Staphylococcal Enterotoxin A

e AasTRACT: Enterotoxin A from a selected high toxin producing (3) chromatography on hydroxylar te,'nd (4) filtration on strain (15-25 Agiml in deep culture) of Staphylococcus aureums Sephadex G-75. Two major and two minor components were designated 13N-2909, has been obtained :n highly purified found by isoelectric focusing. This paucidispersity is attribform as indicated by sedimentation velocity, sedimentation uted to differences in charge. The principal component has i equilibrium, disc electrophoresis, and several immunological an isoionic point of 7.26 at 4'. The enterotoxin has a molecprocedures. The effective dose, ED5,,. by intravenous route ular weigh, of 27,800 as determined by sedimentation eqjilibto produce emesis and diarrhea in inesus monkeys is 0.03 rium. A value of 27,500 was obtaineu from polyacrylamide Mg/kg of animal weight with 95 % covidence limits of 0.017gel electrophoresis in the presence of denaturant. The molecule 0.065 usg/kg. Purification was accomplished by the following is a single polypeptide chain with one disulfide bridge and no steps: (1) removal of the toxin from culture (diluted five timel free sulfhydryl groups. Serine was identified as the C-terminal with water, and adjusted to pH 5.6) with CG-50 carboxy!ic residue but no free N term~nal was found. A complete amino acid resin; (2) chromatography on carboxymethylcellulose; acid analysis is reported.

In 1965 we reported a procedure for purifying staphylococ-Schuell), hydroxylapatite (Bio-Gel HTP, Bio-Rad Laboracal enterotoxin B,' employing chromalographic procedures tories), and Sephadex G-75 were prepared and used as dethat enableo us to purify large quantities of the toxin (lichantz scrib 2d in the stepwise procedure below.All solvents and et al., 1965).Chu et a!.(1966) have described a method for the chemicals were reagent grade.Hydrazine, benzaldehy'e, and purification of enterotoxin A that employs dialysis against phenyl isothiocyanate were prepared as previously described Carbowax to reduce the culture volume and remove some (Spero et al., 1955).The antisera used for immunological tests impurities as an initial step.While this technique is applicable during the first part of the work were made from partially to small volum= of culture, it is not practical for handling purimied *oxin.tAs more highly purified toxin was obtained, volumes necessary for the isolation of gram quantities of antisera with higher specificity were prepared for the Oudin highly purified toxin for chemical, physical, and immunologi-tests.Antisera for the Oakley (1953) tests were prepared cal studies.The present poiper describes a method for purifyfrom toxin in crude culture.ing staphylococcal enterotoxin A that employs chromato-Amino Acid Analysis.The procedures for sample preparagraphic methods that have enabled us to obtain good qu cn-tion, hydrolysis, calibration of the analyzer, the actual analytities of highly purified enterotoxin A. This report also pre-,es, and calculations were the same as previously describe,.sents a determination of the molecular weight, a complete (Spero et al., 1965).The analyzer was increased in sensitivity amino acid analysis of 0t.. protein, and studies on the t.-minal through the incorporation of a 7-to 9-mV resistor card in amino acid r'-iL !es.The implications of n, se findings on the recorder.Since the logarithmic scale is almost linear in the structur-: of this ,xin and a comparison to otner types this range and there is a large varialion in concentrations of of enterotoxin are discussed.
the various amino acids, it became necessary to carry out "epetitive runs with each hydrolysate, using di Terent volumes Materials and Methods of appl.idsample to ensure optimally sized peaks for each an;: -id.Materials.The toxin ,..as produced by culturing Stao,'slo-Cystine was determined as cysteic acid after performic -ci I coccusaureus, strain 13N-2909(Friedman and Howard, 1971), oxidation (Moore, 1963).Free sulfhydryl was detcrmine.by for 24 hr in a medium containing 3% protein hydrolysate the ElIman reagent by the modification of Janatova et al. powder (Mead Johnson), 3% NZ-amine, type NAK (Shef-(1968), in which a very high concentration of reagent is used field), 0.2% yeast extract (Difco), and 0.2% glucose added in the presence of 6 Nt gua-idine hydrochloride.Tryptophan aseptically and adjusted to pH 6.8.The resins, CG-50 (Rohm was measured spectrophotometrirally in 6 mi guanidine hydroand Haas), carboxymethylcellulose, type 20 (Schlcicher & chloride (Edelhoch, 1967).Amide nitrogen was done by the method of Stegemann (1958) except that the ammona analysis was done on the From the Biological Sciences Lahoratories, Fort DeItrick, Frederick, amirco acid analyzer restored to its original lower sensitivity., 1965).The cyanate method of Stark and Smyth (1963) original culture.Usually 60 I of culture or 300 1. of diluted and the Eriksson and Sj6quist (1960) modification of the Edculture was processed at one time with 120 g of the resin.The man method were also employed in attompts to ideaitify an settled resin, containing the toxin and impurities, was placed N-terminal residue.
in a column to accommodate a volume of rcsin 3.5 X 6) cm, Ut&racentrifuge Studies.A Spinco Model E analytical ultrawashed with about one column volume of water, and the toxin centrifuge equipped with absorption optics and an ultraviolet fractionally el-ted with 0.5 M sodium phosphate at pH 6.2 in scanner was used.The sedimentation coefficient vas measured 0.5 M sodium chloride at a flow rate of 6 ml/min.The fracwith schlieren optics, using a eap,',Iary-type doubIc-sector tions, 15 ml each, containing I mg or more of toxin per ml synthetic boundary cell.Equilibrium molecular weight was (based on absorbance at 277 nm using an extinction of 14 and determined with absorption optics in a double-sector cell Oudin tests) were pooled for further nitrification.At this point (Schachman and Edelstein, 1966).Volumes of 0.15 rnl of soluthe pooled iractions amounted to about 0.4% of the culture tion and 0.16 ml of solvent were introduced with a i iicroliter volume, 240 ml, and the yield was about 85-90% (3920 mg of syringe.Scans were made with a chart speed of 5 mm/s~c total protein, 780 mg of toxin: see Table 1).and a scan speed of 0.63 cm/min at a wavelength of 280 nm.
STEP 2. The pool of fractions from step I was dialyzed to The temperature controlling unit was not used in equilibrium reduCe the salts by dialyzing against a weaker buffer of .0-.Y8 runs.Appropriate tempcrat.reswere obtained by adjusting M sodium phosphate at pH 6.0 in 0.008 m sodium chloride.the refrigeration pressure.rhe baseline deflection was mea-This process keeps the toxin in the presence of some of the sured by accelerating the rotor to 40,000 rpm and continuing buffer and increases its stability.The toxin was adsorbed en a the run until the upper portion of the solution was completely column of CM-cellulose using 360 g (50-100 g/g of protein depleted of solute, a period of about 4 hr.We could not disbased on absorb.inceat 277 nm) prepared by soaking in distinguish between scans made at this time and after deceleratilled water I hr, decanting, and suspending it in 0.01 m monotion to the original speed.sodium phosphate solhtion.The pH was adjusted to 6.0 with lsoelectric Focusing.An LKB isoelectric focusing apparatus NaOH, readjusted to 6.U after I hr, and the cellulose placed in was used with a pH gradient between 3 and 10. a column so that the length of the bed volume (1600 ml) is at Gel Electrophoresis.Polyacrylamide disc electrophoresis least 15 times the diameter.The column was kept at 4' and was performed at pH 4.3 as described by Reisfeld et al. (1962).
washed with buffer (0.01 M, pH 6.0) at this temperature.The For the determination of molecular weight in SDS and $toxin solution was cooled to 4', passed through the column mercaptoethanol the method of Weber and Osborn (1969) was at 2 mi/min, and the column was washed with one void volume followed.A Canalco Model 6 inst-ument was employed.
(800 ml) of 0.008 m phosphate in 0.008 M sodium chloride at Ditusion Coefficient.The me, nod of Schantz and Lauffer pH 6.0.The toxin was fractionally eluted at room tempera-(1962) was employed.The protein diffuses into a column of ture with a linear gradient of sodium phosphate.0.01 M, pH •tgar ge which is subsequently extruded and sliced.A plot 6.0 to 0.05 M, pH 6.8 (1800 ml of each buffer) at 3 ml/min.of the protein concentration of each slice rs.distance on arith- The fractions (14 ml each) containing the peak of toxin were metic probability paper permits a ready determination of the selected and pooled for further purification.At this point the dliffusion coefficient.
volume of the pooled fractions was 1640 ml and the purity was about 50%, based on Oudin tests.The recovery in this Results step wa, greater thar 60%.STEP 3. The combined fractions from step 2 were adjusted Purification Procedure.The method is outlined in the followto pH 5.7 and centrifuged to remove suspended material if ing steps.
necessary.The toxin was then adsorbed on a column of hy-STEP I.The fermented culture usually assayed about 15-25 droxylapatite.One gram was used for each 3.5 mg of protein, mg of toxin/ml.When the fermentation was complete, the in this case 280 s. (bed volume 750 ml; column 5 X 38 cm, culture was cooled and passed through a high-speed Sharpies length should be at least six times the diameter).Hydroxylapacentrifuge to remove the cells.The supernatant solution was tite was prepared by suspending it in 0.008 m sodium phosdiluted with four volumes of water and adjusted to pH 5.6.phate at pH 5.7 with occasional stirring for 30 min.The sus-The toxin was removed from the diluted culture with CG-50 pension was allowed to settle so as to decant to a medium resin, eq 'ilibrated in 0,005 m sodium phosphate buffer at slurry.The column was then poured and equilibrated with 0.03 pH 5.6. 1 he equilibrated resin was stirred in the culture for M sodium phosphate buffer at pH 5.7 for at least 24 hr at a about I hr at room temperature and allowed to settle.This flow rate of no greater than 0.5 ml min.The sample was put Approximately 20/yg of toxin was applied to the gel.four times the bed volume for each buffer was used for this gradient.The peaks of protein emerging from the column (5-ml fractions) were located by the absorbance and the enterotoxin by additional Oudin tests.Figure 1 shows a typical mately 98% and the recovery in this step was about 90% elution pattern of toxin and impurities from hydroxylapatite.
(Table I).The fractions that showed a single antigen-antibody line, STeP 4. The selected fractions from step 3 were pooled and with only traces of impurity in the double-diffusion tests (Oakconcentrated by lyophilization to about one-tenth volume, ley tubes), were selected for chromatography on Sephadex.
dialyzed to remove most of the buffer salts, and lyophilized The volume of tiie pooled fractions was about 350 ml and to dryness.The dried preparation was taken up in about I/s 0 th contained 450 mg of protein by absorbance and 441 mg of the original volume of distilled water, dialyzed against 0.05 toxin by Oudin tests.The purity at this point was approxi-NM sodium phosphate at pH 6.8 in I M sodium chloride, and passed through a column of Sephadex G-75 to remove a trace amount of high molecular weight material.The Sephadex column (5 X 85 cm, void volume 530 ml) was made up in the I I [same buffer.This buffer was used to elute the toxin at a flow rate of 0.6 ml/min.The central fractions of the peak (3 ml each) 3.0 of toxin (Figure 2) from this column were pooled and dialyzed against a sodium phosphate buffer at pH 6.8 at a concentra-25 ition to bring the buffer salts to 3-5% of the protein concen-  The ratio of the absorption at 26a) nm t that at 277 nm was tcnimpralure.Tubes 116 125 were selected is' highll pur tfid cuilerotoxin A. based on a single line in Ihe ()tchichteo) aynd (Oakle) 0.45 0.47, confirming that very little, if any, nucleic acid matest,.

Characteristics of tlw
terial is present in the preparation.-Based on values in the first column corrected to 100% recovery.b Omitted from the total.'To correct for the molecular weight difference between OH and NH.-,, 0.989, is subtracted per amide residue.To correct for the mole of water on the terminal amino acids, the molecular weight of water is added.
has an isoelectric point of 7.26.(It should be noted that the Two half-cystine residue were found.The titration for free isoelectric pH obtained in electrofocusing has been bhown to sulfhydryl showed less than 0.05 mole/mole of protein.The be the isoionic point (Vesterberg anu Svensson, 1966).)This tryptophan analysis gave 2.19 residues/molecule a significantly value is appreciably higher than the value of 6.8 obtained by higher value than was found for enterotoxin B. This difference Chu el al. (1966) by means of paper electrophoresis.We feel is reflected in the extinction coefficients, 14.6 for enterotoxin that most of this difference is due to the difference in pK values A and 14.0 for enterotoxin B. for the titrating group in the isoelectric region at the tem-Partial Specific Volume.The partial specific volume of peratures of measurement.From the van't Hoff equation, erterotoxin A was calculated from the weight percentages assuming that Al for the imidazole ionization is 7 kcal (Cohn and partial specific volume of each amino acid (Cohn and and Edsall, 1943), an increase of 0.39 in pH occurs in going Edsall, 1943).A value of 0.732 ml/g was obtained.from 25 to 4*.
Ideniifcation of Terminal Residues.A free N-terminal resi-,ince the enterotoxin was demonstrated to be homogeneous due was not found by either the Edman method or the cyain gel electrophoresis in SDS after unfolding, indicating that nate method.A dinitrophenol-containing spot was found with all the protein present was of the %am, molecular size, it is the Sanger technique.It was located in the DNP-senne-DNPlikely that the four components found by isoelectric focusing threonine area in the two-dimensional system of Levy (1955).differ only in charge.
The absorbance of the spot amounted to about 20% of theo-Amino Acid Analysis.Table If shows a complete amino acid retical.However, when the bicarbonate solution containing analysis of staphylococcal enterotoxin A. The results are the extracted spot was acidified and extracted with ether based on two analyses at three periods of hydrolysis: 24, 48, the material appeared to decompose in the Blackburn-Lowand 96 hr.Serine, threonine, and tyrosine were extrapolated ther (1951) tert-amyl alcohol system, only an extremely faint to zero time according to zero-order kinetics.Only the 48-spot was found where DNP-threonine should be locRted and and 96-hr values for valine and isoleucine were averaged, most of the color was at the R, of dinitrophenol.Authentic Integral values are reported for all the amino acids but even samples of DNP-serine and DNP-threonine were run through a casual inmpection of the column giving actual residue values the identical series of operations with no significant loss.Remakes it evident how small a change is required to change covery from the rert-amyl alcohol system papers was over some of them.A relative error of but 0.25% in the aspartic 70%.It appears that the spot found initially is an artifact.acid value would increase the calculated number of residues It has not been identified.from 38 to 39.
The C-tet minal residue was easily identified by the hydrazi-nolysis technique.An excellent yield of serine was obtained.
No other amino acid was found in significant yield.TAsLE in: Some Properties of Purified Staphylococcal Enterotoxin A.

Discussion
Appearance (freeze-dried) While fluffy powder The big problems in the purification of any bacterial prod-Solubility 1. ery soluble et water uct from the culture involve (a) the increase of the quantities and salt solut'ons in the culture by the selection of higher producing strains or Type of protein Simple (contains amino the improvement of the environmeiia; conditions of cultur-acids only) ing; and (b) the development of a technique to remove the Nitrogen content (%) 16.2 product, usually present in trace quantities.Our first attempts Sedimentation coefficient (s 2 o.,_) (S) 3.03 at culturing (strain 100) produced only 3-5 Ag of enterotoxin Diffusion coefficient (D)-(cm," 9.8 X l0-7 A/ml, but selection of higher producing colonies brought sec-) the production up to 15-25 ug/ml (Friedman and Howard, Moiecular weight (sedimentation 27,800 1971).No change in the immunological properties of the equilibrium) toxin were observed in material elaborated by the higher pro-Molecular weight (gel 27,500 ducing strains. electrophoresis) The key to the isolation procedure was the establishment Molecular weight (s, D) 28,000 of conditions for the removal of the enterotoxin from culture Electrophoresis' Single component by adsorption onto CG-50 and its subsequent elution by strong Isoelectric point (isoelect, 1: 7.26 (major component) buffer and neutral salt.Virtually no enterotoxin was detect-focusing) able in the supernatant culture after the adsorption.Minor Maximum absorption (nm) 277 losses did occur upon elution from the ion-exchange resin Extinction (E'V0,) 14.6 but the overall yield for this step was highly satisfactory.
Toxicity, ED 50 intravenous (ug/kg) 0.03 (0.017-0.065)YUltimate purification was difficult.The peak from Sepha-Determined in agar gel (Schantz and Lauffer, 1962).dex G-75 was nearly symmetrical and showed constant imh Run in polyacrylamide Lel; 6-alanine-acetic acid buffer.munological activity to protein ratios in Oudin assay.How-Rhesus monkeys weighing approximately 3 kg.Numbers in ever, only the central portion of the peak gave a single line parentheses refer to 95 % confidence limits.
in Oakley double diffusion and it was necessary to eliminate the edge fractions which gave multiple lines.
In general the properties of the purified enterotoxin A were in line with the results of Chu et al. (1966) except that theirs contained minor amounts of impurities as indicated by Ouchcomponent while the reverse conversion did not occur.They terlony and Oakley tests.The properties of the enterotoxin suggest that hydrolysis of labile amide groups is the cause of are summarized in Table Ill.
both the heterogeneity and the conversion from one electro-Enterotoxin A is a simple protein composed of a single phoretic form to another.polypeptide chain, containing one disulfide bridge and no free The similarities between the two enterotoxins is striking sulfhydryl groups.It is in these respects identical with enteroand we believe that charge differences due to differing numtoxin B. This is not at all surprising since the two proteins hers of amide groups is the most likely explanation for the have the same biological action and are differentiated priobserved behavior in e6ectrofocusing.The difference in net marily on the basis of their immunological reactivity.It is charge may also be due to noncovalently bonded amino acids.not known whether a three dimensionally intact configura-An analogous explanation, conjugation with RNA or some tion is required for activity or only a relatively short linear other acidic material, has been advanced to explain the chrosection, but it would be expected that changes in amino acid matographic behavior of ribonuclease (Anfinsen and White, composition in either case would be conservative.Indeed, 1961).Finally one must entertain the possibility that these after the enterotoxin A has been sequenced, it might be poscomponents are conformational isomer.sible to eliminate parts of the molecule from active-site in- In the amino acid analysis, changes in almost every resivolvement by comparison of enterotoxins A and B and idendue were observed between A and B toxins but the distributifying those regions where drastic changes have occurred.
tion of ionizable and nonionizable residues is not drastically Fssentially all of the material in the purified preparation is altered.There is a loss of four basic residues and an increase of the same molecular size and has identical immunological of one acidic group, consistent with the lowered isoelectric reactivity.There is, however, a paucidispersity found by isopoint.Quantitatively the most striking change is the decrease elect ic focusing.We believe this does not reflect the presin the number of methionines from eight to two.ence of extraneous protein but rather several different forms The tryptophan value found by the Edelhoch spectrophoof enterotoxin A. Fhe existence of multiple comnonents in tometric method indicates two residues per molecule.We many highly purified proteins is well documented, e.g., rihopreviously reported the same number for enterotoxin B. I luang nuclease, myoglobin.Our initial report on the isolation of and Bergdoll (1970), however, in uctermining the sequence enterotoxin B (Schantz. et al.. 1965) of enterotoxin B using isoelectric focusing.They ohservedl Ws have ý epeated this dctermination with the Edelhoch method two major and at least one minor component.Both major on a %ample of the original material and obtained a value of components were toxic to monkeys and gave identical immu-1.56 and b-elieve this to be a real and accurate estimate.Hownological reactions.Furthermore.in refocusing cperiments.
L'cr, wh%:n the dvttcimination was tarried out on a newly isothe more basic component gave rise to some of the more acidic latcd preparation, wc obtained a value of 1.92.We suggest by the absorbance.The volume of the .2.0 pooled fractions in this case was 36 ml and contained 177 mg of protein by absorbance and virtually the same amount by < 1.5 IOudin tests, indicating a high degree of purity.Ouchterlony tests indicated a purity of better than 99%.Disc electrophore-1.0sis in polyacrylamide gel showed one component (Figure 3).The dialyzed solution of toxin was centrifuged to remove any 0.5s insoluble material and freeze-dried.The overall yield of puri-i fied toxin usually amounted to 20%, based on the Oudin tests on the original culture.The results of each step are summa- , noted a heterogeneity of entcrotoxin B. found only one residue.Our value was based in starch g,' el•'trophoresik.Recently Chang and Dickie on a molecular weight of 35.(XI which when recalculated to (1971) confirmed thr finding of multiple elect rophoretic fornm% the corredt figure of 28,5~(Xi is reduced to 1.60 moles/mole.

TABLE t :
Steps in the Purification of Staphylococcal Enterotoxin A.
M at pH 5.7 at a rate of I ml/min.The best fractions (tubes 85-110) contained 55% of the absorbance and 92% of the toxin.The purity of the pool was 98%.on the column with a flow rate of 0.5 ml/min.The column was then washed with one void volume (75 % of bed volume) of 0.2 M sodium phosphate (pH 5.7).Elution was carried out with a linear phosphate buffer gradient from 0.2 M at pH FIGURE 3: Disc electrophoresis of purified enterotoxin A in polyacrylamide gel (7.5%) in 0-alanine-acetic acid buffer at pH 4.3.5.7 to 0.4 m at pH 5.7 at a flow rate of 1 ml/min.A volume of

TABLE a :
Amino Acid Composition and Integral Residues of Staphvlococml Enterotoxiu A.