1‘“

l

   

,I

It

 
   
 

    

l (

M

DETECi‘lON AND DEFFEREWKATEON OF
CLOSTREDESM BUWLENUM BY
PYRGLYSIS - GAS ‘ LIQUID
CHROMM‘OGRAWY

Thesis for the Degree of M. S.
MICHEGAR STATE Ui‘éEKJ‘ERSiTY
RICKY D. CONE
19-69

LIBRARY ‘

Michigan State
University

      
   

[Ht‘525

r} L1
q

E7” #76

 

 

 

 

ABSTRACT

DETECTION AND DIFFERENTIATION OF
CLOSTRIDIUM BOTULINUM BY
PYROLYSIS-GAS—LIQUID
CHROMATOGRAPHY

 

By

Ricky D. Cone

The objective of this study was to investigate the
feasibility of using pyrolysis-gas-liquid chromatography
(PGLC) for identification and detection of growth and
toxin production by Clostridium botulinum. Ten strains

representing types A, B, and E of g. botulinum were

 

used in the study.

Vegetative cell cultures (10 hr incubation at 32 C)
and spore cultures (36 hr incubation at 32 C) of all 10
strains were grown in Trypticase-peptone—sucrose-yeast
extract medium (TPSY). Five type B strains were also
grown in Multipeptone-sucrose-nutramino acids medium
(MSN). Lyophilized samples of these cultures were
subjected to PGLC analysis and the resulting pyrograms
were examined for differences in elution patterns among
(1) vegetative cells grown in each medium, (2) spores
grown in each medium, (3) spores and vegetative cells
grown in each medium, (A) vegetative cells grown in

different media, and (5) spores grown in different media.

Ricky D. Cone

The possibility of using PGLC to detect growth and

toxin production of all 10 strains of g. botulinum was

 

investigated using a modification of the dialysis sac
culture technique described by Vinet and Fredette

(1951). The inoculum was added to the physiological
saline (0.89%) which was outside the dialysis sac and
nutrients were provided by dialysis of TPSY medium across
the membrane from inside the sac.

The PGLC analysis was conducted for the cultures
grcwn for 5 days at 32 C as well as the dialysate super-
natant fluid (DSF) which remained after centrifugation
of the cultureso Three control samples consisting of
(l) uninoculated DSF, (2) uninoculated DSF plus purified
toxin, and (3) a 1.0 mg sample of partially purified
toxin were analyzed in the same manner as other DSF
samples to assist in determining whether either growth
or toxin production could be detected.

Differences in pyrograms suitable for identifica-
tion purposes at the type level were obtained for spores
and vegetative cells from both media, but only a few
reliable differences were obtained at the strain level.
It was possible to differentiate pyrograms of spores
from those of vegetative cells grown in the same medium,
and distinct differences among cultures of the same

serotype harvested from different media were noted.

Ricky D. Cone

The dialysis sac technique was found to be useful
in detecting growth of g. botulinum. Identification of

the type of g. botulinum which had grown in the medium

 

was possible. Toxin production, however, was not

detectable by PGLC using this technique.

DETECTION AND DIFFERENTIATION OF CLOSTRIDIUM

 

BOTULINUM BY PYROLYSIS-GAS—LIQUID

 

CHROMATOGRAPHY
By

I"3
\J
lA‘T

,t
Ricky Df’Cone

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE
Department of Food Science

1969

To

Janice and Brent

ii

ACKNOWLEDGMENTS

The author wishes to express his appreciation to
his major professor, Dr. R. V. Lechowich, for his con-
tinued interest and guidance throughout this study and
for his aid in preparing the manuscript. The suggestions
of Dr. P. Markakis and Dr. F. R. Peabody in editing this
manuscript are also appreciated.

To his wife, Janice, the author is eSpecially
grateful for her patience and encouragement during this
study.

The author is deeply appreciative of the funds and
facilities provided by Michigan State University and the

United States Public Health Service to aid in this study.

iii

TABLE OF CONTENTS

 

Page
DEDICATION I O O O 0 O O O O O O O O 0 ii
ACKNOWLEDGMENTS . . . . . . . . . . . . . iii
LIST OF TABLES . . . . . . . . . . . . . vi
LIST OF FIGURES . . . . . . . . . . . . . vii
INTRODUCTION . . . . . . . . . . . . . 1
LITERATURE REVIEW . . . . . . . . . . . . 3
Toxin Detection Methods and Culture Differentia-
tion of Clostridium botulinum . . . . 3
Detection, Classification, and Identification of
Microorganisms by Gas—Liquid- Chromatography . . 8
Pyrolysis-Gas—Liquid Chromatography . . . . . IA
METHODS AND MATERIALS . . . . . . . . . . . 18
Microorganisms . . . . . . . . . . . . 18
Culture Media . . . . . . . . . . . . l9
Harvesting Cells and Spores . . . . . . . . 20
Direct Microscopic Counts . . . . . . . . . 21
TOXiCity Assays O O O O O O O 0 O O O O 22
Controls . . . . . . . . . . . . 22
Lyophilization of Samples . . . . . . . . . 23
Preparation of Column Material . . . . . . . 23
Packing and Conditioning Columns . . . . . . 2A
Pyrolysis Apparatus . . . . . . . . . . . 25
Gas Chromatograph . . . . . . . . . . . 25
Analysis of Samples . . . . . . . . . . . 25
Interpretation of Results . . . . . . . . . 26
RESULTS AND DISCUSSION . . . . . . . . . . 28
Effects of Sample Size . . . . . . . . . . 28
Analysis of Vegetative Cells . . . . . . . . 30
Analysis of Spores . . . . . . . . . 37
Comparison of Spores and Vegetative Cells . . . AA
Effect of Medium . . . . . . . . . . . . “5
Detection of Growth and Toxin Production . . . . A6

iv

SUMMARY AND CONCLUSIONS

LITERATURE CITED . .

Page

58

63

Table

LIST OF TABLES

 

 

 

 

 

Page
Direct microscopic counts of vegetative
cells of Clostridium botulinum types A, B,
and E grown in two different media . . . . 36
Direct microscopic counts and per cent
sporulation of cultures of Clostridium
botulinum types A, B, and E grown in two
different media . . . . . . . . . . . 39
Direct microscopic counts and per cent
sporulation of strains of Clostridium
botulinum types A, B, and E using the modi-
fied dialysis sac technique. . . . . . . 50

Toxin titers of various strains of Clostridium
botulinum types A, B, and E grown in duplicate

in Trypticase-peptone-sucrose yeast extract

medium . . . . . . . . . . . . 56

 

 

vi

Figure

LIST OF FIGURES

Pyrograms from vegetative cells of four
strains of Clostridium botulinum types A and
B grown in Trypticase—peptone-sucrose—yeast
extract medium. The capital letters denote
the type and Att is the attenuation . . . .

 

Pyrograms from vegetative cells of six strains
of Clostridium botulinum type E grown in
Trypticase-peptone-sucrose-yeast extract
medium. Att is the attenuation . . . .

 

Pyrograms from vegetative cells of four strains
of Clostridium botulinum type E grown in Multi-
peptone-sucrose-nutramino acids medium. Att

is the attenuation . . . . . . . . .

 

Comparison of pyrograms from Clostridium
botulinum A) spores and B) vegetative cells

of type E grown in Multipeptone-sucrose-
nutramino acids medium; C) spores; and D)
vegetative cells of type A grown in
Trypticase-peptone-sucrose-yeast extract
medium, and E) spores, and F) vegetative cells
of type E grown in Trypticase-peptone- sucrose-
yeast extract medium . . . . . . . . .

 

Pyrograms from spores of four strains of
Clostridium botulinum types A and B grown in
Trypticase-peptone-sucrose-yeast extract
medium. The capital letters denote the type
and Att is the attenuation . . . . . . .

 

Pyrograms from spores of six strains of
Clostridium botulinum type E grown in
Trypticase-peptone— —sucrose— yeast extract
medium. Att is the attenuation . . . . .

 

Pyrograms from spores of four strains of
Clostridium botulinum type E grown in
Multipeptone-sucrose-nutramino acids medium.
Att is the attenuation . . . . . . . .

vii

Page

31

32

33

35

38

Al

A2

Figure Page

8. Pyrograms from cultures of four strains of
Clostridium botulinum types A and B grown
using the modified dialysis sac technique.
The capital letters denote the type and Att
is the attenuation . . . . . . . . . A8

 

9. Pyrograms from cultures of six strains of
Clostridium botulinum type E grown using the
modified dialysis sac technique. Att is the
attenuation . . . . . . . . . . . . “9

 

 

lO. Pyrograms of the dialysate supernatant fluid
remaining after centrifugation of four strains
of Clostridium botulinum type A and type B
from the dialysate from the modified dialysis
sac technique. Capital letters denote the
type and Att is the attenuation . . . . . 51

 

ll. Pyrograms of the dialysate supernatant fluid
remaining after centrifugation of six strains
of Clostridium botulinum type E from the
dialysate from the modified dialysis sac
technique. Att is the attenuation . . . ,. 53

 

12. Comparison of pyrograms for samples consisting
of A) uninoculated dialysate supernatant
fluid (DSF); B) uninoculated DSF plus
partially purified toxin; C) 1.0 mg of par-
tially purified toxin; D) DSF of type E
Clostridium botulinum E) DSF of type A
Clostridium botulinum and F) DSF of type B
Clostridium botulinum. . . . . . . . . BA

 

 

 

 

 

viii

INTRODUCTION

There has been a substantial increase during the
last decade in studies related to the microorganism
Clostridium botulinum. This interest was primarily
brought about by the increase in outbreaks of type B
botulism throughout the world (Iida 32 $1., 1958; Dolman
and Iida, 1963), and in the United States in 1963
(Anonymous, 196A).8 8 A

Botulism is a rare, but highly fatal, disease for
which no adequate clinical diagnostic tool is available.
The study of botulism in foods usually involves (l)
demonstration that the filtrate of a suspected food
sample contains type specific toxin by use of animal
neutralization tests and (2) cultural procedures to
.isclate and identify the causative microorganism. The
use of animals in toxin assay procedures is limited by
(l) the availability of animals to many laboratories,
(2) the cost of the animals, (3) the lack of proper
facilities in most laboratories for handling animals,
(A) variability in response to toxin among species of
animals, and (5) differences in injection methods among
workers. Conventional methods of identification of g.
botulinum are also time consuming and often fail to

provide reliable results.

Since pyrolysis-gas-liquid chromatography has been
used successfully for the rapid identification of other
genera and species of bacteria (Oyama and Carle, 1967;
Reiner, 1967; Reiner and Ewing, 1968), it seemed
feasible that this technique might also be used for

identification of cultures of Q. botulinum. This method

 

would not only provide adequate identification of cul-
tures, it might also greatly decrease the minimum time
of 5 to 7 days required for detection and identification
of cultures by conventional methods.

The use of pyrolysis—gas-liquid-chromatography for
the identification of these microorganisms as well as
detection of their growth and toxin production was inves-
tigated. The effects of the growth medium and of per
cent sporulation on the resulting elution patterns after

pyrolysis were also studied.

LITERATURE REVIEW

Toxin Detection Methods and Culture
Differentiation of Clostridium
botulinum
Most of the methods used in the detection and
identification of clostridia involve the use of the
blackening reaction which occurs when sulfite and iron

are present in the isolation medium. Wilson and Blair

(1924) were the first to use this technique which has

since undergone many modifications. The modification

most commonly used today is the one described by Mossel
(1959) in which 0.05% (w/V) Na2SO3°7H2O and 10 ppm
lpolymyxin B sulfate were added to the medium. Angelotti
gt a1. (1962) reported that the incorporation of sulfa-
diazine and polymyxin in the Mossel based medium con-
taining 0.05% sulfite increased the selectivity of the
Inedium. More recently, Gibbs and Freame (1965) described
‘the use of a differential reinforced medium for isolating
(zlostridia. These workers also include a review of pre-
Krious selective and differential media that have been
Ilsed for clostridia.

The organisms of the group Clostridium botulinum,

hCIwever, are of primary importance to food microbiologists,

arnj the previously mentioned techniques fail to provide

the rapidity and ease which are necessary in isolating
and differentiating them on a routine basis. Morpho-
logical and biochemical characteristics of the organisms
of this group are traditionally used for identification
purposes. These methods are thoroughly discussed in
studies and reviews by Seddon (1922), Meyer and Gunnison
(1928), Schoenhol (1928), Gunnison and Meyer (1929),
Townsend (1929), Robinson (1930), Gunnison, Cummings,
and Meyer (1936), Hazen (1937), Dolman and Chang (1952),
Eales and Turner (1952), and Dolman and Murakami (1961).
Techniques for the isolation of g. botulinum from
heavily contaminated materials have been reviewed by
Kautter and Bartram (1967). McClung (1967) has provided
a review of methods used for isolation and identification

of g. botulinum up to 1966, and Kautter (1968) has com-

 

piled a more recent review of methods of isolation and
identification as well as detection of toxin production

by g. botulinum.

 

Several workers have reported the production of a
precipitate and the formation of a layer having a "pearly"
(iridescent) appearance adjacent to and covering the-
colonies of the different types of g. botulinum on egg
yolk agar medium that may be useful in isolating and
differentiating these bacteria (McClung and Toabe, 19A7;
'Willis and Hobbs, 1958; Willis and Hobbs, 1959). The

Ifiormer workers used Trypticase (BBL)-peptone—agar as the

basal medium for this method while the latter used beef
infusion agar. The use of liver-veal-agar as the basal
medium for the egg yolk reaction was investigated by
Hobbs, Stiebrs and Eklund (1967). They observed that
this medium was less efficient for isolating Q. botulinum
type E while it was just as effective as the other two

in isolating g. botulinum types A and B as well as Q.

 

sporogenes.
The fluorescent antibody technique has proven useful
in detecting thepresenceof C. botulinum microorganisms
in éaasié”matébia1 aswell as in confirming the type of
various isolates detected. Russian workers including
Kalatina (1960) and Bulatova and Kabanova (1960) were
among the first to find this technique useful for identi-

fying types of g. botulinum when they observed that cer-

 

tain types of luminescent sera would react only with

certain types of g. botulinum. Batty and Walker (1963)

 

and Walker and Batty (196”) were able to separate 9.

botulinum into three groups using fluorescent antibody

 

staining: (a) proteolytic types A, B, and F; (b) non-
proteolytic type E and D; and (c) nonproteolytic type

E. Boothroyd and Georgala (196A) also found that
fluorescent antibody staining showed considerable promise
as a tool for isolation and/or identification of g.
‘QQtulinum. These same workers (Georgala and Boothroyd,

1967) later indicated that more work was required on

the serological characteristics of cells, spores, and

flagella of each type before the technique can be widely
accepted. Walker and Batty (1967) further observed that
the fluorescent antibody technique could be used in the
identification of toxic as well as non—toxic strains of

g. botulinum. Ward, Carol, and Garrett (1967), however,

 

were unable to obtain consistent results with respect
to toxin production and fluorescence when mouse-toxin
tests were compared to fluorescent detection of toxin.
Hobbs and Anderson (1967) used polyacrylamide gel
electrophoresis to study relationships between 9. botulinum
and non-toxic clostridia or non-toxic mutants. They
observed that protein patterns of vegetative cell extracts
of all strains of type E examined were essentially the
same and that each of the types A, B and C gave different
characteristic patterns.
Midura g3 g1.(l967) examined the possibility of
'using immunofluorescence for detection of g. botulinum
types A, B, and E. They reported that this technique
could be quite useful as a presumptive indicator of
the presence of g. botulinum cells. Midura g2 g},
(1968) further combined the use of this technique with
animal tests to detect growth and toxin production by

Q. botulinum type E in experimentally inoculated packages

 

Of turkey roll. They were able to obtain a time rela-

tionship between the presence of vegetative cells,

determined by immunofluorescence, and the presence of
toxin as demonstrated by animal tests.

Immunological techniques have also been reported
to be useful in detecting toxins of Q. botulinum. Rycaj
(1956), Yafayer and Chepeler (1961), and Sinitsyn (1961)
indicated that passive hemagglutination of botulinal
antitoxin-coated, tanned sheep red blood cells was as
sensitive in detecting botulinal toxins as animal assays.
Johnson g3 gg, (1966) were able to detect one LD50 mouse
unit of botulinal toxin using hemagglutination. Vermilyea,
Walker and Ayres (1968) have employed the Ouchterlony
immunodiffusion technique to detect toxins of g.

botulinum. They indicated, however that the mouse assay

 

was a more sensitive method for detecting toxin since the
toxin had to be concentrated with Sephadex when the gel
diffusion technique is used.

More recently, Duda and Slack (1969) have demon-
strated toxin production by g. botulinum type A using
electron micrOSCOpy. Upon electron microscopic examina-
‘tion of ultrathin sections of spheroplasts of the organism
Inhich had been exposed to ferritin-labeled type A anti-
tsoxin, they observed the presence of ferritin-labeled
aintibody in the outer spore coats in the vegetative
(zell cytoplasm. No ferritin-labeled antibody was

(Dbserved in the spore cortex or spore core.

Detection, Classification,_and
Identification of Microorganisms
by Gas—Liquid-Chromatography

The utilization of gas-liquid-chromatography (GLC)
as an analytical tool became a reality when James and
Martin (1952) made the first successful separation of
volatile fatty acids. Since that time the use of GLC
has probably developed more rapidly than any other
'analytical technique.

Abel, deSchmertzing, and Peterson (1963) were the
first to investigate the feasibility of using GLC for the
classification of microorganisms. They extracted the
lipid components of the microorganisms, transesterified
them to component carboxylic acid methyl esters, and
resolved the esters by gas chromatography. The elution
patterns which they obtained permitted them to distinguish
between families of Enterobacteriaceae, Bacillaceae,
ygcrococcaceae, and Parvobacteriaceae. Differentiation
of genera and species was however, rather difficult.

Cattaneo g3 2;. (1964) using a similar method of
Gexamining methyl esters of bacterial fatty acids by GLC
inere unable to demonstrate significant differences
eumong strains of the same type of mycobacteria from
\farious sources. They did, however, show noticable

‘Iariation in elution patterns from one type of Myco-

Efigcterium to another.

 

Brown and Cosenza (196A) observed differences in
methyl esters of fatty acids among Gaffkya, Micrococcus,

and Veillonella by GLC. No attempt was made by these

 

workers to differentiate between closely related species.
Yamakawa and Ueta (196“), however, were able to demon-
strate differentiation at the species level in the
genus Neisseria with this technique.

Garner and Gennaro (1965) were able to classify
bacteria more readily than previous workers using GLC.
On the basis of "fingerprints" obtained from GLC
analysis of pyro-distillates of bacterial cultures using
a flame ionization detector, they were able to differ-
entiate between Aerobacter aerogenes, Aerobacter cloacea,
Aerobacter lipolyticus, Escherichia coli, Escherichia

aurescens, and Eschericia anaerogenes. Preparation of

 

the material for GLC analysis, however, was rather
complicated.

Bassette and Claydon (1965) found that GLC
analysis of head-space vapors of volatile microbial
:metabolic products of growth in deodorized milk could
be useful for characterizing bacteria. Pure cultures
(if microorganisms known to have widely different
Ckiaracteristics gave markedly different chromatographic
Péitterns while related bacteria produced similar pat-
tesrns. However, similar patterns were sufficiently

(tifferent to enable differentiation of these bacteria.

IO

Bawdon and Bassette (1966) used this technique to dif-
ferentiate E. ggtt and g. aerogenes.

The preparation of methyl esters from fatty acids
has continued to be an important technique in the dif-
ferentiation of bacteria by GLC. Moss and Lewis (1967)
examined methyl esters of fatty acids extracted from
whole cells of 41 strains representing 13 species of

Clostridium by GLC. Upon both visual and quantitative

 

comparisons of chromatograms they were able to rapidly
differentiate Clostridium perfringens, Clostridium

sporogenes, and Clostridium bifermentans from each other

 

and from 10 other species of clostridia. Steinhauer,
Flentge, and Lechowich (1967) on the basis of their
studies have also supported the suggestion that cellular
lipid patterns could be a valuable means of identifica—
tion of closely related bacteria.

Moss gt gt. (1967) examined the fatty acid methyl
esters of 27 isolates of Corynebacterium acnes by GLC
and observed that each strain possessed a similar pro-
file which was characterized by a large percentage of
0—15 branched—chain acids.

These workers felt that their findings, along with
CIther similarities of these cultures to those of the
genus Propionibacterium, strengthened previous suggestions
tiuat g. ggggg should be classified in the genus Propioni-

Egacterium. Moss gt gt. (1969) used the same technique

11

and further proposed a similarity between anaerobic
corynebacteria and a group of propionibacteria which
contained 0—15 saturated branched chain fatty acids in
the iso-form. The corynebacteria were distinctly dif-
ferent from another group of propionibacteria which con-
tained primarily anteiso fatty acids. Lewis, Weaver,
and Hollis (1968) have also observed a difference in the

fatty acid composition of various species of Neisseria

 

by use of GLC.

Henis, Gould and Alexander (1966) investigated the
possibility of detecting and differentiating microorganisms
by GLC analysis of unique metabolic products which the
organisms produce. Upon examination of ether extracts
of several species of Bacillus as well as individual

strains of E. coli, A_.aerogenes and Pseudomonas aeruginosa

 

using a gas chromatograph equipped with flame ionization
and electron capture detectors, several products were
observed. Both qualitative and quantitative differences
in the production of such products as acetic, propionic,
and butyric acids, ethyl alcohol, diacetyl, acetoin, and

2 ,3-butanediol made it possible to detect differences among
genera, species and strains. Mitruka and Alexander

(1967) further observed the sensitivity of the electron
C=<'9.pture detector in detecting such microbial metabolites

as acetoin and diacetyl in the picogram range.

l2

Mitruka and Alexander (1968) found that GLC
analysis of metabolic products enabled them to detect
the presence of Proteus vulgarus, Streptococcus

faecalis, Streptococcus liquefaciens, _E_. coli, Bacillus

 

cereus and Bacillus popilliae in 2 to 4 hr old cultures
inoculated with less than 1 x 10“ cells per ml. A some-
what longer period of 7 to 12 hr was required for the

detection of products formed by A. aerogenes, E. coli

 

 

K-l2, Staphylococcus aureus, and Salmonella mphimurium.
More recently, Mitruka and Alexander (1969) have
shown that bacterial growth may be even more easily
detected by the use of cometabolism which involves the
use of substrates which are presumed to yield products
detectable at low concentrations. A halogenated
compound which was incorporated in the growth medium was
metabolized by the bacterium but was not a growth sub-
strate, and a 7,000, 20,000, and 1,000 fold increase in

GLC detectability was obtained in _S_. typhimurium,

 

E. gait, and S. aureus respectively.

Differentiation of bacteria has also been accom—
plished by GLC analysis of ether extracts of the metabolic
products produced by washed bacterial cells suspended in
buffered glucose (O'Brien, 1967). The cells in this
experiment were considered to be resting or non—growing
Since the organisms grow poorly or not at all in this

Chemically defined medium.

l3

Cecchini and O'Brien (1968) reported a procedure
to differentiate E. coli from other bacteria by direct
GLC analysis of the culture medium. They were able to

cietect E. coli even in the presence of E. aerogenes

 

21nd other microorganisms using special growth conditions
saince only the E. ggtt produced ethyl alcohol.

Moore, Cato and Holdeman (1966) have observed

‘that fermentation patterns may be useful in characteriz-
:ing and identifying microorganisms. Examination of
:fermentation patterns of 20 species of clostridia by GLC
showed that the relative proportions and concentrations
<>f the alcohols may be useful in this respect.

MacGee (1968) has observed distinct differences
:in elution patterns when alcohol extracts of widely
(iifferent genera of microorganisms were subjected to
(3L0 analysis. Only quantitative or no differences
:41 chromatograms, however, were observed between three
strains of E. g9__l_i_.

Holdeman and Brook (1968) have found that GLC
aiialysis of chloroform extracts of culture fluids in
cxbmbination with GLC analysis of prepared derivatives
(bf'extracted compounds can be used in the identifica-
txion of these compounds. They have tentatively identi-
Ified l8 amines and 23 fatty acids frOm various strains

<Jf clostridia by this technique.

14

One of the most recent methods of characterizing
t>acteria by GLC involves the preparation of trimethylsilyl
(fPMS) derivatives of whole-cell hydrolysates (Farshtchi
zind Moss, 1969). These workers observed significant dif-
:ferences in the TMS profiles of Listeria, Neisseria, and
(Zlostridium as well as Clostridium perfringens (two
isypes) and Clostridium sporogenes. Differences in the
‘two types of g, perfringens were believed to reflect
eeither strain variations or differences among serological

(Dr toxigenic types within the same species.

Pyrolysis-Gas-Liquid Chromatography

The use of pyrolysis as an adjunct to gas chromo-
tsography has recently gained considerable importance.
CDhis technique is particularly useful in the polymer
:industry since it provides the rapidity necessary in the
:identification of polymers and polymer structures
(CPerry, 1967). Dal Nogare and Juvet (1966) have
dcescribed thoroughly the major developments which have
téiken place in the use of pyrolysis—gas-liquid chroma-
tcography (PGLC) as an analytical tool and McKinney (196A)
heis compiled a comprehensive outline of the literature
Guancerning the use of this technique between 1960 and
31963. Perry (1964) has also reviewed the techniques
arni potentialities of PGLC.

Advances in pyrolysis techniques have been

aCkxieved through improvements in apparatus design as well

15

greater control of experimental variables. The former
has been aided by the work of Honaker and Horton (1965)
vvho described a simple and inexpensive constant tempera-
ture pyrolyzer and Cox and Ellis (196A) who used a packed
tubular furnace microreactor. The latter workers were
able to identify polymeric materials and evaluate the
effect of changes in reactor geometry and temperature on
the number of pyrolysis products. Stanford (1965)
described the use of glass capillary tubes Open in one
end to hold the material to be pyrolyzed in the pyrolyzer
coil. Further advances in apparatus design were made
by Barbour (1965) who described the use of a valved
pyrolyzer in which the pyrolysis chamber could be isolated
during pyrolysis. This allowed the pyrolysis products
to be instantaneously injected into the gas chromatograph.
Information concerning the structure of organic com-
pounds was obtained by Beroza and Sarmiento (19614) when
they used a carbon skeleton technique to degrade the
compounds. Burr gt gt. (19614) in another attempt to
identify pyrolysis products found that methane and
benzene were the only pyrolytic products of toluene when
hydrogen was used as the carrier gas. Parsons (196A)
was able to identify pyrolytic materials on the basis
of their molecular weight by using a gas density balance

technique in which the density of the carrier gas was

Varied.

16

The use of PGLC as an analytical tool encompasses
many different areas. Those most commonly studied
include analysis of proteins, peptides, and amino acids
(Winter and Albro, 196A; Merrit and Robertson, 1967),
and degradation of such products as porphyrins (Levy
gt gt., 196A), polymeric materials (Feuerberg and
Weigel, 1964) including cellulosic polymers (Groten,
196A), polyolefins (Cieplinski gt gt., 196A; Kolb gt gt.,
1965; Deur-Siftar, 1967), polymers of polystyrene and
acrylic resins (Esposito, 196A; Esposito and Swann,
1965; Spagnolo, 1968), aliphatic alcohols (Dhont, 196A),
dialkyl phosphates and o,o-dialkyl dithiophosphates
(Hanneman and Porter, 196A), and phthalate esters of
tertiary alcohols (Rutherford and Funk, 196A).

Stack (1967) has investigated the possibility of
using PGLC in dental work. Changes in the organic matrix
of enamel of teeth as they decayed could be detected.
Kirk (1967) has also examined the possibility of using
‘this technique in crime laboratories for identification
(Df unknown materials.

Brodasky (1967) observed that both low and high
13emperature PGLC could be used in the differentiation
(bf selected antibibtics. He further observed that
Structural characterization of antibiotics could be

accomplished using this technique.

l7

Reiner (1965) was the first to investigate the
feasibility of using PGLC for the identification of
bacteria. Upon pyrolysis of cells of E. ggtt (18
different antigenic strains), Shigella sp., group A
Streptococcus pyogenes (A types), and mycobacteria (10
different pathogenic and nonpathogenic forms) each strain
produced its own unique pyrogram and different cultures
of the same strain yielded similar profiles. Reiner
(1967) in a further study using additional microorganisms
which were somewhat difficult to classify, reconfirmed
his earlier report that PGLC could be used in the classi-
fication of microorganisms. Reiner and Ewing (1968)
further observed that this technique could be used to
differentiate two cultures of E. ggtt which differed
only in their flagellar, H, antigens. These antigens
'were composed almost entirely of protein, thus, indicat-
.ing that slight protein differences have a strong
inifluence on pyrograms obtained.

Oyama and Carle (1967) have also reported the use
fo PGLC for the identification of bacteria. They con-

errted the pyrograms which they obtained into.a digital
.fcxrm.and analyzed these with a computer. They obtained
1383? graphs from the computer data which enabled them to
(iIITferentiate among 7 fungi grown in malt extract as

Well as A microorganisms grown in trypticase soy broth.

METHODS AND MATERIALS

Microorganisms

 

The microorganisms studied in this experiment

include 2 strains of Clostridium botulinum type A, 2

 

 

strains of E. botulinum type B, and six strains of E.

botulinum type E (5 toxic strains and l non—toxic variant).

 

Type A strains used were 62 and 78 and type B strains

used were 169 and 213.

The type A and B strains were

all obtained from Dr. C. F. Schmidt, Continental Can

Co., Chicago, Illinois.

their sources were:

Strain

‘Vancouver Herring (VH)

Kalamazoo (KAL)

Seattle Forks (SF)

517

A62A7

The strains of type B used and

18

Source

Dr. C. F. Schmidt
Continental Can Company
Chicago, Illinois

Mr. R. W. Johnston
Food and Drug Administration
Detroit, Michigan

Dr. J. T. Graikoski

Bureau of Commercial
Fisheries

Ann Arbor, Michigan

Mr. Donald A. Kautter
Food and Drug Administration
Washington, D.C.

Mr. R. W. Johnston
Food and Drug Administration
Detroit, Michigan

l9

Strain Source
066B nontoxic (NT) Mr. Haim M. Solomon

Food and Drug Administration
Washington, D. C.

Culture Media

Spores and cells of all 10 strains of E. botulinum
investigated were grown in a medium consisting of 5.0%
Trypticase (BBL), 0.5% peptone (Difco), 0.2% sucrose,
1.0% yeast extract (Difco), and 0.1% sodium thioglycolate
adjusted to pH 7.2 (TPSY).

In addition, 5 type E strains were grown in a medium
consisting of 2.5% Multipeptone (Fisher), 0.2% sucrose,
1% nutramino acids (Fisher) and 0.2% sodium thioglycolate
(MSN). One liter Erlenmeyer flasks containing approxi-
mately 1 liter of the growth medium (TPSY or MSN) were
inoculated with 15 m1 of a 16 to 18 hr actively growing
subculture prepared in the same medium of each of the

E. botulinum strains and incubated at 32 C. Duplicate

 

flask cultures were prepared for each strain.

Toxin production and growth were studied by growing
each of the 10 strains of g. botulinum in 1 liter Erlen-
lneyer flasks filled with physiological saline solution
(0.89%) in which a dialysis sac filled with 250 ml of
'TPSY medium (normal concentration) was suspended. The
:ratio of the medium to saline solution was approximately

1:A (v/v). To prevent carry-over of materials from the

20

inoculation medium to the saline solution, the inoculum
(1A ml of 16 to 18 hr actively growing culture) was
centrifuged 10 min at 1000 x g using a Sorvall RC-2
refrigerated centrifuge. The culture was then resuspended
using 5 ml of sterile distilled water and inoculated into
the saline solution. The flasks were again incubated at
32 C and duplicate flask cultures were prepared from

each strain.

Harvesting Cells and Spores

Cultures incubated at 32 C for 8 to 10 hr in TPSY
and MSN media to obtain vegetative cells were then
immediately cooled to A C to prevent sporulation,
dispensed into 250 m1 centrifuge bottles, and centrifuged
for 15 min at 5000 x g using a Sorvall RC-2 centrifuge
refrigerated to A C. The cells were then washed by
centrifuging as above and resuspended in 200 ml of
sterile distilled water until they appeared clean by
microscopic examination (required 5 to 8 washings). After
the final washing the cells were suspended in 100 ml of
sterile distilled water and stored at 2 0.

Spore crops of E. botulinum type E were grown in
both media and harvested after 2A to 30 hr incubation
when spores were free from their sporangia. Longer
incubation of 30 to 36 hr was necessary to obtain adequate

:Sporulation of type A and B strains. It was then

21

necessary to treat these latter cultures with lysozyme
(0.75 mg lysozyme per ml of 0.1N KCl) for A to 6 hr at
32 C to obtain free spores. All washings and storage
procedures of Spores and vegetative cells were the same.
No attempt was made to separate spores and cells
using the dialysis sac technique. All cultures were
incubated for 5 days and harvested in the manner pre-
viously described. Since the type A and type B cultures
consisted of a considerable percentage of vegetative
cells, lysozyme was not used. The dialysate supernatant
fluid (DSF) was separated from the microorganisms by
centrifugation and retained for subsequent analysis.
The DSF was filtered through a Millipore membrane filter
(0.80 u pore size) to remove any organisms not removed

by centrifugation and stored in glass bottles at 2 C.

Direct Microscopic Count

- Direct microscopic counts were made on all culture
samples using a Petroff Hausser counting chamber. One
ml of each culture was added to either 3 ml or 7 ml of
30% glycerol solution depending on the concentration of
the microorganisms. One 100p of the suspension was
placed under the cover slip on the counting chamber and
the microorganisms in 80 small squares were counted.
Total counts were then calculated using the following

equation:

22

6
_ (Number cells counted)(Dilution)(20)(10 )
total count ’ Number small squares counted

'The direct counts gave an indication of the total count

as well as the per cent sporulation. Refractile bodies

‘were considered to be spores.

Toxicity Assays

Swiss Webster white mice weighing 15 to 20 g were
injected intraperitoneally with 0.1, 0.2, and 0.5 m1 of
serial dilutions of each DSF sample. Prior to injection
the samples were digested for 60 min at 37 C with 1%
trypsin (Difco, 1:250) in an equal volume of 0.05 M
sodium phosphate buffer, pH 6.0, and dilutions made
in 0.05 M sodium phosphate buffer. The highest dilution

at which those mice injected died was used to calculate

the minimum lethal dose (MLD).

Controls

Control samples consisted of DSF treated in the
Same manner as other DSF samples with the exception that
tho inoculum was used. A second control consisted of DSF
tc: which was added partially purified toxin of the VH
EFtrain of E. botulinum type E prepared by Mr. Alexander
E3Inodi, Department of Food Science, Michigan State
LThiversity. The partially purified toxin was added to
'tlie uninoculated DSF to give a final toxin titer of 1.0

X: 106 MLD per ml. A lyophilized sample of the same

23

Ipartially purified toxin was also retained for subsequent

analysis.

Lygphilization of Samples
All cultures as well as the DSF and controls were
lyophilized using a Virtis Freeze Mobile. Approximately
5 ml of each culture or sample was added to a 10 ml
freeze drying ampoule. The contents were then shell-
frozen using a dry ice and alcohol bath and lyophilized

for 8 to 10 hr under a vacuum of 0.025 mm Hg.

Preparation of Column Material

Column material used for pyrolysis-gas—liquid
chromatography (PGLC) analysis consisted of a 15% high
‘temperature stabilized ethylene glycol adipate (EGA)
liquid phase (Analabs, Hamden, Conn.) coated on 90 to 100
Inesh Anakrom ABS (Analabs). Fifteen grams of EGA were
carefully weighed into a 600 ml beaker and dissolved in
several hundred ml of acetone. One hundred grams of the
solid support were then added to the liquid phase solu—
‘tion and stirred for several minutes using a magnetic
:Stirrer. The beaker and contents were placed on a
firteam bath to evaporate the solvent. Constant stirring
“filth a glass rod was done throughout the evaporation to
irisure uniform coating, and a slow moving stream of
rliitrogen was blown over the material to help the solvent

esscape as it vaporized. After the material was as dry

2A

515 could be achieved using this technique, it was placed
:in a round bottom flask which was connected to a Rinco
:rotary vacuum apparatus. A vacuum was applied using a
‘water aspirator, and the remaining solvent removed while
applying a slight amount of steam as the flask turned.
The material was then weighed to determine the per-

centage of liquid phase actually obtained.

Packing and Conditioning Columns

Prior to packing the columns, the packing material
was screened through an 80 mesh wire sieve to remove any
large clumps which might cause inefficient packing. The
packing material was added to the tOp of the column
through a funnel. The column was simultaneously
‘vibrated with a vibrator tool and a vacuum drawn on the
bottom of the column to uniformly pack the material in
the column. Caution was taken, however, not to vibrate
the material too severely since this could break some
of the particles. Glass wool was used in each end of
the column after packing to hold the material in place.

Columns were conditioned for 2 weeks at 2A0 C.
Tfliey were further conditioned overnight during use to
fxacilitate removal of pyrolytic materials retained by the

Column.

25

Pyrolysisggpparatus

A Hewlett-Packard Model 80 Pyrolysis Unit (F & M
Scientific Division, Avondale, Pa.) was used to pyrolyze
'the samples. This instrument utilized an automatic
pyrolysis cycle in which there was a 60 sec delay after
activation of the switch, thus allowing the gas chromato-
graph to stabilize. After the delay period the jaw-style
probe element which contained the sample was energized
for 12 sec to bring about pyrolysis at 1200 C. The
probe was fitted to the injection port of the gas chromato-
graph with an adaptor, thus allowing the volatile com—
ponents to be swept directly into the column upon

pyrolysis.

Gas Chromatograph
A Hewlett—Packard Model 5750 B research gas
chromatograph (F & M Scientific Co., Avondale, Pa.)
eequipped with dual—flame ionization detectors and a
Ddoseley 7128 A dual pen strip chart recorder were used
for analysis of pyrolytic products. A special adapter
Ifias added to the injection port of the chromatograph

tc> facilitate use of the pyrolysis unit.

Analysis of Samples
Lyophilized samples were analyzed by pyrolysis gas

chromatography using the following operating parameters:

26

Column: 72 x 3/16" copper refrigeration tubing
Sample Size: 0.9 - 1.1 mg for spores and cells
1.3 - 1.7 mg for DSF and controls
Coating: 15% high temperature stabilized ethylene
glycol adipate (Analabs)
Support: 90 to 100 mesh Anakrom ABS (Analabs)
Carrier Gas: Helium at pressure of 60 psi
Column A--50 ml per min flow rate
Column B--67 ml per min flow rate
Hydrogen: 12 psi
Air: 33 psi
Temperatures: Pyrolysis—-l200 C for 12 sec
Detectors--250 C
Injection Port--230 C
Column-—programmed at 10 0 per min
from 500 to 2A0 C.
Upper Limit Interval: variable
Chart Speed: 0.25 in per min
Sensitivity and Attenuation: 102 x 8 for spores and

cells; variable for DSF and controls.

Intetpretation of Results

 

Results obtained were interpreted by visual examina-
tiion of the pyrograms. The presence or absence of peaks,

'tlae ratio of peak heights with relationship to each other,

27

and retention times were particularly useful. No
quantitation or identification of peaks was attempted

in these preliminary studies.

RESULTS AND DISCUSSION

Effects of Sample Size

 

Initial examination of several of the pyrograms
appeared to reveal considerable differences in the
quantity of material being eluted between strains and
types as well as relative ratios of peak heights for
each strain. These differences appeared to be caused by
an inability to accurately control the amount of sample
being pyrolyzed. The use of the jaw-style pyrolysis
probe made it impossible to weigh the sample more
accurately than to the nearest 0.1 mg.

The consistency with which the sample was packed
into the probe also appeared to influence the elution
pattern. If the sample was packed too firmly some of
the products were not swept instantaneously into the
column upon pyrolysis. However, if samples were too
loosely packed they were often blown from the probe by
the carrier gas when it was inserted into the injection
port.

Some of the pyrolytic products were retained by
the column support material during extended periods of
use of the columns. The size of the sample as well as

the physical and chemical characteristics of the

28

29

pyrolytic products might have contributed to this
problem. In any event, these products contributed to
baseline drift and often gave the false effect that one
peak was proportionally higher than another. This prob-
lem occurred only occasionally in the pyrograms pre-
sented since it was eliminated by allowing the columns
to condition at 2A0 C for a minimum of one hr twice a
day when the columns were being used constantly. In
addition the columsn were conditioned overnight at 2A0 C.
The pyrograms in which sample size and baseline drift
caused variations are discussed in their respective
sections.

Levy (1967) has also indicated the importance of
the effect of sample size and other Operating parameters
on the pyrolysis products. He observed that secondary
reactions between radicals and fragments formed upon
rupture of primary bonds during pyrolysis were affected
by the quantity of material pyrolyzed. He also indicated
that the rate of decomposition of the material pyrolyzed
was dependent upon the film thickness of the sample in
relationship to the geometry of the source of heat.

Janak (1960) observed that pyrolytic products of
Esamples in the microgram range were often independent
Of’the final temperatures of pyrolysis and gave better
reproducibility from one laboratory to another. Ehrler

and Frijouf (1968) also indicated that sample size as

30

well as final temperature, rate of temperature rise, and
residence time in the reactor, greatly influenced
pyrolysis elution patterns.

The final temperature and rate of temperature rise
were readily controlled by the pyrolysis apparatus used
in this study. Residence time, however, was probably
influenced by sample packing while the sample size, as

stated previously, was difficult to accurately control.

Analysis of Vegetative Cells

Pyrograms obtained upon pyrolysis-gas-liquid
chromatography (PGLC) analysis of vegetative cells of
various types and strains of Clostridium botulinum
are shown in Figures 1 (types A and B), 2 (type E in
Trypticase medium), and 3 (type E in Multipeptone medium).
Examination of elution patterns for types A and B
revealed distinct differences in peaks e, h, o, and r.
Strain 62A was distinguishable from strain 78A and
strains of type B by the large peak 0. The type B
strains were not readily differentiated from each other,
however, the larger quantity of material eluted in peaks
e, h and r allowed them to be differentiated from strains
Of‘type A. There also appeared to be some quantitative
(iifferences in peak d, but these failed to be consistent
for any one strain or type. All other differences

Inentioned were present in duplicate as well as replicate

analyses of the same sample.

31

.c01umzcuuum
orb m1 uu< pcm oa>u osu ouocop mcmuum1 _mu1amu och
.Ez1poe uumcuxo ummm>uom0cuamsuccuamquommu1ua>ch

 

:1 czoLm m can < mma>u Esc1_:uon En1p1cumo1o
mo.mc1mcum Lao; so m__mu m>1umHmmo> EoLw mamcmob>m

1352.21 112: 20:21:13

._ 0c3m1m

“Panamanian wonnmuxuoufific v

 

 

 

 

 

 

 

 

 

 

 

A so
11, 7113311111... _ 2.51% 5?..8
1.1111..____1 1 1. 1. .. 1.1
$111.1: 1.1.1...u s.1111_ .11 1 . 1
111.111 :1 11 .1. 111.1... 16.. a
. .1111... 1 1 . 1 1 . .1.
11.1.1.1 1 1 1 . 1 .. as m
.1: 1 .1 new _ <2 . 4.
1 1 1 .8 m
1 . 1 1
. . . .1 1 1.81m
1 11... H
13 \v 3.2.1111.m m4.
1 1 1. .9 ON .0
1 111%: ._11.1.11.._H . 111121151. 1.1.. .1. m
. .1 .1 1. ._ P .1111... 1... 1 1.3 .3.
.: . 1 _. .11 :1 . _
..._11 . 1 1 1 r...
1 . 11%. 1. <3 _
> a c .118
_ 1 1.

 

 

 

32

.co1umacouum ozu m1 uu< .Ez1poE uumcuxo ammo>nomoLuamsuccuauanomm01ua>ck.21 czocm
m was“ E3c1_:uon Ea1p1cumo_u mo mc1mcum x1m mo m__mu o>1umummm> 50cm msmcm0c>m .N mcsm1u

1352.21 1.2: 20:21.51.

 

 

 

assess a 1. assesses 1. assesses. 1. 1. 1,15
T
j 3 m \N/aj . m bar/3‘31”.”
1.11:... 1111.111. 1.1), .111... 2.1111,; 11.111. .8
,2 .>131..1....1;1. _t_ 1.. 1.113113 7.... 1111.1. 1. 1x11111111“... 111 ..,,1
1.72. 1119:1111 1 11.1. 1 1.1.1.11..:_1.___. 1 1., .1111. 111.11.11.11. . 1.11,,
.11 11,, . (a 1 w 1 .__ 1s ,. 1. .. 1. _1 111 11.1 .8. a
1. 1.1....1....1..11..11.1 11.1. 1 .1. .1 111,1... 1.11.111.11_ 1 :11 1 1 11.. .1 .8
1..1 1 1.1 1,1 .. .1 .1 _1 , . u
.1 n1.- 1 K... _r 1 3 1 w
.11. 1 P . .. - _ r8 M
538 .1 1. As 1 . .5.
1. . .1 -8 m
1. . ..
1 1 1AX= mm
b h n p b h h b |_.s\r p p p p P s p m p p h > r A b BIG ‘1“
.117 . 31:11w >5. 9) :11WJ . 5 21¢ >11...“ no
.17/ 1.11. 31111.? 1.11 11,... 1.9/ 1.1.1. 31...... . 11.3.1111... :1}... .2 .1... 1.1.1.5. 1.1 . :11... .8 9
.11 1,. 11.1.5131: _ o .1 11 1.1%..11111511111 .1. T .11.1.,_>.....11.1v1._.:1... a, m
., 1.1.... ,1. :2 .1. .1 .11 .1. 1.1.... .1 .. 1...... _. .11 m.
. 5.x. __.1 E x} _ ,1 h. C O u D 18 a
k 1: 31 _u
'11; 18
as: an _._>
. .8
1 .

 

33

.c01umacmpum 0;“ m1 uu< .Eamume mn_um oc_Eng::amm0102m
succuama1u_:z c1 c301m u oa>u E:c__:uon 531311umo—u
mo mc1m1um 1:01 10 m__mu m>1umummo> 501$ mEM1m01>m .m m1:m_m

 

13521121 92: 2011215
“11.11.111.21 11 11 1. 11111111111111.1111 11 1 11

 

 

 

 

 

 

 

 

. 1h
1 1\<\ah/ figggw v .€.o
5.11 .. .11 1 .1. .111
1 1111... .1 .1 1:11 111. 1
1.12 1111111 1 .511 111 _ 1.1.. a
1 11. _. ._ 3
. 1111111111 11 1. 1 11 1 m
111111.1 1 1 1 11.8m.
1. _1 1&3 1 151 1
1 1 1 1.8 w
b ,_ ¢ 1
12......
. . r . . . . . m . . . . . . . .o m
.1 1111.. 111.11.311.11... .
1 q 1. 1.1. 1
h 1.> 11 I“ . 21? _M m
1111:1111 1. . 11 «JP1111/111111 . 11.9
1.11 11.1 1 1 1
1.1.1 1.11 .1 1 1 1.3
111. 1.. 1m . 11 11> 1
1 , 2..
1 1
1. :81

 

3H

Pyrograms of type E vegetative cells grown in
Trypticase-peptone—sucrose-yeast extract medium (TPSY)
showed no consistent differences among the 6 strains
examined. There were quantitative differences in peaks
k, n, r, and s, but identification on this basis was not
possible. These quantitative differences might have been
caused by variability in the growth rate of the cultures.
This seemed quite possible upon examination of Table 1.
Strains SF and 517 gave higher direct microscopic counts
and higher concentrations of eluted material in peaks k
and 5. Peak n was also different for these two strains
than for the others examined.

No differences were observed in pyrograms obtained
from 5 strains of type E grown in Multipeptone sucrose—
nutramino acids medium (MSN). Upon initial examination
of the pyrograms in Figure 3 and part B of Figure U,
there appeared to be differences in the relative pro-
portion of peaks p and r to peak v. This was apparently
caused by increased drifting in the baseline between 16
and 22 min retention time. The pyrolysis of a possibly
.greater amount of sample with strains SF and A6247 might
liave also added to this effect. These phenomena were
(iiscussed earlier in the section concerning effects of
sample size.

Vegetative cells of types A and B grown in TPSY

Inedium gave definite differences in pyrograms from those

35

J .E:_ume yomguxo ummo>uomoLosm
nocOHaoanumquHa>gh c_ czoLm m oa»u mo m__mu o>_umuomm> Am pcm moLoam Am ucm "Eamvoe
uumuuxo ammo>ummohusmnocouaoauommu_ua>uk c_ czoLm < oa>u mo m__mu m>_umummo> Ag new
mmuoam Au MEDJvoE muJUM oc_EmLu:cummoLu:mnocouamd_u_az c_ czoLm m oa>w mo m~_mo

 

 

 

 

 

 

 

 

 

 

o>_umummm> Am vcm mmuoam A< E:c__:uon E:_U_Lumo_u sou» msmumou>a mo COmJLmQEou .q mLJme
$952.2. m2: ZOJHZwEm .
sausage a .. camouflage a J. osssuaaa «L. so
n m . h
J... J“ J J
_m JJJJ“. s B 5:.
)jJJJJJJJ/EFJ/JJ_ :JJJJJJJJJ JJJJDJJBJJJJJ _JJJJ. a 953.. DJ? :6“
J 2? JJ JJJ. a .J. ”J J .JJJJJ. JJ v.1 3/ \JJJJJJ... J .J.
J... A JJ L. .J J... .JJJ... J,__ F J . . J__J_ _
. JJJJJ JJJ. J .JJJJJ. JJ.J. J J J .JJJJJJJ JJ To.
JJ J, J. J. J JoJ J J . E “m
.. . J .J. . _ . J ,8 m
J._ c J i
3 JJ 0 J m J
J Jlom W
J _J J r09 3
p - L p b p b » .lellp p p p p p h . [WWI r . F p L L p p . o W
J” o J. JJw , €32.71Mv
/ xx 1):“ g) JJJRJ? j gramme 3 J .JJamJqu3v/J JJthm JION w
. a; 1 J 7) 72) _ :J/ . J. w J_ u ._ .5 J’s. “_ngJ J.. mm
J Js. E K : ;J.J moJe J J h a n7).. Js_meJ JJ . J no
JJ .JJJJJJJJ J J JJ J J: J i. JJJ JJJ.JJ JJJJ 9 w.
JJ J . . JJ 0” J J _.. J J J. 4.J J 3 JJ. JrJJ. _ I 3.
JJ JJ J_J.J J J_ JJ ._J.J. J J J JJ J .J. .. J. J .
JJ , J J. J .. J: _J__ J .
JJ .. J .. J J J J. .w . -0.
J. m J JJU J_ J <
J J J. JJ J -8
J J. J J J J .
J J J J Ax:

 

36

'TABLE l.--Direct micrOSCOpic counts of vegetative cells
of Clostridium botulinum types A, B, and E grown in two
different media.

 

 

Total Count (organisms/ml)

 

 

Type Strain

MSN TPSY
A 62 —- 8.3 x 108
A 78 -— 7.3 X108
B 169 -- 5.9 x 108
B 213 -- 1.6 X 108
E VH 2.4 x 108 5.0 x 108
E KAL 1.2 x 108 7.5 x 108
E SF 1.0 x 108 12.0 x 108
E 517 1.1 X l08 15.0 x 108
E A6247 0.6 x 108 7.5 x 108
E O66BNT -- 5.5 X 108

 

of type E grown in the same medium. Parts D and F of
Figure u illustrate the differences in representative
strains of type A and type E. Since pyrograms of type
A and type B were somewhat alike similar differences
exist between types A and B compared to type E. Con-
siderable variations were noted in peaks a through e
while peaks 1 and n were much larger in type A than in
type E. An additional peak which was not present in
type A and type B appeared between peaks r and s in

type E.

37

Reiner (1967) and Reiner and Ewing (1968) observed
differences at the sub—species level in pyrograms of

microorganisms including strains of Escherichia coli

 

which differed only in their flagellar antigens. Their
results, however, were based on years of research and
only slight differences in one or two peaks.

The fact that few differences were found at the
strain level in this study was difficult to explain. It
appeared, however, that the use of the different column
material from that used by the other workers might have
been a possible answer. On the other hand, identifica-

tion of Q. botulinum at the strain level by PGLC may be

 

more difficult than with other organisms previously
tested. A difference in one peak however, was noted in
strain 62A, and there were definite variations at the
type level which made identification possible on this

basis.

Analysis of Spores
Elution patterns obtained upon PGLC analysis of

spores of g. botulinum types A and B grown in TPSY

 

medium are shown in Figure 5. Strain 78A was readily
distinguishable from other strains of type A and type B
examined by the large quantity of material eluted in

peaks g, i, and 5. Once again, however, these differences
appeared to be more quantitative than real since sporula-

tion of strain 78A (Table 2) was much better than

38

.coJumzcmmum mzu mJ uu< ucm oa>p as“ ouocov

mgouuu_ _mHJQmu och .EJJBoE pumguxm ummm>ummoLosm

nocouaoqummmUJua>Lk CJ czoLm m pcm < moa>u E:c__:uon
E:_v_eumo_u wo mCJmLum Lao» mo mogoam EoLw mEmLmoL>m .m oesm_u

 

 

Bub—422$ w<<_._. ZOEZmEm

 

 

 

 

 

 

 

 

 

 

 

 

 

mmmuxuowfifl oxymmmvwowfiaawoo
J J....... . 3 > 77/ >?».M
JJ J.J...JJJJ “.3... J.J:JJ. J, JJJJJJ J F...“
.J. _ _ . J J... J J.J J._
J J. _ J.J J .JJJ... JJ JJ. .J.J .._J .J . JJJ .J
J... J J J .5... .JJ .. ..
.J . J J J J J. J J. m
J. J J L. a J: ..J .100 m
J _J 3 _JJ J.J JJ 0 I.
J. . 3a J Jab... J. 4.
J J J i8 m
c J J J J. S
J. J proo— mu
. . . . . . . . w . . r . . . . mo 3
J... 4.. ud
. 23 JJ... J.J J.J _JJ. 2
JJJJJJJ J JJJMJJJ JJJJJ. :JJJJJJ J.J..JJ.J_J .8 w
9? E J J JJ fivs .meic J J J N
. J J... J. J J.J , J? J. J L: . J.J x
J .J. J. . J.J.H _J J.J . J J J J19.
. J . J. J J J J J
J ._ a L J _JJ J.
QJ J J Jig
J J .J J. J.
.J .: GOO—u J J (NO J.
J g J.ow
JrOQ

 

 

 

39

 

 

 

 

:.w~ moa x m.: I: I: ezmmmo m
m.mb woa x m.m m.m> moa x m.m szo< m
m.:w woa N :.m m.Hm woa x o.m wam m
:.mw moa x m.m m.wm woa x m.: mm m
m.Hm woa x m.m m.mm moa x m.m Adm m
m.mm mOH x m.m m.©> woa x w.m m> m
:.mm woa x m.m II II mam m
w.mw woa x o.m II II mmH m
m.:m moa x m.m u- I- mp ¢
m.:m woa x o.mH II In mm <

Ayv AHE\mEmH:mwhov A&v AHE\mEmHQmwpov

soapmHSLomm pssoo fiance scandazpoqm uGSOU fiance
wmme 2m: samppm maze

 

Esfivmz mhzuaso

 

 

.mHUmE pampmmmfiv 03p 2H czopw m vcw .m .¢ mmamp ESQHHSpon Eudoahpmoao
mo meSpHSQ mo coapmazmoam pamo mma vcm mpcsoo ofiaoomopoae pomnfimll.m mqm<9

MO

sporulation for the other strains of type A and type B.
Peaks w and 2 were often higher in strain 213B than in
other strains, but this was not always consistent. In
general, no differences were consistent and dependable
enough to be used for identification purposes. Even
strain 62A which had shown some differences at the
vegetative cell stage did not produce a significantly
different spore pyrogram.

Figure 6 illustrates the pyrograms for spores of
type E grown in TPSY medium. With the exception of dif-
ferences in the quantity of sample pyrolyzed, all of the
elution patterns appeared to be the same. Strain 517 did
show a smaller concentration of some components, but this
was probably caused by poor growth since Table 2 indicates
that the total count of this strain was considerably less
than counts for the other strains. The differences in
peak 2 also failed to be sufficiently consistent among
duplicate and replicates of the same culture to be used
for identification purposes.

Elution patterns of type E spores grown in MSN
medium provided some rather interesting results (Figure 7).
Strain VH was readily differentiated from all other type
E strains examined by the tremendous size of peak n. In
addition, peak e was slightly smaller in VH and an
additional peak was present between peaks a' and b'. The

additional peak appeared to have been brought about by

ill

.cOJHmacmuum ozu m_ uu< .E:_poe uumLuxo ummo>uomoLusmnocouamanommu_ua>ch
c_ czonm m oa>u E:c__:uon E:_p_uumo_u mo mCJmLum x_m mo mmuoam 50L» mEmLmoL>m .o oLstm

. JmmSZJJJs m2: 20:sz
sausages«sfisaaaaawsssaaau J. V

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o
- 0
JW . m j J. JW
u m J) a
J . J J JJJJJJJJJJJJJJ... J.J... .J.
P J/ j a JJJ J J J ...J_..J J x JJ waJ
. J J... .J_ JJ
J _. J a.PJJJJ a NJ...J__ > JJJ. J.J. .JJJ 0. -8
J. JJ :J J J J J P “M
JJ J. J J.J. J. J J .38 m.
JJ J J. . J J . .N.
.J J 233 J b... J :2 .
J... J J.J J .. 28 w...
J J 2 J J 1
_ J J r.oo_Jm
. . . . . . . - H . J _ . . . . w . . J . . . . no a
J m h/ J .p eJJ>JW. a
J J. JJJJ JJn/JJJJJJJJJJJJ . J.J.JJJJJJJ J.J?JJJJJ J. J.JJJJJJJJJJJ JJJJJJJJJJJ 8 v...
J. J J .J .. JPJJJ. __._J.J_J _JwJJJJJJ > J. J.J..JJJJJ. .nJJJJJJ
JJ JJJJJJJ/JJJJ fiJ J J.J... J J. JJJJ.. J J..... JJJJJJJJE 1:: J.J.JJJ.
JJ J.J..JJ J.J . J.._ J RJJJJJJJJJJJJJJE J JJ J. _JJJ J.J.JJJJJJJJJJ J.J: J J J.J. .9. x
J .J J. .
J. J.J JJ J J J. J JJJ J.JJ J J .J_ J J J.J .. J .18
._ _J. - J J. J. J J J. J. J
.J JJ .J S2 . J J .1 _J. . a J J J J J J. J; J J 8
. . .J a J J nJJ J.J o J I
m J J J. ,. J... J
J ._ J J J J J
J J. _c J J J 78.

 

H2

.c0wumscouum as“ m_ uu< .E:_uoe mn_um
o:_EmguacnomoLuzmuocOuamaJp.32 cJ czoLm m oa>u E:c__:uon
EszJLumo_u mo chmLum L:o* mo mogoam EoLw mEmLmoL>m .n 0L:m_m

 

 

ambazzs mash Zfi¥¢fithm
fiasuomaau «capmfiaaea J.

0
JM W 0
J}. v J
J J... J Jerou
.J_JJ.JJ 3....J J J.J J.
J. a. .. JD? J.J. .2 JJ J.J 5?
J. J. .KuJJI LJJJ H JJ. OJ .
. . JJJ roe
.3. _J

 

 

J NVNO<

 

BSNOJSE-‘J ETVDS Th'H lNEDEBd

l

 

 

 

 

 

“3

differences in the retention time of peaks eluted after

2“ min. Strain A62M7 first appeared to be different from
the other type E strains grown in MSN, but closer investi-
gation revealed that this was caused by the pyrolysis of
a smaller amount of sample for this strain. Variations

in peaks g and h were initially thought to be caused by
differences in growth and sporulation. Examination of
the data in Table 2, however, failed to indicate any
correlation between these peaks and growth or sporulation.
None of the differences, with the exception of those in
strain VH, were useful for identification purposes. No
attempt was made to determine why strain VH gave a
characteristic elution pattern.

A comparison of pyrograms for spores of representa-
tive cultures of type A and type E grown in TPSY medium
is provided in parts C and E of Figure 4. Distinct and
consistent differences occurred in peaks g, h, i, n, r, s,
x, and 2. These peaks permitted easy differentiation of
“type A spores and thus type B spores from spores of type B.

No studies have previously been undertaken to dif—
1ferentiate strains and types of cultures by pyrolysis of
‘their spores. Differences in elution patterns, however,
‘were expected since these spores were formed from the
same vegetative cells in which differences were observed.
Once again major differences were found at the type level
while only a few differences were noted among strains of

the same type.

uu
Comparison of Spores and
Vegetative Cells

Pyrograms of spores and vegetative cells of
representative strains of Q. botulinum types A and E
are shown in Figure A. Consistent differences were
observed between spores and cells of each type (type B
being similar to type A) of g. botulinum investigated.
Differences which were common among all types included
those in peaks g, n, and 2. Types A and B showed addi-
tional differences in peaks a through d and peaks w, x,
and y. Vegetative cells of all types also had a larger
peak between peaks r and s.

Since spores have been shown to be physiologically
different from vegetative cells (Murrell, 1967), dif-
ferences in their pyrograms were expected. The presence
of such materials as calcium and dipicolinic acid in
spores as well as lysis of the sporangia and formation
of new coat layers in spores, made them completely
different bodies than the cells from which they were
formed.

The presence of spores in cultures may strongly
influence the pyrograms obtained and thus prove to be a
limiting factor in the use of pyrolysis for the identi-
fication of those microorganisms which form spores.
With the exception of strain 78A, however, the sporula-
tion data in Table 2 failed to indicate any correlation

between per cent sporulation and elution patterns. A

U5

closer quantitative examination of those peaks related
to sporulation might have revealed some correlation in
this respect. Peaks g, n, and 2 might have been par-
ticularly useful in this respect.

Since lysis and other physiological changes besides
sporulation occurred in the cultures as they grew older,
apparent differences in spores and cells might also be
partially attributed to the age of the culture. The data
collected in this experiment failed to either support or
contradict this theory since spores and cells of each

strain were harvested at only one time.

Effect of Medium

 

The type of medium in which the cultures were grown
was found to have a pronounced effect on the pyrograms
obtained for both vegetative cells and spores of Q.

botulinum type B. Since types A and B would not grow

 

in MSN medium, only type B strains were used for this
study.

Differences in elution patterns of vegetative
cells grown in MSN and TPSY media are shown in parts B
and F of Figure 4. Distinct quantitative differences,
always characteristic for their particular medium, were
noted in peaks 0, k, m, n, q, s, and 2. Peak 1 was also
absent in the MSN pyrograms while there was a larger
additional peak between peaks r and s in the TPSY pro-

grams.

A6

Parts A and E of Figure A illustrate the differences
in elution patterns for spores grown in MSN and TPSY media
respectively. Several characteristic differences were
noted in the concentrations of peaks a, e, g, n, r, w, and
2. Peak x was absent in spores grown in TPSY and peak 2
was much larger in spores from TPSY.

Oyama and Carle (1967) have also indicated that the
growth medium had an influence on the pyrograms which they
obtained when they grew Candida pulcherrima in malt extract
and in trypticase soy broth. They further observed that
organisms grown in similar media gave similar pyrograms.

The two media used in this experiment showed defi-
nite similarities as well as definite differences in pyro-
grams for the same culture. Since both media are similar
in that they are rich in nitrogen, this might account for
many of the similarities. The differences were probably
caused by the presence of yeast extract as well as varying
concentrations of other nitrogen—rich materials.

Detection of Growth and Toxin
Production

 

An attempt was made to detect growth and toxin
production of various types and strains of g. botulinum
using a modification of the dialysis sac technique
described by Vinet and Fredette (1951). Since the
majority of the toxin should have been present in the

supernatant fluid after growth and centrifugation of the

.u7

cultures, it was proposed that this toxin might be
detectable by PGLC. The dialysis sac technique was used
to limit and control the availability of nutrients to the
cultures, thus preventing possible interference of these
compounds with the pyrograms subsequently obtained. This
was discussed in the previous section of the Results and
Discussion concerning the effect of the medium on pyro-
grams.

No retention value with respect to molecular size
has been provided for dialysis tubing, therefore, it was
not possible to accurately predict what nutrients were
available to the growing cultures except that they had to
be less than approximately 30,000 M.W. Since the nutrients
were limited, some of the cultures failed to grow and
sporulate as readily as others (Table 3). Also, many of
the vegetative cells lysed because of the prolonged period
of incubation for 5 days. Microorganisms which were
harvested were thus not defined as either spores of
cells. In fact, the per cent sporulation data in Table
3 is questionable since many of the vegetative cells had
lysed and thus the data actually represents apparent
sporulation at the end of incubation.

The pyrograms obtained upon PGLC analysis of these
organisms are shown in Figures 8 and 9. They illustrate
the same type of differences as those found in regular

TPSY and MSN media. Strains of type A and type B were

l48

.co_um:couum mg“ m_ uu< cam oa>u
oz» ou0cou mubuub. _mu_qmu ozk .oso_czomp omm m_m>_m_n

 

po_w_noE ozu mc_m: czoLm m pcm <.moa>u E:c__:uon

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E:_pmcumo_u mo mc_mLum L30m mo moL:u_:o Eoum mEMcmoL>m .m oL:m_u
$52.; 92: zozzmma
«Mmmfipmaamwofimmfipmfiflmwoo
w . m
m m
_._ . 3.23.9 .5 . 7. , A
it? 3.. la “3“ a:—_._._. ti)?“ :3 3? A) r8
# . u __ y .27 _,.._ .
2 a . _l sis 7.. :. .A.
b 2 (if D _m 9.” 2.3.9: . . __ «i
g} _ __ x u...” i w _ f wow
i A x . z _i 2, E m _ _ ._. u...
_ K: 3.. _ hi ,. .i M. : J. r . m
m i. i _ a a .1 . z .1. is m
M l. a 3n n .2 _. __.. I. é:
_ m l ; i. _i is w
. a l... i S “a _..
_ W a n.oo~ m"
. . . . . . . . .. . - - r . _ r . . o w
Q5 . : _Mm a...» 91 £¢pimvg m
i I; c ”a. “cg :L .4 _. o l. m
# _ _ a. l 31 JP . r o a: To
.5 . l .: i f ”m .9.
. a c, j g . w
c. _ __ L a... s _
h a : i .8
_. .. 32 W . . _ <3
L W a t . .18
__ .8.

“9

.co_um:couum ozu m_ uu< .m:U_ccoou omm m_m>_m_p no_w_poE ozu
mc_m: czocm m ma>u E::__:uon E:_n_Lumo_u mo mc_mcum x_m mo mmL:u_:o EoLm mEmLmoL>m .m oc:m_m

 

3352.2. m2: 20:23.3

 

 

 

 

 

 

 

 

 

 

 

an on New». A. m w onnmmvw owfi m.— m w mum mwvwowfi & m w mo
m m T
>>. ._m. .._, 7. .uu 3%.).3.”
if. 3.1.9339... .........2: 7... p... .._ ..._. _. J .. EH...) 7.... _. .. ... a. .8
a... . _. .. . .... . m 2...... . _._ .__. .
:/ .._:_. . . ._. .. .52; J. .. a}... :5:,>......_.:.... . ..._ ..__ ._
. 9...? .. . :3... . .._ .._ .. ..._._....._....... .. .. . .9.
4 j. E s . .. _..._..._. E .. .. _V .. _ .1. .: . u...
._ fi_. .. . . ..: 3.. . a 1 L . . . ww
__. ‘ F _ _ f in. .. f L y ._ . . . _ x00 1:
. . . , . . .. .. .. . . .. ... _ _.. .N.
R 2000 .. 1 a _. .m Eh m. m .20. a _
.0 _. c F. m . .. . fiOw W
.0 ..- u .l
w _ w _roo_mm
. p . . . . . . w . . . . . . .. x. .W . + Ll . . . . . w o “M
s 1 3M E. . T %%.m .32 7312......“ mm
_ 5. ~ 1; m% x. a. _ W3 Eg\&fw.c : Ea..0m Mm
1.... ._ H. .- .... .. L _.. ._._. ..
.. .. ._ . ._ 3. e . . .. 43.23.? . .._ . _. m
u f u w .. _ 3x .3 0.. __ ..._.__...»L _._.__ __ 0 10V
5. a L . ~ ‘ . _._ : .. f .. a .
\ __ _ __. _ wk 1.. t f .__...M.. M... a
._ .. . _ , _. . 100
_ . ( fl _ _ _. . m ..
. _ .
.... . .. . #84. T “a H. M. . _ :> M
r L . ___ . . .8
.. . .... .
I r # .._ H. :8

 

 

50

TABLE 3.--Direct microscopic counts and per cent sporula-
tion of strains of Clostridium botulinum types A, B, and
E using the modified dialysis sac technique.

 

 

 

Type Strain Direct Count % Sporulation
A 62 8.u x 107 91.7
A 78 6.8 x 108 82.u
B 169 6.9 x 107 58.3
B 213 3.9 x 108 62.u
E VH 1.5 x 108 8M.8
E KAL u.6 x 108 80.8
E SF 2.8 x 108 69.8
E 517 1.u x 108 75.u
E A6247 8.3 x 107 90.0
E 066BNT 1.u x 108 92.u

 

slightly different in peaks g and u. Strain 62A produced
an additional characteristic peak between peaks m and n.
Elution patterns for strains of type E were identical
with the exception of slight differences in peaks e and m.
Because the cultures were grown under such "nonideal"
conditions, these differences were not reliable enough
to be used for identification purposes.

Elution patterns obtained by PGLC analysis of the
dialysate supernatant fluid (DSF) of cultures of g.

botulinum are illustrated in Figure 10. The DSF of strain

51

.co_um:couum ocu m_ uu< cam ma>u on“ vaccmo

mumuuo_ _mu_amu .mao_c;oou omm m_m>_m_p co_w_poe

mzu EoLw mumm>_m_n ocu EoLm m vcm < moa>u Esc__:uon

E:_p_cumo_u mo mc_mcum Lao» mo co_umm:m_cucmo Loumm
mc_c_mEmL p_:_w ucmumcLoaam mumm>_m_p msu mo mEmcmoL>a .o_ ocsm_m

 

 

352.2. ms: zozzmma

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

O
. . . .. . no
3.. _v u 1
._..... Z .22.. . _.,... 3 M...
, .2... .5... 3.. .. 4 .___. ._ ... .2. s...
162.... .... 3... f1. .37.... .H...__..... E. .._.. .2.
_ _. .. __ ._ _r 1 1. f. .13.... p; a
z . ff .th. Wu J m “T" ....f_” :u .
k f. :._:_ _ .n ... ._. 5: _._ We. a
5. :H. . . . .. . 4...... . m
_m. m .fi. Lrg.g. v.90 mu
. . Mr.a V w LW avg.m . n
y .a_ can 3 r . ....4&k r J
: t; a. a v" "7...... _. _ 18 m
H. .n ...._ . m
. _E r. c. _ .. ..
a .5 . ..... E ._. :69 m
p b b r p r P 5L b p p L F)“ P h “To a
1m Em ajDJ 22$.) 3m wm
... .h .. :. a. s u .9. mb
rm .. .3Eé a W ./ (.23; ; _.._..__._...___ «E...» a. 5m W
._ a é. .: .. .. ..
u _. _ _ l. I _L. h 1...... ‘ _... 1:
f. ._.._ __ W W. ’ .2 z. . Sow
. .. a. W .. . : .z. .
1.4M ._ u M... ..
.. _. .._ .w . _._ ... . .8
_fi .6 moo. @ g Q: _(Nc “
. ._ _ _. f L” . 18
_ .— r .H. w. .
.. . .. . :8.

 

52

78A was distinguished from that of other strains of type
A and type B by the large peak k. Once again this dif-
ference might be quantitative since strain 78A was observed
to grow better than the other strains (Table 3).

Differences in peak 0 of type A DSF and type B DSF
were consistent and could not be correlated with growth
rate. Type A could thus be distinguished from type B
by this peak.‘ Variation in peaks m and u, however, were
not consistent enough for identification purposes.

Pyrograms from the DSF of type E strains are shown
in Figure 11. They-failed to indicate any differences in
elution patterns which were consistent and significant
enough for identification purposes of strains within the
serotype E. The quantitative differences in peaks c,
k, and m were probably caused by variability in the total
amount of growth as well as lysis of many of the vegeta-
tive cells in some cultures.

The differences in pyrograms for the DSF of types
A and B from those of type E DSF were quite distinct.
Parts D, E and F of Figure 12 illustrate the differences
in the DSF elution patterns of types E, A, and B
respectively. These differences were probably caused by
physiological differences between the three types. Since
type A and type B were proteolytic, they produced and
broke down different compounds than type E which was

nonproteolytic and primarily saccharolytic with respect

53

.co_um:couum ecu m_ uu< .ozc_c;oou omm m_m>_m_n nm_m_coe
ecu sob» mumm>_m_p ozu 50cm m ma>u E:c__:uon E:_u_cumo_o mo mc2mcum x_m mo
co_umm:w_uucmo Loumm m:_c_mEoc v_:_m ucmumccmdsm oumm>_m_v ozu mo msmcmoL>d .__ oL:m_m

 

23.52.22 52: zozzmmm

mafioalaam vofifi899.m vommvmomasm vo
_KI.

 

 

 

 

 

 

\u. ..... m S«1,22, m _... 2m 2....
35.. 222.22,... 2222. (22.2%...» _.2_. 2.22222 {2. .. 2...,an _..2 222...
WEB“. 222.722 26.2 I _.2 2.2222... 2 2.2 x 3.231. .. .22. .2. 2 ION
2. 2 o. 222 2r. .- 2 x: 2. 2 2. 22 >3... _... .23..
.2. p .. 22 2.2... 22.2.2. . 2 22 2.2.2.2. a: ... 2. 2.2... 2.22... __
22 _..2 2.2.... a... e 2 22.9... 2_ 2 2 2...... ._._...._. 2 .9.
22.2.2 2. _..... . .. 2. .82....
.2... .22 2. 2.. . 2 2.2.2 _. w
. g. . .— . . 100 no
.. _ a 2 3
538 2 E. . .5. . 2 m
.. _ -Om .d
. _ m
2 2. . _l
. . . . . . . . .8. x
. . p . 22:23". . 22ml. . L3 .7”. 22>)..Wl .%: .vw .222Mro W
.. . ... ..,. ... .1. _..... a
Jig... .222... 2 2.22. r232... 2.2.2.2222... J2.2....2. .22.... . 2...... .8 a.
. .2 2. :2 .2... . . 2 2 2... _..... .. ._ .22... if... m
.2 2... 2.2.3. .2. .. _.2 .22.. 2 .. ... 2.... 2.... . w.
2 a. 22 2. 2 a 2 2. .. .__.. ..o.2 a. .ov 3
E g. 2 E . . 2 .. 2.. .
. . _. . w...
5.34 2 a I...» ___
2 . A5
2 2
2.. 2
. 2 .. .8—

514

 

.E:c__:aon E:_n_gumo_u m ma>u mo mmo Am ucm
”Eac__:uon E:_u_gumo_u < ca>u mo umo Au ME:c__:pon E:_n_Lumo_u m ma>u
mo umo An "c2x0u to_m_gsa >__m_u2ma m0 m5 0.. Au “c2x0u no_$_2:a
>__m_ugma m:_a mmc nmum_=oo:_c: Am ”Ammov v_:_m “cmumcgmasm mumm>_m_v
tmum_:uo:_:: A< mo mc_“m_mcou mm_QEmm Lo» mEmLmoL>a mo c0m2LmQEOQ

 

2852.22 92: 20:25..

.N_.mL:m_m

 

manafiau... «Buchanan... «ofinueau...
fl.

 

 

 

 

 

 

 

 

m .V .fl 2“ W. » WTO
. 1.29.2 3.22.2. 2.222m _.. .2272. 22222..“ S}... .2...
.72 2:.. 22...2. 52 _ 2 p 3 .22 .2{.222 .. 2 .r 2. \2 62.22.0u
.22.... .12.... ...._.... . . 2 2.22... 2 2 322.22.. .. _. .2222...
2 . 2.. .22.... 22... 2 . 2. .. 2...... _... .2 2 _ 2. 2.... ..
_. .2. 2.2.2.. .._ .. 2 . _.....2. . .. . 2 2 .12 -3
.. 2.. 2...... . 2 .. .._.. .. . . . y.
2 2.2 .22... . .. 22 ._ m
m 1.2.22... .2 _ .. . I
. 2 .2222 2 g . . co _m
2.2 .2... 22 n— .. 22 D _.. m .. n:
2.. . 2 .
22’ 3.2;: 2.8 2.. . 2
. .2 2 . . :8. m
. P . ~ . . . p . r . p p . . h » . . 0 11.4.
u 1 u 2
.. .2... . 2.. Egg... 2... .
.v 2 2 _.B . _.z . 2a
25. 22222 ?_:m2r..2 2 ‘22 2xwaf. 2.12 2222.0“ 2%
.2. .2 2.2;... . .. 2 22.2.... N
.2 , _. 2.2 2. _... . _. .21; S
22 2 2 22 222* 22 . _.. 22 . a 3
. . .. :2 .. . _..... {9.
2 2: _. :.. 2 . 22 .
. _. 2 .22 22 .. .2. .
2. _ _. _..2 . 22 _... . .-
. 2 222 . . s
. 2. 22 2 m 2 U < 2
2 2.... _... .. . .8
r . 2.2 2 2..
2 1...... .. 2.. _.
c 2. 2 r.09

 

 

55

to physiology. Types A and B have also been shown to be
markedly less saccharolytic than type E. These physio~
logical differences were described by Dolman (l96u).
Several types of control samples were analyzed to
evaluate the feasibility of using PGLC to detect growth
and toxin production. The elution pattern of the first
control which consisted of an uninoculated sample of the
DSF is shown in part A of Figure 12. The pyrogram was
identical to the one for the DSF of type E (part D,
Figure 12) with the exception that peaks k and m were
considerably smaller in the control. The patterns of
type A and type B were of course quite different. This

indicated that growth of g. botulinum could be detected in

 

a dialyzed medium.
The feasibility of toxin detection by PGLC was
determined by the addition of a partially purified toxin

of Q. botulinum type E strain VH to uninoculated DSF

 

giving a final toxin titer about equal to that of the
actual cultures (Table 4). A 1.0 mg sample of the
toxin was also subjected to PGLC analysis.. These
results are shown in parts B and C of Figure 12.
Essentially all of the pyrolytic products of the
toxin were eluted in small peaks between 0 and 15 min
retention time. Since the amount of toxin actually
present under normal conditions would only be a fraction

of that used for this analysis, it appeared that detection

56

of the actual toxin in the growth medium was impossible.
The toxin peaks were hidden by peaks from other pyrolytic

products.

TABLE N.—-Toxin titers of various strains of Clostridium
botulinum types A, B, and E grown in duplicate in
trypticase—peptone—sucrose-yeast extract medium.

 

 

#1__

Toxin Titer (MLD/ml)

 

 

Type Strain l 2
A 62 2.0 x 105 2.0 x 105
A 78 u.0 x 10” 4.0 x 10“
B 169 1.0 x 102 u.o x 102
B 213 4.0 x 102 u.0 x 102
E VH 2.0 x 10“ 1.0 x 105
E KAL 1.0 x 105 2.0 x 105
E SE 2.0 x 103 2.0 x 103
E 517 1.0 x 103 1.0 x 103
E A62u7 1.0 x 105 u.0 x 10“
E O66BNT not toxic not toxic

 

The fact that some of the peaks in the DSF plus
toxin control were different than those in the uninoculated
DSF control was probably caused by the nature as well as
the amount of toxin which was added. Only those peaks

in the first 11 min of retention time were changed,

57

further indicating that this was the area in which toxin
was eluted but not readily detected under these experi-

mental conditions.

SUMMARY AND CONCLUSIONS

Experiments were conducted to determine the feasibility
of using pyrolysis-gas-liquid chromatography (PGLC) for
the identification of various types and strains of

Clostridium botulinum. The use of this technique for

 

detecting both growth and toxin production by the same
microorganisms was also investigated.

Vegetative cells (10 hr incubation at 32 C) and spores
(36 hr incubation at 32 C) of 10 strains representing types

A, B, and E g. botulinum were grown in Trypticase-peptone-

 

sucrose—yeast extract medium (TPSY) and the resulting cul-
tures analyzed by PGLC. Vegetative cells and spores of
five strains of type E were also harvested from
Multipeptone—sucrose-nutramino acids medium (MSN) and
analyzed in the same manner as those grown in TPSY.
Examination of pyrograms for vegetative cells
revealed differences in elution patterns which were of
sufficient significance to permit differentiation at the
type level. Pyrograms of types A and B indicated only
slight but significant differences from each other. The
pyrograms of type A and B, however, were both quite dif-
ferent from those of type E strains. The only consistent

difference at the strain level occurred in strain 62 of

58

59

type A. It was readily differentiated from other strains
by the size of one of its peaks. Several other slight
differences occurred in elution patterns of pyrograms

of type E vegetative cells_grown in both media, but these
differences appeared to be caused by changes in the amount
of sample pyrolyzed or variations in growth and sporulation
of the cultures. No differences suitable for identifica-
tion purposes were observed in pyrograms of type E strains
grown in either TPSY or MSN media.

Only differences at the type level were detected
among pyrograms of the three types of spores grown in TPSY.
Other variations were believed to be related to the size
of the sample pyrolyzed or differences in growth and per-
centage of sporulation. In any event, these differences
at the strain level were not sufficiently reliable for
identification purposes.

Spores of type E grown in MSN did show consistent
differences at the strain as well as the type level.
Strain VH was easily differentiated from the other strains
of type E.

Pyrograms of spores were different than those of
vegetative cells for all strains studied. This may prove
to be a limiting factor in the use of PGLC for the
identification of microorganisms which form spores, since
the pyrogram appeared to be dependent on the presence or

absence of spores. The age of the culture might also be

 

_——_ ._ _., _.

60

related to these differences in spores and vegetative
cells.

The effect of the growth medium on the elution
patterns of pyrograms from 5 strains of type E was also
observed. Significant differences between pyrograms
from type E strains grown in MSN and TPSY media were
easily detected in each of the strains examined. This
indicated that the growth medium may be a limiting
factor in the use of this technique for identification

of Q. botulinum as well as other microorganisms.

 

A modification of the dialysis sac technique
described by Vinet and Fredette (1951) was used to deter-
mine the feasibility of detecting growth and toxin pro-
duction by PGLC.

The inoculum was added outside the dialysis sac and
nutrients were provided by dialysis across the membrane
from inside the sac. Pyrograms of the cultures harvested
from the growth medium after 5 days incubation at 32 C
revealed only the same type of differences in elution
patterns as those observed for the subcultures grown in
normal flask cultures in MSN and TPSY media. The result-
ing pyrograms were difficult to evaluate since the
unusual conditions under which the cultures were grown
made it practically impossible to obtain a synchronous
relationship between cultures with respect to growth and

sporulation. Lysis of vegetative cells caused by

 

 

 

61

prolonged incubation of cultures might also have con-
tributed to the variability. Reliable differences in
pyrograms once again occurred at the type level but
not at the strain level for the modified dialysis sac
grown cultures.

Comparison of pyrograms obtained from samples of
the dialysate supernatant fluid (DSF) with those obtained
from three control samples that consisted of (1) uninocu-
lated DSF, (2) uninoculated DSF plus partially purified
toxin, and (3) a 1.0 mg sample of partially purified
toxin revealed that PGLC could be used to detect growth

of Q. botulinum in this dialyzed medium. It was even

 

possible to determine which type of g. botulinum had been

 

grown in the medium. Detection of toxin production,
however, was impossible since the toxin was present in
such small quantities with respect to other compounds
that its pyrolytic products were hidden among peaks from
other pyrolytic products of the culture medium.

In general, PGLC appeared to show promise as an

analytical tool for identification of g. botulinum at the

 

type level. The investigation of other column materials
and other Operating parameters, however, is necessary

to determine the usefulness of this technique at the
strain level since this study revealed only a few

reliable differences. Pyrolysis-gas-liquid chromatography

may also be extremely valuable for detection of growth of

 

 

62

g. botulinum. Further studies must be conducted, however,

 

to determine whether growth in actual food products can

be detected.

LITERATURE CITED

63

LITERATURE CITED

Abel, K., H. deSchmertzing, and J. I. Peterson. 1963.
Classification of microorganisms by analysis of
chemical composition. I. Feasibility of utilizing
gas chromatography. J. Bacteriol. §2:1039-1044.

Angelotti, R., H. Hall, M. Foter, and K. Lewis. 1962.
Quantitation of Clostridium perfringens in foods.
Appl. Microbiol. 19:193—199.

Anonymous. 1964. Botulism outbreak from smoked white-
fish. Food Technol. l§:71-74.

Barbour, w. M. 1965. Refinement of pyrolysis tech-
niques for gas chromatography. J. Gas. Chromatog.
iz228-231.

Bassette, R. and T. J. Claydon. 1965. Characterization
of some bacteria by gas chromatographic analysis of
head-space vapors from milk. J. Dairy Sci. 4§:775.

Batty, I. and P. D. Walker. 1963. The differentiation
of Clostridium septicum and Clostridium chauvoei
by the use of fluorescent labeled antibodies. J.
Pathol. Bact. §§:517—521.

 

Bawdon, R. E. and R. Bassette. 1966. Differentiation
of Eschericia coli and Aerobacter aerogenes by
gas liquid chromatography. J. Dairy Sci. 49:
624-627. ""

 

Beroza, M. and R. Sarmiento. 1964. Carbon skeleton
chromatography using hot-wire thermal-conductivity
detection. Anal. Chem. iézl744-1750.

Boothroyd, M. and D. L. Georgala. 1964. Immunofluore—
scent detection of Clostridium botulinum. Nature.
aggz515.

Brodasky, T. F. 1967. Pyrolysis-gas chromatography in
the differentiation and characterization of anti-
biotics. J. Gas Chromatog. §:311-3l8.

64

65

Brown, J. P. and B. J. Cosenza. 1964. Fatty acid com-
position of lipids extracted from three spherical
bacteria. Nature. 204:802.

Bulatova, T. I. and Y. A. Kabanova. 1960. Identification
of botulism pathogen with luminescent sera. J.
Microbiol. i1:403-408.

Burr, J. G., R. A. Meyer, and J. D. Strong. 1964. The
hydrOgen carrier technique for the pyrolysis of
toluene. J. Am. Chem. Soc. Qéz3846-3850.

Cattaneo, C., M. Lucchesi, G. C. DeRitis, C. Pietropaolo,
P. Rossi and G. Frisani. 1964. Gas Chromatographic
separation of fatty acids (Cl2-20) present in
mycobacterial lipids. Ann. lst "Carlo Forlanini"

231: 353-361.

Cecchini, G. L. and R. T. O'Brien. 1968. Detection of
Escherichia coli by gas chromatography. J. Bacteriol.

 

Cieplinski, E. W., L. S. Ettre, B. Kolb, and G. Kemmner.
1964. Pyrolysis-gas chromatography with linearly
programmed temperature packed and open tubular
columns. The thermal degradation of polyolefins.
Z. Anal. Chem. 225:357-371.

Cox, B. C. and B. Ellis. 1964. A microreactor-gas
chromatographic method for the identification of
polymeric materials. Anal. Chem. 16:90-96.

Dal Nogare, S. and R. S. Juvet, Jr. 1966. Gas Chromato—
graphy. Anal. Chem. ;§:72R.

Deur—Siftar, D. 1967. Application of pyrolysis—gas
chromatography for characterization of polyolefins.
J. Gas Chromatog. 5:72-76.

Dhont, J. N. 1964. Identification of aliphatic
alcohols by pyrolysis. Analyst. ng7l-74.

Dolman, C. E. 1964. Growth and metabolic activities
of g. botulinum types. 32 K. H. Lewis and K. C.
Cassel, Jr. (eds), Botulism. Proc. of Symp.,

U. S. Department of Health, Education, and '
Welfare, Cincinnati, Ohio, pp. 43-68.

 

Dolman, C. E. and H. Chang. 1952. Clostridium botulinum
type E. Can. J. Publ. Hlth. [43:38.

66

Dolman, C. E. and H. Iida. 1963. Type E botulism: its
epidemiology, prevention, and specific treatment.
Can. J. Public Health 54:293—308.

Dolman, C. E. and L. Murakami. 1961. Clostridium
botulinum type F with recent observations on other
types. J. Infect. Dis. 109:107-128.

 

 

Duda, J. J. and J. M. Slack. 1969. Toxin production in
Clostridium botulinum as demonstrated by electron
microscopy. J. Bacteriol. 91:900-904.

 

Eales, C. E. and A. W. Turner. 1952. Description of
91. botulinum type D recovered from soil in South
Australia. Aust. J. Exp. Biol. Med. Sci. 10:
295-300.

Ehrler, A. J. and J. A. Frijouf. 1968. Repeatibility
in pyrolysis chromatography. Digest of paper
presented at the Pittsburgh Conf. of Anal. Chem.
and Appl. Spectroscopy. Cleveland, Ohio (1968).

 

Esposito, G. G. 1964. Quantitative pyrolytic gas
chromatography by internal standard. Anal. Chem.
36:2183—2185.

Esposito, G. G. and M. H. Swann. 1965. Application of
pyrolysis and programmed temperature gas chroma—
tography to the analysis of thermo-setting acrylic
coating resins. J. Gas Chromatog. ;:282-284.

Farshtchi, D. and C. W. Moss. 1969. Characterization
of bacteria by gas chromatography: comparison of
trimethylsilyl derivatives of whole-cell
hydrolysates. Appl. Microbiol. 11:262-267.

Feuerberg, H. and H. Weigel. 1964. Ein einfaches
verfahren zur identifizierung organischen
polymerer durch pyrolyse und gaschromatographie.
Z. Anal. Chem. 199:121-160.

Garner, W. and R. M. Gennaro. 1965. Related bacteria
differentiated by gas chromatography. Chem. Eng.
News. 4;:69.

Georgala, D. L. and M. Boothroyd. 1967. Fluorescent
antibody techniques applied to Clostridium botulinum
types A, B, and E. lE.M- Ingram and T. A. Roberts
(eds). Botulism 1966. Proc. Fifth Intern. Symp. on
Food Microbiol. Chapman and Hall, Ltd., London,
pp. 494-502.

67

Gibbs, B. M. and B. Freame. 1965. Methods for the
recovery of Clostridia from foods. J. Appl.
Bacteriol. 2§:95-111.

 

Groten, B. 1964. Application of pyrolysis-gas chromato-
graphy to polymer characterization. Anal. Chem.
36:1206-1212.

Gunnison, J. B., J. R. Cummings, and K. F. Meyer. 1936.
Clostridium botulinum type E. Proc. Soc. Exp. Biol.
Med. 35:278—283.

 

Gunnison, J. and K. F. Meyer. 1929. Cultural study of
an international collection of Clostridium
botulinum and parabotulinum. XXXVIII. J. Infect.
Dis. 45:119-134.

 

 

 

Hanneman, W. W. and R. S. Porter. 1964. The thermal
decomposition of dialkyl phosphates and 0,0-dia1ky1
dithiophosphates. J. Org. Chem. 2252996-2998.

Hazen, E. L. 1937. A strain of B. botulinus not
classified as A, B, or C. J. Infect. Dis. 60:
260-264.

Henis, Y., J. R. Gould, and M. Alexander. 1966. Detection
and identification of bacteria by gas chromatography.
Appl. Microbiol. 145513—524.

Hobbs, G. and A. W. Anderson. 1967. Protein differences
in strains of Clostridium botulinum based on
separation by disc electrophoresis. In. M. Ingram
and T. A. Roberts (eds), Botulism 1966: Proc.

Fifth Intern. Symp of Food’MicrobioI. Chapman and
Hall Ltd, London, pp. 417-423.

 

Hobbs, G., A. Stiebrs, and M. W. Eklund. 1967. Egg yolk
reaction of Clostridium botulinum type E in dif-
ferent basal media. J. Bacteriol. 93:1192.

 

 

Holdeman, L. V. and J. B. Brooks. 1968. Variation among
strains of Clostridium botulinum and related clostridia.
Presentedau;UJNR Conference on Toxic Microorganisms, '
Botulism Conference. Honolulu, Ha.

 

Honaker, C. B. and A. D. Horton. 1965. A simple
pyrolyzer for use with gas chromatography. J.
Gas Chromatog. 3:396-398.

Iida, H., Y. Nakamura, I. Nakagawa, and T. Karashimada.
1958. Additional type E outbreaks in Hokkaido,

68

James, A. T. and A. J. P. Martin. 1952. Gas-liquid
partition chromatography: the separation and micro-
estimation of volatile fatty acids from formic
acid to dodecanoic acid. Biochem. J. 50:679-690.

Janak, J. 1960. Identification of the structures of non-
volatile organic substances by gas chromatography
of pyrolytic products. Nature 185:684-686.

Johnson, H. M., K. Brenner, R. Angelotti, and H. E. Hall.
1966. Serological studies on types A, B, and E
botulinal toxins by passive hemagglutination and
bentonite flocculation. J. Bacteriol. 91:967-974.

Kalitina, T. A. 1960. The detection of Clostridium
botulinum by means of luminescent antibodies.
Communication 1: the production of specific
luminescence in Clostridium botulinum by treatment
with luminescent immune serum. Bull. Biol. Med.
Exp. 49:278-280.

 

 

 

Kautter, D. A. 1968. The detection, identification,
and isolation of Clostridium botulinum. Presented
at UJNR Conference on Toxic Microorganisms,
Botulism Conference, Honolulu, Ha.

 

Kautter, D. A. and M. Bartram. 1967. Problems associated
with the isolation and identification of Clostridium
botulinum type E from heavily contaminated
materials. In M. Ingram and T. A. Roberts (eds),
Botulism 19667 Proc. Fifth Intern. Symp. on
Food Microbiol. Chapman and Hall Ltd., London,
pp. 448-454.

 

 

 

Kirk, P. L. 1967. Identification by means of pyrolysis
products. J. Gas Chromatog. 5:11-14.

Kolb, B., G. Kemmner, K. H. Kaiser, E. W. Cieplinski and
L. S. Ettre. 1965. Pyrolysis—gas chromatography
with linearly programmed temperature packed and open
tubular columns. The thermal degradation of
polyolefins. Z. Anal. Chem. 292:302-312.

Levy, R. L. 1967. Trends and advances in design of
pyrolysis units for gas chromatography. J. Gas
Chromatog. 5:107-113.

Levy, R. L., H. Gesser, E. A. Halevi, and S. Saidman.
1964. Pyrolysis gas chromatography of porphyrins.
J. Gas Chromatog. 2:254-255.

69

Lewis, V. J., R. E. Weaver, and D. G. Hollis. 1968.
Fatty acid composition of Neisseria species as
determined by gas chromatography. J. Bacteriol.
2i: 1’5 0

 

MacGee, J. 1968. Characterization of mammalian tissues
and microorganisms by gas-liquid chromatography.
J. Gas. Chromatog. 6:48-52.

McClung, L. S. 1967. A review of enrichment and isolation
methods. In M. Ingram and T. A. Roberts (eds),
Botulism 1966. Proc. Fifth Intern. Symp. on
Food Microbiol., Chapman and Hall Ltd., London,
pp. 431-442.

 

McClung. L. S. and R. Toabe. 1947. The egg yold plate
reaction for the presumptive diagnosis of Clostridium
sporogenes and certain species of the gangrene and
botulinum groups. J. Bacteriol. 53:139—147.

 

 

McKinney, R. W. 1964. Pyrolysis-gas chromatography: a
bibliography (1960—1963). J. Gas Chromatog. 2:

Merritt, Jr, C. and D. H. Robertson. 1967. The analysis
of proteins, peptides, and amino acids by pyrolysis-
gas chromatography and mass spectrometry. J. Gas
Chromatog. 5:96-98.

Meyer, K. F. and J. F. Gunnison. 1928. Clostridium
botulinum type D Sp. N. Proc. Soc. Exp. Biol.

 

 

Midura, T., Y. Inouye, and H. L. Bodily. 1967. Use of
immunofluorescence to identify Clostridium
botulinum types A, B, and E. Public Health Rept.
(U.S.)'§2:275-279.

Midura, T., C. Taclindo Jr., G. S. Nygaard, H. L. Bodily
and R. M. Wood. 1968. Use of immunofluorescence
and animal tests to detect growth and toxin pro-
duction by Clostridium botulinum type E in food.
Appl. Microbiol. 1§3102—105.

 

 

 

Mitruka, B. M. and M. Alexander. 1967. Ultrasensitive
detection of certain microbial metabolites by gas
chromatography. Anal. Biochem. 29:548-550.

7O

Mitruka, B. M. and M. Alexander. 1968. Rapid and sensi-
tive detection of bacteria by gas chromatography.
Appl. Microbiol. 16:636—640.

Mitruka, B. M. and M. Alexander. 1969. Cometabolism and
gas chromatography for the sensitive detection of
bacteria. Appl. Microbiol. 11:551-555.

Moore, W. E. C., E. P. Cato, and L. V. Holdeman. 1966.
Fermentation patterns of some Clostridium species.
Intern. J. System. Bacteriol. 16:383-415.

 

Moss, C. W., V. R. Dowell, Jr., D. Farshtchi, L. J.
Raines, and W. B. Cherry. 1969. Cultural
characteristics and fatty acid composition of
propionibacteria. J. Bacteriol. 91:561-570.

Moss, C. W., V. R. Dowell, Jr., V. J. Lewis, and M. A.
Schekter. 1967. Cultural characteristics and fatty
acid composition of Corynebacterium acnes. J.
Bacteriol. 94;1300-1305.

Moss, C. W. and V. J. Lewis. 1967. Characterization of
clostridia by gas chromatography. I. Differentia-
tion of species by cellular fatty acids. Appl.
Microbiol. 15:390-397.

Mossel, D. A. A. 1959. Enumeration of sulphite-
reducing clostridia occurring in foods. J. Sci.
Fd. Agric. 19:662-669.

Murrell, W. G. 1967. The biochemistry of the bacterial
endospore. In A. H. Rose and J. F. Wilkinson
(eds), Advances in Microbial Physiology, Academic
Press, London, pp. 133-244.

 

O'Brien, R. T. 1967. Differentiation of bacteria by
gas chromatographic analysis of products of
glucose catabolism. Food Technol. 21:1130-1132.

Oyama, V. I. and G. C. Carle. 1967. Pyrolysis gas
chromatography application to life detection and
chemotaxonomy. J. Gas Chromatog. 5:151—154.

Parsons, J. S. 1964. Bracket method for molecular
weight determination of pyrolysis products using
gas chromatography with a gas density detector.
Anal. Chem. 3651849-1852.

Perry, S. J. 1964. Techniques and potentialities of
pyrolysis-gas chromatography. J. Gas Chromatog.
$54-59.

 

71

Perry, S. G. 1967. An industrial view of pyrolysis- GC.
J. Gas Chromatog. 5:77—80.

Reiner, E. 1965. Identification of bacterial strains
by pyrolysis-gas liquid chromatography. Nature
206:1272-1274.

Reiner, E. 1967. Studies on differentiation of micro-
organisms by pyrolysis-gas-liquid-chromatography.
J. Gas Chromatog. 2:65-67.

Reiner, E. and W. H. Ewing. 1968. Chemotaxonomic
studies of some gram negative bacteria by means of
pyrolysis-gas-liquid chromatography. Nature.
gllgl9l-l94.

Robinson, E. M. 1930. The bacteria of the Clostridium
botulinum C and D types. Union of South Africa,
Department of Agriculture, Report of the Director
of Veterinarian Service and Animal Industry, 16:97.

 

 

Rutherford, K. G. and D. P. C. Funk. 1964. Low
temperature pyrolysis. I. The pyrolysis of
hydrogen phthalate esters of some tertiary
alcohols. Can. J. Chem. 42:2657-2664.

Rycaj, T. 1956. The testing of food for botulinic type
A toxin with the aid of the specific haemagglutina-
tion method. Bull. Acad. Polon. Sci. Classe II.
4:317-320.

Schoenholz, P. 1928. Surface colony types of C1.
botulinum on blood agar. J. Infect. Dis._§2z
HO-h? a

 

Seddon, H. R. 1922. The specific identity of B.
parabotulinus. J. Comp. Path. Ther. 35:275-280.

 

Sinitsyn, V. A. 1961. The use of the indirect hemag-
glutination reaction for detection of botulinus
toxin. Voenno-Med. Zh. 19:65-68.

Spagnolo, F. 1968. Detection and estimation of poly-
styrene in acrylic plastics by pyrolysis-gas
chromatography. J. Gas Chromatog. 6:609-610.

Stack, M. V. 1967. Quantitative resolution of protein
pyrolyzates by gas chromatography. J. Gas Chromatog.

72

Stanford, F. G. 1965. An improved method of pyrolysis
in gas chromatography. Analyst 99:266-269.

Steinhauer, J. E., R. L. Flentge, and R. V. Lechowich.
1967. Lipid patterns of selected microorganisms as
determined by gas-liquid chromatography. Appl.
Microbiol. 12:826-829.

Townsend, C. T. 1929. Comparative study of nontoxic
and toxic strains of Clostridium parabotulinum.
XXXV. J. Infect. Dis. 45:87-95.

Vermilyea, B. L., H. W. Walker, and J. C. Ayres. 1968.
Detection of botulinal toxins by immunodiffusion.
Appl. Microbiol. 16:21-24.

Vinet, G. and V. Fredette. 1951. Apparatus for the
culture of bacteria in cellophane tubes. Science.
114:549—550.

Walker, P. D. and I. Batty. 1964. Fluorescent studies
in the genus Clostridium. II. A rapid method for
differentiating Clostridium botulinum types A, B,
and F, types C, and D, and type E. J. Appl.
Bacteriol. 21:140-142.

 

Walker, P. D. and I. Batty. 1967. The serology of
Clostridium botulinum with reference to fluorescent
staining. InIM.IIngram and T. A. Roberts (eds),
Botulism 1966, Proc. Fifth Intern. Symp. on Food
Microbiol. Chapman and Hall Ltd., London, pp. 482--

493.

Ward, B. Q., B. J. Carroll, and E. S. Garrett. 1967.
Presence of Clostridium botulinum type E in
estuarine waters of the United States Gulf Coast
comparing fluorescent detection and mouse-toxin
methods. In M. Ingram and T. A. Roberts (eds),
Botulism 1966. Proc. Fifth Intern. Symp. on
Food Microbiol., Chapman and Hall, Ltd., London,

pp. 503-5070

Willis, A. T. and G. Hobbs. 1958. Medium for the
identification of clostridia producing opalescence
in egg-yolk emulsions. J. Pathol. Bacteriol.

12:299-305.

Willis, A. T. and G. Hobbs. 1959. Some new media for
the isolation and identification of clostridia. J.
Pathol. Bacteriol. 11:511-521.

 

 

 

 

73

Wilson, W. J. and E. M. Blair. 1924. The application
of sulphite-glucose-iron-agar medium to the
quantitative estimation of B. welchii and other
reducing bacteria in water supplies. J. Pathol.
Bacteriol. Bl:119-121.

Winter, L. N. and P. W. Albro. 1964. Differentiation
of amino acids by gas-liquid chromatography of
their pyrolysis products. J. Gas Chromatog. B:1-6.

Yafayer, R. K. and S. A. Chepeler. 1961. The use of the
indirect haemagglutination test for the detection of
Clostridium botulinum toxin in sausages. Zh.
Mikrobiol. Epidemiol. i Immunobiol. BBz21-25.

 

Yamakawa, T. and N. Ueta. 1964. Gas chromatographic
studies of microbial components. I. Carbohydrate
and fatty acid constitution of Neisseria.

Japan J. Exptl. Med. 14:361-374.