THE JOURNAL 01 BIO~GICAL Cz-mslrsTP.Y Vol. 252, No. 1, Issue of January 10. PP. 137-193. 1977

Printed in U.S.A.

Structure of Tetanus Toxin I. BREAKDOWN OF THE TOXIN MOLECULE AND DISCRIMINATION BETWEEN POLYPEPTIDE

FRAGMENTS (Received for publication, April 21, 1976)

TORSTEN B. HELTING* AND OSWALD ZWISLER

From the Behringwerke AG, D-3550 MarburglLahn, Federal Republic of Germany

Tetanus toxin was digested with papain, yielding one major polypeptide (Fragment C) with a molecular weight corresponding to 47,000 f 5%, thus comprising about one-third of the toxin molecule. Fragment C was antigenically active, atoxic, and stimulated the formation of antibodies neutralizing the lethal action of tetanus toxin in ho. Furthermore, a second split product (Frag- ment B) was isolated from the papain digest, containing two polypeptide chains linked together via a disulflde bond. Fragment B (M, = 95,000 f 5%) was atoxic and showed a reaction of nonidentity with Fragment C on immunodiffusion analysis against tetanus antitoxin.

The basic two-chain structure (heavy and light chain polypeptide, cf. Matsuda, M., and Yoneda, M. (1975) Infect. Zmmun. 12, 1147-1153) of tetanus toxin has been confirmed and the relationship between Fragments B and C within this framework has been established. Fragment C was distin- guished from the light chain by electrophoresis in sodium dodecyl sulfate and by immunodiffusion analysis, indicating that this fragment constitutes a portion of the heavy chain polypeptide. Frag ment B showed a reaction of partial identity with the light as well as the heavy chain from tetanus toxin. Reduction of Fragment B with dithiothreitol followed by gel chromatography yielded a fraction which was indistinguishable from the light chain portion of the toxin molecule. It is concluded that Fragment B comprises the complementary portion of the heavy chain (remaining after scission of the polypeptide bond(s) releasing Fragment C) linked to the light chain by a disulfide bond.

Conflicting evidence regarding the molecular structure of tetanus toxin has emerged during recent years (l-4). On the basis of gel chromatography studies Murphy et al. (2) con- cluded that the toxin consists of a single polypeptide chain with a molecular weight of about 150,000. Bizzini et al. (3) described a number of fragments obtained after nonenzymatic treatment of tetanus toxin and presented a model involving two identical subunits, each consisting of two polypeptide chains with molecular weights corresponding to 21,000 and 52,000, respectively. In contrast, Craven and Dawson (4) pro- posed that the toxin from culture filtrates consists of only two nonidentical polypeptide chains with molecular weights corre- sponding to 55,000 and 100,000, respectively, whereas intracel- lularly located toxin constitutes a single polypeptide chain. Matsuda and Yoneda (5, 6) confirmed and extended these findings, showing that the intracellular toxin may be “nicked’ by mild trypsin treatment, yielding a product with properties indistinguishable from those of the extracellular toxin,

Detailed knowledge on the structure of tetanus toxin may be essential for the elucidation of the mechanism of action of this protein, which exerts profound neurotoxic effects at minute levels of material (1). Recent evidence has accumulated that different sites of the toxin molecule may be implicated in the binding process to brain matter, toxicity, and immunogenicity (7-9).

In the present communication, a highly reproducible method of degrading tetanus toxin with papain to yield an atoxic, immunogenic polypeptide (Fragment 0 is described. In addition, the complementary portion (Fragment B) formed when Fragment C is released from tetanus toxin has been identified. Finally, the polypeptide chain structure of tetanus toxin is contlrmed (4, 5) and a procedure is presented for the preparation of stable derivatives of its composite chains. Evi- dence is submitted that Fragment C is derived from the heavy chain polypeptide of the tetanus toxin molecule. Preliminary reports have appeared (10, 11).

EXPERIMENTAL PROCEDURES

Materials- Papain (EC 3.4.4.10), 30 units/mg, was purchased from Boehringer, Mannheim, Germany, or (type II; 2 units/mg) from Siema Chemical Co.. St. Louis. MO. The latter oroduct was extracted &h cyateine and fractionatei as described by Kimmel and Smith (12). Sephadex or Sepharose gels were obtained from Phannacia Fine Chemicals, Uppsala, Sweden. Ultrogel AcA 44 was a product of LKB, Bromma, Sweden. Molecular weight standard proteins were supplied by Boehringer.

Tetanus toxin was isolated from culture filtrates of Clostridium tetani grown on Latham medium essentially as described (13, 14). Fractions homogeneous on electrophoretic analysis in polyacryi- amide gel were pooled and chromatographed on Sephadex G-100 to

* To whom correspondence should be directed.

187

by guest on March 5, 2019

http://ww

w.jbc.org/

Dow

nloaded from

188 Structure of Tetanus Toxin. I

remove a small amount of partially aggregated protein (see Fig. la). Analytical Methods- Protein was determined according to Lowry

et al. (15) with bovine serum albumin as standard. Polyacrylamide gel electrophoresis in SDS buffer was carried out according to Weber and Osborn (16). Prior to application to the gels, samples were made 1% with respect to SDS and heated in a boiling water bath for 1 min. Double diffusion analysis in agar gel was performed by the method of Ouchterlony (17). Amino acid analysis was performed with a Beck- man multichrom B autoanalyzer by the procedure of Moore et al. (18). Total cysteine was determined as cysteic acid (19) and trypto- phan by the method of Edelhoch (20). Preparative electrophoresis in polyvinylchloride was performed in barbital buffer, pH 8.6, at room temperature for 48 h (4 V/cm).

Preparation ofInsoluble Reagents-Attachment of papain to Seph- arose 4B gel was achieved essentially as described by Cuatrecasas (21). Papain (500 ml; 50 mg/ml) was dialyzed against 0.1 M Na,CO, buffer, pH 10.0, and allowed to react with cyanogen bromide-acti- vated agarose gel (1000 ml; 200 mg of BrCN/ml). The washed, acti- vated gel was stirred gently with the protein for 24 h at 4” and was subsequently washed with 5 liters of 4 M urea, 0.5 M NaCl followed by 20 liters of 0.1 M phosphate buffer, pH 6.5, containing 0.001 M Na,EDTA.

Immunoadsorbents suitable to remove toxic impurities from prep- arations of Fraction C (see below) were prepared as follows. Three times chromatographed Fraction C (10 mg, corresponding to 10,000 Lf) was mixed with horse tetanus antitoxin (10,000 LB and the resulting precipitate was removed by centrifugation. The superna- tant, containing approximately 5,000 Lf when tested against native toxin, was dialyzed against 0.1 M Na,CO,, pH 10.0, and attached to 20 ml of Sepharose gel activated with cyanogen bromide in a manner similar to the procedure described above.

Antisera - Antibodies to tetanus toxoid or fragments derived from tetanus toxin were prepared by a sequence of three subcutaneous injections of the antigen (total amount, 0.3 mg of protein) in Freund’s complete adjuvant. The sera were analyzed by the double diffusion technique and their protective potency was determined as described below.

In Viuo Tests-Toxicity was determined by injection of dilutions of tetanus toxin or fragments into NMRI mice weighing 14 to 16 g. The diluent (0.15 M NaCl) was supplemented with protective colloid to prevent adsorption of the diluted material to glass. It was considered that animals dying on Day 4 following injection had received 1 MLD of tetanus toxin. Immunization with Fragment C was carried out in guinea pigs (350 g) by injection of varying doses of antigen as described in detail in the legend to Table III. The potency of antisera was estimated by the L+ method (22). Dilutions of the sera were mixed with a standard amount of tetanus toxin and the protection observed was compared with that of a reference serum (22).

Digestion of Tetanus Toxin with Papain - The digestion was per- formed essentially as described (10). Experimental details are given in the legend to Fig. 1. All fractions were concentrated and rechro- matographed twice in order to remove impurities.

Fraction B was purified further by passage through a column (1.5 x 10 cm) of Sepharose 4B to which anti-Fragment C serum had been attached* (for details on the coupling procedure see above). In a similar manner, Fraction C was passed through a column with covalently attached tetanus antitoxin which had been prepared by previous absorption with highly purified Fraction C. The eluates of both affinity columns were concentrated and subjected to a final chromatography on a Sephadex G-100 column (5 x 100 cm) yielding atoxic Fragments B and C, respectively.

Preparation of Heavy and Light Chains and Their Stabte Deriua- tioes-The separation of the heavy chain from the light chain of tetanus toxin was performed essentially as described by Matsuda and Yoneda (6). Urea-treated reduced toxin was passed through a column (2.5 x 100 cm) of Ultrogel AcA 44, equilibrated with 2 M urea

1 The abbreviations used are: SDS, sodium dodecyl sulfate; Lf, flocculation unit, i.e. that amount of toxin which shows the fastest flocculation with 1 unit of tetanus antitoxin (22); MLD, minimal lethal dose; LD,, (that amount of toxin required to kill within 4 days 50% of all animals injected; 1 MLD = 1.3 LD,,, cf. Ref. 1); NMRI, National Marine Research Institute mice (according to Festing, M. F. W., ed) (1971) International Index ofLaboratory Animals, 2nd Ed, p. 22, Medical Research Council Laboratory Animal Centre, Car- shalton, Surrey, United Kingdom).

* This affinity column was selected and prepared after pilot stud- ies had indicated the properties of Fragment B.

containing 0.05 M Tris/HCl buffer, pH 8.5, 0.6 M glycine, 0.001 M NazEDTA and 0.001 M dithiothreitol (6). Routinely, a peak contain- ing both polypeptide chains preceded the two peaks corresponding to heavy and light chain, respectively (Fig. 4). Fractions corresponding to the latter peaks were pooled, concentrated, and rechromato- graphed in order to remove impurities revealed by SDS gel electro- phoretic analysis. Removal of urea, however, easily caused irreversi- ble precipitation of the isolated polypeptide chains; therefore, the preparations were kept in 1 M urea or were made 1% with respect to SDS and dialyzed against SDS buffer (16) for electrophoretic analy- sis.

In order to produce fragments suitable for manipulation in physio- logical buffer, tetanus toxin was subjected to treatment with 0.02% formaldehyde at pH 7.8 for 16 h at 4”. After removal of formaldehyde by dialysis against 0.05 M Tris/HCl buffer, pH 8.2, containing 0.001 M

EDTA, the slightly derivatized toxin was subjected to reduction, urea treatment, and chromatography on Ultrogel AcA 44 as de- scribed above for the native toxin. The isolated derivatives were dialyzed against 0.1 M phosphate buffer, pH 7.8, and incubated with 0.05% formaldehyde at room temperature for 72 h. Such a derivative was stable under physiological conditions and elicited the formation of chain-specific antibodies.

RESULTS

Analysis of Papain Digest

Treatment of tetanus toxin with papain at 55” followed by gel chromatography gave the pattern shown in Fig. 1. Five fractions, denoted A to E, were isolated and analyzed sepa- rately.

Fraction A-This fraction was highly toxic (Table I) and could not be distinguished from native tetanus toxin by analy-

sis in conventional or SDS polyacrylamide gel electrophoresis (Fig. 2). Under denaturing conditions, both proteins yielded bands at a molecular weight of about 140,000. Treatment with mercaptoethanol produced two components with molecular weights corresponding to 93,000 f 5% (heavy chain polypep- tide) and 48,000 f 5% (light chain polypeptide), respectively. On immunodiffusion analysis, Fragment A exhibited complete identity with tetanus toxin when tested against horse tetanus antitoxin (Fig. 3).

L 1500 35'00 ’ 55‘00 . 75'00

ELUTION VOLUME (ML)

FIG. 1. Gel chromatography of tetanus toxin on Sephadex G-100 (cf. Ref. 10). The toxin (2.5 g; 15 mg/ml in 0.1 M phosphate buffer, pH 6.5, 0.001 M Na,EDTA, and 0.001 M cysteine-HCl) was treated with soluble papain (40 mg; 30 unit/mg) at 45” for 1 h, after which the temperature was raised to 55” (2 h). Digestion with carrier-bound papain (150 ml of gel/g of tetanus toxin) was performed in a similar manner except that the concentration of the toxin was lowered to 1.5 mg/ml. The digests were concentrated by ultrafiltration and applied to the column (10 x 100 cm). a, native toxin isolated after elution from DEAE-cellulose; b, analysis of papain digest (soluble enzyme); c, same as b except that matrix-bound papain was employed.

by guest on March 5, 2019

http://ww

w.jbc.org/

Dow

nloaded from

Structure of Tetanus Toxin. I 189

TABLE I

Toxicity of fragments from tetanus toxin

For details see “Rxnerimental Procedures.”

Substance Largest amount ad- ministered Toxicity

mg protein MLDlmg protein

Tetanus toxin 0.00001 30 x 106 Fragment A 0.00001 15 x 106 Fragment B 0.1 0 Fragment C 1.0 0 Lieht chain 0.001 5 x 104 Heavy chain 0.001 7 x 104

(-9 __ -

-mm-_-

FIG. 2. a, polyacrylamide gel electrophoresis of tetanus toxin and Fragments A to C from the papain digest. TT, native tetanus toxin; A, Fraction A (cf. Fig. 1); B, Fragment B; C, Fragment C. Samples treated with reducing agent prior to electrophoresis are indicated by +Me. b, gel electrophoresis (7% polyacrylamide) in SDS buffer of tetanus toxin and fragments isolated from the papain digest.

FIG. 3. Double diffusion analysis of tetanus toxin and fragments from the papain digest. ATT, horse tetanus antitoxin; for other abbreviations see the legend to Fig. 2.

Fraction B -This fraction was inadequately separated from Fraction A and was initially considered to be highly toxic (10). However, repeated chromatography on larger columns (1 x 200 cm) revealed that Fraction B contained a protein fragment (M, = 95,000 k 5%) which was virtually devoid of toxicity (Table I) and which showed a reaction of partial identity with tetanus toxin, but a reaction of complete nonidentity with Fragment C on immunodiffusion analysis (Fig. 3). Therefore, removal of minor traces of toxicity could be achieved by pas- sage of Fraction B through a column containing covalently attached anti-Fragment C serum as described under “Experi-

2b0 300 400 ELUTION VOLUME (ML)

FIG. 4. Chromatographic pattern of reduced and urea-treated Fragment B on a column of Ultrogel AcA 44. The material corre- sponding to Regions I to III was pooled and analyzed separately by electrophoresis in SDS buffer (inset; migration toward the anode at the bottom of the gels). LC, light chain polypeptide. The material eluted at 450 ml corresponded to the reducing agent.

TABLE II

Amino acid analysis of fragments from tetanus toxin Fragment C

Fast” Middle’ SlOW Fraggment

mo1147,000 g protein m0ll

95,000 g

Lysine 27.2 26.9 27.7 Histidine 5.1 5.2 4.8 Arginine 10.4 9.8 10.0 Aspartic acid 74.1 75.4 75.2 Threonine 19.5 19.9 19.0 Serine 37.6 38.4 37.2 Glutamic acid 21.4 21.5 20.6 Proline 16.5 15.5 15.3 Glycine 52.7 55.8 53.8 Alanine 28.4 29.6 27.4 Half-cystine 4.5 4.1 4.7 Valine 24.8 26.6 25.6 Methionine 4.4 5.1 4.9 Isoleucine 37.4 34.2 36.0 Leucine 40.9 39.3 40.6 Tyrosine 16.0 16.4 17.2 Phenylalanine 12.9 12.7 13.1 Trvntonhan 6.6 6.3 6.3

protein

65.1 10.6 13.6

120.8 53.3 77.2 75.0 49.8 75.3 56.1

7.4 42.2 13.2 87.3 69.3 32.5 30.5

4.1

a Refers to the relative mobility of the components of Fragment C on polyacrylamide gel electrophoresis (cf. Fig. 2).

mental Procedures.” Fragment B thus isolated was homogene- ous on polyacrylamide gel electrophoresis with or without SDS (Fig. 2), but was split into two narrowly spaced bands (M, = 48,000 + 5% and 45,000 r 5%, respectively) by treatment with detergent and reducing agent prior to the electrophoretic pro- cedure. Treatment of Fragment B with 4 M urea in the pres- ence of 0.05 M dithiothreitol followed by gel chromatography on a column of Ultrogel AcA 44 under the conditions described for the separation of heavy and light chain polypeptide (see “Experimental Procedures”) resolved the composite chains partially (Fig. 4). On SDS gel electrophoresis, material from the tailing edge (Region III, Fig. 4) was homogeneous and corresponded in migration to the light chain polypeptide. Fur- thermore, isolated light chain (see below), mixed with mate- rial from Region III, showed a single band on SDS gel electro- phoresis (inset, Fig. 4). The immunochemical analysis of Frag- ment B is presented in detail below. The amino acid composi- tion is shown in Table II.

by guest on March 5, 2019

http://ww

w.jbc.org/

Dow

nloaded from

190 Structure of Tetanus Toxin. I

Fragment C - This fragment was described in some detail in the preliminary communication (10). On electrophoretic anal- ysis in conventional polyacrylamide gel electrophoresis this fragment exhibited at least three bands (Fig. 2). Fractionation on blocks of polyvinylchloride yielded the separate components in quantities suitable for further analysis. The isolated compo- nents of Fragment C were indistinguishable on immunodiffu- sion analysis against rabbit anti-Fragment C serum or horse tetanus antitoxin. Further, it was not possible to discriminate between the components on the basis of the amino acid compo- sition (Table II). For a consideration of the possible origin of the heterogeneity of Fragment C see “Discussion.”

On SDS gel electrophoresis in 10% acrylamide gel, Frag- ment C was homogeneous and migrated as a protein with a molecular weight of about 45,000. However, when analyzed on 5% acrylamide gels, two narrow bands were observed (Fig. 5o). Analysis in the presence of mercaptoethanol yielded the slow migrating component only, indicating that Fragment C may consist of a single polypeptide chain with one or more partly reduced intrachain disulfide bonds. Reduction of such bonds would allow a more random conformation with an ap

e - *CAT

l - L ?! *ALD

a +Me

b FIG. 5. a, analysis of fragments from tetanus toxin by SDS gel

electrophoresis (5% polyacrylamide). LC, light chain polypeptide; C, Fragment C. Reduced samples are indicated by +Me. b, distinction of Fragment C from the light chain polypeptide by SDS gel electro- phoresis. CAT,. reduced catalase polypeptide (M, = 60,000), ALD. reduced aldolase polypeptide (M, = 40,000).

TABLE III

Protective capacity of Fragment C

Groups of guinea pigs were immunized subcutaneously with 1 ml of antigen preparation as indicated. When A1(OH)I gel was used as an adjuvant, a concentrated antigen solution was added to a 2% suspension of the gel followed by dilution of the adsorbed Fragment C to yield the appropriate protein concentration. The animals were kept for 4 weeks and were subsequently challenged with 22 LDSo of tetanus toxin. The number of survivors was determined on the 5th day following challenge.

Fragment ~o~nmunizing Al( gel Survival rate after chal- lenge

pg proteinldose pgldCW %

2.5 40 100 1.25 20 76.9 0.625 10 64.2

30 100 15 78.6

7.5 66.7 Control, no antigen 0

parent increase in size, thus yielding the slow migrating com- ponent.

Fragment C did not induce any toxic symptoms when in- jected into experimental animals (Table I). Injection of guinea pigs with Fragment C followed by challenge with native teta- nus toxin after 4 weeks revealed that protective immunity had beef elicited (Table III). As can be seen, the degree of protec- tion was highly dependent on Al(OH), gel as an adjuvant and on the amount of antigen administered. Rabbit antisera against Fragment C prepared as described under “Experimen- tal Procedures” routinely contained more than 200 I.U./ml of protective antibodies.

Fraction D-This fraction contained papain according to the data already presented (10). This fraction disappeared from the elution pattern when soluble papain was replaced by ma- trix-bound enzyme (Fig. 1~).

Fraction E - This fraction represented low molecular weight material which was largely dialyzable and no longer toxic.

Isolation of Native and Derivatized Light and Heavy Chain Polypeptide

Fig. 6 shows the gel chromatography pattern obtained when urea-treated reduced tetanus toxin was chromatographed on Ultrogel AcA 44 gel (6). The pattern was practically identical when the toxin was subjected to preliminary treatment with formaldehyde prior to chromatography in urea buffer. The protein fragments were rechromatographed and were essen- tially homogenous when analyzed by SDS gel electrophoresis (Fig. 6). After removal of denaturant by dialysis, the native polypeptides tended to precipitate from the solution, particu- larly at protein concentrations above 0.5 mg/ml. In contrast, the slight derivatization of the toxin by 0.02% formaldehyde, which caused a reduction of the toxicity from 30 x lo6 to about 5 x log MLD/mg of protein, allowed the isolation of modified polypeptide chains suitable for immunization purposes.

Distinction of Light Chain Polypeptide from Fragment C

On polyacrylamide gel electrophoresis in SDS buffer, the isolated light chain polypeptide exhibited properties similar to Fragment C (Fig. 5o). Like the latter fragment, the light chain yielded two closely spaced bands when analyzed in the absence of reducing agent, whereas treatment with mercapto-

‘i c 0.4

P ” 6 6

0.2

200 300 400 ELUTION VOLUME (ML)

FIG. 6. Chromatographic pattern of reduced and urea-treated tet- anus toxin. The fractions were pooled as indicated and subjected to rechromatography on the same column. The inset shows the electro- phoretic analysis (7% polyacrylamide) on SDS gels of the purified fractions (anode located at the bottom of the gels). TT+Me, reduced tetanus toxin; HC, heavy chain polypeptide; LC, light chain polypep- tide. The material eluted at 450 ml corresponded to the reducing agent.

by guest on March 5, 2019

http://ww

w.jbc.org/

Dow

nloaded from

Structure of Tetanus Toxin. 1

ethanol produced the slow component only. Furthermore, the molecular weight range corresponded to that recorded for Fragment C, and a mixture of the two proteins resulted in a broad, poorly resolved band containing both components (Fig. 5b). However, mixtures of each fragment with molecular weight standard proteins allowed a distinction between the two fragments. It is concluded that Fragment C (M, = 47,000 + 5%) migrates slightly ahead of the light chain (M, = 48,000 -t 5%). The latter molecular weight estimate is somewhat below that reported by others (4, 5).

On Ouchterlony double diffusion analysis, light chain gave a single line of precipitation which showed absence of coales- cence with Fragment C but which fused in partial identity with tetanus toxin (Fig. 7A). Furthermore, specific anti-Frag- ment C serum raised in rabbits gave no precipitation with the light chain derivative (Fig. 7B). Conversely, specific anti-light chain polypeptide serum reacted well with tetanus toxin and with the homologous antigen, but failed to precipitate with Fragment C (Fig. 70. Finally, horse tetanus antitoxin ab- sorbed with Fragment C still reacted with the light chain derivative (Fig. 70).

Taken together, the data presented above indicate that Fragment C may be distinguished from the light chain poly- peptide on a physicochemical as well as on an immunochemi- cal basis. It is therefore concluded that Fragment C cannot be derived from the light chain of tetanus toxin; consequently, it must constitute a portion of the heavy chain polypeptide.

Immunochemical Classification of Fragment B

On double diffusion analysis, Fragment B gave a reaction of complete nonidentity with Fragment C and showed a partial coalescence with tetanus toxin (Fig. 3). Similarly, intact Frag- ment B reacted partially identical with light chain derivative (Fig. 8), whereas Region III (cf. Fig. 4) derived from reduced and chromatographed Fragment B, showed a reaction of com-

plete identity with the light chain (Fig. 8). In conjunction with the SDS polyacrylamide gel electrophoresis data shown for Fragment B (Figs. 2 and 4), the immunochemical evidence strongly indicates that Fragment B represents that portion of tetanus toxin which remains after scission by papain of the polypeptide bond(s) releasing Fragment C. The remaining

STRUCTURE OF TETANUS TOXIN

HEAVY CHAIN S ,

cdl LIGHT CHAIN

PAPAIN

+

B

I

FRAGMENT

FRAGMENT B

C

FIG. 9. Schematic model of tetanus toxin and relationship be- tween the fragments discussed. The heavy chain comprises Frag- ment C (stt&ed area) and the complementary portion (cross- hatched), which is linked to the light chain via a disulfide bond. Fragment B corresponds to the cross-hatched area of the heavy chain linked to the light chain (the latter being virtually identical with Region III (Fig. 4)). Whether Fragment C constitutes the NH2- or COOH-terminal portion of the heavy chain remains to be deter- mined.

FIG. 7 (left). A to D, Ouchterlony double diffusion analysis of C. fragments isolated from tetanus toxin against antisera of different FIG. 8 (right). Immunochemical characterization of Fragment B. specificities. TT, native tetanus toxin; LC, light chain derivative; C, ATT, horse tetanus antitoxin; B, Fragment B derived from papain Fragment C from papain-digested tetanus toxin; 1, horse tetanus digestion of tetanus toxin. Re III, Region III derived from Fragment antitoxin; 2, rabbit anti-Fragment C serum; 3, rabbit anti-light B by reductive cleavage followed by chromatography (cf. Fig. 4). For chain serum; and 4, horse tetanus antitoxin absorbed with Fragment other abbreviations see the legend to Fig. 7.

by guest on March 5, 2019

http://ww

w.jbc.org/

Dow

nloaded from

192 Structure of Tetanus Toxin. I

fragment would be composed of the complementary portion of tetanus toxin molecule into three separate, nonoverlapping the heavy chain as well as the intact light chain and would regions which should be more suited for amino acid sequence conceivably exhibit the properties registered for Fragment B. studies than the intact toxin.

DISCUSSION

Apart from theoretical considerations in relation to the study of protein structure in general, the elucidation of the molecular parameters of tetanus toxin is of interest in view of the remarkable toxicity of this protein. This study confirms the two chain structure of tetanus toxin proposed by Craven and Dawson (4) and by Matsuda and Yoneda (5, 6), but is at odds with the conclusions reached by Bizzini et al. (3) who suggested that tetanus toxin is a dimer of two identical sub- units, each consisting of two polypeptide chains. It seems possible that the latter workers based some of their conclu- sions on SDS gel electrophoretic data obtained after prelimi- nary treatment of the protein samples with SDS under rela- tively mild conditions (3). To avoid confusion in SDS gel patterns induced by incomplete reaction of the polypeptide with the detergent, all samples analyzed here by this proce- dure were boiled in SDS prior to electrophoresis.

Native tetanus toxin is remarkably stable against proteoly- sis and only limited amounts of Fragment C may be detected by digestion with papain below 45”. Other proteases also pro- duced small amounts of Fragment C (trypsin, pronase, and subtilisin), but the treatment of tetanus toxin with papain at 55” was by far the most reproducible procedure to prepare large quantities of this substance.

Although a definite conclusion as to the nature of the heter- ogeneity of Fragment C revealed on polyacrylamide gel elec- trophoresis may not be drawn at the present time, the three components observed may conceivably differ by a varying degree of amidation of glutamic or aspartic acid residues, or both. This possibility is supported by the amino acid analysis (Table II) as well as by the fact that gel electrophoresis of Fragment C in formic acid buffer showed a single band only.”

Fragment C and the light chain fragment (6,ll) exhibited a considerable similarity in their SDS gel electrophoresis pat- terns (Fig. 5). As was observed for Fragment C, the light chain polypeptide also yielded a single component if the electropho- retie analysis was carried out immediately after isolation un- der reducing conditions. In contrast, two bands were obtained after prolonged dialysis of the light chain in the absence of reducing agent. Therefore, it is likely that formation of inter- nal disulfide bonds may occur in the light chain polypeptide or its derivative, upon storage under nonreducing conditions. Such a process might confer a more compact conformation to the polypeptide resulting in the appearance of the fast migrat- ing component (Fig. 5).

The complementary relationship between Fragment C and the light chain polypeptide provided the clue to the structure of Fragment B which was overlooked in the preliminary study (10). Since Fragment B apparently consists of light chain polypeptide linked via a disulfide bond to that portion of the heavy chain which is not comprised by Fragment C, the rela- tionship between the papain digestion products and the two chain structure of tetanus toxin may be imagined as shown in Fig. 9. Whether or not Fragment C is located in the NH,- or COOH-terminal end of the heavy chain is at present un- known.

Although Ultrogel AcA 44 under the conditions described by Matsuda and Yoneda (6) provided a far better separation of the light and heavy chain polypeptides than the procedure used initially, involving chromatography on Sephadex G-150 (111, the two composite chains were rather labile and precipitated upon removal of urea. Preliminary treatment with formalde- hyde at a low concentration permitted the subsequent separa- tion of the chains under identical conditions and provided derivatives which were stable in physiological solutions.

The native fragments used for toxicity determinations were kept in 1 M urea at a concentration below 200 pg/ml and were diluted immediately prior to injection into experimental ani- mals. Since the toxicity was drastically reduced by rechroma- tography of isolated light and heavy chain polypeptide, it is reasonable to conclude that each chain per se is atoxic, and that the contaminating complementary polypeptide chain pre- sumably recombined with the main component to form trace amounts of tetanus toxin. Indeed, overloading the SDS poly-

‘acrylamide gels with the purified, mercaptoethanol-treated fractions showed the presence of a faint band corresponding to the migration of the alternate polypeptide chain. Such a re- combination of the polypeptide chains to form the active teta- nus toxin was recently described by Matsuda and Yoneda (23).

Interestingly, the highly purified Fragments B and C from the papain digest, which together apparently comprise all antigenic determinants of the native toxin, were devoid of toxicity in. vivo. Lack of toxicity was observed even when mixtures of Fragments B and C or any combination of these fragments with nontoxic dilutions of the heavy or light chain polypeptide were injected. It seems possible, therefore, that the toxic principle displayed in vivo by the native toxin may involve the cooperative effect by several sites, arranged in a proper manner only in the intact protein molecule. If multiple sites must act in concert to produce the pharmacologic effect noted upon injection of tetanus toxin into experimental ani- mals, antibodies specific for only a portion of the toxic mole- cule may s&ice to protect against the lethal action of this protein. This concept is supported by the data shown in Table III which demonstrate the protective immune response elicited by Fragment C of tetanus toxin. Indeed, preliminary studies with chain-specific antibodies have revealed that all three main regions of the protein described here, 1, Fragment C, 2, the complementary portion of the heavy chain, derived from Fragment B, and 3, the light chain, induce the synthesis of antibodies neutralizing the action of tetanus toxin. The results of these studies will be published elsewhere.

Also in progress are studies using Fab fragments from chain-specific antitetanus toxin sera, with the aim to define the location of determinants important for expression of toxic- ity. One such hypothetical site would comprise that region of tetanus toxin which binds to the brain receptor (presumably a ganglioside-associated structure, cf. Ref. 11. In the following communication, it will be demonstrated that this binding site is probably located on the heavy chain portion of tetanus toxin (24).

Since Fragment B is easily split by reducing agents into its composite polypeptide chains, it is now possible to degrade the

3 T. B. Helting, unpublished observation.

Acknowledgments-We are indebted to Dr. H. Zilg for per- forming the amino acid analyses and to M. Kumpe, K. Massie, H. Muller, D. Reinhold, and H. Spenler for skilled technical assistance.

by guest on March 5, 2019

http://ww

w.jbc.org/

Dow

nloaded from

Structure of Tetanus Toxin. I 193

REFERENCES

1. Van Heyningen, W. E., and Mellanby, J. (1971) in Microbial Toxins (Kadis. M.. Montie. T. C.. and Ail. S. J.. eds) Vol IIb. pp. 69-108, Academic Press, New York ”

2. Murohv. S. G., Plummer. T. H.. and Miller. K. D. (1968) Fed. - _. P~OC. 27, 268’

3. Bizzini, B., Turpin, A., and Raynaud, M. (1973) Naunyn- Schmiedebergs Arch. Pharmacol. 276, 271-288

4. Craven, C. J., and Dawson, D. J. (1973) Biochim. Biophys. Aeta 317, 277-285

5. Matsuda, M., and Yoneda, M. (1974) Biochem. Biophys. Res. Commun. 57, 1257-1262

6. Matsuda, M., and Yoneda, M. (1975) Infect. Zmmun. 12, 1147- 1153

7. Kryzhanovsky, G. N. (1973) Naunyn-Schmiedebergs Arch. Phar- macol. 276, 247-270

8. Habermann, E. (1973) Med. Microbial. Zmmunol. 159, 80-100 9. Robinson, J. P., Picklesimer, J. B., and Puett, D. (1975) J. Biol.

them. 250, 7435-7442 10. Helting, T., and Zwisler, 0. (1974) Biochem. Biophys. Res. Com-

man. 57, 1263-1270 11. Helting, T., and Zwisler, 0. (1975) in Proceedings of the Fourth

International Conference on Tetanus, Dakar, Senegal, April 6 to 12,1975 (Edsall, G., and Triau, R., eds) pp. 639-646, Lips, Lyon

12. Kimmel, J. R., and Smith, E. L. (1954) J. Biol. Chem. 207, 515- 531

13. Latham, W. C., Bent, D. F., and Levine, L. (1962) Appl. Micro- biol. 10, 146-152

14. Sheff, M. F., Perry, M. B., and Zacks, S. I. (1965) Biochim. Biophys. Acta 100, 215-221

15. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193, 265-275

16. Weber, K., and Osborn, M. (1969) J. Biol. Chem. 244,4406-4412 17. Ouchterlony, ii. (1958) Prog. Allergy 5, l-78 18. Moore, S., Spackman, D. H., and Stein, W. H. (1958) Anal.

Biochem. 30, 1185-1189 19. Moore, S. (1963) J. Biol. Chem. 238, 235-237 20. Edelhoch, H. (1967) Biochemistry 6, 1948-1954 21. Cuatrecasas, P. (1970) J. Biol. Chem. 245, 3059-3065 22. Spaun, J., and Lyng, J. (1970) Bull. W. H. 0. 42, 523-534 23. Matsuda, M., and Yoneda, M. (1976) Biochem. Biophys. Res.

Commun. 68, 668-674 24. Helting, T. B., Zwisler, O., and Wiegandt, H. (1977) J. Biol.

Chem. 252, 194-198

by guest on March 5, 2019

http://ww

w.jbc.org/

Dow

nloaded from

T B Helting and O Zwislerbetween polypeptide fragments.

Structure of tetanus toxin. I. Breakdown of the toxin molecule and discrimination

1977, 252:187-193.J. Biol. Chem. 

  http://www.jbc.org/content/252/1/187Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/252/1/187.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on March 5, 2019

http://ww

w.jbc.org/

Dow

nloaded from