Insulin Binding Properties of Normal and Transformed Human Epidermal Cultured Keratinocytes

2 2-2X/R!i/R!i 1-:)2R$ 2. / TilE.)OliHNA I. O F I NV ESTIC:ATIVE Df::RMATOLOt:\', 85::!28- :J:\2, 198!; Copy righr r J98!i by The Wi lliams & Wilkins Co. Vo l. 8.'i, No. -1 Prin ted in U.S.A. Insulin Binding Properties of Normal and Transformed Human Epidermal Cultured Keratinoytes PATRICK VERRANDO, PH.D. AND JEAN-PAUL RTONNE, M.D. Loborot.oire de Reherhe> Dermotolouiques, U.E.l?. de Medeine, Nie, Frane Insulin binding to its reeptors was studied in ultured normal and transformed (A431 line) human epidermal keratinoytes. The speifi binding was a temperaturedependent, saturable proess. Normal keratinoytes possess a mean value of about 8, reeptors per ell. Fifteen hours exposure of the ells to insulin (2 x 1-7 M) lowered their reeptor number (about 65% loss in available sites); these reappeared when the hormone was removed from the ulture medium ("down-regulation" proess). In the A431 epidermoid arinoma ell line, there is a net derease in insulin binding (84% of the initial bound/free hormone ratio in omparison with normal ells) essentially related to a loss in reeptor affinity for insulin. Thus, ultured human keratinoytes whih express insulin reeptors may be a useful tool in understanding skin pathology related to insulin disorders. Insulin binding to its reep tors and s ubsequent ellula r events h a ve been studied in many ell types but not in epiderm a l keratinoytes. Indeed, aurate investigations on the biology of t hese ell s have been m ade reen t ly, sine only in the last few years have they s uessfully been isolated a nd maintained in.vitro in a pure ulture system [1-4]. In suh a system, ins ulin has been found to a llow optimal multipliation of keratinoytes [5] a lthough no de tailed follow-up study has been pe rformed. On t he othe r ha nd, o nly ve ry few of the studies on insulin ation at the ellula r level have been arried out on huma n ell s, with the e xeptio n of dermal fibroblasts or irulating ell s [6-9]. The refo re, in t he light o f these observatio ns, t he a im of this paper was to investigate whether the insulin reeptor ours in human epide rma l keratinoytes, this being t he first neessary ondition fo r a p ossible ation o f this hormone in t he biology o f t he epidermis. 1 n addition, transformed ells (A431 ell line) de rived from an epide rmoid arinoma of the vulva were studied as a n example of the neoplasti state of the same ell type. Manus ri pt reeived,june 2!i, 1984; aepted for publiation April 2::!, 1985. This work was support.ed by the Fondation pour Ia Reherhe Mediale Franais e, the Li b'lle Nationale Fr a n a i se ontre le Cane r, The Assoiation pour Ia Reherhe sur le Ca ner, and Lhe Caisse Mutuelle Regiona le d'assura ne Maladie des CommeranLs el Artisans des Alpes- Maritimes (C. F a ui e r, J. Blleudy). Reprint requests LO: Jean-Paul Ort.onne, M.D., Laboratoire de Reherhes Oermatologiques, U.E.R. de Medeine, Ave nue de Valombrose, 6o:l Nie Cedex, Frane. Abbreviati ons: B: bound B/F: bound/ free EGF: epidermal growth fa tor HC: hydroortiso ne IGF-1: Insulin -like growth falor-j.k, 11 : Dissoiation o nstant. related to the high affinity sites K.,,: Dissoiation o nstant. related to the low aftin itv sites K.,: AfTinit y o nstant fo r t.he empty sites.. Kr: Artinity o nstant for the fill ed sites Chemia/.s MATERIALS AND METHODS [ 115 1] Na (22-25 mci/ ml) was fr om the CEA (Salay, Frane). Porine monoomponent insulin (rystalline form as well as AC TRAPID 4 U/ml used for the hi ghest o nentrations of insulin, i.e., 2.8 X 1-15 M, 8.4 X 1-G M) and gluagon were from Novo (Paris, Frane). Epidermal growt h fator (EGF) was from Collaborative Researh (Sohibo, Paris, Frane ). In sulin B hain is a Boehringe r (Mannheim, B.R..O.) produt. Insulin-like growth fator-] (JGF-l) was ve ry kindly supplied hy Prof. E. Humbel (Biohemishes lnsti tut Der Universitat, Zurih, Sw itze rl and). For oating ulture dishes, o llagen from alf skin was supplied by Sigma (St. Louis, Missouri) whereas ulture produts we re from I<. C. Biologials, In. (Lenexa, Kansas) and fetal alf se rum from Eurobio (Pa ris, Fra ne). All other produts, analyti al grade, were purhased either from Fluka (Buhs, Switze rland) or usual om merial soures. Ce /.1 Culture Human normal epidermal kerat.inoytes were ori ginated from skin of infant foreskin or from adult abdominal surgery. They were ulti vated aordin g to a method lose to that of Liu a nd Karasek [1]. Brief1y, after several was hes wit h saline buffer, the sheets of skin were mined with an eletrial dermatome (Davol In., Cranston, Rhode Island) to remove the max imum of dermis. The sheets bearing epidermis were inubated 45-6 min at 37"C with.25 % (tina! onentration) trypsin in saline buffer; biopsies were transferred into ulture medium ontaining 1% fetal alf serum and the epidermis was sepa rated from the remaining dermis. Then epidermal e lls were gently dissoia ted by the aid of salpels. After the large undissoiated piees of both epidermis and dermis we re disarded by filtration on sterile gauze, filtered e lls were entrifugated and the pellet resuspended in fres h omplete u lture medium. T his was the Dulbeo's modified Eagle's medium o ntaining 1% fetal alf' se rum, l.l x 1-6 M hydroortisone (HC), 1.6 X 1-9 M EGF, and 1 U/ ml peniillin plus 5 1-'g/ml streptomyin. Cells were grown in 4.5-m" wells whih had bee n previously oated with o llagen (referred to as "thin gel" in Liu a nd Karasek's tehnique). The moist inubator atmosphere wa maintained at 37"C and gassed with 7-93 % CO,/ air. Experiments were performed diretly on e ll layers of primary ulture (never passaged), on days 8-11 after seeding a t. a n initial density of' 1'' e lls/em" (allowing a o n flueny state ""7 days after plating). For eah experiment the e ll number per we ll was so red in a Neubauer hemoytometer and usually ranged from 5-6 X 1 5 ells (5.1 5 ells represe nt approximately.68 mg of proteins). Keratinoytes in ulture were ontrolled by indiret immunoflu oresene tehniques. Antisera used were direted against.: total keratin, 67 ko keratin, 55 kd keratin, bullous pemphigoid antige n, and involurin. Two to eve ntually three layers of e lls were usually obse rved. All the e lls were labeled by total antikeratin antiserum; only e lls from the upper layers were labeled by both involurin, an anti-55 kd keratin, whereas bullous pemphigoid serum labeled only t.he "basal'' ells (i.e., in lose ontat with ollagen) (data not shown). Ao rding t.o these riteria, therefore, the behavior of keratinoyt.es in ulture is similar to that in intat epidermis. Furthermore, under these ulture onditions, observations arried out on the ligh t phase ontrast mirosope revealed an homogenous pavement-like aspet. or the e lls (data not. shown ). The A431 line was established by Dr. G. T odaro (National Institute of' Hea lth, Bethesda, Maryland) and was ge nerously supplied by Or. M. Da rmon (C.l.R.O., Sophia Antipolis, Frane). lt was ul tured under the same onditions as normal keratinoytes bm in plasti dishes without. o llagen oating in the medium free of HC and EGF. Culture se ra were analyzed for insulin ontent using a ommerial radioimmunoassay kit. A negligible level was found (<36 pm ) a nd t hus no signifiant interferene with reeptor determination had to be taken into aount. 328

Ot. 1985 INS U LIN RECEPTORS IN HUMAN KERATINOCYTES 329 Iodination of Insulin and Insulin Binding S tudies Ins ulin w a iodin ated to spe ifi ativities of 114-16 Ci/ mmol u s ing a modifiati on of the FreyheL's init ial method [1 ]. B inding studies were performed a foll ows. For experiments at l G"C, ell layers were rinsed 3 times wit h Kr e b - Rin ge r-bi a rb o n a te buffer ph 7.6 (ll8 mm NaCI, 4.7:3 mm KCI, 1.1 8 mm Kl-bPO,, 24.8 mm Na H C O :, 1. 27 mm CaCI", 1.18 mm MgSO, ) ontaining I% (w/v) bovine serum a lbumin. They were then preinuhated in this butler for 3 min allg"c. After re m oval of this solution,.5 ml of fresh buller o ntain ing eit her labeled (64-1 pm : kineti studies) or labeled plus unlabeled hormone (1-1 to 8.4 x JO -" M; equi li brium studies) was added diretly on e ll layers. Assays we re terminated b_v rapid washes (4 X 2 ml for 2 maximum) with ie-old 9 'Yoo (w/ w) NaCl a nd e ll s reovered in.1 M NaOH for 125I ra dioati vit y oun tin g. For :3?"C binding, the same protool was a pplie d, exept preinubation a nd inubation were at 37"C. Insulin_ d eo-r a dation during the b111dmg pen od was estimated at the end of in;ubations by aid preipit ation of an aliquot of inubation medium u s ing 5% (w/ v) t rihloroaeti a id as d e rib e d in Ill]. Analysi. of Binding Data D ata are expressed as speifi binding (in "(, of total) for ompetitioninhibition urves and are also a nalyzed by Satha rd. Curves fi tting in Sath a rd we re arried out by a omputer a lul ation from the ex peri mental va lues of B/ F (bound/ free insulin) and B (bound in sulin) using bot h t he equations of the 2 sites and negative ooperativity models a desribed in [1 2]. The good fit was ensured, in e the o mputer analysis [ 12 ] m ade a urvi linear regression from t he mean ± SD of t he least square for eah ex periment al point. The nonspeifi binding value was obtained from a linear regression between the high o nentrations of free h o rmone in the inuhat ion buffer and int ensity of the radioative s ignal [1 2[. Cell Counting After washes with isotoni phosphate-buttered saline solution, e ll s were r eove red h a n isotoni.1 % (v/v) t.rypsin-.25 % (w/ v) EDTA solut io n and ounted in a Neubauer h e m o ' to m ete r. Dete rmination of!nw ll:n Conent ration al the Derm al-epidermal Juntion Ins u li n onentrat ion at the dermal-epidermal juntion was evalu ated in sution blister lluid of 6 healthy voluntee rs by the sution tehnique of Kiistala and Mustakallio [J :l [. T he blisler lluids were t hen subjeted to a ommerial radioimmunoassay ki t determination for insulin. RESULT S Time -Course of f 2 ''!] lodoinsu.lin Binding Whe n no rma l ke ratinoyte layers we re inubated for diffe r e n t periods of time at 16"C with a t raer a m ount of [ 1 "''1] iodoinsulin, binding of t he li gand reahed a steady state after 4 h and remained onstant at least until 6 h (Fig 1). At t his temperature a nd after t he equilibrium has been reahed, non s peifi binding (measured in t he presene of 8.4 x 1-6 M native po rine insulin) was onst a n t over the inubation time and r a n ged from 23-26% of total binding. Under t hese ondit io n s, [ 12 ''l]iodoinsulin degradation by ell s was very low:.5-.8 % of the init ial amount of hormo ne. Whe n binding was performed a t 37, t he steady state was r eahed more rapidly (1 h only, F ig 1 ); but afterwards, in ontrast at lg, total binding inreased as a funtion of t ime to t h e same extent as n o nspeift b inding, and after 3 h, total a nd n onspeifi b inding dereased (data not s hown). As insulin degr adatio n by ell s represents o nly 5 % of t he hormo ne on entratio n at 3 h, o ne may tentatively suggest t hat t he init ia l inrease in total binding is in pa rt due to la be led insulin whih has been trapped in ell s and whi h was inaessible to the washes (as evidened by t he in rease in no ns peifi b inding). S p eifiity of Insulin Binding To investigate t he speifiity of insulin binding to huma n keratinoytes, ompetit io n studies we re pe rformed using sev- " ::>.n.s 6 5-4.s s 3,.!.. o rl ;; 2 5 a. 1 u '"' "' Q. / " 16 / " / ". I. I / o - o--<'----.,. ;,. _..-,-> {I t/ 1 2 3 4 5 6 1 2 3 Time of inubation (h) FI G L Time-ourse of [ 125 l]monoiodoinsulin binding to ultured normal human keratinoyt.es. In two sepa rate experiments (16 C a nd 3rC), e ll layers (5-7 x 1'' e lls) were inubated with 64 pm ( I 6 C) or 144 pm (37 C) [ 125 l]iodoinsulin in t.he absene (e ) or presene () of 2.8 X 1-r. M unlabeled hormone. At t he indiated time points, reation was stopped and ells we re o lleted and ounted for 12 "! rad ioativity. Eah point was run in dupliate. These experiments a re representative of at least 2 others. Cl 'g1 : -= (/) - 8 7 6 Peptide onentration (-log [M)l FIG 2. Speifiity of [ 12 "l]iodoinsulin binding to normal keratinoyt es. Keratinoytes were allowed to grow on 4.5 m 2 o llagen-oated u lture well s. At day 9 a ft er seeding they were preinubated with 65 pm [ 125 l]iodoin suli n for 4 h at 16"C in the presene of unlabeled insulin (e ), proinsulin ( ), IGF-1 ( ), EGF (), a nd gluagon ("1:\-) at the o nent. rat ions indiated (M in molar). Cells were then reovered a nd ounted for m 1 radioativity. Resul ts a re expressed as the perent of initial speifi binding of the traer a mount of [ 12 ''1]iodoi nsulin (i.e., in the abse ne of unlabeled hormone). Points were made in dupliate. era! peptides against a t raer a m ount (65 pm) of [ 12 "J]iodoinsulin. Fig 2 s hows that pept ides whose sequene is lose to t hat of insulin are able to displae t he hormo ne from its reeptor but t hat t he required on en tratio n of t hese peptides is highe r than t hat of insulin itself. In t hese experiments, to obtain 5% of insulin binding inhibitio n 13-fo ld more (i.e., 79.4 X 1-9 M) a nd 32-fold more (i.e., 199.5 X 1- " M) of proinsulin a nd IG F - 1, r espetively, a re required. EGF a nd gluagon, whi h are pept ides not related to t he insulin-like fa mily m oleules, were una ble to inhibit the binding of [ 12 "I]iodoinsulin although for t he highest onentration (1 J-1M) of gluagon a s light inhibition was seen, ompatible with a minor ontaminatio n of t he om meria l gluagon with insulin. In separa te experiments, B -ha in of insulin a nd redued glutathio ne h ad no ompetit ive effets aga inst insulin (data not s hown). Steady-State Studies of Insulin Binding to Normal /(aa,tinoytes In o rde r to estimate the number of insulin reeptors pe r n o rma l keratinoyte as well as t he binding affinity of t he ligand fo r t hese reeptors, equilibrium studies were arried out at 16 and 37 as desribed in Ma.terial.s a.nd Methods. I

33 VE RRANDO AND ORTONNE The model funtions (see Materials and Methods ) were fi tted on experimental data as B/F vs B (Satha rd). Curvilinear plot with upward onavity was obtained, whih was ompatible with t he three l assial models: (a) two lasses of independent family of binding sites [14]; (b) negative ooperativity among one lass of sites [15 j; or () both (a) and (b) operative. Fig 3 shows the Satha rd representation of binding data in t he ase of hypothesis (a) and fo r experiments at 16 C. Nearly the same graph may be obtained by assuming the existene of ooperativity among one family of sites (data not shown). _So, in ase of hypothesis (a) it was found that human keratmoytes 111 ulture possess at 16C about 8, ± 1, (mean ± SEM) total sites whih may be divided into about 6, ± 22 "high affini ty" sites (K, 11 ==.36 ±.6 nm ) and 74, ± 1, "low affini ty" sites (K t! == 17.25 ± 1.75 nm ). In t he ase of negative ooperativity the alulation from the sa me experimental data gives a total site number of about 87, ± 8,, an affinity o nstant of the empty sites, K. = (2.95 ±.25) X 1 M- 1 an affinity onstant of the fill ed sites, Kr = (.32 ±.1) X lor M- 1 and an interation fator a = Kr/ K equal to.131 ±.1 2. The o mpetition-inhibit ion urves shown in Fig 4, prov ide evidene that there is no signifiant hange in the insulin 6 - s 4[\ : 3.\ 1:t f....,... ::J so 1 1SO B o und insulin (Bl fmol/ 1 6 ells FI G 3. Satha rd pl ot of in sulin binding at l 6C to norma l huma n keratinoytes. Cells l aye r (4.!) x 1' ells) we re p reinubated for 4 h at l 6C with 7 p M l ' ' l]i odo in s ulin a nd vari ous one nt rations (1 - '" M to 8.4 x 1-6 M) o f un labeled hormone. T he bindi ng data we re alulated fo r 1 O" ells as stated in Materials and Methods a nd plotted he re. Po i nt we re made in dupliate. This experimen t is representative of at least 2 othe rs. N,. 8 H u:.... e 6.... 4.E.5,... " 2 'I m ' I 5 13 1 9 8 7 6 lneulln onentrntlon.,_log(m}l FIG 4. Com petit ion-in hi bit ion urves of insulin binding to norma l huma n ke ratinoyts at JG < ) a nd :n (e ). Cells layers (4 -<1.5 x JO'' ells) we re inubated for 4 h at JG"C or 1 h at 37"C wit h approx i mately 7 pm of [' 2 '' 1\ iodoinsulin a nd va ri ous onent rations of unlabeled hormone ind iated ( M, in molar). The speifi bind ing is exp ressed as R/ F ratio, orreted fo r 1" ells. Poin ts were made in dupliate. T his experi me nt is rep resentative of a l least 2 others. Vol. 85, No. 4 binding harateristis at 37 (the urves were almost superimposable, and have the same E Cr.o lose to 1 nm ). "Down Regulation" of Insulin Reeptors T o evaluate the down-regulation proess in ultured normal keratinoytes, onfluent ell layers (day 11 after seeding) were supplied for 15 h wit h omplete ulture medium wi t h or wit hout 2 x 1-7 M insulin. Several washes and 3 preinubations (45 min eah) in assay buffer at 37"C allowed maximum removal of the remaining insulin bound to the ells. This was ont rolled by performing binding studies (at 16. C) of a traer amount of [ 12 r'i]iodoinsulin (7 pm ) on ell layers previously not exposed (ont rol) or exposed for 2 min, 5 min, and 15 h to 2 X 1-7 M insulin and then washed as desribed above. The amount of traer speifially bound by ells exposed 2 and 5 min to insulin represented 16% and 84 %, respetively, of the total speifi amount of t he traer bound to ont rol ells (represent ing in this experiment 1.6% of total insulin onentration). This perentage was only 43 % in the ase of 15-h exposure of ells to insulin. Thus it was o nluded that the total insulin binding did not hange signifia ntly after short exposure to the hormone, so that the effi ieny of the washes was ensured. From a simila r experiment, Sathard plots of the binding data (Fig 5) show that in insulin-treated ells there is a de rease in hormone binding mainly due to a loss of keratinoyte reeptors. The reovery of reeptors after down regulation was then studied as follow. After the ells were exposed for 15 h to 2 X 1-7 M insulin, t he hormone was removed aording to the same proedure as above. Then speifi binding of a traer amount (9 pm ) of [ 12 "l]iodoinsulin was evaluated at 16C from - 15 h after the down-regulation period, both for ont rol ells (whih were not previously inubated with 2 X 1-7 M insulin) and for down -regulated ells. The inset in Fig 5 shows that to time zero the level of insulin speifially bound by down -regulated ells m loor----------------. ;,,\((.. 'ii N o 2 - " /. --r- )( 1.1 ;:,'- /.s 2 3 li-,;- i \\.s \ '\ T itn h.,, " so 1 1SO Bound insulin (B) fmol/ 1 6 ells Ft. 5. Satha rd plot and reovery of insulin binding of down regulated human keratinoytes. In a same expe riment, onl1uent keratinoyte layers (day ll a fter seeding) were inubated for 1.5 h at 37 in t he presene (e ) or absene () of 2 x 1-7 M insulin. Cells were then extensively washed as ment ioned in text, a nd assayed fo r binding for 4 h at l 6C with 8 p M [' 25 l]iodoinsulin a nd 1- 'o to 8.4 x 1-6 M of unlabeled hormone. Satha rd representation o f' binding data orreted for 1 6 ells is shown here. Points were made in dup liate. 1 nset. After t he ells we re "down regulated" by 15 h exposure (3TC) to 2 x 1-7 M in sulin, t he reovery of insulin binding from t ime (2 X 1-7 M insulin removed) was evaluated as desribed in text. From a total t raer amount (9 pm ) of [' 2 ''l]iodoinsulin in inubation mi xture, speifi bind ing (lg.c) to down-regulated ells is evaluated as t he perent of t hose related to ontrol ell s (not previously exposed to 2 X 1-7 M insulin).

Ot. 1985 INSULIN RECEPTORS IN H UMAN KERATINOCYTES 331 represents only 33% of t he level bound by ontrol ell s (1.6% of total amount of insulin in t his experiment). T he reovery of the binding response inreased wit h time, and after 15 h, was greater than 8%, showing then that reeptors reappeared after t he down-regulation proess. Ins ulin Reeptors in the A431 Epidermoid Cell Line As A431 ells were ul tured wit hout EGF and H C, previous studies on normal keratinoytes were performed to determine whether EGF (1.6 nm ) and HC (1.1 pm) have any effet on the in sulin binding properties of t he insulin reeptor. We observed that without EGF and HC morphologi as well as growth harateristis of normal ells were affeted to a great extent after 3 days in ulture, so t hat no valid omparison following studies o uld be done. Therefore, normal e lls were grown (7-8 days) with EGF a nd HC a nd 4 days before binding experi ments t hese two fators were removed from t he omplete ult ure medium. Insulin binding was t hen evaluated at 16"C for 4 onentrations of hormone both for ontrol normal ells (always g rown with EGF a nd H C) and for 4 days EGF / H C depleted normal e lls. Table I shows t hat no hange ourred in insulin binding for t he two o nditions. Thus normal keratinoytes grown with or wi t hout EGF and HC (at the onentra- TABLE I. Compa rison nf insulin binding to /; eratin.oytes as a fun tion Culture o ndi tions With EGF and HC of ell ui lure onditions Jnr--ulin onentration in binding assay (nm ) 86 X 1-:1. 1 1 Withou t EGF and HC 4 days before assay Keratinoytes were g-rown for 1 days with epidermal growth fator 2.2 2.4.'3.3 16.9 272.2 4.7 19.5 228. (EGF) (1.6 nm ) and hyd roortisone (HC) (l.l 11M) and then either not depleted or depleted with EGF and HC for 4 additional days. The results are given in fmo l of insulin spe ifially bound per 1" e lls for 4 h at 16"C. Points were made in t ripliate. 4 no 5 rl )(. 6 '.. " u:- I 3 ' 3. 2 @. \ \.s... s 2 1/) \. \..:!..4,, 1., ' 1 ' :::> ([) I ""' 5 1 Bound Bound inulin (B) fmol./1 zll5 F I G 6. Comparison of' in su lin binding to normal and transformed human kerutinoy t.es. Transformed human keratinoyt.e (A431 line) layers (9.5 X 1" e lls) were as:ayed for insulin binding at 16"C. They were inubated in the prese ne of' 14 4 pm of l 1 "'' 1J iodoinsulin and various onentrations (1-1 M to RA x 1-" M) of unlabeled hormone, as already des ribed for normal e ll s. Sathard plot of' the binding data is given in the n-'<'l. In anothe r experiment, insu lin binding was perfo rmed under the sa me onditions on normal keratinoyte layers as des ribed before. Sathard plots f'or 1" normal (e ) and transformed t ) ells are then litt.ed using the sa me drawin g sale to allow om par ison. Points we re made in dupliate. tions used) behave identially wit h regard to insulin binding. Sine t he ont rol ells, i.e., normal keratinoytes, were not affeted by t hese two fators, t he insulin binding of t hese ontrol ell s may be ompared with t hat of A431 ell s, even if t hey were grown wit h EGF and H C and the other not. For the A431 ells, equilibrium of insulin binding was found to be reahed after 3-4 h at 16"C (data not shown) a nd in F ig 6, Sathard plots at 16"C are given for both normal and transformed keratinoytes. One an see t hat there is a dramati derease (84%) in insulin binding. In t his representative experiment speifi binding of the traer amount of [ 12 ''I]iodoinsulin (144 pm) represents only.7% per 1 6 ells of the total free hormone, against usually about 4-5% per 1 6 ells for norma l keratinoytes. This is t he resul t of a derease in reeptor affinity (Kd 1 = 4.5 nm ; K, 12 = 77 nm) rather than a derease in t heir total number. DISCUSSION The study presented here learly demonstrates that ultured huma n keratinoytes possess speifi reeptors for insulin. Kineti propert ies (Fig 1) of t he speifi insulin binding, espeially wit h regard to temperature, are similar to t hose observed in other insulin target ells suh as hepatoytes [1 2]. For keratinoyt.es, t he speifiity of reeptors is also similar to t hose observed in other ell system [8,12,16-18]. Inidently, peptides wit h similar sequenes to insulin, i. e., IGF-I or proinsulin, ross-reat wit h insulin fo r t he binding to t he reeptor. Nevert heless, t he reeptor affinity fo r t hese peptides is less t han for insulin itself sine a onentration 1- to 3-fold higher is required for half maximal inhibitio n of [ ' '' l]i o do in s ulin binding. In ontrast, peptides suh as EGF or gluagon, not related to insulin-like growth fators or proinsulin, did not show any ompetitive effets wit h insulin. The binding data of insulin analyzed aording to t he Satha rd met hod show a urvilinear fit at 16"C. Reeptor number per keratinoyte, approximately 8,, is lower than in the main insuiin target ells suh as adipoytes (25, sites/ell [19]) but higher than in other ells not t ruly related to t he insulin spetrum of ation like irulating ells (2-38 sites for erythroytes, 7, sites for mo noytes [9]). Competitioninhibition urves performed at eithe r 16aC or 37 C, are si milar (the same maximum apaity binding and the same ECw of 1 nm), so that t he binding apaity seems to be not affeted by temperature. It is interesting to note t hat human keratinoytes behave differently than rat hepatoytes wh ih ex hibit at 37"C a diminution in total sites, related to t he disappearane of t he lower affinity omponent of binding [12]. The affinity onstants of t he keratinoyte reeptor are lose to those of other ell types suh as hepatoytes [1 2] as evidened by Kd, in the range of 1 nm, K <1 2 ""' 5 K <, a nd K, approx im ately 3.1" M - 1, with the Ke equivalent to 6- to 1-fold Kr. As a funtional harateristi of the insulin reeptor in teration, it has been observed that t here is a lose inverse relationship between t he a mbient onentration of insulin and that of its ell surfae reeptor [1 6,2,21 ]. This phenomenon, a lled "down-rerulation " is also observed in ul tured keratinoytes (65% l os of aa ilab l e reeptors) (Fig 5) after 15-h exposure to insulin (2 X 1-7 M). If insulin is removed from t he extraellular medium, t he number of reeptors beome replenished, 5% of t hem are found again after 3 h and t heir initial level is reahed at least 15 h later. The epidermoid arinoma A431 line, as a ellular tool, is able to show t hat epidermal insulin reeptors may be modified in pathologi ellular states. In these ells there is a drasti derease of in sulin binding, probably due to lowering of reeptor affinity rather than a loss of reeptor moleul es (Fig 6). From this point of view, t hese t ra nsformed keratinoytes behave differently with respet to insulin b inding and EGF binding. Indeed, from the works of Fabriant et al [22 ] and O'Keefe et al [23 ], it appears t hat there was a great inrease in A431 EGF

332 VERRANDO AND ORTONNE reeptors when ompared to normal keratinoytes (3-4 X 1'' sites/ell vs 2 X lor, sites/ell). The exat signifiane of this and of the loss in insulin binding for t hese ells remains to be determined. Considering t he results of Baldwin et al [8] whih showed only 7, reeptors per dermal human fibroblast, it ould be onluded t hat the epidermal human keratinoyte, whih presents 1-fold more reeptors, may be impliated in a possible insulin ation in the skin. Although no evidene for this yet exi sts, we have shown using t he sution blister tehnique on huma n subjets t hat in vivo insulin an ross t he dermalepidermal juntion to reah keratinoytes within epidermis. Indeed, in the sution blister 11uids of 6 volunteers, we have fo und a physiologi mean onent ration of 1.3 X w-o M insulin (data not shown). Furthermore, some authors [24,25] have also shown that in t he ase of t he s kin lesions of aanthosis nigrians some patients displayed insulin resistane. In their derma l fibroblasts or monoytes, insulin reeptors were eit he r less numerous or showed a less affinity for the hormone, depending on t he linial type of t he illness. However, further work is needed to determine whether t he presene of t he hormone in skin, partiularly in epidermis, has physiologi importane and whether its binding to reeptors is aompanied by a biologi response. ln t his way, ultured keratinoytes t hat express insulin reeptors represent a useful tool for future investigations. T he aulbors a re indebted Lo Dr. J. Dolais-Kita bgi a nd Dr. M. Fehlma nn for helpful dirussions. REFERENCES l. Liu SC, Ka rasek M: Isolation a nd growth of adul t human epidermal keratinoytes in ell ulture.. ) Invest Dermatol 71:157-162, 1978 2. Eisinge r M,' Solee J,!-Iefton.JM, Darzynkiewiz Z, Chiao.Jy..;, De Harven E: Huma n epide rm al ell ul tures: growth a nd ddlerent iation in the abse ne of dermal omponents or medium supplements. Pro Nat! Aad Si USA 76:531-5344, 1979 :3. Peehl OM, Ha m RG: Clonal growt h of human kerat inoytes with small amou nts of dialysed serum. In Vitro 16:516-525, 198 4. Prunieras M, R egnier M, Woodley DL: Methods for ultivation of kerat inoyles wi t h air-l iq uid in terfae.,j Invest Dennatol 81 (suppl ):28s- :33s, 1983!). T sao MC Walt ha ll B-J, H a m RG: Clonal growth of normal human ep id e r a l kerati noytes in a defin ed medium. J Cell Physiol 11 :219-229, 1982 6. Thomopoulos P, Roths -1: ln ulin reepto rs in normal a nd transformed fibroblasts: relationship to growth a nd transformatio n. Cell 8:417-42:3, 1976. 7. Shilling EE, Rehler MM, Grunfeld C, Rosenberg AM: Pnmary defe t. of insulin reeptors in skin libroblast ul tured from an infant with leprehaunism a nd insulin resistane. Pro Nat\ Aal Si USA 7(1:<>877-5881, 1979 8. Ba ldw in D. Prine M, Tsai P,.Johnson C, Lotan R, Rubenstein AH, Olei'sky.JM: Insulin binding, in terna li zation, a nd reeptor Vol. 85, No. 4 regu lation in ultured huma n fibroblasts. Am,J Physiol 241 (Endorinol Metab 4):E2!Jl- E26, 1981 9. Thomopoulos P: Les reept.eu rs insuliniques des ellules sanguines irula ntes et leur etude en linique humaine. Ann Enlonnol (Paris) 42:381-389, 1981 1. Freyhet P: Insulin reeptors, Reeptors Researh, vol 9, pt _2. Edited by M Bleher. New York, Marel Dekker, 1976, pp 38:>- 428 11. De Meyts P: Insulin and growth hormone reeptors, Methods in Reeptor Researh, vol 9, pt 1. Edited by M Bleher. New York, Marel Dekker, 1976, pp 31-383 12. Fehlmann M, Morin, K itabgi P, Freyhet P: Insulin and gluagon reeptors of iso lated rat hepatoytes: o mpa rison between hor mone binding and amino aid t ransport stimulation. Endori nology 19:253-261, 1981. 13. Kiistala U, Mustakallio KK: Dermoepidermal separation with sution: eletron mirosopi and histohemial study of initial events of blistering on huma n skin.. ) Invest Dermatol 48:466-477, 1967 14. Polett RJ, Standayert ML, Haase BA: Insulin binding to the human lymphoyte reeptor. Evaluation of the negative ooperativi ty model.,j Bioi Chem 252:5828-5834, 1977 15. De Meyts P, Biano AR, Roth J: Site-site in terations among insulin reeptors. Charaterization of t he negative ooperativit.y. J Bioi Chem 251:1877-1888, 1976 16. Mott DM, Howard BV, Bennett PH: Stoihiometri binding and regulation of insulin reeptors on human diploid fibroblast using physiologi insulin levels. J Biol Chem 254:8762-8767, 1979 17. King GL, K ahn CR, Rehler MM, Nissley SP: Diret demonstra t ion of sepa rate reeptors for growth and metaboli ativities of insulin and multipliation-st imulation ativity (an insulin like growth fator) using a ntibodies to the insulin reepto r. J Cli n Invest 66:1 3-14, 198 18. Zapf J, Shoenle E, Froesh ER: Insulin-like growth fator I and fl: some biologial ations a nd reeptor binding harateristis of two purified onstituents of nonsuppressible insulin-like ativity of human serum. Eur,J Biohem 87:285-296, 1978 19. Olefsky,JM: Dereased insulin binding to adipoytes and irulat ing monoytes from obese subjets. J Clin Invest 57:1165-1169, 1976 2. Gavin,JR lll, Roth J, Neville OM, De Meyts P, Buell N: Insulin dependent regulation of insulin reeptor onent rations: a diret demonstration in e ll ulture. Pro Nat\ Aad Si USA 71 :84-88, 1984 21. BarRS, Gorden P, Roth.), Kahn CR, De Meyts P: Flutuations in the affinity and onentration of insulin reeptors on i rulating monoytes of obese patients. Effets of starvation, refeeding and dieting.,j Clin Invest 58:112:3-1133, 1976 22. Fabriant RN, De La rgo JE, Todaro GJ: Nerve growth fator reeptors on huma n mela noma ells in ulture. Pro Nat] Aad Si USA 74:565-569, 1977 23. O'Keefe E, Battin T, Payne R: Epidermal growth fator reeptor in human epidermal ell s: diret demonstration in ul tured ells. J In vest Dermatol 78:482-487, 1982 24. Ba r RS, Muggeo M, Roth.), l< ahn R, H av rankova J, Imperato MeG in ley J: Insulin resistane, aanthosis nigri ans, and normal insulin reeptors in a young woman: ev idene for a postreeptor defet. J Clin Endorinol Metab 47:62-625, 1978 25. Ka hn R, Podskalny J: Demonstration of a primary (? geneti) defet in insulin reeptors in fibroblast from a patient with the syndrome of insulin resistane a nd aanthosis nigri ans type A. J Clin Endorinol Metab 5:11.39-1141, 198