THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 250, No. 18, Issue of September 25, pp. 7435-7442, 1975

Printed in U.S.A.

Tetanus Toxin THE EFFECT OF CHEMICAL MODIFICATIONS ON TOXICITY, IMMUNOGENICITY, AND CONFORMATION*

(Received for publication, June 10, 1974, and in revised form, April 23, 1975)

JOHN P. ROBINSON,+ JOHN B. PICKLESIMER, AND DAVID PUEW§

From the Departments of Microbiology and Biochemistry, Vanderbilt University, Nashville, Tennessee 37232

Tetanus toxin has been isolated from the extract of Clostridium tetani and analyzed for purity using various methods, e.g. sedimentation velocity, gel filtration, polyacrylamide gel electrophoresis, immuno- electrophoresis, and immunodiffusion. The homogeneous toxin, characterized by a minimum lethal dose of 10 pg (18- to 20-g mouse), was judged to be of high purity.

The amino acid composition was determined and found to be in good agreement with reported values for both filtrate and extract toxin. The corrected sedimentation coefficient, sgO,,,, was found to be 7.5, and the molecular weight was estimated to be 150.000. These values agree closely with those reported by others.

The toxin was modified using the conventional formaldehyde reaction to produce toxoid and both reductive methylation and carbamylation, which are highly specific for lysyl residues. Under certain reaction conditions, carbamylation of the toxin completely eliminated toxicity. Whereas reductive methylation yielded a high degree of conversion of lysine to dimethyllysine and monomethyllysine, the toxicity, albeit greatly reduced, was never completely eliminated.

The circular dichroic spectrum of each chemically modified toxin was obtained, resolved into Gaussian components, and compared with that of native toxin, which is estimated to contain about 20% LY helix and 23% p structure. The far ultraviolet circular dichroic spectra of toxin and toxoid were characterized by negative extrema at 208 nm and 217 nm attributable to ordered secondary structure, and toxin also exhibited a distinct shoulder at 223 nm. Carbamylated toxin and methylated toxin were characterized by negative extrema at 210 nm and 206 nm, respectively, and both exhibited shoulders at 216 to 217 nm and 223 nm. The toxin and derivatives exhibited multiple negative extrema above 250 nm which were assigned to the various aromatic residues. There were differences in the spectra of the toxin and derivatives over the entire wavelength region, thus suggesting changes in the local environment of various chromophores. In particular, the rotational strengths of many of the bands assigned to tryptophan, tyrosine, and phenylalanine were altered in the derivatives. Also, in the far ultraviolet region of the circular dichroic spectrum, the data were suggestive of some reduction in the amount of both (Y helix and p structure in the derivatives. However, there was no evidence of extensive conformational changes, e.g. unfolding, in the modified toxins. Presently, it is not known if the small conformational differences between toxin and toxoid are important in the loss of toxicity with the retention of immunogenicity in the derivative.

The modification data are consistent with the hypothesis that separate amino acid residues are

involved in toxicity and immunogenicity.

Tetanus toxin is a potent neurotoxin that acts both in the central nervous system and peripherally in skeletal tissue. In

*This work was supported by Health, Education, and Welfare Contract FDA 73-36. A portion of this research was presented at the 74th Annual Meeting of the American Society for Microbiology, Chicago (May 12-17, 1974).

$ To whom correspondence should be addressed at the Department of Microbiology.

P Recipient of United States Public Health Service Research Career Development Award AM-00055 and of a Camille and Henry Dreyfus Teacher-Scholar Award.

the central nervous system, the toxin may interfere with the transmission of inhibitory impulses, while in skeletal tissue, the effects appear to involve the relaxing system, perhaps by interfering with calcium transport (1).

There are many conflicting reports in the literature on such molecular characteristics of the toxin as the sedimentation coefficient and the molecular weight. However, recent evidence indicates that the toxin has a molecular weight of 141,000 to 148,000 (2-4), and is composed of two nonidentical chains (4, 5). Bizzini et al. (4) have suggested that the toxin is a dimer of

7435

by guest on March 28, 2018

http://ww

w.jbc.org/

Dow

nloaded from

7436

molecular weight 146,000, with the monomer being composed of two polypeptide chains of molecular weights 21,000 and 52,000 that are covalently linked by a single disulfide. They proposed that the association of the two-chain monomer units to form the dimer is via strong noncovalent interactions. This quaternary structure accounts for many of the molecular weights that have been found when the toxin is exposed to various chemical conditions.

The amino acid composition of tetanus toxin (2, 3, 6) and of the heavy and light chains (4,5) has been reported. NH,-termi- nal analyses have given variable results. For example, Bizzini et al. (4) found a terminal leucine, while Murphy et al. (3) were unable to detect an NH,-terminal residue.

We recently reported the CD spectrum of tetanus toxin, and the data suggested the presence of about 20% (Y helix and 23% p structure (7). Other than this communication, there are no experimental data available on the conformation of tetanus toxin. Such studies are of general interest in comparative protein structural work and are, of course, essential for a full understanding of the mechanism of toxin action. Also, of both scientific and clinical interest is the mechanism of conversion of toxin to toxoid. In particular, little is known regarding the relative effects of specific side chain modification and confor- mational alterations on toxicity and immunogenicity.

We have prepared cell extract tetanus toxin and checked its purity using gel filtration, polyacrylamide gel electrophoresis, sedimentation velocity, immunoelectrophoresis, immunodiffu- sion, and toxicity measurements. The toxin was characterized with regard to its amino acid composition, molecular weight, and circular dichroic properties. Also, the toxin has been modified by the classical formaldehyde reaction to prepare toxcid and by two lysine-specific methods, reductive methyla-

tion and carbamylation. These derivatives have been evalu- ated for the extent of lysine modification, toxicity, and immunogenicity whenever possible, and conformational changes using CD spectroscopy.

MATERIALS AND METHODS

Chemicals

All chemicals used in this work were reagent grade unless otherwise noted. Acrylamide and N,N’-methylenebisacrylamide were obtained from Eastman Organic Chemicals, formaldehyde and potassium cyanate were purchased from Fisher Scientific Co., and Coomassie blue and sodium dodecyl sulfate were supplied by Sigma Chemical Co.

The following materials were used in chromatographic procedures: Sephadex G-50 (fine) and Sephadex G-100 (from Pharmacia Fine Chemicals) and diethylaminoethyl-cellulose, Whatman DE52 (from H. Reeve Angel and Co., Inc.).

Preparation of Purified Toxin

Clostridium tetani, Massachusetts strain, was kindly provided by Mr. W. C. Latham, Institute of Laboratories, Massachusetts Depart- ment of Public Health, Boston, Mass. Cultures were grown in 30- to 40-liter quantities as described by Murphy and Miller (8). Purification was also carried out according to the report of these authors.

Analytical Procedures

Homogeneity-All purified toxin preparations were examined for homogeneity by several methods. Polyacrylamide gel electrophoresis was carried out both according to Davis (9) and without spacer gels according to Hjerten et al. (10). When no stacking or spacer gels were used, samples containing 0.3 M sucrose were layered on top of the resolving gels in 0.01 M Tris-glycine buffer, pH 8.9. A constant current of 3 ma/gel was used with buffer and gel compositions as follows: 0.05 M

Tris-glycine, pH 8.9, was used in a continuous buffer system with the separation gel containing 7.5% acrylamide and 0.25% N,N’-

methylenebisacrylamide. The use of spacer and sample gels with a discontinuous buffer system produced similar results.

Immunodiffusion was done using purified and crude toxin prepara- tions with antisera prepared against both purified and crude toxoid (11). The latter was prepared as described by Dawson and Mauritzen (12). Antiserum was prepared by subcutaneous injection of 5 mg of the toxoid, in Freund’s adjuvant, in rabbits. After 3 to 4 weeks, the rabbits were again injected subcutaneously with 5 mg of the toxoid in the absence of adjuvant. Antisera were collected by heart puncture 1 week later. Immunoelectrophoresis was carried out according to methods reported by Grabar and Williams (13).

Sedimentation velocity experiments were performed using a model E ultracentrifuge with schlieren optics. Purified toxin was examined at several concentrations ranging from 1 to 15 mg/ml of protein in 0.2 M phosphate buffer at pH 6.0. The samples were centrifuged at 60,000 rpm using electronic speed control.

Molecular Weight-Low speed sedimentation equilibrium runs were done according to the method of Van Holde and Baldwin (14) in a Spinco model E analytical ultracentrifuge using the Rayleigh interfer- ence optical system. The apparent molecular weight was evaluated using Svedberg’s equation:

M, = [2 RT/( 1 ~ VP),] .dlnc/dr* (1)

The symbols have their usual meaning and the value for the partial specific volume of the purified toxin was calculated (15) from the measured amino acid composition.

Amino Acid Analysis-The amino acid content of purified tetanus toxin was determined by the method of Moore and Stein (16) on a Beckman-Spinco model 120 amino acid analyzer equipped with an Infotronics model CRS-12-AB integrator. Tryptophan was determined as described by Lui and Chang (17) by hydrolysis of the sample in 4 N sulfonic acid with 0.2% tryptamine at 115’ for 22 hours. After neutralization of the hydrolysis mixture, the analysis was completed as usual. Cysteine was determined as cysteic acid after performic acid oxidation according to Moore (18). Values for serine were calculated by extrapolation to zero time after hydrolysis times of 20, 40, 70, and 140 hours. Surprisingly, threonine reached a maximum value at 40 hours and did not decrease with longer hydrolysis times.

Circular Dichroism-The CD spectra were determined with a Cary 60 spectropolarimeter equipped with a CD attachment. The instru- ment was calibrated with d-lo-camphorsulfonic acid and the perform- ance was checked using various standard proteins. Between about 240 to 310 nm, path lengths of 5 and 10 mm were used, and protein concentrations were 1 to 1.5 mg/ml. Below 240 nm, path lengths of 0.5 and 1 mm were used and protein concentrations were 0.1 to 0.2 mg/ml. The full scale range was generally 20 or 40 millidegrees and a time constant of 3 s was used. The reported spectra were obtained at ambient temperature and represent the average of several scans for samples and base-lines. In all cases, toxin and derivatives were dissolved in 0.2 M phosphate buffer, pH 6 to 7. The reduced mean residue ellipticity was based on an average residue molecular weight of 114.5 as determined from the amino acid composition, and protein concentrations were estimated by the Lowry method using bovine serum albumin as a standard. Curve resolution into Gaussian compo- nents was achieved as described elsewhere (19), and the corresponding rotational strengths were determined from the Gaussian parameters (20).

Preparation of Toxoid and Modified Toxins

Reaction with Formaldehyde-Toxoid was prepared with formalde- hyde on both crude cell extracts and purified toxin according to the method of Dawson and Mauritzen (12). This procedure consists of exposure of samples to 0.35% formaldehyde at pH 7.5 for 2 days at room temperature followed by 3 days at 37”.

Toxoids were examined for residual toxicity by injecting 0.1 ml (approximately 100 to 400 rg) of undiluted sample intramuscularly in mice. The mice were observed for 4 days for signs of paralysis or death. Only 18- to 20-g Swiss strain ICR mice, purchased from Harlan Industries, Cumberland, Indiana, were used in these assays.

Reductive Methyl&ion-Borohydride reductive methylation of ly- sine on purified tetanus toxin was done according to published methods (21, 22). Purified toxin was dialyzed into 0.2 M borate buffer at pH 9.0 and divided into two portions. One portion was exposed to formaldehyde in a 4.fold molar excess over the lysine residues present. This was followed within 1 min by the addition to both aliquots of sodium borohydride in an equal molar amount to the lysines present.

by guest on March 28, 2018

http://ww

w.jbc.org/

Dow

nloaded from

7437

This was followed still 1 min later by the addition of another equal molar quantity of sodium borohydride. Thus, one aliquot received both formaldehyde and the reducing agent while the control was exposed only to the reducing agent.

After dialysis to remove excess aldehyde and reducing agent, both preparations were subjected to examination for toxicity and amino acid analysis to determine the extent of methylation of lysine residues in the toxin. The amino acid analysis for methylated lysine was run by Method 1 of Kuehl and Adelstein (23). Four experiments were completed. In the first, reductive alkylation was carried out as just described. In Experiment 2, after completing the reaction once, the mixture was dialyzed against borate buffer to remove excess dialyzable reactants and the process was repeated. In Experiments 3 and 4, this process was completed three and six times, respectively. After final dialysis, all .of the reaction products were examined for toxicity (by mouse injection) and amino acid analysis for monomethyllysine and dimethyllysine. Controls were run for Experiments 1 and 4 by omitting formaldehyde from the reaction mixtures.

Carbamylation-Purified toxin was reacted with potassium cyanate at varying temperature and pH. The reaction proceeds favorably in aqueous solution at neutral or alkaline pH (24). The following experiments were performed: (a) toxin was dialyzed into 0.2 M borate buffer at pH 9.0 and reacted. with 5% potassium cyanate at 50’ for 18 hours; (b) toxin was exposed to 5% potassium cyanate at 37’ for 18 hours in 0.1 M phosphate buffer at pH 7.0; (c) toxin was exposed to the same conditions as b above, except the temperature was 25”. Control preparations for each experiment were treated identically except that the cyanate was omitted.

RESULTS

Homogeneity of Purified Toxin-Purified toxin was exam- ined for homogeneity by several methods. Each procedure indicated that the purified preparations contained a single, highly purified component.

Fig. IA illustrates the final purification step through Sepha- dex G-100, equilibrated and developed with 0.1 M phosphate buffer (pH 7.5), where the optical density at 280 nm of each fraction is given. Fig. 1B shows the results of polyacrylamide gel electrophoresis of samples on the leading edge, the center and the trailing edge of the peak. All preparations were examined to ensure that no tetanalysin was present (25, 26). If contaminating protein was present in the pooled fractions, the preparation was rechromatographed through the last two purification steps to remove the contaminant.

Fig. 2A represents the examination of purified toxin by immunodiffusion. The center well contains rabbit antiserum to crude cell extract toxoid. Well b contains crude cell extract, while Well c contains purified toxin. It is apparent that while the cell extract contains more than one antigenic component, the purified toxin contains only one. As an additional test of homogeneity of our purified preparations, rabbit antisera against purified tetanus toxoid were examined using cell extract and purified toxin as antigens. Fig. 2B illustrates that antisera against purified toxoid gave a single precipitin band in immunodiffusion against purified or crude antigens. These experiments indicate that by immunodiffusion, our purified toxin preparations contain a single, homogeneous component.

Fig. 3 illustrates the results of examination of our purified toxin preparations by immunoelectrophoresis. Well B con- tained purified toxin, while Well A contained crude cell extract. The trough contained antiserum against crude cell extract toxoid and gave only one precipitin band with Well B, while Well A shows more than one component. When antiserum against purified toxoid was added to the trough, only one precipitin band formed with the components of both wells. These results indicate that by this method our purified toxin preparations show homogeneity.

Fig. 4 shows the results of examining purified toxin in the

FRACTION NUMBER

FIG. 1. A, final purification of tetanus toxin on Sephadex G-100 equilibrated and developed with 0.1 M phosphate buffer, pH 7.5. Two columns (5 x 100 cm) were run in series using reverse flow. Five-milli- liter fractions were collected and monitored using absorption at 280 nm. B, polyacrylamide gel electrophoresis was run using 20 rg of protein/gel on samples from the leading edge, center, and trailing edge of the peak illustrated in Fig. 1A. Sample 1 represents a fraction from the leading edge; 2, 3, and 4 denote fractions from the center of the peak; and 5 represents a fraction from the trailing edge of the peak. Gels were stained with Coomassie blue and decolorized according to the method of Davis (9).

iA

:B

FIG. 2 (upper). A, immunodiffusion of antiserum to crude toxoid (prepared from crude cell extract) with crude toxin and the highly purified toxin. Well a contains antiserum to crude toxoid, Well b contains crude toxin, and Well c contains purified toxin. B, immuno- diffusion where Well a contains antiserum to toxoid prepared from purified toxin, Well b contains crude extract toxin, and Well c contains purified toxin. Antisera were prepared in rabbits as described in the text.

FIG. 3 (lower). Immunoelectrophoresis of crude toxin and purified toxin with antiserum to crude toxoid. The antigen toxoids were added in lo- to 20-&g quantities. The crude preparation (A) shows multiple bands while the pure toxin (B) shows only one.

analytical ultracentrifuge. Again, this method indicates ho- mogeneity in our purified toxin. Several runs were made at concentrations ranging from 1.5 to 15 mg/ml of toxin. The sedimentation constant was plotted as a function of concentra- tion. At zero concentration, the corrected value of slO,u, was calculated to be 7.5 S. This value agrees well with the 7.6 to 7.8 S reported by Mangalo et al. (2). The sedimentation experi- ments were run at rotor speeds of 60,000 rpm using schlieren optics and electronic speed control. The S values were corrected

by guest on March 28, 2018

http://ww

w.jbc.org/

Dow

nloaded from

7438

FIG. 4. Sedimentation of tetanus toxin at 60,000 rpm using schlieren optics. The cell contained a 7 mg/ml solution of toxin in 0.2 M phosphate buffer at pH 6.0. Photographs were made at 16-min intervals after the desired speed was achieved, and this particular figure represents an exposure made at 48 min. Patterns at the bottom of the cell are probably due to polyphosphates in the buffer.

to 20” and also to the viscosity of the buffers with reference to water.

Molecular Weight of Purified Toxin-The molecular weight of the toxin was estimated from several low speed sedimenta- tion equilibrium runs (14). The results indicate a weight average molecular weight of 150,000 + 10,000. This range agrees well with the molecular weights reported by other laboratories (2-4). The hydrodynamic data indicate that the toxin preparation probably contains some aggregated material.

Amino Acid Composition-The amino acid composition of the toxin is presented in Table I and the values are in good agreement with those reported by other workers (2,3,6). These data are based on analyses which were conducted using varying hydrolysis times. Interestingly, threonine reached a maximum value after 40 hours of hydrolysis and did not change apprecia- bly thereafter. Stable amino acids such as valine and leucine reached maximal values after 40 hours of hydrolysis.

Toxin Modifications-The effect of various chemical modifi- cations of tetanus toxin was examined and the extent of lysine modification is summarized in Tables II and III. Purified tetanus toxin was reacted with formaldehyde at neutral or slightly alkaline pH to prepare toxoid, and the toxicity results are presented in Table IV. It is readily apparent that interac- tion of toxin with formaldehyde at neutral pH at 25’ followed by 37” completely destroys toxicity as, of course, expected (12).

Reductive methylation (21, 22) of 74% of the lysine residues in the toxin molecule, however, does not completely eliminate the toxic properties. Tables III and IV show that methylation of 65% of the lysine residues has about the same effect on toxicity as methylation of 74%, while methylation of 19% of lysine does not reduce toxicity so dramatically.

We examined the effect of reacting the lysyl residues of toxin with potassium cyanate under varying conditions (cf. “Mate- rials and Methods”). Table IV shows that carbamylation of toxin effectively reduces toxicity. At pH 9.0 and 50”, toxicity was destroyed after 18 hours, while at lower temperatures and neutral pH, the preparations retained only traces of toxicity. Opalescence of the modified toxin solutions was noted follow- ing modification at both 37’ and 50’.

Circular Dichroic Spectra of Native and Modified Tox- in-The resolved near and far ultraviolet CD spectrum of tetanus toxin has been reported elsewhere (7). The spectrum is shown in Fig. 5 and the band characteristics and tentative

Amino acid analysis of tetanus toxin0

Amino acid Mel %

Lysine 8.50 Histidine 0.97 Arginine 2.37 Tryptophan 0.66 Aspartic acid 16.81 Threonine 5.41 Serine 7.54 Glutamic acid 9.18 Proline 4.23 Glycine 5.18 Alanine 4.22 Half-cystineb 0.84 Valine 4.23 Methionine 1.67 Isoleucina 8.47 Leucine 9.12 Tyrosine 5.87 Phenylalanine 4.23

“These represent the average of eight independent determinations with the exceptions of Trp and Yz Cys, which are based on the average of two determinations each. The value for serine is based on an extrapolation to zero time.

b Determined as cysteic acid.

TALILE II Extent of conversion of lysine to homocitrulline in tetanus toxin by

carbamylation

Carbamylation % Unreacted % Modified conditions” lysine lysine”

Control 100 0 pH 7.0, 25“ 41 59 pH 7.0,37’ 23 77 pH 9.0,50’ 19 81

a See text for full description of the reaction of potassium cyanate with toxin.

b The lysine and homocitrulline contents were determined by amino acid analysis on hydrolysates from 0.5-mg quantities of toxin and carbamylated toxin. The results are reported as the per cent of unreacted and modified (homocitrulline) lysine in each preparation,

assignments we reported (7) are given in Table V. The near ultraviolet CD spectrum of proteins is due to the aromatic amino acid residues and to disulfides (27,28) and the spectrum in this region is quite sensitive to the local environment and the constraints on these groups. Disulfides generally give broad, shallow bands centered at about 270 nm and are frequently not resolved in proteins containing tyrosyl and tryptophanyl resi- dues. Thus, it is not surprising that these groups were unde- tected.

Using the far ultraviolet CD spectra reported by Chen et al. (29) as reference, the degree of secondary structure was estimated using two independent methods (7). One is a least squares analysis which compares the ellipticity at several wavelengths with that of standard proteins, and the other involves the ratio of the rotational strength for the conforma- tion-dependent transitions (e.g. 222 to 223 nm for the a helix and 214 to 216 nm for j3 structure) to the corresponding value of standard proteins (30). We found tetanus toxin to contain about 20% cy helix, 23% /3 structure, and thus is 57% aperiodic (7). This represents a significant amount of ordered secondary

by guest on March 28, 2018

http://ww

w.jbc.org/

Dow

nloaded from

TABLE III Extent of lysine modification in tetanus toxin by reductive

methylation

Cycles” L y s&e) b % I Unreacted % Modified

Lys(Wtb lysine lysine

0 0 0 100 0 1 1.8 11.4 80.8 19.2 3 8.2 57.1 34.7 65.3 6 6.1 67.5 26.4 73.6

“Represents the number of times toxin was subjected to reductive methylation (see text for details).

b Lys(Me) and Lys(Me), denote monomethyllysine and dimethylly- sine, respectively. The values are normalized to the total recovered lysine, monomethyllysine, and dimethyllysine. Hydrolysates from 2 mg of toxin or methylated toxin were analyzed by Method 1 of Kuehl and Adelstein (23).

TABLE IV

Toxicity” of tetanus toxin derivatives

I Fold dilution Toxin derivative

0

Toxoid (per formaldehyde) . 0 Carbamylated toxin (25”, pH 7) +& Carbamylated toxin (37’, pH 7) + Carbamylated toxin (50”, pH 9) 0 Methylated toxin (1 cycle) + Methylated toxin (3 cycles) . + Methylated toxin (6 cycles) +

i?

0 0 0 0 + + +

-

F 0 0 0 0 + 0 0

-

iF - 0 0 0 0 + 0 0

-

ii5 I I I

G F - - 0 0 0 0 0 -c

0 0 -c

0 0 -c

+ + 0 + 0 0 + 0 0

- - -

a The assay is based on paralysis and death in 20-g mice following injections of 0.1 ml of toxin solution into hind muscle; observations were up to 4 days. Dilutions were made from solutions containing 1 to 4 mg of protein/ml. The symbols 0 and + indicate no effect and a lethal dose, respectively. Toxin controls, i.e. toxin which had been exposed to the same experimental conditions in the absence of modifying agent, were assayed for each derivative at each dilution and with but one exception, i.e. the control for carbamylated toxin at 50’ and pH 9, lethality was observed.

b Paralysis followed by death on 5th day. c Not determined.

h.mu

FIG. 5. The CD spectrum of tetanus toxin. The mean residue ellipticity is in units of deg.cm*/dmol. The spectrum shown has been resolved (cf. Table V) and the details are given elsewhere (7). The arrows denote the positions of the resolved bands and the tentative assignments are indicated; a and j3 refer to (Y helix and j3 structure, respectively.

structure in the protein. As discussed elsewhere (7), these estimates of ordered secondary structure have not been cor- rected for the contribution of the toxin aromatic residues to the far ultraviolet CD spectrum.

Figs. 6 and 7 show the resolved CD spectrum of the toxoid

7439

TABLE V Rotational strengths” of resolved circular dichroic spectrum of tetanus

toxin and toxoid

Tentative assignments A,, nm

Toxin

R x 10”

Toxoid

L nm R x 10”

TrP - -

TrP 300.0 -0.99

‘b 293.5 -8.09

Trp 287 -4.75 ‘b 283 -6.48 % 279.5 -2.44 Tyr 276 -9.07 Tyr 272 -1.51 Phe 268 - 12.43 Phe 260.5 -4.26 Phe 254 -8.02 0 229 +113.16 Peptide’ 223 -4200 Peptidec 215 - 2070 Peptidec 207 -1800

303 -0.65 298.5 -0.58 293 - 12.00 286.3 -4.20 283 -4.45

- - 217 - 17.40 272.6 -1.92 268.5 -6.62 262.5 -7.71 256.3 -4.24 236 -471 221 - 3460 216.5 -1200 206.5 -3760

“The rotational strength in cgs units for each resolved Gaussian band was calculated from the relationship, R 3 (1.23 x lo-‘*) x [&‘I x A/X,, where [0”] is the mean residue ellipticity at the extremum, X, is the wavelength at which the extremum occurs, and A is the band width from X, to the wavelength where [B] = [e’ye. The R values are based on mean residue ellipticity; to convert to molecular ellipticity simply multiply by 1265, the estimated number of residues in toxin. The data for toxin have been reported elsewhere (7).

“This resolved band probably has contributions from all three aromatic groups and perhaps peptide groups as well.

c The 223-nm and 215-nm bands are assigned to the n-A* transition of peptide chromophores in the (Y helical conformation and B structure, respectively. The 207-nm band arises mainly from a R-K* component of the (Y helix; however, other conformation-dependent ?T-?T* transitions contribute in this region.

I I I I I

TOXO’D w , , 1””

I 200 220 240 260 260 300

A. nm

FIG. 6. The resolved CD spectrum of tetanus toxoid.

of each toxoid-resolved band is given in Table V, and the helical content of toxin and the various derivatives is presented in Table VI. Relative to native toxin, there are small changes in the rotational strengths of many of the resolved bands. (For brevity, the values are listed for the toxoid only.) In the near ultraviolet region, this indicates that the local environment of several aromatic amino acid residues is haltered. For example, the near ultraviolet CD difference spectrum for toxin and toxoid (i.e. [O] . toxin - [O] .formaldehyde-produced toxoid versus X) is shown in Fig. 8 and the positions of the extrema indicate alterations in the vicinity of tryptophanyl, tyrosyl, and phenylalanyl residues. In addition, there are changes in

and the modified toxins, respectively. The rotational strength the far ultraviolet CD spectra, indicating that the chemical

by guest on March 28, 2018

http://ww

w.jbc.org/

Dow

nloaded from

7440

0

-4

-8

[e] x 10-s 0

-4

-8

CARBAMYI

METHYL

200 220 240 260 280 300

: I ATED

J?, TOXIN

1 TOXIN

l,nm

FIG. 7. The resolved CD spectrum of carbamylated and of methyl- ated toxin.

TABLE VI

Per cent helical content0 of tetanus toxin and derivatives

Protein a Helix j3 Structure Aperiodic

0

-20

-40

lel 0

-20

-40

260 260 300

x .nm FIG. 8. The near ultraviolet difference CD spectrum of tetanus

toxin and toxoid. The toxoid ellipticity (Fig. 6) was subtracted from the +oxin ellipticity (Fig. 5) at each wavelength, and the resulting difference is denoted as A [e].

Toxin 20 l 0 23 zt 3 57 f 3 Toxoid 17 f 1 21 f 7 62 zt 8 Carbamylated toxin 17 f 4 16 f 4 67 zt 8 Methylated toxin 10 f 3 12 f 1 78 f 4

a These values were estimated from CD spectra using two methods as described in the text. The results are presented as the mean of the two; the mean f the number given represents the two estimates of helical content. The aperiodic content is simply the per cent remaining after the (Y helix and j3 structure are accounted for.

modifications reduce somewhat the amount of ordered second- ary structure.

DISCUSSION

The results of the hydrodynamic, electrophoretic, immuno- logical and biological characterization of our tetanus toxin preparation demonstrate that the sample is highly purified. The molecular weight we obtained is in good agreement with the reported values (2-4).

The amino acid composition of tetanus toxin reported herein is in good agreement with other reports (2,3,6). Using the data given in Table I, we have calculated the polarity (31) and hydrophobicity (32-35) of tetanus toxin.

Capaldi and Vanderkooi (31) analyzed the polarity of a large number of soluble and membrane proteins. They classified Asp, Asn, Glu, Gln, Lys, Ser, Arg, Thr, and His as polar residues and the other amino acids as nonpolar. Of the 205 soluble proteins that were considered, 85% had polarities (i.e. per cent polar residues) of 47 + 6%, whereas membrane proteins were generally characterized by lower polarities. A similar calculation on tetanus toxin yields a value of 51%, thus suggesting that it is a high polarity soluble protein. It is noteworthy that an analysis of the amino acid composition of high and low molecular weight chains of tetanus toxin (4, 5) gave values of 51 to 52%, and 50%, respectively, thus indicating similar polarities.

A somewhat more quantitative approach (33-35) can be taken by using side chain free energies of transfer (e.g. from an organic solvent to water) obtained from solubility measure- ments of amino acids in water and in organic solvents (14, 32). With this method, one calculates the hydrophobicity by multi- plying the residue percentage for each side chain by its respective transfer free energy and then summing over all side chains comprising the protein. Using the same values as Goldsack (34), we find a hydrophobicity of 1167 cal for tetanus toxin. This is comparable, albeit slightly higher, to the following hydrophobicities which represent an average based on an analysis of the composition of each protein from several species: 1158, 1157, 1133, 1127, 1103, 1097, and 1081 cal (+ approximately 10%) for hemoglobin, insulin, malate dehydro- genase, lactate dehydrogenase, cytochrome c, myoglobin, and glutamate dehydrogenase, respectively (34). Using the recent transfer free energies reported by Nozaki and Tanford (32) for several amino acid side chains and assuming a zero transfer free energy for His, Lys, and Arg, we obtain a hydrophobicity of 907 cal for tetanus toxin. Hence, the qualitative polarity index suggests that tetanus toxin, like its constituent chains, is quite polar relative to other soluble proteins. Moreover, the semi- quantitative hydrophobicity index indicates the toxin to be comparable to various classes of soluble proteins.

The tetanus toxin modification studies reported herein consisted of specific alteration of lysyl residues by either reductive methylation or carbamylation. The less specific reaction with formaldehyde to produce toxoid was also used. As expected, the latter yielded an atoxic, immunogenic product.

Methylation of some 74% of the lysyl residues greatly reduced, but did not eliminate, toxicity. The remaining toxicity may be related to the nature of the modification, since the reduction of the Schiff base intermediate results in the substitution of methyl groups for hydrogens in the t-amino groups. Such a modification causes no significant charge alteration in the protein at neutral pH, and does not greatly change the stereochemical nature of the lysyl side chain. Due to the remaining toxicity in the methylated toxin, the immuno- genic properties of this derivative could not be examined.

Carbamylation of a similar number of lysyl residues in the toxin completely destroyed toxicity. The introduction of this group into the toxin would be expected to produce more pronounced changes in the protein than reductive methylation due to its size and the alteration of charge. Although atoxic, the carbamylated derivative failed to stimulate detectable precipi-

by guest on March 28, 2018

http://ww

w.jbc.org/

Dow

nloaded from

tating antibodies in a single rabbit. Although this indicates a loss of immunogenicity with carbamylation, the results are not conclusive.

It has been reported by others that maleylation of the lysine t-amino groups in tetanus toxin produced an atoxic, nonimmu- nogenic derivative (4). Dinitrophenylation of toxin was re- ported to produce similar results, although the product was reported to flocculate with antitoxoid serum (36). More re- cently, Bizzini et al. (37) reported on the effect of modification of tyrosyl residues of tetanus toxin by highly selective nitration with tetranitromethane. The properties of the product of this reaction varied depending on the extent of nitration of tyrosine. When only a few nitro groups were introduced, the molecule was rendered atoxic and immunogenic. With sufficient nitra- tion, both immunogenicity and toxicity were destroyed.

Summarizing these reports and our modification data, it appears that the toxicity of tetanus toxin can be reduced or eliminated by the specific modification of either lysyl or tyrosyl residues. These residues may either be part of the site of toxin activity or their modification may produce conformational changes in the protein which affects the geometry of the site(s). There appear to be separate, but less available or less reactive, lysyl and tyrosyl groups involved in the immunogenicity of the protein. The production of toxoid appears to involve the direct modification of 1 or more of these residues or alterations in the molecular conformation which are a consequence of these modifications.

A detailed comparison of the CD spectrum between 250 to 310 nm of each toxin derivative with native toxin indicates that the local environment is altered in the vicinity of various aromatic chromophores. This could result either from small conformational changes due to the modified groups, or it could reflect a different local chemical environment arising directly from the modified group. The latter condition is, of course, dependent on close proximity of reactive lysyl side chains to the various aromatic chromophores. In the formaldehyde- treated toxoid, but not in the carbamylated or methylated toxin, one or more tyrosyl groups are probably modified, and this would be expected to alter somewhat the rotational strength of the resolved bands between about 272 and 283 nm. It is noteworthy that the modification of tyrosyl residues in tetanus toxin appears to cause conformational changes as evidenced by an increased number of titratable sulfhydryl groups per molecule of toxin after the modification (37). It is, therefore, not yet apparent whether loss of toxicity is due to the direct modification of these residues or to other changes in the molecule which the modifications produce. It will be of interest to modify other types of side chains and assay the derivatives for toxicity.

Our analysis of the far ultraviolet CD spectrum suggests that the toxin derivatives are less helical than native toxin. The major difference was found in the methylated derivative. This is somewhat surprising, since dimethyllysine and monometh- yllysine are chemically more related to lysine than carbamylly- sine. Lysine itself is classified as only a weak (Y helix former (38), and it is possible that the methyl groups decrease the intrinsic helix potential of lysine more than the carbamyl groups. Also, it is possible that some denaturation occurred during the repeated reductive methylation steps. This could account for the reduced ellipticity noted over the entire range of wavelengths examined.

It is noteworthy that of the derivatives we have investigated, the CD spectrum of the formaldehyde-produced toxoid most

7441

closely resembled that of native toxin. The small differences observed are consistent with the occurrence of subtle confor- mation changes when the toxic compound is rendered atoxic, but remains immunogenic. It is not known if these conforma- tional changes are important in the loss of toxicity. Our data clearly demonstrate, however, that formaldehyde-mediated toxoid production does not involve extensive conformational alterations.

The concept of a toxin “active site(s)” and “immunogenic site(s)” occupying discrete regions within a well defined conformational structure provides a convenient model for additional investigations aimed at elucidating the molecular mechanisms involved in the conversion of toxin to toxoid.

Acknowledgments-The authors are grateful to Mr. W. C. Latham of the Massachusetts Health Department, whose advice and encouragement greatly accelerated the initiation of toxin production and purification. They also thank Dr. L. A. Holladay for his valuable assistance in the CD curve resolution and interpretation and for many helpful discussions regarding molecular weight determinations.

REFERENCES

1. Zacks, S. I., and Sheff, M. F. (1971) in Neuropoisons, Their Pathological Actions (Simpson, L. L., ed) Vol. 1, pp. 225-262, Plenum Press, New York

2. Mangalo, R., Bizzini, B., Turpin, A., and Raynaud, M. (1968) Biochim. Biophys. Acta 168, 583-584

3. Murphy, S. G., Plummer, T. H., and Miller, K. D. (1968) Fed. Proc. 27, 268

4. Bizzini, B., Turpin, A., and Raynaud, M. (1973) Arch. Z’harmacol. 276, 271-288

5. Craven. C. J.. and Dawson. D. J. (1973) Biochim. Biorhvs. Acta

6. 317, i77-2i5

. I

Bizzini, B., Blass, J., Turpin, A., and Raynaud, M. (1970) Eur. J. Biochem. 17, loo-105

7.

8. 9.

10.

Robinson, J. P., Holladay, L. A., Picklesimer, J. B., and Puett, D. (1974) Mol. Cell. Biochem. 5, 147-151

Murphy, S. G., and Miller, K. D. (1967) J. Bacterial. 94,580-585 Davis, B. J. (1964) Ann. N. Y. Acad. Sci. 121, 404-427 Hjerten, S., Jerstedt, S., and Tiselius, A. (1965) Anal. Biochem.

11, 214-223 11. Ouchterlony, 0. (1949) Acta Pathol. Microbial. Stand. 26,507-515 12. Dawson. D. J.. and Mauritzen. C. M. (1967) AL&. J. Biol. Sci. 20.

13. 253-263 ’

Grabar, P., and Williams, C. A. (1953) Biochim Biophys. Acta 10, 193-194

14. Van Holde, K. E., and Baldwin, R. L. (1958) J. Phys. Chen. 62, 734-743

15.

16. 17. 18. 19.

Cohn, E. J., and Edsall, J. T. (1943) Proteins, Amino-Acids and Peptides, p. 370, Reinhold Publishing Co., New York

Moore, S., and Stein, W. H. (1963) Methods Enzymol. 6,819-831 Lui, T. Y., and Chang, Y. H. (1971) J. Biol. Chem. 246.2842-2848 Moore, S. (1963) J. Biol. Chem. 238, 235-237 Zahler, W. L., Puett, D., and Fleischer, S. (1972) Biochim.

Biophys. Acta 255, 365-379 20. 21. 22. 23.

Puett, D. (1972) Biochemistry 11, 1980-1990 Means, G. E., and Feeney, R. E. (1968) Biochemistry 7,2192-2201 Rice. R. H.. and Means. G. E. (1971) J. Biol. Chem. 246.831-832 Kuehl, W. ‘M., and Adelstein, R. S.‘(l969) Biochem. Biophys. Res.

Commun. 37, 59-65 24. 25. 26.

Stark, G. R. (1972) Methods Enzymol. 25B, 103-120 Hardegree, M. (1965) Proc. Sot. Exp. Biol. Med. 119, 405-408 Hardegree, M., Palmer, A. E., and Duffin, N. (1971) J. Infect. Dis.

123, 51-60 27. Adler, A. J., Greenfield, N. J. and Fasman, G. D. (1973) Methods

Enzymol. 27D, 675-735 28. 29.

30.

Strickland, E. H. (1974) CRC Critical Rev. Biochem. 113-175 Chen, Y.-H., Yang, J. T., and Martinez, H. M. (1972) Biochemistry

11, 4120-4131 Puett, D., Ascoli, M., and Holladay, L. A. (1974) in Hormone

Binding and Target Cell Activation in the Testis (Dufau, M. L.,

by guest on March 28, 2018

http://ww

w.jbc.org/

Dow

nloaded from

7442

and Means, A. R., eds) pp. 109-124, Plenum Press, New York 35. Bigelow, C. (1967) J. Theor. Biol. 16, 187-211 31. Capaldi, R. A., and Vanderkooi, G. (1974) Proc. N&l. Acad. Sci.

U. S. A. 69, 930-932 36. Raynaud, M., Blass, J., and Turpin, A. (1957) C. R. Acad. Sci. [D]

(Paris) 245, 862-863 32. Nozaki, Y., and Tanford, C. (1971) J. Biol. Chem. 246,2211-2217 37. Bizzini, B., Turpin, A., and Raynaud, M. (1973) Eur. J. Biochem. 33. Tanford, C. (1962) J. Am. Chem. Sot. 84, 4240-4247 39, 171-181 34. Goldsack, D. E. (1970) Biopolymers 9,247-252 38. Chou, P. Y., and Fasman, G. D. (1974) Biochemistry 13, 211-222

by guest on March 28, 2018

http://ww

w.jbc.org/

Dow

nloaded from

J P Robinson, J B Picklesimer and D Puettand conformation.

Tetanus toxin. The effect of chemical modifications on toxicity, immunogenicity,

1975, 250:7435-7442.J. Biol. Chem. 

  http://www.jbc.org/content/250/18/7435Access 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/250/18/7435.full.html#ref-list-1

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

by guest on March 28, 2018

http://ww

w.jbc.org/

Dow

nloaded from