Fry age and size effects on immersion immunization of brook trout, Salvelinus fontinalis Mitchell, against infectious pancreatic necrosis virus

Journal of Fish Diseases 1990, 13, 113-125

Fry age and size effects on immersion immunization ofbrook trout, Salvelinus fontinalis Mitchell, againstinfectious pancreatic necrosis virus

L. M. BOOTLAND, P. DOBOS & R. M. W. STEVENSON Department ofMicrobiology, College of Biological Sciences, University of Guelph, Ontario, Canada

Abstract. Brook trout, Salvelinus fontinalis Mitchell, fry were divided into eight age groupsof 1-8 weeks post-hatch (wph) and immunized by a single direet immersion in formalin-inactivated infectious pancreatic neerosis virus (IPNV). After a direet-immersion IPNV

ehallenge given 4 weeks later, only fr>' immunized at 2, 3 and 6wph showed protection. Therelative per cent survival 60 days after IPNV challenge was highest in 2 and 3 wph fry(45-50%) and deereased as fish increased in age or size. The same response was obtainedone year later when four age groups of fry, from 1 to 4wph, were immunized with adifferent serotype and dose of IPNV. The 2 and 3wph fry had mean weights between 49 and60mg at immunization. Killed vaccines administered by immersion have not previously beenreported as inducing proteetion in salmonid fry of such low weights. Analysis of the growthof the fry suggests that proteetion against IPNV requires immunization in the eleuthero-embryo phase, during the time of slow weight gain. This appears to be in direct contrast tothe stage of ontogeny and weight growth rate required for successful immunization againstthe baeterial pathogen Vibrio anguillarum. Although immunization with two IPNV serotypesreduced mortalities from challenges with these same virus isolates, it did not prevent IPNV

infection of fry in any age group.

Introduction

Infectious pancreatic necrosis virus (IPNV) can cause high mortality rates in young salmonids.IPNV mortalities are highest in fry at first feeding, but become negligible by approximately6 months of age (Frantsi & Savan 1971; Dorson &. Torchy 1981). For immunization to beuseful as a means of protecting these fish, a vaccine would have to be administered at a veryyoung age and induce a response rapidly. McAllister (1984) reported that 10-day post-hatchbrook trout, Salvelinus fontinalis Mitchell, fry could be protected against IPNV by vaccinationwith a live, avirulent IPNV isolate. However, the effect of the live IPNV vaccine may have beendue to interference rather than an immune response. Live IPNV vaccines have not alwaysproduced protection (Hill, Dorson & Dixon 1980) and, in any case, there would be difficultiesassociated with licensing a live viral vaccine.

Inactivated IPNV vaccines administered by intraperitoneal injection were successfully usedto protect 4 to 8wph rainbow trout (Dorson 1977; Hill et al. 1980), but injection is not aconvenient method for large numbers of small fish. Immersion would be the method of choice,but it was not successful in IPNV trials carried out by Dorson (1977) and Hill et al. (1980). Inone other case, immunization of brook trout fry by immersion in three very high doses ofinactivated IPNV (up to 187 fig viral protein per ml) appeared to exacerbate the effect ofsubsequent challenge (BooUand, Stevenson & Dobos 1986). The fry in that study were under

Correspondence: Dr. R. M. W. Stevenson, Department of Microbiology, College of Biologieal Sciences,University of Guelph, Guelph, Ontario, Canada NIG 2W1.

113

114 L. M. Bootland ct al.

the usual weight for their age, but the result points out one of the potential risks of immunizingvery young fish against IFNV, which is that the fish may not be immunocompetent. Thelymphoid system of salmonids is immature at the time of hatching (Grace & Manning 1980)and exposure to antigens at this stage may lead to tolerance instead of protection. Tolerancehas been induced in young carp and rainbow trout by injection of mammalian thymus (T)-dcpcndent antigens (Van Loon, Van Oosterom & Van Muiswinkel 1980; Manning, Grace &Secombes 1982), but not to T-independent antigens such as whole Gram-negative bacteria(Tatner & Home 1984). As viruses appear to be T-dependent antigens (Onions 1983), it ispossible that high concentrations of IPNV indueed a state of toleranee (Bootland et al. 1986).

The objective of this study was to expand the previous IPNV immersion immunizationstudy, in order to determine the effect of fish age and weight at the time of immunization onsubsequent protection. The results of the two independent trials reported here demonstratethat, when brook trout fry are immunized by direct immersion in formalin-inactivated IPNV,

the age and weight of the fry combine to have a significant and somewhat unexpected effect onthe induction of protection, which has to be taken into consideration in the development ofviral vaccines.

Materials and methods

Virus and vaccine preparations

The two virulent field isolates of IPNV used in this study were the previously describedVR-299 serotype Quebec isolate (Bootland et al. 1986) and an isolate from Idaho rainbowtrout. The latter isolate was kindly provided by Dr R. Buseh, and belongs to a new, hithertoundesignated serotype (Dr E. Nagy, personal communication).

Monolayer cultures of chinook salmon embryo (CHSE-214) cells were grown in 1-9-1 glassroller bottles at 18°C in 150ml Eagle's minimum essential medium with Earle salts (Gibco),supplemented with 6% foetal bovine serum, 200 IU/ml penicillin, and 200;i^g/ml streptomycin(cMEM). The cells were infeeted with IPNV at a low muItipHcity of infection and virus washarvested when complete cytopathie effect (CPE) was observed, usually in 3—4 days. Tissueculture passage 5 of each IPNV isolate was concentrated by polyethylene glycol-NaCl precipi-tation, then purified by Freon extraction and isopycnic CsCl gradient centrifugation as de-scribed by Chang, MacDonald & Yamamoto (1978). The purified virus was dialyzed againstTNE (lOmM Tris, lmMEDTA, 0-1M NaCl, pH 7-5) for 3h at 4°C to remove CsCl.

The virus titre of eaeh purified isolate was determined by plaque assays using 14 mMHEPES buffer in the overlay (Wolf & Quimby 1973) and expressed as plaque-forming units(pfu/ml). Purified IPNV was inactivated by formalin (Bootland et al. 1986) and proteinconcentrations determined by the method of Lowry, Rosebrough, Farr & Randall (1951).

Fish experiments

Two lots of genetically related brook trout fry were obtained by spawning first and secondgeneration stocks of laboratory-reared broodstock. Eggs from each mating cross were rearedseparately in fibreglass egg trays with flowing well water at 7 - 8 5°C. After hatching, each ofthe progeny groups were maintained separately in 55-1 plastic holding tanks. The tank hds

Immunization against IPNV 115

were partially covered to provide low light intensity, and flowing, aerated well water wassupplied at 10 ± TC. When fry became positively phototactic with the onset of feeding at3 weeks post-hatch (wph), covers were removed from tank lids and fry were fed with acommercially prepared trout diet (Martin Feed Mills Ltd, Elniira, Ontario, Canada). Fishmaintenance and experimental procedures were in accordance with the principles of theCanadian Council on Animal Care (1984).

Fry immuuization

Two fry immunization trials were performed one year apart. The experimental fish used ineach trial were progeny that had hatched on the same day from eggs of one broodstock pair.For each of the 8 weeks in trial I, 210 fry were randomly selected from the holding tank. Ofthese, 10 fry were measured for total length (snout to tail tip) and wet weights determined. Agroup of 100 fry were immunized by a 45min bath in 1-14^g/ml inactivated Idaho IPNV (Igfish per 12 ml, 1 37 x 10 * ' particles iPNv/ml) at 12°C, while the unimmunized control groupconsisted of 100 fish immersed in water only. The groups of fish were placed in separate 9-1plastic tanks supplied with flowing, aerated well water at 10 ± TC, and mortalities weremonitored daily for 60 days. For trial II, the same protocol was used with four age groups (1 —4 wph), but fry were immunized with the Quebec isolate, at a 30 times lower dose of antigen(0-038/^g/ml, 4-54 x 10* particles iPNv/ml).

Fry challenge

Control and immunized fish from each age group were challenged 4 weeks post-immunization(wpi) by immersing 90 fish for 5h in id"" pfu/ml Idaho (trial I) or Quebec (trial II) IPNV,

using a fish density of 150 g/1 at 12°C. The remaining fry in each group were measured andweighed. After fry were challenged and returned to the tanks, the effluent water was disinfectedby contact with sodium hypochlorite at a final concentration of 10ppm for a minimum of 2h.Mortalities were monitored daily for 60 days post-challenge (dpc), then the survivors weresacrificed and their wet weight measured. For each group, the total of all daily mortalitiesoccurring within 60dpc is reported as the cumulative percent mortality (C%M). Protection wasexpressed as the relative percent survival (RPS) at 60 dpc, using the formula given by Amend(1981):

% vaccinate mortalityRPS = 1 - % control mortality

X 100

The X" t^st with Yates correction (Wardlaw 1985) at P < 0-05 was used to evaluate thesimilarity in C%M 60dpc between control and immunized fry within each age group.

Detection of IPNV in fry

All mortalities and sacrificed fry were tested individually for IPNV infection. Each was placed ina plastic Whirl-Pac bag with sufficient cMEM to give a 1/50 (w/v) dilution, then homogenizedwith a Stomacher (Seward Labs, UK). The samples were eentrifuged at 1200^ for 15 min at4X, passed through a 0-45 fim membrane filter, and stored at 4°C until processed. For eachfish sample, 0-lml of the 1/50 and 1/100 dilutions were added in duplicate to 0-1 ml of CHSE-214 eells in wells of 96-well plastic microtitre plates (Nunc). The monolayers were grown at

L. M. Bootland et al.

18°C for 14 days and observed periodically for CPE. Medium from each well (35i(/l) was blind-passaged onto a new cell monolayer and cells were again examined for CPE for 14 days.Staphyloeoccal coagglutination was used to confirm that CPE was caused by IPNV (Kimura,Yoshimizu & Yasuda 1984).

F/.s7/ growth

Fish growth in length and weight was analyzed to determine if immunization influenced growthrates and, conversely, if fish growth rates were related to the ability to induce protection.Pooled r-tests (P < 0-05) were used to test for significant differences in fish weight and lengthwithin and between specific age groups for the two trials.

Simple linear regression analysis (Wardlaw 1985) was used on logarithmically transformed(logui) wet weights (mg) and total lengths (mm) of individual fish measured at immunization orchallenge to calculate the relationship of fish length and weight to fish age (1-12wph). Theslope of each regression line is an expression of fish growth rate. The significance of all slopes(HQ: b=O) and regressions were tested at a significance level of 5%, using r-tests and analysisof variance (ANOVA) respectively (Wardlaw 1985). Residual plots were examined to determineif more than one regression line was necessary to fit the data for eaeh trial.

Results

Mortalities from immunization to challenge

For the 28 days following immunization and prior to challenge, cumulative per cent mortalitiesin each fry age group were generally less than 10%. The exception was for the 5 wph fish intrial I, where mortalities were 15% for the control fish and 11% for the immunized fish.Overall, the immunized fish did not suffer higher mortalities than the control groups. Mortalitiesoccurred within 7 days after immunization (dpi) except in the ease of the 1 wph control fish intrial II, when the majority of the mortalities occurred at 23 dpi, near the onset of feeding, IPNV

was not isolated from any of the prechallenge mortalities.

Relative per cent survival after challenge

In trial I, only fry immunized at 2, 3 or 6wph were protected against a waterborne IPNV

challenge. The immunized fry in these three age groups had significantly lower C%M at 60dpcthan the corresponding unimmunized control fish (Table 1). Relative percent survival (RPS)was highest for the 2 and 3 wph age groups. For the 1, 4, 5, 7 and 8wph age groups, there wasno significant difference in C%M at 60dpe between the immunized and control fish and RPSwas low (Table 1). In trial II, carried out one year later, only l -4wph fish were tested forimmunization-induced protection, as the objective was to protect fish at the earliest possibleage. In the repeated trial, once again, only the fish vaccinated at 2 and 3 wph were protected asC%M at 60dpe was significantly lower for immunized fish compared to control fish (Table 1).

The RPS values for all age groups in trial I were higher than for the corresponding agegroups in trial 11, which had been immunized with a lower vaccine dose (Table 1). In trial II,challenge of 1 wph fry with Quebec IPNV resulted in a significantly higher C%M over the 60 dpcthan did challenge with the Idaho IPNV in trial I (Fig. 1). For 2, 3 and 4 wph control groups, the

Immunization against IPNV 111

Table 1. Cumulative per eent mortality and relative per cent survival 60 days after challenge for brook trout fryimmunized by immersion in formalin-inactivated tPNV*

Immunizationage (wph)

12345678

Control

2688744976525031

Trial I

C%M

Immunized

2744t41^517032^4136

RPS

- 45045

- 48

3818

-16

Trial II

C%M

Control

76737753

Immunized

84

62^56

RPS

-111919

• - 6

* Groups of 90 brook trout fry were immunized at 1-8wph with Idaho IPNV (trial I) or a30-times-lower dose of Quebee IPNV (trial II). C%M and RPS values were calculated asdescribed in 'Methods'.' Indicates a signifieant difference in mortalities between control and immunized groups(P < 0 05).

10 20 30 40 50 60

100

80

60

40

20

0

b) 2 wph

10 20 30 40 50 60

oE<u

100 I- ( c} 3 wph

30

(d )4wph100

80

60

40

20

40 50 60 0 10 20 30 40 50 60Time (days post-challenge)

Figure 1. Cumulative per eent mortalities in brook trout fry after bath-challenge with IPNV, given 4 weeks afterbath-immunization. The fry were immunized at (a) 1 wph; (b) 2wph; (c) 3wph; and (d) 4wph. The results of twoindependent trials have been superimposed. In trial I, fry were vaccinated and challenged with Idaho [PNV(triangles); in trial 11, vaeeination and challenges were with Quebec IPNV (circles). The open symbols indicategroups of fry vaccinated with inactivated virus; closed symbols are unimmunized control groups (n = 90 fish pergroup).

118 /.. M. Bootland ct al.

differences in C%M over the bOdpc between the two trials became less pronounced withincreasing fish age (Fig. I).

Prevalence of //'N\' infection

The mortalities in the control and immunized fry groups were 96-100% IPNV infected in trial I(Table 2) and 98-100% infected in trial II (data not shown). For sacrificed fry, 93-100% ofthe l -4wph groups were infected in trial 1 (Table 2) and 100% in trial II. For fry sacrificed at(M)dpc in the 4-<Swph groups, the prevalence of iPNV-infection generally decreased withincreasing fish age (Table 2). In both trials, the prevalence of IPNV infection in control andimmunized fry was not significantly different within any age group nor over all age groups(Table 2). Although immunization reduced mortalities in the 2, 3 and 6wph age groups, it didnot prevent the establishment of IPNV infection in any of the 8 age groups.

Table 2. Prevalence of IPNV infection in immunized and unimmunized brook trout fry after IPNV challenge(trial I)

Percentage of fish infected with IPNV

Mortalities to 60dpc: Sacrificed at 60dpc:Immunization

10099100100100

(44)(66)(47)(44)(28)

10096100100100

(46)(56)(27)(37)(32)

9483615052

(16)(12)(23)(22)(21)

10067633965

(15)(12)(24)(23)(20)

age (wph)' Control Immunized Control Immunized

1 100 (24)^ % (24) 100 (15) 93 (15)2 99 (79) 9S (40) 100 (13) 93 (15)3 100 (66) 100 (38) 95 (20) 100 (20)45678

Mean percentage 99-5 (398) 98-7 (300) 76-1 (142) 75-4 (142)of infected fish

" Brook trout fr)' immunized at 1-8 wph by immersion in formalin- inactivated Idaho IPNV werebath-challenged 4 weeks later and mortalities collected for 60 days (dpc). Fish were testedindividually for IPNV infection.'' Bracketed numbers for each group indicate the number of fish tested for IPNV.

Fish growth

In both trials, immunized and control fish within each age group were not significantly differentin mean total length or mean wet weight at challenge or when sacrificed (data not shown).Therefore, to obtain mean weights and lengths for age groups, the individual fish weights andlengths of both control and immunized fry within each age group were pooled. In both trials,each age group at least doubled in mean weight during the 4 weeks from immunization tochallenge (Table 3), but they did not double in mean length (Table 4), indicating fish grewfaster in weight than length.

The statistics derived from linear regression analysis of fish log lengths (L) and log weights(W) versus age in weeks (A) are shown in Table 5. The regression equations can be obtainedfrom the information in Table 5 by using the general equations:

Immunization against IPNV 119

log L = a + bAand log W = a H- bA.

All slopes (b), regressions and correlation coefficients (r) were significant (P < 0-05). In eachtrial, only one regression line was necessary for length versus age data, indicating that fishlength increased linearly at a constant rate with age. Length growth rates were similar for thetwo trials. Unlike the length growth rates, weight growth rates were not constant with fish age.Two regression lines were necessary to fit logio weight versus age data, with an increase ingrowth rate occurring at 3 wph in trial I (Fig. 2) and trial II (Table 5). Fry used in trial II had afaster weight growth rate (b) than did fry in trial I.

The best protection against IPNV was obtained when fry were immunized in the weightrange of 49—60mg during the period of slow weight gain (2 and 3 wph). If fish were outside

Figure 2. Scatterplot of logmweight (mg) versus age (wph) ofbrook trout fr>' fromimmunization trial I. Statisticaldata for the graph is given inTable 5. Data points representindividual fish, and the graphcombines the results from all fish,regardless of the immunization orchallenge treatment given.

3-4

3-1

o 2-8

2-5 •

2-2 -

-i 1-9 •

6 -

, ' 1 •

-

i l l

1 '

1 1

1 ' 1 '

• ^

I/- '

I 1 1 r

1 ' 1 ' .

/ ^

-

1 1 1 1 '

0 2 4 6 8 10 12Age {weeks post-hatch)

14

Table 3. Mean wet weights of eight age groups of brook irout fry at IPNV immunization and at challenge 4 weekslater

Immunizationage (wph)

12345678

Mean wet

At immunization («—10)

Trial

50 ±53 ±56 ±70 ±78 ±

135 ±185 ±260 ±

1

2111269

24

Trial II

46 ± 249 ± 360 ± 264 ± 2

ndndndnd

weight

105153190282302543713812

(mg ± SE):

At challenge {n

Trial I

± 6(11)± 6 (10)± 8(17)± 15 ( 5)± 30 (11)± 18 (18)± 41 (18)± 38 (14)

rTrial II

120 .154 :207 :269

± 10 ( 8)± 7(17)± 10 (17)± 27 ( 5)

ndndndnd

'^ Sample size was n=\0 fish per group at immunization but varied (bracketednumbers) at challenge.^ Results arc for two independent trials, run one year apart. In trial 11, no fish wereimmunized between 5 and 8 wph (nd).

120 L. M. Bootland et al.

4. Mean total lengths of brook trout fry at the time of IPNV immunization and at challenge 4 weeks later

Immunizationage (wph)

12345678

Mean total

At immunization {n= 10)'*

Trial l''

13-5 ± 0-516-7 ± 0-218-8 ± 0-520-4 ± 0-221-4 ± 0-225-6 ± 0-328-0 ± 0-530-0 ± 0-7

Trial II

15-7 ± 0-217-6 ± 0-419-8 ± 0-420-9 ± 0-3

ndndndnd

length (mm ± SE):

At challenge {n

Trial 1 {n)

23-8 ± 0-6(11)27-1 ± 0-4(10}28-6 ± 0-4(17)33-0 ± 0-6 ( 5)33-5 ± 1-4 ( 9)38-4 ± 0-4(18)42-2 ± 0-6(18)44-7 ± 0-7 (14)

)

Trial 11 (

25-325-0:28-830-6 :

± 0-8±0-3± 0-5± 0-8ndndndnd

n)

( 8)(17)(17)( 5)

^ Sample size was rt^IO fish per group at immunization but varied (bracketed numbers)at challenge. Lengths and weights (Table 3) were obtained from the same groups of fish.• Results are for two independent trials, run one year apart. In trial II, no fish wereimmunized between 5 and 8 wph (nd).

Table 5. Statistics describing the linear regressions of logjo weights and log 20 lengths versus brook trout age forimmunization trials

Statistical parameters^

Variables «

Trial IWeight v' age (1-3 wph)Weight vs age (3-12 wph)Length vs age (1-12 wph)

Trial IIWeight vs age (1-3 wph)Weight vs age (3-8wph)Length vs age ( l -8wph)

29164183

296687

1-651-30113

1-591-321-16

0-0320-1370-045

0-0620-1420-041

0-4480-9340-945

0-4210-8530-899

5-0756-855-9

4-4319-327-5

25-723093129

19-6371756

0-0280-0970-034

0-0610-0920-031

^ Values are: n = number of fish; a = y intercept; b = slope of regression line; r = coeffieient of detennination;t = t statistic for testing Ho: b = O; F = F statistic for testing the significance of the regression; S = standarderror of the estimate. Calculations were as described in 'Methods'.

this weight range and had a faster rate of weight growth than the 2 and 3wph fry, they werenot protected after IPNV challenge, with the exception of the 6wph fry of trial I (Table 1).

Discussion

Brook trout fry were protected against IPNV by direct immersion in inactivated IPNV, but onlywhen the fry had been immunized over a narrow age- weight range. Fry immunized at 2 and 3weeks post-hatch (wph), when their mean weight range was 49—60mg, had the highest levelsof protection. The 6 wph fish in trial I showed slight protection, but the rcprodudbility of the

Immunization against IPNV 121

effect was not tested. This age group was not included in trial II as the objective in immunizingfry is to protect them at as early an age as possible. Although the levels of protection observedwere low, this is the only report of successful protection of salmonid fry against IPNV byimmersion in inactivated IPNV.

Vaccination of young trout by direct immersion in formalin-inactivated IPNV did not haveany obvious deleterious effects. Overall mortalities prior to challenge were not significantlydifferent for immunized and control fry, and at the time of IPNV challenge, control andimmunized fry within each age group were equivalent in weight and length. Both the Idahoand Quebec isolates of IFNV, representing two different serotypes, were capable of inducingprotection to challenge with the homologous strain. Cross-protection between serotypes andnon-specific protection against other fish viruses were not tested. The concentration of theIPNV antigen used to immunize appeared to influence protection levels, as the RPS for 1 -4 wph fry in trial I was higher than those in trial II, in which a 30-times lower dose of antigenwas used. However, the virus isolates used may have also influenced the levels of protection.

Cumulative per cent mortalities (C%M) of lwph fry were higher in challenges with theQuebec IPNV than with Idaho IPNV, but the differences in C%M between the two trialsbeeame less evident with increasing fish age. The variation in C%M between the two trials maybe due to differences in virulence between the two IPNV isolates. Frantsi & Savan (1971) haveshown that IPNV isolates do vary in virulence, with the differences becoming less evident asthe age of infected fish increases. Differences in susceptibility of the two lots of brook troutused is a less likely reason for the variation as the two lots of fish were genetically related andwere reared under the same conditions.

Previous trials have tested the efficacy of immunization against the three major salmonidviruses, IPNV (Dorson 1977; Hill et al. 1980; Bootland et al. 1986), infectious haematopoieticnecrosis virus (IHNV) (Fryer, Rohovec, Tebbit, McMichael & Pilcher 1976; Gilmore,Engelking, Manning & Leong 1988) and Egtved virus (Vestergaard Jorgensen 1976, 1981; DeKinkelin & Bearzotti 1981). However, these studies have not considered the influence of fishage or weight, and only rarely are both the weight and age at immunization specified.Generally, only salmonids in a single age group (4 wph or older) and weighing 0-4 g or morehave been tested. The only study that clearly specifies that fry younger than 4 wph wereimmunized is that of McAllister (1984), in which 10-day post-hatch (dph) brook trout weresuccessfully protected by direct immersion in live attenuated IPNV.

The lack of IPNV protection of the immunized 4 to 8 wph brook trout in the present studyagrees with previous immunization trials in which inactivated ipm' was administered by directimmersion or hyperosmotic infiltration to 4 wph or older salmonids (Dorson 1977; Hill et al.1980; Bootland et al. 1986). However, inactivated IPNV did protect 4 wph or older rainbowtrout when injected intraperitoneally (Dorson 1977; Hill et al. 1980), suggesting that fish of thisage were immunocompetent. The apparent discrepancy in the ability to protect 4wph or oldersalmonids with inactivated IPNV could be due to the two different routes of antigen adminis-tration. Similarly, for IHNV and Egtved virus, better protection is obtained by intraperitonealinjection of inactivated virus than that afforded by bath immunization (Amend 1976; DeKinkelin & LeBerre 1977). Recently Gilmore et al. (1988) have successfully protected 0-4grainbow trout (of an unspecified age) by direct immersion in a subunit IHNV vaccine. Thisdemonstrates that salmonids can be protected against viruses by direct immersion immunization,although high antigen concentrations may be needed.

In studies with bacterial pathogens, salmonid immunocompetence and RPS appear to in-crease more as a function of weight than age (Johnson, Fiynn & Amend 1982; Tatner & Horne

122 /.. M. Bootland a ;il.

S3, U)84). The lowest reported weight at which rainbow trout have been immunized andshown some subsequent evidence of protection against a bacterial pathogen is 140 mg at 2 wph(Tatnor iK: Horno 1 >83). A second trial, using the same methods, found a rather differentcritical point, 4(S1 mg at lOvvph (Tatner &. Hornc 1984), suggesting that the minimum weight atwhich salmonids become immunologically responsive to bacteria is not yet clearly established.For IPNV, the minimum weight at which brook trout fry could be successfully immunized was53mg in trial 1 and 49mg in trial 11 at 2 wph, the lowest reported immunization weights ofsalmonids for which protection against bacterial or viral pathogens has been achieved. Theminimum weight and age requirement necessary for induction of protection against virusesmay be lower than that needed for protection against bacteria, but this possibility has yet to beinvestigated.

It was surprising that only fry immunized over a narrow weight-age range were protectedagainst IPNV. It was also unexpected that RPS did not increase with fish age or weight, as thisis in contrast to results of bacterial immunization trials. The 1 wph fry were equivalent inweight to the 2 wph fry at immunization, but only the 2 wph fry were protected. This suggestedthat fish weight at immunization was not the only factor determining if protection was induced.Two other parameters that may influence the outcome of fish immunization are the stage inontogeny and the rate of growth at the time of immunization. In this study, fish growth rates interms of length and weight were used to assess whether these, or the stage of fry developmentthey reflect, were related to the ability to induce IPNV protection. As the length growth rateremained constant for 1 — 12wph brook trout, this growth parameter was not related to IPNV

protection. Based on the characteristics described by Balon (1980) for dividing brook troutontogeny into periods and phases, and the weight growth rates, the 1—3 wph fry in this studywere in the eleutheroembryo phase of the embryo period. These fry were not feeding, werenegatively phototactic, and had a slow weight growth rate. At approximately 3wph, fryentered the alevin period, characterized by positive phototaxis, onset of feeding and a switchto a rapid weight growth rate. The fry immunized at 8 wph were still in the alevin period. Theobserved increase in weight growth rate at the transition from embryos to alevins is inagreement with Balon's (1980) study of brook trout ontogeny. The success of immunization inprotecting brook trout fry showed a strong correlation with the stage in ontogeny and/orweight growth rates at immunization. Immunized 2 and 3 wph eleutheroembryos with a slowweight growth rate were protected, but immunized alevins with a faster weight growth ratewere not.

To determine if weight growth rate also influenced immunization-induced protection againstbacteria, the weights and ages of rainbow trout vaccinated against V. anguillarum wereanalysed, using the values reported by Tatner & Horne (1983, 1984). In the 1983 study, fishweight growth rate increased at 146mg (4 wph), but the transition to a faster weight growthrate did not occur until 481 mg (10 wph) in the 1984 experiment. In both trials, fish were onlyprotected if they were immunized at the time of, or after, the transition to a faster rate ofweight growth. Thus, fish weight growth rates appear to be important determinants for fishprotection with both IPNV and V. anguillarum immunization, but in contrasting ways. Success-ful protection required a slow weight growth rate for IPNV immunization and a faster weightgrowth rate in the case of V. anguillarum vaccination. Further comparisons are needed todetermine if these contrasting responses are general requirements for viral and bacterialimmunizations respectively.

The reason why only brook trout immunized in the eleutheroembryo phase, with coicomi-tant slow growth rates, were successfully protected against IPNV remains unknown. H^ er.

Immunization against IPNV 123

it may relate to the growth and development of the lymphoid organs. Lymphoid organs are notonly important in the immune response but they are also involved in IPNV pathogenesis as thevirus infects the spleen, kidney and leucocytes of salmonids (Reno, Darley & Savan 1978;Swanson & Gillespie 1982; Yu, MacDonald & Moore 1982). Growth and development of theiymphoid organs of young brook trout have not been studied, but in rainbow trout thelymphoid organs reach a maximum percentage of body weight two months after hatching, at abody weight of 0-24g (Tatner & Manning 1983; Weatherly & Gill 1983). The body weights ofthe rainbow trout in the Tatner & Manning (1983) study suggest a transition to a faster weightgrowth rate at 2 months post-hatching. Lymphoid organs may reach a maximum percentage ofbody weight at or near the time of the transition in weight growth rate. If this is true,salmonids may need to be immunized prior to maximum lymphoid organ weight to induceIPNV protection.

Although the lwph fry were eleutheroembryos with a slow weight growth rate, they werenot protected by IPNV immunization. The lwph fry in this study would still have had a thickmucus layer covering the gills and skin as yolk sac re-absorption was not complete (Blackstock& Pickering 1982). Also, the lwph fry probably had not switched from cutaneous to branchialrespiration (Balon 1980). As the gills are probably the major portal of entry for IPNV, similarto the findings for Egtved virus (Chilmonczyk 1980), it is likely that there was little IPNV

antigen uptake by lwph fry. The 2wph fry would probably have had higher uptake of virusantigen as they would have switched to branchial respiration and the mucus layer would bethinner as yolk sac reabsorption was nearly complete. Prevention of antigen uptake may be anearly non-specific defence mechanism to protect young fish from exposure to foreign antigens.

The stage in ontogeny thus appears to be a significant factor in the outcome of fryimmunization. An approximation of this stage can be made by clearly defining the fish age andweight, or weight growth rates, and water temperature. These parameters should all bespecified when reporting immunization trial results, as they would facilitate comparisonsbetween immunization trials. For example, 8-5 wph brook trout fry used in a previous IPNV

immunization trial (Bootland et al. 1986) weighed the same at challenge (0'15g) as the 2 wphfry that were protected in the current study. Based on weight alone, one would predict thatthese fry should have been protected. Instead, immunization with a dose of inactivated IPNV

(i-Sl iig/m\), similar to the dose used in this study, induced a tolerance-like state afterchallenge. However, when both age and weight are considered, it is evident that the 8-5 wphfry were impaired in growth and possibly also in immunoiogical maturity, which may accountfor the observed tolerance-like state.

As direct immersion did induce protection in these trials, albeit at low levels, there is somepromise that an effective fry IPNV vaccine can be developed. The results of this study do notallow a distinction to be made between protection induced by a specific immune response toIPNV (that was most effectively induced in the eleutheroembryo stage) and protection due toactivation or enhancement of non-specific defence mechanisms. Increasing the concentration ofimmunizing antigen might help resolve this and also improve RPS results, but the amounts ofprotein required would make an expression vector for cloned IPNV antigens essential. Theimmunization strategy itself is not yet commercially feasible as it is not practical to immunizefry over such a narrow weight-age range. In addition, immunization did not prevent IPNV

infection of fry, hence horizontal transmission would not be prevented. Rational developmentof viral vaccines for fry requires further work to help us understand how the developingimmune system of fry responds to viral antigens, and what factors affect the induction ofprotection by immunization.

124 /.. M. Bootland et al.

Acknowledgmeots

This research was supported by Strategic Research Grant G1641 from the Natural Sciences andEngineering Research Council of Canada. Technical assistance from Dawn MacDonald andstatistical advice from Robert Korver were greatly appreciated.

References

Amend D. F. (1976) Prevention and control of viral diseases of salmonids. Journal of the Fisheries ResearchBoard of Canada 33, 1059-1066.

Amend D. F. (1981) Potency testing of fish vaccines. Developments in Biological Standardization 49, 447-454.Balon E. K. (1980) Early ontogeny of the brook charr, Salvelinus (Baione) fontinalis. In: Charrs: Salmonid

Fishes of the Genus Salvelinus (cd. by E. K. Balon), pp. 631-666. Dr W. Junk bv. The Hague.Blackstock N. & Pickering A. D. (1982) Changes in the concentration and histoehemistry of epidermal mucous

cells during the alevin and fry stages of the brown trout Salmo trutta. Journal of Zoology (London) 197,463-471.

Bootland L. M., Stevenson R. M. W. & Dobos P. (1986) Experimental induction of the carrier state in yearlingbrook trout: a model challenge protocol for IPNV immunization. Veterinary Immunology and Immuno-pathology 12, 365-372.

Chang N., MacDonald R. D. & Yamamoto T. (1978) Purification of infectious pancreatic necrosis (IPN) virusand comparison of polypeptide composition of different isolates. Canadian Journal of Microbiology 2A,19-27.

Chilmonczyk S. (1980) Some aspects of trout gill structure in relation to Egtved virus infection and defensemechanisms. In: Fish Diseases (ed. by W. Ahne), pp. 18-22. Springer-Verlag, Berlin.

De Kinkelin P. & Bearzotti M. (1981) Immunization of rainbow trout against viral haemorrhagic septicaemia(VHS) with a thermoresistant variant of the virus. Developments in Biological Standardization 49, 431—439.

De Kinkelin P. & LeBerre M. (1977) Demonstration de la protection de la truite arc-en-ciel contre la SHV, parI'administration d'un virus inactive. Bulletin de TOfftce Jnternational des Epizooties 87, 401-402.

Dorson M. (1977) Vaccination trials of rainbow trout fry against infectious pancreatic necrosis. Bulletin derOffice International des Epizooties 87, 405—4('J6.

Dorson M. & Torchy C. (1981) The influenec offish age and water temperature on mortalities of rainbow trout,Salmo gairdneri Richardson, caused by a European strain of infectious panereatie necrosis virus. Journal ofFish Diseases 4. 213-221.

Frantsi C. & Savan M. (1971) Infectious pancreatic necrosis virus temperature and age factors in mortahty.Journal of Wildlife Diseases 7, 249-255.

Fryer J. L., Rohovec J. S., Tebbit G. L., McMichael J. S. & Pilcher K. S. (1976) Vaccination for control ofinfectious diseases in Paeific salmon. Fish Pathology 10, 155-164.

Gilmore R. D. Jr, Engelking H. M., Manning D. S. & Leong J. C. (1988) Expression in Escherichia coli of anepitope of the glyeoprotein of infectious haematopoietic necrosis virus protects against viral challenge. BiolTechnology 6, 295-300.

Grace M. F. & Manning M. J. (1980) Histogenesis of the lymphoid organs in rainbow trout, Salmo gairdneriRich. 1836. Developmental and Comparative Immunology 4, 255—264.

Fiill B. J., Dorson M. & Dixon P. F. (1980) Studies on immunization of trout against IPN. In: Fish Diseases (ed.by W. Ahne), pp. 29-36. Springer-Verlag, Berhn.

Johnson K. A., Flynn J. K. & Amend D. F. (1982) Onset of immunity in salmonid fry vaccinated by directimmersion in Vibrio anguillarum and Y. ruckeri baeterins. Journal of Fish Diseases 5, 197-205.

Kimura T., Yoshimizu M. & Yasuda H. (1984) Rapid, simple serological diagnosis of infectious pancreaticnecrosis by coagglutination test using antibody-sensitized staphylococci. Fish Pathology 19, 25-33.

Lowry O. Fl., Rosebrough N. J., Farr A. L. & Randall R. J. (1951) Protein measurement with the Fohn phenolreagent. Journal of Biological Chemistry 193, 265—275.

Manning M. J., Grace M. F. & Secombes C. J. (1982) Developmental aspeets of immunity and tolerance in fish.In: Microbial Diseases of Fish (ed. by R. J. Roberts), pp. 31-46. Academic Press, New York.

McAllister P. E. (1984) Infectious pancreatic necrosis immunization. Fish Health Section/American FisheriesSociety Newsletter 12, 6.

Onions D. E. (1983) The immune response to virus infections. Veterinary Immunology and Immunopathology 4,237-277.

Immunization against IPNV 125

Reno P. W., Darlcy S. & Savan M. (1978) Infeetious pancreatic neerosis: experimental induetion of a earnerstate in trout. Journal of the Fisheries Research Board of Canada 55, 1451-1456.

Swanson R. N. & Gillespie J. H. (1982) Isolation of infeetious pancreatic neerosis virus from blood and bloodcomponents of experimentally infected trout. Canadian Journal of Fisheries and Aquatic Sciences 39,225-228.

Tatner M. F. & Horne M. T. (1983) Susceptibility and immunity to Vibrio anguillarum in post-hatching rainbowtrout fry, Salmo gairdneri Riehardson 1836. Developmental and Comparative Immunology 7, 465-472.

Tatner M. F. & Horne M. T. (1984) The effects of early exposure to K anguillarum vaccine on the immuneresponse of the fry of the rainbow trout, Salmo gairdneri Richardson. Aquaculture 41, 193-202.

Tatner M. F. & Manning M. J. (1983) Growth of the lymphoid organs in rainbow trout, Salmo gairdneri fromone to fifteen months of age. Journal of Zoology (London) 199, 503-520.

Van Loon J. J. A., Van Oosterom R. & Van Muiswinkel W. B. (1980) Development of the immune system incarp {Cyprinus carpio). In: Aspects of Developmental and Comparative Immunology (ed. by J. B. Solomon),pp. 469-470. Pergamon Press, Oxford.

Vestergaard Jorgcnsen P. E. (1976) Partial resistance of rainbow trout {Salmo gairdneri) to viral haemorrhagieseptieaemia (VHS) following exposure to non-virulent Egtved virus. Nordisk Veterinaennedicin 28,570-571.

Vestergaard Jorgensen P. E. (1981) Experimental VHS-vaccine in field trials. Bulletin of the European Associationof Fish Pathologists 1, 13-14.

Wardlaw A. C. (1985) Practical Statistics for Experimental Biologists. John Wiley and Sons, London.Weatherley A. H. & Gill H. S. (1983) Relative growth of tissues at different somatic growth rates in rainbow

trout Salmo gairdneri Richardson. Journal of Fish Biology 22, 43—60.Wolf K. & Quimby M. C. (1973) Fish viruses: buffers and methods for piaquing eight agents under normal

atmosphere. Applied Microbiology 25, 659-664.Yu K. K. Y., Maedonald R. D. & Moore A. R. (1982) Replication of infeetious pancreatie necrosis virus in

trout leucocytes and detection of the carrier state. Journal of Fish Diseases 5, 401-410.