cammmmmwc wmmmw ma MGLECULAE mmmggam OF mg; a gqgmsgzgfi £91m: mm: mm: ‘E‘hm {‘09 $0 Dogma :2? 3h. 0.. WCBIMN STAKE L‘NWEESET Aiexancier Emodi E969 L [B A? A R Y .' -1. ° hlxu. I Q“ n S {‘3 EC Umvc-rsi ty . Mg." -.o m This is to certify that the thesis entitled ”Chromatographic Purification and Molecular Dimensions of Type E Clostridium botulinum Tox in" presented ,hg Alexander Emodi has been accepted towards fulfillment of the requirements for -1 /" ' .7 f2 t/‘i'C—K‘6 Vida/3L- /C ichard V. Lechowich Major professor Date July 10, 1969 0-169 i M - at”... ._.,' q5,, Q ,‘ 1- ¢ $W,1M'.usx.x:m$iq :' . ‘ «8'0 W95! “flair“??? .11 W. 7“W Pi,“ . no. ’ , ram am '21:.- x W“ a». w m.:u~.qdn «rennin-b - m ‘ - .30 s» m u mach‘ue £1111:va . -- A... fin: “M ant-taro om 153mm u: «4»: an" '02 fl 4?. no Mina toxin-m W 3"" Mung the heater“! and}: If wanna“. at _‘ ;.JILf18<3 on aux oellulelo am. at who remain an» I 1- “4.; -tatien nu: Main out!“ a. mvusmuou ad, ' f: I: , . .. ..: ' s. .J'. 1 '1, ".331? x _ 'J u: o . 1 2m: {WALC amid on (unto-mud on new sac-s ::“’ .2 my ez-m ante with m “we mefii -T‘-~-~~1‘ . "‘ .v '-- ' - . "iv-3' " v I i . f iii}? ' f ’g‘k‘“ ; 1;} ,9)... I. ' - . 4'“ it; Min/fol}. " it. .1 “mid: h} new!” mu m_ _ .- _ . _ warm fraction «glut. 31-1: M ' " ._ i ‘ , .9; _ - ' - - : ‘ ~ " -. “I ‘r -. '1 I .--‘1P'I‘T ‘ ' '1'; -l t t ‘ O-.‘§ - '..>- ' 3‘57“, < '1‘.\\-<‘.k't 7““:- o_ . '1 ‘ ' - C ABSTRACT CHROMATOGRAPHIC PURIFICATION AND MOLECULAR DIMENSIONS OF TYPE E ELOSTRIDIUM BOTULINUM TOXIN BY Alexander Emodi The objective of this study was to clarify some of the discrepancies relative to the homogeneity, aggregation, and molecular weight of type E botulism toxins. The Clostridium botulinum type B Vancouver Herring strain used for toxin production was grown in a physiolog- ical saline solution in which a dialysis sac filled with culture medium was suspended. Cultures were incubated for seven days at 30 C. The extracellular toxin was separated by removing the bacterial cells by centrifugation, and purified on DEAE cellulose colums at pH 6.0 following pre- cipitation with ammonium sulfate at 60% saturation, and dialysis. The toxic material was fractionated on DEAE cellu- lose columns into toxic and non-toxic components repre— sented by separate peaks. Disc electrophoresis of the toxic fraction on pH 4.3 gels showed three bands, two of Alexander Emodi which were not toxic. Passage of the toxic fraction through a Sephadex G-200 column resulted in two fractions with equal toxicity. Both fractions were eluted after cy- tochrome c, therefore the molecular weights were less than 12,200. Final purification of the toxic fraction was car- ried out on CM-Sephadex C—50. The toxic fraction was re- tarded by the gel and released when 0.5 M NaCl was applied to the column. Disc electrophoresis of the toxic fraction showed only one band indicating that the purified toxin was homogeneous with respect to the number of components pres- ent. The toxic fraction was eluted in one peak, after cy— tochrome c, on a calibrated Sephadex G-200 column. There- fore, the estimated molecular weight of the purified toxin was less than 12,200. A test for hemagglutinating activity of the toxic fractions gave negative results, while nonpurified type A and B toxins used as controls were positive. A possible aggregation effect of toxin with ribo- nuclease was observed when toxin purified on DEAE cellu- lose was treated with the enzyme. The preparation was separated into three peak fractions on a calibrated Sephadex G-200 column. An additional highly toxic com- pound of large molecular weight (approximately 240,000) was obtained in addition to the two small molecular weight fractions. However, the large molecular compound was un- stable and disappeared after 3 to 4 days of storage at 4 C. Alexander Emodi The adsorption or coupling of the toxin to blue dextran, to red blood cells and to bacterial cells was ob- served when purified toxin mixed with those large molecular weight compounds on Sephadex G-200 columns eluted with them in the void volume. A possible aggregation phenomenon of the toxic pro- tein was apparent at pH 7.0 and pH 9.1. Toxic material purified on DEAE cellulose columns eluted on Sephadex G-200 columns throughout the entire fractionation range of the G-200 gel. The toxin probably existed in the forms of ag- gregates of different sizes. CHROMATOGRAPHIC PURIFICATION AND MOLECULAR DIMENSIONS OF TYPE E CLOSTRIDIUM BOTULINUM TOXIN BY Alexander Emodi A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Food Science 1969 ACKNOWLEDGEMENTS The author wishes to express his appreciation to his major professor Dr. R. V. Lechowich, for his continued interest and guidance throughout this study; to Dr. J. R. Brunner for his suggestions and for the use of his labora- tory facilities; to H. A. Lillevik, L. G. Harmon and C. M. Stine for serving as members of the guidance committee; and to Margaret Dynnik for her valuable help in the labor- atory. To his wife, Magda, the author is especially grateful for her encouragement during the course of his study. Appreciation is extended to Dr. B. S. Schweigert, Chairman, Department of Food Science, for his interest in this program, to Michigan State University for the facili- ties which were provided and to the United States Public Health Service to make this study possible. ii Chapter TABLE OF CONTENTS INTRODUCTION . . . . . . . LITERATURE REVIEW . . . . Toxin Isolation and Purification Toxin Potency, Homogeneity, and Molecular Size . . . Composition and Active- TOXin O O O O O O 0 MATERIALS AND METHODS . . Microorganism . . Culture medium . . Toxicity assays . Analytical methods Purification procedures Site of Treatment of dialysis sacs Determination of sedimentation coefficients . . Concentration of toxin Disc electrophoresis . Test for hemagglutinating activity Estimation of molecular size Preparation of buffers RESULTS AND DISCUSSION . . Toxin Production . . . the Toxin Isolation and Purification Homogeneity of the Toxin . Estimation of Size . . S WRY O O O O O O O O 0 LITERATURE CITED . . . . . . . APPENDIX iii Page 12 17 17 17 18 18 19 22 23 23 23 24 25 26 27 27 30 43 53 65 70 77 LIST OF FIGURES Figure Page 1. Schematic diagram of the purification procedure of type E Clostridium botulinum toxin . . . . . . . . . . . . . 20 2. Elution patterns of Type E VH toxin on DEAE cellulose with 0.05 M sodium phosphate buffer, ph 6.0. The major component contains the toxin . . . . . . 33 3. Elution of 2 ml of type E Kalamazoo toxin on DEAE cellulose with 0.05 M sodium acetate buffer, pH 4.5. The first three components contain the majority of the toxicity . . . . . . . . . . . . . 35 4. Elution of 2 m1 of type E Kalamazoo toxin on DEAE cellulose with 0.05 M sodium phosphate buffer, pH 6.0. The third component contains the toxin . . . . . . 36 5. Elution of 1.5 ml of DEAE purified type E VH toxin on CM-Sephadex C-50. The peak eluting after NaCl application contains the toxic component . . . . . . 39 6. Ultracentrifuge photographs of DEAE pur- ified type E VH toxin at 59780 rpm, 0 min and 32 min after reaching full speed, in 0.05 M sodium phosphate buffer, pH 6.0 . . . . . . . . . . . . . 44 7. Replicate disc electrophoresis patterns of purified type E VH toxin on pH 4.3, 7.0 % acrylamide gel using pH 5.0 B alanine-acetic acid buffer and a cur- rent of 3 mAmp with the cathode at the bottom. The top band contains the toxin . . . . . . . . . . . . . . . . 46 iv Figure 10. ll. 12. Table 1. Disc electrophoresis patterns of type E VH toxin, B. before and A. after CM-Sephadex purification on pH 4.3, 7.0 % acrylamide gel with pH 5.0 B alanine-acetic acid buffer and a current of 3 mAmp . . . . . . . . . Elution of DEAE purified type E VH toxin on Sephadex G-200 column with 0.05 M sodium phOSphate buffer at pH 6.0 . . . . . . . . . . . . . . Molecular weights of toxic fractions as determined on a Sephadex G-200 column (1 x 28 cm). Proteins were eluted with 0.05 M sodium phos- phate buffer pH 6.0. The molecular weights of proteins are stated within the parentheses . . . . . . Elution of type E VH toxin mixed with blue dextran on Sephadex G-200 column with 0.05 M Sodium phos- phate buffer, pH 6.0 . . . . . . . Disc electrophoresis patterns of the toxic and non-toxic fractions eluted from DEAE cellulose after RNase treatment A. Toxic fraction B. Non-toxic fraction . . . . . . LIST OF TABLES Toxin production of the VH strain in "dialyzed" TPSY medium . . . . . . . Page 50 52 55 59 62 Page 29 INTRODUCTION Botulism has been known as a disease for over a century, but its association with food poisoning was not established until Van Ermengem described and isolated the causative organism in the 1890's. Since then a large number of cases have been recorded and many publications have been written on the subject throughout the world, in- dicating the high level of interest the disease has aroused. Most of the reported cases implicated underproces- sed home-prepared foods as the causative agents and Gov- ernment agencies continue to emphasize the importance of safe thermal processes in food preservation. The safe processing of commercial products is also a concern, since the presence of toxin in these products could affect very large numbers of consumers. Human botulism is caused by the microorganisms of Clostridium botulinum types A, B, or E, however, it should be noted that the other types C, D, and F are also involved on rare occasions. The microorganisms are Gram-positive, spore forming rods, motile by means of peritrichous flagella. They produce powerful neurotoxins which upon ingestion or injec- tion may cause death. Death usually results from respira- tory paralysis, since the toxins act on the acetylcholine release mechanism. The full toxicity of type E toxin be- comes available only after treatment with trypsin. Type A and B toxins also possess hemagglutinating properties which can be retained without retention of tox- icity. The hemagglutinating activity of type E toxin has not been demonstrated and relatively few data are available for type E toxin. Numerous attempts have been made to purify these toxins and to determine their chemical composition or molecular structures. The purification procedures yielded products of reported high and uniform potency with widely different molecular weights, and questionable homogeneity. Purification of the toxin formed by type E 9123- tridium botulinum has been aarried out by several tech- niques. Gordon 23 El: (1957) precipitated the toxin with ethanol and used ionized cellulose columns for purifica- tion. Gerwing 23 a1. (1964) used ammonium sulfate to pre— cipitate the toxin, further purified the concentrated di- alysate on DEAE cellulose columns, and calculated a molecu- lar weight of 18,600 (320W = 1.708). Sakaguchi and Sakaguchi (1967) digested the "precursor" toxin first with ribonuclease and then used CM-Sephadex and Sephadex G-200 columns for purification and to estimate molecular weight. Their results, in contrast to those of Gerwing 33 31., indicated a uniform molecular weight of 200,000 or more. Due to the above discrepancies, attempts were made to clarify some of the questions relative to the homogene- ity, aggregation, and molecular weight of the type B toxins. LITERATURE REVIEW Toxin Isolation and Purification The isolation and purification of Clostridium botu- linum toxins have involved a variety of techniques such as precipitation with acid, shaking with chloroform, precipi- tation with ethanol, high speed centrifugation, extraction with salt solutions, precipitation and crystallization from salt solutions,-and recently chromatography on cellu- lose compounds. One of the earliest isolation procedures, which was developed by Brieger and Kempner (1987), used zinc chloride to precipitate type B §3botulinum toxin. Sommer, Sommer and Meyer (1926) described a gentler method in which type A toxin was adsorbed to aluminum hydroxide and then eluted with ammonium phosphate. Snipe and Sommer (1928) used an acid precipitation procedure for concentrat- ing the toxin, and thus provided a basis from which puri- fication procedures have proceeded. None of these early methods produced highly purified toxins. The potency of these products was little more than one-thousandth of the toxicity reported by Lamanna, Eklund, and McElroy (1946), and Lamanna, McElroy and Eklund (1946) who isolated type A toxin in crystalline form with a specific activity of 2.2 x 108 minimum lethal dose (MLD) per mg N. These workers used a procedure of precipitation with acid, followed by shaking with chloroform and final crystallization from an ammonium sulfate solution. Abrams, Kegeles and Hottle (1946) also isolated type A toxin in crystalline form by precipitating the toxin with sodium sulfate instead of shaking with chloroform. Duff gt 21. (1957) reported an improved method for isolating type A toxin that used acid precipitation at pH 3.5 with 3 N sulfuric acid, extraction of the toxin in 0.075 M calcium chloride at pH 6.5, a second acid precipitation at pH 3.7 with 1 N hydrochloric acid, a subsequent precipitation with 15 % alcohol at -5 C, and final crystallization from 0.9 M ammonium sulfate solution. Cardella gt El. (1958) precipitated type C toxin from the culture with 25 % ethanol, extracted the toxin from this precipitate with 0.05 M calcium chloride at pH 5.0, and reprecipitated the toxin with 15 % ethanol at -5 C. Cardella e£_al. (1960) also purified type D toxin by slightly modifying the procedure used for type C toxin. The type D toxin after the initial ethanol precipitation was extracted with 0.075 M calcium chloride at pH 6.5 and reprecipitated with 10 % ethanol at -5 C. Gordon et a1. (1957) purified crude type E toxins by precipitating with 25 % ethanol in the cold, then ex- tracted the toxin from the precipitate with 0.075 M cal- cium chloride solution at pH 6.0, and reprecipitated with 25 % ethanol at -5 C. Fiock, Yarinsky and Duff (1961) described the production, purification and conversion to toxoid of type B toxin treated with trypsin. They treated the culture with 0.1 % trypsin (pH 6.0) for 2 hr at 37 C. Purification was accomplished by precipitation with ammo- nium sulfate at 60 % saturation, extraction with calcium chloride, and reprecipitation with ethanol in the cold. The purified toxin was then converted to toxoid by incuba- tion with formalin adsorbed on aluminum phosphate. Ger- wing, Dolman and Arnott (1961) prepared highly purified type E toxin by modifying the method described by Gordon §E_§l. (1957). They added 95 % ethanol to toxic filtrates to produce a final concentration of 35 % ethanol. After centrifugation the precipitate was resuspended in 0.05 M sodium acetate buffer (pH 6.0) and further purified on DEAE cellulose columns. Gerwing 33 21. (1964) further modified this purification procedure by using ammonium sulfate as an initial precipitating agent instead of ethanol. Sakaguchi and Sakaguchi (1961) and Sakaguchi, Sakaguchi and Imai (1964) extracted type E toxin from bac- terial cells with l M acetate buffer at pH 6.0. The toxin was precipitated from the extract with ammonium sulfate to obtain 50 % saturation, centrifuged, and the precipitate redissolved in 0.05 M acetate buffer. This precipitation procedure was repeated three times, and was followed by chromatography on Sephadex G-25, digestion, with ribonu- clease and further purification on CM - Sephadex C-50. Toxin Potency, Homogeneity, and Molecular Size The purification procedures described yielded products of high and uniform potency. The specific activ- ity for type A toxin ranged from 2.2 x 108 to 2.7 x 108 MLD per mg N. For type B toxin the range of specific activities was from 1.25 x 108 to 2.6 x 108 MLD per mg N. The data for type E toxin are less constant. Sakaguchi and Sakaguchi (1961) reported a potency of 1.7 x 105 MLD per mg N, while Gerwing, Dolman and Arnott (1961, 1962), and Gerwing EE.E£° (1964) obtained a potency of 6.0 x 105 and 7.5 x 106 MLD per mg N, when ammonium sulfate was sub- stituted for ethanol as a precipitating agent. However, a considerable variation exists in the reported molecular weights. Lamanna,-Eklund and McElroy (1946), Lamanna, McElroy and Eklund (1946) and Putnam, Lamanna and Sharp (1946) determined a molecular weight of 900,000 to 1,000,000 for type A toxin. On the basis of cysteine and cystine contents the minimum molecular weight of crystal- line toxin was calculated as 45,000 (Buehler, Schantz, and Lamanna, 1947). The high molecular weight toxin concept was supported by Duff et_§l, (1957) who obtained a value of 320W = 14.5,_implying that the molecular weight is in- deed very large. Gerwing, Dolman and Bains (1965) reported the iso- lation of a toxic moiety of low molecular weight for type A. The toxic substance was shown to be homogeneous using electrophoresis and ultracentrifugation and a calculated molecular weight of 12,200 was obtained. The homogeneity of type A toxin obtained in cry- stalline form by Lamanna, McElroy and Eklund (1946) re- mained unquestioned for 20 years. However, observations made by Lamanna (1948) showed that the crystalline toxin also contained a hemagglutinin which could be dissociated from the toxin with a two to three-fold increase in spe- cific activity (Lamanna and Lowenthal, 1951; Lowenthal and Lamanna, 1951). Wagman and Bateman (1951, 1953) and Wagman (1954) later found the molecular size and shape of purified crystalline type A toxin was dependent upon the pH and salt concentration of aqueous buffers used as solvents. They isolated toxic fragments formed at alkaline pH that had a molecular weight about 70,000. Heckley, Hildebrand and Lamanna (1960) showed that toxin appearing in the lymph of orally poisoned rats had a significantly lower sedimentation constant than that ob- tained with undissociated crystalline toxin, which had a molecular weight of 900,000. Similar results were reported on the toxin appearing in the plasma and lymph of rabbits injected intravenously by Hildebrand, Lamanna and Heckly (1961). Schantz and Lauffer (1962) studied the diffusion rate of type A toxin and estimated the molecular weight between 10,000 and 20,000.. Wagman (1963) found that puri— fied type A toxin irreversibly dissociated at pH 9.2 with little loss in toxicity and formed a principal and a minor component with sedimentation constants of 7 S and 13.7 S. No reaggregation of the 7 S toxin was observed upon read- justment of the pH to 3.8. Therefore, he suggested that these products are components of large aggregates. The diffusibility of significant quantities of these toxic fragments through cellulose dialysing membranes was ob- served in several experiments but not in others. Boroff and DasGupta (1964, 1966), DasGupta, Boroff and Rothstein (1966), and DasGupta and Boroff (1968) ques- tioned the homogeneity of the crystalline toxin and they demonstrated that chromatographically the toxin can be re- solved into a toxic (a) and into a hemagglutinin (B) frac— tion. The toxic fraction had a molecular weight of 150,000 at the physiological pH range of 7.0 to 7.4. The hemag- glutinin appeared to exist in three forms of aggregation (81, 82 and 83) with molecular weights of 81 = 290,000, 82 = 500,000 and 83 = 900,000. Boroff 23.31. (1966) 10 analyzed crystalline type A toxin in the ultracentrifuge for homogeneity and the a fraction was analyzed for molecu- lar weight. The crystalline whole toxin resolved into two components with S values of approximately 7 S and 13.5 S. The a fraction was homogeneous and a molecular weight of 128,000 was determined. A molecular weight of 60,000 was reported by Lamanna and Glassman (1947) for type D toxin using membrane diffusion techniques. Duff gt a1, (1957) found purified B toxin was essentially homogeneous when examined in the analytical ultracentrifuge. The toxin consisted of a main component of 14.9 S and a diffusely sedimenting minor com— ponent with an S value of 10.9. These results are similar to those reported by Wagman and Bateman (1951). They de- termined by sedimentation and diffusion measurements that the molecular weight of the major component in a similar preparation of type B toxin was about 500,000. Gerwing 23' El. (1966) showed that purified type B toxin was mono- phoretic and monodisperse in the ultracentrifuge and on the basis of the biophysical studies and amino acid anal- ysis a molecular weight between 9,000 and 10,000 was cal- culated. However, the homogeneity of the purified B toxin was seriously questioned by Boroff, DasGupta and Fleck (1968). They examined the toxin and separated it into two components, a toxic and a non-toxic component. The 11 molecular weight of the toxic component was 100,000 or greater and resembled the component isolated from the cry- stalline toxin of type A. Recently Gerwing, Morrell and Nitz (1968) demonstrated that the toxin was synthesized in the later stage of logarithmic growth and was released in- to the supernatant fluid of the culture during lysis of the cells. Sedimentation studies of the intracellular toxin showed one fraction with low molecular weight (S = 2.5) and low specific activity, and one fraction with high molecular weight (S = 15.3) and with specific activity similar to that isolated from culture lysates. The high molecular weight toxin was composed of an aggregate of small sub-units. The presents of two toxic units was also shown in type A toxin by Hauschild and Hilsheimer (1968). They found two equally toxic units with molecular weights of 400,000 to 450,000 and 150,000 to 200,000. Only two estimations of the molecular weight of type E g, botulinum toxin are available. Sakaguchi, Sakaguchi and Imai (1964) reported that purified type E toxin did not penetrate Sephadex G-200 gel, therefore, the molecular weight must be 200,000 or larger. Ultracen- trifugal studies on purified type E toxin done by Fiock, Yarinsky and Duff (1961) showed that the partially puri- fied fraction produced by their procedure was not homoge- neous. Two components with calculated S values of 12.5 12 and 4.7 were observed. The more quickly sedimenting boundary represented the major component. Gerwing, Dolman and Arnott (1962) also analysed a highly purified prepara— tion of type B botulinum toxin ultracentrifugally. They obtained sedimentation constants of 5.6 for the toxic com- ponent and 1.1 for the non-toxic component. Gerwing 32 a1. (1964) later reported that a highly purified type E toxin was electrophoretically and ultracentrifugally homogeneous and possessed a calculated molecular weight of 18,600. Recalculation of the molecular weight by Gerwing, Dolman and Ko (1965) based on the analysis of amino acid residues showed that the molecular weight of the pure toxin was around 14,000 to 16,000 or roughly 20 % less than the fig- ure based on ultracentrifugal analysis. Kitamura, Sakaguchi and Sakaguchi (1967) found the purified material below pH 6.0 homogeneous with 520W = 12.3. However, this material above pH 7.0 resolved into two frac- tions, a toxic and a non-toxic fraction each with S values of 7.3 Composition and Active Site of the Toxin Buehler, Schantz and Lamanna (1947) studied the amino acid composition of type A toxin by microbiological assay. They found that the total number of amino acid residues were 7754 and that cysteine was the limiting amino acid. Analysis showed a total sulfur content of 0.436 % 13 and was accounted for in the methionine, cysteine and half— cystine content. The nitrogen content of the toxin was 16.2 % which was slightly higher than the 14.2 % nitrogen in the toxin reported by Abrams, Kegeles and Hottle (1946). Lamanna (1959) defined the toxins as simple globu- lar proteins composed exclusively of amino acids. He des- cribed type A crystalline toxin as a white odorless protein with "unknown taste" and with an isoelectric point about. pH 5.6. . Stefayne et a1. (1964) using chromatographic tech- niques for amino acid determination of the toxin found agreement with Buehler's report for 14 amino acids. Some values for tyrosine, phenylalanine, tryptophan, serine and glycine were different. Schantz and Spero (1957) studied the active site of the toxin by reacting the toxin with ketene. They found that 98 % of the toxicity was lost when 19 % of the amino groups had reacted. Nitrous acid also deactivated the toxin rapidly. Spero and Schantz (1957) interpreted this as a deamination effect. Weil g£_§l. (1957) found a loss of toxicity due to photo-oxidation when methylene blue was present. Boroff and Fitzgerald (1958) and Boroff (1959) reported that when toxin is treated with alkali, with cer- tain iron slats, or with antisera to the toxin, there is a decrease in flourescence which correlates with the result- ing decrease in toxicity. They suggested that the structure l4 responsible for flourescence and toxicity is the same or related. Schantz, Stefanye and Spero (1960) refuted this opinion. They inactivated the toxin with urea without the parallel reduction in flourescence. Spero (1958) reported that the loss of toxicity at alkaline pH is caused by the ionization of the e-amino groups of lysine. Toxin dis- solved in buffered solutions at a temperature of 15 C and pH up to 10.58 was very stable for 3 hr, but 50 % of the toxicity was lost at pH 10.97 within minutes, and at pH 11.2 all toxicity was lost in 1 to 2 minutes. Coleman and Meyer (1922) observed that viable type A spores heated at 80 C for 30 min still produced toxin. Grecz and Lin (1967) made a distinction between heat sen- sitive vegetative toxin and heat resistant spore toxin. The vegetative toxin was inactivated by temperatures as low as 70 C within 15 seconds, while spore toxin was heat resistant and was not destroyed by heating for 10 min at 80 C. To date heat resistant spore toxin has been found in type A but not with types B and E (Kempe and Graikoski 1962, and Schneider, Grecz and Anellis 1962). Type B and E toxins released in the medium were denatured by heating for 10 min at 80 C. 15 Cartwright and Lauffer (1958) observed that at pH 6.9 toxin was detoxified in minutes at 50 C, while at 40 C the toxin was stable for more than 1 hour. Boroff and DasGupta (1964) reacted type A toxin with 2-hydroxy-5-nitro-benzy1bromide (HNBB) and contended that inactivation was due to-tryptophan modification, since HNBB reacts specifically with tryptophan. Actually HNBB reacts with cysteine also (Koshland, Harkhanis and Latham, 1964). I Gerwing gt Ei- (1966a, 1966b) found-17 amino acids and approximately 85 amino acid residues in type B toxin. Arginine was identified as the single N—terminal amino acid residue. They concluded that cysteine played a critical role in toxicity since toxicity was rapidly lost when cysteine residues reacted with para-chloromercuricbenzoate (PCMB), or N-(4 dimethyl-amino-3,5-dinitropheny1)-ma1eimide (DDPM). They found no tryptophan in type B toxin. Gerwing, Dolman and Ko (1965) reported the loss of 18 of the amino acid residues present in the original pro- tein when type E toxin was activated by trypsin. The N- terminal amino acid residue was glycine in the non-acti- vated toxin and arginine in the trypsin treated toxin. The sequence data for type A, B and E botulinum toxins were published by Gerwing, Mitchell and Van Alstyne (1967). They concluded that the cysteine residue, in com- bination with other amino acid residues which may be close 16 to it in terms of protein tertiary structure, is involved in the toxic activity of these proteins. Due to the above discrepancies, the need for addi- tional studies is apparent. MATERIALS AND METHODS Microorganism. Clostridium botulinum type E Van- couver Herring (VH) strain was used for toxin production. The strain was obtained from Dr. C. F. Schmidt, Continental Can Company, Chicago, Illinois. Culture medium. A medium consisting of 5.0 % tryp- ticase (B.B.L.), 0.5 % peptone (Difco), 1.0 % yeast extract (Difco), 0.2 % sucrose, and 0.02 % sodium thioglycollate at pH 7.2 (TPSY) was used. This medium ahd previously been found to yield excellent growth, sporulation and toxin production by type E strains of C, botulinum. The organisms were grown and toxin was produced by a modified version of the sac method of Vinet and Fredette (1951). Two- or 3-liter Erlenmeyer flasks were filled with physiological saline solution and a dialysis sac filled with medium was suspended in each flask. The ratio of the medium to saline solution was approximately 1:4. An inoculum of 15 ml of a 12 to 16 hr actively growing culture was introduced into the saline solution and incu- bated at 30 C for 7 days. Toxin assays and protein deter- minations were carried out daily. 17 18 Toxicity assays. White mice weighing 15 to 20 g were injected intraperitoneally with 0.5, 0.2, and 0.1 ml of serially diluted toxic preparations. Samples were di- gested with 1.0 % trypsin (Difco, 1:250) in an equal volume of 0.05 M sodium phosphate buffer at pH 6.0. Further di- lutions were made in the same phosphate buffer. Two mice were injected per dilution. The highest dilution at which the mice died within 96 hr was used for mouse lethal dose (MLD) calculations. Analytical methods. Procedures used to obtain protein concentrations varied according to the accuracy that was required. The spectrophotometric method (Warburg and Chris- tian, 1942) was used for a rapid estimation of protein concentration of the fractions. Protein was determined spectrophotometrically by measuring the absorption of light at wavelengths of 280 and 260 nanometers, with a correc- tion for the nucleic acid content from the data given by Warburg and Christian (1942) for known mixtures of cry- stalline enolase and yeast nucleic acid (Merck). The de- termination was carried out in silica cells (d = 1.0 cm). One aliquot of the protein solution was suitably diluted with water or buffer to a volume of 3.0 ml and the absor- bance was read on a Beckman DBG spectrophotometer (Beckman Instruments, Fullerton, California) at 280 and 260 nm 19 against a blank containing the solvent. The protein or nucleic acid concentration was then obtained from the nomograph. The Biuret reaction (Gornall 33 al,, 1949) was ap— plied to determine protein concentration in the partially purified toxic supernatants. The method of Lowry gt 31. (1951) was used to es- timate the protein concentration in the purified fractions. Total nitrogen determinations were carried out by the micro-Kjeldahl method (Kabat and Mayer, 1948) on the purified concentrated toxins used for ultracentrifugal work. Purification procedures. Purification procedures were modified from the procedures of Gerwing gt 31. (1946). A schematic flow sheet of the procedures used in these studies is presented in Figure 1. The bacterial cells were separated from the toxin by centrifugation at 5,000 x g for 20 min at 4 C. The toxic supernatnat was then filtered through a Millipore membrane filter with a mean pore size of 0.45 um, and the toxin was precipitated over- night at 4 C with crystalline ammonium sulfate [(NH4)ZSO4] at 60 % saturation. The precipitate was then collected by centrifugation for 20 min at 10,000 x g at 4 C, dissolved in 0.05 M sodium phosphate buffer at pH 6.0, and dialysed overnight at 44C. Any insoluble material present after 20 Type E VH strain grown for 7 days in dialysate at 30 C 2 x 105 MLD toxicity per ml or 1.3 x 104 MLD toxicity per mg protein Centrifuged at 5,000 x g for 20 min at 4 C J Toxic sfipernatant Cells Filtered through 0.45 pm Millipore filter Precipitated with 60 % (NH4)2SO4 at 4 C Centrifuged at 10,000 x g for 20 min at 4 C VV Precipitate dissolved in 0.05 M sodium phosphate buffer at pH 6.0 Dialysed overnight at 4C 5.7 x 105 MLD toxicity per mg protein Purified on DEAE (DEAE purified fraction) \ 3’ Toxic fraction Non-toxic fraction 1.04 x 106 MLD toxicity per mg protein Purified on CM - Sephadex C - 50 ¢7' 7$ Toxic fraction eluted with Non-toxic fractions 0.5 M NaCl (CM purified fraction) 5.0 x 106 MLD toxicity per mg protein Figure 1. Schematic diagram of the purification procedure of type E Clostridium botulinum toxin 21 dialysis was removed by centrifugation. This crude toxin preparation, approximately 1/40 of the original volume was than further purified on diethyl-aminoethyl (DEAE) cellu- lose columns (1 x 28 cm). The DEAE cellulose (Brown Co., Berlin, N. H.) was pretreated with 2.0 N NaCl for 24 hr at 4 C and packed into 1 cm diameter columns with slight pres- sure to a height of 28 to 30 cm. The packed columns were washed with 1.0 N HCl until they became acidic, and then equilibrated with 0.05 M sodium phosphate buffer at pH 6.0. Quantities of 4.0 m1 of the partially purified pre- parations were placed on the columns and eluted at 4 C with the phosphate buffer used for equilibrating the columns. The fractions were collected with a LBK 7000A UltroRac fraction collector. Protein was detected by a Gilson UV detector at 280 nm and monitored with a Model SR Sargent Recorder. The elution profile of the protein exhibited one. major peak and one minor peak. The fraction represented by the major peak contained all the toxin, therefore, these samples were collected and kept at 4 C for further purification. Application of up to a 0.5 M NaCl gradient on the columns did not elute more toxin. The approximate protein content of the major frac- tion was determined by the spectrophotometric method. The nucleic acid content was also calculated from the adsorbance at 280/260 nm. 22 Final purification of the toxin was achieved by placing 1 to 2 ml of the major sample on a CM-Sephadex C-50 column (1 x 28 cm). Toxin was retarded on the column and eluted with 0.5 M NaCl in the buffer. Other final purification techniques also evaluated were: a.' Disc electrophoresis was carried out in gels at pH 4.3 as subsequently described in the disc electro- phoresis methodology section. The proteins were separated and then the parts of the gels containing the toxic pro- tein were cut out and toxin was extracted overnight as described. b. Disc electrophoresis in gels at pH 4.3 was carried out until the two non-toxic proteins ran off the gels. Then the end of the tubes were capped with dialysis membranes and electrophoresis was continued until the toxic protein migrated off the gels. The toxin was driven in this manner into the few drops of buffer occupying the space between the gel and the dialysis membrane. Treatment of dialysis sacs. A11 dialysis sacs used in the experiments were boiled for 5 min in 0.01 M ethylenediaminetetra-acetic acid (EDTA) solution at pH 7.0 (Gerwing §£_al., 1964) to prevent the inactivation of toxin by surface-active agents. The bags were washed in dis- tilled water after they were boiled. 23 Determination of sedimentation coefficients. A Beckman-Spinco model E analytical centrifuge equipped with a synthetic boundary cell was used for determining the sedimentation coefficient. Determinations were made at 20 C for 40 min at 59,780 rpm using protein concentrations of 3.85 mg and 6.12 mg protein per ml of 0.05 M sodium phos-. phate buffer at pH 6.0. Concentration of toxin. The toxic preparation was concentrated by evaporation for the centrifugal studies. Dialysis bags were filled with DEAE purified toxic fractions and were suspended in front of a fan at 4 C. Evaporation was completed when approximately 90 % of the liquid was lost. The concentrates were dialyzed overnight against 0.05 M phosphate buffer at pH 6.0 and the nitrogen content was determined by micro-Kjeldahl determination. Protein concentrations were calculated by using the equation pro- tein = N x 6.25. Disc electrophoresis. Disc electrophoresis of the toxic fractions was performed in pH 4.3, 7.0 % acrylamide gel, using pH 5.0 B-alanine-acetic acid buffer, and in pH 8.9, 7.0 % acrylamide gel using pH 8.3 tris-glycine buffer with a modification of the David (1964) technique. The concentrations of stock solutions and preparations of work- ing solutions were different from those of Davis and are given below. A sample of 50 to 100 mg of protein was 24 placed on the gel and electrophoresis was carried out at 4 C with a current of 3 m. Amp. per tube, applied to an analytical temperature regulated disc electrophoresis ap- paratus (Polyanalyst, Buchler Instruments, Fort Lee, N.J.). The current was obtained from a power supply regulated with a Heathkit variable voltage regulator, Model PS-3. (Heath Co., Benton Harbor, Mich.). Two tubes per sample were used for electrophoresis: one was stained to detect the bands while the other one was used for toxin assay. Por- tions of the unstained gel corresponding to the bands on the stained gel were cut out, macerated, and extracted ! overnight at 4 C in physiological saline phosphate buffer at pH 6.0 (l to 2 ratio), then trypsinized and assayed for toxin. Stock solutions for poSitively and negatively charged proteins were prepared and mixed as listed in the appendix. Test for hemagglutinating activity. The rapid test of DasGupta and Boroff (1968) was used in which a drop of human red blood cells in suspension was mixed with a drop of partially purified toxin on a microscopic slide. Agglutination was observed with or without the aid of a microscope. Quantitative estimation of hemagglutination was done by adding 0.5 m1 of a 2.0 % suspension of human red 25 blood cells in physiological saline to 0.5 ml of serially diluted purified toxin (1 to 10 then 2 fold dilutions). The mixtures were incubated at 37 C for 60 min and then at 4 C overnight. Reactions were considered as positive in which all cells were clumped at the bottom of the tube and resuspended as distinct granules, flakes, or flocculent masses in an otherwise clear fluid. The most dilute sample which caused cell clumping was recorded. 'Estimation of molecular size. The size of the toxic fractions was estimated by gel filtration (Andrews 1965). Columns were prepared as follows: to Sephadex G- 200 gel (water regain 20 i 2 g/g) distilled water was added and the gel was allowed to swell on a boiling water bath for 5 hr. The hydrated gel was deaerated under vacuum and the columns (1 x 28 cm) were filled at 4 C and equilibrated with 0.05 M sodium phosphate buffer at pH 6.0. The void volume was determined with blue dextran 2,000 (M.W. 2 x 106). Blue dextran (0.2 g) was dissolved in 25 ml of dis- tilled H20 and 0.05 to 0.1 ml was applied on the column. The column was calibrated with hemoglobin (M.W. 6.7 x 104), cytochrome c (M.W. 1.22 x 104), and aldolase (M.W. 1.4 to 1.5 x 105). Solutions of 4 mg/ml of hemoglobin, 5 mg/ml of cy- tochrome c and 4 mg/ml of aldolase were prepared and 0.05 to 0.1 ml of each were applied on the column. The elution a. 26 volumes (Ve) of each of these compounds were carefully measured and plotted versus their molecular weights on semi—log paper. A straight line relationship was obtained. The molecular weights of the unknown toxic proteins were determined by obtaining their elution volumes on the column and then determining their molecular weights from the standard plot. The sources of proteins used were as follows: human hemoglobin, (lot No. T-3131, Mann Research Labora- tories, N.Y., N.Y.); aldolase, (rabbit muscle, lot No. U- 114, Mann Res. Lab., N.Y., N.Y.); cytochrome c, (equine heart, Type III, lot No. 87B-7131, Sigma Chemical Co., St. Louis, Missouri). Blue Dextran 2,000 and Sephadex G-200 were obtained from Pharmacia Chemical Company, Uppsala, Sweden. Preparation of buffers. Phosphate buffers were prepared by titrating 0.05 M Na2HP04 with 0.05 M NaH2P04 to the required pH. The Tris-HCl buffers were prepared by titrating 0.05 M Tris with 0.05 M HCl to the required pH. RESULTS AND DISCUSSION Toxin Production The modified dialysis sac procedure was used for growth and toxin production of type E Clostridium bgtgr liggm. The procedure was modified in that the saline con- taining the inoculum was placed outside and the medium was placed inside the sac. The original sac method was used successfully for toxin production by Gerwing, Dolman and Arnott (1961). They obtained toxic filtrates with 3.2 x 103 MLD activity 4 per mg N, or 2.0 x 10 MLD activity per mg protein. The toxic filtrate produced by our modification had a slightly 4 MLD per mg protein, however, lower activity of 1.3 x 10 our method did not require the construction of an apparatus. We harvested the toxin after 7 days of incubation while they separated their toxin from the cells after-5 days of incubation. We found a difference in toxin production among the type E strains. The Kalamazoo strain reached a peak of toxin production after 4 to 5 days of incubation. Due to the differences in methods, in number of days of incubation and in strains used for toxin production, the 27 28 method to obtain maximum toxin production must be deter- mined for each strain. The sac technique has several advantages over the usual bottle or flask technique. The bacteria are grown and toxin is produced in a dialysate, therefore, the pro- duced toxin has fewer impurities than the toxin produced in bottle or flask cultures. The toxins produced in the saline are free from nondialysable constituents in the medium, and the large amount of impurities in the medium do not interfere with the purification procedures. The nutrients constantly diffuse into the saline and provide a renewed supply of nutritive substances for growth. Some of the dialysable metabolic products which might restrain growth are removed. Thus much heavier growth could be ob- tained. The Erlenmeyer flasks were capped with aluminum foil which allowed the gas produced to escape and eliminated the danger of a flask explosion by seating of the screw cap. The medium in the sacs was used at l, 2, or 4-fold concentrations. Best results were obtained when the 2 fold concentration of TPSY medium was used. There were fewer numbers of bacteria produced in normal TPSY medium than when a 2 fold concentration was used. The 4 fold concen- tration was unsatisfactory because the highly concentrated medium enhanced diffusion of water into the bags and con- stant leakage occurred at the top of the sacs. Thus a 29 route for bacterial contamination was opened and a constant loss of growth medium was also encountered. Therefore, the medium at 2 fold concentration was used for toxin pro- duction. Two milliliters of sample were removed daily for toxin and protein assays. The amount of toxin produced in the cultures steadily increased during the first 7 days of incubation (Table 1). A slight reduction in the amount of Table 1.--Toxin production of the VH strain in "dialyzed" TPSY medium Days of incubation Toxin in dialysate at 32 C in MDL/ml 102 103 104 102 10 105 log 10 105 u>m~aoxma>dlwrd Hboouoaspwokap xxxxxxxxx toxin.produced was noticed on the 8th and 9th days of in- cubation, therefore, toxin was separated from the cells on the 7th day of incubation by centrifugation at 5,000 x g at.4 C. The activity of the toxin was measured in MLD rather than in LD50. The purpose of the toxin assay was to find the time when the cultures contained the maximum amount of toxin and for this purpose determination of MLD 30 was quite satisfactory. Considerably fewer mice were re- quired for the MLD titration than the LDSO' and the labor involved was also greatly reduced. The appearance of the cultures was followed by ob- serving a representative sample of the cultures daily with the aid of the microscope fitted with phase contrast optics. The cultures consisted of mostly vegetative cells and very few spores on the first and second day of incuba- tion. About 40 % sporulation and a great number of actively growing vegatative cells were observed on the third day. The number of growing vegetative cells decreased sharply on i the fifth and sixth days of incubation, and the number of Y free spores also greatly increased. Cell disintegration and accumulation of cell debris was observed from the third day on. Gerwing, Morrell and Nitz (1968) studied the syn- thesis and nature of toxin of type B E, botulinum in grow- ing cells and demonstrated that the toxin was synthesized in the latter stage of logarithmic growth and was released into the supernatant fluid during lysis of the cells. Toxin Isolation and Purification A sufficient number of flasks were prepared so that approximately 14 liters of toxic supernatant was obtained and the toxic material was precipitated as indicated in Figure l, with 60 % saturation of crystalline ammonium sul- fate (NH4)ZSO4. The toxic precipitate appeared as a 31 whitish flocculent material which settled out when left overnight at 4 C. The precipitate was sedimented by cen- trifugation and the sedimented toxin material was dissolved in 0.05 M sodium phosphate buffer at pH 6.0. The-toxic material was water soluble and dissolved very easily in the pH 4.5 to 7.0 buffers used in our studies. The partially purified material, approximately 1/20 to 1/40 of the original culture volume, contained 5.7 x 5 MLD toxicity per mg protein and was not ready for fur- 10 ther purification of DEAE cellulose columns. The toxic. material was dialyzed overnight at 4 C against the phos- phate buffer to remove all traces of (NH4)ZSO4. Most of the purification was done at 4 C to prevent a loss in toxin activity. The dialysate was assayed for toxicity to deter- mine if any toxin diffused through the sac. Ordinary di— alysing sacs will permit materials with molecular weights below 30,000 to pass through. Throughout the course of our studies toxin in the dialysate was never observed after precipitation with (NH4)ZSO4. The medium in the dialysis sac suspended in the saline solution in which the cultures were growing also never contained any toxicity. Therefore, the toxic material produced by the bacteria and the active material precipitated by (NH4)ZSO4 were probably large enough not to pass through the membrane. The partially purified toxic material separated on DEAE cellulose columns into a major fraction and a minor 32 fraction (Figure 2). The major peak samples were highly toxic and contained 1.0 x 106 MLD toxicity per mg protein. The specific activity of this purified material was approx- imately 50 % higher than that prepared by (NH4)ZSO4 pre- cipitation. The smaller fraction was only slightly toxic and contained 1 x 103 MLD toxicity per mg protein. Since the small fraction followed the major fraction fairly closely, probably some overlapping occurred that caused the negligible activity of this fraction. The separation of the toxin of type E strain Iwanai into two components using DEAE cellulose was reported by Gerwing gt El. (1964). They obtained a highly toxic frac- tion containing 7.5 x 106 MLD per mg N, and a non-toxic fraction. Our results are in agreement with theirs, except they obtained a higher specific activity of the toxin. This might be explained by the difference in buffer systems used. Sakaguchi, Sakaguchi and Imai (1964) pointed out that the toxin was approximately 5—fold less active when pH 7.5 phOSphate buffer was used for elution than when acetate buffer at pH 6.0 was used. Other variants such as strain differences should also be taken into account. The principle involved in DEAE cellulose purifica- tion is that the toxic material elutes frontally while most of the other components are strongly adsorbed to the acidified DEAE cellulose.and are released when the pH is altered or when the molarity of the eluent is changed. .35 .30 E .25 c O (I) N u .20 m m o c B H .15 o E .10 .05 Figure 2. 33 20 30 40 50 ML eluent Elution patterns of Type E VH toxin on DEAE cellulose with 0.05 M sodium phosphate buffer, pH 6.0. The major component contains the toxin 34 Absorbance of the major peak fractions at 280/260 nm yielded ratios of 1.06 to 1.35 showing that the isolated, purified toxic fractions were essentially all protein and that the amount of nucleic acid present in the preparation was approximately 0.9 to 3.0 %. The nucleic acid content was computed from the data given by Warburg and Christian (1964) for known mixtures of crystalline enolase and yeast nucleic acid (Merck). The DEAE columns were equilibrated with 0.05 M sodium phosphate buffer at pH 6.0 and elution was carried out with the same buffer. The use of this buffer was pre- ferred over 0.05 M sodium acetate buffers at pH 5.5 and 4.5 since it was observed in this laboratory that toxic materi- al produced by type E Kalamazoo strain after purification with (NH4)ZSO4 separated into four components on DEAE cellulose columns. The first three peak fractions were all highly toxic; 5 x 105 MLD/m1, 2 x 105 MLD/ml, and 1 x 105 MLD/ml when pH 4.5 or 5.5 sodium acetate buffer was used for elution (Figure 3), and no definite differences in total toxicity could be observed in any of the three fractions. The fourth fraction had a very low toxicity of 2 x 103 MLD/m1. 4 MLD/ml was obtained only Toxic activity of 7 x 10 in the third peak fraction when the same toxic material was eluted with pH 6.0 phosphate buffer (Figure 4). The other three fractions contained no activity. Although the .25- .20~ 010‘ Absorbance at 280 nm 005 " Figure 3. 35 20 30 40 50 60 ML eluent Elution of 2 m1 of type E Kalamazoo toxin on DEAE cellulose with 0.05 M sodium acetate buffer, pH 4.5. The first three components contain the majority of the toxicity 36 045‘ .40. .35. o o o N N U o L11 0 n l : Absorbance at 280 nm i.- U! J .10- 005 d J ql— I —' j I 1 ' 20 30 4o 50 60 7o éo do ML eluent Figure 4. Elution of 2 m1 of type E Kalamazoo toxin on DEAE cellulose with 0.05 M sodium phosphate buffer, pH 6.0. The third component contains the toxin 37 number of peaks were the same with both buffers, the heights and areas of the peaks were different. The non- toxic fraction produced a much larger peak with phosphate buffer elution than with acetate buffer elution. The slight hump that appeared with phosphate buffer elution be- tween the third and fourth peak did not appear in the ace- tate buffer elution pattern. However, as more portions of partially purified toxins were placed on the column this non-toxic compound became larger and was regarded more and more and the hump became discernible. The hump in Figure 3 could probably be overlapped by the larger peak. The differences in elution patterns might be explained by a difference in charge of the toxic protein at pH 4.5 and at pH 6.0. The difference in the elution patterns of the toxin of the VB strain produced similar toxin elution profiles containing two components only one of which was toxic when both acetate and phosphate elution buffers were used. How- ever, the decision to use phosphate buffer was made be- cause of the need for a standard purification procedure which could be applicable to other type E strains. There- fore, all subsequent initial purifications on DEAE were carried out with phosphate buffer at pH 6.0. Toxin eluted with phosphate buffer showed about 5- fold less toxicity than toxin eluted with sodium acetate buffer. This phenomenon was also observed by Sakaguchi, 38 Sakaguchi and Imai (1964) who found that potencies of toxins eluted by pH 7.5 phosphate buffer were lower than those obtained with acetate buffer at pH 6.0. This may reflect a lower degree of stability at the higher pH, or sodium acetate buffer may keep the toxin stable for a longer time than phosphate buffer. Application of a NaCl gradient on the columns up to 0.5 M concentration did not elute more toxic material. Therefore, the toxic fractions were pooled and kept for further purification. A second precipitation with (NH4)ZSO4 and purification of the toxic fractions on DEAE was also investigated. However, this additional procedure did not increase the degree of purification. Disc electrophoresis of the fractions indicated three bands in both fractions. Both fractions produced the same pattern of one dense and two slight bands. Therefore, this second purification treatment was not repeated. Disc electrophoresis of the fractions will be discussed in detail later. Final purification of the toxic fraction was car- ried out on CM-Sephadex C-50 gel. This ion exchange gel retarded the toxin when eluted with phosphate buffer at pH 6.0. The toxin strongly adsorbed to the acidified gel, while some of the other non-toxic proteins were eluted with the buffer. Toxin was released from the gel in one small fraction when 0.5 M NaCl was applied to the column (Figure 5). Toxin assay showed that this toxic fraction Absorbance at 280 nm 39 '20' 0.5 M NaCl I l 015‘ l I l l 010‘ I l | I 005- : l l l T I T I ‘F 1 A 10 20 30 10 20 ML Eluent Figure 5. Elution of 1.5 ml of DEAE purified type E VH toxin on CM-Sephadex C-SO. The peak eluting after NaCl application contains the toxic com- ponent 40 7 MLD toxicity contained approximately 5 x 106 to 1 x 10 per mg protein, which was a give to tenfold increase in activity when compared to the activity obtained after DEAE chromatography. A continuous increase in activity of the toxin was demonstrated throughout our purification pro- cedures. Our final preparation contained a toxic activity 6 to 1 x 107 MLD per mg protein. Gerwing at 31. of 5 x 10 (1964) with similar procedures obtained 7.5 x 106 MLD toxicity per mg N, or 4.7 x 107 MLD toxicity per mg pro- tein, while Kitamura, Sakaguchi and Sakaguchi (1967) with a different procedure obtained 5 to 10 x 107 MLD toxicity per mg N or 3.1 to 6.2 x 108 MLD toxicity per mg protein. The latter workers obtained the highest activity, but their starting material and purification procedures were differ- ent. They extracted the toxin from the bacterial cells with phosphate buffer and purified only the extracellular toxin. rTheir purification procedure of the toxin involved RN-ase digestion, CM-Sephadex purification and Sephadex G-200 gel filtration. They used acetate buffer at pH 6.0 while Gerwing 25 31. (1964) used acetate buffers at pH 4.5 and 5.5 and we used phosphate buffer at pH 6.0 throughout the purification procedures. Kitamura, Sakaguchi and Sakaguchi (1967) used the Hazen strain for toxin production. Our attempt to grow this strain in TPSY medium or in the peptone-glucose-yeast 41 extract medium used by them was unsuccessful. This strain grew very poorly in both media and produced little toxin. The CM-Sephadex C-50 column when packed and equil— ibrated with phosphate buffer at pH 6.0 had a height of ap- proximately 28 cm. CM-Sephadex gels are supplied in their sodium forms. They are strongly hydrophilic and swell in water and salt solutions. Application of NaCl to the column shrunk the gel to a height of 18 to 20 cm. Because of this excessive shrinking and swelling property it was hard to work with this gel since it was difficult to obtain a constant flow rate. Therefore, other final purification procedures were also investigated. Disc gel electrophoresis purification procedures use the principle that proteins are separated because they possess a different net charge at different pH and the rate of migration of the molecules in an electric field depends upon the net charge. The first or simpler pro- cedure involved the separation of the proteins by disc gel electrophoresis, followed by extraction of the toxic pro- tein from the gel. The portions of the gels containing the toxin were cut out and extracted overnight in a 1:2 mix- ture of physiological saline and phosphate buffer at pH 6.0. A highly purified toxin preparation could be obtained in this manner since the toxin readily diffuses from the gel into buffered saline. Some loss of protein could occur with this method since the overnight extraction may not 42 extract all the protein. The use of a longer extraction time, such as two or more days, may result in the loss of samples through microbial contamination. The second method employed a procedure that util- ized the difference in rate of migration of the proteins in the gel. The non-toxic proteins pass through the gel first, then the tube is removed and capped with a dialysis membrane and the slowest moving toxic protein is captured in the few drops of buffer occupying the space between the gel and the dialysis membrane. The dialysis membrane is placed on the end of the tubes after all non-toxic proteins have migrated into the buffer system. This method worked well but through leakage around the membrane the loss of protein is possible, and some care is required in placing the dialysis membranes on the end of the tubes without en- closing an air bubble. The membrane material must be a type that would not allow smaller molecules to diffuse through, therefore, cellulose sausage casing was used. It was observed during our experimentation that purified toxin diffused through ordinary dialysis bags. Since dialysis bag materials do not carry any specification as to the retention of molecular sizes, the diffusion of smaller colecular compounds through the membranes could be estab- lished only on a trial and error basis. Cellulose sausage casing will retain materials with molecular weights above 4,000. 43 The diffusion of purified toxin through the dialy- sis membranes used in most of our toxin production was ob- served upon several occasions. This indicated that these purified molecules were smaller than those present in toxic cultures or those obtained after precipitation with (NH4)ZSO4. It is possible that the toxin in its original form exists as a large molecule, or as an aggregate, or it may combine with some large non-toxic protein. Chromatog- raphy on DEAE cellulose may fractionate the protein into smaller peptide units, or it may disaggregate the large aggregates into their original small units. Homogeneity of the Toxin The toxin fraction purified on DEAE cellulose ap- peared to be homogeneous with respect to molecular size when examined by ultracentrifugation (Figure 6). One slowly moving component was observed and a sedimentation coefficient of Sobs = 0.6 was calculated. However, at the beginning of the centrifugation two small components were also observed which later diffused. This may have been due to the presence of very small amounts of impurities or the presence of some aggregated materials. But since microscopic appearance of the protein solutions showed no impurities or aggregated materials, it was possible that small amounts of other proteins were also present in the preparation. Therefore, the partially purified preparations, Figure 6. 44 Ultracentrifuge photographs of DEAE purified type E VH toxin at 59780 rpm, 0 min and 32 min after reaching full speed, in 0.05 M sodium phosphate buffer, pH 6.0 45 previously examined ultracentrifugally were tested by disc electrOphoresis to determine if the preparation was homogeneous with respect to number of components present. Disc electrOphoresis was carried out in 8.0 % acrylamide gels at first, but a decision to change to 7.0 % gels was made after it was observed that the proteins moved very little after 2 hr of electrophoresis. Electrophoresis carried out in gels at pH 8.9 produced a substance with low electrophoretic mobility. It was difficult to ob- serve whether the material consisted of one or more frac- tions due to poor resolution. The low electrophoretic mo- bility suggested that either the protein or proteins had very few net charges at this pH or the size of the pro- teins was large. Disc electrophoresis was also carried out in gels at pH 4.3. The concentrate separated into three fractions in these acid gels (Figure 7) and one dense and two faint fractions were observed. Assay of the frac- tions for toxin showed that only the one fraction with the lowest electrophoretic mobility was toxic. Acid gel disc electrOphoresis was performed also on unconcentrated DEAE purified fractions not used for ultracentrifugal studies and the same three materials resulted. A few milliliters of the preparation were treated with 4.0 and 8.0 M urea to ensure that three different proteins were present in the purified preparation. Urea causes an alteration in the conformation of proteins and competes with the hydrogen bond in the native protein and produces an unfolding. Disc Iii Figure 7. 46 \ p... «a o, Replicate disc electrOphoresis patterns of pur- ified type E VH toxin on pH 4.3, 7.0 % acrylamide gel using pH 5.0 B alanine-acetic acid buffer and a current of 3 mAmp with the cathode at the bot- tom. The top band contains the toxin 47 electrophoresis carried out in the preparation treated with urea produced similar results to those observed without the urea treatment. One major and two minor fractions were observed. Since urea quickly detoxifies the toxin, the toxicity of the fraction with the lowest electr0phore- tic mobility was not established. It is possible that urea by unfolding the molecule also affects the toxic site or sites thereby causing the loss of activity. The loss of toxicity of type A toxin after 6 M urea treatment was previously reported by Schantz, Stefanye and Spero (1960). Disc electrophoresis established that three dif- ferent proteins were indeed present in the DEAE purified preparation, and that the preparation was not homogeneous with respect to the number of components present. Only one toxic protein, the one with the lowest electrophoretic mobility, was detected. Although the shape and size of the protein molecule influence the absolute rate of electrophoretic migration, the major factor is the net charge. It is most likely that the toxin has the least number of net charges, but it may also be possible that it is a large molecule and moves more slowly through the gel than the other proteins. The disc electrophoresis data thus suggest the pos- sibility that when the molecular size and homogeneity of the preparation was determined with the analytical ultra- centrifuge, the protein with the small sedimentation coef- ficient was not the toxic protein but represented instead 48 the major non-toxic protein producing the large dense band on the gel. The toxic protein could have been simply un- observable because it was present in very low concentration in the preparation. Since its exact proportion of toxin in the purified preparation is unknown, it is quite possi- ble that even if samples with higher protein concentration would be analyzed centrifugally, the toxic protein would not be identifiable by an absorption peak. Gerwing gt_§l, (1964) using similar purification techniques, analyzed the purified material with the ultracentrifuge using a protein concentration of l %, and observed the presence of a single major homogeneous component. A rapidly sedimenting com- ponent producing a small shoulder was also observed and accounted for as a small percentage of apparent impurities. The peak moved slowly with a sedimentation coefficient of 520W = 1.70 S and our findings are in agreement with theirs in respect tx>indication of a small protein molecule. However, Kitamura, Sakaguchi and Sakaguchi (1967) with their different purification method obtained a purified toxic product which appeared to be homogeneous below pH 6.0 with an 520W = 12.3 5. Their result_was in good agree- ment with an SZOW of about 11.5 reported by Sakaguchi and Sakaguchi (1967) for a purified homogeneous toxic product. It was also observed during our disc electrophore- sis experiments that if the concentration of protein ap- plied to the gels was decreased the two faint components 49 could not be observed and only a sharp band representing the major protein would be present. This proved that it is quite possible to fail to detect the toxin and to come to the conclusion that a homogeneous compound is present, even when the preparation is not homogeneous. Very poor resolution was obtained when the basic gel was used for disc electrophoresis. It is possible that at basic pH the proteins carried very few net charges and they were close to their isoelectric points. The iso- electric point of type E toxin is not known yet. Crystal- line type A toxin is reported to have an isoelectric point at pH 5.6 (Putnam, Lamanna and Sharp, 1946 and 1948). The toxin purified on CM-Sephadex was also examined using disc electrophoresis. Electrophoresis was conducted using pH 4.3 gel, and only one component was obtained (Figure 8). This indicated that the preparation contained one protein.with a very low electrophoretic mobility. Toxin assay of the gel showed that the protein was highly active and it traveled with the same mobility as the toxic fraction on the control. The DEAE cellulose partially pur- ified toxin was used as a control. It would appear that with this final purification step the separation of the toxin molecule from other proteins was achieved, and a toxic product homogeneous with respect to the number of components was obtained. 50 A 8 Figure 8. Disc electrophoresis patterns of type E VH toxin, B. before and A. after CM-Sephadex pur- ification on pH 4.3, 7.0 % acrylamide gel with pH 5.0 B alanine-acetic acid buffer and a cur— rent of 3 mAmp 51 Homogeneity of the fractions purified on DEAE and on CM-Sephadex were also tested on Sephadex G-200 gel columns. Molecules larger than the largest pores of the hydrated Sephadex cannot penetrate the gel particles and, therefore pass through the bed in the liquid phase outside the particles and are thus eluted first. Smaller molecules, however, penetrate the gel par- ticles to a varying extent depending on their size and shape, and are eluted from a Sephadex bed in order of de- creasing molecular size. Therefore, Sephadex G—200 column chromatography should indicate whether a size difference exists among the proteins present in the material purified on DEAE cellu- lose and it would also give an estimate of the size of these proteins. Rechromatography of the toxic components eluted from DEAE cellulose on a Sephadex G-200 column (1 x 28 cm) produced two fractions of equal toxicity (Figure 9). This indicated that the preparation was not homogeneous with respect to molecular size. Toxin assay of the eluent occasionally showed that toxic material appeared in the eluent before any material absorbed at 280 nm could be detected with our apparatus. The protein content of these eluates was so low that the detector could not detect changes in absorbancy. Since this phenomenon could not always be observed, it is possible 52 .20- 8 (v.15- 4.) m (D O 5.10. .Q 34 O é’ 005‘ 'flv l I T 10 15 20 25 ML eluent Figure 9. Elution of DEAE purified type E VH toxin on Sephadex G-200 column with 0.05 M sodium phosphate buffer at pH 6.0 53 that the toxin may aggregate or disaggregate with time into smaller or larger units depending on the treatment to which it is subjected. The low protein content of these eluates suggested that the toxin is a protein with very high toxic activity. Eluates were found containing 5 x 104 MLD activ- ity per m1, while absorbance at 280 nm was practically zero. The fraction purified on CM-Sephadex showed the presence of only one component after passing through a Sephadex G-200 column. Toxin assay showed that the peak fraction was toxic. Therefore, this fraction on Sephadex G—200 proved to be homogeneous with respect to molecular size. Estimation of Size The size of the fraction purified on DEAE was first estimated using the ultracentrifuge. However, as was dis- cussed previously, the small protein indicated by the sedi- mentation coefficient of Sobs = 0.6 may not have been the toxic protein. Furthermore, the disc electrophoresis data showed that the protein preparations used for ultracen- trifugal studies were not homogeneous. Therefore, the data obtained with the ultracentrifuge should be accepted with reservation when the size of the toxic molecule is estimated. 54 Disc electrophoresis indicated that the toxic fraction in the preparation purified on DEAE was present in low concentration, and calibrated Sephadex G-200 columns (1 x 28 cm) were found more suitable than the ul- tracentrifugal studies for estimating molecular size. Rechromatography of the toxic fractions eluted from DEAE cellulose on calibrated Sephadex G-200 columns produced two fractions of equal toxicity (Figure 9). Both peaks were eluted after cytochrome c (Figure 10), indicat- ing that the molecular weights of the substances were less than 12,200. Elution of these fractions was carried out with phosphate buffer at pH 6.0. Fractions eluted with pH 7.0 phosphate buffer exhibited the same two peaks in their elution profile and the elution volumes were in close agreement with those obtained at pH 6.0. Evidently no change in size occurred with the change of pH. The fraction purified on CM-Sephadex was then ap- plied to a calibrated G-200 column and only one material eluted after cytochrome c (Figure 10). Therefore, the mo- lecular weight of this toxic protein was less than 12,200. Gerwing gt 31. (1964) obtained a toxic protein with a molecular weight of 18,600. Recalculation of the molecular weight by Gerwing, Dolman and Ko (1965) based on the amino acid residues showed that the molecular weight of their pure toxin was around 14,000 to 16,000 or roughly 20 % less than the figure based on ultracentrifugal 55 momonusonom may cflcuw3 poumum mum mcamuoum mo mucmflmz Hoasooaofi one .o.m mm ummmsn oumnmmonm EsfioOm z mo.o nuflz oousao ono3 mcwouonm .AEU mm x av :Esaoo oomlw xooosmmm o co oocfiauouoo mm mcofluomnm owxou mo munmwo3 Hmasooaoz BEUHMZ m¢qbumqoz 0H moa ooa _ . —-...~F P . —r-.Ph. _ _ FPF-pu Amoa x NV cmuuxmo osam A.umouu mmmzmv saxoe Amoa x mv.av mmmaooaé Avoa x h.wv cHQOHmOEmm .oa musmflm Avoa x NN.HV o .uao Axoomsmmm AM¢WQVH cofipomum waoa ISUV :onu whom Amamovm ceauomum oflxoe r 0H Im UT 3A I ma iom 56 analysis. These two reports are in agreement with our finding that botulinum toxin has a small molecular weight. A purified toxin of large molecular weight such as reported by Sakaguchi and Sakaguchi (1967) with S20W = 11.5 S, and Kitamura, Sakaguchi and Sakaguchi (1967) with 820W = 12.3 S, was never obtained in any of our studies. The concept that type E‘Q. botulinum toxin has a small molecular weight was supported by the fact that when fractions purified on DEAE were dialyzed against phosphate buffer at pH 6.0 the buffer became toxic, indicating that toxin molecules diffused through the dialysis membrane. The dialysis bags used were known to retain compounds with molecular weights of 30,000 or more. The toxic materials diffusing through the membrane must thus have had molecular weights less than 30,000. Several times during our experimentation a possible aggregation phenomenon was observed. Toxin purified on DEAE cellulose and placed on DEAE cellulose columns at pH 7.0 eluted in a sharp frontal peak and when this toxic fraction was placed on a calibrated Sephadex G-200 column at pH 7.0 for size estimation, the toxin started to elute before cytochrome c and eluted constantly until the exclu- sion limit was reached. Since no peak was obtained, it is reasonable to assume that the toxin was in form of aggre- gates of different sizes and it is possible that at basic pH the toxin would aggregate into larger sizes. This 57 aggregation phenomenon was also investigated at pH 8.5 and pH 9.1. The fraction purified on DEAE were rechromatographed on DEAE columns equilibrated with 0.05 M Tris-HCl buffers at pH 8.5 and pH 9.1 and eluted with the same buffers. No toxic material was eluted from the columns at pH 8.5. Ap- plication of a sodium chloride gradient up to 0.5 M concen- tration on the column also did not elute toxic material. Since the type E toxin is very unstable at this pH it was quite possible that by the time it was eluted it had lost all its activity, or that at this pH the toxic protein was very close to its isoelectric point and may have been in- soluble. Studies using disc gel electrophoresis also showed that a toxic preparation in pH 8.9 gel moved very little, and thus these proteins at this pH probably had very few net charges. The toxic preparation chromatographed on DEAE cellulose columns at pH 9.1 separated into three peak fractions. Only the first peak fraction possessed toxic activity. When this toxic sample was placed on a Sephadex G-200 column, the toxic protein started to elute almost at the void volume and eluted continuously until the lower ex- clusion limit was reached. No real toxic fraction was ob- served, except for the two fractions which eluted after cytochrome c. The continuous elution of toxic material within the exclusion limits of the gel may indicate that the protein 58 is in forms of aggregates of many sizes. DasGupta and Boroff (1968) separated type A toxin into a and B compo? nents on DEAE cellulose columns with Tris-HCl buffers at pH 8.0. Both components were large molecules with molecu- lar weights of 150,000 and 500,000. But since type A toxin is stable at basic pH while type B is not, the determina- tion of stable large molecular aggregates or small disag- gregates of type E toxin cannot be performed with good scientific techniques until type E toxin can be stabilized at basic pH. The adsorption of toxin to a large molecule was also observed when toxic fractions purified on DEAE were mixed with blue dextran. The toxin eluted with blue dex- tran in the void volume (Figure 11). The elution of type A toxin simultaneously with dextran was observed by Hauschild and Hilsheimer (1968). Whether dextran has an aggregating effect on the toxin, or the toxin simply ad- sorbs to dextran is not known. The toxin eluted with dex- 5 tran was highly active, contained 2 x 10 MLD toxicity per m1 while the two toxic fractions eluting after cytochrome 3 and 1 x 103 contained 2 x 10 MLD of activity per ml, respectively. Since the protein content of our blue dex- tran fraction was very low and was not measurable with the protein determinations used in the course of these studies, it is reasonable to assume that the toxin is a highly active protein, and adsorbs on blue dextran in a pure form. 59 .30. Blue Dextran .25- O (D 08.20. 4,; 2 x 105 MLD/m1 0) U C.‘ 3.15- H O m 3 2 2 x 10 MLg/ml .10. 1 x 10 MLD/ml 2 x 105 MLD/ml 5 x 104 MLD/m1 .05« a“) , . . . 5 10 15 20 25 ML eluent I Figure 11. Elution of type E VH toxin mixed with blue dex- tran on Sephadex G-200 column with 0.05 M sodium phosphate buffer, pH 6.0 60 This adsorption or coupling phenomenon of the toxin to large molecular entities was also observed with red blood cells and with bacterial cells. Toxin prepara- tions were mixed with red blood cells and eluted on a Sephadex G-200 column. The red blood cells eluted in the void volume and were highly toxic. Presumably, toxin was adsorbed to the cells. The two substances eluting after cytochrome c were also present but only carried slight toxic activity. It was also noticed that the purified material be- came contaminated very easily and supported microbial growth. This contaminated toxic material was passed through a Sephadex G-200 column and a highly toxic materi- al appeared in the void volume while the two fractions that subsequently eluted were less toxic. Contaminated toxic material filtered through a Millipore filter did not contain this toxic fraction at the void volume. The Mil- lipore filter apparently removed the microbial cells, but should have permitted the toxin to pass through even if it was in an aggregated form separate from the cells. It is unlikely that bacteria growing in the toxic material would incorporate the toxic protein without any change, or.with- out destroying its toxicity. The best probable explanation is that toxin is adsorbed to the cell walls. Since a purified toxin was never found in the course of our studies, similar to the one (M.W. 200,000) 61 reported by Sakaguchi and Sakaguchi (1967) after treatment of their toxic material with RNase, the possibility of an aggregating effect of the enzyme was investigated. The toxic material purified on DEAE containing approximately 0.9 to 3.0 % of nucleic acid was treated with 10 ug/ml RNase at room temperature. The digestion was completed in about 2 hr. This mixture digested with RNase was now 100 % protein, as demonstrated by the absorbance at 280/260 nm. The preparation eluted as two fractions on a DEAE cellulose column with phosphate buffer at pH 6.0. Both fractions eluted before the NaCl gradient application. The gradient produced no additional absorption peaks. Assay for toxin showed that the first component contained more than 90 % of the toxin. Samples of the toxic and non-toxic fractions were placed on pH 4.3 disc gels and when electrophoresis was carried out as before, the toxic component yielded four fractions while the non-toxic component yielded.two frac- tions (Figure 12) equivalent to the third and fourth frac- tion of the toxic sample. The third fraction of the non- toxic sample was extremely faint but was observable. Toxin assay of the components showed that the slowest moving ma- terial contained essentially all the toxin. This result correlates well with our previous findings. The toxic pro- tein in disc gels at pH 4.3 always moved very slowly and always had the lowest electrophoretic mobility. The toxic ‘\ component was also missing from the non-toxic fraction. Figure 12. 62 keg: g r ‘r "“ Disc electrophoresis patterns of the toxic and non-toxic fractions eluted from DEAE cellulose after RNase treatment A. Toxic fraction B. Non-toxic fraction _63 Rechromatography on a Sephadex G-200 column of the preparation treated with RNase produced the usual two frac- tions with equal toxicity of 2 to 4 x 104 MLD per mg pro- tein and a small fraction with an activity of about 1.3 x 105 MLD per mg protein. This fraction had an elution volume of 14.12 ml (Figure 10) or an estimated molecular weight of 240,000. However, this toxic material was not stable. Preparations kept at 4 C for 3 to 5 days or longer contained this fraction no longer. This may suggest that RNase had an aggregation effect on the toxin, but with time the toxin disaggregated to its smaller size. Test for hemagglutinating activity of the sample purified on DEAE gave negative results, while unpurified type A and B toxins used as controls were found to contain hemagglutinin. DasGupta and Boroff (1968) separated type A toxin into a toxic and a non-toxic fraction on DEAE cellulose columns with Tris-HCl buffers at pH 8.0. The non-toxic fraction was a powerful hemagglutinin. The pres- ence of hemagglutinin with type E toxin has not yet been demonstrated. Kitamura, Sakaguchi and Sakaguchi (1967) separated type B toxin to a toxic a and a non—toxic B frac- tion and tested the 8 fraction for hemagglutinating activ- ity. No activity was detected. Since type B toxins differ from type A and B toxins in several properties, it is quite possible that the lack of demonstrable hemagglutinin is one more differentiating property. Conversely, type B toxin 64 has a hemagglutinin fraction, the activity of this compo- nent may be very weak without trypsin activation and can not be observed. The purification procedure employed in our studies produced a highly purified type E toxin. Studies of the toxin should now be expanded to the toxins produced by the other type E strains as well as those of the other toxin types of g, botulinum. The scope of the experiments should be increased so that sufficient quantities of purified toxin can be obtained for chemical studies. SUMMARY The isolation, purification, homogeneity and molec- ular size of type E Clostridium botulinum toxin was studied. The Vancouver Herring strain was used for toxin production. Cultures of this strain were grown in physio- logical saline and nutrients required for the microorganisms were supplied in dialysis bags filled with trypticase-pep- tone-yeast extract-sucrose (TPSY) medium suspended in the saline solution. The toxins produced by this method were free from nondialysable medium constituents. The maximum amount of toxin was produced by the sixth to seventh days of incubation. Cells were separated from the toxin by centrifugation and the toxin was precipitated from the su- pernatant with crystalline ammonium sulfate at 60 % satur ration. The precipitate after dialysis against phosphate buffer was redissolved after centrifugation in 0.05 M phosphate buffer at pH 6.0. The activity of this partially 5 MLD per mg protein. purified toxic material was 5.7 x 10 Further purification was accomplished by applying the toxic material to a DEAE cellulose column at pH 6.0, which re- sulted in a major toxic fraction with an activity of 1.0 x 106 MLD per mg protein and a minor slightly toxic fraction. 65 66 Absorbance of the toxic fraction at 280/260 nm yielded ratios of 1.06 to 1.35, indicating the presence in the purified material of approximately 0.9 to 3.0 % nucleic acid. The purified fraction appeared homogeneous when examined by ultracentrifugation. One very slowly moving component was observed and a sedimentation coefficient of S = 0.6 was calculated. Apparently the component ob- obs served with the centrifuge represented a very small mole- cule. Disc electrophoresis studies of the protein prep- aration used for the ultracentrifugal experiments showed that the preparation consisted of three different entities. Material producing one dense and two slight bands were observed in pH 4.3 gels. Toxin assay indicated that the material that produced minor band with the lowest electro- phoretic mobility contained the toxin. It is possible that the toxic protein is present in the preparation in a very minor proportion and that the peak produced with the ultra- centrifuge represented the major portion of the protein fraction, while there was no evidence of the minor toxic protein because of its low concentration. The purified toxin applied to a calibrated Sephadex G-200 column separated into two equally toxic fractions. Both fractions eluted after cytochrome c, suggesting that their molecular weights were less than 12,200. 67 The disc electrophoresis and the gel filtration experiments showed that the toxic material purified on DEAE was not homogeneous with respect to the number and size of components. Final purification of the toxic fraction was done on CM-Sephadex C-50 columns at pH 6.0. This acidified gel retarded the toxin until the molarity of the eluant was changed by adding 0.5 M NaCl to the eluting buffer. The eluted toxic fraction now contained an activity of 5 x106 MLD per mg protein and eluted as one absorption peak. Disc electrophoresis studies of the doubly chroma- tographed toxin showed only one toxic component that pro- duced only one band in the pH 4.3 gel. The molecular weight of this toxic protein was es- timated by the gel filtration technique. The toxin eluted as a one peak fraction after cytochrome c, indicating that it had a molecular weight of less than 12,200. The disc electrophoresis and the gel filtration results suggested that the toxin purified on CM-Sephadex was not essentially homogeneous with respect to number of components and molecular size. A possible aggregation phenomenon of the toxic protein was observed when elution on DEAE cellulose columns were carried out with buffers at pH 7.0 and 9.1. Toxic fractions purified on DEAE cellulose columns with these buffers and rechromatographed on Sephadex G-200 columns 68 eluted irregularly throughout the entire fractionation range of the G—200 gel. This may indicate that the toxin at basic pH existed in forms of aggregates of different sizes. The adsorption or coupling of the toxin to blue dextran, to red blood cells, and to bacterial cells was demonstrated. These large molecular entities eluted in the void volume on Sephadex G-200 columns. Purified toxin mixed with these large molecular weight compounds eluted with them in the void volume. However, if bacterial cells and red blood cells were removed from the mixture by Mil- lipore filtration no toxic material appeared at the void volume. A possible aggregation effect of RNase was ob- served when toxin purified on DEAE was treated with the enzyme. The preparation treated with RNase eluted in a toxic and a non-toxic fraction on DEAE columns at pH 6.0. The toxic fraction separated into four bands upon electro- phoresis in pH 4.3 gels, while the non-toxic fraction separated.into two bands. The two slowest moving compo- nents, one of them representing the toxin, were absent in the non-toxic fraction. Rechromatography of the preparation treated with RNase was performed on calibrated Sephadex G-200 columns for size estimation, and an additional highly toxic frac- tion as well as the two equally toxic fractions were 69 produced. This highly toxic fraction eluted before aldolase and its estimated molecular weight was about 240,000. It appears that RNase treatment aggregated the toxin into a large unstable molecular aggregate, which upon storage for 3 to 4 days at 4 C disappeared by prob- ably again disaggregating into small units. Tests for hemagglutinating activity of type E toxin were negative, while crude type A and B toxin prep- arations were found to possess hemagglutinating activity as previously reported. The toxin of Clostridium botulinum type E Vancouver Herring strain was thus shown to be a low molecular weight protein of less than 12,000, based upon ammonium sulfate precipitation, DEAE cellulose and CM-Sephadex C-50 column chromatography. The aggregation of the toxin to large molecular weight substances was indicated and should be the subject of additional investigation. LITERATURE CITED LITERATURE CITED Abrams, A., G. Kegeles and G. A. Hottle. 1946. The puri- fication of toxin from Clostridium botulinum type A. J. Biol. Chem. 164, 63-79. Andrews, P. 1965. The gel-filtration behaviour of pro- teins related to their molecular weights over a wide range. Biochem. J. 22, 595-605. Boroff, D. A. 1959. Studies of the toxin of Clostridium botulinum. IV. Fluorescence of the toxin of Clostri- dium botulinum and its relation to toxicity. Intern. Arch. Allergy. 12, 74-90. Boroff, D. A. and B. R. DasGupta. 1964. Study of the toxin of Clostridium botulinum. VII. Relation of tryptophan to the biological activity of the toxin. J. Biol. Chem. 222, 3694-3698. Boroff, D. A. and B. R. DasGupta. 1966. Study of the toxin of Clostridium botulinum. Effects of 2-hydroxy- 5-nitrobenzy1-bromide on the biological activity of botulinum toxin. Biochem. Biophys. Acta. 111, 289-292. Boroff, D. A., B. R. DasGupta and U. S. Fleck. 1968. Homogeneity and molecular weight of toxin of Clostri- dium botulinum type B. J. Bacteriol. 22, 1738-1744. Boroff, D. A. and J. E. Fitzgerald. 1958. Fluorescence of the toxin of Clostridium botulinum and its relation to toxicity. Nature. 181, 751-752. Boroff, D. A. and U. Fleck. 1966. Statistical analysis of a rapid in vivo method for the titration of the toxin of Clostridium botulinum. J. Bacteriol. 22, 1580-1581. Boroff, D. A., R. Townend, U. Fleck and B. R. DasGupta. 1966. Ultracentrifugal analysis of the crystalline toxin and isolated fractions of g1gstridium botulinum type A. J. Biol. Chem. 221, 5165-5167. 70 71 Brieger, L. and W. Kempner. 1897. Beitrag zue 1ehre von der fleischvergiftung. Dt. Med. Wschr. 22, 521-526. Buehler, H. J., E. J. Schantz and C. Lamanna. 1947. The elemental amino acid composition of Clostridium botu- linum type A toxin. J. Biol, Chem. 169, 295-302. Cardella, M. A., J. T. Duff, C. Gottfried and J. S. Begel. 1958. Studies on immunity to toxins of Clostridium botulinum. IV. Production and purification of type C toxin for conversion to toxoid. J. Bacteriol. 75, 360-365. _ Cardella, M. A., J. T. Duff, B. H. wingfield and C. Gott- fried. 1960. Studies on immunity to toxins of Clos- tridium botulinum. VI. Purification and detoxifica- tion of type D toxin and the immunological response to toxoid. J. Bacteriol. 22, 372-378. Cartwright, T. E. and M. A. Lauffer. 1958. Temperature effects on botulinum A toxin. Proc. Soc. Exptl. Biol. Med. 22, 327-330. Coleman, G. E. and K. F. Meyer. 1922. Some observations on the pathogencity of 2, botulinus 2'. J. Infect. Dis. 21, 622-624. Duff, J. T., J. Klerer, R. H. Bibler, D. E. Moore, C. Gott- fried and G. G. Wright. 1957. Studies on immunity to toxins of Clostridium botulinum. II. Production and purification of type B toxin for toxoid. J. Bacteriol. 22, 597-601. Duff, J. T., G. G. Wright, J. Klerer, D. E. Moore and R. H. Bibler. 1957. Studies on immunity to toxins of Clos- tridium botulinum. I. A simplified procedure for iso- lation of type A toxin. J. Bacteriol. 22, 42-47. DasGupta, B. R. and D. A. Boroff. 1968. Separation of toxin and hemagglutinin from crystalline toxin of Clos- tridium botulinum type A by anion exchange chromatog- raphy and determination of their dimensions by gel fil- tration. J. Biol. Chem. 222, 1065-1072. DasGupta, B. R., D. A. Boroff and E. Rothstein. 1966. Chromatographic fractionation of the crystalline toxin of Clostridium botulinum type A. Biochem. Biophys. Res. Comm. 22, 750-755. 72 Davis, B. J. 1964. Disc electrophoresis. II. Method and application to human serum proteins. Annals of the New York Academy of Sciences. 121, 404-427. Fiock, M. A., A. Yarinsky and J. T. Duff. 1961. Studies on immunity to toxins of Clostridium botulinum. VII. Purification and detoxification of trypsin-activated type E toxin . J. Bacteriol. 22, 66-71. Gerwing, J., C. E. Dolman and D. A. Arnott. 1961. Puri- fication and activation of Clostridium botulinum type E toxin. J. Bacteriol. 21, 819-822. Gerwing, J., C. E. Dolman and D. A. Arnott. 1962. Acti- vation phenomenon of Clostridium botulinum type E toxin. J. Bacteriol. 22, 302-306. Gerwing, J., C. E. Dolman and H. S. Bains. 1965. Isola- tion and characterization of a toxic moiety of low molecular weight from Clostridium botulinum type A. J. Bacteriol. 22, 1383-1386. Gerwing, J., C. E. Dolman, D. V. Kason and J. H. Tremaine. 1966. Purification and characterization of Clostridium botulinum type B. toxin. J. Bacteriol. 21, 484-487. Gerwing, J., C. E. Dolman and A. Ko. 1965. Mechanism of tryptic activation of Clostridium botulinum type E toxin. J. Bacteriol. 22, 1176-1179. Gerwing, J., C. E. Dolman, M. E. Reichmann and H. S. Bains. 1964. Purification and molecular weight determination of Clostridium botulinum type E toxin. J. Bacteriol. 22, 216-219. Gerwing, J., A. Ko, D. Van Alstyne and J. H. Tremaine. 1966. Studies on the active region of botulinus toxins. I. Involvement and characterization of a peptide in the region of the single cysteine residue in botulinus toxins types A, B, and E. Biochem. Biophys. Acta. 111, 487-489. Gerwing, J., B. Mitchell and D. Van Alstyne. 1967. Stud- ies on the active region of botulinus toxins. II. Isolation and amino acid sequence of the cysteine-con- taining tryptic peptides in botulinus toxins types A, B, and E. Biochem. Biophys. Acta. 122, 363-365. Gerwing, J., R. W. Morrell and R. N. Nitz. 1968. Intra- cellular synthesis of Clostridium botulinum type B 73 toxin. I. Demonstration of toxin synthesis and sedi- mentation studies on toxic products. J. Bacteriol. 22, 22-27. Gordon, M., M. Fiock, A. Yarinsky and J. T. Duff. 1957. Studies on immunity to toxins of Clostridium botulinum. III. Preparation, purification and detoxification of type E toxin. J. Bacteriol. 12, 533-538. Gornall, G., G. Allan, Charles J. Bardowill and Maxima M. David. 1949. Determination of serum proteins by means of the Biuret Reaction. J. Biol. Chem. 177, 751-766. A. Lin. 1967. Properties of heat resis- tant toxin in spores of Clostridium botulinum 33A. "Botulism 1966" Chapman and Hall, London, pp. 302-311. Hauschild, A. H. W. Grecz, N. and C. and R. Hilsheimer. 1968. Hetero- geneity of Clostridium botulinum type A toxin. Can. J. Microbiolf12, 805-807. Heckly, R. J., G. J. Hildebrand and C. Lamanna. 1960. On the size of the toxic particle passing the intestinal barrier in botulism. J. Exptl. Med. 111, 745-749. Hildebrand, G. J., C. Lamanna and R. J. Heckly. 1961. Distribution and particle size of type A botulinum toxin in body fluids of intravenously injected rab- bits. Proc. Soc. Exptl. Biol. Med. 122, 284-289. Kabat, E. A. and M. M. Mayer. chemistry, pp. 282-291. 1948. Experimental immuno- field, Ill. Charles C. Thomas, Spring— Kempe, L. L. and J. T. Graikoski. 1962. Gamma ray steril- ization and residual toxicity studies of ground beef innoculated with spores of Clostridium botulinum. Appl. Microbiol. 12, 31-35. Kitamura, M., S. Sakaguchi and G. Sakaguchi. 1967. Dis- sociation of Clostrigium botulinum type E toxin. Bio- chem. Biophys. Res. Comm. 22, 892-897. Koshland, D. E., Y. O. Karkhanis and H. G. Latham. 1964. An environmentally sensitive reagent with selectivity for the tryptophan residue in proteins. J. Am. Chem. Soc. 22, 1448-1451. Lamanna, C. 1948. Hemagglutination by botulinal toxin. Proc. Soc. Exptl. Biol. Med. 22, 332-336. 74 Lamanna, C. 1959. The most poisonous poison. Science. 130, 763-772. Lamanna, C., H. W. Eklund and O. E. McElroy. 1946. Botu- linum toxin (type A): including a study of shaking with chloroform as a step in the isolation procedure. J. Bacteriol. 22, 1-13. Lamanna, C. and H. N. Glassman. 1947. The isolation of type B botulinum toxin. J. Bacteriol. 22, 575-584. Lamanna, C., O. E. McElroy and H. W. Eklund. 1946. The purification and crystallization of Clostridium botu- linum type A toxin. Science. 103, 613-614. Lowenthal, J. P. and C. Lamanna. 1951. Factors affecting the botulinal hemagglutination reaction and the rela- tionship between hemagglutinating activity and toxicity of toxin preparation. Am. J. Hyg. 22, 342-353. Lowry, O. H., N. J. Rosenbrough, A. L. Farr and R. J. Ran- dall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265-275. Putnam, F. W., C. Lamanna and D. G. Sharp. 1946. Molecu- 1ar weight and homogeneity of crystalline botulinus A toxin. J. Biol. Chem. 165, 735-736. Putnam, F. W., C. Lamanna and D. G. Sharp. 1948. Physio- Chemical properties of crystalline Clostridium botu- linum type A toxin. J. Biol. Chem. 176, 401-412. Sakaguchi, G. and S. Sakaguchi. 1961. A simple method for purification of type E botulinal toxin from the precursor extract of the bacterial cells. Jap. J. Med. Sci. and Biol. 12, 243-248. Sakaguchi, G. and S. Sakaguchi. 1967. Some observation on activation of Clostridium botulinum type E toxin by trypsin. "Botulism 1966" Chapman and Hall, London, pp. 226-277. Sakaguchi, G., S. Sakaguchi and N. Imai. 1964. Compara- tive gel filtration of toxin precursor and trypsin- activated toxin of Clostridium botulinum type E. J. Bacteriol. 22, 401-407. Schantz, E. J. and M. A. Lauffer. 1962. Diffusion measure- ments in agar gel. Biochemistry. 1, 658-663. 75 Schantz, E. J. and L. Spero. 1957. The reaction of botu- linum toxin type A with ketene. J. Am. Chem. Soc. 22, 1623-1625. Schantz, E. J., D. Stefanye and L. Spero. 1960. Observa- tions of the fluorescence and toxicity of botulinum toxin. J. Biol. Chem. 235, 3489-3491. Schneider, M. D., N. Grecz and A. Anellis. 1962. Concen- trated cultures of Clostridium botulinum spores in di- alysis sacs. 3. Toxic spores of Clostridium botulin- BE: United States Army Quartermaster Food and Con- tainer Institute. Project Nr. 7-84-01-002. Report No. 3. Snipe, P. T. and H. Sommer. 1928. Studies on botulinus toxin. 3. Acid precipitation of botulinus toxin. J. Infect. Dis. 22. 152-160. Sommer, E. W., H. Sommer and K. F. Meyer. 1926. The pur- ification of botulinum toxin. J. Infect. Dis. 22, 345-348. Spero, L. 1958. The alkaline inactivation of botulinum toxin. Arch. Biochem. Biophys. 22, 484-491. Spero, L. and E. J. Schantz. 1957. The reaction of botu- linum toxin type A with nitrous acid. J. Am. Chem. Soc. 12, 1625-1628. Stefanye, D., R. T. Iwamasa, E. J. Schantz and L. Spero. 1964. Effect of quanidinium salts on the toxicity of botulinum toxin. Biochim. et Biophys. Acta. 22, 412- 415. Van Alstyne, D., J. Gerwing and J. G. Tremaine. 1966. Amino acid composition of Clostridium botulinum type A toxin. J. Bacteriol. 22, 796-797. Vinet, G. and V. Fredette. 1951. Apparatus for the cul- ture of bacteria in cellophane tubes. Science. 114, 549-550. Wagman, J. 1954. Isolation and sedimentation study of low molecular weight forms of type A botulinus toxin. Arch. Biochem. Biophys. 22, 104-112. Wagman, J. 1963. Low molecular weight forms of type A botulinum toxin. II. Action of pepsin on intact and dissociated toxin. Arch. Biochem. Biophys. 100, 414- 421. 76 Wagman, J. and J. B. Bateman. 1951. The behaviour of the botulinus toxins in the ultracentrifuge. Arch. Bio- chem. Biophys. 21, 424-430. Wagman, J. and J. B. Bateman. 1953. Botulinum type A toxin: Properties of a dissociated product. Arch. Biochem. Bi0phys. 22, 375-383. Warburg, O. and W. Christian. 1942. Isolierung und kristallisation des garungsferments enolase. Biochem. Z. 310, 384-421. Weil, L., T. S. Seibles, L. Spero and E. J. Schantz. 1957. Photooxidation of crystalline Clostridium botulinum type A toxin in the presence of methylene blue. Arch. Biochem. Biophys. 22, 308-313. APPENDIX APPENDIX Stock Solutions for Negatively C arge Proteins A. To 48 m1 of 1 N HCl, 36.3 g of tris (hydroxy- methyl) aminomethane (TRIS), 0.23 ml of N, N, N', N'- Tetramethylethylenediamine (TEMED) and H20 were added to make 100 ml (pH 8.9 to 9.0). B. To 48 m1 of 1 N HCl, 5.98 g of TRIS, 0.46 ml of TEMED and H 0 were added to make 100 ml (pH 6.6 to 6.8). 2 C. To 60 g of acrylamide, 0.4 g of N, N'-Methyl- enebisacrylamide (BIS) and H 0 were added to make 135 ml. 2 D. To 10 g of acrylamide, 2.5 g of BIS and H O 2 were added to make 100 m1. E. To 4 mg of riboflavin, H O was added to make 2 100 ml. F. Catalyst: To 0.14 g of ammonium persulfate, H20 was added to make 100 m1 (prepared fresh weekly). G. Tracking Dye: 0.001 % bromphenol blue in H20. H. Protein stain: To 250 ml H O, 250 m1 methanol, 2 50 ml glacial acetic acid, and 2 g amido black were added. I. Buffer: To 6.0 g of TRIS, 28.8 g of glycine and H20 were added to make 1,000 m1 (pH 8.3). A 1/10 di- lution was used. 77 78 Stock Solutions for Positively Charged Proteins AA. To 48 ml of 1 N KOH, 17.2 ml of acetic acid, 4.0 ml of TEMED and H 0 were added to make 100 ml (pH 4.3). 2 BB. To 48 ml of 1 N KOH, 2.89 ml of acetic acid, 0.46 ml of TEMED, and H 0 were added to make 100 ml (pH 2 6.7). CC. Catalyst: To 0.28 g of ammonium persulfate H20 was added to make 100 ml (prepared fresh weekly). DD. Buffer: To 31.2 g of B-alanine, 8.0 ml of acetic acid and H20 were added to make 1,000 ml (pH 5.0). A 1/10 dilution was used. Working solutions were prepared as follows: Lower gel; 7.0 % - 1 part of A or AA, 1.4 parts of C and 2.1 parts of H O 2 Lower gel; 8.0 % - 1 part of A or AA, 1.6 parts of C and 1.9 parts of H20 were combined with catalyst in a 1:1 ratio. Upper gel; 1 part of B or BB, 2 parts of D, 1 part of E and 2 parts of H20 were mixed. MICHIGAN STATE UNIVERSITY LIBRARIES HHII HI ”III II 3 129313056 1694