- Interest in the distribution of Clostridium botulinum type E
was heightened by the sudden outbreak of human botulism from
smoked whitefish chubs and canned tuna fish in 1963. The question
arose as to how widely the organism is distributed among fish
and shellfish in the Northwest and what potential hazard exists
for the consumer of fish products. This sporeforming anaerobic
orgnism is heat sensitive and had eluded detection in other
surveys where heat shock had been used to eliminate non
sporeforming contaminants. More recent study using other
techniques than heat to facilitate recovery has shown this organism
to be widespread, especially in the marine environment.
This study was undertaken to find the incidence and
distribution of C. botulinum in the marine organisms and
environment of the Pacific Northwest and the food products
All species of fish were examined by incubating the gills
and viscera individually in tryptone-peptone-glucose medium
anaerobically at 28 C for four days and testing the culture filtrate
for mouse toxicity by intraperitoneal injection. Toxic filtrates
were typed by retesting them in mice protected by specific botulinal
antitoxin of type A, B, E, or F.
Among salmonid fish the proportions of specimens of each
species yielding toxic filtrates were as follows. Sockeye salmon
from the Columbia River, 14 of 59 (23.7%); Chinook salmon from
the Columbia River, 19 of 106 (18.0%); Chinook salmon from the
Pacific Ocean, 1 of 18 (5,6%); Coho salmon from the Columbia
River, 10 of 19 (34.4%); Coho salmon from the Pacific Ocean,
13 of 186 (7.0%); Steelhead trout from the Alsea River, 7 of 37
(19.0%). About one-third of the toxic cultural filtrates were
successfully typed and proven to contain botulinal toxin. Most of
them proved to be type E toxin but 3 were type A, 3 were type
B and one, a comparatively new type, type F, was isolated
from a Sockeye salmon in the Columbia River.
Pure cultures of Clostridium botulinum type E were
isolated from 18 specimens and one specimen yielded a pure
culture of type F from a sockeye salmon. This was the second
time this type had been isolated. In all of the experimental
groups the proportion of fish producing toxigenic cultures was
significantly higher in those taken in the two rivers than those
of the same species taken from the ocean waters.
"Bottom fish" represented by Cod, Sole, Grouper and members
of the Sebastodes group were also tested in the manner described
above. The number of specimens yielding toxic filtrates were
28 of 157 (17.8%). When grouped according to location at which
the fish were caught, those near the mouth of the Columbia
River produced a greater percentage of toxic filtrates than did
those caught off the open shore line. The results were as follows:
Bottom fish from Astoria, 23 of 70 (32.8%), Botton fish from
Coos Bay, 5 of 87 (5.6%). Sturgeon specimens produced 3 of 24
(12.5%) toxic filtrates. Most of the species contained type E;
however, one type A and one type B were found on typing, with
about one-third of the toxic filtrates being successfully typed.
Environmental swab samples from the "deep sea" fillet
processing plants produced 3 of 39 (7.7%) toxic filtrates. None
of the 53 samples taken in the salmon processing plants produced
Shellfish were collected along the ocean beach and in the
estuaries. Three to five shellfish were combined into a single
specimen and treated as described. All shellfish obtained from
the estuaries demonstrated a higher percentage of toxic filtrates
than those obtained from the ocean beach. The results were as
follows: Razor clams, 11 of 75 (14.6%), Cockle clams, 12 of 15
(80.0%); Softshell clams, 8 of 12 (66.4%); Littleneck clams,
4 of 11(36.2%); Horseneck clams, 1 of 3 (33.3%); Oysters, 6 of
19 (31.6%); Dungeness crabs, 17 of 24 (71%). Only the razor
clams were collected exclusively from the ocean beach.
Loss of toxicity on holding mixed cultures at -15 C while
awaiting typing was a continual problem. This accounts for
only one-third of toxic filtrates being successfully typed.
Electrophoretic analysis of the total bacterial proteins was
carried out on cell sonicates and cell free culture filtrates by
first growing cells for four days at 28 C anaerobically. The cells
were separated, washed and disrupted with ultrasonic energy.
The cell free culture filtrate was concentrated 10 fold by dialysis
against polyethylene glycol 4000. Both the toxigenic organisms
and the toxic filtrate demonstrated an extra common protein band
in the upper third of the electrophoretic pattern not present in
the nontoxic spectra. This band might represent the type E toxin.
Differences could also be noted in the number of protein bands
in the lower third of the patterns in different nontoxigenic strains
and also when the toxigenic and nontoxigenic strains were compared.
This could suggest an association with a phage in the toxigenic