276
observed that the partially purifi ed H lysin contained detectable levels of GCAT. This
enzyme possesses some similarities to the H lysin, e.g. molecular weight of 23.2
and 25.9 kDa, respectively (Buckley et al. 1982 ). Yet, the molecular weights were
much smaller than the 200 kDa size of “salmolysin”, the haemolytic toxin described
by Nomura et al. ( 1988 ). The ionic strengths needed for the elution of GCAT and H
lysin from ion-exchange gels were similar. These factors may complicate the isolation of pure H lysin, assuming that GCAT and the haemolysins are separate entities,
which appears to be the case. Thus, GCAT has not been reported to possess haemolytic activity, and is stable at room temperature (Buckley et al. 1982 ). Membrane
fi ltration of the preparation failed to remove GCAT, whereas H lysin activity was
lost after the procedure (Titball and Munn 1981 ). Other observations of H lysin
activity have indicated that haemolysis of horse erythrocytes occurs in two steps,
namely a fi rst stage in which there is no detectable cell lysis (this was termed the
pre-lytic stage), followed by a second phase involving haemoglobin release and
disruption of the cell membrane. Binding of the H lysin to the erythrocytes during
the pre-lytic stage does not occur. Together with the observation of an optimum
temperature of 25–33 °C for lysis, this suggests that the H lysin has enzymic action
on the erythrocyte membrane. Nevertheless, fi sh injected with H lysin appeared to
be unaffected, despite an apparent toxicity to rainbow trout gonad tissue cell lines.
Titball and Munn ( 1981 ) concluded that the failure of H lysin to elicit a response in
the fi sh experiments was explained by the use of an unsuitable route of administration or the injection of too low a quantity of the material. In addition, these authors
argued that possibly H lysin is non-toxic to fi sh, with no important role in the pathological process.
Several investigators have explained the relationship of haemolysins and proteases to virulence by using different strains of Aer. salmonicida. For example,
Hackett et al. ( 1984 ) studied the possibility of a plasmid-encoded origin for these
extracellular enzymes. Signifi cantly, the team concluded that the loss of proteolytic
and haemolytic activity in variants of wild-type Aer. salmonicida, obtained by
treatment with ethidium bromide, did not correlate with loss of plasmid
DNA. Moreover, there was no apparent change in the LD 100 between the virulent
wild-type strain and its protease-haemolysin defi cient variant. This implied that the
extracellular activities were not essential for virulence or, indeed, pathogenicity, at
least with regard to the acute form of furunculosis in rainbow trout. Two clones
derived from another virulent strain, one of which was negative for protease and
haemolysin production whereas the second derivative was positive for these attributes, were avirulent (LD 50 increased by greater than four orders of magnitude).
This was an important observation, denoting that attenuation of a virulent strain
occurred without loss of the A-layer, plasmids or extracellular proteolytic and haemolytic activities. For this reason, Hackett et al. ( 1984 ) concluded that virulence
was attributable to other, as yet unknown, factors. Titball and Munn ( 1985b ) also
studied the effects of quantitative differences in virulence on the production of
potential toxins by Aer. salmonicida. The release of ECP, i.e. proteases and haemolysins, by virulent strains and their avirulent attenuated derivatives (differing only in
the presence or absence of the A-layer) could not be linked directly to virulence.
5 Aeromonadaceae Representative (Aeromonas salmonicida)
observed that the partially purifi ed H lysin contained detectable levels of GCAT. This
enzyme possesses some similarities to the H lysin, e.g. molecular weight of 23.2
and 25.9 kDa, respectively (Buckley et al. 1982 ). Yet, the molecular weights were
much smaller than the 200 kDa size of “salmolysin”, the haemolytic toxin described
by Nomura et al. ( 1988 ). The ionic strengths needed for the elution of GCAT and H
lysin from ion-exchange gels were similar. These factors may complicate the isolation of pure H lysin, assuming that GCAT and the haemolysins are separate entities,
which appears to be the case. Thus, GCAT has not been reported to possess haemolytic activity, and is stable at room temperature (Buckley et al. 1982 ). Membrane
fi ltration of the preparation failed to remove GCAT, whereas H lysin activity was
lost after the procedure (Titball and Munn 1981 ). Other observations of H lysin
activity have indicated that haemolysis of horse erythrocytes occurs in two steps,
namely a fi rst stage in which there is no detectable cell lysis (this was termed the
pre-lytic stage), followed by a second phase involving haemoglobin release and
disruption of the cell membrane. Binding of the H lysin to the erythrocytes during
the pre-lytic stage does not occur. Together with the observation of an optimum
temperature of 25–33 °C for lysis, this suggests that the H lysin has enzymic action
on the erythrocyte membrane. Nevertheless, fi sh injected with H lysin appeared to
be unaffected, despite an apparent toxicity to rainbow trout gonad tissue cell lines.
Titball and Munn ( 1981 ) concluded that the failure of H lysin to elicit a response in
the fi sh experiments was explained by the use of an unsuitable route of administration or the injection of too low a quantity of the material. In addition, these authors
argued that possibly H lysin is non-toxic to fi sh, with no important role in the pathological process.
Several investigators have explained the relationship of haemolysins and proteases to virulence by using different strains of Aer. salmonicida. For example,
Hackett et al. ( 1984 ) studied the possibility of a plasmid-encoded origin for these
extracellular enzymes. Signifi cantly, the team concluded that the loss of proteolytic
and haemolytic activity in variants of wild-type Aer. salmonicida, obtained by
treatment with ethidium bromide, did not correlate with loss of plasmid
DNA. Moreover, there was no apparent change in the LD 100 between the virulent
wild-type strain and its protease-haemolysin defi cient variant. This implied that the
extracellular activities were not essential for virulence or, indeed, pathogenicity, at
least with regard to the acute form of furunculosis in rainbow trout. Two clones
derived from another virulent strain, one of which was negative for protease and
haemolysin production whereas the second derivative was positive for these attributes, were avirulent (LD 50 increased by greater than four orders of magnitude).
This was an important observation, denoting that attenuation of a virulent strain
occurred without loss of the A-layer, plasmids or extracellular proteolytic and haemolytic activities. For this reason, Hackett et al. ( 1984 ) concluded that virulence
was attributable to other, as yet unknown, factors. Titball and Munn ( 1985b ) also
studied the effects of quantitative differences in virulence on the production of
potential toxins by Aer. salmonicida. The release of ECP, i.e. proteases and haemolysins, by virulent strains and their avirulent attenuated derivatives (differing only in
the presence or absence of the A-layer) could not be linked directly to virulence.
5 Aeromonadaceae Representative (Aeromonas salmonicida)