Aquaculture production has grown 5-fold over the last three decades with a total of 82.1 million tons of aquatic animals produced in 2018, making it the fastest growing food sector in the world (Boyd and McNevin, 2015; FAO, 2020). Measuring by resource requirements of production, aquaculture can achieve a significantly reduced carbon footprint and is therefore considered a sustainable food source for the growing world population (Hai et al., 2018; Boyd et al., 2020; MacLeod et al., 2020). Atlantic salmon and rainbow trout are ranked first and second for global production volume for farmed marine finfish species, respectively (FAO, 2020), making salmonids one of the most successfully farmed fishes (Iversen et al., 2020). In New Zealand, Chinook or king salmon, Oncorhynchus tshawytscha, is one of the highest valued aquaculture products with a total annual revenue of NZ$ 254 million and growing (Aquaculture New Zealand, 2021a,b).

Aquaculture growth and rapid domestication of aquatic organisms come with an increased risk of diverse diseases triggered through elevated stress levels due to stocking densities and potential suboptimal environmental conditions (e.g., due to sea temperature rise or anomalies; Bateman et al., 2021; Feidantsis et al., 2021). For example, more than 20 potentially pathogenic taxa (viruses, bacteria, and parasites) have been recorded from farmed O. tshawytscha in New Zealand among which NZ Rickettsia-like organisms, Tenacibaculum maritimum, and Yersinia ruckeri (serogroup O1b) are actively managed through preventative measures such as controlled area notices and vaccines (Lane et al., 2020).

Fish disease surveillance usually involves fish, water, or sediment samples sent to commercial laboratories for testing using bacteriology, virology, histology, and molecular assays (Khor et al., 2021). A variety of molecular tools have been developed for disease surveillance from fish samples. For example, fluorescence in situ and other hybridization techniques have been used to screen for proliferative kidney disease (PKX) in salmonids (Morris et al., 2000); DNA microarrays coupled with conventional PCR have been used for herpesvirus and pathogenic Flavobacterium species in fish (Lievens et al., 2011), RNA viruses were targeted with reverse transcriptase PCR in shrimp (Poulos and Lightner, 2006) and many-related technologies such as real-time PCR, loop mediated isothermal amplification (LAMP), random amplified polymorphic DNA (RAPD), restriction/amplified fragment length polymorphism (R/AFLP), and genotyping have been used for pathogen detection in aquaculture (Kim et al., 2017). Overall, efficient, sensitive, and cost-effective methods guarantee competitiveness on the national and international market (Bozzi et al., 2021; Mordecai et al., 2021) but need to overcome complicated and timely protocols leading to inefficient real-life applicability (Law et al., 2014).

Novel molecular-based assays using droplet digital Polymerase Chain Reaction (ddPCR) can produce immediate results and high sample throughput without compromising detection sensitivity (Miotke et al., 2014; Cao et al., 2020). While traditional PCR technologies are currently applied by New Zealand’s Ministry for Primary Industries (MPI) for salmon aquaculture surveillance (Brosnahan, 2020), the advantages of ddPCR are now fully acknowledged and targeted assays are being used for routine monitoring and commercial applications across medicine to biosecurity (Rowlands et al., 2019; Wood et al., 2019; Kiefer et al., 2020; Lewin et al., 2020; Netzer et al., 2021; Orioles et al., 2022). The digital droplet PCR system QX100/QX200 (Bio-Rad, California, United States) is based on partitioning each sample (e.g., extracted DNA from fish tissue) into approximately 20,000 individual droplets, with each small reaction volume containing a single target DNA fragment, which minimizes inhibition (Mazaika and Homsy, 2014; Nathan et al., 2014). Additionally, partitioning into droplets enables absolute quantification of the targeted gene fragments to be conducted through direct measurement of DNA copy numbers, removing the need for replicates and standard curve extrapolation (Hindson et al., 2011). Finally, the possibility to multiplex several target genes into a single ddPCR reaction through two optical channels and adjusting the fluorescence signal of the different targets allows for significant time and cost savings when large datasets are being processed (Hughesman et al., 2016; Lewin et al., 2020).

Endemic pathogens have been identified as immediate and emerging concerns for New Zealand’s salmon aquaculture industry, given its freedom from exotic and notifiable disease agents (Aquaculture New Zealand, 2021a,b). For example, in 2015, up to 70% of salmon summer mortalities in the Marlborough Sounds of New Zealand were associated with bacterial pathogens including New Zealand Rickettsia-like organism (NZ-RLO) and T. maritimum (see Brosnahan et al., 2016). At least two strains of NZ-RLO have been associated with clinically diseased fishes; NZ RLO1, which shares 100% homology with Tasmanian RLO, and NZ-RLO 2 which is suggested to be the more virulent of the two strains (Brosnahan et al., 2019). Tenacibaculum maritimum is a Gram-negative filamentous bacterium that causes ulcerative skin disease, tenacibaculosis, and has been associated with high mortality in marine fishes (Avendaño-Herrera et al., 2006; Chapela et al., 2017). Virulence is likely associated with extreme environmental conditions such as high water temperatures and/or co-infections (Diggles, 2016). Another bacterial pathogen of concern in New Zealand O. tshawytscha farms is an endemic strain of Y. ruckeri (serotype O1b), which causes enteric red-mouth disease. This bacterium originates in freshwater hatcheries but can persist in fish following transfer to marine farms (Tobback et al., 2007; Chapela et al., 2018; Lewin et al., 2020). Collectively, these four bacteria cause disease responsible for multi-billion dollar losses globally (Assefa and Abunna, 2018); therefore, cost-effective diagnostics tools are required to enable early detection and appropriate management responses to outbreaks (Brosnahan, 2020).

The aim of this study was to design and validate a novel ddPCR tetraplex assay for priority salmon diseases for use in commercial applications. The exemplar species used for the assay included New Zealand Rickettsia-like organism 1 (NZ-RLO1), NZ-RLO2, T. maritimum, and Y. ruckeri. Following validation, the tetraplex assay was applied as part of a commercial fish disease surveillance program in New Zealand for 1 year to demonstrate the applicability of tetraplex tools for the salmonid aquaculture industry.

The ddPCR tetraplex assay developed in this study provides a rapid, cost-effective, and reliable screening tool for four primary aquaculture pathogens with commercial relevance and application. Other studies have proven multiple pathogens to be screened in single real-time PCR assays and their implementation as a cost-efficient disease monitoring tool (Chapela et al., 2018; Peters et al., 2018).

For this molecular diagnostic tool to become an accredited test, it needs to undergo appropriate validation (Thalinger et al., 2020). The primers and probes used in this study have been previously published and checked for sensitivity and specificity (see Table 1 and references therein) and were then optimized to work in combination with a ddPCR instrument. While the approach with different probe concentrations (i.e., the assay having 100% FAM for target 1, 100% HEX for target 2, a mix of 70% FAM and 30% HEX for target 3, and a mix of 70% HEX and 30% FAM for target 4) revealed good separation between the target organisms (Supplementary Figure 1), we found an amplitude assay was easier for pipetting purposes and clearer result output especially when dealing with unclean signals (“rain”) from a degraded sample. In vitro testing on reference tissue was achieved by successfully amplifying reference cultures of the pathogens individually and in combination. Additionally, gBlocks™ were spiked into uncontaminated, presumably healthy, fish tissue DNA to check for any inhibitory effects. No inhibition was detected. This test was applied commercially for salmon farm surveillance for 12 months in conjunction with additional validation via bacterial culture techniques and standard and qPCR confirmatory analysis by MPI. Successful and true positive detections were achieved using our tetraplex assay on field samples that also aligned with other studies, e.g., T. maritimum detections in Bateman et al. (2021).

Diagnostic tools can never achieve 100% sensitivity and specificity (Assefa and Abunna, 2018). For example, cultivation-based methods will only detect bacteria that are able to replicate under the provided conditions and most aquatic microorganisms (>99%) are unable to be cultivated using standard methods (Netzer et al., 2021). NZ-RLO1 and NZ-RLO2 are fastidious, and good-quality DNA extracts from cultures could not be achieved for molecular purposes; thus, artificially constructed DNA oligos (gBlocks™) were used for assay optimization. By combining these screening technologies, cross-validating outcomes and interpreting them in the context of their designed application, developed tools need to be continuously optimized.

During this study, three important observations gave us novel insights for New Zealand aquaculture surveillance. First, the T. maritimum specific primers and probe used in this study revealed low positive signal for a non-typical T. maritimum-like bacterial colony on agar plates, which was then confirmed as T. dicentrarchi through 16S rRNA gene sequencing. Fringuelli et al. (2012) did not include this strain for specificity testing as it is simply not possible to validate primers on all existing bacterial strains (Collins et al., 2006). We tested the assay on several T. dicentrarchi strains and received three positive signals out of 19 tested strains. Further strain confirmation by sequencing and intraspecific variability are needed to confirm cross-reactivity or wrongly identified strains/cross-contamination with T. maritimum.

Our second observation of cross-reactivity of Y. ruckeri with closely related Serratia species (both Yersiniaceae) was also not included as part of the specificity testing by Carson and Wilson (2009) and Ghosh et al. (2016). We discovered through a standard GenBank BLAST that the Y. ruckeri forward primer might have been developed on an incorrectly deposited reference sequence (NR_119063.1) that included a single base-pair mistake at the 5′-end of the sequence, dating back to 1993 when sequencing technology was in its beginnings and prone to erroneous nucleotide outputs. This theory is supported by the fact that none of the recent deposited sequences for Y. ruckeri identified that particular nucleotide which is located at the 3′-end of our forward primer. DNA polymerase requires the 3′ base of a primer to form appropriate hydrogen bonds to initiate polymerization and might experience loss in sensitivity through nucleotide ambiguities (van Pelt-Verkuil et al., 2008).

Finally, a high number of positive detections of NZ-RLO 1 following inactivated (DNA) vaccination on the fish farm revealed that testing DNA with the ddPCR specific assays would not be able to differentiate between a vaccinated fish or a true infection, reported previously (Laurin et al., 2020). Investigating RNA or other viability tests such as PEMAX that focus on live pathogens could help circumvent this problem at least for inactivated vaccines (Brosnahan et al., 2020).

Optimizing the tetraplex assay further will involve designing more specific primers to avoid cross-reactions between strains and correcting other primer issues, including the design of new primers for emerging agents, such as T. dicentrarchi and Serratia strains. Novel ddPCR technologies such as the QX600 AutoDG Flex from Bio-Rad are already on the market that work on four optical channels for multiplexing up to eight samples which could be used to test eight pathogens simultaneously. Overall, there is substantial interest in adopting approaches that allow for point-of-need and in-field surveillance (Peters et al., 2018). Future technologies such as the portable Oxford Nanopore Sequencing or the NS2 Nucleic Sensing System¹ which allows for in-field (water) multi-probe ddPCR analyses are promising for these purposes.

The baseline ddPCR tetraplex assay presented here can be repurposed and adjusted to advanced technologies, including new pathogens of interest or analytical methods, which will foster scientific research output and expand the molecular surveillance toolbox. For example, applying occupancy and co-occurrence modeling approaches on the detection signals of different pathogens enables detection probabilities even if all samples return negative (Willoughby et al., 2016; Laurin et al., 2020) and provides a cost-efficient tool for aquaculture surveillance (Farrell et al., 2021).

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

All fish sampling was done with approval from the Nelson Marlborough Institute of Technology Animal Ethics Committee (AEC 2018 CAW01). Written informed consent was obtained from the owners for the participation of their animals in this study.

UA, JS, and XP designed the research. UA, TA, and KK performed the research. UA, CB, JS, and KH contributed the reagents and analytical tools. UA analyzed the data. UA, JS, KH, and XP wrote the paper. All authors contributed to the article and approved the submitted version.

This work was funded by the New Zealand Ministry of Business, Innovation and Employment, under the program: Aquaculture Health to Maximise Productivity and Security (CAWX1707).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.