Evidence for Histamine in the Urticating Hairs of Hylesia Moths*

Scott M . Dinehart, M.D., Joseph L. Jorizzo, M .D., Nicholas A. Soter, M .D., Nopadon Noppakun, M.D. , William R . Voss, D.V.M.,t James A . Hokanson, Ph.D., and Edgar B. Smith, M.D. Department of Dermatology (S MD, JLJ , NN. EBS). the Animal Care Center (WRV). and the Biostatisti ca l Component. Cancer Center UAH), The University of Texas Medical Branch, Galveston, Texas; and Department of Dermatology (NAS). New York University School of Medicine, New York, New York. U.S.A.

An urticarial dermatosis after contact with the urtica tin g les ia extract (HE) into the backs of cynomolgus monkeys. hairs of the adult female Hyles ia moth may occur by several This whealin g res ponse was suppressed by pretreatment of mech a nisms including the intradermal inj ection of inflam the animals with diphenhydraminc hydrochloride, but not matory mcdiators through the urticatin g hairs. Extracts by pretrea tment with indometha ci n. Histologic examina were prepared from whole moths, urticatin g hairs, and tions showed a perivascul ar lymphocytic infiltrate around other moth parts. Each of these extracts was subjected to dilated capillaries without evidence of mas t cell degranu.:. a radioenzy mc assay for histamine. Histamine was present \ation in H E-injected sites but not in controls. These find in extracts made from whole moths and from urtica ting in gs provide evidence that histamine may be the mediator hairs. Extracts madc from other moth parts contained no res ponsible for the urticarial lesions seen after contact with histamin e. C utaneous wheals occurred after intradermal Hyles ia moths. ) I/1l1 est Dernwtol 88:691- 693, 1987 injections of histamine and various concentrations of Hy-

ermatitis occurring after contact with urtica ting hairs portunity to exa mine crew members from an oil tanker that was of th e adult female Hyles ia moth has been well doc infested by swarms of Hylesia moths at the port of Caripito, umented [1-3]. The eruption is pruritic, occurs on Venezuela 11] . The presence of a predominantly urtica ri al eruption exposed surfaces, and is commonl y urtica ri al. The in most of the crew suggested the possibl e role of hista mine as a condition is generall y unresponsive to various top toxic substance contained in the urtica ting hairs. Dica l and systemic medica ti ons; however, spontaneous resolution Meth ods for identifyin g and charac terizin g mediators of cu occu rs in about one week, provided contact with moths or moth taneous inflammation have been described [4-6]. The present pa rts is discontinued. study examines the actions of extracts of Hylesia mo ths in monkey Several mechanisms for the producti on of the dermatitis have skin over a wide dose range in the presence and abse nce of an been proposed and include intracutaneous injection of toxic sub H I-receptor antagonist, diphenhydramine hydrochloride, and the stan ce(s) through the holl ow urticating hairs, direct irritant effect cyclooxygenase inhibitor, indomethacin , to indirectly assess the of the h airs, and hypersensitivity to insect antigen(s) [1 ,31. More presence of mediator ac tivity. Using a radioenzy me assay, we than one mechanism may be involved. examined extracts of whole moths, w ings and legs, and urticatin g Dern1atitis produced by Hylesia moths is frequent in Latin hairs for th e presence of hista mine. Am.erica but rare in the United States. We recently had the op- MATERIALS AND METHODS Subjects Indirect mediator studies employed healthy ad ult cy Manuscript received August 19. 1986; accepted for publication No nomolgus monkeys (Macaca JasciCil laris). ve mber 25, 1986. * Read before the Southern Region of The Society fo r In vesti gative Intracutaneous Injections After sedation w ith i. m. ketamine Dennatology. Inc., New O rlea ns, Louisiana, February 7. 1986 and at the Annual Meetin g of The Society for In ves ti gative Dermatology, In c.. hydrochl o ride (Ketalar®) at a dose of 15 mg/ kg, the backs of 4 Washington, D.C., May 3. 1986. monkeys were shaved, washed with isopropyl alcohol, and then tPresent address: University of Texas System Cancer Center. Science inj ected intradermall y usin g a 30-ga uge needle w ith 50 J.Ll of di Pa rk, Veterinary Resources Division, Route 2, Box 151-B I. Bastrop, lutions of Hylesia extract (HE), phosphate-buffered saline, pH 7.6 Texas 78602. (PBS), hi stamine, and buffer consisting of 1.0 M sodium chl o ride Reprint requests to: Scott Dinehart, M. D .. Department of Dermatol and 25% ethylene glycol (EG-N aCl) . Inj ections w ere ca rried out ogy, T h e University of Texas Med ica l Branch, Ga lveston. Texas 77550. at 2.5-min intervals in a double-blind random fashion according Abbreviations: to our published method [5 ,6]. Two perpendicular diameters of ANOVA: anal ysis o f variance th e wheal, including the longest axis and ignorin g small pseu EG-N aCI: buffer consisting of 1. 0 M sodium chl oride and 25% dopodia, were measured between 5 and 7 min, w hen the wheal ethylene glycol HE: /-/),l csia extract was fully developed. The size of the w heal was given as th.e mean K-W: Kruskal-Walli s of the maximal diameters. Wheals measuring less than 7.5 mm PBS: phosphate-buffe red sa line, pH = 7.6 median diameter were not considered a response, as this was the

691 692 DINEHART ET AL THE JOURNAL OF INVEST IGATIVE DERMATOLOGY

mean diameter of the injection bleb. The ambient temperature Statistical Analysis Analysis of variance (ANOVA) followed was 22-24°C. Evidence for delayed hypersensitivity at the sites by contrast analysis were the primary statistical methods em of cutaneous injection was assessed at 48 and 96 h. ployed. The Statistical Package for the Social Sciences was used Pharmacologic Treatment Baseline dose response curves of for all computations [1 0]. w hea l response were obtain ed after injections as described above RESULTS on day 1. All 4 monkeys received 3.0 mg/kg of th e H ,-receptor blocker diphenh ydramine hydrochloride i.m. the morning of the Local Effects in Monkey Skin Results of HE monkey injec experiment on day 3. Four hours later, another 1.5 mg/kg i.m. tion experiments plotted on log 2 paper are summarized in Fig dose was given. T he experiment began 2 h after the second dose 1. Wheals produced by vehicle control injections approximated and lasted 45 min to 1 h. All 4 monkeys received 4.0 mg/kg of 8 mm, which was about the size of an injection bleb. Histamin e indo methacin orally the morning of the experiment on day 5. injection w hea ls approached 12 mm. As the concentration of HE Four hours later, another 4.0 mg/kg oral dose was given. The was increased, wheal size generall y increased. After pretreatment experiment began 2 h after the second dose and lasted 45 min to with diphenhydramine hydrochloride, wheal formation was gen 1 h. Dosages of diphenhydramine hydrochloride and indometh erally suppressed to the levels of control injections. For day 1 (no acin were given in accordance with guidelines for nonhuman pharmacologic pretreatment) a one-way ANOVA was performed primates. to compare PBS and EG-NaCI to the 4 dilutions of HE. Because these res ults were stati sticall y significant (F = 8.13, P < 0.01), Solutions Appropriate dilutions of injected substances were PBS and EG-NaCI were then compared with all the dilutions made up freshly and sterili zed by an acryli c copolymer membrane grouped together. As both the Bartlett-Box F and Cochran's C filter 0.20 J.Lm (Gelman Acrodisc@ no. 4192, Gelman Scientific, were not signifi cant, the pooled variance estimate was used. The Ann Arbor, Michi gan). In each experiment, response to HE di contrast analysis was statistically signifi cant (t = 3.37, P < 0.01). luted 1:1, 1 :2, 1 :4, and 1:16 with PBS was assessed. The response These two analyses suggest that PBS and EG-NaCI produce val to 0.05 mg histamine/50 J.LI was a positive control. The res ponses ues different than vehicle containing any amount of HE. Phos to 50 J.LI of PBS and EG-NaCI were negative controls. All so phate-buffered saline and EG-NaCI were compared w ith hista lutions were measured at pH 7.5-7.6. mine of day 1. These results were statistically signifi ca nt (t = The HE had been prepared as follows: the bodies of each of 48.8, P < 0.01). These results were also consistent with the more 10 adult moths, the separated hairs, and the wings of 1 moth conservative Kruskal-Wallis (K-W) procedure. For PBS and EG were weighed and minced in plastic tubes containing 1.0 ml of N aCI vs dilutions of HE, the K- W chi-square was 13.0 (p < 0.01) EG-NaCI [7]. After boiling for 10 min, the tubes were sedimented and for comparison of PBS with histamine the K-W chi-square at 400 g for 10 min, and the supernatant fractions were frozen at was 5.4 (p < 0.02) . -20°C. The initial two-way ANOV A across' days and dilutions indi Assessment of.Histamine Duplicate aliquots were assessed for cated both significant main (F = 17.5, P < 0.01) and interraction th eir histamine content with a radioenzyme assay [8,9]. One ali (F = 8.4, P < 0.01) effects (Fig 1). Comparison of day 1 (no quot was in cubated with an excess amount of diamine oxidase pharmacologic pretreatment) and day 3 (pretreatment with di (Sigma C hemical Co., St. Louis, Missouri) for 30 min at 37"C phenhydramine) measurements indicated significant differences after which its histamine content was determined. between no pharmacologic pretrea tment and diphenhydramine Histologic Examinations After sedation w ith i.m. ketamine pretrea tment at the HE dilution of 1:2 (t = 5.11, P < 0.016) and hydrochloride, skin biopsy specimens (3-mm diameter) were taken for histamine (t = 6.5, P < 0.01) (Fig 1) . without additional local anesthes ia from the center of induced Pretreatment with indomethacin (day 5) did not have significant w hea ls between 10 and 20 min after injection of HE. Control effect on wheal response to any dose of HE or to histamine. No biopsy specimens were taken from the sites of injection of PBS, evidence of delayed hypersensitivity at 48 and 96 h was seen at EG-NaCI, and from normal, uninjected skin from the monkey's any injection site. back. Each specimen was examined for mast cell s and mast cell Histamine Content of Moths The histamine content (Table deg ranulation using li ght and electron microscopy by published I) ranged from 7-38 ng per whole organism [mean 19 ± 6 (SD) methods [5,6]. and ± 1.9 (SEM)] and from 0.10-0.50 ng/mg wt [mean 0.24 ± 0.01 (SO) and ± 0.003 (SEM)]. Preincubation with diamine ox idase destroyed 100% of the histamine. Urticating hairs separately 14.0

12.0 Table I. Histamine Content in Extracts of Hylesia Moths E 10.0 ..s Histamine Conten t Wei ght ~ 8.0 Moth Bodies (mg) ng/Wholc Organism ng/mg wt ::::i: ~ P<.Ol P<.03 P<.Ol 1 78 29 0.37 C 8.0 Z 2 78 23 0.29 3" 86 38 0.50 ~ 4.0 a - OFF DIPHENHYDRAMINE ::::i: 4 102 22 0.22 a - ON DIPHENHYDRAMINE 5 97 13 0.13 2.0 EG-NeCI- tOM NeCI AND 26% ETHYlENE GlYCOL CONmOL 6 76 7 0.1 0 PBNS - PHOSPHATE BUFFERED NORMAL SAUNE CONmOL 7 76 35 0.46 0 .0 8 92 8 0.10 1:1 1:2 1:4 1:16 9 92 8 0.10 DILUTION 10 91 8 0.10 Mean 19 0.24 Figure 1. Dose response curves for the maximum diameter of wheals Standard deviation 6.0 0.01 produ ced on the backs of cy nomolgus monkeys by intradermal inj ection Standard error of the mea n 1. 0 0.003 of va ri ous co ncentrations of Hylesia extract are shown without phar Hair 30 9 0.30 macologic pretreatment (circles) and after pretreatment with diphenhy Wings 6 0 0 dramine hydrochl oride (sq ll ares). Histamine is a positive control. Phos ph ate-buffered sa line and EG-NaCI are negative controls. "After diamine oxidase: ng/wholc organism = 0; ng/mg wt = O. VOL. 88, NO.6 JUNE 1987 HI STAMINE IN MOTHS 693

an alyzed contained 0.3 ng/mg wt. The wings did not contain seen in specimens inj ected w ith HE are consistent with the in histaITline. Radioenzyme assessment for histamine was 70.6 ng/ml tracutaneous effects of the inj ection of histam ine into normal skin. 111 the 1:1 HE. One interesting finding is the presence of endothelial cell degen eration, which may represent a toxic effect of the extract. The Histologic Examinations discovery of endothelial cell degeneration was focal in nature and Light Microscopy: Control biopsy specimens from skin injected not associated w ith clinical purpura. with PBS, EG-NaCl, and normal skin from the backs of the We have provided evidence that histamine participates in the monkeys showed no abnormal histologic findings. Biopsy spec production of Hyles ia dermatitis. Other m ediators or mechanisms Imens taken from extract injection sites showed a perivascular are no t excluded. Persistence of the eruption clinically, despite lymphocytic infiltrate around sli ghtly dilated ca pillari es in the ri gorous antihistamine therapy, suggests that o ther mechanisms Up p er papillary dermis. Endothelial cell swelling and mild peri of pathogenesis m ay be involved . It is known that the urticating vascular edema were also present. No eosinophils were seen. hairs can penetrate and rem ain in the skin for long periods of time GieITlsa-stai ned sections did not show an increase number of mast and possibly cause irritation by m echanical means [1] . Urticating cells or evidence of m ast cell degranulation. hairs were filtered from our extracts prior to injection. Electron Microscopy: Ultrastructural examinati on of sections from HE-injected sites showed a perivascular ly mphocyti c infiltrate We thallk Regilla Sco tt o-Chillllici for expert tee/mica/ assista ll ce alld Pat MorellO around blood vessels in the papillary derrnis. Degeneration of the fo ,· typillg th e "Jalll/script. end o theli al cells of the ca pillaries was observed along with wid ening o f gaps between endothelial cells. So m e endothelial cells appeared to separate from the basem ent m embrane. All m as t cells REFERENCES examined appeared normal and were not degranulated. 1. Dinehart SM, Archer ME, WolfJE, McGaveran MH, Reitz C, Smith EB : Caripito itch: dermatitis from contact with H y/esia moths. ] DISCUSSION Am Acad Dermatol 13:743-747, 1985 Skin diseases resulting from contact with members of the order 2. Hill WR, Rubenstein AD. Kovacs ]R: Dermatitis resulting from Lepidoptera (butterfli es and moths) have been known since ancient contact with moths (Genu s Hylesia) . ]AMA 138:737-740, 1948 times [11] . M any studies have been performed to isolate and 3. Zaias N, loannides G. Taplin D: Dermatitis from contact with moths char act erize the agents responsible for lepidopteric dermatitis (Genus Hy/esia). ]AMA 207:525-527, 1969 [7, 12-15]. Some of these experiments have shown histamine to 4. Greaves MW, Black KA: Identification of mediators of inflamma be present in the hairs or poison glands of certain species [7,12] . tion, in In ves ti ga tive Techniques in Dermatology. Edited by R Marks. St. Louis, Blackwell Scientific Publications, 1979, pp 94-103 DUCOITlbs et al suspected the presence of histamine in the hollow urtica ting hairs of the adult female Hylesia moth beca use of the 5. ]orizzo ]L, Coutts AA, Eady RA], Greaves MW: Vascular responses of human skin to inj ection of substance P and mechanism of action. app earance of wheals and a demonstration of increased vascular Eur] Pharmacol 87:67-76, 1983 permeability after injection of an HE into the skin of guinea pigs 6. Coutts AA, ]ori zzo JL, Eady RA], Greaves MW, Burnstock G: [16]. Adenosine triphosphate-evoked vascular changes in human skin: Our studies confirm those of Ducombs et al that injection of mechanism of action. Eur] Pharmacol 76:39 1-401, 1981 extracts made from the urticating hairs of Hyles ia moths causes 7. Shama SK, Etkind PH, Odell TM, Canada AT, Finn AM , Soter cuta n eous wheal formation. Additionally, we show that the size NA: Gypsy-moth-ca terpillar dermatitis. N Engl ] Med 306: of this response is dose-dependent. The negative responses from 1300-1301, 1982 extracting substances (PBS and EG-NaCl) eliminate the possi 8. Beaven MA , Jacobsen S, Horakova Z: Modifi cation of the enzymati c bility that these solutions cause the response seen to HE. isotopic assay of hista mine and its appli ca tion to measurement of The H,-antihistamine, diphenhydramine hydrochloride, given histamine in tissue, serum and urine. Clin Chim Acta 37:91 -1 03, in doses adequate to suppress histamine, significantly reduces 1972 wh eal formation to HE at a concentration of 1 :2. This finding 9. Shaff AB , Beaven MA: Increased sensitivity of the enzymatic iso supports our h ypothesis that the wheaIing reacti ons to HE are at topic assay of histamin e: meas urement of histamine in pl as ma and least in part m ediated by histamine. The almost complete suppres serum. Anal Biochem 94:425-430, 1979 sion of all cutaneous response to HE at higher concentrations by 10. Nie NH, Hull CH, Jenkin s ] G, Steinbrenner K, Bent DH: SPSS: diph enh ydramine suggests that histamine is the predominant me Statistica l Pa ckage for the Social Sciences . New York. McG raw diator present in our extract. Although wheal response to 1:1 HE Hill , 1977 is la r ge, we do not consider its suppression by diphenhydramine It. Aber R, DeMeJfi T, Gill T, Healey B, McCa rthy MA , Oswell N , statistically significant, probably because of a large SD; the small Ruhig W, Speziale H, Witte EJ: Rash illness associated with gypsy number of subjects tested might have caused this high SD. The moth caterpill ars-Pennsy lvan ia. Morb Mortal Wkly Rep 31: 169-170, 1982 fi nding that indomethacin produces no significant effect o n wheal form a tion suggests that products of arachidonic acid metabolism 12. Valle ]R, Picarelli ZP, Prado JL: Hista mine content and pharma cological properties of crude extracts from setae of urticatin g cat do not playa major role in the production of Hylesia dermatitis. erpill ars . Arch ln t Pharmacol 98:324- 327, 1954 Radioenzym e assay has been shown to be a sensitive and specific 13. Go ldman L, Sawyer F, Levine A, Goldman], Goldman S, Sp in anger method for the identification of histamine [8,9]. This assay showed J: Investigative studies of skin irritations from ca terpillars. J In vest histaITline in extracts m ade from the urticating hairs of Hylesia Dermatol 34:67-79, 1960 moths (Table I) . Histamine was found in much greater quantities 14. Ziprokowski L, Rolant F: Study of the toxin from the poison hai rs froITl extracts made from the urticating hairs than extracts made of Thallllletopoea Wi/killsolli caterpillars. J In ves t Dermatol 46: from other moth parts, suggesting that the majority of the his 439-445, 1966 tamine found in Hylesia moths is present in the urticating hairs. 15 . Ziprokowski L, Rolant F: Study of the toxin from poison hai rs of The presence of histamine in the extract m ade from urticating T/wllllletopoea Wi/ki"solli ca terpillars. J Invest Dermatol 58:274-277, hairs suggests that it m ay play an important role in the production 1972 of Hylesia dermatitis. 16. Ducombs G, Lam y M, Michel M, Pradinaud R, ] amet P, Vincendeau Light and electron microscopic studies of biopsy specimens did P, Maleville] , Texier L: La papillonite de Guyane Francaise. Etude not show mast cell degranulation. The histologic findings of vaso clinique et epidemioIogique. Ann Dermatol Venereol 11 0:809-816, dilation, widening of endothelial gaps, and perivascular edema 1983