Separation and Characterization of Heavy and Light Chains from CZostridium botulinum Type C Toxin and Their Reconstitution*

(Received for publication, June 18, 1980, and in revised form, October 22, 1980)

Bunei Syuto and Shuichiro Kubo From the Department of Biochemistry, Faculty of Veterinary Medicine, Hokkaido University, Sapporo 060, Japan

Clostridium botulinum type C toxin consists of a heavy and a light chain with molecular weights of 98,000 and 53,000, respectively, which are linked by one disulfide bond. The two components were separated from each other by quaternary aminoethyl Sephadex A-50 column chromatography by stepwise elution with NaCl in 21.5 mM borax-45 mM sodium dihydrogen phos- phate buffer, pH 8.0, containing 5% 2-mercaptoethanol at 0°C. The purified components had different amino acid compositions and antigenicities, and the toxicity of the toxin was neutralized completely by either anti- heavy chain Fab or anti-light chain Fab.

The two components could be reconstituted to form an active molecule with recovered toxicity which var- ied according to the method used. Maximum recovery was obtained in a system in which the intersubunit S- S bond was first formed in the presence of high concen- tration of neutral salts, after which the concentration of salt was gradually decreased. The reconstituted preparation was highly toxic and had the same prop- erties as the parental toxin on chromatography, sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and immunodiffusion. By the use of three perturbants, the fractions of exposed tryptophans and tyrosines of the preparation were found to be almost the same as that of the parental toxin.

Clostridium botulinum is known to produce the immuno- logically distinct neurotoxin types A, B, C, D, E, F, and G. Most of the toxins have a molecular weight of about 150,000 and consist of a heavy chain (Mr - 100,000) and a light chain ( M , - 50,000) linked by disulfide bond(s) (1-5). These common molecular features suggest that the action mechanism of all of the types is the same.

Recent evidence has demonstrated that different subunits of tetanus neurotoxin (Mr = 150,000) participate in the binding process and in toxicity. The subunits are different in immu- nogenicity (6-10). Similarities of the molecular structure and action on neuronal cells of botulinum neurotoxin and tetanus neurotoxin (11) prompted us to obtain more information on the two-chain structure of botulinum neurotoxin. This infor- mation will contribute to the elucidation of the molecular mechanism of its action on organisms.

In this paper we describe a highly reproducible method for the separation of the heavy and light chains from the type C toxin, the characteristics of these subunits, and finally, the

*?‘his work was supported in part by Grant 556192 from the Ministry of Education, Science and Culture of Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduer- tisement” in accordance with I 8 U.S.C. Section 1794 solely to indicate this fact.

mode of their reconstitution to form an active molecule. Preliminary reports have previously been made (12, 13).

EXPERIMENTAL PROCEDURES

Organism and Toxin Production-C. botulinum type C strain Stockholm was enriched as reported previously (14). The toxin was produced by the cultivation method reported previously (3).

Toxicity Assay-White mice (dd line) weighing about 20 g were injected intravenously with 0.1-ml portions of the toxin in 0.1 M sodium phosphate buffer a t pH 6.9. The time-to-death method (15) was applied using 6 mice for each determination.

Toxin Preparation-The toxin was purified by the method re- ported previously ( 3 ) with modifications. A step of gel fdtration with Sephadex G-200 was inserted after the three chromatographic pro- cedures and before the anti-hemagglutinin Sepharose 4B column, and the final step of quaternary aminoethyl (QAE) Sephadex A-50 chro- matography was 0-mitted. The toxin fraction eluting from the third column was concentrated by dialysis against saturated ammonium sulfate solution, and the dissolved toxin was passed through a column (3.2 X 55 cm) of Sephadex G-200 (superfine) equilibrated with Buffer A‘ to remove the large amount of hemagglutinin activity present. The collected fraction was then applied to a column (1.2 X 10 cm) of anti- hemagglutinin binding Sepharose 4B equilibrated with Buffer A. The toxin did not bind to the absorbent, whereas the remaining residual hemagglutinin activity was removed completely. The toxin was ho- mogenous by SDS’-polyacrylamide gel electrophoresis and the im- munodiffusion test.

Prepurations of Anti-Heavy Chain (or Anti-Light Chain) F a b a n d Anti-Heauy Chain (or Anti-Light Chain) Binding Sepharose 4B-A 300-mg portion of lyophilized horse anti-type C crude toxin serum (Ministry of Supply Station, England) was dissolved in 20 ml of 0.01 M sodium phosphate buffer a t pH 8.0 and applied to a column (1 X 3 cm) of heavy chain binding Sepharose, which was prepared by cou- pling 3 mg of purified heavy chain and 3.5 ml of CNBr-activated Sepharose 4B by the method of Livingston (16). After the column- bound anti-heavy chain IgG was washed with 10 column volumes of 0.01 M sodium phosphate buffer a t pH 8.0 to remove anti-light chain IgG and other contaminants, anti-heavy chain IgG was eluted with 3 M KSCN and dialyzed against Buffer A. The immunodiffusion test did not detect any anti-light chain in the eluate. Anti-heavy chain Fab was then prepared from the papain digest of this anti-heavy chain by the method of Porter (17), and, using Livingston’s method

of 3 mg of the IgG with 3.5 ml of CNBr-activated Sepharose 4B. (16) anti-heavy chain binding Sepharose was obtained by the coupling

Anti-light chain IgG was also prepared from the anti-type C crude toxin by using light chain binding Sepharose 4B, as described in the method above. After determining that no anti-heavy chain was pres- ent in the eluted anti-light chain fraction, anti-light chain Fab and anti-light chain binding Sepharose 4B were prepared from the IgG by the methods of Porter (17) and Livingston (16), respectively.

Amino Acid Analysis-Analyses were made with a Nihon Uensi JLC-GAH amino acid analyzer. Hydrolyses were performed in glass- distilled HCI a t 110 “C for 20 h on the heavy and the light chains and for 20, 40, and 72 h on the toxin. Corrections were made for the

1 Buffer A was prepared as Kolthoffs buffer by mixing in the proportion of 55 ml of 0.05 M borax and 45 ml of 0.10 M NaHzPO, to give pH 8.0 at 18°C.

-? The abbreviations used are: SDS, sodium dodecyl sulfate; MLD, minimal lethal dose.

3712

Structure and Toxicity of Clostridium botulinum Type C Toxin 3713

unstable and slowly liberated amino acids in the toxin from the results of the kinetics of hydrolysis (18).

In the presence of 1% SDS the free sulthydryl group was reacted with a 150" excess of 5,5"dithiobis(2-nitrobenzoic acid) over total amounts of half-cystine residues of toxin following the method of Baillie and Horowitz ( 19).

Protein Determination-Protein was routinely estimated by the method of Lowry et al. (20) with bovine serum albumin as standard. The concentrations of toxin and the heavy and light chains at neutral pH were routinely estimated by absorbance at 258 nm usingA !':.,,, = 14.18 (3),A!';.", = 12.02. and A!';.", = 9.03, respectively. The absorbance coefficient of each subunit was calculated from the amino acid com- position. During chromatographic separation of the heavy and light chains from the toxin, the protein was estimated densitometrically after SDS-polyacrylamide gel electrophoresis by the method of Kahn and Hubin (21).

Solvent Perturbation Method-Difference spectral measurements were taken using a Hitachi 556 recording spectrophotometer equipped with a base-line corrector and a scale-expanding attachment to give 0.1 absorbance unit equal to full scale. Two pairs of matched I-cm cells were used instead of the tandem double cells. All liquid pertur- bants were diluted to 40% (v/v) immediately before use. and 43.05 g of glucose was dissolved in water to 100 ml for a final 40V solution (w/w). A stock solution of protein in Buffer A and its control solution containing the same constituents, except protein, were prepared. Carefully measured 2-ml portions of stock protein solution were added, the first to 2 ml of water, and the second to 2 ml of the particular 40% perturbant was employed. The exact same procedures were applied to the control solution. The spectra were obtained within 15 min after mixing and equilibrating the solvents at 25°C. All measurements were made in triplicate. N-Acetyltryptophan ethyl ester (Ac-Trp-OEE) and N-acetyltyrosine ethyl ester (Ac-Tyr-ONE) were used as model compounds, and their concentrations were deter- mined spectrophotometrically using the following molar extinction coefficients, 5570 M 'cm I for Ac-Trp-OEE at 280 nm and 1415 M 'cm I for Ac-Tyr-OEE at 275.5 nm. These values were determined in the presence of half-concentration of Buffer A.

RESULTS

Separation and Purification of Heavy a n d Light Chains from Toxin-Fig. 1 shows the separation profile of the heavy and light chains obtained from the purified toxin after QAE- Sephadex A-50 chromatography. After determining the eluted fractions by SDS-polyacrylamide gel electrophoresis, it was found that the light chain in the first peak and the heavy chain in the third peak were cross-contaminated with traces

Elution Volume (ml)

FIG. 1. Chromatographic separation of heavy and light chains from toxin on a column (1.2 X 8 cm) of QAE-Sephadex A-50 at 0 OC. The toxin (15 mg) was applied to a column previously equilibrated with Buffer A and washed with 18 ml of the same buffer at a flow rate of 10 ml/h. Stepwise elution with additions to the above buffer was performed according to the following schedule. 1) 18 ml of 50 mM dithiothreitol in Buffer A was used at a flow rate of 10 ml/h. 2) After application of 5 ml of 5% 2-mercaptoethanol in Buffer A, the column was aged for 5 h, and then the elution was continued with 36 ml of the same solution at a flow rate of 6 ml/h. 3) 36 ml of 0.12 M NaCI-5% 2-mercaptoethanol in Buffer A was used a t a flow rate of 15 ml/h. 4) 36 ml of 0.2 M NaCI-58 2mercaptoethanol in Buffer A was used at a flow rate of 15 ml/h. Each fraction was 6 ml.

of undissociated toxin of the second peak. Each chain fraction was further purified by rechromatography on the column previously equilibrated with Buffer A containing 54 2-mer- captoethanol. The overall yields of both the heavy and light chains were about 70% of parental toxin. Fig. 2A shows that each of these chains was homogeneous. The scant trace of toxicity, 200-500 MLD/mg, which corresponded to 0.5-1.1 x lo-" of the original toxicity of the parental toxin (4.4 X 10' MLD/mg), could not be removed from each chain component at this stage by the chromatography. However, these contam- inating toxicities were completely removed from the heavy and light chains preparations by passing them through col- umns of anti-light chain and anti-heavy chain binding Seph- arose 4B, respectively.

Amino Acid Compositions of Heauy a n d Light Chains- The amino acid compositions of heavy chain, light chain, and toxin are presented in Table I together with those of type B toxin for comparison. Ultracentrifugal analysis revealed that the toxin had a molecular weight of 141,000. The toxin disso- ciated into light and heavy chains with molecular weights of 53,000 and 98,000, respectively, by SDS-polyacrylamide elec- trophoresis in the presence of 2-mercaptoethanol (3). The molecular weights of the subunits in Table I were corrected by multiplying the values by 141,000/151,000. The sum of all of the amino acid residues in the heavy and light chains agreed well with that in the parental toxin. The amino acid distri- bution pattern of the heavy chain differed from that of the light chain, in that the heavy chain contained predominantly acidic and neutral amino acids, and the light chain contained more of the basic amino acids. There were 4 mol and 2 mol of half-cystine residue in the heavy chain and the light chain, respectively. When the toxin was treated with 5,5'-dithiobis(2- nitrobenzoic acid) in the presence of SDS, 3.8 mol (4 mol) of

A B

1 2 3 4 1 2 3 4 FIG. 2. Sodium dodecyl sulfate-polyacrylamide gel electro-

phoretic patterns of toxin, reconstituted toxin, and heavy and light chains. A, parental toxin. 1, toxin in the absence of 2-mercap- toethanol; 2, toxin in the presence of 2-mercaptoethanol: 3, heayv chain; 4, light chain. Protein (50 pg) was loaded on a gel. B. reconsti- tuted toxin from heavy and light chains. 1, reconstituted toxin in the absence of 2-mercaptoethanol; 2, reconstituted toxin in the presence of 2-mercaptoethanol; 3, heavy chain used for the reconstitution; 4. light chain used for the reconstitution. Protein (20 pg) was loaded on a gel. The electrophoresis was carried out by the method of Weber et al. (22).

3714 Structure and Toxicity of Clostridium botulinum Type C Toxin

TARLE I Amino acid composition of heavy chain, light chain, and toxin

The data was obtained from a single analysis of heavy and light chains and dual analyses of toxin. The values in the square brackets represent the sum of the residues in the heavy and light chains.

Heavy chain I.iaht chain Toxin Tipe 13 toxin (2)

Amino acid

- -~~_____ Lysine Histidine Arginine Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Valine Methionine Isoleucine Leucine Tyrosine Phenylalanine Tryptophan Half-cystine Total amide

53 6.7 5

28 0.6 3.6

148 18.8 43 5.5 60 7.6 80 19

10.2 2.4

40 5.1 25 3.2 42 5.3 15 1 .9 73 9.3 57 7.3 46 5.8 42 5.3

7' 0.9 4" 0.5

I<esidues Per cent t o per 5O.(XM) total resi-

R dues (4:lIi)

28 6.4 3 1 . 1

20 4.7 76 28

17.4

29 6.4 6.8

39 25

8.9

20 5.7 4.6

18 21

4.1 4.8

5 1 . 1 38 8.7 32 22

7.3

26 5.0 6.0

2' 2"

0.5 0.5

- -~

Residues per 141.000 R

78 [81] IO r lo l 47 i48i

225 [224]

88" [89]

41 [44]

44 [43]

78" [71]

116 [I191

59 [SO]

65" [63] 16 ['LO]

114* [ I l l ] 91 [89] 71 [68] 70 [68] 11' [9] fi" [S]

(147)

Per cent to to- tal residues

Residues per Per cent to to-

,(j7,(o I: tal residues

I1434 b

6.3 0.8 3.8

18.3 6.3 7.2 9.4 3.3 4.8 3.6 5.3 1.3 9.3 7.4 5.8 5.7 0.9 0.5

(1 1.9)

123 8.6 9 0.6

45 3.1 229 16.0 65 4.6

103 7.2 149 10.4 45 3.1 70 4.9 54 3.7 64 4.4 27 1.9

147 10.2 99 81

6.9 5.7

77 36

5.4 2.5

1 1 0.8 (156) (10.9)

Polar amino acid Acidic and neutral 53.3 Basic 10.9

Nonpolar amino acid 35.6

49.7 12.1 38.2

52.3 10.9 36.8

49.6 12.3 38 .1

" Extrapolated to zero time. ' I Values are for 72-h hydrolysis.

" Iletermined as cysteic acid after oxidation with performic acid (24). Determined by spectrophotometric method of Goodwin and Morton (23).

FIG. 3. Immunodiffusion of toxin, reconstituted toxin, heavy and light chains. Center ulell, horse anti-type C crude toxin serum (Ministry of Supply Station, England); I , toxin; 2 and 5, heavy chain; 3 and 6, light chain; 4, reconstituted toxin from heavy and light chains. Protein (10 p g ) was added to wells 1-4. The gel was Buffer A containing 1% special agar B (Wako) at pH 8.0.

the free " S H group were titrated. The amino acid distribution pattern of the toxin was similar to that of the type B toxin with the exception of tryptophan.

Immunological Properties-In the double diffusion test, all of the toxin and the heavy and light chains produced a single precipitin line with the horse anti-type C toxin serum (Fig. 3). The precipitin lines of the heavy and light chains formed spurs with those of the parental toxin. However, the two lines crossed without interaction, indicating that the two chains differed in antigenicity. Table I1 shows the neutralization of toxin by anti-subunit Fab. The toxicity of the parental toxin was neutralized completely by either anti-heavy chain Fab or anti-light chain Fab, suggesting that the heavy and light chains were integral for the toxicity of the toxin molecule.

Reconstitution of Toxin Molecule from Heavy a n d Light

TABLE I1 Neutralization of toxin hy anti-heal'y chain Fah and anti-light

chain Fah Each mixture was made up of 0.1 ml with 0.05 ml of toxin (5 X 10'

MLD, 1.14 pg) in Buffer A a t pH 8.0 and 0.05 ml of anti-subunit Fab (25 pg) in the same buffer. After incubation at 30°C for 30 min. the toxicitv was assaved bv the time-to-death method (15).

Mixture Toxicity M L I )

Toxin + anti-heavy chain Fab 0 Toxin + anti-light chain Fab 0 Toxin + buffer 5 x IO4

Chains-Table 111 shows the method of reconstitution and the recovered toxicity from an equimolar mixture of the two subunits. The recovered toxicity varied according to the method. Maximum recovery was obtained in the system (Ta- ble 111, sample 6) in which the intersubunit S-S bond was formed first in the presence of high concentration of neutral salts, followed by the gradual removal of neutral salts. The reconstituted toxin in the system was further purified by chromatography on a QAE-Sephadex A-50 column (Fig. 4). The most toxic fraction was eluted at 0.1 M NaCl in the same aliquot as the parental toxin under similar conditions of NaCl gradient. The toxicity per of the peak fractions reached the level of 2.2 X 10' MLD/Az;,, which corresponds to 70% of the specific toxicity of the parental toxin, 3.1 X 10' MLD/Az;n (3). This major protein fraction having values of MLD/Az'" = 2.2 X 10' showed the same mobility as the parental toxin in SDS-polyacrylamide gel electrophoresis without 2-mercapto- ethanol, and it was separated into a equimolar ratio of heavy and light chains in the presence of 2-mercaptoethanol during

Sam- Dialyzing system P'e

1. 20 mM DTTh-5 mM NHlHCOa for 48 1 h, Buffer A at pH 8.0 for 72 h 2. 1 Saturated (NH,)2SO, solution at pH

1 NH4HC0.3 for 24 h, Buffer A at pH 8.0 for 24 h, 20 mM DTT-5 mM

1 8.0 for 72 h 3. 5 mM NHAHCOn for 48 h, Buffer A at

pH 8.0 for 72 h

8.0 for 24 h, 5 mM NHdHC0.j for 24 h, Buffer A at pH 8.0 for 72 h

mM NH4HCOs for 24 h, Buffer A at pH 8.0 for 72 h

6. Saturated ( N H M O I a t p H 8.0 for 24 h, 1 M (NH,),SO, at pH 8.0 for 24 h, 5 mM NH,HCO1 for 24 h. Buffer A

4. Saturated (NH&&04 solution at p H

5. 1 M (NH,)rSO, at pH 8.0 for 24 h, 5

I at pH 8.0 for 72-h

Toxicity"

MLDlAm x 10:

0.05-0.10

0.10-0.15

0.20-0.30

1.00-1.30

1.20-1.50

1.50-2.20

~~

tecovered"

B

1.6-3.2

3.2-4.8

6.5-9.7

32.2-41.9

38.7-48.4

48.4-71.0

i " Values were obtained after reconstitution experiments done in

triplicate. DTT, dithiothreitol.

_.-

30 60 90 Elutiwl Volume (ml)

FIG. 4. Chromatographic purification of the reconstituted toxin of sample 6, Table I11 on a column (1.2 X 8 cm) of QAE- Sephadex A-50. Protein (10 mg) was applied t.o the column at a flow rate of 15 ml/h. Elution was performed with a linear gradient of 0-0.2 M NaCl in Buffer A, at 5"C, and at a flow rate of 12 ml/h. Each aliquot was 3 ml.

electrophoresis (Fig. 2B). The molecule also showed the same pattern as the parental toxin in immunodiffusion (Fig. 3).

Low recovery toxicities were obtained in the systems in which the formation of the S-S bond was retarded fist by the addition of a reducing reagent and/or by the rapid removal of the effects of high concentration of neutral salts (samples 1,2, and 3 in Table 111).

Degree of Exposure of Chromophores-The solvent pertur- bation difference spectra of the parental toxin and the major protein fraction having values of MLD/An;* = 2.2 X 10' was measured with dimethylsulfoxide, glycerol, and glucose. The degree of exposed tyrosine and tryptophan chromophores was determined by solving the two simultaneous equations re- ported by Herskovits and Sorensen (25, 26) and by adjusting

o i J 2 70 280 790 300

Wavelength ( n m 1

0 270 2 80 290 3 00

Wavelengthhm)

FIG. 5. Comparison of solvent perturbation difference spec- tra of toxins produced by 20% dimethylsulfoxide with calcu- lated curves based on the results of Table IV. --, experimental data; --- calculated curves. The medium contained half-concentra- tions of Buffer A at pH 8.0 and 3.2 X IO"' M toxin. A , native toxin. B. reconstituted molecule which attained the high recovered toxicity (2.2 X 10' MLD/Au").

the difference spectral profile of the toxin with the appropriate combinations of tyrosine and tryptophan curves based on model experiments suggested by Donovan (27).

Fig. 5 shows comparisons of the difference spectra of the toxin and the reconstituted molecule with the calculated curves of model compounds in 20% dimethylsulfoxide as an example. The curves of proteins were well fitted to those of the model compounds with the exception of the poor fit around and below 280 nm, which may have been caused by the preferential interaction of the perturbant with the poly- peptide chains, as suggested by Herskovits and Laskowski (28) and Riddiford (29). Table IV summarizes the data ob- tained from the parental toxin and the reconstituted molecule with the three perturbants. The exposed chromophore frac- tions of the reconstituted toxin were almost the same as that of the parental toxin, except the value of tryptophan in 20% glucose. No apparent gradation in exposure with increasing perturbant diameter was observed in either toxin.

3716 Structure and Toxicity of Clostridium botulinum Type C Toxin

TABLE IV Difference spectralparameters of native toxin (3.1 X 10' MLD/Azid and reconstituted molecule with high recovered toxicity (2.2 X 10'

MLD/Aud

Molar absorptivity difference, A m " ~ ~ ~ ~ ~ ~ ~ ~ ~ $ ! Fraction of residues ex- based on eaiculation ~~

posed" _______

Perturbants Mean diameter Calculated

291.5-292.5 284-286.5 291.5-292.5 284-286.5 Trp T F Trp Tyr nm nm nm nm

A Native 20% Dimethylsulfoxide 4.0 3903 7236 3835 7190 3.0 32.0 0.27 0.45 20% Glycerol 5.2 1120 2399 1119 2494 4.0 20.0 0.36 0.28 20% Glucose 7.2 530 1495 454 1895 5.5 26.0 0.50 0.37

Reconstituted 20% Dimethylsulfoxide 4.0 3872 6919 3877 6916 3.0 31.0 0.27 0.44 20% Glycerol 5.2 1174 2634 1169 2623 4.0 23.0 0.36 0.32 20% Glucose 7.2 127 1333 122 1326 4.0 25.0 0.36 0.35

I' Calculations were based on 11 tryptophans and 71 tyrosines per 141,000 g of protein.

DISCUSSION

Kozaki et al. (30) have succeeded recently in separating two fragments with molecular weights of 111,000 and 59,000 from urea-treated trypsinized type B neurotoxin (Mr = 170,000) by gel fitration in the presence of urea and a reducing agent. Their fragments contained 0.04-0.4% of the original toxicity of the parental toxin, and the large fragment retained two minor fragments which may have derived from the parental toxin digested by trypsin. In their study, the recovery of toxicity by reconstituting their fragments was 34% of the parental toxin. Our purification procedure presented here effectively eliminated the original toxicity in the two subunits to 0.5-1.1 X lo-' of the parental toxin and yielded homogenous subunits with an overall recovery of 70%. The recovery of specific toxicity by reconstituting the subunits was about 70% of the parental toxin after purifying the reconstituted mole- cule. In the separation and reconstitution procedures, the presence of a high concentration of reducing agents and the adjustment of the column temperature to 0 "C were essential; otherwise, the yield of the recovered two subunits decreased, and the light chain aggregated and, thus, lost its solubility. Although type B and C toxins were similar in molecular construction and amino acid composition, type C toxin lost its toxicity in the presence of urea and became aggregated after removal of urea by dialysis. There was no basis upon which we could attribute the two distinct behavior patterns in urea to differences in the species or the toxins treated with or without trypsin.

The heavy and light chains differed from each other in amino acid composition and antigenicity. The amino acid composition and the chromatographic elution pattern showed that the heavy chain was more acidic than the light chain. There were six half-cystine residues in the parental toxin, four in the heavy chain, and two in the light chain. There were four titratable-SH groups per mol of the toxin. These find- ings show that the toxin consists of a heavy and a light chain with three free-SH groups and one free-SH group, respec- tively, and that they are combined by one disulfide bond. Accordingly, from the results of the complete neutralization of the toxicity by each of the anti-subunit Fab fragments, it was clear that each of the subunits was essential for the toxicity of the toxin. These two subunits could be reconsti- tuted to form an active molecule by a method in which the intersubunit S-S bond was first formed, and then the ionic effects of dissolved salts of the charged groups on the toxin molecule were gradually decreased.

In the solvent perturbation different spectra of the parental

toxin and the reconstituted molecule which was formed from a 1:l ratio of the subunits, the calculated values of the exposed fraction of chromophores showed no apparent decrease with increase of the perturbant diameter. These results suggest that a small fraction of the chromophores is partly buried, similar to the partly buried groups in ribonuclease reported by Herskovits and Laskowski (31). In our study the exposed residues in the parental toxin were 3-5 tryptophyls and 20-32 tyrosyls, and perhaps 1-2 tryptophyls and 12 tyrosyls were partly buried. In the reconstituted molecule 3-4 tryptophyls and 23-32 tyrosyls were exposed, and perhaps 1 tryptophyl and 8 tyrosyls were partly buried. Since the toxin contains a large number of tyrosine residues, 71 mol of tyrosine per mol of toxin, the calculated values of exposed tyrosine varied. From the solvent perturbation studies we conclude that the exposed surfaces of the subunits of the reconstituted molecule are very similar to that of the parental toxin.

Most bacterial exotoxins, e.g. cholera toxin, tetanus toxin, and diphtherial toxin, are composed of two functionally dif- ferent protomers linked by a disulfide. One of the protomers interacts with the binding site of the cell surface and the other shows the actual toxicity to the cell (32). On the basis of the botulinum toxin-induced blockade of neuromuscular trans- mission and the inhibition of the blockade by anti-toxin, Sirnpson has proposed that there is an initial binding step which does not impair transmission (33). Our results show clearly that C. botulinum toxin is formed by two different subunits linked by a disulfide and that these chains are integral for the toxicity of the toxin molecule. These subunits may be useful for determining the steps in the relationship of function to toxicity.

REFERENCES 1. DasGupta, B. R., and Sugiyama, H. (1972) Biochem. Biophys.

2. Beers, W. H., and Reich, E. (1969) J. Biol. Chem. 244,4473-4479 3. Syuto, B., and Kubo, S. (1977) Appl. Enuiron. Microbiol. 33,400-

4. Miyazaki, S., Iwasaki, M., and Sakaguchi, G. (1977) Zfect. Im-

5. Yang, K. H., and Sugiyama, H. (1975) Appl. Microbiol. 29, 598-

6. Kryshanovsky, G. N. (1973) Naunyn-Schmiedeberg's Arch. Phar-

7. Habermann, E. (1973) Med. Microbiol. Immunol. 159, 80-100 8. Robinson, J. P., Picklesimer, J. B., and Puett, D. (1975) J. B i d .

9. Helting, T. B., and Zwisler, 0. (1977) J. Biol. Chern. 252, 187-193 10. Helting, T. B., Zwisler, O., and Wiegandt, H. (1977) J. Biol. Chem.

Res. Commun. 48, 108-112

405

mun. 17,395-401

603

macol. 276,247-270

Chem. 250,7435-7442

Structure and Toxicity of Clostridium botulinum Type C Toxin 3717

11.

12.

13.

14. 15. 16. 17. 18. 19.

20.

21. 22.

252, 194-198 Bigalke, H., Dimpfel, W., and Haberman, E. (1978) Naunyn-

Syuto, B., and Kubo, S. (1978) Jpn. J. Med. Sci. Biol. 31, 169-

Syuto, B., and Kubo, S. (1979) Jpn. J. Med. Sci. Biol. 32, 132-

Syuto, B., and Kubo, S. (1972) Jpn. J. Vet. Res. 20, 19-30 Boroff, D. A., and Fleck, U. (1966) J. Bucteriol. 92, 1580-1581 Livingston, D. M. (1974) Methods Enzymol. 34, 723-731 Porter, R. R. (1959) Biochem. J. 73, 119-126 Moore, S., and Stein, W. H. (1968) Methods Enzymol. 6, 819-831 Baillie, R. D., and Horowitz, P. M. (1976) Biochim. Biophys. Acta

Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.

Kahn, R., and Rubin, R. W. (1975) Anal. Biochem. 67, 347-352 Weber, K., Pringle, J. R., and Osborn, M. (1972) Methods En-

Schmiedeberg’s Arch. Pharmacol. 303, 133-138

1 70

133

249, 383-390

(1951) J. Biol. Chem. 193,265-275

23.

24. 25.

26.

27. 28.

29. 30.

31.

32. 33.

zymol. 26,3-27

632 Goodwin, T. W., and Morton, R. A. (1946) Biochem. J . 40,628-

Hirs, C. H. W. (1956) J. Biol. Chem. 219,611-621 Herskovits, T. T., and Sorensen, M. (1968) Biochemistry 7,2523-

Herskovits, T. T., and Sorensen, M. (1968) Biochemistry 7, 2533-

Donovan, J . W. (1964) Biochemistry 3, 67-74 Herskovits, T. T., and Laskowski, M., Jr. (1962) J. Biol. Chem.

Riddiford, L. M. (1966) J. B i d . Chem. 241, 2792-2802 Kozaki, S., Miyasaki, S., and Sakaguchi, G. (1977) Infect. Immun.

Herskovits, T. T., and Laskowski, M., Jr. (1968) J. Biol. Chem.

Wiegandt, H. (1979) Ado. Cytopharmacol. 3, 17-25 Sirnpson, L. L. (1979) Adu. Cytopharmucol. 3, 27-34

2532

2542

237, 3418-3422

18, 761-766

243,2123-2129