An Investigation of the Ability of Antipsoriatic Drugs to Inhibit Calnlodulin Activity: A Possible Mode of Action of Dithranol (Anthralin)*

William F. G. Tucker, M .B., M .R.C.P., Sheila Mac N eil, Ph.D., Rebecca A. Dawson, H .N.D., S. Tomlinson, M .D ., F.R.C.P. , and Stanley S. Bleehen, M .B., F.R.C.P. Department of Dermatology, Royal Hallamshire Hospital (WFGT , SSB), Sheffield, and Department of Medi cine, Clinica l Sciences Centre. N orthern General Hospital (S MacN , RAD. ST). Sheffield. U . K.

Epidermal calmodulin (CaM) has been reported to be el­ hydrocortisone and crude coal tar showed minimal C aM evated in psoriasis and to decrease following clea rance of inhibitory activity. Dithranol (anthralin), however, whether psoriasis with treatment. We set out to inves ti gate whether freshly prepared or oxidized, produced substantial inhibi­ any of the principle drugs used in the trea tment of psoriasis tion of CaM activity and was demonstrated to be a potent had inherent CaM antagonist activity. Utilizing a CaM­ competitive antagonist of CaM, suggesting another pos­ activated phosphodiesterase we have demonstrated that even sible therapeutic mode of action of dithra nol in psoriasis . at very hi gh concentrations, the systemic drugs etretinate, J Invest D ermatol 87:232-235, 1986 methotrexate, and 8-methoxypsoralen, and the topical agents

evels of the intracellular calcium binding protein cal­ trexate, and 8-methoxypsoralen to inhibit the CaM ac tivation of modulin (CaM) have been shown to be elevated in both a CaM-sensitive enzyme, beef hea rt phosphodies terase. lesionQl [1-5] and nonl esional [2-5] psori ati c epidermis. MATERIALS AND METHODS CaM regulates a w ide ra nge of intracellular processes, several of whi ch have been reported to be altered in Assessment of the Effect of Drugs on Calmodulin Lpsori as is; thus CaM acti vates a number of enzymes including Activity Calmodulin ac tivity was meas ured based on its acti­ cyclic nucleotide phosphodi es terase, phospholipase A2, and or­ va ti on of a CaM-dependent beef heart phosphodiesterase (Boeh­ nithine decarboxylase. whose levels and those of their products ringer Mannheim, London) as described previ ously [11] . Assays are known to be abnormal in pso ri as is [6- 8]. In addition, the conta ined in a fin al reacti on mixture of 400 J.LI, 40 mM Tris-HCI, epidermis is hyperproliferati ve in psori as is, and there is a consid­ pH 7.0 at 37°C, 4 mM 2-mcrca ptoethanol, 5 mM M gClz, erable body of evidence that CaM is essential for the regul ation 3H-labeled cyclic AMP (2 x 105 cpmltube) (Amersham Inter­ of cell di vision [9]. An alterati on in the regul ati on of CaM acti vity nati onal Limited, Bucks, U .K .), 100 J.LM cyclic AMP (Boehringer in the psoria tic epidermis would therefore help to expl ain a num­ Mannheim, London), 25 J.LM C aCh, and pure CaM and drugs as ber of seemingly unrelated abnormalities. required. Pure CaM was prepared from pig brain by the method We have recently shown that a successful res ponse to trea tment of Kakiuchi et al [12] . in psori as is, by a variety of es tablished treatment regimens, is Incubati ons were for 15 min at 37°C. For assess ment of the associated with a decrease in epidermal CaM activity to within potential inhibitory cffects of the drugs, assays contained suffi­ the normal ra nge for control volunteers with little change in those cient CaM (12.5 ng/400 J.Ll incubation) to produce 60- 80% of the individuals where CaM levels were already low [10]. This latter maximum response to CaM . The antipsoriatic drugs tes ted were study led us to consider whether any of the es tablished topical etretinate (Roche Products Limited, Welwyn Garden City, Herts, and systemic methods of trea tment fo r psori as is (for some of U . K.), methotrexate (Lederl e Laboratories Division, Cyanam id which the mode of action in psori asis is not clearl y es tablished) of Grea t Britain Ltd.), 8-methoxypsoralen (supplied as a gift from had any inherent CaM antagoni st ac ti vity. Accordingly we have U pj ohn Ltd., Fleming Way, Crawley, Wes t Sussex), crude coal inves ti gated the ability of 3 topical agents, dithranol, hydrocor­ tar (Thornton & Ross, Linthwaite Laboratories, Huddersfi eld), tisone, and crude coal tar, and 3 systemic drugs , etretin ate, metho- dithranol (Staveley C hemicals Ltd. , Staveley Works, Chester­ fi eld, Derbys hire), and hydrocortisone acetate (The Boots C om­ pany PLC, N ottingham). Two oxidative products of dithranol, Manusc ri pt received October 8, 1985; accepted for publica tion February 1,8-dihydroxyanthraquinone and dithranol dehydrodimer, were 7, 1986. also kindly supplied by Dr. M . Whitefi eld of Dermal Laboratorie T his work has been supported by the Psori as is Association of Great Ltd., Hitchin, Hertfordshire. The CaM antagonist N-(6-amino­ Britain and the Well come T rust. hexyl)-5-chloro-1-naphthalene sulphonamide (W7) (supplied as a * A preli m in ary account of th is work was presented at the Annual gift from Dr. M . Blackburn, Department of C hemistry, Sheffi eld Meetin g of the British Society for In ves ti gative Dermatology. Oxford, University) was also used for comparati ve purposes in this study. Se ptember 1985 (B r J Dermatol 11 3:794-795, 1985). For etretinate and 8-methoxypsoralen it w as necessary to solu­ Reprint req ues ts to: Dr. William F. G. Tucker. T he Unive rsity of Sheffield. Academic Division of Medicin e. Royal Hallamshire Hospital. bilize these drugs first in DMSO, and 2% DMSO was accordingly Sheffield S10 2J F, England. added to all incubations with these drugs. C rude coal tar and Abbreviations: dithranol were solubilized first in ethanol and 2. 5% ethanol wa CaM: calmodulin present throughout all incubations with these drugs. (DMSO at W7: N-(6-aminohexyl)-5-chl oro-1-naphthalene sulfonamide 2% and ethanol at 2. 5% signifi ca ntly lowered both CaM-inde-

0022-202X/86/S03.50 Copyri ght © 1986 by T he Society fo r In vesti gative Dermatology, Inc.

232 VOL. 87. N O.2 AUGUST 1986 C A LMODULIN INHIBITION BY ANTHRALIN 233

pendent and CaM-dependent phosphodiesterase activity. T he re­ 10 sponse of the enzyme to CaM was not itself affected .) Each drug was tested over a wide range of concentrations (all (} 9 J were tested up to 2.5 X lO - g/liter and higher where solubility ~ permitted) in o rder to include those concentrations li kely to be "- 8 '"c: achieved systemically o r locall y in clinical usage. ~ E Incubations were performed in triplica te and each experiment ~ III 7 "- was repeated on a minimum of 3 occasions. "IV -'tI Student's [-test was used to evaluate the statistical significance ., 6 In" ,.." of the effects of the drugs on CaM activity. ~ '0 !., .t;,.. 5 RESULTS :;;" ~ 0 4 ~ '"~ The m ajority of the anti psoriatic drugs studied showed minimal .,a. ..: CaM inhibitory activity even at very hi g h concentrati ons (see Fi g 0 .!! ~ U 3 1) . Thus, methotrexate at 2.5 x 1O - J gil iter, etretinate at 2.5 X ,.. 3 1 '" 10- g/liter, 8-methoxypsoralen at 2.5 x 10- g/liter, hydro­ '0" 2 cortisone at 2.5 X 10- 1 g/Jiter, and crudc coal tar at 2.5 X 10- 2 E .: 0 g/iiter inhibited CaM-dependent activation of beef heart phos­ . pliodies terase by 20% or less. (Although 8-methoxypsoralen had no CaM antagonist properties in its own ri ght, we sho uld point 0 i 00. 1 10 100 100010,000 out that it is normall y given clinica ll y in combination w ith UV A irradiation). SolLibility problems with etretinatc and coal tar pre­ n9 Calmodulin/ Incubation vented testing of thcse subst<1 nces at any hig her concentrations. The deg ree of inhibition achieved at these concentrations varied Figure 2. T he effect of dithr~nol on the abi lit y of CaM to stimulate beef from experiment to experiment and, overall , did not achieve heart Ca M-depend ent phosphodiesterase activit y. In cubations contain ed 2 statistical signifIcance (p > 0.05). Dithrano l, however, signifi­ Ca M alone (closed circle) or CaM plus 2.5 x 10 - g/liter freshly made up dithranol (opei l trit/ lIs le) or Ca M plus 2.5 x 10 - 2 glliter dithranol all owed cantly inhibited CaM activity (Fig 1) with 50% inhibition of to stand in th e sun fo r approxim ately 3 weeks (opell circle). CaM-dependcnt phosphodiesterase activity occurring at 5.5 ± 0.04 x lO - J g/liter (± SEM, n = 3 experiments), equivalent to 24 Ji-M (p < 0.005). In comparison, in these experiments the CaM 2 antagonist W7 produced 50% inhibition (ICsu) at 1 X 10- g/liter, 10,10' -bi-9[1 OH J-anthrone). We cxamined th e CaM anragonist equivalent to 29 Ji-M . pro penies of fres hly prepared dithranol and of oxidized dithrailOl We examined the inhibitory effect of dithrano l (1 ,8-dihydroxy- solutions and found bo th inhibited CaM acti vity to a similar 9-anthrone) in g reater detail as shown in Figs 2 and 3. Dithranol extent, w ith the sole diffcrence being that fres h dithranol also exposed to air and sunlight quite quickly becomes discolored as inhibited basal phosphodies terasc activity which is CaM-indc­ it is oxidized to danthron (1,8-dihydroxyanthraquinone) and pendent (Fig 2). Oxidized dithranol solutions did not affect basal dianthrone, a dithranol dehydrodimer (1,8, 1 ',8 '-tctrahydroxy- enzyme activity even up to 2.5 x 10 - 2 g/liter. . We confirmed thc CaM-antagonist properties of oxidized dith­ ranol solutio ns by examining pure preparations of the 2 most C0111mon oxidation products of dithranol-danthron and the dith­ 100 "'iiii,:,:, ranol dchydrodimer. The former was equipotent w ith dithranol in inhibiting CaM activity; the latter appeared to be sli ghtly less 9 0

80 e "- c 'C 0 6 " 7 0 •,.. '0 ... 2 I 5 ~ 6 0 '0,.. ~ .. :?: e Q, ,.. 4 ~ 50 2 N (j ..: C .. • I*tw.nol 2.1 1 10~· gi L ;:)• E 4 0 ' .2 3 'S U,.. e 'C "-c 0 ... ~.. e 3 0 '0 E 'iii u e 2 0

10

0 oJ. 1/caM (ng CaM/ 400 ,.1 Ineubatlon)-1 0 10'5 10'4 gms/ L Figure 3. Kinetic analysis of the dithranol-indu ced inhibition of acti­ vation of beef heart phosphodiesterase by pure Ca M. Ph osphodiesterase Figure 1. The effects of methotrexate (1) . etretinate (2). 8-methoxy­ acti vity was measured in the presence of 0, 2.5 x 10 - 4 g/liter, and psoralen (3). dithranol (4), crude coa l tar (.5), hydrocortisone (6) , and the 2.5 x 'lO - J g/liter dithranol over a range of Ca M concentrati ons. Data CaM antagonist N-(6-aminohexyl)-5-chloro-l-naphthalene sulfonamide arc plolted as 1 I V vs 1/CaM. where V is nmol cy clic AMP hydrolyzed (W7) (7) on CaM-activated phosphodiesterase. The results shown are of min - I m U phos ph odiesterase and Ca M conccntration is ng of Ca M per single representative ex periments for each drug (repeated on at least 2 400 J.LI in cubation. Each poillf rep resents th e mean of triplica te determi­ other occasions). Data from separate experiments are co mbined by des­ nations. The in crease in slope due to 2.5 X IO - J gi l iter dithranol was ignating the increa se in phosphodiesterase activity due to CaM (Ca M­ used to calculate Ki using th e rel ationship that this in crease represents independcnt enzyme activity of approximately 20% is subtracted) as \ 00%. 1 + IIIIK i . 234 TUCKEH ET AL THE JO URNAL OF INVESTIGATIV E DERMATOLOGY potcnt, although thc extremel y low water solubility of thc dc­ in psoriasis-for example, the spec ies of dithranol and its produc ts hydrodimcr made working with this form extrcmcly difficult. which have been found to be active against glucose-6-phosphate In creasing conccntrations of CaM reduced the inhibition produccd dehydrogenase [251 are different fr o m those which inhibit ON A by dithranol whereas increasing conccntrations of calcium (up to synthesis and induce cytotoxicity [26,27] . 1 mM) (on w hi ch the enzym e is also dependent) did no t affcct It has recently been reported that mitochondrial fun ction is the the inhibitio n. Kineti c analysis of the inhibition using a double­ nlost sensitive target in the cell for dithranol activity (an effect reciprocal Lincweaver-Burk plot (Fig 3) showcd that the V""" mimicked to a lesser extent by the dithranol dimer but not b y (y axis intercept) va lues for thc CaM-stimulated enzyme activity danthron), suggesting that this may be dithranol's mode of action in the absence of drug and in thc presence of 2.5 X 10 - 4 and clinically [28] . However, dithranol has also been reported to in­ 2.5 x 10- 3 g/liter dithranol were the sa me and that the apparent hibit D NA repair [29] (danthron much less so) and very recently K", (s lope = K", /VIlI :o x ) of CaM decreased w ith in creasing con­ a role for CaM in DNA repair has also been proposed [30]. centratio ns of dithrano l. This is consistent w ith dithranol acting Accordin gly, we would argue against premature discarding of as a competitive inhibitor of CaM. The in crease in slope in the any new information about the biologic properties of dithranol double-reciprocal plot produced by 2.5 X 10 - 3 g/liter dithranol and its oxidative products. was used to calculate the inhibito r constant K; [1 3]. using the Several drugs whose clinical modes of action are not full y es­ relati o nship that this in crease in slope is equivalent to 1 + [1] 1K ; tablished have recently been found to possess CaM antagonist [1 01. This gave an inhibitor constant for dithranol of2.5 x 10 - 3 pro perties (for example T amoxifen [31 D, which may contribute g/liter. to their therapeutic actions. Caution must be exercised, however , since despite their CaM antagonist activity, some drugs such as D ISCUSSION propranolol, in practice, exacerbate psoriasis. In the case of pro­ We have previously repo rted that bo th biologica lly active and pranolol, the explanation may be that inhibition of adenylate radioimmunoassayablc levels of CaM are in creased 2- to 3-fold cyclase by this drug, with a consequent fa ll in intracellular cyclic in the psoriatic plaque, and are also sli ghtly in creased in the un­ AMP levels, takes precedence over a weak CaM inhibition [32]. in volved epidermis in psoriasis [2,4]. O ur finding that there is an N o reduction in disease severity is seen among coincidentally intrinsic elevation in the nonles ional epidermis has now been psychotic psoriatic patients treated with large doses of oral Or confirmed by 2 othcr groups, Fairley ct al [3] and Mizumoto et parenteral phenothiazines. Hence, it is possible that, in order to al [5 j. T hc 2- to 3-fold elevation in CaM seen by these workers have an anti psoriatic effect, a sufficient .concentration of a CaM and by us in the psori ati c pl aq ue is of a similar order to that antagonist can be achieved safely only by topical application. reportcd in virally transformed cell s [14]. and to the CaM levels The discovery that dithranol, an es tablished topical treatment we h ~ ve observed in va ri o us neoplas ti c tissues. Cellular hyper­ fo r psori as is, possesses CaM antagonist properties which may at proliferation alo ne, however, does no t seem to explain this ele­ least partiall y explain its therapeutic efficacy, nevertheless pro­ vario n, sin ce we have found no significant increase in CaM levcls vides a stimulus to investigate the use of less toxic CaM antag­ in normal control epidermis induced to proliferate by plastic tape onists in the management of this disease. stripping, or in lymphocytes stim.ul ated to divide by phytohem­ agglutinin [4 1. REFERENCES Whether or not in creased CaM acti vity in psoriatic epidermis 1. Van de Kerkhof PCM, Van Erp PEJ: Calm odulin levels are grossly is onc of the ini tiating biochemical changes in psorias is, we have eleva ted in th e psoriatic lesion. Br J Dermatol 108:217-21 8, 1983 shown that clinical improvement in psoriasis is associated with 2. Tucker WFG, MacNei l S, Bleehen SS, Tomlinson S: Biologically decreased CaM activity in the lesional epidermis [1 0] (CaM ac­ active calm odulin levels arc eleva ted in both involved and unin­ ti vity in the uninvolved epidermis remained slightly elevated). volved epidermis in psori as is. J In ves t Dermatol 82:298- 299, 1984 The possibility that some of the current psoriasis trea tment reg­ 3. Fairl ey JA , Marcelo CL, Hogan VA , Voorhees JJ: In creased cal­ imens might affect CaM acti vity directly seemed worthy of ex­ modulin levels in psoriasis an,d low Ca + + regulated mouse epi­ amin ation, particul arl y as a wide variety of structurally dissimilar dermal keratinocyte c ultu~es . J Inves t DermatoI 84: 195-1 98, 1985 drugs ca n inhibit CaM . 4. MacNei l S, Tucker WFG, Dawson RA, Bleehen SS, Tomlinson S: O ur res ults show that dithranol alo ne, out of the 6 drugs tested, The ca lm odulin content of the epidermis in psoriasis. Clin Sci had any direct CaM antagonist activity as determined by thc 69:681-686, 1985 ability of these drugs to inhibit CaM-dependent beef hea rt phos­ 5. Mi zumoto T, Has him oto Y, Hirokama M, Ohkumo N, lizuka H, phodiestcrase activity. Dithranol has been shown in this study to Ohkawara A: Calm odulin activities are signifi ca ntly increased in be a potent compctitive inhibitor of CaM at concentrations within both uninvolved and involved epidermis in psoriasis. J Invest Der­ the rangc li kely to be encountered in clinical usage. The m ech­ matol 85:450-452, 1985 anisn; of dithranol's therapeutic acti o n in psoriasis has remained 6, Marcelo CL, Duell EA, Sta wiski MA , Ailderson TF, Voorhees JJ: obscurc for many years. Dithrano l has been reported to inhibit Cyclic nucl eotides in psoriatic and normal keratomed epidermis. glu cose-6-phosphate dehydrogenase [15] and several glycolytic J Invest Dermatol 72:20-24, 1979 enzy mes in vitro [1 6], to reduce mitotic activity and DNA syn­ 7. Forster S, lIderton E, Summerl y R, Yardley HF: The level of ph os­ thesis [1 7-201, and to act as an uncoupler of oxidative phos­ ph olipase A2 ac ti vit y is raised in the uninvolved epidermis of pso­ , rias is. Br J Dermatol 108 :1 03-105, 1983 phorylation in mitochrondria [21]. The ability to inhibit CaM would secm to be equall y impo rtant and may also expl ain the 8. Russel l DH, Combest WL, Due ll EA, Stawiski MA, Anderson TF, Voorhees JJ: Glucocorticoid inhibits elevated polyamine synthesis effects of dithranol o n ON A synthesis which is known to be in pso ria sis. J Invest Dermatol 71 :177- 181, 1978 inhibited by CaM antagonists [22]. However, oxidized forms of 9. Vei gl ML, Vanaman TC, Sedwick WD : Calcium and calm odulin dithranol are· repo rted to be less effective therapeutica lly [19,23] in cell growth and transformation. Biochim Biophys Acta 738:21-48, and we find fresh and oxidized forms to be equipotent in inhib­ 1984 iting CaM activity. One is led to the conclusion that either the 10. Tucker WFG, Mac Neil S, Dawson RA , Tomlinson S, Bleehen SS: CaM antagonist acti vi ty o f the drugs is irrelevant to their action Calm odulin levels in psoriasis: the effect of treatment. Acta Der­ in psori asis or that oxidized forms of dithranol should also be mato-Venereologica (Stockh), in press effective clinica ll y. It has been pointed o ut that the inherent in­ 11. Mac Neil S, Walker SW, Senior HJ , Blcchen SS, Tomlin so n S: Effect stability of dithranol which leads in vitro and possibly in vivo to of ex tracellular ca lm odulin and ca lm odulin antagonists on B1 6 the formation of o ther derivatives makes it difficult to determine melanoma cell growth . J Invest Dermatol 83: 15-1 9, 1984 w hich is the active species in the skin [24]. Not all of the biologic 12. Kaki uchi S, Sobue K, Yamaza ki R, KambayashiJ, Sakon M, Kosaki acti ons of dithranol and its oxidative products ca n be accom­ G: Lack of tissue specificity of ca lmodulin: a rapid and hi gh-yield modated in a consistent coherent pi cture of how dithranol acts purifica ti on meth od. FEBS Lett 20 126 (2):203-207, 1981 V O L. 87, N O.2 AUGU ST 1986 ALMODULI N IN HIBIT ION BY ANTHRALIN 235

13. Lchningcr AE: Biochemistry, 2d ed. N ew Yo rk, Worth, 1975, pp psoriasis wi th dithranol (a nthralin). Br J Dermatol 84:282-289, 198-199 197 1 14. Zendeg ui J G, Z ielinski RE, Watterson DM, Van Eldik LJ: Bi osyn­ 24. Shroot 13 , Schae fer H, Juhlin L, Greaves MW: Edi tori al: Anthralin­ thesis of calmodulin in no rmal and virus-transfo rmed chi ck el11- th e chall enge. Br J Dermatol 105:3-5, 1981 bryo fibroblasts. Mol Cell BioI 4:883-889, 1984 25. Cavey D, Caron J C , Shroot 13: Anthralin : chemi ca l in stability and 15 . Raa b WI': Dithranol (a nthralin) versus triacctoxyanthracene. Br J glu cose-6-ph osphatc dehydrogenase inhibitio n. Br J Dermatol Dermatol 95: 193-196, 1976 105 (S uppl 20): 15-19, 198 1 16. Rassner G: Enzy maki vitJtshcmmung in vitro durch Dithranol (Cig­ 26. J acq ues Y, Reichert U : Effects of anthralin and an alogues in growth nolin). Arch Dermatol Forsch 243:47-5 1, 1972 and IJH Ith ymidine in co rporation in human skin fi broblasts. Br J 17. Braun-Falco 0, Burg G, Schoefinin s HH: Uber die Wirking von Dermatol 105 (S upp1 20) :45-58, 198 1 Dithranol (Cignolin) bei Pso riasis vul garis. Cyto-und hi stochcm­ 27. Dykes "J , Roberts GP, Marks H: Anthralin cytotoxicit y. Br J Dcr­ ische Unter-suchungen. Arch Dermatol Fo rsch 24 1: 217-236, 197 1 matoll05 (S uppI 20):43-44, 1981 18. COX AJ , Watson W: Histologica l va ri ati ons in les ions of psori as is. 28. Reichert U , Jacques Y, Grangeret M, Schm.idt R: Antirespiratory Arch Dermatol 106:503-506, 1972 and anti proliferative acti vity of anthrali n in cu ltured human ker­ 19. Fisher LB , Maibach HI : T he effect of anthralin and its deri va ti ves atinocytes. J In vest Dermatol 84: 130-134, 1985 on epidermal cell kinetics. J Inves t Dermatol 64:338-341 , 1975 29. C lark JM, I-b nawalt PC: Inhibition of DNA replication an d repa ir 20. Lidcn S, Michacl sson G: Dithra nol (anthralin) in psori as is. T he effec t in cultured human cell s by anthralin or danthron. Br J Dermatol on DNA-synthes is, g ranular la yer and parakeratosis. Br J Der­ 105 (S uppl 20):49-50, 198 1 matol 91 :447-456, 1974 30. C hafoul eas JG, Bolton WE, Mean s An. : Potentia tion of bl eomycin 21. M o rliere P, Dubertrct L, Mdo T , Sa E, Sa lct C, Fosse M, Sa ntus leth ali ty by anrica lmoduli n drugs: a role fo r ca lmodulin in DNA R: The effect of anthralin (dithranol) on mitochondria. Br J Der­ repa ir. Science 224: 1346- 1348, 1984 matol 11 2:509-515, 1985 31. Lam H-Y: T amoxifen is a calmodulin antagoni st in the acti va tion 22. Hidaka H , Sa saki Y, Tanaka T, Endo T, O hno S, Fujii Y, N aga ta of cAMP phosphodi esterase. Biochell1 Bioph ys Res Commun T : N-(6-Aminohexyl)-5-chl oro-1-naphthalcne sulfonamide, a ca l­ 11 8:27-32, 1984 modulin antagoni st, inhibits cell proli fe rati on. Proc N atl Aca d Sci 32. O rcn berg E K, Wilkinson DI: E ffec t of f3 -adrenergic recepto r block­ USA 78:4354- 4357, 198 1 ade or refr actoriness induced by isoproterenol on growth of kc­ 23. Comaish S, Smith J. Sevill e RH: Factors affec ting the cl ea rance of ratinocytes ;1/ ,,;11"0 . Br J Dermatol 107(S uppl 23): 11 9-124, 1982