,

12

2006). Reverse transcriptase quantitative PCR (RT-qPCR) methods are also widely used to estimate gene copy number for RNA viruses (Raso et al,

13

2014), including for IHNV (Dhar et al., 2008; Liu et al., 2008; Overturf et al., 2001; Purcell et al., 2006; Purcell et al., 2013; Yue et al., 2008). However,

14

the quantification of nucleic acids by qPCR depends on a calibration (standard) curve and there is no consistent standard material for every qPCR

15

method. Marked variation in assay performance characteristics and in materials used as calibration standards may prevent agreement between different

16

laboratories, even when testing identical material.

17

Recently, diagnostic assays based on droplet digital polymerase chain reaction (ddPCR) technology are increasingly being used for detecting nucleic

18

acids of pathogens (Coudray-Meunier et al., 2015; Kiselinova et al., 2014; Racki et al., 2014; Sedlak et al., 2014; Yang et al., 2014), including RNA

19

viruses (Coudray-Meunier et al., 2015; Racki et al., 2013; Sedlak et al., 2014). The ddPCR has many potential advantages over qPCR, as it provides an

20

accurate, sensitive and specific measure of target DNA molecules without the need for a standard curve (Hayden et al., 2015). The ddPCR is an

21

end-point measurement and the signal is measured only after finishing the PCR amplification reaction. The ddPCR has also been shown to have

22

increased precision over qPCR (Hindson et al., 2013), and improved sensitivity to detect rare targets at low copy numbers (Hindson et al., 2011; Whale

23

et al., 2012). To our knowledge, the study presented here is the first report of a RT-ddPCR assay for the detection of IHNV RNA.

1

The aim of this study was to evaluate the potential of RT-ddPCR for accurate quantification of IHNV RNA copy number. Performance of the

2

RT-ddPCR method, using a commercialized droplet digital PCR platform, was compared with an established RT-qPCR assay for IHNV that targets the

3

N gene. We compared the analytical properties of specificity and sensitivity (limit of detection) for both the RT-ddPCR and RT-qPCR assays, and

4

evaluated agreement between the two tests. We also assessed the suitability of RT-ddPCR and RT-qPCR for quantitative detection of IHNV in tissues

5

of infected juvenile rainbow trout (albino phenotype) and arctic char (Salvelinus alpinus).

6

2. Material and methods

7

2.1. Virus

8

The IHNV strain Ch20101008 (passage 3) was isolated from a diseased brook trout (Salvelinus fontinalis) in 2010 in Jilin province in China (Jia et

9

al., 2014; Jia et al., 2013). The Ch20101008 isolate was propagated in Epithelioma papulosum cyprini (EPC) cell line (Fijan et al., 1983) at 15 °C

10

following OIE methods (OIE, 2015). The titer of the Ch20101008 isolate was determined by plaque assay using monolayers of EPC cells as previously

11

described (Batts et al., 1989). To evaluate analytical specificity of the RT-ddPCR and RT-qPCR assays, cell culture supernatant from a total of 13

12

IHNV strains representing the global diversity of IHNV was obtained, along with isolates of other fish viruses including, spring viremia of carp virus

13

(SVCV), viral hemorrhagic septicemia virus (VHSV), infectious pancreatic necrosis virus (IPNV), infectious salmon anemia virus (ISAV), hirame

14

rhabdovirus (HIRRV) and salmonid alphavirus (SAV) (Table 1). The ability of both assays to detect IHNV in the presence of other fish viruses was

15

tested by mixing an equal volume of IHNV cDNA (2.5 μl) with cDNA from one of the six viruses listed above.

16

2.2. RNA isolation

17

Viral RNA was isolated using the RNeasy Mini Kit (QIAGEN, CA, USA) according to the manufacturer’s instructions. For viral supernatant, 300

18

μl was extracted with a final elution volume of 50 μl. For tissues, the material was weighed and approximately 30 mg was homogenized in RLT buffer

19

(Qiagen, Germany) using a bead lysing matrix (MP Biomedicals, Germany), followed by elution in 50 μl. The remainder of the procedure was as

20

described by the manufacturer. The negative and blank controls for the extraction procedure consisted of non-infected fish tissues, non-infected cell

21

culture supernatant or nuclease-free water. The isolated RNA was stored at -80 °C until used.

1

2.3. Reverse transcription

2

Reverse transcription was initiated by incubating 8 μl RNA with 1 μl random primer, 1 μl 10 mM dNTP and 10 μl RNAse free water at 65 °C for 5

3

min followed by 5 min on ice. Next, 4 μl 5 × First-strand buffer, 1 μl RNaseOUTTM (Invitrogen, USA), 1 μl 10 mM DTT (Invitrogen, USA) and 1 μl

4

200 U / μl of SuperScriptTM Ⅲ RT (Invitrogen, USA) were added to a final volume of 20 μl. Tubes were incubated at 25 °C for 5 minutes, 50 °C for

5

60 minutes, and then at 70 °C for 15 minutes. The cDNA was stored at -80 °C until used.

6

2.4. Primers and probes

7

The RT-qPCR and RT-ddPCR assays used a published set of primers and probe that targets a conserved region (positions 798 – 879) of the IHNV N

8

gene (Table 2; Purcell et al., 2013). An arbitrary tag sequence probe was also included in the RT-qPCR assay that would bind to a site on the artificial

9

positive control (APC) plasmid to detect any contamination due to plasmid DNA (Snow et al., 2009)

10

2.5. RT-ddPCR

11

IHNV was quantified using the QX100TM Droplet DigitalTM PCR system (Bio-Rad, Pleasanton, CA). The RT-ddPCR reaction mixture consisted

12

of 10 μl of a 2 × ddPCR supermix (Bio-Rad, USA), each IHNV specific primer and probe at 500 nmol, and 2.5 μl of sample cDNA in a final volume of

13

20 μl. Each sample was applied in three technical replicate tubes to a final 20 μl reaction volume; copy number of an individual sample was based on

14

the mean of these three replicate measures. The entire reaction mixture was loaded into a disposable plastic cartridge (Bio-Rad, USA) together with 70

15

μl of droplet generation oil (Bio-Rad, USA) and placed in the droplet generator (Bio-Rad, USA). After processing, the droplets generated from each

16

sample were transferred to a 96-well PCR plate (Bio-Rad, USA). PCR amplification was carried out on the T100 thermal cycler (Bio-Rad, USA) using

17

a thermal profile of beginning at 95 °C for 10 min, followed by 40 cycles of 94 °C for 30 s and 60 °C for 45 s, 1 cycle of 98 °C for 10 min, and ending

18

at 12 °C. After amplification, the plate was loaded on the droplet reader (Bio-Rad, USA) and the droplets from each well of the plate were read

19

automatically at a rate of 32 wells/hour.

20

The RT-ddPCR data were analyzed with QuantaSoft TM analysis software (Bio-Rad, USA), and presented as the number of copies per μl of

1

RT-ddPCR mixture. The raw quantitative output (RNA copy number in the input sample) was normalized to either IHNV N gene copies / µl (viral

2

supernatant) or copies / g (tissue samples). Positive droplets, containing amplification products, were discriminated from negative droplets by applying

3

a fluorescence amplitude threshold in the QuantaSoft software (Bio-Rad, CA, USA). The threshold was set manually at the highest point of the

4

negative droplet cluster, as visualized using both the fluorescence amplitude vs. event number and the histogram of events vs. amplitude data streams,

5

on the FAM channel. A sample was considered positive by the assay if detectable amplification was observed in at least 2 of 3 technical replicates.

6

2.6. RT-qPCR

7

The RT-qPCR assay was performed using a 7500 Fast Real-Time PCR system (Applied Biosystems, CA, USA). Each sample was tested in triplicate

8

in a final 20 μl reaction volume; copy number of an individual sample was based on the mean of these three replicate measures. Each reaction

9

contained 10 μL of 2 × QuantiNovaTM (QN; Qiagen, Germany) Probe PCR master Mix, 0.2 μl of QN ROX Reference Dye, each IHNV specific primer

10

and FAM® labeled probe at 500 nmol, the VIC® labeled arbitrary tag sequence probe at 200 nmol, and 2.5 μl of sample cDNA in a final volume of 20

11

μl. The thermocycler parameters included 2 min at 95 °C, followed by 5 s at 95 °C, 45 s at 60 °C for 40 cycles. The IHNV APC plasmid for producing

12

RT-qPCR standard curves has been previously described (Purcell et al., 2013). Briefly, the APC plasmid was synthesized by Sangon Biotech Co., Ltd.

13

(Shanghai) and cloned into the pUC57 vector. The copy number calculations of APC plasmid are as previously described (Jia et al., 2010) and the

14

starting concentration of APC plasmid was 1.6 × 1011 copies / μl. The plasmid DNA was linearized by Nco I restriction enzyme, diluted to standard

15

copies per μl, then serially diluted in 10 - fold increments down to 1 copy per μl. All plasmid DNA was stored as aliquots at -80 °C until used. The

16

standard curves for each assay were made from separate master stocks from which serial dilutions (1.6×107 - 103 copies / μl dilution series of APC)

17

were made to obtain a 5 - point standard curve. Positive and negative controls and standard curves were included on all runs. Threshold and baselines

18

were set automatically by the software. Criteria to consider a sample positive for IHNV RNA and normalization of copy number were as described

19

above.

20

2.7. Serial dilutions of IHNV supernatant to assess analytical sensitivity

21

Viral supernatant containing the Ch20101008 strain of IHNV was used to create a 10 - fold dilution series in cell culture medium. After dilution, the

1

concentration of IHNV used in the PCR reactions ranged from 86,400 to 0.02 PFU / μl (Table 3). Procedures for RNA extraction, cDNA synthesis and

2

RT-qPCR or RT-ddPCR were performed as described above.

3

2.8. Viral challenge of rainbow trout and arctic char

4

Juvenile rainbow trout (albino phenotype; also called ‘golden trout’) and arctic char were reared and challenged with IHNV strain Ch20101008

5

under a protocol approved by the Institutional Animal Care and Use Committee (IACUC) of the Beijing Fisheries Research Institute, Beijing, China.

6

The fish were generously donated by Dr. Lin Luo (Beijing Fisheries Research Institute, China) and reared in sand-filtered spring water at a constant

7

temperature of 12 – 14 °C. The size of fish was approximately 15 g. For each species, fish were divided into two tanks (one mock and one virally

8

infected group) with each tank housing 20 fish. Fish were anesthetized with tricaine methanesulfonate (MS222; Argent Technologies), and injected in

9

the peritoneal cavity with either 100 μl medium (mock) or medium containing 3.1 × 104 TCID50 / ml of virus. After the challenge, the fish were held at

10

10 °C for the duration of the experiment. Fish were fed to satiation during the challenge with a commercial dry-pellet diet. Four or five fish per

11

treatment (mock and virally infected) were sampled at 72, 120, 168 and 240 hours post infection (hpi). The brain, gill, spleen, kidney, liver, heart,

12

intestines, muscle and swim bladder were collected and stored in -80 °C for RNA extraction. The tissues were pooled by organ type and treatment at

13

each time point and RNA extracted for analysis by the RT-ddPCR and RT-qPCR assays.

14

2.9. Statistical Analysis

15

All IHNV N gene copy numbers were log10 transformed and statistical analysis was performed using the SPSS software v13.0 (IBM) or InStat 3.0

16

(GraphPad). Linear regression was used to analyze the standard curves for RT-qPCR quantification. The association between IHNV N gene copies

17

estimated by RT-ddPCR and RT-qPCR or between gene copies and PFUs was evaluated by Pearson’s correlation analysis.

18

3. Results

19

3.1. Comparative analytical specificity and sensitivity

1

The RT-ddPCR and RT-qPCR assays detected all 13 IHNV strains that included representatives of all five global genogroups (Table 1). Neither

2

assay amplified template from VHSV, HIRRV, SVCV, IPNV, ISAV, or SAV (Table 1). The results indicated 100 % analytical specificity for both the

3

RT-ddPCR and RT-qPCR assays. The assay also specifically detected IHNV in simulated mixed infection samples, as evidenced by only minor

4

fluctuation in estimated copy numbers obtained from the mixed samples. The IHNV cDNA sample tested alone yielded an estimated log10 3.8 ±

5

0.01 and 5.4 ± 0.03 copies per reaction (by RT-ddPCR and RT-qPCR, respectively); when IHNV cDNA was mixed in equal volumes with cDNA

6

from VHSV, HIRRV, SVCV, IPNV, ISAV and SAV, the log10 copy estimates were 3.8 ± 0.01, 3.9 ± 0.03, 3.8 ± 0.01, 3.8 ± 0.02, 3.8 ±

7

0.05 and 3.7 ± 0.02 (P = 1.0) by RT-ddPCR and 5.4 ± 0.00, 5.7 ± 0.03, 5.8 ± 0.03, 5.5 ± 0.00, 5.6 ± 0.00, and 5.4 ± 0.03 (P = 0.39) by

8

RT-qPCR, respectively.

9

Serially diluted cDNA derived from IHNV supernatant demonstrated good PCR efficiency (E) for both the RT-ddPCR and RT-qPCR assays of

10

108.7 % and 103.1 %, respectively, across the range tested (Fig. 1 and Table 3). Estimated plaque assay and RT-qPCR quantities were of similar

11

magnitude with significant correlation between IHNV cDNA standard copy number and the IHNV titer (R2 = 0.99; P < 0.001; degrees of freedom (df) =

12

25). The RT-ddPCR assay yielded a similarly high correlation (R2 = 0.99; P < 0.001; df = 25) between plaque assay titer and estimated copy number,

13

but the estimated N gene copy number was approximately a log lower than the plaque assay estimates (Table 3). The RT-qPCR reliably detected virus

14

in dilutions containing from 86,400 to 2.2 PFU / μl, but the dilution containing 0.2 PFU / μl produced a positive reaction in only two of three replicate

15

wells (Table 3). The RT-ddPCR also reliably detected IHNV over the same range, but only one replicate well was positive at the 0.2 PFU / μl dilution.

16

Because the criterion to consider a sample positive was detectable amplification in two of three replicate wells, the limit of detection (LOD) for the

17

RT-qPCR assay was considered to be 0.2 PFU / μl, while the LOD for the RT-ddPCR was considered to be 2.2 PFU / μl (Table 3). The dynamic range

18

of the RT-qPCR assay was ~ 5 log with linearity observed in dilutions spanning 1.6 × 105 – 6.6 N gene copies / μl. The dynamic range of the

19

RT-ddPCR was ~ 3 logs with linearity observed in dilutions spanning 5.6 × 103 – 0.6 N gene copies / μl (Table 3 and Fig. 1).

20

The RT-qPCR assay produced estimates of IHNV N gene copies / μl that were approximately 6 to 11-fold (mean 8.4 - fold) greater than did the

21

RT-ddPCR assay (Table 3). Significant correlation was observed between log copies / μl estimated by RT-qPCR and RT-ddPCR (P < 0.001; R2 = 0.98;

22

df = 25). The precision of these estimates of gene copy number, as measured by the coefficient of variation (% CV), was similar between the two

23

assays at higher copy numbers (e.g. > 21,600.0 copies / μl), but the RT-ddPCR had slightly higher CV values at the lowest comparable copy number

1

(e.g. 2.2 copies / μl) relative to the RT-qPCR.

2

3.2. Evaluation of RT-ddPCR and RT-qPCR assays for the detection of IHNV RNA in infected fish tissues

3

In rainbow trout, IHNV RNA was detected in all tissues examined at all time points except for the 72 hpi muscle RNA sample that was not positive

4

by either assay and the 168 hpi gill sample that was not positive by RT-qPCR (Fig. 2A, Supplemental Table 1). No IHNV RNA was detected in the

5

mock control rainbow trout (data not shown). The no template control (NTC) lacked detectable amplification by RT-qPCR but 2 positive droplets

6

(equating to 0.10 – 0.12 copies / reaction) was observed in one run of the RT-ddPCR assay (that included the 72 hpi arctic char samples). Estimates of

7

the IHNV N gene copy number in the positive samples ranged from 2.3×104 – 2.1×108 copies / g by RT-ddPCR while the RT-qPCR values in the

8

positive samples ranged from 8.3 × 105 to 1.3 × 1010 (Supplemental Table 1). At the earliest time point, the highest copy number was found in the swim

9

bladder while the heart had the highest copy numbers at the final time point. The estimated IHNV N gene copy numbers in the samples that tested

10

positive by both assays were significantly correlated (P < 0.001; R2 = 0.93; df = 32). The RT-ddPCR detected IHNV in 35/36 (97.2 %) pools of tissues

11

from infected fish while the RT-qPCR detected virus in 34/36 (94.4 %) pools; overall agreement between the two assays was 97.2 %. There was no

12

consistent trend to suggest that the RT-ddPCR was any more or less precise at estimating gene copy number relative to the RT-qPCR assay (as

13

measured by the CV produced among technical replicates; Supplemental Table 1).

14

In arctic char, IHNV RNA was detected in all tissues of exposed fish but not at all time points (Fig. 2B; Supplemental Table 2). No IHNV RNA was

15

detected in the tissues from mock exposed (control) arctic char or in blank controls (data not shown). IHNV RNA was detected in 29/36 (80.6 %)

16

samples by the RT-ddPCR with the estimated number of N gene copies in the positive samples ranging from 2.3 × 103 to 1.3 × 106 copies / g. The

17

RT-qPCR detected IHNV RNA in 28/36 (77.8 %) of the samples and the number of N gene copies ranged from 4.8 × 104 to 1.8 × 107 copies / g. The

18

highest IHNV copy number at 72 hpi was in the swim bladder; interestingly, no IHNV RNA was detected in the kidney at the early time points (Fig.

19

2B). In general, the viral load in arctic char was lower than in rainbow trout (Fig. 2). Among the 36 pools of tissues from infected char, there was

20

69.4 % agreement among the two assays with two samples testing negative by both assays, five samples testing negative by RT-qPCR and six samples

21

testing negative by RT-ddPCR. The estimated IHNV N gene copy numbers in the samples that tested positive by both assays were significantly

22

correlated (P < 0.001; R2 = 0.70; df = 21). There was no general trend towards reduced variation in quantity estimates in the RT-ddPCR relative to the

23

RT-qPCR (as evidenced by the CV of the technical replicate wells; Supplemental Table 2).

1

4. Discussion

2

The RT-ddPCR and RT-qPCR assays reported in this study utilized the IHNV N gene primers and probe sequences from a previously published

3

reverse transcriptase real-time PCR (RT-rPCR) assay that has been subjected to both analytical and diagnostic validation (Purcell et al., 2013). The

4

distinction between the use of the term RT-qPCR in the present study and the RT-rPCR term used by Purcell et al. (2013), was that the previous study

5

did not emphasize copy number estimation for diagnostic purposes and interpreted diagnostic test samples as positive or negative (Purcell et al., 2013).

6

The IHNV RT-rPCR assay was shown previously to have good analytical specificity (ASp), as evidenced by its ability to detect 30 distinct strains of

7

the virus that span its known geographic and phylogenetic range along with a lack of detection of several closely related rhabdoviruses. Additionally,

8

the RT-rPCR was reported to have high analytical sensitivity (ASe) with a limit of detection between 5 to 70 copies depending on the template. In the

9

present study, we also demonstrated 100 % ASp of the RT-ddPCR and RT-qPCR assays using 13 phylogenetically diverse strains of IHNV, 3 related

10

rhabdoviruses and 3 unrelated fish viruses. We observed no inhibition in the ability of either assay to detect IHNV in the presence of other viruses

11

using simulated mixed infection samples. Both assays also had good ASe with the RT-ddPCR detecting viral RNA in cell culture supernatant diluted to

12

2.2 PFU / µl while the RT-qPCR detected RNA in supernatant diluted to 0.2 PFU / µl. The IHNV N gene copy number estimated by RT-qPCR was

13

approximately 8.2 - folder higher than estimates obtained by RT-ddPCR. Nonetheless, the two assays detected viral RNA copy numbers in the range of

14

0.6 to the 6.6 copies / µl, which is more sensitive than the range reported by Purcell et al. (2013). The RT-qPCR assay had a linear dynamic range of

15

approximately five orders of magnitude. The droplet digital PCR system works by counting the number of positive and negative droplets among

16

approximately 20,000 droplets per reaction and analyzing the data by applying a Poisson distribution, so the reduced dynamic range of a droplet digital

17

PCR system can be a limitation (Hindson et al., 2011). As predicted, the RT-ddPCR had a more compressed dynamic range (~3 orders of magnitude).

18

Both assays had comparable precision with more variation in quantity estimates at lower copy numbers but less variation at higher copy numbers.

19

Moderate to high agreement was observed between the RT-ddPCR and RT-qPCR assays when applied to IHNV infected fish tissues. Tissues

20

were derived from rainbow trout (albino phenotype) and arctic char; arctic char are not a principle host for IHNV (OIE, 2015) and this species

21

appeared less susceptible in our challenge studies. In rainbow trout, most of the fish tested positive (94 – 97 %) with a high agreement between the two

22

assays (97.2 %). In arctic char, fewer fish tested positive (78 – 81 %) with generally lower virus copy numbers; a more moderate agreement (69.4 %)

1

was observed between the two tests. All of the arctic char samples with disconcordant test results had very low copy numbers and were near the LOD

2

for both assays. However, samples near the LOD should be interpreted cautiously as some negative control samples may contain single positive

3

droplets by RT-ddPCR leading to false positive results (Kiselinova et al., 2014). In the present study, negative template controls (NTCs) were included

4

in all runs and any positive events within NTCs were subtracted when we analyzed the test groups. These false positive events were detected randomly,

5

and the positive droplets exhibited different fluorescence height. Nonetheless, the RT-ddPCR may yield false positive test results when used to

6

quantify low viral loads near or below the LOD (Kiselinova et al., 2014; Majumdar et al., 2015; Pender et al., 2015; Strain et al., 2013).

7

One of the primary advantages for ddPCR is accurate quantification of gene copy numbers in a nucleic acid sample, without being dependent on

8

external calibrators (Dong et al., 2014; Hayden et al., 2015). In the present study, the estimates of N gene copy numbers produced by the RT-qPCR and

9

RT-ddPCR assays were significantly correlated, but the RT-qPCR assay produced consistently higher estimates (e.g. ~8 fold for viral supernatant).

10

This discrepancy has previously been reported by others (Hayden et al., 2013; Strain et al., 2013; Yang et al., 2014). The RT-ddPCR and RT-qPCR

11

assays estimate copy number in a fundamentally different manner with the former assay relying on counting positive droplets and a determination of

12

copy number with Poisson statistics and the latter assay utilizing a standard curve based on a reference standard (e.g. plasmid DNA). One potential

13

cause for the copy number discrepancies could be errors introduced by the use of spectrophotometric determination of the nucleic acid concentration

14

leading to an overestimation of the plasmid copy number in the reference standard (Corbisier et al., 2015; Sanders et al., 2013). The requirement of the

15

RT-qPCR for an exogenous standard curve, can impact quantification with this method if the input concentration is inaccurate (Yang et al., 2014).

16

Additionally, primer-template or probe-template mismatches may have a direct effect on RT-qPCR-based quantification (Lefever et al., 2013), but the

17

RT-ddPCR assay is less susceptible to these effects (Strain et al., 2013). It is typically assumed that RT-ddPCR copy number is a more accurate

18

estimate of the true copy number. For IHNV, as for several other analogous assays, DNA standards are usually obtained using dilutions of recombinant

19

plasmids containing IHNV gene sequences (Purcell et al., 2013; Purcell et al., 2006). There is not a certified reference material for IHNV

20

quantification, which might lead to a discrepancy in the RT-qPCR quantification of IHNV among different laboratories. Marked variation in assay

21

performance characteristics and in materials used as calibration standards may prevent agreement between different laboratories, even when testing

22

identical material (Dong et al., 2014; Hayden et al., 2013). This is one reason that many diagnostic laboratories prefer to interpret real-time PCR results

1

as simply a positive or negative test result (Purcell et al., 2013).

2

Previous studies have compared infectious units (determined by plaque assay) to virus load (determined as target gene copy number) (Purcell et

3

al., 2006). For instance, using an RT-qPCR specifically targeting the negative-sense RNA genome of IHNV, it was determined that the genome copy

4

number in the liver of infected fish was approximately 8,000 - fold higher than the infectious titer (PFU), similar to that reported for other virus

5

systems (Purcell et al., 2006). In the present study, using viral supernatant that should contain a high ratio of infectious to non-infectious virions, the

6

target copy number determined by RT-qPCR was ~ 8-fold higher than the number of PFUs determined by plaque assay while the copy number

7

determined by RT-ddPCR was lower at ~0.3-fold the PFU values, suggesting the RT-ddPCR may be underestimating the true copy number. This range

8

was similar to that reported for VHSV genotype IVb where the number of VHSV N gene copies as estimated by RT-qPCR ranged from 0.4- to 1.3-fold

9

different than the number of PFUs determined by a plaque assay (Hope et al., 2010). Several factors have been shown to influence this ratio such as the

10

levels of messenger RNA in the sample (Fleige et al., 2006), number of detective interfering particles (Thompson et al., 2010), and extraction and

11

reverse transcriptase efficiencies (Bengtsson et al., 2008; Hindson et al., 2013).

12

In summary, nanoliter-sized droplet technology paired with digital PCR (ddPCR) holds promise for precise, sensitive and stable quantification of

13

IHNV loads in fish sample, without the need of standard curves. Further work is needed to establish if RT-ddPCR quantity estimates are more

14

repeatable within and among laboratories. Further comparisons of RT-ddPCR with the current reference procedures are needed to more effectively

15

evaluate the possibility of using ddPCR to routinely monitor IHNV in various fish samples.

16

Acknowledgement

17

This work was funded by the World Organization for Animal Health (OIE) Laboratory Twinning Project for Infectious Hematopoietic Necrosis

18

Virus, Shenzhen Scientific Research Project (GJHZ20150316155421411) and the Quality Inspection Programs of Scientific Research Project (No.

19

2012IK032, 201210055, 2015IK265). We would like to thank Jinshun Yang for her technical assistance. Use of any trade, firm, or product names is for

20

descriptive purposes only and does not imply endorsement by the U.S. Government.

1

Reference

2 3

Batts, W.N. and Winton, J.R., 1989. Enhanced detection of infectious hematopoietic necrosis virus and other fish viruses by pretreatment of cell monolayers with polyethylene glycol. J. Aquat. Anim. Health 1, 284-290.

4 5

Bengtsson, M., Hemberg, M., Rorsman, P. and Stahlberg, A., 2008. Quantification of mRNA in single cells and modelling of RT-qPCR induced noise. BMC Mol. Biol. 9(1), 63.

6 7 8

Corbisier, P., Pinheiro, L., Mazoua, S., Kortekaas, A.M., Chung, P.Y., Gerganova, T., Roebben, G., Emons, H. and Emslie, K., 2015. DNA copy number concentration measured by digital and droplet digital quantitative PCR using certified reference materials. Anal. Bioanal. Chem. 407(7), 1831-1840.

9 10

Coudray-Meunier, C., Fraisse, A., Martin-Latil, S., Guillier, L., Delannoy, S., Fach, P., Perelle, S., 2015. A comparative study of digital RT-PCR and RT-qPCR for quantification of Hepatitis A virus and Norovirus in lettuce and water samples. Int. J. Food Microbiol. 201, 17-26.

11 12

Dhar, A.K., Bowers, R.M., Licon, K.S. and Lapatra, S.E., 2008. Detection and quantification of infectious hematopoietic necrosis virus in rainbow trout (Oncorhynchus mykiss) by SYBR Green real-time reverse transcriptase-polymerase chain reaction. J. Virol. Methods 147(1), 157-166.

13 14

Dong, L., Meng, Y., Wang, J. and Liu, Y., 2014. Evaluation of droplet digital PCR for characterizing plasmid reference material used for quantifying ammonia oxidizers and denitrifiers. Anal. Bioanal. Chem. 406(6), 1701-1712.

15 16

Enzmann, P.J., Kurath, G., Fichtner, D., Bergmann, S. M., 2005. Infectious hematopoietic necrosis virus: monophyletic origin of European isolates from North American genogroup M. Dis. Aquat. Org. 66(3), 187-195.

17 18

Fendrick, J.L., Groberg, W.J., Leong, J.C., 1982. Comparative sensitivity of five fish cell lines to wild type infectious haematopoietic necrosis virus from two Oregon sources. J. Fish. Dis. 5(2), 87-95.

19 20

Fijan, N., Sulimanovi, D., Bearzotti, M., Muzini, D., Zwillenberg, L.O., Chilmonczyk, S., Vautherot, J.F., De Kinkelin, P., 1983. Some properties of the epithelioma papulosum cyprini (EPC) cell line from carp (Cyprinus carpio), Ann. de l'Institut Pasteur / Virol. 134, 207-220.

21

Fleige, S. and Pfaffl, M.W., 2006. RNA integrity and the effect on the real-time qRT-PCR performance. Mol. Aspects Med. 27(2), 126-139.

22 23

Hayden, R.T., Gu, Z., Ingersoll, J., Abdul-Ali, D., Shi, L., Pounds, S., Caliendo, A.M., 2013. Comparison of droplet digital PCR to real-time PCR for quantitative detection of cytomegalovirus. J. Clin. Microbiol. 51(2), 540-546.

24 25

Hayden, R.T., Gu, Z., Sam, S.S., Sun, Y., Tang, L., Pounds, S., Caliendo, A.M., 2015. Comparative evaluation of three commercial quantitative cytomegalovirus standards by use of digital and real-time PCR. J. Clin. Microbiol. 53(5), 1500-1505.

26 27 28 29

Hindson, B.J., Ness, K.D., Masquelier, D.A., Belgrader, P., Heredia, N.J., Makarewicz, A.J., Bright, I.J., Lucero, M.Y., Hiddessen, A.L., Legler, T.C., Kitano, T.K., Hodel, M.R., Petersen, J.F., Wyatt, P.W., Steenblock, E.R., Shah, P.H., Bousse, L.J., Troup, C.B., Mellen, J.C., Wittmann, D.K., Erndt, N.G., Cauley, T.H., Koehler, R.T., So, A.P., Dube, S., Rose, K.A., Montesclaros, L., Wang, S., Stumbo, D.P., Hodges, S.P., Romine, S., Milanovich, F.P., White, H.E., Regan, J.F., Karlin-Neumann, G.A., Hindson, C.M., Saxonov, S., Colston, B.W., 2011. High-throughput droplet

digital PCR system for absolute quantitation of DNA copy number. Anal. Chem. 83(22), 8604-8610.

1 2 3

Hindson, C.M., Chevillet, J.R., Briggs, H.A., Gallichotte, E.N., Ruf, I.K., Hindson, B.J., Vessella, R.L., Tewari, M., 2013. Absolute quantification by droplet digital PCR versus analog real-time PCR. Nature Methods 10(10), 1003-1005.

4 5

Hope, K.M., Casey, R.N., Groocock, G.H., Getchell, R.G., Bowser, P.R., Casey, J.W., 2010. Comparison of quantitative RT-PCR with cell culture to detect viral hemorrhagic septicemia virus (VHSV) IVb infections in the Great Lakes. J. Aquat. Anim. Health 22(1), 50-61.

6 7 8

Jia, P., Zheng X.C., Shi X.J., Kan S.F., Wang J.J., He J.Q., Zheng W., Yu L., Lan W.S., Hua Q.Y., Liu H., Jin N.Y., 2014. Determination of the complete genome sequence of infectious hematopoietic necrosis virus (IHNV) Ch20101008 and viral molecular evolution in China. Infect. Genet. Evol. 27, 418-431.

9 10

Jia, P., Jin N.Y., Li, X., Zhu, G.Z., Liu, Y., Gao, P., Xu, X.H., Yang, E.C., Meng, R.Z., Kan, S.F., 2010. Establishment of fluorescent real-time quantitative RT-PCR assay for detection of genetype 4 hepatitis E virus in swine stools. Chin. J. Virol. 26(1), 98-103.

11 12

Jia, P., Zhu, X., Zheng, W., Zheng, X.C., Shi, X.J., Lan, W.S., Kan, S.F., Hua, Q.Y., Liu, H., Chen, X.X., 2013. Isolation and genetic typing of infectious hematopoietic necrosis virus from cultured brook trout (Salvelinus fontinalis) in China. Bull, Eur, Ass. Fish Pathol, 33(5), 150-157.

13 14

Kiselinova, M., Pasternak, A.O., De Spiegelaere, W., Vogelaers, D., Berkhout, B. Vandekerckhove, L., 2014. Comparison of droplet digital PCR and seminested real-time PCR for quantification of cell-associated HIV-1 RNA. PLoS One 9(1), e85999.

15 16

Kurath, G., Garver, K.A., Troyer, R.M., Emmenegger, E.J., Einer-Jensen, K., Anderson, E. D., 2003. Phylogeography of infectious haematopoietic necrosis virus in North America. J. Gen. Virol. 84, 803-814.

17 18

Lefever, S., Pattyn, F., Hellemans, J., Vandesompele, J., 2013. Single-nucleotide polymorphisms and other mismatches reduce performance of quantitative PCR assays. Clin. Chem. 59, 1470-1480.

19 20

Liu, Z., Teng, Y., Liu, H., Jiang, Y., Xie, X., Li, H., Lv, J., Gao, L., He, J., Shi, X., 2008. Simultaneous detection of three fish rhabdoviruses using multiplex real-time quantitative RT-PCR assay. J. Virol. Methods 149(1), 103-109.

21 22

Majumdar, N., Wessel, T. Marks, J., 2015. Digital PCR modeling for maximal sensitivity, dynamic range and measurement precision. PLoS One 10(3), e0118833.

23 24

Nishizawa, T., Kinosha, S., Kim, W-S., Higashi, S., Yoshimizu, M., 2006. Nucleotide diversity of Japanese isolates of infectious hematopoietic necrosis virus (IHNV) based on the glycoprotein gene. Dis. Aquat. Org. 71(3), 267-272.

25

OIE, 2015. Infectious haematopoietic necrosis, Manual of Diagnostic Tests for Aquatic Animals. World Organization for Animal Health, Paris.

26 27

OIE,

28 29

Overturf, K., LaPatra, S., Powell, M.,, 2001. Real-time PCR for the detection and quantitative analysis of IHNV in salmonids. J. Fish Dis. 24(6), 325-333.

30

Pender, A., Garcia-Murillas, I., Rana, S., Cutts, R.J., Kelly, G., Fenwick, K., Kozarewa, I., Gonzalez de Castro, D., Bhosle, J., O'Brien, M., Turner,

2016. WAHIS Interface, Disease information of infectious haematopoietic http://www.oie.int/wahis_2/public/wahid.php/Reviewreport/Review/viewsummary?fupser=&dothis=&reportid=20560

necrosis.

1 2

N.C., Popat, S., Downward, J., 2015. Efficient genotyping of KRAS mutant non-small cell lung cancer using a multiplexed droplet digital PCR approach. PLoS One 10(9), e0139074.

3 4

Purcell, M.K., Hart, S.A., Kurath, G., Winton, J.R., 2006. Strand-specific, real-time RT-PCR assays for quantification of genomic and positive-sense RNAs of the fish rhabdovirus, Infectious hematopoietic necrosis virus. J. Virol. Methods 132(1), 18-24.

5 6

Purcell, M.K., Thompson, R.L., Garver, K.A., Hawley, L.M., Batts, W.N., Sprague, L., Sampson, C., Winton, J.R., 2013. Universal reverse-transcriptase real-time PCR for infectious hematopoietic necrosis virus (IHNV). Dis. Aquat. Org. 106(2), 103-15.

7 8 9

Rački N., Morisset, D., Gutierrez-Aguirre, I., Ravnikar, M., 2014. One-step RT-droplet digital PCR: a breakthrough in the quantification of waterborne RNA viruses. Anal Bioanal Chem 406(3), 661-7. Raso, A., Biassoni, R., 2014. Twenty years of qPCR: a mature technology? Methods Mol. Biol. 1160, 1-3.

10

Sanders, R., Mason, D.J., Foy, C.A., Huggett, J.F., 2013. Evaluation of digital PCR for absolute RNA quantification. PLoS One 8(9), e75296.

11 12

Sedlak, R.H., Cook, L., Huang, M.L., Magaret, A., Zerr, D.M., Boeckh, M., Jerome, K.R., 2014. Identification of chromosomally integrated human herpesvirus 6 by droplet digital PCR. Clin. Chem. 60(5), 765-772.

13 14

Snow, M., McKay, P., Matejusova, I., 2009. Development of a widely applicable positive control strategy to support detection of infectious salmon anaemia virus (ISAV) using Taqman real-time PCR. J. Fish Dis. 32(2), 151-156.

15 16

Strain, M.C., Lada, S.M., Luong, T., Rought, S.E., Gianella, S., Terry, V.H., Spina, C.A., Woelk, C.H., Richman, D.D., 2013. Highly precise measurement of HIV DNA by droplet digital PCR. PLoS One 8(4), e55943.

17 18

Te, S.H., Chen, E.Y., Gin, K.Y., 2015. Comparison of Quantitative PCR and Droplet Digital PCR Multiplex assays for two genera of bloom-forming cyanobacteria, Cylindrospermopsis and Microcystis. Appl. Environ. Microbiol. 81(15), 5203-5211.

19

Thompson, K.A., Yin, J., 2010. Population dynamics of an RNA virus and its defective interfering particles in passage cultures. Virol. J. 7(1), 257-267.

20 21

Whale, A.S., Huggett, J.F., Cowen, S., Speirs, V., Shaw, J., Ellison, S., Foy, C.A., Scott, D.J., 2012. Comparison of microfluidic digital PCR and conventional quantitative PCR for measuring copy number variation. Nucleic Acids Res. 40(11), e82.

22

Wolf, K., 1988. Infectious hematopoietic necrosis. In: Fish Viruses and Fish Viral Diseases. Cornell University Press, Ithaca, New York.

23 24

Yang, R., Paparini, A., Monis, P., Ryan, U., 2014. Comparison of next-generation droplet digital PCR (ddPCR) with quantitative PCR (qPCR) for enumeration of Cryptosporidium oocysts in faecal samples. Int. J. Parasitol. 44(14), 1105-1113.

25 26

Yue, Z., Liu, H., Liang, C., Gao, H., Xu, B., Deng, M., Jiang Y., 2008. Real-time quantitative RT-PCR assay for detection of IHNV in Fish. Acta. Hydrobiol. Sinica 32(1), 91-95.

27 28

1 2 3 4

Figure legends

5

Fig. 1. Comparison of copy number estimated by RT-ddPCR and RT-qPCR for serially diluted cDNA derived from cell culture fluid containing IHNV

6

(strain Ch20101008). The log10-transformed viral titer (PFU / µl) of serially diluted IHNV supernatant was plotted against the corresponding

7

log10-transformed cDNA copy numbers determined by either RT-ddPCR or RT-qPCR.

8

Fig. 2. IHNV N gene copies (log10) in the tissues of (A) rainbow trout or (B) arctic char exposed to IHNV detected by either the RT-ddPCR or

9

RT-qPCR assay. Each bar represents a pool of 4 or 5 fish (pool tested as a single sample in each assay).

1

1

1

1

Table 1

2

Specificity of the RT-ddPCR and RT-qPCR assays Viral species

IHNV

VHSV HIRRV IPNV SVCV ISAV SAV

Genogroup

Subgroup

Isolate

Geographic area

J J J U U M M M M

Nagano Shizuoka Nagano C P N A B C

Ch20101008 RtPy91 RtNag06a RB1 BLk94 LR80 WRAC 220-90 C30

China Korea Japan Oregon USA WA, USA Washington USA Idaho USA Idaho USA Idaho, USA

2010 1991 2006 1975 1994 1980 1982 1990 1991

M

D

Mer95

CRB, WA

1995

L

I

C18

California USA

1991

L E I Ia 1

II a -

FR0031 4008 DK613 080113 578 461 Glesvaer2/90 F93-125

California USA Italy Denmark Shandong, China Spain China Norway Ireland

2000 1989 1991 2008 2004 2004 1995 1993

3

a

4

of positive to negative nanoliter droplets.

5

b

Isolation year

GenBank accession no. KJ421216 AB288204 AB510195 U50401 DQ164100 L40878 L40883 GQ413939 AF237984 Unsubmitted (Breyta et al., 2016) Unsubmitted (Purcell et al., 2013) DQ164102 Unsubmitted (Bovo et al., 1987) AY546593 FJ376982 AJ489228 DQ097384 HQ259676 AJ316244

Mean detected targetsa (copies / μl) by RT-ddPCR 7.1×102 1.3×103 5.5×103 5.7×103 4.7×102 8.2×104 2.3×102 2.2×102 5.5×103

Mean detected targetsb (copies / μl) by RT-qPCR 4.1×103 4.6×104 2.8×105 5.4×105 2.3×104 1.4×106 8.9×103 6.1×103 1.9×105

1.3×103

3.9×104

2.0×103

7.1×104

3.8×102 4.2×102 0.00 0.00 0.00 0.00 0.00 0.00

1.4×104 1.5×104 0.00 0.00 0.00 0.00 0.00 0.00

Number of target IHNV RNA molecules in each microliter (copies / μl) of viral supernatant ss measured by RT-ddPCR and estimated by applying a Poisson distribution to the ratio

Number of target IHNV RNA molecules (copies / μl) as measured by RT-qPCR and estimated from a standard curve where y = 3.8x + 43.3, R2 = 0.998.

1 2 3

Table 2 Primers and probes used in this study Name

Sequence

IHNV N 796F primer

AGAGCCAAGGCACTGTGCG

IHNV N 875R primer

TTCTTTGCGGCTTGGTTGA

IHNV N 818MGB probe

FAM-TGAGACTGAGCGGGACA-NFQ/MGB

Arbitrary tag sequence probe

VIC-ACCGTCTAGCATCCAGT-NFQ/MGB

4 5 6 7 8 Sample a (PFU / μl )

86400

43200.0 21600.0

10800.0 5400.0 2160.0

216.0 21.6

2.2 0.2 0.02

Table 3 Comparative sensitivity of the RT-ddPCR and RT-qPCR assays Positive droplets / number of droplets analyzed (in 20 μl) b 12575 / 12591 11672 / 11701 8717 / 8730 11125 / 11183 9263 / 9291 11448 / 11474 10873 / 10936 9732 / 9785 9909 / 9980 10575 / 11837 9344 / 10567 8055 / 9080 5776 / 9036 5754 / 8756 5309 / 8366 5966 / 12974 4821 / 12028 5150 / 11716 655 / 11129 670 / 10241 834 / 11872 69 / 12395 81 / 13785 86 / 13257 10 / 12174 7 / 14201 4 / 13567 0 / 13892 1 / 14257 0 / 13499 0 / 12400 0 / 13340 0 / 12375

RT-ddPCR Detected targets c (RNA copies / μl) 7330.0 6590.0 7150.0 5780.0 6380.0 6690.0 5700.0 5730.0 5460.0 2460.0 2370.0 2400.0 1120.0 1180.0 1110.0 677.0 563.0 636.0 66.7 74.4 80.0 6.1 6.5 7.1 0.9 0.5 0.3 bd f 1.5 bd f bd f bd f bd f

RT-qPCR Mean ± SD

CV (%)

7023.3 ± 385.9

5.5

6283.3 ±462.6

7.4

5630.0 ±148.0

2.6

2410.0 ±45.8

1.9

1136.7 ±37.9

3.3

625.3 ±57.7

9.2

73.7 ± 6.7

9.1

6.6 ±0.5

7.9

0.6 ±0.3

49.9

/

/

/

/

22 / 23

Cq values d

Detected targets (RNA copies / μl) e

19.6 19.5 19.5 20.2 20.2 20.1 21.3 21.2 21.3 22.9 22.9 22.8 24.1 24.2 24.2

160650.60 172968.86 169803.46 114504.8 111719.3 117359.6 55717.2 58891.6 56407.5 21588.6 20678.0 22678.5 10061.8 9519.4 9578.2 3517.7 4085.4 4148.7 683.5 599.9 597.3 55.4 57.6 42.7 5.0 5.7 9.0 3.2 1.7 bd f bd f bd f bd f

25.8 25.6 25.6

28.5 28.7 28.7 32.6 32.5 33.0 36.5 36.2 35.5 37.2 38.3 bd f bd f bd f bd f

Mean ± SD

CV (%)

167807.6 ± 6397.1

3.8

114527.9 ±2820.2

2.5

57005.4 ±1669.5

2.9

21648.4 ±1001.6

4.6

9719.8 ±297.6

3.1

3917.3 ±347.5

8.9

626.9 ±49.1

7.8

51.9 ±8.1

15.5

6.6 ±2.1

32.0

2.4 ±1.1

44.4

/

/

0 / 11886 0 / 14094 0 / 13191

ddH2O

1 2 3 4 5 6 7 8 9 10 11 12

bd f bd f bd f

/

a

/

bd f bd f bd f

bd f bd f bd f

/

Titer of IHNV (strain Ch20101008) in plaque forming units (PFU) per µl as measured by plaque assay. b The number of droplets showing positive signal for IHNV / the total number of droplets in RT-ddPCR reaction. c Number of target IHNV RNA molecules per microliter (copies / μl) in a virus dilution series as measured by RT-ddPCR and estimated by applying a Poisson distribution to the ratio of positive to negative nanoliter droplets. d Cq: quantification cycle values for RT-qPCR. e Number of target IHNV RNA molecules (copies / μl) in a virus dilution series as measured by RT-qPCR and estimated from a standard curve where y = 3.8x + 43.3, R2 = 0.998. f bd = below detection (negative reaction).

13

23 / 23

/