ORIGINAL ARTICLE

Origin of the Epidermal Calcium Gradient: Regulationby Barrier Status and Role of Active vs Passive Mechanisms

Peter M. Elias, Sung K. Ahn,n Barbara E. Brown, Debra Crumrine, and Kenneth R. FeingoldDermatology and Medical Services,Veterans Administration Medical Center and Departments of Dermatology and Medicine, University of CaliforniaSchool of Medicine, San Francisco, CA, U.S.A. nDepartment of Dermatology,Yonsei University,Wonju College of Medicine,Wonju, Korea

Mammalian epidermis displays a characteristic calciumgradient, with low calcium levels in the lower, basal,and spinous epidermal layers, whereas calcium levelsincrease progressively towards the outer stratum granu-losum, and declining again in the stratum corneum. Asthe calcium gradient disappears after acute permeabil-ity barrier disruption, and returns after 6 h in parallelwith barrier recovery, barrier function (through restric-tion of transcutaneous water movement) could regulatethe formation of the epidermal calcium gradient. Twotypes of experiments con¢rmed the role of barrierstatus in regulating the calcium gradient: (i) either avapor-permeable membrane (Gore-Texs) or an emolli-ent (Vaselines), applied after acute barrier disruption,immediately restored barrier function, while accelerat-ing the return of the calcium gradient, and (ii) in con-trast, applications of lovastatin, a cholesterol synthesisinhibitor, which delayed barrier recovery and retardedthe return of the calcium gradient. We next askedwhether the calcium gradient is formed/maintained by

passive and/or active mechanisms. Previous studies havedemonstrated that cold exposure (41C) blocks perme-ability barrier recovery after acute disruption. Here,we abrogated the barrier with tape-stripping, and thencompared barrier recovery and restoration of the cal-cium gradient in hairless mice exposed to 41C externaltemperatures, with and without occlusion with Gore-Texs. Although low levels of returned calciumthroughout the epidermis, acutely disrupted, unoc-cluded, cold-exposed sites showed neither barrier re-covery nor reappearance of the calcium gradient at 5 h.In contrast, acutely disrupted, cold-exposed sites, cov-ered with Gore-Texs, likewise displayed little barrierrecovery, but the calcium gradient largely returned by3 h. These results show that (i) barrier status regulatesformation of the calcium gradient, and (ii) passive pro-cesses alone can account for the formation/maintenanceof the calcium gradient. Key words: calcium/cold/di¡eren-tiation/epidermis/transepidermal water loss/ultrastructure. JInvest Dermatol 119:1269 ^1274, 2002

The gradient of calcium (Ca) within the epidermis,with highest levels in the stratum granulosum (SG)and lowest levels in the basal cell layer (Menon et al,1985b), is very important for both permeability bar-rier homeostasis and epidermal di¡erentiation. The

Ca gradient disappears after acute barrier disruption by organicsolvents, detergents, or tape-stripping, reappearing after 6 h inparallel with barrier restitution (Menon et al, 1992a; Mao-Qianget al, 1997; Mauro et al, 1998). Decreased Ca levels in the outerepidermis signal the rapid, lamellar body (LB) secretoryresponse to acute barrier disruption, which is the initial step inthe recovery response. Moreover, secretion and barrier recoveryare inhibited when acutely perturbed sites are exposed to highCa (Lee et al, 1992, 1994; Mao-Qiang et al, 1997). Rapid secretionof preformed LB from outer SG cells is the initial response thatultimately leads to barrier restoration (Menon et al, 1992b; Eliaset al, 1998a). This early response is followed by: (i) stimulation of

lipid synthesis in the underlying epidermis (Menon et al, 1985a;Grubauer et al, 1987); (ii) further production/secretion of new LB(Menon et al, 1992a); and (iii) organization of secreted andprocessed lipids into mature lamellar membrane structures (re-viewed in Elias and Menon, 1991). Together, these metabolicresponses lead to a rapid restoration of barrier function (Grubaueret al, 1989). Yet, the central role of Ca is shown by the obser-vation that LB secretion can be manipulated by changes in Ca,independent of barrier abrogation (Menon et al, 1994; Lee et al,1998).In addition to signaling the LB secretory response, Ca is a well-

known regulator of keratinocyte di¡erentiation (Yuspa et al, 1989;Bikle et al, 1996). Many of the unique proteins required for kerati-nocyte di¡erentiation, including involucrin, small proline-richproteins, loricrin, ¢laggrin, and keratins, are regulated in vitro bychanges in Ca at a transcriptional, and in some cases at a post-tran-scriptional level (e.g., processing of pro¢laggrin to ¢laggrin, andactivation of transglutaminase) (Polakowski and Goldsmith, 1991;Simon, 1994).We showed recently that modulation of Ca levels inthe outer epidermis regulates these di¡erentiation-linked processes,not only in vitro, but also in vivo (Elias et al, 2002). Thus, modula-tions in epidermal Ca regulate events lead to the generation ofboth corneocytes and the extracellular matrix in a coordinate, butinverse, fashion.Despite indirect evidence that supports a link between barrier

status and Ca gradient formation, the basis for the epidermal Ca

Reprint requests to: Peter M. Elias, MD, Dermatology Service (190),Veterans A¡air Medical Center, 4150 Clement Street, San Francisco,CA 94121, U.S.A. Email: eliaspm@itsa.ucsf.eduAbbreviations: CE, corni¢ed envelope; LB, lamellar bodies; OsO4,

osmium tetroxide; SC, stratum corneum; SG, stratum granulosum; TEWL,transepidermal water loss.

Manuscript received March 29, 2002; revised August 8, 2002; acceptedfor publication September 12, 2002

0022-202X/02/$15.00 � Copyrightr 2002 by The Society for Investigative Dermatology, Inc.

1269

gradient is not known. The indirect evidence in support of thisassociation can be summarized as follows. As noted above, thegradient is lost when the barrier is either acutely or chronicallydisrupted, returning to normal as Ca percolates into the epider-mis from deeper skin layers, in parallel with normalization ofbarrier function (Menon et al, 1992b; Mao-Qiang et al, 1997;Mauro et al, 1998). Second, the Ca gradient appears in fetal skinin parallel with the appearance of a competent permeability bar-rier (Elias et al, 1998b). Third, we have shown recently that whenbarrier status is modulated by sustained exposure to either an in-creased or decreased external humidity (Denda et al, 1998), the Cagradient changes accordingly (Elias et al, 2002). Finally, diseasescharacterized by an abnormal barrier, such as experimental essen-tial fatty acid de¢ciency and psoriasis, are accompanied by a lossof the Ca gradient (Menon and Elias, 1991; Menon et al, 1994; Pal-lon et al, 1996).Whereas passive movement of Ca with transcuta-neous water £ux can account, in theory, for changes in the Cagradient, the epidermis possesses a broad array of energy-depen-dent membrane transporters (Mauro et al, 1993; Bale and Toro,2000; Missiaen et al, 2000), which could facilitate formation/maintenance of the Ca gradient.We report here experiments that(i) show directly the link between barrier status and the epider-mal Ca gradient, and (ii) demonstrate that passive (di¡usional)mechanisms alone can account for Ca gradient formation.

MATERIALS AND METHODS

Materials Male hairless mice, aged 8^12 wk old, purchased from CharlesRiver Laboratories (Philadelphia, PA), were used for these studies. Baselinetransepidermal water loss (TEWL) levels were measured prior toexperiments, with an electrolytic water analyzer (Meeco Co.,Warrington,PA) (Menon et al, 1985a). Skin temperatures were measured using a digital,multipurpose thermometer (model BAT-10, Physitemp. Instr., Inc.,Clifton, NJ). Gore-Texs was produced byW.L. Gore and Associates, andpurchased at Nairains Outdoor Fabrics and Sewing Specialists (Berkeley,CA). Petroleum jelly (petrolatum, Vaselines brand) was obtained fromBlue Cross Laboratories (Santa Clara, CA).

Experimental procedures In the ¢rst set of experiments, the barrierwas disrupted by sequential tape-stripping (TEWLZ4 mg per cm2 perh), and the reappearance of the Ca gradient at room temperature wascompared in tape-stripped plus Gore-Texs-wrapped, petrolatum-treated,and untreated sites at 0, 3, and 6 h, as described previously (Grubaueret al, 1989; Halkier-Sorensen et al, 1995; Mao-Qiang et al, 1995). Skintemperatures were measured, and the kinetics of barrier recovery werecompared over occluded and unoccluded sites, at the same time-points.In a second set of experiments, male hairless mice were tape-stripped on

both £anks until TEWLZ4 mg per cm2 per h. Immediately after acutedisruption, 50 ml of lovastatin [2.5% wt/vol in propylene glycol/ethanol(7 : 3 vol)] was applied to a 2.5 cm2 area on one £ank, whereas thecontralateral £ank was treated with vehicle alone (Feingold et al, 1990).TEWL recovery rates then were compared over lovastatin and vehicle-treated sites 3 and 6 h after barrier disruption, and biopsies were obtainedfor Ca localization studies (see below).In the third set of cold-exposure experiments, the following

experimental groups of animals were studied while under chloral hydrateanesthesia (n¼ 6 each): (i) tape-stripped followed by cold exposure (41C)for 5 h, and (ii) tape-stripped, followed by immediate application of aGore-Texs membrane, and cold exposure for 5 h. Cold exposure (areasof 3.14 cm2) was achieved by continuous surface applications of icednormal saline. To control for separate e¡ects of external moisture alone,we treated a separate group of animals with normal saline alone either atroom temperature (23^251C), or at a normal skin temperature (331C). Tocontrol further for the possibility of Ca leakage out of the tissue into thecold saline solution, another set of animals were tape-stripped and allowedto recover at an air temperature cold enough to maintain skin surfacetemperature at between 4 and 101C by keeping the mouse in closeproximity to a liquid nitrogen-¢lled copper tube (n¼ 5).While the skinwas exposed to the ice slurry or cold air, the contralateral side of theanimals were warmed constantly in a water bath (371C), to maintain anormal core temperature. After 5 h, the Gore-Texs wrap was removed,skin temperatures were recorded, and TEWL rates were then measured,followed by skin biopsy for Ca localization (see below). In allexperiments, the ambient air temperature was 23^251C and the relativehumidity was 35^43%.

Statistical comparisons were carried out using the Student’s t-test, andANOVA analysis, where di¡erences between three or more groups werecompared.

Electron microscopy Glutaraldehyde-¢xed specimens were processed,for standard electron microscopy, utilizing osmium tetroxide (OsO4)post¢xation in parallel with samples for Ca localization. Ca was localizedby ion-capture cytochemistry, as described previously (Menon et al, 1985a,1992a). This method has been validated repeatedly for the assessment ofsites of relatively high vs low Ca concentrations by a variety ofquantitative methods, e.g., electron probe microanalysis (Forslind et al,1985) and proton-induced X-ray emission (Forslind et al, 1984, 1985, 1995;Mauro et al, 1998). Brie£y, the samples were minced to o0.1 mm3, andimmersed in a precooled ¢xative (o41C), which contained 2%glutaraldehyde, 2% formaldehyde, 90 mM potassium oxalate, and 1.4%sucrose, pH 7.4, remaining overnight in the ice-cold ¢xative in the dark,the samples then were post¢xed in 1% OsO4, containing 2% potassiumpyroantimonite, at 41C in the dark for 2 h, rinsed in cold distilled water(adjusted to pH 10), followed by graded ethanol washes, and embeddingin an Epon-epoxy resin mixture. Ultrathin (60^80 mm) sections weredoubled stained with uranyl acetate and lead citrate, and examined in aZeiss 10A electron microscope, operated at 60 kV. The status of the Cagradient was evaluated by an uninvolved investigator, who examined,coded, randomized micrographs from each experiment.

RESULTS

The calcium gradient returns rapidly under vapor-permeable dressings In normal murine epidermis, the basaland spinous layers contain few Ca precipitates, with aprogressive increase towards the outer nucleated layers of theepidermis, reaching the highest density within the outer SGlayer (Menon et al, 1985a). As described previously (Menon et al,1992a), barrier disruption by tape-stripping caused an immediatedisappearance of the epidermal Ca gradient, with markeddepletion of both intracellular and extracellular Ca evidentin the outer nucleated layers of the epidermis (Fig 1). At2^3 h after tape-stripping, the extracellular and cytosoliccompartments were still largely depleted of Ca precipitates(Fig 1B), but by 5^6 h, a partial return of Ca precipitates wasevident both in the extracellular and cytosolic compartments ofall epidermal nucleated cell layers (not shown; see Menon et al,1992b). Between 12 and 24 h after tape-stripping, the epidermalCa distribution almost resembled normal (not shown).As Ca could not yet be demonstrated in the epidermis at 3 h,

we utilized this time-point to assess whether the immediate,arti¢cial restoration of permeability barrier function, with eithera vapor-permeable dressing (Gore-Texs) or an emollient(petrolatum), would accelerate the reappearance of the Cagradient. Prior studies have shown that blockade of all watermovement with a vapor-impermeable (Latexs) membranecompletely inhibits the return of Ca (Menon et al, 1992a). Incontrast, when barrier function is restored immediately, withlow rates of transcutaneous water movement (Grubauer et al,1989; Ghadially et al, 1992; Halkier-Sorensen et al, 1995; Mao-Qiang et al, 1995), using either Gore-Texs or petrolatum, the Cagradient largely returns by 3 h (Fig 1C,D vs 1B). These resultsshow that rapid imposition of an arti¢cial barrier, withimmediate reduction of transcutaneous water £ux to low levels,accelerates the reappearance of the epidermal Ca gradient.

Lipid synthesis-induced delay in barrier recovery results in adelayed reappearance of the Ca gradient The previous groupof experiments show that immediate, arti¢cial restoration of thepermeability barrier, with restricted, but ongoing water loss,accelerates return of the Ca gradient. We next asked, instead,whether experimentally induced retardation of barrier recoverywould cause a parallel delay in the reappearance of the Cagradient. As described previously (Feingold et al, 1990), a singleapplication of lovastatin again caused a signi¢cant delay inbarrier recovery in comparison with recovery rates overcontralateral, vehicle-treated sites at 3 and 6 h, respectively (35and 25% delay at 3 and 6 h, respectively). Accordingly and in

1270 ELIAS ETAL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

parallel, there was little or no return of the epidermal Ca gradientat either 3 or 6 h, whereas vehicle-treated sites demonstrated analmost replete Ca gradient by 6 h (Fig 2A vs 2B). Whereas theocclusion studies demonstrate the accelerated reappearance of theCa gradient with arti¢cial imposition of a barrier, these resultsshow, conversely, that an experimentally induced delay in barrierrecovery results in a parallel delay in reappearance of the Cagradient.

Cold exposure also blocks barrier recovery and retardsreturn of the Ca gradient Prior studies showed that exposureto reduced external temperatures delays barrier recovery(Halkier-Sorensen et al, 1995). Therefore, we next employed thismodel to again assess the barrier^Ca link. Hairless male micewere treated on one £ank with tape-stripping and cold vs roomtemperature exposure for 5 h. At room temperature, tape-stripping is followed by a rapid recovery phase, leading to 60^70% recovery of TEWL by 5^6 h (Fig 3), with nearnormalization of barrier function by 24 h (not shown). Incontrast, both cold air-exposed £anks and cold slurry-exposed£anks demonstrated a marked delay in barrier recovery (Fig 3).By 5 h of tape-stripping plus cold exposure, moderate quantitiesof Ca precipitates had returned to the epidermis (Fig 4C vs 4B),but the amounts were less than levels in tape-stripped sites thatwere allowed to recover at room temperature (Fig 4C vs 4A).Yet, whereas substantial Ca returned at 41C, it did not form atypical gradient (Fig 4A vs 4C); i.e., levels of precipitate wereequal in all nucleated cell layers, and some Ca precipitates evenappeared in the stratum corneum (SC) (Fig 4C;Table I). These

results show, in a second, unrelated model, that when barrierrecovery is delayed, Ca reappears, but gradient formation isretarded.

Arti¢cial barrier restoration results in return of a Cagradient in cold-exposed skin We next asked whether thecold-induced delay in reappearance of the Ca gradient re£ects arequirement for active (metabolic) processes in gradientreformation, or whether it re£ects the absence of a barrier. Ifpassive (barrier-related) processes alone su⁄ce, then occlusionwith a vapor-permeable membrane, such as Gore-Texs, shouldfacilitate reappearance of the Ca gradient, even when metabolicprocesses are largely inhibited at 41C. Gore-Texs was applied totape-stripped skin sites for 5 h during exposure to the reducedtemperatures. Although the Gore-Texs-occluded sites displayeda 2^31C higher skin temperature than sites exposed to coldalone, there was a signi¢cant delay in barrier recovery in theunderlying epidermis after removal of the Gore-Texs wrap (10vs 55% recovery for room temperature; Fig 3), indicating thatcold exposure inhibits metabolic responses required for barrierrepair, even under a Gore-Texs membrane. Even at reducedtemperatures, and despite signi¢cant inhibition of barrierrecovery, tape-stripped sites, occluded with Gore-Texs andexposed to cold air for 5 h, demonstrated substantialreformation of the Ca gradient, in comparison with cold-exposed, nonoccluded sites, which display Ca dispersedthroughout the epidermis (Fig 4D vs 4C). Furthermore, in bothof the immersion controls, i.e., in tape-stripped sites exposed to asaline solution at either 23 or 331C for 5 h, barrier recovery was

Figure1. Reappearance of Ca gradient is accelerated when barrierfunction is arti¢cially restored. (A) Normal (Nm) epidermal Ca gradi-ent, with highest density of electron-dense precipitates in SG, and low le-vels in both SC, and subjacent stratum spinosum (SS) and stratum basale(SB). (B) Three hours after tape-stripping (TS; note removal of most SC),little or no Ca is present in epidermis, although Ca precipitates can be seenin the dermis (D). (C,D) When barrier is arti¢cially restored, immediatelyafter tape-stripping by application of either Gore-Texs (G) or petrolatum(Vaseline,V) wrap, substantial Ca precipitates are present in epidermis. Scalebars¼ 2 mm. Pyroantimonate precipitation, followed by OsO4 post¢xation.

Figure 2. Reappearance of Ca gradient is delayed when barrierrecovery is delayed. (A) Six hours after tape-stripping (TS) and a singlelovastatin (Lov) application, little or no Ca precipitates are evident in epi-dermis. (B) In contrast, inTS plus vehicle (Veh)-treated sites, substantial Cahas returned (arrows). Scale bar¼1 mm. Pyroantimonate precipitation, fol-lowed by OsO4 post¢xation.

BARRIER STATUS REGULATES EPIDERMAL CALCIUM GRADIENT 1271VOL. 119, NO. 6 DECEMBER 2002

normal (Fig 3), and the Ca gradient was comparable with that innormal epidermis (Table I). Together, these results demonstratethat, despite cold exposure su⁄cient to inhibit the metabolicresponses required for barrier recovery, a Ca gradient reappearswith imposition of an arti¢cial, vapor-permeable barrier.

DISCUSSION

Several correlative studies support a relationship between barrierstatus and the epidermal Ca gradient, but none clearly demon-strate the interdependence of the two. These indirect studies haveshown that: (i) the Ca gradient disappears after acute barrier ab-rogation, independent of the method of disruption, and returnsin parallel with barrier recovery (Menon et al, 1992a; Mao-Qianget al, 1997; Mauro et al, 1998); (ii) during fetal epidermal develop-ment, a competent barrier and a Ca gradient appear simulta-neously (Elias et al, 1998b); (iii) in organotypic keratinocytecultures, a Ca gradient appears as barrier function improves(Vicanova et al, 1998); and ¢nally (iv) with sustained extremes inenvironmental humidity (Denda et al, 1998), the Ca gradientchanges in parallel with the humidity-induced modulations inbarrier function (Elias et al, 2002). Here, we performedadditional manipulations that show directly the interdependenceof the epidermal Ca gradient and barrier status (Fig 5). The addi-tional correlative approaches comprised two unrelated models inwhich barrier recovery is delayed after acute abrogation; i.e., ap-

plication of the lipid synthesis inhibitor, lovastatin (Feingold et al,1990) and exposure of perturbed sites to cold (Halkier-Sorensenet al, 1995). In both models, not only was barrier recovery delayed,but reappearance of a Ca gradient was retarded, as well.Direct evidence for the relationship between barrier status and

the Ca gradient was shown again, using the acute model, withand without vapor-permeable occlusion.When acutely perturbedsites were allowed to recover normally at room temperature, bar-rier recovery was still in its initial stages at 3 h (E 35% recovery),with 50% recovery by 6 h (Mao-Qiang et al, 1995).Whereas theCa gradient is lost after acute perturbations, it does not begin toreform until 6 h (Menon et al, 1992b; Mauro et al, 1998).We there-fore exploited the 3 h time-point to ascertain whether arti¢cialbarrier restoration with two, unrelated methods of partial occlu-sion, petrolatum or Gore-Texs, would accelerate reappearance of

Figure 3. Barrier recovery is delayed at 41C, with or without occlu-sion. Sets of normal hairless mice were tape-stripped until TEWL4 4 mgper cm2 per h, and either allowed to recover at room temperature, at 41Cby immersion in iced normal saline (NS), exposure to cold air (4^101C), oroccluded with Gore-Texs, followed by application of iced NS. Datashown represent mean7SEM of the percent barrier recovery at 5 h (n¼ 5or 6 for each group).

Figure 4. Arti¢cial restoration of barrier accelerates return of Cagradient, even at 41C. (A) Five hours after tape-stripping (TS) and airexposure at 231C, substantial Ca has returned to epidermis. (B) Immedi-ately after TS and prior to exposure to cold air (Pre: ^41C), the Ca gradientis absent; i.e., levels of precipitate in SG and stratum spinosum (SS) arevery low. A few precipitates are present in the basal layer (SB). In contrast,tape-stripped epidermis, maintained at 4^101C (cold air) reveals substantialreturn of Ca to epidermis, with a few precipitates reaching the SC, but nogradient formation. (C) In contrast, tape-stripped, cold-exposed sites, oc-cluded with Gore-Texs (GT) reveal a substantial return of the Ca gradient;i.e., higher levels in the SG than in the basal/spinous layers. Scale bars: (A) 2mm, (B^D) 1.0 mm. Pyroantimonate precipitation, followed by OsO4 post-¢xation.

Table I. Epidermal Ca distribution 5 h after tape-stripping7occlusion. Epidermal calcium distribution in normal and tape-stripped7Gore-Texs-occluded mice, followed by cold (4^101C), warm or room temperature (air) liquid exposure. Both cold-and warm-exposed sites were exposed to solutions or air alone, as in Materials and Methods. Semiquantitative scale ranges from 0(absence of calcium precipitates) to density þ þ þ þ (maximum). Skin temperatures in cold-exposed animals, occluded with

Gore-Texs were E2^31C higher (12^131C) than in animals with cold-exposed, nonoccluded sites (8^101C)

Warm exposure (331C) Room temperature (231C) Cold exposure (iced normal saline) Cold exposure (cold air)

Tape-stripped Tape-stripped þ Tape-stripped Tape-strippedLayer Tape-stripped Normal skin Tape-stripped Normal skin alone Gore-Texs alone Gore-Texs

SC 0/þ 0 0/þ 0 þ 0/þ 0/þ 0/þSG þ þ /þ þ þ þ þ þ þ /þ þ þ þ þ þ þ þ þ þ þ þ þ /þ þ þ þ þ þ /þ þStratum spinosum þ þ þ þ þ þ þ þ þ þ þStratum basale þ þ þ þ þ þ þ þ þ

1272 ELIAS ETAL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

the Ca gradient. Indeed, we showed that the Ca gradient substan-tially reappears at 3 h, with restoration of barrier function byeither method, whereas no gradient was yet present in air-ex-posed controls. It is important to note, however, that neither bar-rier recovery (Grubauer et al, 1989) nor reappearance of a Cagradient (Menon et al, 1992b) occur under a vapor-impermeable(e.g., Latexs) membrane. Moreover, application of neitherGore-Texs nor petrolatum accelerate endogenous barrier recov-ery; recovery is actually impeded under occlusive dressings in in-verse proportion to their vapor-permeability (Grubauer et al,1989; Proksch et al, 1991). Hence, the accelerated reappearance ofthe Ca gradient is dependent upon the vapor-permeability ofthe occlusive dressing, with restriction of transcutaneous watermovement to low, but measurable rates: vapor-permeable mem-brane (Grubauer et al, 1989; Proksch et al, 1991) and petrolatum(Ghadially et al, 1992; Halkier-Sorensen et al, 1995; Mao-Qianget al, 1995). It is likely that, whereas a competent barrier allowslow rates of transcutaneous water loss, it impedes the outwardmovement of ions. Thus, ongoing water loss, with restriction ofion movement, together explain the epidermal Ca gradient.Finally, we performed experiments that address the question:

Are active metabolic processes required for Ca gradient forma-tion, or do passive mechanisms su⁄ce? The epidermis possessesa wide array of energy-dependent Ca pumps (Mauro et al, 1993;Bale and Toro, 2000; Missiaen et al, 2000), which in theory, couldregulate Ca gradient formation. Our results strongly suggest thatpassive processes alone can account for gradient formation, be-cause the Ca gradient reappears when barrier recovery is arti¢-cially corrected in cold-exposed sites that are occluded withGore-Texs. The success of these experiments requires that skintemperatures remain su⁄ciently depressed under the Gore-Texs

wrap, so that metabolic activity is largely shut down. The smalldegree of barrier recovery that occurred in Gore-Texs-wrapped,cold-exposed vs cold-exposed, nonoccluded sites can beexplained by a slightly higher skin temperature. But the substan-tial inhibition of barrier recovery under the Gore-Texs is consis-tent with a virtual shut-down of metabolic activity due to coldexposure. Thus, the Gore-Texs studies support a largely orpurely passive mechanism for Ca gradient formation; i.e., passivemechanisms alone appear to su⁄ce for formation of the Cagradient.The potential importance of the relationship between barrier

function and the Ca gradient is shown in Fig 5.With improvingbarrier function and further restriction of transcutaneous waterloss, the levels of Ca in the outer epidermis increase, leading toaccelerated expression of di¡erentiation-speci¢c proteins, furtherimproving barrier function. Direct evidence for this positivefeedback loop comes from studies where barrier function im-proves in response to a decline in environmental humidity (Den-da et al, 1998), i.e., the Ca gradient is accentuated, and expressionof di¡erentiation-speci¢c proteins increase at a lower humidity(Elias et al, 2002). Conversely, prolonged exposure to an elevatedhumidity abrogates barrier function (Denda et al, 1998), with lossof the Ca gradient and decreased di¡erentiation (Elias et al, in

press). Finally, TEWL is increased in EFAD (essential fatty acidde¢ciency), as well as in diseases such as psoriasis and atopic der-matitis, in association with loss of the Ca gradient (Menon andElias, 1991; Menon et al, 1994; Pallon et al, 1996), decreased di¡er-entiation (Detmar et al, 1990; Schroeder et al, 1992; Ishida-Yama-moto and Iizuka, 1995; Ishida-Yamamoto et al, 1996; Fujimotoet al, 1997), and an increased risk of disease ampli¢cation (Fig 5)(Elias et al, 1999).In summary, we have provided direct evidence that permeabil-

ity barrier status regulates the formation of the epidermal Ca gra-dient, and that passive di¡usion, rather than active mechanisms,appears to su⁄ce to explain gradient formation. This barrier-dominant model is consistent with situations where TEWL iseither normal or abnormal, providing a unifying hypothesis fora central role for barrier function in regulating epidermal func-tion, as well as the pathogenesis of dermatoses characterized byabnormal barrier function.

Jerelyn Magnusson andJennifer Quevedo capably prepared the manuscript, and ForrestChoy prepared several micrographs for publication. These studies were supported byNIH grants AR19098 and AR39448(PP), and the Medical Research Service, De-partment ofVeterans A¡airs.

REFERENCES

Bale SJ, Toro JR: Genetic basis of Darier-White disease: bad pumps cause bumps.J Cutan Med Surg 4:103^106, 2000

Bikle DD, Ratnam A, MauroT, Harris J, Pillai S: Changes in calcium responsivenessand handling during keratinocyte di¡erentiation. Potential role of the calciumreceptor. J Clin Invest 97:1085^1093, 1996

Denda M, Sato J, MasudaY, et al: Exposure to a dry environment enhances epidermalpermeability barrier function. J Invest Dermatol 111:858^863, 1998

Detmar M, Mayer-da-Silva A, Stadler R, Orfanos CE.: Initial hyperpro-liferation and incomplete terminal di¡erentiation of cultured humankeratinocytes from lesional and uninvolved psoriatic skin. Acta Derm Venereol70:295^299, 1990

Elias PM, Menon GK: Structural and lipid biochemical correlates of the epidermalpermeability barrier. Adv Lipid Res 24:1^26, 1991

Elias PM, Cullander C, Mauro T, Rassner U, Komuves L, Brown BE, Menon GK:The secretory granular cell: the outermost granular cell as a specialized secre-tory cell. J Invest Dermatol Symp Proc 3:87^100, 1998a

Elias PM, Nau P, Hanley K, et al: Formation of the epidermal calcium gradientcoincides with key milestones of barrier ontogenesis in the rodent. J InvestDermatol 110:399^404, 1998b

Elias PM,Wood LC, Feingold KR: Epidermal pathogenesis of in£ammatory derma-toses. AmJ Contact Dermat 10:119^126, 1999

Elias PM, Ahn SK, Brown B, et al: Modulations in the epidermal calcium gradientregulate epidermal di¡erentiation. J Invest Dermatol 119 (5) 2002

Feingold KR, Man MQ, Menon GK, Cho SS, Brown BE, Elias PM: Cholesterolsynthesis is required for cutaneous barrier function in mice. J Clin Invest86:1738^1745, 1990

Forslind B, Roomans GM, Carlsson LE, Malmqvist KG, Akselsson KR: Elementalanalysis on freeze-dried sections of human skin: studies by electron microp-robe and particle induced X-ray emission analysis. Scanning Electron Microsc(Pt):755^759, 1984

Forslind B, Kunst L, Malmqvist KG, Carlsson LE, Roomans GM: Quantitativecorrelative proton and electron microprobe analysis of biological specimens.Histochemistry 82(5):425^427, 1985

Forslind B, Roomans GM, Carlsson LE, Malmqvist KG, Akselsson KR: Elementalanalysis on freeze-dried sections of human skin: studies by electron microbeand particle induced x-ray emission analysis. Scanning Electron Microsc (Pt2):755^759, 1995

FujimotoW, Nakanishi G, Arata J, Jetten AM: Di¡erential expression of human cor-ni¢n alpha and beta in squamous di¡erentiating epithelial tissues and severalskin lesions. J Invest Dermatol 108:200^204, 1997

Ghadially R, Halkier-Sorensen L, Elias PM: E¡ects of petrolatum on stratum cor-neum structure and function. J Am Acad Dermatol 26:387^396, 1992

Grubauer G, Feingold KR, Elias PM: Relationship of epidermal lipogenesis to cuta-neous barrier function. J Lipid Res 28:746^752, 1987

Grubauer G, Elias PM, Feingold KR:Transepidermal water loss. the signal for recov-ery of barrier structure and function. J Lipid Res 30:323^333, 1989

Halkier-Sorensen L, Menon GK, Elias PM, Thestrup-Pedersen K, Feingold KR:Cutaneous barrier function after cold exposure in hairless mice: a model to

Figure 5. Pathophysiologic implications of a barrier-dominated Cagradient.

BARRIER STATUS REGULATES EPIDERMAL CALCIUM GRADIENT 1273VOL. 119, NO. 6 DECEMBER 2002

demonstrate how cold interferes with barrier homeostasis among workers inthe ¢sh-processing industry. Br J Dermatol 132:391^401, 1995

Ishida-Yamamoto A, Iizuka H: Di¡erences in involucrin immunolabeling withincorni¢ed cell envelopes in normal and psoriatic epidermis. J Invest Dermatol104:391^395, 1995

Ishida-Yamamoto A, Eady RA,Watt FM, Roop DR, Hohl D, Iizuka H: Immunoe-lectron microscopic analysis of corni¢ed cell envelope formation in normaland psoriatic epidermis. J Histochem Cytochem 44:167^175, 1996

Lee SH, Elias PM, Proksch E, Menon GK, Mao-Quiang M, Feingold KR: Calciumand potassium are important regulators of barrier homeostasis in murineepidermis. J Clin Invest 89:530^538, 1992

Lee SH, Elias PM, Feingold KR, Mauro T: A role for ions in barrier recovery afteracute perturbation. J Invest Dermatol 102:976^979, 1994

Lee SH, Choi EH, Feingold KR, Jiang S, Ahn SK: Iontophoresis itself on hairlessmouse skin induces the loss of the epidermal calcium gradient without skinbarrier impairment. J Invest Dermatol 111:39^43, 1998

Mao-Qiang M, Brown BE, Wu-Pong S, Feingold KR, Elias PM: Exogenousnonphysiologic vs physiologic lipids. Divergent mechanisms for correction ofpermeability barrier dysfunction. Arch Dermatol 131:809^816, 1995

Mao-Qiang M, MauroT, Bench G,Warren R, Elias PM, Feingold KR: Calcium andpotassium inhibit barrier recovery after disruption, independent of the type ofinsult in hairless mice. Exp Dermatol 6:36^40, 1997

MauroTM, Issero¡ RR, Lasarow R, Pappone PA: Ion channels are linked to di¡er-entiation in keratinocytes. J Membr Biol 132:201^209, 1993

Mauro T, Bench G, Sidderas-Haddad E, Feingold K, Elias P, Cullander C: Acutebarrier perturbation abolishes the Ca2þ and Kþ gradients in murine epider-mis: quantitative measurement using PIXE. J Invest Dermatol 111:1198^1201,1998

Menon GK, Elias PM: Ultrastructural localization of calcium in psoriatic and nor-mal human epidermis. Arch Dermatol 127:57^63, 1991

Menon GK, Feingold KR, Moser AH, Brown BE, Elias PM: De novo sterologenesisin the skin. II: Regulation by cutaneous barrier requirements. J Lipid Res26:418^427, 1985a

Menon GK, Grayson S, Elias PM: Ionic calcium reservoirs in mammalian epidermis:ultrastructural localization by ion-capture cytochemistry. J Invest Dermatol84:508^512, 1985b

Menon GK, Elias PM, Lee SH, Feingold KR: Localization of calcium in murineepidermis following disruption and repair of the permeability barrier. CellTis-sue Res 270:503^512, 1992a

Menon GK, Feingold KR, Elias PM: Lamellar body secretory response to barrierdisruption. J Invest Dermatol 98:279^289, 1992b

Menon GK, Elias PM, Feingold KR: Integrity of the permeability barrier is crucialfor maintenance of the epidermal calcium gradient. Br J Dermatol 130:139^147,1994

Missiaen L, Robberecht W, van den Bosch L, et al: Abnormal intracellular Ca(2þ )homeostasis and disease. Cell Calcium 28:1^21, 2000

Pallon J, Malmqvist KG,Werner-LindeY, Forslind B: PIXE analysis of pathologicalskin with special reference to psoriasis and atopic dry skin. Cell Mol Biol 42:111^118, 1996

Polakowski RR, Goldsmith LA: The cell envelope and transglutaminase. In: Gold-smith LA (ed.). Biochemistry, Physiology, and Molecular Biology of the Skin. Oxford:Oxford University of Press, 1991; pp 168^201

Proksch E, Feingold KR, Man MQ, Elias PM: Barrier function regulates epidermalDNA synthesis. J Clin Invest 87:1668^1673, 1991

SchroederWT,Thacher SM, Stewart-Galetka S, et al: Type I keratinocyte transgluta-minase. expression in human skin and psoriasis. J Invest Dermatol 99:27^34, 1992

Simon M: The epidermal corni¢ed envelope and its precursors. In: Leigh IM, LaneEBWatt FM (eds).The Keratinocyte Handbook. Cambridge: Cambridge Univer-sity of Press, 1994 pp 107^160

Vicanova J, Boelsma E, Mommaas AM, et al: Normalization of epidermal calciumdistribution pro¢le in reconstructed human epidermis is related to improve-ment of terminal di¡erentiation and stratum corneum barrier formation.J Invest Dermatol 111:97^106, 1998

Yuspa SH, Kilkenny AE, Steinert PM, Roop DR: Expression of murine epidermaldi¡erentiation markers is tightly regulated by restricted extracellular calciumconcentrations in vitro. J Cell Biol 109:1207^1217, 1989

1274 ELIAS ETAL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY