ORIGINAL PAPER

Matthias Fischer Æ Thomas William Æ Peter Helmbold

Johannes Wohlrab Æ Wolfgang Ch. Marsch

Expression of epidermal N -methyl-D-aspartate receptors (NMDAR1)depends on formation of the granular layer—analysis in diseaseswith parakeratotic cornification

Received: 9 January 2004 / Revised: 12 June 2004 / Accepted: 22 July 2004 / Published online: 25 August 2004� Springer-Verlag 2004

Abstract Ionotrope glutamate receptors of theN-methyl-D-aspartate (NMDA) receptor type areexpressed on keratinocytes and influence the intracel-lular calcium concentration. The importance ofNMDA receptors in pathophysiological processes inthe skin is, however, still unclear. Epidermal distri-bution patterns of NMDA receptors were investigatedin dermatoses with parakeratotic cornification (psori-asis vulgaris and verrucae vulgares) and compared tothe expression of filaggrin. The expression of NMDAreceptors (R1 component) in paraffin-embedded nor-mal epidermis (n=22), psoriasis vulgaris (n=21) andverrucae vulgares (n=23) was examined and evaluatedby means of digital image analysis. For quantitativecharacterization of the distribution patterns, a quo-tient was formed of the expression in the stratumgranulosum and stratum basale (‘‘NMDA ratio’’). Thedistribution of NMDAR1 was compared to theimmunohistochemical expression of filaggrin. Addi-tionally the expression of filaggrin was investigated inHaCaT cells after treatment with the NMDA receptorantagonist MK-801. NMDA receptors were demon-strated in the epidermis of all preparations. In healthyskin, the highest receptor density was found in thestratum granulosum. This distribution pattern wasbasically also present in the dermatoses examined.Thus, the occurrence of parakeratosis in psoriasisvulgaris, but not in verrucae vulgares, was character-ized by a significant reduction in the NMDA ratio(reduced expression of NMDAR1 in the upper epi-dermis). The immunohistochemical distribution of

filaggrin was similar to that of NMDAR1. In HaCaTcells MK-801 suppressed the expression of filaggrin.NMDA receptors are expressed in human epidermisunder physiological conditions especially in the stra-tum granulosum. Their reduced expression withinparakeratotic epidermis in psoriasis vulgaris may beevidence of impaired intracellular calcium influx in thisdisease.

Introduction

Under physiological conditions, the intraepidermal cal-cium concentration, especially the sharp increase in theintracellular calcium concentration in the stratumgranulosum [11, 12], is an important regulatory factor inepidermal differentiation and cornification. The calciumrequired for this may be taken both from intracellularstores [2, 15] and from the extracellular space [11, 12].The mechanism involved in calcium influx from theextracellular space into the keratinocytes has been onlyinadequately examined thus far. Among the possibleaccess pathways are the ion channel proteins such as thefamily of ionotropic glutamate receptors, which aresubdivided according to their selective agonists inAMPA (amino-3-hydroxy-5-methyl-4-isoxazolpropio-nate), kainate and N-methyl-D-aspartate (NMDA)receptors [20]. The first evidence of regulation of theintracellular calcium concentration of keratinocytes byinotropic receptors of the NMDA receptor type wasreported recently [6, 7]. In culture, native keratinocytes[6] and HaCaT cells [6, 7] show an elevated intracellularcalcium concentration after stimulation by the selectiveagonist NMDA. Moreover, an elevated NMDA recep-tor density has been found in the upper segments of thevital epidermis of hairless mice [7]. We have already beenable to immunohistochemically demonstrate a similardistribution pattern with a maximum NMDA receptordensity in the stratum granulosum in human skin [6].This receptor distribution correlates well with the

M. Fischer (&) Æ T. William Æ P. Helmbold Æ J. WohlrabW. Ch. MarschDepartment of Dermatology and Venerology,Martin-Luther-University Halle-Wittenberg,Ernst-Kromayer-Str. 5, 06097 Halle (Saale),GermanyE-mail: matthias.fischer@medizin.uni-halle.deTel.: +49-345-5572685Fax: +49-345-5573384

Arch Dermatol Res (2004) 296: 157–162DOI 10.1007/s00403-004-0505-0

intraepidermal calcium gradient in normal epidermisdescribed by Menon and Elias [13]. Regulation of thekeratinocytic calcium concentration by NMDA recep-tors appears especially important for cornifying epithe-lia, since in epithelia that do not cornify (such as oralmucosa) there is an equal distribution of NMDAreceptors over the entire epithelium [6].

In contrast to physiological terminal differentiation,the possible importance of NMDA receptors in impairedcornification has not yet been examined. The histologi-cal phenomenon of parakeratosis, which is defined as theoccurrence of cell nuclei within the stratum corneum[19], was chosen as a model for the possible pathologicalrole of epidermal NMDA receptors. Parakeratosis waschosen for two reasons. First, the development ofparakeratosis is associated, at least in psoriasis vulgaris,with a pathologically altered epidermal calcium gradient[13]. Moreover, it has been demonstrated that cultivatedkeratinocytes of psoriasis patients also have a defect intransmembrane calcium influx [11]. Therefore it seems tobe evident that (psoriatic) parakeratosis might be due toan altered intraepidermal calcium gradient, based on areduced transmembrane calcium influx from the extra-cellular space. Second, parakeratotic cornification issometimes but not always accompanied by the presenceof a stratum granulosum. While parakeratosis in psori-asis vulgaris is associated with a loss of the stratumgranulosum, this layer is more pronounced in verrucaevulgares [16, 21]. Possible differences in the expressionpattern of epidermal NMDA receptors could thereforebe helpful in understanding the role of the granular layerin the keratinization impairment in these dermatoses.

Materials and methods

Immunohistological examinations

Paraffin-embedded skin sections of patients with psori-asis vulgaris (plaque-type, n=21) and verrucae vulgares(n=23) were examined retrospectively. The diagnoseswere based on clinical parameters and histologicalexamination of preparations stained with hematoxylin-eosin (HE). The results were compared with those frompreparations of healthy skin (n=22) obtained duringreduction surgery and follow-up resections of variousskin tumors [6]. All preparations were fixed immediatelyafter collection in 4% buffered formalin and embeddedin paraffin.

Since the NMDA receptor consists of a constant R1unit (NR1), combined with four variable NR2 subunits,an antibody was selected which includes the NR1 com-ponent and thus all NMDA receptors. For this, amonoclonal mouse antibody was selected (Clone 54.1;BD PharMingen; Heidelberg, Germany). This antibodyis directed against a fusion protein (amino acids 660–811of NR1), which represents the intracellular loop betweenregions III and IV of NR1 [18]. Negative controls, inwhich the primary antibody was replaced with an isotype

IgG2a antibody (BD PharMingen; Heidelberg, Ger-many; Cat. No. 550339), were treated in the same way.

The NMDA receptors in the paraffin slides wereimaged using protocols adapted from those used inneuropathology [1, 4]. Paraffin sections (5 lm) weredeparaffinized and antigen demasked (10 m M citratebuffer, microwave 600 W 3·5 min/trypsin 0.001%,37�C, 15 min). To reduce nonspecific backgroundstaining, the preparations were then incubated in 1%milk powder for 15 min at room temperature. This wasfollowed by further incubation with the above mouseanti-NMDAR1-Ak (1:500, 60 min, 37�C) and thenovernight at room temperature. For further treatmentthe ABC method (Vectastain Elite ABC Kit; VectorLaboratories, Burlingame, Calif.) was used with AEC(3-amino-9-ethyl carbazole) as chromogen (AEC sub-strate-chromogen system; DAKO Corporation, Glost-rup, Denmark) according to the manufacturer’sinstructions. The preparations were then counterstainedwith hemalum and embedded.

The sections were evaluated by digital image analysisusing SigmaScan Pro version 4.01 software. For thisthree 10·10 lm areas in each epidermis layer weremeasured in representative segments of the preparations,and the results summed to provide a mean value. Thearea (pixels) of red–brown colored segments was definedas the measure of NMDAR1-positive reaction. Thecoefficient of variation of the method (84 sections, threesingle measurements each) was 32.1% for the individualmeasurements and 25% for the triple measurements.According to these data, 100 pixels was defined as thelowest threshold value.

The individual epidermal layers (stratum basale,stratum spinosum, stratum granulosum, stratum corne-um) were identified visually. If the stratum granulosumcould not be clearly identified in the presence of para-keratosis, the immediately subcorneal cell layer was usedfor the assessment. To compare the patient groups, thevalues from the stratum granulosum were expressed as aratio in relation to those from the stratum basale (theNMDA ratio). The extent of parakeratosis was recordedsemiquantitatively. Parakeratotic cornification of>50% in relation to the epidermal surface was rated assevere and of <50% of the epidermal surface as mild.The results were analyzed using the Mann-Whitney andWilcoxon nonparametric tests using SPSS 11.0.1 soft-ware (SPSS, Chicago, Ill.).

In addition, for comparative imaging, the differenti-ation marker filaggrin was visualized immunohisto-chemically in serial slides using a monoclonal mouseantibody (filaggrin (Ab-1), clone FLG01; OncogeneResearch Products, San Diego, Calif.). Pretreatment andstaining was carried out following the manufacturer’sinstructions. To demask the antigen, the preparationswere heated in 10 m M citrate buffer (pH 6.0) for20 min. After cooling, the slides were incubated with theprimary antibody at a dilution of 1:200. Further treat-ment was by the ABC method with AEC as chromogen,as described above.

158

Cell culture

HaCaT keratinocytes were cultivated in Dulbecco’smodified Eagle’s medium (DMEM) supplemented with1% antibiotic-antimycotic (10,000 U/ml penicillin G,10 mg/ml streptomycin, 0.25 pg/ml amphotericin B) and10% fetal calf serum (FCS), and maintained at 37�C in ahumidified atmosphere containing 5% CO2. The cellswere seeded at a density of 0.8·106 in culture dishes andcultivated in calcium-rich DMEM (1.8 m M Ca). Every2 days, when the medium was changed, the NMDAreceptors were selectively blocked with MK-801. Thefinal concentration of MK-801 selected was 50 lMsince, using the crystal violet method [8] which enablesestimation of the cytotoxic potential of active agents,MK-801 can be anticipated to be cytotoxic in HaCaTcells at concentrations of 200 lM or more [6]. The cellswere examined immunohistochemically on the 14th dayof culture for expression of filaggrin as described above.

Supplementary tests of the influence of MK-801 onfilaggrin expression in native human keratinocytes couldnot be performed. In the crystal violet method, MK-801in serum-free medium was cytotoxic after 48 h at con-centrations of 50 lM or more. Due to the lower sensi-tivity to MK 801 of native keratinocytes compared toHaCaT cells [6, 14], no effect would be anticipated atconcentrations lower than 50 lM.

Results

NMDA receptors (NMDAR1) were demonstrated in allpreparations. Healthy epidermis was NMDAR1-posi-tive in all vital layers, with a particularly high expressionin the stratum granulosum (Fig. 1). Patients with pso-riasis vulgaris and verrucae vulgares (all preparations)showed a comparable distribution pattern of NMDAreceptors, with particular emphasis in the granular layer.

The NMDA ratio (NMDA expression stratum granu-losum/NMDA expression stratum basale) of healthyepidermis was 6.3±4.0 (mean±SD). There were nosignificant differences between normal epidermis, psori-asis vulgaris (5.3±3.2, all preparations) and verrucaevulgares (5.0±2.8, all preparations).

In contrast, there was a considerable difference inNMDA expression between highly parakeratotic skinareas and skin with mild or no parakeratosis. In psori-asis vulgaris, preparations showing severe parakeratosis(n=13) had significantly lower levels of NMDA receptorexpression in the upper epidermis than those showingmild or no parakeratosis (n=8; Figs. 2 and 3). Thelower NMDAR1 expression was accompanied by areduction or loss of the stratum granulosum. In contrastto psoriasis vulgaris, in verrucae vulgares, preparationsshowing severe parakeratosis (n=6; NMDA ratio7.1±2.8) had higher levels of NMDA receptor expres-sion than those showing mild parakeratosis (n=16;NMDA ratio 4.1±2.3; P=0.065). Furthermore, a stra-

Fig. 1 a Immunohistochemistry of NMDAR1 in healthy epider-mis: maximal NMDAR1 expression in the stratum granulosum.b Negative control

Fig. 2 NMDA ratio in patients with psoriasis vulgaris in relationto the extent of parakeratosis. The ratio was significantly different(P<0.01) between preparations showing parakeratosis of <50%of the surface (n=8) and those showing parakeratosis of >50% ofthe surface (n=13)

159

tum granulosum could be identified in all preparationsof verrucae vulgares (Fig. 4).

Comparative imaging of filaggrin and NMDAR1showed a similar distribution pattern in the stratumgranulosum. In particular, in analogy to NMDAR1, areduction or loss of filaggrin was found in the presenceof parakeratosis (Fig. 5).

Compared to controls, treatment with MK-801 led tovisibly reduced keratinization in HaCaT cells. More-over, on the 14th day, there was expression of filaggrinin untreated controls, while the MK-801-treated cellsshowed no positive staining (Fig. 6).

Discussion

In the present study, a reduced expression of NMDAreceptors in the upper epidermis was demonstrated in

psoriasis vulgaris in the presence of parakeratosis. Sinceevidence has recently been found that the NMDAreceptor is a factor in calcium homeostasis of keratino-cytes [6, 7], the altered immunohistochemical expressionfound in the studies reported here is the first evidence ofa potential pathogenetic role of epidermal NMDAreceptors. Clearly, their expression depends especially onthe formation of the stratum granulosum. Thus, prepa-rations of verrucae vulgares, with both parakeratosisand a pronounced granular layer [16] showed NMDAratios comparable to those of normal skin. In contrast toverrucae vulgares, the parakeratosis of psoriasis vulgarisis associated with a loss of the stratum granulosum [21].This loss of the granular layer is accompanied by a re-duced expression of NMDAR1.

It seems likely that the loss of NMDA receptors is aconsequence and not a cause of psoriasis vulgaris. Inpsoriasis, the NMDA quotient of all preparations is notsignificantly different from that of controls. Reducedexpression is associated first with the onset of paraker-atosis. The reduced expression of the NMDA receptorsfound in our study would fit well with the altered epi-dermal calcium gradient in psoriasis in which the cal-cium peak is missing in the upper epidermis [10, 13].However, it is only the calcium peak due to calciuminflux from the extracellular space that is consideredessential for normal differentiation and cornification ofthe epidermis [3, 9, 17]. Whether the known defect intransmembrane calcium influx in psoriatic keratinocytes[11] is attributable to NMDA receptors cannot yet bedefinitively assessed, but it seems possible at least inparakeratotic cornification, because the expression ofvarious differentiation markers such as filaggrin is acalcium-dependent process [5].

The agreement between the immunohistochemicalexpression of NMDAR1 and that of filaggrin, and theabsence of filaggrin expression after blocking the

Fig. 4 Immunohistochemistry of NMDAR1 in verruca vulgaris,showing maximal NMDAR1 expression in the stratum granulosum

Fig. 3 Immunohistochemistry of NMDAR1 in psoriasis vulgaris,showing a reduction in the NMDAR1-positive stratum granulosumin the presence of parakeratosis (closed arrowheads) (open arrow-heads normal distribution of NMDAR1)

Fig. 5 Immunohistochemistry of filaggrin in psoriasis vulgaris(same patient as in Fig. 3), showing a reduction in the filaggrin-positive granular layer in the presence of parakeratosis (closedarrowheads) (open arrowheads normal distribution of filaggrin)

160

NMDA receptors, support the assumption that NMDAreceptors in the epidermis influence differentiation.Furthermore, we have been able to suppress theexpression of the differentiation marker CK10 in HaCaTcells by means of selective blockade of NMDA receptorswith MK-801 [6]. This permits the conclusion thatNMDA receptors play a role in the differentiation andcornification of keratinocytes. By contrast, recent resultsreported by Fuziwara et al. [7] suggest that the NMDAreceptor is associated with proliferation of epidermalkeratinocytes. In contrast to these findings, we could notdemonstrate any influence of NMDA on the prolifera-tion of HaCaT cells [6].

Whether the different morphology of the nuclei ofparakeratotic cells is due to differences in calciumhomeostasis is not yet clear. However, this appearspossible, since parakeratosis in verrucae vulgares ischaracterized by large, round strong basophil nuclei [16],while stripe-like nucleic remains are seen in psoriasis[21].

The results presented here show that the expression ofthe NMDA receptor occurs especially in the stratumgranulosum. In diseases in which no stratum granulo-sum is formed, the lack of NMDA receptors could be offunctional importance because of the resulting change inthe epidermal calcium gradient leading to parakeratoticcornification.

References

1. Bayer TA, Wiestler OD, Wolf HK (1995) Hippocampal loss ofN-methyl-D-aspartate receptor subunit 1 mRNA in chronictemporal lobe epilepsy. Acta Neuropathol 89:446–450

2. Bikle DD (1996) 1,25(OH)2D3-modulated calcium inducedkeratinocyte differentiation. J Invest Dermatol Symp Proc1:22–27

3. Bikle DD, Ratnam A, Mauro T, Harris J, Pillai S (1996)Changes in calcium responsiveness and handling duringkeratinocyte differentiation. Potential role of the calciumreceptor. J Clin Invest 97:1085–1093

4. Blumcke I,WolfHK,Hof PR,Morrison JH,Wiestler OD (1995)Regional distribution of the AMPA glutamate receptor subunitsGluR2(4) in human hippocampus. Brain Res 682:239–244

5. Elias PM, Ahn SK, DendaM, Brown BE, Crumrine D, KimutaiLK, Komuves L, Lee SH, Feingold KR (2002) Modulations inepidermal calcium regulate the expression of differentiation-specific markers. J Invest Dermatol 119:1128–1136

6. Fischer M, Glanz D, William T, Klapperstueck T, Wohlrab J,Marsch WCh (2004) N-Methyl-D-aspartate-receptors influencethe intracellular calcium concentration of keratinocytes. ExpDermatol 13:512–519

7. Fuziwara S, Inoue K, Denda M (2003) NMDA-type glutamatereceptor is associated with cutaneous barrier homeostasis. JInvest Dermatol 120:1023–1029

8. Gillies RJ, Didier N, Denton M (1986) Determination of cellnumber in monolayer cultures. Anal Biochem 159:109–113

9. Hennings H, Holbrook K, Steinert P, Yuspa S (1980) Growthand differentiation of mouse epidermal cells in culture: effectsof extracellular calcium. Curr Probl Dermatol 10:3–25

10. Hwang SM, Ahn SK, Menon GK, Choi EH, Lee SH (2001)Basis of occlusive therapy in psoriasis: correcting defects inpermeability barrier and calcium gradient. Int J Dermatol40:223–231

11. Karvonen SL, Korkiamaki T, Yla-Outinen H, Nissinen M,Teerikangas H, Pummi K, Karvonen J, Peltonen J (2000)Psoriasis and altered calcium metabolism: downregulated ca-pacitative calcium influx and defective calcium-mediated cellsignaling in cultured psoriatic keratinocytes. J Invest Dermatol114:693–700

12. Lansdown ABG (2002) Calcium: a potential central regulatorin wound healing in the skin. Wound Repair Regen 10:271–285

13. Menon GK, Elias PM (1991) Ultrastructural localization ofcalcium in psoriatic and normal human epidermis. Arch Der-matol 127:57–63

14. Morhenn VB, Waleh NS, Mansbridge JN, Unson D, ZolotorevA, Cline P, Toll L (1994) Evidence for an NMDA receptorsubunit in human keratinocytes and rat cardiocytes. Eur JPharmacol 268:409–414

15. Oda Y, Tu CL, Chang W, Crumrine D, Komuves L, Mauro T,Elias PM, Bikle DD (2000) The calcium sensing receptor and itsalternatively spliced form in murine epidermal differentiation. JBiol Chem 275:1183–1190

16. Penneys N (1997) Diseases caused by viruses. In: Elder D,Elenitsas R, Jaworsky C, Johnson B (eds) Lever’s histopa-thology of the skin, 8th edn. Lippincott-Raven, PhiladelphiaNew York, p 569

Fig. 6a, b In vitro functional study. HaCaT cells after 14 days ofculture in calcium-rich DMEM. Blockade of the NMDA receptorby treatment with 50 lM MK-801 (b) suppresses the expression offilaggrin compared to control (a)

161

17. Sharpe GR, Gillespie JI, Greenwell JR (1989) An increase inintracellular free calcium is an early event during differentiationof cultured human keratinocytes. FEBS Lett 254:25–28

18. Siegel SJ, Brose N, Janssen WG, Gasic GP, Jahn R, Heine-mann SF, Morrison JH (1994) Regional, cellular, and ultra-structural distribution of N-methyl-D-aspartate receptorsubunit 1 in monkey hippocampus. Proc Natl Acad Sci U S A91:564–568

19. Steigleder GK, Gans O (1964) Pathologische Reaktionen derEpidermis. In: Gans O, Steigleder GK (eds) Normale und

pathologische Anatomie der Haut II. Springer, Berlin Gottin-gen Heidelberg New York, pp 184–189

20. Stephenson FA (2001) Subunit characterization of NMDAreceptors. Curr Drug Targets 2:233–239

21. Toussaint S, Kamino H (1997) Noninfectious erythematouspapular, and squamous diseases of the skin. In: Elder D, Ele-nitsas R, Jaworsky C, Johnson B (eds) Lever’s histopathologyof the skin, 8th edn. Lippincott-Raven, Philadelphia NewYork, p 151

162