Vol. 50, No. 3INFECTION AND IMMUNITY, Dec. 1985, p. 869-8760019-9567/85/120869-08$02.00/0Copyright ©) 1985, American Society for Microbiology

Skin Reactivity of Unsensitized Monkeys upon Challenge withStaphylococcal Enterotoxin B: A New Approach for Investigating

the Site of Toxin ActionPETER H. SCHEUBER,' JOCHEN R. GOLECKI,- BOTHO KICKHOFEN,' DORIS SCHEEL,' GEORG BECK,' AND

DIETRICH K. HAMMER'*Max-Planck-Institut fiir Immunbiologie,' and Institlt Biologie II, Mikrobiologie, Unii'ersitiit Feiblurg,, D-7800 Freibirg,

Federal Reputblic of Germaniy

Received 14 Junel985/Accepted 19 August 1985

The correlation between skin tests and emetic responses in unsensitized monkeys was used to elucidate thecellular site of action of staphylococcal enterotoxin B (SEB). Evidence is presented that SEB administeredintradermally provoked immediate-type skin reactions associated with mild degranulation of cutaneous mastcells. The cytoplasma showed signs of synthetic and metabolic activity, with formation of vesicles and increasedprominence of mitochondria. Carboxymethylation of histidine residues of SEB altered the molecule (cSEB)from more alkaline components to more acidic species with increased microheterogeneity. This modificationcaused a loss in toxicity and completely abrogated the skin-sensitizing activity without changing theimmunological specificity. cSEB, however, could compete with SEB for binding sites on the target cell surface.Previously, compound 48/80-treated skin sites behaved refractively to challenge with SEB, indicating thatmediators from cutaneous mast cells are required for SEB-induced skin reactions. Skin reactions as well asemetic responses challenged with SEB were completely inhibited by H2 receptor antagonists and calciumchannel blockers but not by HI antihistamine or competitive antagonists of serotonin. This new approachprovides a model for investigating the mechanisms of SEB action.

Staphylococcal enterotoxin (SE), consisting of a singlepolypeptide chain, is responsible for the most prevalent typeof foodborne debilating enteric intoxication in humans.

Administration of SE to monkeys or human volunteersproduces emesis and diarrhea, the classic symptoms of foodpoisoning (1), whereas laboratory animals show little if anyclinical effect to peroral challenge with SEB. There aresequential morphologic (11) and ultrastructural (16) alter-ations associated with the oral administration of SE, and ithas been suggested that the toxin-induced vomiting responsefollows stimulation of local neural receptors in the gastroin-testinal tract (4).Although considerable efforts have been expended on

attempts to clarify the pathogenesis of enterotoxemia, thetarget cells and subcellular structures involved in the intes-tinal site of SEB action still remain obscure.

In this report a new in vivo model is described for studiesof SEB function by the direct skin test in Cynomolgusmonkeys. A series of experiments provide evidence thatSEB causes immediate-type skin reactions by degranulationof cutaneous mast cells. By using pharmacological agents, itcould be established that the SEB-induced cutaneous hyper-sensitivity reaction and most notably the enteric intoxicationas well are completely inhibited by H2 competitive antago-nists of histamine and by calcium channel blockers. There-fore, the skin reactivity of SEB in monkeys may be avaluable tool for studies of the biologic function of the toxinand H2 receptor antagonists, and calcium channel blockersmay prove to be a very useful prophylaxis or treatment ofenteritic intoxications after challenge with SEB.

* Corresponding author.

MATERIALS AND METHODS

Chemicals. Diphenhydramine (Benadryl) and ketaminehydrochloride (Ketavet) (Parke, Davis & Co., Detroit,Mich.), cimetidine hydrochloride (Tagamet; Smith Kline,Dauelsberg GmbH, Gottingen, Federal Republic of Ger-many), methysergide (Sandoz Pharmaceuticals, Basel,Switzerland), doxantrazole (Burroughs Wellcome Co.,Research Triangle Park, N.C.), diltiazem hydrochloride(Godecke AG, Berlin, Federal Republic of Germany),histamine dihydrochloride, compound 48/80, bovine serumalbumin, and bromoacetic acid (Sigma Chemical Co., St.Louis, Mo.), methylene blue and azur II (E. Merck AG,Darmstadt, Federal Republic of Germany), Evans blue andAlu gel S (Serva, Heidelberg, Federal Republic of Germany),cyclophosphamide (Endoxan; Asta, Brackwede, FederalRepublic of Germany), and methanesulfonic acid (Merck)were obtained from the indicated sources.

Animals. Cynomolgus monkeys (Macaca fisciciudaris),weighing 2.5 to 3.0 kg each, kindly provided by BayerischeLandesimpfanstalt, Munich, Federal Republic of Germany,were selected from a colony of primates that were free ofdiseases and of SEB-specific antibodies. Laboratory animalswere cared for in accordance with published animal protec-tion procedures (1Sa). BALB/c (H-2d) mice, LEW (RT-11),and BS (RT-1v) rats were obtained from the breeding stockof the Max Planck Institute, and female strain 13 guinea pigswere supplied by the National Institutes of Health,Bethesda, Md., and further propagated by strict brother-sister mating. All animals were maintained on standard dietand water ad libitum and used routinely at 6 to 8 weeks(mice), 10 to 12 weeks (rats), or 3 months (guinea pigs) ofage.

Toxins. SEB was highly purified by the method of Ende et

869

on May 30, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

870 SCHEUBER ET AL.

al. (5). Carboxymethylation of SEB (cSEB) was effected bythe method of Harris and Hill (7) by mixing 0.4 Mbromoacetic acid (adjusted to pH 7.0 with 5 N NaOH) withan equal volume of 0.6% SEB in 0.5 M potassium phosphatebuffer (pH 7.0). The solutions were incubated in the dark atroom temperature for 7 days. At the end of the reactionperiod, the solution was dialyzed at 4°C against severalchanges of distilled water, and the protein concentration wasdetermined by UV absorbance with absorbancy values ofA1 m = 14.4 at 277 nm (19). Amino acid analysis wasperformed on a Chromakon 500 analyzer (Kontron, Munich,Federal Republic of Germany) after hydrolysis in 6 N HCI(Ultrapur; Merck) for 24 h at 110°C. The integration con-stants for carboxymethylated derivatives of histidine usedwere described previously (3).

Immunization. For primary immunization, BALB/c miceand LEW rats were injected intraperitoneally with SEB (1and 10 jLg, respectively) mixed with 4 mg of Alu gel S in atotal volume of 0.5 ml of phosphate-buffered saline (PBS).After 28-day intervals all animals were pretreated intraperi-toneally with 175 mg of cyclophosphamide per kg, second-arily challenged with SEB 2 days later, and bled sequentiallyfor determination of immunoglobulin E (IgE) anti-SEB anti-body.

Strain 13 guinea pigs were primed intraperitoneally with 10jg of SEB adsorbed on 4 mg of Alu gel S in a total volumeof 1.0 ml and challenged three times at 30-day intervals withthe same dose. All guinea pigs were pretreated intraperito-neally with 100 mg of cyclophosphamide per kg 2 daysbefore each antigenic exposure.

Passive cutaneous anaphylaxis (PCA). Mouse and rat IgEantibodies were titrated in the dorsal skin of female BS ratsthat were sensitized intradermally with 0.1 ml of antiserumserially diluted twofold in PBS supplemented with 0.03%bovine serum albumin. After 72 h they were challengedintravenously with 1 ml of PBS containing 200 jig of SEBand 1% Evans blue. The titer was expressed as a reciprocalof serum dilutions yielding 5-mm-diameter reactions.Guinea pig antisera at appropriate dilutions were injected

intradermally into the dorsal skin surface of the homologousstrain. After sensitization periods of 72 h, the animals werechallenged by intravenous injection of 200 jig of SEB in 1 mlof PBS containing 1% Evans blue. The resulting blueingreactions were read 30 min after challenge and recorded as areciprocal of the highest dilution giving a threshold (5-mm-diameter) reaction.

In another series of experiments, monkeys were passivelysensitized by intradermal injection of 0.1 ml of serial dilu-tions of rat IgE anti-SEB antibodies into the ventral skinsurface. After 72 h the monkeys received 2 ml of 1% Evansblue intravenously and were challenged with 0.05-ml por-tions of SEB or cSEB (40 jig/ml) into sensitized skinsites. The size of blueing reactions was measured 30 minlater.

Direct skin test. A series of skin testing experiments wereperformed in unsensitized monkeys. Prior to any form ofchallenge, each monkey received 2 ml of 1% Evans blue dyeintravenously. Immediately after this, duplicate 0.05-mlsamples of serial dilutions of SEB or cSEB, compound 48/80(500 jig/ml), and PBS as controls were injected intradermallyinto the anterior aspect of the monkey. To evaluate the effectof compound 48/80 on the response to intradermal SEB,48/80 (25 Vig) was injected, followed by rechallenge at a 24-hinterval with the same compound or with a dosage of S jig ofSEB into the same skin site. Inhibition experiments wereperformed by mixing cSEB dissolved in PBS with SEB at

different molar ratios. The mixture (0.05 ml) was tested forits ability to elicit immediate-type reactions in unsensitizedskin sites of monkeys. The size of any blueing reaction wasmeasured 15 min later.

Additionally, highly sensitized BALB/c mice or strain 13guinea pigs received 0.2 ml (mice) or 1 ml (guinea pigs) of 1%Evans blue intravenously. Immediately after this, 0.05 to 0.1ml of SEB or cSEB serially diluted 10-fold in PBS containing0.03% bovine serum albumin was carefully injectedintradermally into dorsal skin. Each test was done in 5 to 10recipients, and reactions appearing 15 min after injectionwere read and recorded as the dilution of SEB evokingthreshold (5-mm-diameter) skin reactivity.

Experimental enterotoxemia. The monkeys used in thisstudy for skin tests or gastric intubation were anesthetizedby intramuscular injections of ketamine hydrochloride (11mg/kg) and maintained unconscious for 15 min.

Experimental enterotoxemia was induced by administra-tion of S jig of SEB per kg in 5 ml of PBS by gastric tube, andthe clinical course was followed for a period of 24 h.

Pharmacologic inhibition. To investigate potential inhibi-tion of SEB-induced skin reaction or enterotoxemia inunsensitized monkeys, as well as in sensitized BALB/cmice, a number of relevant pharmacologic agents wereadministered before and after SEB challenge. Diphenhydra-mine, an Hi antagonist of histamine, and cimetidine, an H2competitive antagonist of histamine, were injected at a doseof 1 mg/kg intravenously 1 h before intradermal and 15 minprior to and 1, 2, and 3 h after intragastric intubation withSEB.Methysergide (100 mg), a serotonin antagonist, was dis-

solved in 1 ml of 1 N methanesulfonic acid by heating to 60°Cand adjusted to a total volume of 10 ml with distilled watercontaining a final concentration of 5% glucose. This stocksolution was diluted in PBS and injected by gastric tube at adose of 30 jig/kg 15 min prior to exposure of SEB. Addition-ally, the following agents were administered with the dosageand the route of application as indicated: doxantrazole, 3mg/kg; diltiazem, 0.3 mg/kg. These compounds were deliv-ered in a 3-ml volume in PBS by gastric tube either 60 or 15min before SEB challenge.

Alternatively, diltiazem was injected intravenously at adose of 0.3 mg/kg 15 min before intradermal and 15 minprior to and 1, 2, and 3 h after intragastric intubation withSEB.The effect of diphenhydramine, cimetidine, methysergide,

and diltiazem on the response to intradermal SEB in highlysensitized BALB/c mice was tested by intravenous admin-istration of the drugs 15 min prior to challenge at theindicated dosage. In additional experiments the response toSEB was examined after intragastric intubation of doxan-trazole 60 min before challenge.

Biopsy and processing of tissue for ultrastructural examina-tion. Surgically excised skin specimens were obtained 15 minafter skin tests under anesthesia. The skin sample wasprepared for light and electron microscopy as describedpreviously (12). Briefly, semithin sections (0.5 to 1 jim) forlight microscopy and thin sections (50 nm) for electronmicroscopy were cut from tissues embedded by the methodof Spurr (22) that were fixed in Karnovsky fixative (10)diluted 1:4 in 0.1 M cacodylate buffer containing 3.4%sucrose. The osmolarity of the buffer supplemented withsucrose was within the range of 300 to 330 mosmol. Mastcells were identified in semithin sections by staining withmethylene blue-azur 11 (18) for 45 s at 600C. Blocks of tissuewhich contained confirmed mast cells were then trimmed,

INFECT. IMMUN.

on May 30, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

SKIN REACTIONS TO SEB 871

a b

-9.0

-8.0

-7.0

-6.0

5.0

FIG. 1. Conversion of SEB to more acidic species (cSEB) afterexposure to bromoacetic acid (pH 7.0) for 7 days. The samples wereseparated by isoelectric focusing between pH 3 and 10. Lanes: a,SEB; b, cSEB.

thin sectioned, stained with uranyl acetate and lead citrate,and studied with a Philips EM 400 microscope.

Isoelectric focusing. Isoelectric focusing was performed inthin-layer polyacrylamide gels (24) with carrier ampholytes(pH 3 to 10; LKB Instruments, Inc., Bromma, Sweden).Data analysis. Values reported are the mean + the stan-

dard error of the mean.

RESULTS

Carboxymethylation of SEB. Previous studies had estab-lished that carboxymethylation of SEA caused a significantreduction in toxicity without changing the immunologicalspecificity of the molecule (23). Obviously, the reaction ofbromoacetic acid with SEB was similar to that of SEA andspecific for histidine residues under the conditions used. Ofthe six histidine residues present in our SEB preparation (5),4.96 residues were derivatized to 3-monocarboxymethylhis-tidine and 1.05 residues were derivatized to 1,3-carboxy-

methylhistidine. There was no histidine residue refractive tocarboxymethylation. This chemical modification of SEBcaused a conversion from alkaline components to moreacidic species and an increase in microheterogeneity (Fig.1).To compare the emetic efficiency of SEB and cSEB, the

toxin was administered to Cynomolgus monkeys by gastrictube. Whereas the emetic 100% effective dose of SEB wasfound to be less than 5 jig/kg, cSEB consistently failed toevoke emetic responses even when five times the effectivedose of SEB was used.

Skin reactions elicited by SEB and cSEB. In an attempt todelineate the target cells possibly involved in the site of toxinaction, the effectiveness of SEB in eliciting skin reactions inhighly sensitized guinea pigs and BALB/c mice as well as inunsensitized Cynomolgus monkeys was compared.An intense and persistent IgE anti-SEB response was

established in BALB/c mice (PCA titer, 2048) and strain 13guinea pigs (PCA titer, 8192) pretreated with cyclo-phosphamide, followed by two to four sensitizing doses of 2to 10 ,ug of SEB. From the experimental details and theresults given in Table 1, it is apparent that SEB was effectivein eliciting immediate-type skin reactions in sensitizedBALB/c mice and guinea pigs in a range of 10-9 to 10-12 M,and the sensitivity of the assay system correlated stronglywith the serum IgE antibody level.SEB, modified by carboxymethylation (cSEB), resulting

in a complete loss in the toxic activity, however, showedonly a slight reduction in its efficiency to elicit skin reactionsin sensitized animals. In contrast, SEB consistently failed toprovoke skin reactions in unsensitized guinea pigs andBALB/c mice.The most notable feature, however, is the fact that com-

paratively SEB behaved effectively in promoting immediate-type skin reactions at equimolar toxin solutions in unsensi-tized monkeys proved to be free of any anti-SEB antibody.Chemically modified cSEB, however, was incapable ofevoking skin reactions in unsensitized monkeys when testedat doses up to 10-4 M.

Effect of compound 48/80 and cSEB on SEB-induced skinreactions. To establish whether the hypersensitivity-likereaction in monkeys was due to a direct effect of SEB oncutaneous mast cells, the observation was utilized that mastcells initially exposed to a mast cell-degranulating agentbecome unresponsive to a subsequent exposure to a secondchallenging agent (17). Skin sites of monkeys passivelysensitized with rat IgE anti-SEB antibody developed apotent immediate-type reaction following challenge with

TABLE 1. Comparison of skin reactions promoted by SEB or cSEB in highly sensitized strain 13 guinea pigs, BALB/c mice, andunsensitized cynomolgus monkeys"

IgE anti-SEB Skin reaction (>5 mm) promoted by following concn (M):(PCA) titer 10-6 1o-7 1o-, io-9 10-"' 10-" 102 1O-

Strain 13 guinea pigs 8192 SEB + + + + + + +cSEB + + + + + + _ _

BALB/c mice 2048 SEB + + + +cSEB + + + - - - - -

Cynomolgus monkeys None SEB + + + +cSEB' - -

"The results are representative of eight separate experiments.6 SEB in unsensitized guinea pigs and mice consistently failed to provoke skin reactions.No skin reaction, even at 1o-4 M.

VOL. 50, 1985

on May 30, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

872 SCHEUBER ET AL.

Pretreatment

None

None

48/80

None

Buffer

48 /80

48/80

None

None

Cha llenge

SEB

c SEB

SEB

SEB

48/80

SEB

48/80

c SEB+ SEB (1:1 )

c SEBt SEB (5:1 )

Skin reactions(mm diameter)

5 10 15 20 25. . .

FIG. 2. Inhibitory effect of cSEB and compound 48/80 on SEB-induced skin reaction in monkeys. Sensitized skin sites were prepared byintradermal injection of rat IgE anti-SEB and after a sensitization period of 72 h challenged with SEB. Inhibition of reactions in presensitizedand unsensitized skin sites was affected by either an injection of cSEB and then challenging SEB at a molar ratio of 1:1 and 5:1 simultaneouslyor an injection of 25 ,ug of compound 48/80 into presensitized or unsensitized skin sites followed by a challenge with the same activator orwith 5 pLg of SEB at a 24-h interval. Data of the mean + the standard error of the mean of four separate experiments are presented.

SEB (Fig. 2). From the results, it is also apparent thatexposure of sensitized skin sites to an equimolar solution ofcSEB resulted in only a slight reduction of the skin reaction.Immediate-type reactions in sensitized skin sites, however,were completely inhibited by pretreatment with compound48/80 24 h before SEB challenge. Analogously unsensitizedskin sites initially exposed to compound 48/80 became totallyunresponsive to a subsequent exposure to the same activatoras well as to SEB challenge. Further, there is evidence thatskin responses to histamine (1 jig) were not significantlyinfluenced by previous compound 48/80 treatment as com-pared with responses at control sites. Unsensitized skin sitespretreated with buffer, however, were fully capable of asubsequent response to compound 48/80. To determinewhether cSEB, which is incapable of evoking reactions inunsensitized skin sites, and SEB were interacting with thesame putative receptor on the target cell, inhibition experi-ments were performed. It is apparent from the data (Fig. 2)that the immediate skin reaction following challenge withSEB was completely inhibited in the presence of cSEB,when used at a 5-times molar excess. There was no signifi-cant inhibition noted, however, when the concentrations ofSEB and inhibitor were equimolar.

Effect of drugs on skin reaction and emetic response afterSEB challenge. The ability of SEB to elicit an immediate-typeskin reaction in unsensitized monkeys was more closelyexamined in an attempt to establish the release of vasoactiveamines from mast cells by pharmacological treatment.The data (Fig. 3) show clearly that the immediate skin

response in highly sensitized BALB/c mice upon challengewith SEB was totally abrogated by treatment with theserotonin antagonist methysergide at a dose of 30 xug/kg. Incontrast, diphenhydramine and cimetidine, the Hi- and

H2-competitive antagonists of histamine, had no significanteffect on the skin response, when compared with the controlvalue. However, diltiazem, a calcium channel blocker, at adose of 0.3 mg/kg completely prevented SEB-induced skinreactions in sensitized mice. The intragastric administrationof doxantrazole at 3 mg/kg abrogated the skin response aswell.

In unsensitized monkeys, diphenhydramine caused a sub-stantial inhibition of the immediate skin reaction but failed tohave any influence on the emetic response following chal-lenge with SEB by gastric tube.

Pretreatment with cimetidine, a selective blocker of theH2 receptor, however, completely and consistently pre-vented emesis and diarrhea.Methysergide at a dose of 30 kg/kg, giving optimal thera-

peutic levels in human plasma (15), did not have anyinhibitory effect when administered 15 min prior to SEBchallenge by gastric tube. The immediate skin reaction aswell as the emetic response following exposure to SEB,however, was totally abrogated by the calcium channelblocker diltiazem (0.3 mg/kg) and doxantrazole (3 mg/kg)administered by intragastric intubation.

Ultrastructural changes in skin reactions after SEB chal-lenge. The cytoplasm of mast cells in control biopsies arepacked with typical electron-dense granules of uniform sizebut variable shape (Fig. 4A). Rarely, a few scattered gran-ules have lost the bulk of their normally dense contents.Mast cells from the skin reaction challenged with SEB showgranule alteration including limited fusion of granule mem-branes, some loss of electron-dense particle content, andlimited fusion of adjacent granules (Fig. 4B). Several gran-ules are composed of a dense central zone surrounded by aless dense granule substance with lamellar structure (Fig.

Skin sitessensitizedwith IgE

antibody to

SEB

SEB

SEB

None

None

None

None

None

No ne

0 0 0 0

--- * 0.*-0.0 0.*wxe.*.*.* * * * * e-.. e.... - - - -1r,

0-

INFECT. IMMUN.

on May 30, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

SKIN REACTIONS TO SEB 873

Skin reactions (mm diameter)

Sensitized BALB/cmice Unsensitized monkeys

Drug treatment(animals per group)

None (8 )

Diphenhydramine (6 )

Cimetidine

Methysergide

Doxantrazole

Diltiazem

(6)

(6)

(6 )

(6)

5 10 15 20 251 1 1 -1 I~~~~~~~~~~~~~~~~~~~~

(10)

(4)

.rwrrwmwYVWWWYYY.( 4 )

( 4)

0

(6)1(4)

5 10 15, 1

Emetic response

10 /10

4/4

0/4

4 / 4

0/6

0/4

FIG. 3. Effect of pharmacologically active drugs on skin reactions of actively sensitized BALB/c mice and unsensitized monkeys as wellas on the emetic response of the latter following exposure to SEB. BALB/c mice and monkeys were injected intravenously with Evans blueand subsequently challenged with SEB intradermally in a dose range of 1 to 10 ,ug. Enterotoxemia was induced by 5 p.g of SEB per kgadministered by gastric tube. The animals were treated with the drugs, and the reactions were evaluated as described in Materials andMethods. Emetic response is documented by the number of monkeys showing positive reactions versus the number of animals challenged.Data of the mean ± the standard error of the mean of 4 to 10 separate experiments are presented.

4C). The cytoplasm contains numerous mitochondria andvesicles partly associated with the granules. Extrusion ofintact mast cells through the plasma membrane was notobserved but may have been a relatively rare and rapidevent, difficult to capture in static micrographs. This isopposed to extensive degranulation of mast cells exposed bycompound 48/80. There is an extensive loss of granuledensity, and the changed granules were situated in largemembrane-bounded labyrinthine cavities (Fig. 4E). In manyplaces, however, the cavities appeared as closed vacuoles.The nuclei are pycnotic, and the cell membranes cannot beidentified.Mast cells in skin sites challenged with cSEB (Fig. 4D),

however, had no distinctly different morphology from thoseof control skin (Fig. 4A).

DISCUSSIONIn the present study we attempted to elucidate the patho-

logical mechanisms involved in the site of action of SEB.The first approach was based on the finding that car-boxymethylation of enterotoxin A resulted in a loss oftoxicity in the absence of a gross conformational change(23). Analogous treatment of SEB with bromoacetate at pH7.0 for 7 days modified all histidine residues and abrogatedthe toxicity completely. After carboxymethylation of SEB,however, there was a conversion from more alkaline com-ponents to more acidic species with increased micro-heterogeneity of the molecule.

Further, the results clearly show that cSEB is almostequally effective as the native molecule in eliciting immedi-ate skin reactions in guinea pigs and BALB/c mice highlysensitized with SEB-specific IgE antibody. This is also truefor monkeys whose skin sites were prepared by injectionwith IgE anti-SEB antibody. Therefore, it is reasonable toconclude that carboxymethylation of SEB results only in aslight reduction in the ability of the derivative to react withIg1 antibodies. This finding agrees well with existing data,

where carboxymethylated enterotoxin A showed no loss inthe ability to precipitate with antibody to the native toxin(23). Of particular importance was the finding that lowconcentrations (10-' M) of SEB were highly efficient inpromoting immediate-type skin reactions in unsensitizedmonkeys proved to be free of any anti-SEB antibody,whereas cSEB consistently failed to do so even wheninjected at concentrations of up to 10-4 M. These observa-tions lead us to favor the hypothesis that triggering of mastcells by a nonimmunological stimulus depends on the re-quirement for positive charge of the active peptide (2). Thebasis for the loss in skin sensitizing activity by carboxy-methylation of SEB, may involve modified amino acidresidues located within or near a (toxic) site that binds totarget cells.However, SEB was completely ineffective in inducing

comparable reactions in skin sites of unsensitized guineapigs or BALB/c mice. Since the ability of SEB to promoteimmediate skin reactions and enteritic intoxication is obvi-ously restricted to primates, it might be assumed that mo-lecular regions on SEB, so far not clearly defined, show ahigh degree of specificity, in so far as the reaction with aputative receptor on the target cell is concerned. To obtainmore direct evidence in support of a mast cell triggeringprocess in unsensitized monkeys following SEB challenge,one approach was based on mast cell desensitization. Thus,refractory states have been reported against immediate-typereactions, supporting the occurrence of specific desensitiza-tion in addition to mediator depletion (6). The present resultsprovide compelling evidence that treatment of presensitizedor unsensitized skin sites with compound 48/80 completelyabolished the immediate response to a subsequent challengewith SEB, indicating that mediators of cutaneous mast cellsare required. It was ascertained that vascular reactivity wasnot significantly reduced by comparing the response tohistamine at a pretreated site versus a control site. Thisargues against the possibility that a histamine liberator-

-PR-020 0 0 0 0 0 * * 0 0 0 -6-----* * 0 * * " " 0 * * , 0 *-0* .0. .*****041*0 0 0 * * 0 * 0 *.* e 0 * * * * * 0 * 0 # * 010 0 00 0 0 * 0 * * * * * # 0 0 * 0 & 0 oio9% X 0 : & 0

VOL. 50, 1985

C

on May 30, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

874 SCHEUBER ET AL.

A._

*I%.* 'k 1*,Aja t

-A 44'e S1i'1 1:

s*l

-u_4,

-. 4

a,, ,44.,

$ .*.p ** a

,..',, *' *"* .f*' 0, *S

B4.X,::-V4It

K*S,. ~ ;| .^,$

FIG. 4. (A) Electron micrograph of a typical mast cell in thecontrol skin site of an unsensitized monkey injected with PBS 15min prior to sample fixation. The cytoplasm contains typical elec-tron-dense granules, which only rarely have lost dense contents. (B)Mast cell from the skin site challenged with SEB, showing granuleswhich appear swollen and composed of a dense central zonesurrounded by less dense granule substance with lamellar structure(C, arrow). Some affected granules intercommunicate to form ag-gregates. The cytoplasm contains numerous mitochondria (arrowhead) and vesicles (arrows) partly associated with the granules. (D,facing page) Mast cell from the skin site challenged with cSEB doesnot appear to be distinctly different from that of control sites. (E,facing page) Mast cell from the skin of a compound 48/80-challengedrecipient, showing extensive degranulation with characteristic lossof granule density. Numerous granules intercommunicate to formaggregates. The nucleus is pyknotic, and the cell membrane cannotbe identified. Bars in all portions of figure represent 1 pLm.

INFECT. IMMUN.

44'--MOM

.194swup- .1r,i!.. .'--i-.1

VOW'. ...-aNEEL Almk

on May 30, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

SKIN REACTIONS TO SEB 875

D .4i0eloll60

0

I.4

. t t-,.9 . Ah- &

4L~

rY111o s. ieei.% # vv~~~~%

*

.kl ...,I

-~~~i'..q

induced inhibition of vascular response contributes to thedemonstrated abrogation of the immediate skin reactionfollowing exposure to SEB.There is evidence that SE are potent mitogens activating

T-lymphocytes to secrete gamma interferon (13). Since otherlymphocyte mitogens are known to trigger mast cells for aconcentration-dependent, noncytotoxic release of histamine(21), one may speculate that SEB acts similarly.That SEB acts by affecting cutaneous mast cells is further

reflected in the ultrastructural morphology of mast cells atsites of skin reactions. Whereas mast cells in control biop-sies exhibited characteristically electron-dense granules ofuniform size, mast cell granules from SEB-challenged skinsites show alterations including limited fusion of adjacentgranules and some loss of dense contents. Many of thesecells had prominent mitochondria, ribosomes, and manyvesicles, indicating active synthesis. These observationspoint to the possibility that SEB-induced activation of mastcells is accompanied by synthetic and metabolic activity.Mast cells of skin sites challenged with cSEB, however, hadno distinctly different morphology from those of control skinsites. Massive extrusion of granules from the cells, as occursin anaphylactic degranulation of mast cells, was rarelyobserved.We have presented detailed evidence for the failure of

cSEB to elicit immediate skin reactions in unsensitizedmonkeys. It is apparent from the results, however, that

cSEB can in fact specifically inhibit the response to thenative SEB in unsensitized skin sites. The exact mechanismby which cSEB exerts its inhibiting effect has not beendelineated. However, it is reasonable to assume that cSEBcompetitively antagonizes the action of native SEB onbinding sites of mast cells, but in contrast to the nativemolecule, cSEB is proved to be incapable of promotingactivation signals.The current study has further demonstrated that skin

reactions in sensitized BALB/c mice and unsensitized mon-keys following challenge with SEB differed with regard tothe vasoactive amines released. Thus, the serotonin antago-nist methysergide totally blocked the immediate-type reac-tion in sensitized mice, whereas Hi and H2 histaminereceptor antagonists failed to inhibit the skin response sig-nificantly. In this context, however, it should be consideredthat the peripheral vessels of mice are rather insensitive tohistamine and quite sensitive to serotonin (20). These datacould be suggestive of a deficiency of functional histaminereceptors in mice accounting for the inefficiency of histamineantagonists. In addition, pretreatment with the calciumchannel blocker diltiazem totally abrogated skin reactions insensitized BALB/c mice after SEB challenge, most probablyby preventing the transporting mechanisms or channel con-ductance for calcium ions (14). Based on these studies, theuse of pharmacologic agents should also provide a morecomplete assessment of the skin reactivity upon challenge

VOL. 50, 1985

-1.

W-71-111,

on May 30, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

876 SCHEUBER ET AL.

with SEB in unsensitized monkeys. It is apparent from thedata that treatment of monkeys with the H2 antihistaminecimetidine most notably inhibited skin reactions, as well ascompletely protecting against the emetic response, followingSEB challenge. However, despite significant reduction ofthe cutaneous reaction, diphenhydramine, an Hi competi-tive antagonist, had no effect on the emetic response to SEBadministered by gastric tube.Although SEB-induced immediate skin response is consis-

tent with mast cell activation, so far there is no definite proofthat histamine is the principle mediator of these reactions.Rather, the lack of ability of H2 receptor antagonists alone toinhibit the skin response to histamine or the PCA reactionsto antigen in monkeys (8) argues against the predominance ofhistamine, and the possibility must be considered that me-diators other than histamine are important in the pathogen-esis of SEB-induced skin reaction. However, the results ofHutchroft et al. (8) must be interpreted cautiously; it isapparent from our observation that the H2 antagonist cimet-idine at a dose of 1 mg/kg clearly inhibited the wheal-and-flare response to intradermal histamine in baboons (P. H.Scheuber, unpublished data). In contrast, the finding that H2antihistamine completely prevented the emetic response aswell upon SEB challenge was surprising. The mechanismaccounting for this observation is not defined by the presentdata; however, there is some indication that atypical(mucosal) mast cells may also be involved in the intestinalsite of action of SEB (P. H. Scheuber, H. Rang, J. Golecki,and D. K. Hammer, manuscript in preparation). Further, theinhibitory effect of doxantrazole and the calcium channelblocker diltiazem on both the cutaneous reaction and theemetic response suggests that a predominant step in thesequence of biochemical events following SEB challenge isthe role of calcium ions as a second messenger mediating theaction of the ligand (9).

In conclusion, the association between skin test reactivityand emetic response following challenge with SEB is a newapproach and may provide a model for investigation of theoperative mechanism in enterotoxemia.

ACKNOWLEDGMENTS

This work was supported by grants from the FraunhoferGesellschaft, Munich, Federal Republic of Germany.We are grateful to Helga Drache, Brigitte Sextl, and Mathias

Wiesner for expert technical assistance and Rosemary Schneider forpreparing the manuscript.

LITERATURE CITED

1. Bergdoll, M. S. 1970. Enterotoxins, p. 265-326. In S. J. Ajl,T. C. Montie, and S. Kadis (ed.), Microbial toxins. AcademicPress, Inc., New York.

2. Chu, F. S., E. Crary, and M. S. Bergdoll. 1969. Chemicalmodification of amino groups in staphylococcal enterotoxin B.Biochemistry 8:2890-2896.

3. Crestfield, A. M., W. H. Stein, and S. Moore. 1963. Alkylationand identification of the histidine residues at the active site ofribonuclease. J. Biol. Chem. 238:2413-2419.

4. Elwell, M. R., C. T. Liu, R. 0. Spertzel, and W. R. Beisel. 1975.Mechanisms of oral staphylococcal enterotoxin B-inducedemesis in the monkey. Proc. Soc. Exp. Biol. Med. 148:424-427.

5. Ende, I., G. Terplan, B. Kickhofen, and D. K. Hammer. 1983.Chromatofocusing: a new method for purification of staphylo-coccal enterotoxin B and Cl. Appl. Environ. Microbiol.46:1323-1330.

6. Feinberg, S. M., A. R. Feinberg, and F. Lee. 1973. Hypersensi-tivity responses in monkeys. VI. P-K, histamine, 48/80 re-sponses after challenge with heterologous antigen or the otheragents. Clin. Allergy 3:403-409.

7. Harris, C. M., and R. L. Hill. 1969. The carboxymethylation ofhuman metmyoglobin. J. Biol. Chem. 244:2195-2203.

8. Hutchcroft, B. J., E. G. Moore, and R. P. Orange. 1979. Theeffects of HI and H2 receptor antagonism on the response ofmonkey skin to intradermal histamine, reverse-type anaphyl-axis, and passive cutaneous anaphylaxis. J. Allergy Clin. Im-munol. 63:376-382.

9. Ishizaka, T., F. Hirata, K. Ishizaka, and J. Axelrod. 1980.Stimulation of phospholipid methylation, Ca'2 influx and hista-mine release by bridging of IgE receptors on mast cells. Proc.NatI. Acad. Sci. USA 77:1903-1906.

10. Karnovsky, M. 1965. A formaldehyde-glutaraldehyde fixative ofhigh osmolarity for use in electron microscopy. J. Cell. Biol.27:137A-138A.

11. Kent, T. H. 1966. Staphylococcal enterotoxin gastroenteritis inrhesus monkeys. Am. J. Pathol. 48:387-398.

12. Kraeuterkops, S., H. van Loveren, R. W. Rosenstein, W. Ptak,and P. W. Askenase. 1984. Mast cell activation and vascularalterations in immediate hypersensitivity-like reactions inducedby a T cell-derived antigen-binding factor. Lab. Invest.50:421-434.

13. Langford, M. P., G. J. Stanton, and H. NM. Johnson. 1978.Biological effects of staphylococcal enterotoxin A on humanperipheral lymphocytes. Infect. Immur. 22:62-68.

14. Mazurek, N., H. Schindler, T. Schurholz, and I. Pecht. 1984. Thecromolyn binding protein constitutes the Ca'2 channel ofbasophils opening upon immunological stimulus. Proc. Natl.Acad. Sci. USA 81:6841-6845.

15. Meier, J., and E. Schreier. 1976. Human plasma levels of someantimigraine drugs. Headache 16:96-104.

15a.Merkenschlager, M., and W. Wilk. 1979. Schriftenreihe furVersuchstierkunde, vol. VI. Parey-Verlag, Berlin-Hamburg,Federal Republic of Germany.

16. Merrill, T. G., and H. Sprinz. 1968. The effect of staphylococcalenterotoxin on the fine structure of the monkey jejunum. Lab.Invest. 18:114-123.

17. Morrison, D. C., J. F. Roser, C. G. Cochrane, and P. M. Henson.1975. Two distinct mechanisms for the initiation of mast celldegranulation. Int. Arch Allergy Appl. Immunol. 49:172-178.

18. Richardson, K. C., L. Jarett, and E. H. Finke. 1960. Embeddingin epoxy resins for ultrathin sectioning in electron microscopy.Stain Technol. 35:313-323.

19. Schantz, E. J., W. G. Roessler, J. Wagman, L. Spero, D. A.Dunnery, and M. S. Bergdoll. 1965. Purification of staphylococ-cal enterotoxin B. Biochemistry 4:1011-1016.

20. Schwartz, A., P. W. Askenase, and R. K. Gershon. 1977. Theeffect of locally injected vasoactive amines on the elicitation ofdelayed-type hypersensitivity. J. Immunol. 118:159-165.

21. Siraganian, P. A., and R. P. Siraganian. 1974. Basophil activa-tion by concanavalin A: characteristics of the reaction. J.Immunol. 112:2117-2125.

22. Spurr, A. R. 1969. A low viscosity epoxy resin embeddingmedium for electron microscopy. J. Ultrastruct. Res. 26:31-43.

23. Stelma, G. N., and M. S. Bergdoll. 1982. In activation ofstaphylococcal enterotoxin A by chemical modification.Biochem. Biophys. Res. Commun. 105:121-126.

24. Vesterberg, 0. 1973. Isoelectric focusing of proteins in thin layerof polyacrylamide gel. Science Tools 20:22-29.

INFECT. IMMUN.

on May 30, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/