M H Allen

Endothelial leucocyte adhesion molecule-1 (ELAM-1) expression in cutaneous inflammation.

Endothelial leucocyte adhesion molecule‐1 (ELAM‐1) is a recently described endothelial surface glycoprotein which is inducible by interleukin 1 (IL‐1). tumour necrosis factor‐alpha (TNF‐α) or bacterial lipopolysaccharide (LPS). Using an immunohistochemical technique and a monoclonal antibody (1.2B6) specific for ELAM‐1 we have found marked vascular endothelial expression of ELAM‐1 in many cutaneous inflammatory disorders, including allergic contact dermatitis, atopic dermatitis and psoriasis, and in dermal infiltrates associated with benign, premalignant and malignant keratinocyte proliferation. In normal skin, minimal levels of ELAM‐1 expression were detected. In psoriasis, double immunoenzyme staining studies revealed a close spatial relationship between ELAM‐1 expression and neutrophil margination, suggesting a functional link. Recombinant human interferon‐7 (30 μg) injected intradermally in normal adult human volunteers did not substantially upregulate ELAM‐1 in contrast to its marked effect on intercellular adhesion molecule‐1 (ICAM‐1) expression, indicating that this cytokine is probably not involved in ELAM‐1 induction in vivo. These results indicate that ELAM‐1 is widely induced in cutaneous inflammation with a time course of expression that is longer than that observed in vitro.

Tumour necrosis factor alpha is pro-inflammatory in normal human skin and modulates cutaneous adhesion molecule expression

Tumour necrosis factor a (TNF‐α) is a potent immunoregulatory cytokine produced by many cutaneous cells, including kcratinocytes, mast cells and Langerhans cells. To explore its potential role in inflammatory skin disease, we have studied immunohistochemically the effects of intradermal recombinant human TNF‐α (rHuTNF‐α) on cutaneous inflammatory cells, adhesion molecules and Langerhans cells in normal human skin. Volunteers received rHuTNF‐α 100U (group A), 5000 U (group B), or 100 U daily for 5 days (group C), and biopsies were taken at 6 h (groups A and B), or 6 h after the final injection (group C). An inflammatory cell infiltrate developed in all cases: following single injections of either 100 or 5000 U rHuTNF‐α this was predominantly neutrophilic, whereas following multiple injections of 100 U few neutrophils were seen, although many lymphocytes (CD3+, CD44) were present. In all groups there was an increase in cells of monocyte/macrophage lineage (CD36)+. TNF‐α induced a dose‐ and time‐dependent decrease in CDla+ epidermal Langerhans cell numbers and an increase in dermal CDla4 cells, suggesting migration of Langerhans cells away from the epidermis. TNF‐α induced endothelial E‐selectin, intercellular adhesion molecule‐1 (ICAM‐1) and vascular cell adhesion molecule‐1 (VCAM‐1) in all groups, and adhesion molecule expression by interstitial dermal dendritic cells (ICAM‐1 and VCAM‐1) and keratinocytes (ICAM‐1) was observed. These findings indicate that TNF‐α is a potent modulator of cutaneous immune function in vivo, and this central role in the cutaneous immune response suggests that TNF‐α may be an attractive target for therapeutic inhibition.

Alterations induced in normal human skin by in vivo interferon-gamma

In a study of the direct effects of interferon‐gamma (IFN‐γ) on normal human skin, healthy adult male volunteers received either 3 μg (n=4) or 30 μg (n=9) of recombinant IFN‐γ administered intradermally over 3 days. Biopsies were taken on day 6 and histopathological examination of fixed paraffin‐embedded sections from sites which had received 30 μg IFN‐γ revealed a moderate perivascular lymphohistiocytic dermal infiltrate with mast cells. Immunophenotyping of 5 μm cryostat sections demonstrated that 3 μg IFN‐γ induced keratinocyte HLA‐DR expression in the absence of any significant infiltrate. More intense keratinocyte HLA‐DR expression was produced by 30 μg IFN‐γ in all specimens, with HLA‐DP concurrently expressed in three biopsies. The ratio of CD4:CD8 cells within the infiltrate was approximately 3:1. CD1+ cells within the epidermis were markedly depleted by 30μg IFN‐γ, while CD1‐labelled cells were observed in the dermal perivascular infiltrate. Intradermal IFN‐γ induces similar immunopathological changes to those observed in many of the inflammatory dermatoses.

Abnormal expression of epidermal growth factor receptor in cutaneous epithelial tumours

Epidermal growth factor (EGF) and transforming growth factor alpha (TGFα) are important keratinocyte mitogens. Their effects are mediated by a cell membrane receptor (EGFR), quantitative and qualitative abnormalities of which may be responsible for deranged keratinocyte proliferation and differentiation. We have therefore examined EGFR expression immunohistochemically in a variety of benign and malignant epithelial neoplasms using monoclonal antibodies to the extracellular and intracellular receptor domains. In benign tumours (virus wart, seborrhoeic keratosis, keratoacanthoma), there was an ordered pattern of EGFR expression. In malignant tumours (basal and squamous cell carcinoma), there was loss of membrane labelling and cytoplasmic accumulation of the receptor. In premalignant proliferations, there was loss of membrane receptor with either absent cytoplasmic EGFR (actinic keratosis) or cytoplasmic receptor accumulation (Bowens disease). Evidence of truncated receptors was not found. We suggest that dysregulation of the EGFR may be important in the development of cutaneous epithelial malignancies but that grossly abnormal forms of the receptor do not occur.

T‐cell inducer populations in cutaneous inflammation: a predominance of T helper‐inducer lymphocytes (THi) in the infiltrate of inflammatory dermatoses

The mononuclear infiltrate found in a variety of inflammatory dermatoses was characterized by a predominance of T helper‐inducer lymphocytes (THi), CD4+/CD45RA −/CD45RO+, a population of cells responsible for maintaining and promoting immune reactions. Only small numbers of T‐suppressor‐inducer lymphocytes (TSi), CD4+/CD45RA+/CD45RO−, cells responsible for inducing CD8 suppressor‐effector cells to ‘down regulate’ immune reactions, were seen. The predominance of CD4+ THi lymphocytes was common to all dermatoses studied and suggests a common final pathway in chronic cutaneous inflammation, irrespective of initial causative factors.

Expression of E-cadherin in human epidermal non-melanoma cutaneous tumours

Summary E‐cadherin is a calcium‐sensitive, cell‐to‐cell, adhesion molecule that is expressed widely in normal human epithelial tissue. Abnormal expression has been described in colorectal, breast and nasopharyngeal squamous cell carcinomas, where loss of E‐cadherin is associated with an increased metastatic potential. We have examined, by standard immunohistochemical techniques using the monoclonal antibody HECD‐1 (E‐cadherin monoclonal antibody), the distribution of E‐cadherin in normal human skin and in non‐melanoma neoplastic lesions. In the normal epidermis, E‐cadherin was strongly expressed on the surface of keratinocytes and specialized epithelial structures. Staining was absent from the lower pole of basal keratinocytes in contact with the basement membrane. Weak cytoplasmic staining was also noted in basal keratinocytes. No reactivity was demonstrated in dermal structures. The assessment of cutaneous tumours demonstrated an altered pattern of staining in most cases. Cell surface expression was reduced in 28 of 30 cases of basal cell carcinomas (BCC). Twenty showed an additional feature of positive staining on the dermal aspect of peripheral cells of tumour lobules. In squamous cell carcinomas (SCC) (n= 16), surface expression was attenuated in eight and absent in a further four. Strong surface expression, similar to normal skin was seen in all examples of Bowens disease (n= 6), viral wart (n= 3), seborrhoeic keratosis (n= 3) and actinic keratosis (n= 4). This study demonstrates that, in BCC and SCC, but not in premalignant lesions, cell‐surface expression of E‐cadherin is reduced, consistent with the observation that the loss of E‐cadherin is associated with tumour invasion.

Haplotype analysis of distantly related populations implicates corneodesmosin in psoriasis susceptibility

Psoriasis (MIM *177900) is a hyperproliferative skin disorder, characterised by inflammatory cell dermal infiltration, disruption of keratinocyte terminal differentiation, and premature desquamation of the stratum corneum.1 Although the disease pathogenesis is poorly understood, it is well recognised that genetic factors underlie psoriasis susceptibility, as documented by family clustering of the disease and increased concordance rates in monozygotic twin pairs.2,3 To date, genome wide scans have identified seven distinct psoriasis susceptibility regions ( PSORS 1-7 4) and have provided compelling evidence for a major role of the PSORS1 locus on chromosome 6p21.5–9 Linkage disequilibrium (LD) based studies of microsatellite maps have refined the PSORS1 boundaries to a 150 kb interval10–12 harbouring three sets of coding polymorphisms that have repeatedly shown association with psoriasis: HLA-Cw*0602, HCR*269, and CDSN*TTC.13–18 In a recent analysis of a UK family cohort, we have shown that the most common psoriasis susceptibility chromosomes (cluster E) carry HCR*269, CDSN*TTC, and two SNPs (defined as n7 and n9) lying proximal to HLA-C and showing extremely significant disease association. We also observed a rare haplotype (cluster D), marginally over-transmitted to affected patients and likely to have originated from cluster E by a double recombination event, replacing HCR*269 while preserving SNPs 7/9 and CDSN*TTC.19 This observation suggested that CDSN SNPs might be required to confer psoriasis susceptibility, in conjunction with determinants lying in the HLA-C genomic region. CDSN is also a very attractive biological candidate, since corneodesmosin is a desmosomal protein involved in keratinocyte cohesion/desquamation.20–22 With this study, we have sought to validate the hypothesis of CDSN involvement in psoriasis by: (1) confirming that cluster D is a risk haplotype, through the analysis of an independent population, originating from the Gujurat region of northern India; and (2) defining CDSN genetic variation …

Keratinocyte expression of OKM5 antigen in inflammatory cutaneous disease

Keratinocyte expression of the monocyte/macrophage surface antigens defined by OKM1 and OKM5 antibodies (Ortho Diagnostics) was examined using the peroxidase anti‐peroxidase immunohistochemical technique. A range of inflammatory cutaneous disorders were investigated, including lichen planus, psoriasis and atopic dermatitis. Positive suprabasal keratinocyte expression of OKM5 antigen was observed in all disorders, while keratinocyte staining with OKM1 antibody was consistently negative. These results provide further evidence that keratinocytes may play an important role in cutaneous immune responses. Furthermore, they are consistent with the recent observation that HLA‐DR positive keratinocytes may modulate cutaneous immunological reactions by inducing T‐cell unresponsiveness.

Preferential adherence of T lymphocytes and neutrophils to psoriatic epidermis

Summary T lymphocytes and neutrophils accumulate in psoriatic epidermis. To determine whether the epidermis plays an active role in this process through the production of cellular adhesion factors, leucocyte adherence to lesional psoriasis was compared with normal skin in a modified frozen‐section adhesion assay. Lymphocyte and neutrophil suspensions were prepared by standard Ficoll–Hypaque techniques from peripheral blood of normal volunteers and overlaid on to glutaraldehyde‐fixed 8‐μm cryostat sections of skin. Adhesion of phorbol ester‐activated T lymphocytes to the epidermis was significantly greater in psoriasis compared with normal skin (P<0.01). Adhesion was absent (a) at 7°C, (b) in the presence of EDTA and (c) in the absence of lymphocyte activation. Immunostaining demonstrated that all adherent lymphocytes were CD3+ve (i.e. T cells). Likewise, neutrophils adhered more prominently to psoriatic epidermis. Adhesion was most prominent at the tips of dermal papillae, corresponding to areas of maximal intercellular adhesion molecule‐1 (ICAM‐1) expression. Both neutrophils and lymphocytes adhered to dermal papillary vascular endothelium. These studies provide functional data that psoriatic epidermal cells are actively involved in leucocyte adherence. The distribution of adhesion suggests that both ICAM‐1‐dependent and independent mechanisms are involved.

Antigenic profile of human acrosyringium

Summary The demonstration of HLA‐DR on human acrosyringium has led to the suggestion that eccrine epithelium, through its interaction with certain molecules, might play an active role in epidermal immune responses. An immunohistochemical study was undertaken to identify the antigenic profile of acrosyringium in normal skin and following the intradermal administration of a T‐lymphocyte‐derived cytokine, interferon gamma (TFN‐γ). Acrosyringium in normal skin, in contrast to interappendageal epidermis, was found to lack CDla+ Langerhans cells. However, antigens CD36 (OKM5) and Ll (MAC387) were uniquely expressed by keratinocytes immediately adjacent to the distal portion of acrosyringium. Constitutive expression of each class II MHC antigen, namely HLA‐DR, DP and DQ was observed on luminal acrosyringial cells. EBMl 1 antigen (CD68), a mononucler cell determinant, was similarly expressed on acrosyringial epithelium in normal skin. Following intradermal administration of IFN‐γ. Intercellular adhesion molecule‐1 (ICAM‐I) (CD54) was induced on acrosyringial epithelium and the expression of HLA‐DR was intensified. A range of other markers including CD3, CD4, CD8, CDl la, CDl l b and CDl5 were not expressed by acrosyringium either in normal skin or after administration of IFN‐γ.