Aeromonas hydrophila NRC 505 produced extracellular substances which were capable of causing pathological effects when injected into trout. Proteolytic activity and haemolytic activity of the extracellular products, and the effect on fish, were lost on heating. The extracellular substances from strain G35, a protease-deficient mutant, were significantly more toxic to both rainbow trout (Salmo gairdneri) and speckled trout (Salvelinus fontinalis) than the analogous preparation from the parental strain NRC 505. The response of speckled trout injected intraperitoneally with dilutions of the extracellular preparations implicates haemolytic activity as a significant lethality factor.

151The major extracellular proteases of Aeromonas hydrophila NRC 505 and ATCC 7966 are a thermostable metallo-protease (TSMP) and a thermolabile serine-protease (TLSP), respectively. The purified enzymes differed in sensitivity to heating at 56 "C for 30 min, in response to the inhibitors EDTA and phenylmethylsulphonyl fluoride (PMSF), and were antigenically distinct. When crude extracellular products (ECP) of 47 strains of A . hydrophila were screened, 27 strains produced both TSMP and TLSP. TSMP was the only extracellular protease produced by 19 strains, whereas only A . hydrophila ATCC 7966 produced TLSP as its sole protease. This distribution of protease enzymes among strains explains conflicting reports from studies in which single strains of A . hydrophila were examined. The TLSP of A. hydrophila was similar to the protease of A.-salmonicida; Either or both TSMP and TLSP were produced by some strains of A. sobria and A'. caviae.

Twenty-five strains of Renibacterium salmoninarum, from diverse geographical locations, were all hydrophobic when tested by ammonium sulphate aggregation and adherence to hydrocarbons. In addition, all strains caused haemagglutination of rabbit erythrocytes, but not of erythrocytes from salmonid fish. The hydrophobic and haemagglutinating properties were further characterized for R. salmoninarum ATCC 33209 (the type strain). Treating the bacteria with protease K or trypsin decreased hydrophobicity but had no effect on the ability to haemagglutinate rabbit erythrocytes. Heating the bacteria to 62 or 100 "C reduced hydrophobicity and removed the haemagglutinin from the cell surface. The haemagglutinin may be the heat-stable bacterial antigen extracted from the tissue of infected fish in serodiagnostic procedures.

Intraperitoneal injection of virulent infectious pancreatic necrosis virus (IPNV) into yearling brook trout Salvelinus fontinalis induced an asynlptomatic, chronic w u s infection that persisted for at least 76 wk post-lnlection (wpl) in spite of the production of a strong humoral immune response. At 8 wpi, 100 % of the fish in one of the injected groups were IPNV camers, as determined by organ and feces samphng. At 76 wpi. 95 % of the fish remained IPNV carriers. Vertical transmission of IPNV to progeny occurred, but the fry mortality rate, the prevalence of IPNV-infechon, and mean IPNV titres were low. Hence, vertical transm~ssion 1s not recommended a s a criterion for evaluating the efficacy of broodstock immunization. The IPNV carner state in yearlings that had survived a &rect mmersion in IPNV as fry was equivalent to that induced in yearlings by injection, in terms of the number of IPNV-infected organs per fish and the prevalence and titre of IPNV in the visceral organs when fish were sampled at 73 and 76 wpi, respectively. However, the Injected yearlings mounted a stronger humoral immune response to IPNV that the fish surviving the direct unmersion. Smce the IPNV camer state induced by injechon was nearly indistinguishable from the natural IPNV carner state, intraperitoneal injechon can be used as a challenge protocol for broodstock immunization tnals.

Viral haemorrhagic septicaemia virus (VHSV) in the Great Lakes has had a dramatic impact on fish husbandry because of the implications of the presence of a reportable disease. Experimental infections with VHSV IVb were conducted in rainbow trout, Oncorhynchus mykiss (Walbaum), and fathead minnows, Pimphales promelas (Rafinesque), to examine their susceptibility and the clinical impact of infection. Triplicate groups of rainbow trout (n = 40) were injected intraperitoneally (i.p.) with 100 microL 10(6.5)50% tissue culture infective doses (TCID(50)) or waterborne exposed to graded doses (10(4.5), 10(6.5), and 10(8.5) TCID(50) mL(-1)) of VHSV IVb. Duplicate groups of fathead minnows (n = 15) were i.p. injected with (10(6.5) TCID(50) 100 microL) or waterborne exposed (10(6.5) TCID(50) mL(-1)). All experiments were performed with single-pass well water maintained at 12 degrees C. Following either i.p. or waterborne exposure, VHSV RNA was detectable in both rainbow trout and fathead minnows by nested reverse transcription polymerase chain reaction (nRT-PCR) as early as 4-7 days post-infection (p.i.). Infected fathead minnow and rainbow trout exhibited lesions characteristic of VHS at 9 and 15 days p.i., respectively. Route of exposure had little effect on the onset of clinical signs. Cumulative mean mortality in rainbow trout was 4.4%, 2.6%, 2.6% and less than 1% in the i.p., high, medium and low dose waterborne exposures, respectively. Cumulative average mortality of 50% and 13% occurred in i.p. and waterborne-exposed fathead minnows, respectively. VHSV was detected from pooled rainbow trout tissue by RT-PCR and virus isolation at 38 days p.i., but not at 74 days p.i., regardless of the exposure route. Immunohistochemistry (IHC) with a rabbit antibody to VHSV IVb revealed the viral tissue tropisms following infection, with the identification of viral antigen in myocardium and necrotic branchial epithelium of both species and in gonadal tissue of fathead minnows. Rainbow trout, but not fathead minnows, are relatively refractory to experimental infection with VHSV IVb.