THE CHEMICAL NATTJRE OF SCARLET FEVER TOXIN*

BY E. S. GUZMAN BARRON, GEORGE F. DICK, AND

CARL M. LYMAN

(From the Lasker Foundation for Medical Research and the Department of Medicine of the University of Chicago, Chicago)

(Received for publication, August 16, 1940)

There is considerable confusion in the literature regarding scarlet fever toxin and “toxic or poisonous” substances obtainable from cultures of scarlet fever streptococci. In 1934, for instance, Hooker and Follensby (1) described an A and B toxin of scarlet fever; in 1936 Kodama (2) separated three substances from strep- tococci which gave skin reactions; and in 1938 Eaton (3) in a review rallied to the opinion of the plurality of the toxin.

The substance under discussion here is the specific soluble toxin of scarlet fever. There is no evidence thus far that more than one such substance exists and in order to demonstrate its presence the following tests should be made: (1) The substance should be destroyed by boiling; (2) suitable dilutions injected into the skin of numerous human subjects should give positive reactions when- ever a positive reaction is given by one skin test dose of a standard scarlet fever toxin obtained from streptococci which have produced typical experimental scarlet fever in man, and give negative reactions when the reaction to the standard toxin is negative; (3) after giving a positive reaction in a human subject who gives a positive test with standard scarlet fever toxin, it should, in proper dilution, give a negative test in the same individual when mixed with a suitable amount of specific scarlet fever antitoxin; (4) it should produce nausea, vomiting, fever, malaise, and a typical scarlet fever rash in susceptible human subjects when given hypodermically in a single dose of 1000 to 3000 skin test doses.

The substance discussed here has been subjected to all of these

* Read before the meeting of the American Society for Clinical Investi- gation at Atlantic City, May 6, 1940.

267

by guest on June 25, 2018http://w

ww

.jbc.org/D

ownloaded from

Scarlet Fever Toxin

tests and is, therefore, specific scarlet fever toxin. It was prepared according to the method given by Dick and Boor (4). The toxin, a white, fluffy substance, dissolved readily in water, giving a yellow color in concentrated solutions, and contains from 20,000 to 30,000 skin test doses per mg.

From a study of the chemical properties of this purified toxin, as reported in this paper, the conclusion is drawn that scarlet fever toxin is a protein of small molecular weight, some amino groups of which are essential for its activity. Furthermore, a study of the electrophoretic mobility of the toxin at different hydrogen ion concentrations has shown that the toxin may be freed by electrophoresis of some nitrogenous impurities still present. In all the experiments the activity was tested with the skin test for susceptibility to scarlet fever as developed by Dick and Dick (5).

Ej’ect of Temperature-It is generally agreed that scarlet fever toxin is very resistant to heat. To test the effect of temperature, the purified toxin (0.5 mg. per cc. dissolved in phosphate buffer of pH 7.4) was heated for 1 hour at different temperatures. Heat- ing for 1 hour without shaking up to 90”, and for 45 minutes up to 100” (boiling water bath) caused no loss of activity. Bubbling either purified nitrogen or oxygen through the solution when heated in a boiling water bath resulted in loss of activity. Ap- parently, agitation at this temperature produced denaturation of the toxin as agitation at lower temperatures produces denaturation of proteins of higher molecular weight (for example, hemoglobin).

E$ect of pH-It is generally believed that scarlet fever toxin is very unstable in alkaline solution, although no study of the in- fluence of hydrogen ion concentration on the activity of the toxin has been reported. To test this influence, the universal buffer of Theorell and Stenhagen (6) was used from pH 2.08 to 11.77; by the use of this buffer, the chemical nature of the electrolytes was kept identical. The toxin solutions (1 mg. of toxin dissolved in 1 cc. of water plus 4 cc. of buffer) were kept for 24 hours at room temperature (&25”). They were then neutralized, diluted to the proper dilutions for skin tests, and filtered through Jena fritted glass filters for bacteriological filtration. The pH values of the solutions were measured by both the glass and hydrogen electrodes at 25”. As can be seen in Table I, the toxin was extremely re-

by guest on June 25, 2018http://w

ww

.jbc.org/D

ownloaded from

Rarron, Dick, and Lyman 269

s&ant to changes in the hydrogen ion concentration, the stability being greater in acid solutions than in alkaline solutions; it did not lose its activity after 24 hours in 0.1 N HCl (pH l.OS), but it was destroyed when kept in 0.1 N NaOH or in buffer solutions of a pH value of 11.77.

Effect of Ketene-Acetylation with ketene has proved to be useful in determining the importance of amino groups for biological activity. It is a highly specific acetylating agent for aqueous solutions of proteins, and several investigators (Herriott and Nor-

TABLE I

Effect of pH on Activity of Scarlet Fever Toxin The toxin solution containing 1 mg. per cc. was kept for 24 hours at room

temperature (+25”); buffer, universal of Theorell and Stenhagen.

PH Size of skin reaction

1.08 2.08 3.01 3.97 5.05 6.03 7.05 7.99 9.03

10.00 11.18 11.77

nm.

22 X 16 (faint) 28 X 36 (bright) 38X40 I‘ 46X32 “ 38 X 32 “ 34 X 42 “ 25X28 “ 33 x 30 “ 32x30 ‘I 35X23 “ 35X25 I‘ Negative

throp (7), Stern and White (8), Pappenheimer (9)) have shown that primary amino groups react most rapidly, while the OH groups of tyrosine are acetylated at a much slower rate. With glucosamine in aqueous solution, only the NH2 groups react with ketene, forming N-acetyl glucosamine (Bergmann and Stern (lo)), while the OH groups in carbohydrates resulting from hy- drolysis of egg albumin are not acetylated (Neuberger (11)). Ketene destroys the activity of diphtheria toxin (Pappenheimer (9)) and of tetanus toxin (Velluz (12)), and, to some extent, the toxicity of gonococcus and meningococcus cells (Boor and Miller (13)). To test the effect of ketene on scarlet fever toxin, 50 mg.

by guest on June 25, 2018http://w

ww

.jbc.org/D

ownloaded from

1 X 8OMl dilution

nun.

27 X 25 Negative

‘I “ “

270 Scarlet Fever Toxin

of the toxin were dissolved in 10 cc. of 2 M acetate buffer of pH 5.06, put into a cellophane bag, and kept inside a 1 liter beaker containing the same acetate buffer to avoid changes in the hydro- gen ion concentration inside the bag. Ketene was generated in the apparatus devised by Herriott (14).l The ketene vapor, before reaching the toxin solution, was purified by being passed through a.flask cooled with ether and COz snow; the gas was bubbled at a rate of about 4 cc. per minute and samples were withdrawn at different intervals (5, 10,20, and 40 minutes). The solutions were then dialyzed overnight in cellophane bags at 3” in running dis- tilled water, made neutral, and tested at two dilutions, one 10

TABLE II Effect of Ketene on Activity of Scarlet Fever Toxin

50 mg. were dissolved in 10 cc. of 2 M acetate buffer, pH 5.06. Ketene was bubbled at the rate of 4 cc. per minute.

Time after k&me treatment

min.

Control 5

10 20 40

Size of skin reaction

1 X 800 dilution

mnt.

20 X 28 (bright) 18 X 15 (medium bright) 10 x 10 “ “ Negative

times stronger than the other. The toxin used as control was treated identically, except for the ketene treatment (Table II). The toxin, at a dilution of 1 X 8000 was completely inactive at the end of 5 minutes. There was some activity at a dilution of 1 X 800, but it was completely lost at the end of 40 minutes. This loss of activity, after a short treatment with ketene, may be taken as an indication that ketene destroys by acetylat#ion the primary amino groups essential for the activity of the toxin.

E$ect of Nitrous Acid-The action of nitrous acid on amino groups has been known for a long time, and was used by Levene and Van Slyke (15) for the determination of amino acids. As shown by Philpot and Small (16), nitrous acid resembles ketene in

1 We express our thanks to Dr. A. K. Boor for his kind loan of the ketene generator.

by guest on June 25, 2018http://w

ww

.jbc.org/D

ownloaded from

Barron, Dick, and Lyman 271

first attacking the primary amino groups in the protein molecule. If ketene destroyed the activity of scarlet fever toxin by acety- lation of amino groups, nitrous acid would also destroy it. Such indeed was the case. The experiments were performed at pH 4.07 in acetate buffer. To 1.15 cc. of 1 N acetic acid were added 3.85 cc. of HzO, 5 cc. of a toxin solution containing 4 mg. per cc., and 345 mg. of NaN02 (final concentration 0.5 M). After stand- ing half an hour at room temperature, the solution was neutralize& dialyzed overnight as indicated, and tested. For control, a t.oxin solution was used in which the NaNOz was replaced by an equimolecular concentration of NaCl and the acetic acid replaced by acetate buffer wit,h a final pH value of 4.06. Treatment with

TABLE III E$ect of Nitrous Acid and Iodine on Activity of Scarlet Fever Toxin

Substance PH

Control. Sodium nitrite (0.5 &I). Control. . . . . Iodine (0.002 N).

“ (0.002 “). I‘ (0.002 1‘). .

4.06 4.07 7.02 7.02 4.63 3.01

Time of action

h.

0.5

6.0 6.0 6.0

- Sire of skin reaction

?mn.

20 X 18 (bright) Negative 30 X 32 (bright) Negative

‘I

20 X 20 (bright)

nitrous acid destroyed the activity of the toxin completely, thus giving more support to the conclusion drawn from the experiments with ketene (Table III).

E$ect of Iodine-Iodine may react with proteins either as an oxidizing agent, or as a substitution agent, its activity as an oxidiz- ing agent being increased as the hydrogen ion concentration in- creases, while its activity as a substitution agent diminishes under the same conditions. The effect of iodine (0.002 N) on scarlet fever toxin was studied at three pH values, 7.02,4.63, and 3.01, the solutions being kept in the presence of iodine for 6 hours at 25”. At the end of this time the iodine was destroyed by the addition of cysteine, the solution was diluted, dialyzed overnight, and the skin tests performed (Table III). The toxin was destroyed by iodine at pH values of 7.02 and 4.63, while at a pH value of 3.01 it still retained activity (control, 30 X 32 mm.; with iodine, 20 X

by guest on June 25, 2018http://w

ww

.jbc.org/D

ownloaded from

272 Scarlet Fever Toxin

20 mm.). Doubtless, this iodine inactivation is due to action of the reagent on the amino groups of the toxin more than to strict oxidizing action, for when the oxidizing power was increased by making the solution more acid there was no inactivation.

E$ect of Porphy-indin and Other Oxidixing and Reducing Xub- stances-Porphyrindin, a dye of highly positive oxidation-reduction potential (E’o at pH 7.0 = +0.57 volt) was recommended by

TABLE IV Effect of Porphyrin&n and Other Oxidizing and Reducing Substances on

Activity of Scarlet Fever Toxin at pH 7.06

Substance

Control, .............................

Porphyrindin (0.001 M). ..............

Control. ............................ Hz02 (0.000 M) .......................

cU& (0.00015 M). ...................

Oxidized glutathione (0.033 5%). ....... Iodoacetic acid (0.001 RI). ............ Cysteine (0.023 M). ...................

Glutathione (0.023 M) ................ Na&04 (0.02 M) ..................... H2S (saturated). ..................... Pt + Hz ............................. Iodoacetamide (0.01 M). ............. Alloxan (0.05 M) .................. Caffeine (0.01 M). .................... Sulfanilamide (0.005 M). ..............

“ ultraviolet irradiation (0.005 M) ...........................

Time of action

~-.- hrs. I-

/

0.5 ! I

4.5 / 4.0 4.0 4.0 4.0 4.0 7.0 5.0

/

6.0 4.0 / 5.0 j 4.0 0.1

0.1 /

Size of skin reaction

nzm. 35 X 30 (bright) Kegat,ive 30 X 30 (bright) 28 x 13 “ 22x20 IL 28X15 “ 35x30 “ 19 x 20 “ 28 X 21 “ 20X23 “ 28 x 20 ‘I 29 x 22 “ 35x30 (’ 28X25 “ 33x35 “ 35 X 25 (faint)

30X20 “

Kuhn and Desnuelle (17) as a reagent for the rapid oxidation of -SH groups in the protein molecule. As such it has been ex- tensively used by Greenstein (18). The effect of porphyrindin, kindly supplied by Dr. L. Hellerman, was studied at pH 7.06 by treating the toxin with the dye (final concentration, 0.001 M) for half an hour, the skin tests being performed after dialysis and proper dilution. Porphyrindin, under such conditions, destroyed the activity of the toxin (Table IV).

The inactivation of the toxin by porphyrindin made it necessary

by guest on June 25, 2018http://w

ww

.jbc.org/D

ownloaded from

Barron, Dick, and Lyman 273

to study the effect of substances which combine with either the -SH or -S-S- groups of proteins, inasmuch as the presence of these groups as the active constituents of certain hormones (insulin, du Vigneaud (19)), enzymes (urea, Perlzweig (20), papain, Hellerman and Perkins (21), cholinesterase, Nachmansohn and Lederer (22)), and toxins (snake venoms, Micheel and Emde (23), Binet, Weller, and Robillard (24)) has been recently established. Of the compounds known to react easily with either -SH or -S-S- groups under physiological conditions (at neutral reac- tion and temperatures not above 38”) the following were used: as reagents for -SH groups, CuCIZ (0.00015 M), CHJCOOH (0.01 M), CHzICONHz (0.01 M), oxidized glutathione (0.023 M), alloxan (0.05 M), caffeine (0.01 M), Hz02 (0.009 M); as reagents for-S- S- groups, cysteine (0.02 M), glutathione (0.023 M), NazSz04 (0.02 M), H2S (saturated), and Pt black plus Hz (Table IV). Since none of these substances had any effect on t’he activity of the toxin, it must be concluded that no active -SH or -S-S- groups are present. The inhibiting action of porphyrindin must therefore be attributed to de&ruction of some active group other than a sulfhydryl group. To elucidate this question, a number of sub- stances containing primary amine groups (glycine, phenylalanine, histidine, t)ryptophane, tyrosine, t,yramine, histamine, 1,4- diaminobutane, diethylamine) were treated with porphyrindin at pH 7.33, 25”, a ratio of IO parts of amine to 1 of dye being used. With tyrosine and tyramine the reaction was immediate, as shown by prompt and complete reduction of the dye; with tryptophane, the dye was half reduced at the end of 1.75 hours. Porphyrindin was reduced by tyrosine and tyramine, presumably by oxidation of their OH groups. From these experiments it is suggested that tyrosine may be one of t)he active groups of the toxin, destroyed by ketene by action on its NH2 group and nitrous acid, by prophy- rindin and ketene by action on its OH group.

Carpenter and Barbour (25) reported tha.t the toxins from Staphylococcus aureus and Clostridium welchii were inactivated by treatment with sulfanilamide. Neit,her sulfanilamide hor the oxidation product obbained by treatment with ultraviolet light had any effect on the activity of scarlet fever toxin (Table IV).

Effect of Pepsin and, Trypsin-Contradictory reports have been published on the effect of trypsin on the activity of scarlet fever

by guest on June 25, 2018http://w

ww

.jbc.org/D

ownloaded from

274 Scarlet Fever Toxin

toxin. Huntoon (26) states that the toxin is destroyed by trypsin, but Kodama (2) on preparing his toxin submitted the solution to the action of trypsin to hydrolyze the protein impurities. To test the effect of trypsin under optimum conditions for the activity of the enzyme, toxin solutions in 0.01 M Na2HP04, containing 1 mg. per cc., were incubated at 38” with trypsin (0.13 mg. of active powder per cc. of toxin solution) for 3, 6, 9, 24, and 48 hours. At the end of the incubation period the solutions were neutralized with 0.01 M KHzPOd, diluted, and tested. When the incubation period was extended to 48 hours, more trypsin was added after 24 hours incubation.2 In no case did trypsin destroy the activity

TABLE V

Effect of Pepsin and Trypsin on Activity of Scarlet Fever Toxin Incubated at 38’.

Substance

Control. Pepsin. .

‘I . Control. Pepsin...................... Control. . Trypsin.

-

-- PH

3.01 3.01 3.01 3.01 3.01 8.5 8.5

Time of action

hrs.

12 6

12 24 24 24 24

Size of skin reaction .-

nml. 35 X 25 (bright) 20x20 “ 30x25 ‘I 15 X 10 (faint) 15 X 18 “ 35 X 30 (bright) 38X22 ‘I

of the toxin. The same results were found when pepsin was used as the protcolytic enzyme. The experiments were performed at a pH value optimum for enzyme activity, namely 3.01 (Table V).

Isoelectric Point of Scarlet Fever Toxin-The electrophoretic mobility of the scarlet fever toxin was determined with Theorell’s cataphoresis apparatus for test purposes (27) at 3.5” to avoid any danger of convection currents. The current intensity was kept at 5 milliamperes and the experiments terminated at the end of 3 hours. The toxin, dissolved in the buffer solution, was in- troduced in the U part of the apparatus, leaving five cells in both branches of the apparatus free for migration of the toxin. At the end of the experiment, the rubber disks were moved SO as to

* In these experiments, trypsin and pepsin solutions were previously tested for skin reactions and were found to give negative reactions.

by guest on June 25, 2018http://w

ww

.jbc.org/D

ownloaded from

Barron, Dick, and Lyman 275

separate the different cells; the fluid of the cells was pipetted off, and the nitrogen content determined. After neutralization, dilution, and filtration, the skin tests were performed with the contents of each cell. The electrophoretic mobility of the toxin was studied at five pH values (Table VI), 4.56, 5.24, 5.76, 6.08, 6.98. (The pH values quoted here are those calculated from measurements made at 25”, and by the use of the temperature coefficients given by Clark (28).) The toxin migrated towards

TABLE VI Electrophoretic Mobility of Scarlet Fever Toxin As Detemined with Theorell’s

Cataphoresis Apparatus T = 3.5”, + = positive test for toxin activity, - = negative test for

toxin activity, U = mg. per cent of Nz remaining in the U part of the cataphoresis apparatus. The figures represent mg. per cent of Nz in the cell.

Anode Cathode

PH Cell Cell Cell

1 2 3 Cgll Cell

6 Cell Cyll Cfl $11

6 -_- ~~-- ---

4.56 2.942.45 1.94 2.02 1.56 1.480.7 0.310.24 +++ +++---

5.24 1.480.33 0.25 0.25 0.25 ? 2.0 1.481.3$ ++-++----

5.76 1.060.19 0.19 0.31 0.51 0.292.021.942.9( +-- - - -+++

6.08 1.020.17 0.15 0.06 0.15 0.052.941.940.81 ++---- + + +

6.98 0.730.38 0.11 0.31 0.25 0.133.561.841.3 _-- - - -+++

Cfl 1 cp 1 Cfl / v

---- 0.13 0.19 0.25 2.96 - - - -

0.62 0.25 0.25 3.1 - - - _

2.10 0.41 2.54 + + -

0.32 0.18 0.32 3.05 - - -

0.71 0.48 0.78 + + -

the cathode from pH 5.76 on, and towards the anode from pH 5.24 to greater acidity. At pH 5.24, when the toxin migrated to the anode, some nitrogenous impurity migrated in the opposite direction, towards the cathode. At pH 5.24 and 4.56, when the toxin migrated to the anode, the average N content of the cells was 2.66 per cent, while the N content of the cells in the cathode branch, with no toxin, was 1.94 per cent. Electrophoresis of the toxin at pH 5.0 will therefore allow further purification of the toxin, for it will withdraw this nitrogenous impurity. Experiments in this direction are in progress.

by guest on June 25, 2018http://w

ww

.jbc.org/D

ownloaded from

276 Scarlet Fever Toxin

The isoelectric point of the toxin, as determined by calculation of the ionic mobility from the data given in Table V was 5.55 (Fig. 1). This value is different from that given by Shinn (29), who reported it to be between pH 7.0 and 7.5; his experiments were carried out without proper control of the conditions necessary to perform electrophoretic studies, conditions studied carefully by Tiselius (30).

Ultrafiltration of Scarlet Fever‘ Toxin-To estimate roughly the size of the scarlet fever toxin, the ultrafiltration of toxin solutions through graded collodion membranes was performed. The membranes were prepared according to the methods described by

I ta

0

-2

-4

B P" 6 ?:

FIG. 1. Isoelectric point of scarlet fever toxin. Temperature, 3.5”. Ordinate, ionic mobility X 106; abscissa, pH values.

Elford (31) and Bauer and Hughes (32) in a room kept at constant temperature and humidity (24“ and 60 to 65 per cent relative humidity); the degree of humidity was diminished to that existing in the room for the preparat,ion of membranes of small port size. The membranes prepared from parlodion manufactured by the Mallinckrodt Chemical Works were similar in property and size of pores to those described by Bauer and Hughes. The average pore size of the membranes was calculated by an application of Poiseuille’s law governing the passage of water through a capil- lary tube.3 To test these membranes the following protein solu- tions of varied molecular weight were used: hemoglobin, prepared according to the method of Sidwell et al. (33), myoglobin, according

8 For the equations used to calculate the pore size of membranes, see Bauer and Hughes (32).

by guest on June 25, 2018http://w

ww

.jbc.org/D

ownloaded from

Barron, Dick, and Lyman 277

to the method of Theorell (34), cytochrome q4 according to the method of Keilin and Hartree (35), clupein, purchased from The British Drug Hauses, Ltd. Filtrations were carried out under a pressure of about 780 mm. of Hg, the filter being connected to a nitrogen tank. The toxin filtered through all the membranes that let hemoglobin (mol. wt,. SS,OOO), carbon monoxide myoglobin (mol. wt,. 17,500), and cytochromc c (mol. wt. 13,000) pass through; moreover, it filtered through a cellophane membrane with a pore size less than 1 rnp. The toxin did not filter through

TABLE VII Ultrafiltration of Scarlet Fever Toxin

The size of the control reaction was 40 X 35 mm., bright.

‘ItP 71.6 39.2 16.3 8.4 3.1 2.1 1.3

-

-

64% moi. wt.

+++ +++ +++ +++

0 0 0 0 0

-

(

_

-

PROXY 17,500’

mol. wt.

+++ +++ +++ +++ +++

0 0 0 0

-

E

-

-

cyto- lhrome c,

13,ocm mol. wt.

+++ +++ +++ +++

++ + 0 0 0

-

- _

-

C~glW&

mol. wt.

+++ +++ ++-I- +++ +++ +++ +++ +++ +++

Toxin, aim of skin reaction

nun.

42 X 40 (bright) 35x30 (‘ 36 X 28 “ 30X22 “ 28X20 ” 27 X 18 “ 20 X 25 (medium bright) 30 X 20 (faint) Negative

* Cellophane from casings. i Cellophane No. 600.

cellophane No. 600 although clupein (mol. wt. from 2000 to 4000) did (Table VII). We are well aware that the ultrafiltration method through collodion membranes, even if the membranes were of uniform pore size, can be applied only to molecules of spherical or globular configuraCon and to substances that are not adsorbed by the membrane. However, since the toxin passes through membranes that retain cytochrome c, while it does not pass through membranes which let clupein through, the evidence is strong in favor of the assumption that the scarlet fever toxin

4 Care was taken to perform the ultrafiltrations at hydrogen ion concen- trations optimum for non-adsorption by the membranes.

by guest on June 25, 2018http://w

ww

.jbc.org/D

ownloaded from

278 Scarlet Fever Toxin

is a small protein with a molecular weight less than that of cyto- chrome c (13,000) and perhaps more than that of clupein (2000 to 4000). It may be recalled that the molecular weight of diph- theria toxin is about 72,000 (Lundgren, Pappenheimer, and Wil- liams (36)) and that of crotoxin about 33,000 (Slotka and Fraenkel-Conrat (37)).

Some Components of Scarlet Fever Toxin-Dick and Boor (4) reported that the purified toxin used in these experiments con- tained 11.83 and 1.04 gm. per cent of nitrogen by form01 titration, i.e. free amino nitrogen. Korschun et al. (38), who prepared a toxin by alcohol precipitation, reported a total nitrogen content

FIG. 2. Color intensity of toxin solutions with the orcinol method after warming at different times. Abscissa, time in minutes; ordinat,e, extinc- tion measurement; Curve 1, readings with Filter 547; Curve 2, readings with Filter S53.

of 10.02 per cent and Stock (39) in a preliminary note reported a total nitrogen of 7 per cent. Undoubtedly, Stock was dealing with a substance not identical with the scarlet fever toxin as previously defined, since he found no free amino nitrogen, while the toxin as prepared by Dick and Boor (4), Korschun (38), and Veldee (40) does contain free amino nitrogen. As found by Korschun et al. (38), the toxin gave a red color with the biuret reaction, an indication that there is a peptide linkage and that the length of the peptide chain is small. It gave a positive Millon test, an indication of the presence of tyrosine in the molecule, and a faintly positive Hopkins-Cole test. It also gave a strongly positive Molisch reaction.

For the determination of the carbohydrate content, two methods were used: (1) titration with copper with the micromethod of

by guest on June 25, 2018http://w

ww

.jbc.org/D

ownloaded from

Barron, Dick, and Lyman 279

Somogyi (41) after previous digestion for 3 hours with 5 N HzS04 (see Lyman and Barron for the details of the procedure (42)), and (2) the calorimetric orcinol method of Sorensen and Haugaard (43) as modified by Hewitt (44). The carbohydrate content by the first method was 1.54 gm. per cent; by the second method, 1.43 gm. per cent. On measuring the extinction coefficient of the blue solutions obtained after heating the toxin with the orcinol reagent at different intervals of time, the curve obtained with Filters 547 and 553 of the Zeiss Pulfrich photometer was similar to that re- ported by Sorensen (45) for mixtures containing 3 parts of mannose and 1 of galactose (Fig. 2). The carbohydrate content of human serum is 3.2 per cent; that of horse serum, 2.4 per cent; of albumin, 1.08 per cent; of seroglycoid, 8.4 per cent; of ovomucoid, 20 per

TABLE VIII Some Components of Scarlet Fever Toxin

Substance Gm. per cent

Total nitrogen (Kjeldahl). Nitrogen (form01 titration). . . . Total carbohydrate (copper titration).

‘I ‘I (orcinol, calorimetric) Glucosamine (calorimetric). . . . .

11.33 1.04 1.54 1.43 0.73

cent (Hewitt (46)). Thus the carbohydrate content of toxin is as low as that of proteins with the lowest carbohydrate content.

The glucosamine content of the scarlet fever toxin was also very low, 0.73 gm. per cent. The toxin, dissolved in 4 N HCl, was subjected to 8 hours hydrolysis in a sealed tube in a boiling water bath. After neutralization, the determination of glucosamine was performed in a Zeiss Pulfrich photometer with Filter 553 by the calorimetric method of Elson and Morgan (47) modified by Hewitt (46). These low figures for carbohydrate and glucos- amine content differ radically from Stock’s preliminary report (39) on the erythrogenic toxin of Streptococcus scarlatinae in which it is stated that his substance contained large quantities of carbo- hydrates and glucosamine.

Since the toxin contained no phosphorus (no color reaction was obtained when the determination of total phosphates was at- tempted by the calorimetric method of Whitehorn (48)), it may be concluded that no nucleoproteins exist in the toxin (Table VIII).

by guest on June 25, 2018http://w

ww

.jbc.org/D

ownloaded from

280 Scarlet Fever Toxin

DISCUSSION

From a study of the chemical properties of scarlet fever toxin described in this paper, namely the nitrogen content, the presence of free amino nitrogen, the resistance to high temperatures and to wide changes in the hydrogen ion concenbration, the precipitation by high concentrations of ammonium sulfate, the color of the biuret reaction, and the passage by ultrafiltration through col- lodion membranes which do not let cytochrome c through, it is logical to conclude that the toxin is a protein of small molecular weight. The toxin is not only resistant to heat and to changes of pH value, but also to the action of active proteolytic enzymes such as pepsin and trypsin. The loss of activity of the toxin by brief treatment with ketene, nitrous acid, and iodine in neutral solutions is good evidence in favor of the theory that the activity is related to the presence of amino groups in the protein molecule, amino groups which arc destroyed by these agents. As porphyrin- din had been used for the estimation of certain mercaptans and of the sulfhydryl groups of certain proteins, the destruction of toxin activity by this reagent made it necessary to conduct a series of experiments with substances known to combine either with sulfhydryl or the disulfide groups. When none of these substances destroyed the activity of the toxin, it was found that porphyrindin was readily reduced by tyrosine, tyramine, and to some extent by tryptophane. Porphyrindin must therefore be rejected as a reagent for the detection of sulfhydryl groups of proteins, and the inactivation of the toxin by the dye may be attributed to its action on the OH group of tyrosine. (The positive Millon and Hopkins- Cole reactions in toxin arc indicative of the presence of tyrosine and tryptophane in the toxin.)

The experiments on the electrophoretic mobility of the toxin are furt,her evidence of its prot,cin nature, for whether the toxin was made to migrate to the anode or to the cathode branch of the cell, a large portion of the nitrogen was retained in the active material. By assuming that the pure toxin contains 10 per cent nitrogen, it is calculated from these electrophoretic experiments that the material is 35 per cent pure. Since at pH values some- what lower than those of the isoelectric point of the toxin (pH 5.55 at 3.5”) the protein impurity migrates towards the cathode

by guest on June 25, 2018http://w

ww

.jbc.org/D

ownloaded from

Barron, Dick, and Lyman

branch of the cataphoresis apparatus, there is here an easy method for further purification of the toxin.

The low carbohydrate and glucosamine content of the toxin, as well as the absence of phosphorus, speaks against, considering it as a conjugated protein.

SUMMARY

Scarlet fever toxin is very resistant to the action of high tempera- tures (up to 100’ for 15 minutes), wide variations in the hydrogen ion concentration (from pH 1.08 to ll.Ol), proteolytic enzymes (pepsin and trypsin), and a number of oxidizing and reducing agents. Its activity is destroyed by substances known to react with amino groups, such as ketene and nitrous acid. Evidence has been presented to support the view that inactivation produced by iodine and porphyrindin is a.lso due to reaction with amino groups. From ultrafiltration with membranes of graded porosity it is concluded t,hat the toxin is a protein of small molecular weight, between 13,000 and 4000. The isoelectric point of the toxin is 5.55 at 3.5”. As the carbohydrate and glucosamine content of the toxin is very low, and there is no phosphorus, it is concluded that the toxin is not a conjugated protein.

BIBLIOGRAPHY

1. Hooker, S. B., and Follensby, E. M., J. Immunol., 27, 177 (1934). 2. Kodama, T., Kitasato Arch. Exp. Med., 13, 101 (1936). 3. Eaton, M. D., Bact. Rev., 2, 3 (1938). 4. Dick, G. F., and Boor, A. K., .I. Infect. Dis., 67, 164 (1935). 5. Dick, G. F., and Dick, G. H., Scarlet fever, Chicago (1938). 6. Theorell, T., and Stenhagen, E., Biochem. Z., 299, 416 (1938). 7. Herriott, R. M., and Northrop, J. H., .I. Gen. Physiol., 18, 35 (1934). 8. Stern, K. G., and White, A., .I. Biol. Chem., 122, 371 (1937-38). 9. Pappenheimer, A. M., Jr., .I. Biol. Chem., 126, 201 (1938).

10. Bergmann, M., and Stern, F., Ber. them. Ges., 63, 437 (1930). 11. Neuberger, A., Biochem. J., 32, 1443 (1938). 12. Velluz, L., Compt. rend. Sot. biol., 127, 35 (1938). 13. Boor, A. K., and Miller, P., PTOC. Sot. Exp. BioZ. and Med., 40, 512

(1939). 14. Herriott, R. M., .I. Gen. Physiol., 18, 69 (1934). 15. Levene, P. A., and Van Slyke, D. D., J. BioZ. Chem., 12, 301 (1912). 16. Philpot, J. St. L., and Small, P. A., Biochem. J., 32,542 (1938). 17. Kuhn, R., and Desnuelle, P., Z. physiol. Chem., 261, 14 (1938). 18. Greenstein, J. P., J. BioZ. Chem., 130,519 (1939).

by guest on June 25, 2018http://w

ww

.jbc.org/D

ownloaded from

Scarlet Fever Toxin

19. du Vigneaud, V., J. Biol. Chem., 76, 393 (1927). 20. Perlzweig, W., Science, 76, 435 (1932). 21. Hellerman, L., and Perkins, M. E., J. Biol. Chem., 107, 241 (1934). 22. Nachmansohn, D., and Lederer, E., Bull. Sot. chim. biol., 21,797 (1939). 23. Micheel, F., and Emde, H., Ber. &em. Ges., 72, 1724 (1939). 24. Binet, L., Weller, G., and Robillard, E., Corn@. rend. Sot. biol., 131,

954 (1939). 25. Carpenter, C. M., and Barbour, G. M., Proc. Sot. Exp. Biol. and Med.,

41, 354 (1939). 26. Huntoon, F. M., Proc. Sot. Exp. Biol. and Med., 21,513 (1924). 27. Theorell, H., Biochem. Z., 276, 1 (1934). 28. Clark, W. M., The determination of hydrogen ions, Baltimore, 3rd

edition (1928). 29. Shinn, L. E., J. Infect. Dis., 49, 281 (1931). 30. Tiselius, A., Inaugural dissertation, Upsala (1930). 31. Elford, W. J., J. Path. and Bad., 34, 505 (1931). 32. Bauer, J. H., and Hughes, T. P., J. Gen. Physiol., 18, 143 (1934). 33. Sidwell, A. E., Jr., Munch, R. H., Barron, E. S. G., and Hogness, T. R.,

J. Biol. Chem., 123, 335 (1938). 34. Theorell, H., Biochem. Z., 262, 1 (1932). 35. Keilin, D., and Hartree, E. F., Proc. Roy. Sot. London, Series B, 122,

298 (1937). 36. Lundgren, H. P., Pappenheimer, A. M., Jr., and Williams, J. W., J. Am.

Chem. Sot., 61,533 (1939). 37. Slotka, K. H., and Fraenkel-Conrat, H. L., Ber. them. Ges., 71, 1076

(1938). 38. Korschun, S. W., Krestownikowa, W. A., and Rjachina, E. M., 2.

Immunitiitsjorsch. u. exp. Therap., 61, 289 (1929). 39. Stock, A. H., Am. J. Path., 13, 618 (1937). 40. Veldee, M. V., Pub. Health Rep., U. S. P. H. S., 63, 909 (1938). 41. Somogyi, M., J. Biol. Chem., 117,771 (1937). 42. Lyman, C. L., and Barron, E. S. G., J. Biol. Chem., 132, 293 (1940). 43. Serensen, M., and Haugaard, G., Biochem. Z., 260,247 (1933). 44. Hewitt, L. E., Biochem. J., 30, 2229 (1936). 45. S$rensen, M., Compt.-rend. trav. Lab. Cadsberg, 20, No. 3 (1934). 46. Hewitt, L. E., Biochem. J., 32, 1554 (1938). 47. Elson, L. A., and Morgan, T. J., Biochem. J., 27, 1824 (1933). 48. Whitehorn, J. C., J. Biol. Chem., 62, 133 (1924-25).

by guest on June 25, 2018http://w

ww

.jbc.org/D

ownloaded from

M. LymanE. S. Guzman Barron, George F. Dick and Carl

SCARLET FEVER TOXINTHE CHEMICAL NATURE OF

1941, 137:267-282.J. Biol. Chem. 

  http://www.jbc.org/content/137/1/267.citation

Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

alerts to choose from all of JBC's e-mailClick here

  tml#ref-list-1

http://www.jbc.org/content/137/1/267.citation.full.haccessed free atThis article cites 0 references, 0 of which can be

by guest on June 25, 2018http://w

ww

.jbc.org/D

ownloaded from