Significance and Pathogenesis of Basal Keratinocyte Herniations in Psoriasis

Madalene C. Y. Heng, M.B., F.R.A.C.P., Suni G. Kloss, B.A. , Craig S. Kuehn, B.S., and David G. Chase, Ph.D. Department of Medicin e, Division of Dermatology, U CLA San Fernando Vall ey Internal Medicine Program, Veterans Administration Medica l Center, Sepul veda, California, U .S.A. .

U sing transmISSIOn electron microscopy, we studied, ria sis are associated with electron-lucent areas suggestive quantitatively, b asal keratinocyte h erniations (BKH) in re of proteolytic enzyme release. Their apparent association lation to the other basem ent membrane zone changes in with Langerhans cells, neutrophils, macrophages, and psoriatic lesions of varying clinical activity, and in p so endothelial cells may point to these cells as the source of riasiform skin diseases. BKH appears to correlate with dis proteolytic enzymes in psoriasis. BKH may prove to be a ease activity. They do not occur p assively as a result of the useful marker for clinical psoriasis. J Invest D ermatol 87:362- formation of gaps in the basal lamina. BKH in active p so- 366, 1986

ytoplasmic processes fro m basal keratinocytes pro completely resolved lesions, marked only by hyperpigm entation truding into the dermis through gaps in the basal (6 biopsies), and from uninvolved skin (10 biopsies) at least 10 lamina, have been 'observed by several in vestigators cm from the nea res t psoriati c lesion. [1 ,2] in involved psoriatic skin. We have subse Controls: The control population provided biopsies of normal . quently referred to these as basa l keratinocyte her skin from 4 nonpsoriati c patients (age- and sex-matched) , and Cniations or BKH [3,4]. However, their role in psori asis remains from the involved skin of 10 patients with the following pso unclear. Loss of integrity of the keratinocyte basement membrane riasiform skin diseases: seborrheic dermatitis (2 patients), pityri has been described in psoriasis [1 ,5-7]. and BKH have been thought asis rubra pilaris (1 patient), and 1 patient each with the following: to occur passively as a consequence of the formati on of ga ps in contact eczema (rubber chemicals), generalized eczema from the the basal lamina [2]. In this study, we have attempted to show, ophylline, pustular eczema palms and soles, primary irritant der by a series of studies, that the frequency of BKH correlated with matitis from Campho-phenique, contact dermatitis (nickel). li clinical activity; that the presence ofBKH could serve as a marker chen simplex chronicus , and li chen planus. for active psorias is; that BKH in active disease were associated with electron-lucent areas, suggestive of proteolytic autodiges Light and Electron Microscopy Four-millimeter punch biop tion; and that their apparent association with Langerhans cell s, sies were taken and divided 10ngitudinal1y with a sharp scalpel. neutrophils, dermal macrophages, and endothelial cell s suggests One hiM of each biopsy was fixed in 10% neutral buffered for that these cells, w hi ch are know n producers of proteolytic en malin and processed for light microscopy. The paraffin sections zymes, may playa role in BKH formation. were stained with hematoxylin and eosin. The other half w as fixed in 2.5% glutaraldehyde, buffered to pH 7.3 with 0. 1 M PATIENTS AND METHODS sodium phosphate, postfixed in osmium tetroxide, treated en bloc with tanmc acid, dehydrated in alcohol and propylene oxide, and Patient Population All the patients and controls attended the embedded in a mixture of Epon 812 and Araldite 502. Silver Outpatient Dermatology Clinics at the Veterans Administration sections were cut on a Sorval MT 2B ultramicrotome with a Medical Center, at Sepulveda, California. They were divided into diamond knife (Dupont) and examined with a Philips EM 201 the following groups. . transmission electron microscope. Patiel'lls: Biopsies were obtained fro m 36 psoriatic patients. Clin Analysis of Data The fr equency of BKH was determined as ical detai ls have been summarized in Table I. Biopsies were taken follows. A series of slightly overlapping electron micrographs from acute, untreated psoriatic plaques (4 patients), pustular le were taken of a 150-250 ,uin length of basement membrane for sions (3 patients), eruptive and/or generalized psoriasis (3 pa each patient and control, and printed at a constant magnification ti ents), treated patients with plaques without areas of resolution of 17,000 x. The basement membrane lengths were measured on (5 patients), treated patients with plaques containing areas of res a graphics tablet connected to an Apple II plus microcomputer. olution (3 patients), resolving generali zed psoriasis (2 patients), The number of herniati ons and other basement membrane zone (BMZ) changes were counted as the length of basal lamina was measured with the graphics tablet. Manuscript received July 31, 1985 ; accepted for publication February 18, 1986. Supported by a Veterans Adminis tration Merit Proposal Grant. RESULTS Reprint requests to: Madalene C. Y. Heng, M .B., Division of Der matology (1 11 F) , Veterans Administration M edica l Center, 16111 Plum Analysis of the data obtained from psoriatic lesions of varying mer Street, Sepul veda, California 91343. clinical activity shows that BKH (Fig la) were most numerous Abbreviations: in untreated eruptive, p!lstuJar, generalized psoriasis (9.1 ± 1.4 BKH: basal keratinocyte h ern i a~ i on(s) BKH per 200,um basal lamina length), less numerous (3.8 ± 0.7 BMZ: basement membrane zone BKH per 200 ,um basal lamina length) in treated psoriatic lesions

362 VOL. 87, NO.3 SEPTEMBER 1986 BASAL KERATINOCYTE HERNIATIONS IN PSORIASIS 363

Table I. Frequency of Basal Keratinocyte H erniation (BKH) and Basement Membrane Zone Abnormalities in Psoriasis and Controls Lamina Lucida Basal Lamina C hanges Dilatations BKH (no.l200 f..Lm basa l lamina length) N o. of Patients (no.l200 f..Lm basal (no.l200 f..Lm basal C linical Status (biopsies) Gaps Thinning Reduplication lamina length) lamina length)

Psoriasis: Treatedlresolving 10 13.5 :±: 1. 6 34.9 :±: 4.9 30.0 :±: 2.6 21.4 :±: 2.1 3.8 :±: 0.7 Untreated 10 40.6 :±: 1. 9 76.5 ± 5.3 11.1 :±: 1.3 41.0 :±: 1.7 9.1 :±: 1.4 Completely resolved 6 5.4 ± 0.6 19.8 ± 0.9 59.3 :±: 0.9 5.4 :±: 0.8 0 Uninvolved 10 2.2 :±: 0.2 9.7 ± 0.9 10.8 :±: 0.9 4.9 :±: 0.4 0 Controls (nonpsoriatics): Pityriasis rubra pilaris 1 52.2 102. 6 3.6 51.0 1 Eczema (all types) 6 13.6 26.2 33.8 16.0 0 Lichen planus 1 12.0 29.0 39.6 18.7 0 Normal skin 4 0.2 1.0 1.1 0.8 0

without areas of resolution, very infrequent in treated psoriatic the purpose of statistical analysis, the psoriatic group was divided plaques with areas of resolution (1.2 ± 0.4 BKH per 200 ,um into 4 subgroups: involved untreated, treatedlresolving, com epidermal length), and absent in areas that showed complete clin pletely resolved, and uninvolved. An analysis of variance shows ical resolution. They were absent from uninvolved psoriatic skin, that the treated group is significantly different (p < 0.01) from from normal non psoriatic skin, and from a number of common that of the treated, resolved, and uninvolved groups with regard dermatoses. These resuits have been summarized in Table I. For to BKH frequency.

Figure 1. a, Electron micrograph showing a basal keratinocyte herniation (BKH) con sisting of a cytoplasmic process from the basal keratinocyte (BK), protruding through a gap (between arrowheads) in the basal lam ina, into the dermis (D) . Note the e1ectron lucent area (A) adjacent to the BKH, sugges tive of proteolytic autodigestion. LL = dil atation of the lamina lucida; LD = lamina densa. Bar = 1 f..L. b, Electron micrograph of a healing psoriatic plaque, showing re duplication of the lamina densa (LD) . Note also a gap (between arrowheads) in the basal lamina, and multiple dilatations of the lam ina lucida (LL). E = epidermis, D = der m is. Bar = 1 f..L. c, Electron micrograph of a treated psoriatic plaque (steroids), show ing a BKH of the central variety, protruding into the dermis (D) through a gap (between arrowheads) in the lamina densa (LD). Note localized dilatations (LL) of the lamina lu cida. BK = basal keratinocyte. Bar = 1 f..L. 364 HENG ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

An examination of the distribution of BKH in relation to the were also found in uninvolved and healed psori atic skin as well other BMZ changes, such as gaps (Fig 1b) , thinned areas (Fig 1a ) as in a number of control diseases, and are, therefore, not as useful in the lamina densa, and dilatations of the lamina lucida (Fig 1b) as BKH as markers for the presence of the clinical psoriatic lesion. revealed that altho ugh these BMZ abnormalities were, in general, The presence of BKH also in pityriasis rubra pilaris, a disease most numerous in the involved psoriatic lesions, there was no characteri zed by hyperproli fera tion as is psoriasis, and the pres correlation between the number of gaps in the lamina densa and ence of BKH-like structures in a num ber of tumors [8-11) would the frequency ofBKH found (Table I). These BMZ changes were suggest that BKH may playa role in hyperproliferation. By pro also found in all the control psoriasiform lesions studied. In pit viding a means whereby epidermal and stro mal elements ca n react yriasis rosea, BKH were noted, but in a fre quency far lower than more intimately, over a larger surface area, it is possible that that found in psoriasis. Many gaps were noted, particul arl y in BKH m ay result in enhancem ent of the epidermal-stromal in seborrheic dermatitis, w ithout the presence of a single BKH. In teraction that results in chronicity of the clinical lesion. The BKH psoriasis and many of the control diseases, Langerhans cell s, lym may provide perpetuation of epidermal-stromal interaction that phocytes, and neutrophils were seen in the process of crossing results in chronicity of the clinical lesion. the basement membrane, although such observations were more The widespread occurrence of gaps in the lamina densa and frequently found in psoriasis. Multilayered basal lamina (redu other BMZ abnormalities in psoriasis and in the psorias iform plication, Fig 1 b) were most prominent in the resolving and com diseases studied suggests that these changes may be the nonspecific pletely resolved psoriatic lesions (Table I) , and were frequently consequence of the inflammatory process. Proteolytic enzym es seen in the control biopsies (Table I) . released from macrophages, neutrophils, an d endothelial cells The BKH observed in biopsies from the patients were further [14-1 6) are capable of causing basement membrane dissolution examined and divided in to 2 subgroups accordin g to whether or and may be responsible for ga ps in the basal lamina. In addition , not they were associated with electron-lucent areas in their im chymotrypsin-like enzymes released from the epidermis and dam mediate dermal aspect (Figs 2,3). T he data have been summarized aged mast cells in the dermis may result in dilatory changes ob in Table II. It was found that virtuall y all the BKH fro m active, served in the lamina lucida [17). The predominance of BKH in untreated psoriasis were associated w ith electron-lucent areas in psoriatic lesions suggests the value of these structures as a marker their dermal aspect. Surprisi ngly, even BKH from healing pso for active psoriasis, and investigation into their formation m ay, riatic lesions were associated with similar electron-lucent areas in therefore, lead to a deeper understanding of psoriasis. all but 3 structures (Table II ). The 3 BKH with minimal or no The work of Sugrue and Hay [1 8) suggests that proteolytic surrounding electron lucency (Fig 1e) were found in 1 patient who enzymes may have been responsible for the formation of BKH had received prolonged topical steroid therapy. like structures. U sing corneal epithelia in ti ssue culture, these It was also noted that BKH were associated w ith the follo wing investigators found that when a trypsin-collagenase mixture was cell types: (1) Langerhans cell s in the process of crossing the basal added to their culture medium, the basal surface of their epithelia membrane; (2) lymphocytes in the process of crossing the base developed persistent "blebs" that protruded through gaps in the ment membrane; (3) derm al macrophages; (4) neutrophils (Fig basa l lamina. In addition, they observed that when they added 2); and (5) endothelial cells (Fig 3). M any of these cells had cy laminin, fib ronectin, and type I collagen, i.e., the in gredients that toplas mic extensions toward the BKH. Occasionally, BKH were have been digested by their enzyme mixture, to their culture noted to be in contact with locali zed areas of electron lu cency, medium, the blebs retracted spontaneously within 2-6 h, re which were lined by cytoplasm from the neutrophil (Fig 2) or gardless of the continued presence of gaps in the basal lamina. In endotheli al cell (Fig 3). addition, there is evidence that the release of proteolytic enzymes in psoriasis may be excessive [1 9,20). Our observations of the intimate association of BKH with locali zed areas of electron lu DISCUSSION cency, apparently produced by neutrophils (Fig 2) and endothelial Our data in Table I show that BKH frequency correlates with cells (Fig 3), are in support of these concepts. We suggest that clinical activity in psoriatic lesions. Although there was some the localized areas of electron lucency may represent areas of correlation of the other basement mem brane changes with clinical autodigestion by proteolytic enzymes released from these cells. activity, the gaps in the basal lamina and lamina lucida dilatations Further support of the involvement of proteolytic enzymes in

Figure 2. Electron photomicrograph showing a BKH closely associated with a localized electron-lucent area (P), lined by cytoplasm fro m an underlying neutrophil (N). The gap in the basal lamina is indi cated by the space between arrowheads. E = epidermis. Bar = 1 Jl. . VOL. 87. NO.3 SEPTEMBER 1986 BASAL KERATINOC YTE HERNIATIONS IN PSORIA SIS 365

Figure 3. Electron photomicrograph showin g close contact and association be tween a BKH, and localized electron-lu cent areas (P) lined by cytoplasm from a dermal endothelial cell (E). A = basal ke ratinocyte. The gap in the basal lamina is indicated by the space betweell arrowheads. Bar = 1 J.L .

disease activity in psoriasis is provided by our recent observations It is of interest that individual endothelial ce ll s migrating toward [21] that psoriatic individuals with comcomitant ai-antitrypsin an angiogenic stimulus ha ve been noted not to possess a basement deficiency had a more severe manifestation of the disease. This membrane [24], and we ha ve reported their presence in psoriasis proteolytic inhibitor, together with az-macroglobulin, provides induced by tape-stripping [25]. Endothelial cells are also ca pable more than 90% of the antiproteolytic activity in serum [22] . We of degrading type [ collagen [26]. which is thought to playa have, moreover, observed that in these psoriatic patients with central role in maintaining the integrity of the basement mem associated ai-antitrypsin deficiency, the morphology of some of brane [1 8,27]. predominantly by reducing the degradation of basal the BKH were noted to be altered to be broader at the base or lamina proteoglycan [27]. It is interesting to note in tape-stripped to consist of multiple polypoid protrusions into the dermis [23]. uninvolved psoriatic skin that although Langerhans cell s and lym lending further support to the concept that the release of excessive phocytes were noted to cross the basement membrane as early as proteolytic enzymes m ay playa role in BKH formation in pso 2 nl.in after tape-stripping [25]. BKH were observed only after naSlS. 1-2 weeks. On the other hand, endotheli al cell proliferation and

Table II. Association of Basal Keratinocyte Herniations (BKH) and Gaps in Basal Lamina (BL) with Electron-Lucent Areas in Dermis

Gaps in BL BKH BL Length BL Length No.l200 J.LlTI No.l200 J.Lm Not Associated Not Associated Association with with ElL Association with with ElL ElL Areas' Areas ElL Areas Areas Psoriasis Involved: Untreated 4 46 58" o Treated. o 76 9 3 Uninvolved o 17 o o Resolved o 14 o o Seborrheic Dermatitis 2 34 o o Eczema (all types) 2 51 o o Lichen planus o 6 o o Pi tyriasis rubra pi laris o 27 I o 'ElL = electron-lucent. ' Highly significant (p < 0.001). 366 HENG ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY intrae pidermal neutrophils were noted just prior to th e formation 13. Johnson-Muller B, Gross J: Regul ation of cornea l coll agenase pro of BKH [25], suggesting, perhaps, that these cellular types may duction: epithelial-stromal cell interactions. Proc Nat! Aca d Sci playa deflI1itive role in BKH formation. More studies explorin g USA 75:4417-4421,1978 these conce pts are, therefore, indicated. 14. Mainardi CL, Dixit SN, Kang AH: Degradation of type IV (base ment membrane) coll agen by a protein ase isolated from human polymorphonuclea r leukocyte granules. J BioI C hem 10:5434-5441, We ackllowledg e til e editorial skills of Eletlll or M. Rllssell . 1980 15. Mainardi C L, Seyer JM, Kang AH: Type-specific collagenolysis: a type V collagen degrading enzyme from macrophages. Biochem REFERENCES Biophys Res Commun 97:1108-1115, 1980 1. Brody I: T he ultrastructure of the epidermis in psoriasis vulga ri s as 16. Ka lebi c T, Garbisa S, Glaser B, Liotta LA: Basement membrane revealed by el ectron microscopy. 1. The dermo-epidermal junc colla gen: degradation by migrating endothelial cells. Science tion and the stratum basaJe in parakeratosis without keratohyalin. 221:281-283, 1983 J UItras truct Res 6:304-323, 1962 17. Fraki ]E, Schechter MM, Laza rus WS: Human skin protineses as 2. Cox AJ: The dermal-epidermal juncti on in psoriasis. J Invest Der inflammatory mediators. Br J Dermatol Suppl 15:72-76, 1983 matol 53:428-435, 1969 18. Sugrue SP, Hay ED: Response of basa l epithelial cell surface and 3. Heng MC Y, Kuehn CS, Kloss SG, C hase DG: Basa l keratinocyte cytoskeleton to solubilized extracellular matrix molecules. J Cell herniations in psori asis. Clin Res 32:17 A, 1984 Bioi 91 :45-54, 1981 4. Heng MCY, Kuehn CS, Kloss SG, C hase DG: Basal keratinocyte 19. Stuttgen G, Hofmann N , Simmich W: Die Proteolyse normaler und herniations and associated BMZ changes in psoriasis. C lin Res pathologis ch vedinderter Haut durch Endopeptidasen. Arch Klin 32:138A, 1984 Exp Dermatol 205:381-388, 1957 5. Brody I: Alterations of clinically normal skin of ea rly eruptive guttate 20. Fraki JE, Hopsu-Havu VK: Human skin proteases. Fractionation of psoriasis. A light on electron microscopic study, J C utan Pathol psoriatic scale proteases and separation of a plas minogen acti vator 5:219-233, 1979 and a histone lysin g protease. Arch Dermatol Res 256: 113-126, 6. Gay S, Kresin a TF, Gay R, Miller EJ, Montes LF: Immunohisto 1976 chemica l demonstration of basement membrane collagen in nor 21. Heng MCY, Moy RL, Lieberman J: Alpha I-antitrypsin deficiency mal human skin and in psori as is. J C utan Pathol 6:91-95, 1979 in severe psori asis. Br J Dermatol 112:129-133, 1985 7. Wilborn WH, Montes LF: Ultras tructural changes in psoriatic epi 22. Rindernecht H, Geokus MC: On the physiological role of alpha 2- dermis following anthralin treatment. J C utan PathoI1 :132-150, macroglobulin. Biochim Biophys Acta 295:233-244, 1973 1974 23. Heng M CY, Kloss SG: Electron mi c ro ~ copic features in psoriatic 8. Luibel FJ, Sanders E, Ashworth CT: An electron microscopic study patients with alpha I-antitrypsin deficiency. J Inves t Dermatol of ca ncer i11 situ and invas ive ca rcinoma of cervix uteri. Cancer -87:59-64, 1986 Res 20:357-361, 1960 24. Schoefl GI: Studies on inflammation. III Growing ca pillaries : their 9. Sugar F: An electron microscopic study of ea rly invasive growth in structure and permea bility. Virchows Anat [A] 337:97-141, 1963 human skin tumors and laryngeal ca rcinoma. Eur J Cancer 4:33-38, 25. Heng MCY, Kloss SG, Kuehn C S, Chase DG: The sequence of 1968 events in psoriatic plaque formation after tape-stripping. Br J Der 10. Frei JV: T he fin e structure of the base ment membrane in epidermal matol 11 2:517-532, 1985 tumors. J Cell Bioi 15:335-342, 1962 26. Moscatelli D, Jaffe E', Rifkin DB: Tetradecanoyl phorbol acetate 11. Woods DA, Smith CJ: Ultrastructure of the dermal-ep id ermaljunc stimulates latent collagenase production by cultured human endo tion in experimentall y induced tumors and human oral lesions. J thelial cells. Cell 20:343-351 , 1980 In ves t Dermatol 52:259-363, 1969 27. David G, Bernfield M: T ype I collagen reduces the degradation of 12. Grillo HC , Gross J: Coll agenolytic activity during mammalian wound basa l lamina proteoglycan by mammary epithelial cells. J Cell Bioi repair. Dev Bioi 15:300-317, 1967 91 :281-286, 1981