Annals ofBurns and Fire Disasters - vol. VIII - n. 4 - December 1995

CD36 IS ONE OF THE IMMUNOLOGICAL MARKERS EXPRESSED BY KERATINOCYTES IN ACTIVE HYPERTROPHIC SCARS

Castagnoli C.,(1) Stella M .,(2) Menegatti E,(1) Trombotto C.,(1) Calcagni M.,(2) Magliacani G .,(2) Teich Alasia S. ' (2,3) Alessio M.(4)

1 Centre for Immunogenetics and Experimental Oncology, CNR, Turin, Italy
2 Department of Plastic Surgery and Burn Unit, Trauma Centre, Turin
3 Piedmont Burn Studies and Research Foundation, Turin
4 DIBIT, San Raffaele Scientific Institute, Milan

SUMMARY. The pathogenesis of hypertrophic scars is largely unknown, although there is considerable evidence to suggest that many of its features are analogous to inflammatory dermatoses, such as psoriasis, scleroderma and lichen planus. An aberrant expression of HLA Class 11 and ICAM-1 molecules on keratinocytes is reported in these forms of dermatosis. We have already demonstrated in hypertrophic scars that these activation markers are ectopically present in all layers of the epidermis. Here we show that CD36, a membrane glycoprotein normally expressed on platelets, monocytes and endothelial cells, is expressed on epidermal keratinocytes in the stratum granulosurn and stratum spinosum of active hypertrophic scars but not in normotrophic scars or normal skin. These resulis, combined with the local alteration of the biosynthesis of some cytokines in hypertrophic scars and the ectopic expression of HLA 11 and ICAM- I molecules on keratinocytes and fibroblasts in the tissue involved, are consistent with a pivotal role of an altered cellular immune response in this disease.

The pathogenesis of hypertrophic scars remains controversial. These scars are characterized by hyperproduction of collagen for an extended period of time, hyperplasia and increased cell turnover.' It has recently been suggested that immunological factors may play a major role in the normal processes of wound healing and tissue remodelling.' A disruption of these mechanisms is thought to be involved in pathological wound healing, as in certain forms of dermatosis, e.g. lichen plarms, scleroderma and psoriasis. Several immunological abnormities are reported in these dermatoses, including the anomalous expression on keratinocytes of HLA DR, ICAM-I and CD36 molecules
We have already demonstrated that in hypertrophic scars the tissue involved contains an increased number of activated T lymphocytes (70% of infiltrates) compared to normotrophic scars (35% of infiltrates) .6 In a recent work it has been shown that in hypertrophic tissues there is a local alteration in the biosynthesis of some cytokines.1,1 The anomalous expression of HLA DR and ICAM-1 (CD54) molecules on keratinocytes and fibroblasts in hypertrophic tissues has also been demonstrated.
CD36 is an 88 kDa membrane glycoprotein expressed by the erythrocyte precursors, mature monocytes, platelets, endothelial cells of the microvasculature, and mammary epithelial cells. In platelets this molecule has been shown to serve as the receptor for two extracellular matrix proteins, collagen and thrombospondin I." Additional aspects of CD36 are as follows: i) it is expressed by macrophages involved in phagocytosis of neutrophils and CD4+ and CD8+ T-lymphocytes undergoing apostosis;1 1,12 ii) it is a receptor for oxidized low-density lipoprotein on macrophages; 11 iii) it is implicated in the binding and transport of fatty acids by foam cells.
The role of CD36 as a cell surface receptor has been extended to a signal transduction molecule in platelets and monocytes, being associated in platelets to protein tyrosine klinases of the pp60-,,, gene family.
In this study we investigate the expression of CD36 molecules in hypertrophic scars and their correlation with other activation markers previously demonstrated to be ectopically expressed on keratinocyles. We also analyse the correlation between CD 36 and other classic markers of macrophages and dendritic cells in the tissues involved.

Biopsies were taken after informed consent from 25 patients (15 women and 10 men, aged 23-60 years) subjecled to plastic surgery under general anaesthesia for the correction of extensive hypertrophic scars consequent to thermal injury. The burned surface area (BSA) of the patients corresponded to 10-40% of the body.
Hypertrophic scars were still present at least one year post-trauma despite continuous compressive therapy, massage and physiotherapy. The scars were raised, erythematons, and often painful, with variable degrees of skin retraction affecting articular motility; at the surface they showed telangectasies, little bullae and sometimes trophic ulcers. The overall judgement was that they were active lesions with no sign of regression.
Control samples included five specimens of normal skin taken from patients undergoing corrective plastic surgery procedures, and ten specimens of normotrophic scars from informed consent patients undergoing surgery for reconstructive purposes.

Tissue specimens
Five-millimeter punch biopsies of hypertrophic scars and controls were obtained from different anatomical sites of the body. Tissue samples were snap-frozen in cold isopentane (-70 'C) and mounted in OCT 4583 embedding compound. Five-micron-thick cryostat sections were cut in serial sections and transferred to microscope slides. The slides were air-dried and stored at -80 'C.

Monoclonal antibodies (MoAb) and conventional antisera
The monoclonal antibodies used in this work were: the anti-CD36 MoAb NL07 (IgM);" the anti-HLA-DR, DP MoAb Hot 214 (IgGI), the anti-HLA-DR, DP, DQ MoAb AA3.84, the anti-ICAMA MoAb CL203.4, and the V1727516 [1 and references therein]; the anti-CD 11 c MoAb Leu M5 (lgG2b, Becton~Dickinson); the anti-CD11b MoAb OKM1 (IgGI, Ortho-Diagno sties); and the anti-CD1a MoAb OKT6 (lgGI, Ortho-Di agno sties). Rabbit antimouse lg and peroxidase-conjugated swine anti-rabbit lg were from Dakopatts, Copenhagen, Denmark.

Immunoenzymatic staining procedures
Sequential tissue sections were incubated with MoAb and stained using a three-stage immunoperoxidase reaction (PAP) as previously described.' MoAb were titrated so as to yield maximal specific staining and minimal non-specific or background staining. Endogenous peroxidase activity was inhibited by the addition of methyl alcohol and 0.033% hydrogen peroxide. The specificity of the immunostaining was evaluated by replacing the primary antibody with non immune ascites. Slides were examined double-blind.

Serial sections of hypertrophic scar biopsies from 25 patients, plus sections of normotrophic scar biopsies from ten individuals and five specimens of normal skin, were stained with anti-CD36, anti-ICAM-1, anti-HLA Class 11, anti-CD1 lb, anti-CD1 lc and CD1a with a three-stage immunoperoxidase reaction.
All hypertrophic scar samples showed strong positivity for the CD36 antigen in suprabasal keratinocytes, in particular in the stratum granulosum and stratum spinosum of the epidermis (Figs. ]a, b). The reactivity of the antiCD36 MoAb was sometimes localized in foci in these strata (Fig. 1b). CD36 immunoperoxidase staining was always negative in the basal layer of epidermis and in the stratum corneum (Figs. ]a, b). The pattern of CD36 expression was not dependent on the thickness of the epidermal layer observed in different scar samples.
In all tested serial sections of hypertrophic scars, the reactivity pattern of anti-CD36 MoAb was not comparable to the pattern of anti-HLA Class 11 and anti-ICAM-1 MoAbs (Figs. 2a, b, c). The HLA Class 11 and ICAM-1 antigens were highly expressed on the keratinocyte cell membrane in the basal layer and sometimes in the mid~epi~ dermal zone (Figs. 2a, b), while their expression decreased in the upper layers of epidermis.The CD 1 lb antigen, a known macrophage marker, was never found on keratinocytes of any epidermal layers (Fig. 3). Similar results were obtained when anti-CD I I c MoAb was used (data not shown).In the dermis of all tested hypertrophic scar samples, CD36+ cells were more abundant than in normotrophic scars and normal skin biopsies. CD36+ cells were widely distributed in the subpapillary compartment and in reticular dermis (fig. 4). The immunoreactivity pattern of anti-CD36 MoAb was very similar to patterns seen in serial sections stained with anti-CD11b and CDD I I c MoAbs (Fig. 3). In the dermis of hypertrophic scar samples, the CD36 antigen was also expressed on endothelial cells of the microvasculature (Fig. 4). The vascular endothelial cells, in addition to the expression of CD36 molecules, showed strong reactivity with anti HLA Class 11 and anti-ICAM-1 MoAbs (data not shown). Scattered epidermal-branched Langerhans-like cells (DR+ and CD I a+) found in serial sections of hypertrophic scars did not express CD36 antigens (data not shown). Normal skin and normotrophic scar biopsies were exa mined as control. In contrast to hypertrODhic scar samples.

Figs. la, b - Photomicrograph of hypertrophic scar section stained with anti-CD36 MoAb. CD36 expression can he observed in the suprabasal layer of epidermis (a) and sometimes in the same strata in the foci (b). Expression of CD36 antigens was also observed on endothelial cells of the dermal microvasculature (a) (x 250; scale bar 50 mm). Fig. 2 - Photomicrograph of hypertrophic scar serial sections stained with anti-CD36 (a), anti-HLA class fi (b) and anti-ICAM-1 (c) MoAbs. CD36 expression was present in the suprabasal layer of the epidermis (a) while HLA class 11 and ICAM-1 expressions can be observed in the basal layer as well as in the suprabasal layer of the epidermis (b, c). The HLA class II and ICAM- 1 expressions decreased in keratinocytes of the upper layers. (a, b x 250, scale bar 50 min; c x 200, scale bar 60 min). Fig. 3 - Photomicrograph of hypertrophic scar section stained with antiCD I lb MoAb. Keratitiocytes of full thickness of epidermis did not show expression of CD11b antigen while expression was present on dermal infiltrating cells (x 250, scale bar 50mm). Fig. 4 - Photomicrograph of hypertrophic scar dermal section stained with anti-CD36 MoAb. Expression of CD36 antigens was observed on vascular endothelial cells (arrows) and on infiltrating cells (x 300, scale bar 25 mm). Fig. 5 - Photomicrograph of normotrophic scar section stained with antiCD36 MoAb. CD36 molecule expression was not detectable within the epidermis (x 200, scale bar 60 mm).

CD36 molecules were not detected in the epidermis in normotrophic scar samples (Fig 5). In normal skin sections the reactivity pattern of anti-CD36 MoAb was identical to that seen for normotrophic scar samples. Table I summarizes the results of MoAb reactivity on nonhaematopoietic cells.

The results presented in this work show that CD36 is ectopically expressed on keratinocytes of hypertrophic scars as in other skin pathologies such as psoriasis, lichen planus and pemphigus vulgaris. These data further support our hypothesis of an involvement of the immune system in hypertrophic scarring.
A recent interesting work has reported that when normal keratinocytes are treated in vitro with g-interferon (gIFN) they express the same molecules (ICAM-1, HLADR) that are observed in hypertrophic sear tissues and they are also capable of providing co-stimulatory signals to T cells .21 These keratinocytes, seem to play a fundamental role as accessory cells, delivering co-stimulatory signals to T cells engaged by antigens or superantigens and influencing the development of different T cell immune responses.
It is tempting to speculate that also in hypertrophic scarring the CD36, DR and ICAM-1 molecules" expressed on keratinocytes might act as co-stimulatory molecules recruiting and activating T cells. In hypertrophic sear bi-

It has to be underlined that the expression of HLA DR, ICAM-1 and CD36 molecules is not comparable: CD36 is detected only in the stratum granulosum and stratum spinosum while the others are detected in all epidermal layers. The different expression pattern of CD36, HLA class II and ICAM-1 could suggest that keratinocytes in hypertrophic scars evolve through sequential stages of differentiation, probably playing different roles in the immune response in scars.
In conclusion, in hypertrophic scars the ectopic expression on keratinocytes of HLA class 11, ICAM-1 and CD36 molecules suggests a cross-talk with the immunocompetent cells, underlying the involvement of the immune system in pathological scarring. However, the regulation of the ectopic expression and the role of CD36 antigen in keratinocytes of hypertrophic scars still remain to be defined.

RESUME. La pathogenèse des cicatrices hypertrophiques n'est pas bien connue, même si les évidences indiquent que beaucoup de ses aspects charactéristiques sont analogues à celle des dermatoses inflammatoires, comme le psoriasis, la sclérodermie et le lichen planus. Dans ces formes de dermatose une expression aberrante des molécules HLA de classe Il et ICAM-1 sur les kératinocytes a été observée. Dans les cicatrices hypertrophiques nous avons déjà démontré que ces marqueurs de l'activation sont présents ectopiquement dans toutes les strates de l'épiderme. Ici nous montrons que la CD 36, une glycoprotéine membraneuse normalement exprimée sur les plaquettes, les monocytes et les cellules endothéliales, est exprimée sur les kératinocytes épidermiques dans le stratum granulosum et le stratum spinosum des cicatrices hypertrophiques actives mais non dans les cicatrices normotrophiques ou la peau normale. Ces résultats, alliés à l'altération locale de la biosynthèse de certains cytokines dans les cicatrices hypertrophiques et l'expression ectopique des molécules HLA Il et 1CAM-1 sur les kératinocytes et les fibroblastes du tissu intéressé, sont compatibles avec un rôle central d'une immunoréponse cellulaire altérée dans cette maladie.

Acknowledgements. This work was supported by the Fondazione Piemontese per gli Studi e le Ricerche sulle Ustioni (Piedmont Burn Studies and Research Foundation). Dr C. Castagnoli was supported by a fellowship from the same Foundation. We thank Dr G. Ponzio for his help with the photomicrographs and Dr R. Sitia for his critical reading of the manuscript.

1. Stella M., Castagnoli C., Magliacani G., Richiardi P.: Fisiopatologia delta cicatrizzazione patologica (revisione bibliografica). Riv. Ital. Chir. Plast., 21: 199-208, 1989.
2. Janssen de Limpens A.M.P., Cormane R.H.: Studies on the immunologic aspects of keloid and hypertrophic scars. Arch. Dermat. Res., 274: 259-64, 1982.
3. Vejlsgaard G.L., Ralfkiaer E., Avnstorp C., Czajkowski M., Marlin S.D., Rothlein R.: Kinetics and characterization of ICAM-1 expression on keratinocytes in various inflammatory skin lesions and malignant cutaneous lymphomas. J. Am. Acad. Dermatol., 20: 782-92, 1989.
4. Griffiths C.E.M., Voorhees J.J., Nickoloff B.J.: Characterization of ICAM-1 and HLA-DR expression in normal and inflamed skin:modulation by recombinant gamma interferon and tumor necrosis factor. J. Am. Acad. Dermatol., 20: 617-24, 1989.
5. Lisby S., Ralfkiaer E., Hansen E.R., Vejlsgaard G.L.: Keratinocyte and epidermal leukocyte expression of CD36 (OKM5) in benign and malignant skin diseases. Acta Derm. Venereol., 70: 18-22, 1990.
6. Castagnoli C., Stella M., Berthod C., Magliacani G., Momigliano Richiardi P.: TNF production in hypertrophic scarring. Cell Immunol., 147: 51-4, 1993.
7. Castagnoli C., Stella M., Magliacani G., Richiardi P.: The role of TNF alpha and beta cytokines in scar hypertrophy in bum patients: an immumohistochemical study. Ann. Medit. Burns Club, 8: 23-7,1995.
8. Peruccio D., Castagnoli C., Stella M., D'Alfonso S., Momigliano Richiardi P., Magliacani G., Teich Alasia S.: Altered biosynthesis of TNF alpha is involved in post-burn hypertrophic scars. Burns, 20: 118-21, 1994.
9. Castagnoli C., Stella M., Magliacani G., Teich Alasia S., Momigliano Richiardi P.: Anomalous expression of HLA class 11 molecules on keratinocytes and fibroblasts in hypertrophic scars consequent to thermal injury. Clin. Exp. Immunol., 82: 350-5, 1990.
10. Castagnoli C., Stella M., Magliacani G., Ferrone S., Momigliano Richiardi P.: Similar ectopic expression of ICAM- I and HLA Class 11 molecules in hypertrophic scars following thermal injury. Bums, 20: 430-3, 1994.
11. Greenwalt D.E., Lipsky R.H., Ockenhouse C.F., Ikeda H., Tandon N.N., Jamieson G.A. ~ Membrane glycoprotein CD36: a review of its roles in adherence, signal transduction and transfusion medicine. Blood, 80: 1105-15, 1992.
12. Savill J., Hogg N., Ren Y., Haslett C.: Thrombospondin cooperates with CD36 and vitronectin receptor in macrophage recognition of neutrophils undergoing apoptosis. J. Clin. Invest., 90: 1513-22, 1992.
13. Akbar A.N., Savill J., Gombert W., Bofill M., Borthwick N.J., Whitelaw F., Grundy J.E., Janossy G., Salmon M.: The specific recognition by macrophages of CD8+, CD45RO+ T cells undergoing apoptosis: a mechanism for T-cell clearance during resolution of viral infections. J. Exp. Med., 180: 1943-7, 1994.
14. Endemann G., Stanton L.W., Madden K.S., Bryant C.M., White R.T., Protter A.A.: CD36 is a receptor for oxidized low density lipoprotein. J. Biol. Chem., 268: 11811-6, 1993.
15. Abumrad N.A., EI-Maghrabi M., Amri E., Lopez E., Grimaldi P.A.:Cloning of a rat adipocyte membrane protein implicated in binding or transport of long-chain fatty acids that is induced during pre-adi pocyte differentiation. J. Biol. Chem., 268: 17665-8, 1993.
16. Alessio M., Greco N.J., Primo L., Ghigo D., Bosia A., Tandon N.N., Ockenhouse C.F., Jamieson G.A., Malavasi F.: Platelet activation and inhibition of malarial cytoadherence by the anti-CD36 IgM monoclonal antibody NL07. Blood, 82: 3637-47, 1993.
17. Ockenhouse C.F., Magowan C., Chulay J.D.: Activation of monocytes and platelets by monoclonal antibodies or malaria-infected erythrocytes binding to the CD36 surface receptor in vitro. J. Clin. 24.Invest., 84: 468-75, 1989.
18. Trezzini C., Jungi T.W., Spycher M.O., Maly F.E., Rao P.: Human monocyte CD36 and CD16 are signaling molecules. Evidence from studies using antibody -induced chemiluminescence as a tool to probe signal transduction. Immunology, 71: 29-37, 1990.
19. Alessio M., Ghigo D., Garbarino G., Geuna M., Malavasi F.: Analysis of the human CD36 leucocyte differentiation antigen by means of the monoclonal antibody NL07. Cell Immunol., 137: 487 500,1991.
20. Barker J.N.W.N., Markey A.C., Allen M.H., MacDonald D.M.: Keratinocyte expression of OKM5 antigen in inflammatory cutaneous disease. Brit. J. Derm:, 120: 613-8, 1989.
21. Viac J., Chardonnet Y.: Immunocompetent cells and epithelial cell modifications in molluscum contagiosum. J. Cutan. Pathol., 17: 202-5, 1990.
22. Nickoloff B.J., Turka L.A.: Immunological functions of non-professional antigen-presenting cells: new insight from studies of T-cell interactions with keratinocytes. Immunol. Today, 15: 464-9, 1994.
23. Nikoloff B.J., Mitra R.S., Green J., Zheng X-G., Shimizu Y., Thompson C., Turka L.A.: Accessory cell function of keratinocytes for superantigens. J. Immunol., 150: 2148-59, 1993.
24. Castagnoli C., Trombotto C., Stella M., Calcagni M., Momigliano Richiardi P., Magliacani G.: Interferon garnma and interferon gamma receptor in post-burn hypertrophic scars: is it a remission marker? Abs. VI Cong. E.B.A., 1995.