Development of Improved Diagnostics for Renibacterium Salmoninarum Causative Agent of Bacterial Kidney Disease in Atlantic Salmon

RPC Report Number:  FFA-J3077-1

Client: 
Cooke Aquaculture Inc.
Dr. Leighanne Hawkins
874 Main Street
St. George, NB  E5H 1E6

Submitted By:
Dr. Benjamin Forward
Head, Food, Fisheries and Aquaculture Department
RPC
921 College Hill Road
Fredericton, NB  E3B 6Z9
Tel: 506.452-1365
Fax: 506.452.1395
EMAIL: ben.forward@rpc.ca

August 12, 2009

Table of contents

• EXECUTIVE SUMMARY
• INTRODUCTION
• MATERIALS AND METHODS
• RESULTS AND DISCUSSION
• CONCLUSIONS
• REFERENCES

AUTHORIZATION

EXECUTIVE SUMMARY

Bacterial Kidney Disease (BKD) is a disease of salmon caused by Renibacterium salmoninarum, a gram positive, obligate intracellular pathogen. The disease is slow to develop with outward signs being variable and it is responsible for considerable economic losses to the Aquaculture industry in Eastern Canada. Control of the pathogen is best achieved by vigilant surveillance and testing of fish using IFAT, bacterial culture and the polymerase chain reaction (PCR) followed by appropriate management of infected fish. However disparity between diagnostic assays and among laboratories has made it clear that improved methods (higher through-put, more specificity and sensitivity) are needed to detect the pathogen and to manage the disease. The objective of this project was to develop improved diagnostics for R. salmoninarum through the development and validation of best sampling practices and diagnostic assays. 

Towards this end a new qPCR assay specific for R. salmoninarum was developed and tested across multiple sample types. The new qPCR assay, IFAT, and bacteriological culture were employed on a large number of samples and sample types to determine best diagnostic assays and sampling practices. However, the ability to draw a clear conclusion as to best practices was complicated due to an extremely low prevalence of R. salmoninarum in the sampled organisms. All samples were negative when tested with IFAT and culture. qPCR detected R. salmoninarum in only two samples, one from a salmon parr kidney and the other from a broodstock egg sample. Each of these samples originated in fish from the facility with a history of BKD. The results indicate that the prevalence of this disease agent is very low in the year class of fish sampled and suggest that qPCR is the most sensitive method for detection of the pathogen. However, the low prevalence of positive samples precluded statistical validation of test suitability or sampling methodology. 

INTRODUCTION

Bacterial Kidney Disease (BKD) infections in net pen raised Atlantic salmon in the Bay of Fundy have become more prevalent in recent years. Indeed, in 2005 economic losses from the disease surpassed those caused by Infectious Salmon Anaemia. BKD is caused by the obligate intracellular pathogen Renibacterium salmoninarum and is manifested as a chronic systemic infection in salmonid fish characterized by granulomatous lesions in the kidney and other organs. The disease is slow to develop with outward signs being variable and occurring primarily near the terminal stages. Control can be problematical since the pathogen can be transmitted both horizontally in water and vertically through infected gametes. Treatment has been complicated by the fact that the organism can survive intracellularly, ‘hiding’ from chemotheraputants and the immune system of the fish and management of the disease has been through vigilant surveillance of moribunds using IFAT, bacterial culture and the polymerase chain reaction (PCR). Indeed, the intracellular and gram positive nature of the pathogen complicates diagnostics and much effort has gone into development and study of diagnostic methods for R. salmoninarum (see for example, Griffiths et al., 1991, Pasho et al., 1987, Miriam et al.,1997, Jansson et al., 1996). Culture is considered the gold standard, but results can take weeks (as many as 4-12 weeks) to appear and fish may need to be moved in the interim (Benediktsdottir et al., 1991). If the test result is positive, the pathogen may have had chance to spread to new fish. IFAT and PCR are relatively quick, but results from two tests are not always consistent within a lab or between laboratories (see the recent study by Bruno et al. 2007). In addition duplicate samples sent to different labs for the same diagnostic labs can on occasion give different results, and within a lab results from IFAT and PCR can give different results. This is in part due to the intracellular nature of the pathogen, but may also be due to differences in sample type (different portions of the kidney may contain different levels of pathogen) and sample storage prior to processing (the pathogen may degrade if in sub-optimum storage). Anecdotally, the inconsistencies may give rise to false negatives hindering optimum fish health management: apparently negative fish in the hatchery often succumb to BKD shortly after transfer to salt water suggesting that the diagnostics are not detecting R. salmoninarum as well as they should to allow effective containment of the pathogen, or management of the disease.

Clearly, the industry is in need of improved diagnostic methods for detection of R. salmoninarum. The methods should be developed taking into account best sampling practices for eachtype of sample (location of sample, storage and treatment of sample prior to analysis).Diagnostic methods should be sensitive and specific for R. salmoninarum and their limits should bewell understood. Methods for high throughput detection of BKD should also beinvestigated to allow for rapid screening of ovarian fluids or other samples duringspawning. The use of improved diagnostics for R. salmoninarum in a surveillance program willreduce the incidence of BKD in the industry, resulting in improved fish heath, greaterproductivity and economic success.

BKD is now the single greatest fish health challenge to the culture of Atlantic salmon in Eastern Canada. In the absence of an effective vaccine, surveillance for the causative agent, Renibacterium salmonicida, followed by appropriate management is the only feasible way of controlling the spread of disease. However, current surveillance programs (sampling and diagnostics) are not always effective at detecting R. salmoninarum and thus management of the disease is not as successful as it should be. Diagnostics using culture, while considered the ‘gold standard’ is slow to produce results--results are rarely available before 6 weeks. While the faster methods of Indirect Fluorescent Antibody Technique (IFAT) and (quantitative) Polymerase Chain Reaction ((q)PCR) which detect R. salmoninarum antigens or DNA respectively do not always produce consistent results within or between laboratories. This may be due to differences in sampling, sample storage, or tissue heterogeneity as much as the tests themselves, but whatever the reason there is a clear need for improved diagnostic methods for the detection of R. salmoninarum. This project will address differences in sampling, sample storage and diagnostic methods used in surveillance for R. salmoninarum with a view to identifying the best practices for diagnostic surveillance of R. salmoninarum in the hatchery, in the marine environment and in broodstock. 

MATERIALS & METHODS

Fish Sampling
Salmon were collected from two different fresh water hatcheries prior to transfer to sea water, and then again after transfer to marine sites. The two hatcheries included one with a history of BKD infections (Fresh water 1 (FW1); Elmsville) and one with no history of disease (Fresh water 2 (FW2); South Oromocto Lake). Fish sampling for FW1 occurred 9 weeks prior to seawater transfer and again 11 weeks post transfer to a site in Beaver Harbour (Salt water 2 (SW2)). Fish sampling for FW2 occurred ~8 weeks prior to transfer and again ~13 weeks post transfer to another site in Beaver Harbour (Salt water 1 (SW1)). Moribund/suspect fish were sampled where possible. In total, one hundred fish were sampled from each hatchery & marine site (400 fish in the case of two populations). Fish were collected in bags placed on ice and transferred to the laboratory for necropsy and tissue sampling.

To compare diagnostic tests in brood fish, tissue samples from 60 broodstock fish were collected at the time of spawning from the Elmsville hatchery.

Tissue Collection & DNA Extraction
For each smolt, the head kidney was collected and divided into two. One half was stored in RNAlater at -20°C and the other was homogenized in 1X sterile hanks solution and plated on SKDM media then subsequently stored at -20°C. A section of anterior kidney was collected and imprinted onto duplicate prewashed slides for IFAT analysis. A sterile swab from along the entire length of the anterior kidney was also collected and smeared onto duplicate prewashed slides for IFAT analysis.

For broodstock fish, eggs, kidney, blood, and ovarian fluid from 60 fish were collected at the hatchery and placed on ice for transport to the laboratory for processing. Blood was collected in 5ml heparinized vacutainers, ~15ml of eggs were collected in sterile Whirl-Pak bags, ovarian fluid was collected in 5ml non-heparinized vacutainers, and kidney was collected in sterile Whirl-Pac bags. Once at the lab, 10µl of each blood sample was mixed with 190µl sterile Phosphate buffered saline (PBS) and frozen at -20°C. Eggs were homogenized by crushing and 50µl of the homogenate was mixed with 150 µl of PBS and frozen at -20°C. Fifty microliters of ovarian fluid was mixed with 150µl PBS and frozen at -20°C for further analysis. At the same time, aliquots of each tissue type were also smeared onto duplicate prewashed slides for IFAT analysis. The remainder of all sample types was then archived at -20°C. 

For DNA extraction of blood, the DNeasy Blood and tissue kit (Qiagen) was used following the blood protocol. The same kit was also used for extraction of egg, ovarian fluid, and kidney samples (50 mg) except that the Animal Tissue Protocol was followed as per the manufacturer’s instructions with the following modification. Following an overnight tissue lysis step, an enzymatic lysis step (with lysozyme) was performed according to the recipe provided in kit instructions (Appendix E) for 30 minutes at 37°C. Two hundred microliters of buffer AL was then added and the sample incubated at 70°C for 30 min. The remainder of the protocol followed from step 4 of the Animal Tissue Protocol in the manufacturer’s instructions. 

Real-time qPCR
The following method was based on that describe by{Bruno, 2007 #239} using the same primers and probes but incorporating both MSA2 and Elf1a primers and probes into a single multiplex reaction. Further, the assay was also designed to be quantitative by incorporating a standard curve which utilized a DNA plasmid containing the MAS2 gene in known amounts. Each reaction consisted of the following except where indicated otherwise: 2.5 µl of dH2O, 12.5 µl 2X TaqMan Universal PCR Master mix (Applied Biosystems), 1 µl Msa2F primer (22.5 µM): 5’-GGA GCA ACT CCG GTT ACT GGT A-3’, 1 µl Msa2R primer (22.5 µM): 5’-TGG CCG TCC TTG AAC CAT-3’, 1 µl Msa2 probe (6.25 µM): 6FAM - TGG TCT GGC GAC AAC AAC ACG TAT GGT – MGBNFQ, 1 µl Elf1aF primer (2 µM): 5’-GGC CAG ATC TCC CAG GGC TAT-3’, 1 µl Elf1aR primer (2 µM): 5’-TGA ACT TGC AGG CGA TGT GA-3’, 1 µl Elf1a probe (6.25 µM): VIC - CCT GTG CTG GAT TGC CAT ACT G – MGBNFQ, and 5 µl of DNA.  Cycling conditions included one cycle of 2 min at 50°C, one cycle of 10 min at 95°C followed by 45 cycles of 15 sec at 95°C and 1 min at 60°C. All reactions were conducted in duplicate on a ABI 7000 Signal Detection System using the Sequence Detection Software (V1.2.3; Applied Biosystems). 

IFAT
Head kidney homogenate, blood, ovarian fluid, anterior kidney, and a swab of the anterior kidney was collected and smeared or imprinted on a cleaned glass slide for IFAT analysis according to the method described in the Canadian Fish Health, Manual of Compliance.  A minimum of 50 fields from each sample was read at 1000X magnification.

Bacteriological Plating
A sample of the head kidney was homogenized with 1X Hanks and plated on Selective Kidney Disease Medium (SKDM) obtained from the Atlantic Veterinary College, PE. The samples were incubated at 14°C for eight weeks and examined for signs of bacterial growth. Any growth present on plates was verified by IFAT.

RESULTS AND DISCUSSION

The objective of this project was to improve diagnostics for R. salmoninarum by identifying best practices for detection of R. salmoninarum in hatcheries, marine sites and broodstock. To achieve this objective, samples of kidney from fish at two freshwater sites were tested using 3 different methodologies (IFAT, culture and qPCR) both before and after transfer to seawater. Also, multiple tissues from broodstock held at one of the freshwater sites were examined using IFAT and qPCR to determine the best tissue and method for detection. 

The results of analyses detected very few positive carriers of R. salmoninarum in any of the fish populations tested thereby precluding the ability to draw firm statistical based conclusions as to best practices and methods for detection. All culture and IFAT tests were negative for R. salmoninarum in all samples and qPCR detected only two positive samples – an egg sample from Fish 18 of the broodstock and a head kidney homogenate sample from fish 59 of the Elmsville freshwater hatchery (see Appendix A-J). In all qPCR assays, each sample was examined in duplicate and was only scored positive when both replicates gave a positive signal. The amount in egg sample 18 equated to approximately 10 R. salmoninarum cells per 50 µl of egg homogenate. In the positive kidney sample, 59, the amount of R. salmoninarum cells equated to approximately 2 per 50 mg of head kidney tissue. The RNALater counterpart of the FW1 samples were misplaced thereby precluding testing to determine if RNAlater storage had any influence on detection. As a result, a larger number of FW1 homogenate samples were chosen to be tested since all others had tested negative by other methods. As the proposal called for a total of 50 fish to be tested, it is possible that a larger selection would have yielded more positive fish. However, given that all other detection methods yielded no positive results, any positives detect by qPCR would likely have been infrequent. 

Unfortunately IFAT analysis in egg samples was not possible as stained slides containing egg homogenate produced very high background levels. Any positive staining resulting from R. salmoninarum was not visible against this high background. While not planned in the original proposal, bacteriological culture was performed on brood kidney and was also negative for all broodfish samples. 

CONCLUSIONS

In summary, the low prevalence of R. salmoninarum detected in the samples precluded a detailed analysis to firmly determine the best sampling methodology and detection method. Unfortunately this is the type of risk which one undertakes when doing field studies such as this, and it should remain a focus for future studies. Where possible, multiple tissue samples should be collected and analysed retrospectively in cases in which R. salmoninarum is detected. 

While the qPCR method did detect two positive samples which were not detected by the other two methods, a higher number of positive samples would have been preferable to more confidently conclude the method is more sensitive to other diagnostic methods. As the qPCR method is quantitative, information regarding the levels of R. salmoninarum cells in the samples was determined. The levels detected were very low and approaching the lower detection limits of the assay. Such low levels, together with the lack of detection of other methods, do suggest this method to be more sensitive and is in agreement with previously published research.

As of August 10, 2009, all fish transferred to seawater still have shown no signs of BKD (Dr. Leighanne Hawkins, personal communication). This result is in general agreement with the low detection in the samples analysed and suggests that the prevalence of positive fish in fresh water does not necessarily pose a significant risk for the development of disease once transferred to sea cages. However, screening a larger number of fish using qPCR may yield more insight into the prevalence of carrier fish in the hatchery and assessment for the potential of disease development upon transfer to sea. 

Analysis of broodstock fish suggests that IFAT of egg samples is not be best method to identify the presence of R. salmoninarum. As bacterial culture takes such a long time it is also not an idea method for screening of brood fish. Although further work is needed to support these preliminary findings, our work here suggests qPCR is the most sensitive and rapid method for detection of R. salmoninarum in broodstock, and as such we recommend qPCR screening of broodstock fish, and maintaining separation of fertilized egg batches until diagnostic results are known.

REFERENCES

Benediktsdottir E., Helgason S., and Gundmundsdottir S (1991) Incubation time for the cultivation of Renibacterium salmoninarum from Atlantic salmon, Salmo salar L., broodfish J. Fish Diseases 14: 97-102

Bruno D., Collet B., Turnbull A., Kilburn R., Walker A., Pendrey D., McIntosh A., Urquhart K., Taylor G., (2007) Evaluation and development of diagnostic method for Renibacterium salmoninarum causing bacterial kidney disease (BKD) in the UK Aquaculture 269:114-122

Griffiths SG., Oliver G., Fildes J., Lynch WH., (1991) Comparison of western blot, direct fluorescent antibody and drop-plate culture methods for the detection of Renibacterium salmoninarum in Atlantic salmon (Salmo salar L). Aquaculture 97: 117-129

Jansson K., Hongslo T., Hoglund J., Ljungberg O., (1996) Comparative evaluation of bacterial culture and two ELISA techniques for the detection of Renibacterium salmoninarum antigens in salmonid kidney tissues Dis. Aquat Org 27: 197-206

Miriam A., Griffiths SG., Lovely JE., Lynch WH (1997) PCR and probe-PCR assays to monitor broodstock Atlantic salmon (Salmo salar L.,) ovarian fluid and kidney tissue for the presence of DNA of the fish pathogen Renibacterium salmoninarum J. Clin. Microbiol 35:1322-1326

Pasho RJ, Elliott DG., Mallett RW., Mulcahy D., (1987) Comparison of five techniques for the detection of Renibacterium salmoninarum in coho salmon. Tran Am Fish Soc 11:882-890

Zou J., Carrington A., Collet B., Dijkstra JM., Bols N., Secombes CJ., (2006) Identification and bioactives of interferon gamma in rainbow trout Onchohynchus mykiss: the first Th 1 type cytokine characterized functionally in fish. J. Immunol 175: 2484-2494