Evaluation of Rainbow Trout as a Model for use in Studies on Pathogenesis of the Branchial Microsporidian salmonae

DAVID J. SPEARE, DVM, DVSC, GARTH J. ARSENAULT, ENV TECH, AND MELANIE A. BUOTE, BS

Abstract _ Loma salmonae is an economically important gill microsporidian pathogen of pen-reared chinook and coho salmon. Chinook and coho salmon are generally poorly suited for use in laboratory studies because of their high mortality rates when infected with L. salmonae and their high-level of susceptibility to other infectious diseases. Using gill tissue from chinook salmon that contained mature xenomas laden with L. salmonae spores, we successfully transmitted the infection to rainbow trout. The infection developed in an identical manner and over a similar time course in trout as for chinook salmon. In contrast, we were unable to transmit the infection to other candidate salmonid species, including Atlantic salmon, brook trout, or arctic charr. Gill tissue from experimentally infected rainbow trout was then used to successfully transmit the parasite to other trout. Horizontal transmission was documented from infected to naive tankmates. Analysis of these results indicated that L. salmonae can have a complete life cycle in trout and produce viable spores. Although abundant xenomas developed in the gills of infected trout, the fish did not have clinical signs and there were no fatalities. We concluded that use of rainbow trout offers several key advantages for study of the pathobiologic characteristics of L. salmonae.

Gill disease associated with the microsporidian parasite Loma mission of the organism, gill material from infected chinook salmonae is an emerging problem affecting the marine aquacul- salmon was obtained from another research worker (Dr. M. L. ture production of Pacific salmon, most notably chinook salmon Kent, Department of Fisheries and Oceans, Pacific Biological ( tshawytscha) and, to a lesser extent, coho salmon Station, Nanaimo, British Columbia, Canada ). Gill arches from (O. kisutch) (1–6). Parasites develop in the vasculature of the naturally or experimentally infected fish were removed, placed gills to form large spore-laden xenomas (2–6). Loma salmonae in Earl’s balanced salt solution, and transported on ice. Material also infect rainbow trout (O. mykiss) (7–9); however, in contrast was kept chilled and used 3 days after harvest. to the Pacific salmon, the clinical effects on these fish vary con- In subsequent studies, spores were obtained from gills of in- siderably (9). Furthermore, on the basis of unpublished fected rainbow trout. Spores were obtained 7 to 8 weeks after observations from diagnostic materials sent to the Fish Health fish were exposed to spores during the previous (initial) study. Unit of the Atlantic Veterinary College, L. salmonae is often de- Fish were housed in water at a temperature of 14.58C. Fish with tected in groups of trout that do not have a history of clinical abundant xenomas were selected, euthanatized with an overdose disease, whereas this is not the clinical history for chinook salmon. of benzocaine dissolved in water (100 mg of benzocaine/L of Because of the restrictive nature of current regulations per- water), gill arches were removed, and tissues were used 1 to 3 h taining to the use of therapeutic agents in aquaculture, licenced after collection. therapeutic agents are not available for treating fish infected Fish: All procedures involving live fish were performed accord- with L. salmonae. Accordingly, there is a need to study the ing to guidelines of the Canadian Council on Animal Care. pathobiologic characteristics of L. salmonae in an effort to un- Specific protocols were reviewed by the institutional animal care cover management factors that could limit the economic effects committee (University of Prince Edward Island, Charlottetown, of infection. Laboratory studies of this disease would benefit by Prince Edward Island, Canada). the availability of a species other than chinook salmon, even Juvenile rainbow trout (mean weight, 4 g) were obtained from though they are the principal aquaculture species affected by L. a certified disease-free (specific pathogens) commercial hatch- salmonae. Chinook salmon have proven to be difficult to work ery on Prince Edward Island which did not have a history of L. with in laboratory studies (10) and often become diseased or salmonae infection. Fifteen fish were arbitrarily selected from die because of a range of other endemic infectious diseases, par- 1,000 received and were examined histologically to rule out L. ticularly when the fish are stressed. Rainbow trout, in contrast, salmonae or other diseases not included in federal certification are a relatively hardy fish that are frequently used in laboratory procedures. During the studies, fish weighed from 4 to > 147 g. studies (11). They have been the salmonid species used as a Variations in weight were due to growth, because the various model for many studies, and their requirements and physiologic studies were conducted sequentially. One hundred 3-g Atlantic responses are well documented (11). salmon (Salmo salar) and 3-g brook trout (Salvelinus fontinalis) Our purpose was to evaluate the suitability of rainbow trout as were also acquired from the hatchery where we purchased the an animal model for L. salmonae infections with particular em- trout. One hundred 6-g Arctic charr (Salvelinus alpinus) were phasis on the ability of this parasite to complete its life cycle in a obtained from another commercial hatchery, also with certified manner similar to that seen in naturally and experimentally in- disease-free status. Samples of these other species were exam- fected chinook salmon. ined 4 weeks after their arrival to detect any evidence of L. Materials and Methods salmonae or L. fontinalis. Methods of infection: Fish were exposed to infected gill mate- Source of Loma salmonae spores: In the initial studies on trans- rial by two ways. A feeding method was devised on the basis of Department of Pathology & Microbiology, Atlantic Veterinary College, University previously published methods (6). Approximately 2 g of gill tis- of Prince Edward Island, Charlottetown, Prince Edward Island, Canada C1A 4P3 sue was minced into pieces approximately 0.2 cm2. This tissue

55 was then fed to 20 to 30 fish in a tank. An intubation method was native salmonid species (two non-Oncorhynchus genera). Using devised on the basis of information gained during personal com- spores derived from infected trout, we intubated 20 Atlantic munications (12). Filament cartilage was dissected from salmon, 20 brook trout, and 20 arctic charr. Twenty naive rain- harvested gills, which were then manually crushed. Fish were bow trout were also intubated and used as control fish. Infection manually restrained or lightly anesthetized in a solution of ben- status of fish was monitored weekly through 9 weeks. At week 7, zocaine (40 mg of benzocaine/L of water), and 0.1 to 0.2 ml of samples were collected for histologic assessment to determine the gill tissue-diluent was administered, using gastric intubation. whether xenomas or other evidence of disease were evident in Water supply, housing and characteristics: On arrival, fish were nonbranchial tissues in these test species. During this study, fish stocked in 70-L (salmon, brook trout, arctic charr) or 500-L fi- were housed for 14 weeks in a single 70-L tank supplied with berglass tanks (trout) that received water pumped from a well at freshwater at a temperature of 14.58C. 108C. Fish were fed a recommended amount (13) of a commer- Study 4. To assess whether the final number of xenomas that cially available diet until they entered a study. All studies were developed on the gills of infected fish was correlated with the conducted in 70-L circular tanks supplied with freshwater (2.7 infective dose received, we infected 75 rainbow trout via intuba- L/ min). Dissolved oxygen concentrations were maintained be- tion with 3 dilutions (1:1, 1:10, 1:100) of a preparation of spores. tween 8 and 11 mg/L. For studies conducted with the water A fourth group of 75 fish was intubated with water only. The 1:1 temperature of 14.58C, water from the well was heated and mixed dilution yielded a dose of approximately 4.8 X 106 spores/fish, in a header tank with ambient-temperature water until 14.58C determined on the basis of calculations derived from enumerat- was reached. To avoid supersaturation problems, heated water ing the number of xenomas used and a volumetric calculation was degassed through a trickle column before use. of the reported theoretical maximum number of spores per Evaluating infection status: For studies conducted by using (8). Seven weeks after intubation, fish were euthana- water at a temperature of 14.58C, the infection status of fish was tized by immersion in an overdose of benzocaine. The second assessed two or more times, usually during the period 5 to 7 gill arch on the left side of each fish was removed, and number weeks after initial exposure to infective material. For studies con- of xenomas was counted, using a compound stereomicroscope. ducted by using water at colder temperatures or when species Number of gill filaments on the arch was also counted. Data other than rainbow trout were used, more-frequent screening were reported as number of xenomas per filament to compen- during a longer period was used. During each assessment, fish sate for any differences in gill surface area that may have resulted were anesthetized in benzocaine (60 mg of benzocaine/L of from potential differential growth rates between groups of fish water). The operculum was lifted, and the gills were examined, exposed to differing doses. using a dissecting microscope. The entire surface of the second Effect of dilution on number of xenomas formed was assessed gill arch of each fish was examined for evidence of xenomas. through a one-way ANOVA with a Tukey-Kramer multiple com- Fish were then placed in water that did not contain benzocaine. parison test. Association between dose and outcome was assessed Fish were allowed to recover fully, and then were returned to with linear regression analysis and assessment of the correla- their tanks. During several of the initial studies, some fish were tion coefficient. euthanatized with an overdose of benzocaine, and gill tissue was harvested and assessed histologically. Results Transmission studies: Study 1: At a water temperature of 14.58C, 18/20 (90%) and 19 Study 1. Susceptibility of trout (mean weight, 7.2 g) to L. of 20 (95%) rainbow trout exposed by intubation or feeding, re- salmonae spores harvested from chinook salmon was determined spectively, of spores from chinook salmon developed typical o at two water temperatures: 10.0 or 14.5 C. For each tempera- xenomas by week 5 (Table 1). A decreased proportion of fish with ture, 40 fish were used: 20 were fed minced gill tissue and 20 xenomas at week 8 indicated that xenomas were rupturing by that were intubated. The objective was to determine whether tem- time. Histopathologic features of samples collected at weeks 6 and perature of water, in this range, had an effect on the ability of L. 10 were consistent with those described for naturally or experi- salmonae to develop to the point of xenoma production. mentally induced infections of chinook and coho salmon (4–6). Study 2. Two experiments were conducted, using water at a At week 6, these features included numerous xenomas within the temperature of 14.58C, to assess whether the life cycle of L. lamellae (Figure 1) and central venous sinusoid (Figure 2) of the salmonae could be completed in rainbow trout. Viability of spores filaments. Fewer xenomas were in filament arterioles, and only collected from trout that were infected in study 1 was determined rarely was a xenoma found in the heart or spleen. A pattern of through their ability to infect naive rainbow trout. Two tanks, multifocal granulomatous proliferative branchitis accompanied by each containing 35 rainbow trout (mean weight, 35 g), and two fibrinoid arteritis was detected in samples collected at week 10, tanks, each containing 12 rainbow trout (mean weight, 114 g), two weeks after xenomas had begun to rupture (Figure 3). During were used. One tank of each size of trout was fed gill material, whereas trout in the other tanks were intubated. Each fish was Table 1. Ratio of trout that developed xenomas after being fed or assessed 0, 5, and 7 weeks after exposure. After the final evalua- intubated with Loma salmonae-infected chinook salmon gills and tion with the dissecting microscope, gills and visceral organs were housed at two temperatures. collected from 3 confirmed-positive infected fish of each tank for histologic examination. Gills and visceral organs were fixed weeks 108C 14.58C after Intubated Fed Intubated Fed in buffered 10% formalin, routinely processed for histologic exposure examination, and stained with hematoxylin and eosin. During another experiment, we determined whether there 2 N/A N/A 0/20 0/20 would be natural horizontal transmission between infected and 3 0/20 0/20 0/20 0/20 naive rainbow trout. We monitored weekly the infection status 5 0/2 * 0/2 * 18/20 19/20 of 60 naive rainbow trout housed as a cohort control group in 5.5 0/18 0/18 6/6 * 6/6* 8 0/6 * 0/6 * 7/14 6/14 tanks with 60 rainbow trout that had been infected via intuba- 10 0/12 0/12 1/6 * 0/6 * tion with L. salmonae. 12 0/11 0/12 0/8 0/8 Study 3. Our objective in study 3 was to provide additional in- formation about the host range (permissive for the formation of *Determined on the basis of histologic assessment. Values represent No. xenomas) of L. salmonae by examining the susceptibility of alter- of fish infected/No. of fish exposed. N/A = not assessed.

56 weeks 5 and 8. Additionally, xenomas were not detected in vis- ceral organs, and visceral tissues did not have lesions suggestive of disease processes. Study 2: We documented that spores collected from infected rainbow trout of study 1 were capable of inducing xenoma for- mation by week 5 in two size groups of trout (mean weight of 35 and 114 g), with 59 of 70 (84%) 35-g fish and 21 of 24 (88%) 114-g fish becoming positive by week 7 (Table 2). Susceptibility did not differ between the two sizes of fish. In the subsequent experiment of study 2 all trout exposed to spores had xenomas by week 6 and were free of xenomas by week 11, whereas unex- posed cohabitant trout were free of xenomas until week 12, when 31/57 (54%) of them were positive (Table 3). Study 3: We did not detect evidence of L. salmonae infection in Atlantic salmon, brook trout, or arctic charr. In contrast, all FIG. 1. Section of gill filament from a trout 6 weeks after exposure to of the rainbow trout developed xenomas by week 5 (Table 4). Loma salmonae. A mature xenoma, containing spores of L. salmonae, is evident within a pillar cell of the proximal channel of a gill lamella. Histologic examination of the three negative species failed to H&E stain; bar = 40 mm. reveal xenomas in visceral organs. Study 4: Number of xenomas that formed on the gills of intu- bated trout was strongly associated with spore dose (Table 5). Linear regression analysis yielded a correlation coefficient of 0.35. The 314.9% increase in xenoma count between the 1:10 and 1:100 dose dilutions was significant (P < 0.01). However, the 13.8% increase between the 1:1 and 1:10 dose dilutions was not significant. Xenomas did not form in the control group.

Discussion Analysis of results of our study revealed that rainbow trout are a useful replacement model for future studies on the pathobiologic characteristics of L. salmonae infections. The in- fectivity trials documented that rainbow trout (genus Oncorhynchus), in contrast to other commonly used salmonid species in the genera Salmo and Salvelinus, was permissive and

Table 2. Ratio for 2 size classes of trout that developed xenomas after FIG. 2. Section of gill filament from a trout 6 weeks after exposure to being intubated or fed Loma salmonae infected trout gills and housed Loma salmonae. A mature xenoma has developed within an endothelial at a water temperature of 14.58C. cell of the central venous sinusoid of the gill filament. H&E stain; bar = 40 mm. weeks 35-g fish 114-g fish after Intubated Fed Intubated Fed exposure 0 0/35 0/35 0/12 0/12 5 11/35 6/35 7/12 5/12 7 32/35 27/35 11/12 10/12 Values represent No. of fish infected/No. of fish exposed

Table 3. Ratio of trout that developed xenomas after being exposed to Loma salmonae-infected trout gills by gastric intubation or by cohabitating with intubated fish weeks after Fish exposed Naive cohabitating initiating via intubation fish study 0 0/60 0/60 1 0/59 0/60 FIG. 3. Section of a gill from a trout 10 weeks after exposure to Loma 2 0/59 0/60 salmonae. Xenomas have ruptured and a multifocal granulomatous branchitis, affecting the central venous sinusoid of gill filaments, has 4 0/59 0/60 developed. H&E stain; bar = 200 mm 5 0/58 0/60 6 56/56 0/60 the course of the infection, fish did not have abnormal behavior 7 56/56 0/59 or clinical manifestations of disease. 8 34/56 0/59 At a water temperature of 108C, xenoma formation was not 9 17/49 0/58 evident, even by week 12. These fish were sporadically exam- 10 6/49 0/58 ined during the subsequent 2 months, and xenomas were not 11 0/49 0/58 detected. 12 0/49 31/57 Branchial lesions were not detected in samples collected at Values represent No. of infected fish/No. of fish exposed.

57 Table 4. Susceptibility of four species of salmonids to xenoma helial cells. If the sporoplasm were to undergo multiple rounds formation after exposure to Loma salmonae-infected trout gill tissue by of replication prior to reaching the final site in the gills, we would gastric intubation. not predict that dose dilution would be closely matched to weeks rainbow Atlantic brook arctic xenoma outcome, unlike when the 1:10 and 1:100 dilutions were after trout salmon trout charr compared. The relatively minor increase in xenoma count af- exposure forded by the 1:1 dilution suggested that we are reaching a maximum amount of disease expression, as would be seen if there 0 0/20 0/20 0/20 0/20 was competition for binding or entrance sites at some key point 5 20/20 0/20 0/20 0/12 7 N/A 0/20 0/20 0/6 during the parasite’s development cycle. 9 N/A 0/15 0/15 0/4 Analysis of results of our study also points to the regulatory role that water temperature has on L. salmonae, because we were un- Values represent No. of infected fish/No. of fish exposed. N/A - not able to infect fish housed in water at a temperature of 108C. Water assessed. temperature is a common promoting or limiting influence on fish pathogens, and has been used for control of the myxosporidian Table 5. Correlation between infective dose and number of xenomas agent responsible for proliferative kidney disease (14). Effects of per filament for rainbow trout exposed to Loma salmonae-infected water temperature have not yet been described for L. salmonae; trout gill tissue by gastric intubation analysis of our results suggested that manipulation of water tem- dose n Mean No. of SEM perature may be a means by which this parasite can be controlled. dilution xenomas/filament 1: 1 24 0.7405a 0.1546 Acknowledgments a 1: 10 25 0.659 0.137 This study was funded through the Natural Sciences and En- b 1: 100 25 0.209 0.064 gineering Research Council (Canada) Strategic Grants Program. 0:1 25 0.0c 0.0 a,b,cValues with different letters differ significantly (P < 0.05). References 1. Canning, E. U. and J. Lom. 1986. The of fish., readily prone to the parasite. This confirms several reports in p. 17–172, In Canning, E. U., and J. Lom (ed.), The which investigators documented existence of a gill microsporidia of vertebrates. Academic Press, Inc. Orlando, Fla. microsporidian with spore characteristics resembling those of 2. Hauck, A. K. 1984. A mortality and associated tissue reac- L. salmonae affecting hatchery-reared rainbow trout (7–9). Our tions of chinook salmon, Oncorhynchus tshawytscha inability to transmit L. salmonae to either of the Salvelinus spe- (Walbaum) caused by the microsporidian Loma sp. J. Fish cies of salmonids adds further evidence to the conclusions of Dis. 7:217–229. Morrison and Sprague (8), who indicated that the gill 3. Magor, B. G. 1987. First report of Loma sp. (Microsporidia) microsporidian pathogen, which has been described in brook in juvenile coho salmon (Oncorhynchus kisutch) from trout, is another species (Loma fontinalis). The host range of L. Vancouver Island, British Columbia. Can. J. Zool. 65:751– salmonae may be largely restricted to species within the genus 752. Oncorhynchus. We are not concluding that is impossible to infect 4. Kent, M. L, D. G. Elliot, J. M. Groff, et al. 1989. Loma other salmonid species with L. salmonae; however, it appears that salmonae (Protozoa: Microspora) infections in seawater they would be unreliable test species. reared coho salmon Oncorhynchus kisutch. Aquaculture Adding support for selecting rainbow trout as a replacement 80:211–222. model is the fact that we determined the parasite’s life cycle can 5. Speare, D. J., J. Brackett, and H. W. Ferguson. 1989. Se- be completed in trout and infective spores are produced. The quential pathology of the gills of coho salmon with a time frame from exposure to spores until xenomas were formed combined diatom and microsporidian gill infection. Can. was similar to that described for chinook salmon (6), with initial Vet. J. 30:571–575. formation evident by week 5 and rupture of xenomas by week 7 6. Kent, M. L., S. C. Dawe, and D. J. Speare. 1995. Transmis- or 8. Analysis of results of our cohabitation study added further sion of Loma salmonae (Microsporea) to chinook salmon in confirmation to the time-frame conclusions, because xenomas sea water. Can. Vet. J. 36:98–101. were first detected by week 12 in control fish, 5 weeks after we 7. Wales, J. H. and H. Wolf. 1955. Three protozoan diseases had the first evidence of xenoma rupture in their experimen- of trout in California. Calif. Fish Game. 41:183–187. tally infected tankmates. In general, pathobiologic characteristics 8. Morrison, C. M. and V. Sprague. 1983. Loma salmonae (Putz, of L. salmonae in trout and chinook salmon are similar in that Hoffman and Dunbar, 1965) in the rainbow trout, Salmo xenomas form almost exclusively in the gills (2,3,6), and bran- gairdneri Richardson, and L. fontinalis sp. nov. chial inflammation is delayed until the period of xenoma rupture (Microsporidia) in the brook trout Salvelinus fontinalis (5,6). In contrast to the severe clinical signs that are generally (Mitchell). J. Fish Dis. 6:345–353. seen in infected chinook salmon, it was remarkable that fish in 9. Markey, P. T., V. S. Blazer, M. S. Ewing, et al. 1994. Loma our studies appeared and behaved clinically normal throughout sp. in salmonids from eastern United States: associated le- the entire cycle of infection. sions in rainbow trout. J. Aquat. Anim. Health. 6:318–328. The character of the dose response trend may have implica- 10. Kent, M.L. 1995. Personal communication tions regarding the pathogenesis of this disease. Whereas the 11. Gall, G. A. E. and P. A. Crandell. 1992. The rainbow trout. difference in xenoma count between the 1:100 and 1:10 dilu- Aquaculture. 100:1–10. tions of the spore preparation were close to being consistent 12. Shaw, R. 1996. Unpublished data. with the dilution factor, a much smaller insignificant increment 13. Cho, C. Y. 1992. Feeding systems for rainbow trout and in xenoma count was produced when the dilution decreased other salmonids with reference to current estimates of en- from 1:10 to 1:1. Relatively little is known about sequential steps ergy and protein requirements. Aquaculture. 100:107–123. in the life cycle of L. salmonae. Once a spore is ingested and the 14. Kent, M. L., and R. P. Hedrick. 1986. Development of the sporoplasm crosses the intestinal mucosa, its fate is unknown PKX myxosporean in rainbow trout, Salmo gairdneri. Dis. until the point at which it is detected in transformed gill endot- Aquat. Org. 1:169–182. 58