Academia.eduAcademia.edu

Dendritic cells in the dermal infiltrate of S�zary syndrome

1986, Histopathology

Dendritic cells in the dermal infiltrate of Sezary syndrome The dermal infiltrates of four patients with the SCzary syndrome were studied by electron microscopy and the data were evaluated quantitatively. The nuclear contour index of lymphocytes was calculated, and many turnour cells had an index greater than 6.5. Dendritic cells were found in all cases. The dendritic cells contained smooth and rough endoplasmic reticulum, moderately well-developed Golgi apparatus, scanty lysosomes and many thin and intermediate filaments; their surface was scalloped with numerous vesicles. Birbeck granules were not found in the cytoplasm of dendritic cells. Dendritic cells comprised 24% of infiltrating cells and were interspersed with lymphocytes; 75% of the lymphocytes were in contact with dendritic cells; 35% of the lymphocytes in contact with dendritic cells had a nuclear contour index higher than 6.5 and 76% had a nuclear contour index higher than 5. The data strongly suggest a functional relationship between lymphocytes and dendritic cells in the dermal infiltrate of SCzary syndrome. They are discussed in relation to the hypothesis that the disease is a consequence of chronic immune stimulation.

zyxwvuts zyxwvu zyxwvu Histopathofogy 1986, 10, 25-36 zyx Dendritic cells in the dermal infiltrate of Sezary syndrome P.ROMAGNOL1, S.MORETTI*,A.FATTOROSS1-t & B.GIANNOTTI* Istituto di Istologia e Embriologia Generale, Universita di Firenze, Italia, * Clinica Derrnatologica II, Universita di Firenze, Italia and TCattedra di Allergolgia e Irnrnunologia Clinica, Universita di Rorna ‘ L a Sapienza’, Italia zyxwvut Accepted for publication 17 June 1985 ROMACNOLI P.,MORETTI S . , FATTOROSSI A. & G I A N O B. ~(1986) Histopathofogy 10, 25-36 Dendritic cells in the dermal infiltrate of Sezary syndrome The dermal infiltrates of four patients with the SCzary syndrome were studied by electron microscopy and the data were evaluated quantitatively. The nuclear contour index of lymphocytes was calculated, and many turnour cells had an index greater than 6.5. Dendritic cells were found in all cases. The dendritic cells contained smooth and rough endoplasmic reticulum, moderately well-developed Golgi apparatus, scanty lysosomes and many thin and intermediate filaments; their surface was scalloped with numerous vesicles. Birbeck granules were not found in the cytoplasm of dendritic cells. Dendritic cells comprised 24% of infiltrating cells and were interspersed with lymphocytes; 75% of the lymphocytes were in contact with dendritic cells; 35% of the lymphocytes in contact with dendritic cells had a nuclear contour index higher than 6.5 and 76% had a nuclear contour index higher than 5 . The data strongly suggest a functional relationship between lymphocytes and dendritic cells in the dermal infiltrate of SCzary syndrome. They are discussed in relation to the hypothesis that the disease is a consequence of chronic immune stimulation. zyxwv Keywords: SCzary syndrome, dendritic cells, electron microscopy Introduction Dendritic cells is the term used to define a group of specialized cells which play a critical role as accessory cells in thymus-dependent immune responses (Nussenzweig & Steinman 1980, Van Voorhis et al. 1982). This group includes veiled cells of afferent lymph, interdigitating reticulum cells of T-cell areas in lymph nodes and spleen, Langerhans cells and at least some of the epidermal and dermal Address for correspondence: Dr P.Romagnoli, Istituto di Istologia e Ernbriologia Generale, V.le Pierraccini 18,1-50139 Firenze, Italia. 25 26 zyxwvutsrq zyxwv zyxw P.Rornagnoli et al. indeterminate cells (Rowden, Phillips & Lewis 1979, Balfour et al. 1981, Hoefsmith, Duijvestijn & Kamperdijk 1982, Knight et al. 1983). Dendritic cells seem to be a common feature in the dermal infiltrate of mycosis fungoides (MF) (Jimbow, Chiba & Horikoshi 1982); when, as often occurs in this infiltrate, dendritic cells contain Birbeck granules, they are identified as Langerhans cells (De Panfilis 1983). Dendritic cells have also been described in the dermal infiltrate of four patients with Sezary syndrome (SS) (Goos 1976, Van der Loo et al. 1979; Iwahara & Hashimoto 1984), but these authors do not state whether Birbeck granules were found. Moreover, it is not known whether dendritic cells, with or without Birbeck granules, are a constant finding in the dermal infiltrate of SS. In view of the critical role of dendritic cells as accessory cells in antigen presentation (Van Voorhis et al. 1982, Van Voorhis, Witmer & Steinman 1983), we have studied the ultrastructural features of dendritic cells in the dermal infiltrate of four patients with SS, concentrating on the types of these cells and the extent of their relationships with infiltrating lymphocytes. Clinical studies Four Caucasian patients observed at the Clinic of Dermatology of Florence from 1980 to 1983 were studied. Before the onset of treatment, each patient had a skin biopsy examined on light and electron microscopy and had phenotypic and functional studies carried out on peripheral blood lymphocytes. CASE1 S.J., a 56-year-old woman, had generalized exfoliative erythroderma, plaques on the legs and extreme diffuse itching that had been present for 10 years. Superficial lymphadenopathy was also present. A skin biopsy revealed a mononuclear infiltrate of lymphocytes with cerebriform nuclei in the upper dermis and in the epidermis. Pautrier’s microabscesses were also found. In the peripheral blood, leucocytes numbered 50X 103/mm3, of which 10% were neutrophils and 90% lymphocytes, most of which had cerebriform nuclei. Lymphocytes with cerebriform nuclei were found in lymph node and bone marrow biopsy on light microscopical examination. CASE2 B.F., a 71-year-old woman had generalized itching and diffuse erythroderma that had been present for 6 years. A skin biopsy showed lymphoid cells with cerebriform nuclei diffusely infiltrating the dermis and forming nests in the epidermis. In the peripheral blood, leucocytes numbered 16x 103/mm3, of which zyxw zy zyxw zyxwvu Dendritic cells in Stzary syndrome 27 14% were neutrophils, 2% eosinophils, 81% lymphocytes; virtually all lymphocytes had cerebriform nuclei. CASE3 S.V., a 53-year-old male, had localized erythema of the arms that had been present for 10 years with the development of an exfoliative, itching erythroderma more recently. Superficial lymphadenopathy was also present. A skin biopsy revealed an infiltrate, mainly made up of lymphocytes with cerebriform nuclei, in the dermis and epidermis. Peripheral blood leucocytes numbered 9.5X 10’/mm3, of which 19% were neutrophils, 2% eosinophils, 76% lymphocytes and 2% monocytes; all the lymphocytes had cerebriform nuclei. A n axillary lymph node biopsy showed a dermopathic hyperplasia. zyxwvut CASE4 B.B., a 60-year-old male had diffuse erythroderma and extreme itching that had been present for three years. Generalized lymphadenopathy was also evident. A skin biopsy showed dermal and epidermal infiltration by Lymphoid cells with cerebriform nuclei. Peripheral blood leucocytes numbered 40 X lO3/rnm3,of which 36% were neutrophils and 60% lymphocytes, all with cerebriform nuclei. Specific involvement of lymph nodes and bone marrow was demonstrated by light microscopy. Cells from this patient were also tested for the presence of human T-cell leukaemidlymphoma virus (HTLV- 1) DNA integrated sequences by restriction enzyme analysis according to Manzari et al. (1984). Such sequences were not found however. Immunological studies Studies on circulating lymphocytes gave the same results in all the patients and a common description is therefore given. At least 90% of circulating mononuclear cells were able to form E rosettes and reacted with OKT3 and OKT4 monoclonal antibodies. Less than 3% of circulating mononuclear cells reacted with OKT8 monoclonal antibody and less than 2% with OKMI monoclonal antibody. Virtually no cell reacted with OKT6 monoclonal antibody. All the monoclonal antibodies were from Ortho Diagnostic Systems, Rarutan, NJ, USA. Less than 6% of circulating mononuclear cells had surface immunoglobulins. The following functional activities were investigated: 1 the response to the polyclonal activators phytohemagglutinin, concanavalin-A and pokeweed mitogen (PWM) in a 72 hour stimulation assay; 2 the response to allogeneic cells from normal donors in a one-way mixed lymphocyte culture assay; 3 the helper or suppressor capacity in a PWM-driven immunoglobulin in-vitro synthesis assay (Romagnani et al. 1982). 28 zyxwvutsr zyxwvutsr zyx zy zyxw P. Romagnoli et al. In all the patients the responses to polyclonal activators and allogeneic cells were markedly impaired and cooperation with B lymphocytes was defective compared with controls. More detailed immunological data on patients S.J., B.F. and S.V. have been reported previously (Fattorossi et al. 1984). Electron microscopy For this study, skin biopsies were fixed either in 4% glutaraldehyde, or 270 glutaraldehyde and 2% formaldehyde, in 0.1 M cacodylate buffer, pH 7.3, postfixed in 1% Os04 in 0.1 M phosphate buffer, pH 7.3, and embedded in Epon 812 (Luft 1961). Ultrathin sections were stained with uranyl acetate and lead citrate and observed with electron microscopes, Siemens Elmiskop 1 and 102, at 80 kV. For a quantitative evaluation of the cell types in the dermal infiltrate and their reciprocal relationships one ultrathin section was taken at random from the papillary dermis of each patient (including the basal layer of the epidermis), and all the cells in the dermal infiltrate were photographed. The area of section ranged between 1x lo4 and 5 X lo4 ,urn2. The number of cells photographed ranged between 32 and 92 per case (total number 265). The nuclei of both lymphocytes and dendritic cells were counted separately. The ratio of lymphocytes touching the cell body or cytoplasmic projections of dendritic cells to the total number of infiltrating lymphocytes was calculated. The nuclear contour index (NCI) of all the infiltrating lymphocytes was evaluated, by intersection and point counting, with an isotropic curvilinear test lattice (Weibel 1979). The lymphocytes were divided into three classes according to this index (Stolz et al. 1983): normal lymphocytes, with an index of 5 or less, intermediate lymphocytes, with an index between 5 and 6.5, and Sezary cells, with an index higher than 6.5. In all the patients, the dermal infiltrate comprised two cell types, lymphocytes and dendritic cells (Figure 1, Table 1). The majority of lymphocytes had a nuclear contour index higher than 5 ; their cytoplasm contained few organelles and short microvilli were present on their surface (Figure 1). The dendritic cells had a spindle-shaped body and long cytoplasmic projections. The nucleus was large, often indented, with a thin peripheral rim of condensed chromatin and one nucleolus. The cytoplasm contained smooth and rough endoplasmic reticulum and coated vesicles, a moderately well-developed Golgi apparatus, scanty, zyxwvut Table 1. Dermal infiltrate in Sezary syndrome Percentage of dendritic cells Percentage of lymphocytes in contact with dendritic cells S.V. S.J. B.B. B.F. 26.1 32.8 18.R 19.3 82.4 82.0 73.1 61.2 Mean 24.3 74.7 Dendritic cells in Stzary syndrome zy zy 29 zyxwvutsr zyxwvuts Figure 1. Dermal infiltrate of the patient S.V. DC=dendritic cells which are interspersed with lymphocytes, many of which have cerebriform nuclei. The asterisks indicate cytoplasmic projections of dendritic cells. Cap: capillary. ~ 7 0 0 0 . 30 zyxwvutsrq zyxwvu zyxwvut P. Romagnoli et al. zyxwvutsr zyxwvutsrq Figure 2. Cytoplasm of a dermal dendritic cell of the patient S.J. Golgi apparatus (Go), smooth tubules and vesicles and dense bodies of various size and shape are shown. Three dense bodies have a clear halo between the limiting membrane and the dense content (arrows). The arrow heads point to the cell mern brane. x.55 000. homogeneous, moderately to intensely electron-dense bodies (presumably primary lysosomes) , some multivesicular bodies, occasional secondary lysosomes and residual bodies, and many filaments of thin and intermediate types (Figures 2 and 3). Birbeck granules were not seen in these cells, despite a careful search. The surface of these dendritic cells was scalloped with numerous vesicles, some coated. Most of these vesicles were located on the cell surface facing lymphocytes (Figure 3). No virus-like particle was identified with certainty in the dendritic cells. In all patients dendritic cells were regularly interspersed with lymphocytes; a small area containing lymphocytes, but no dendritic cells, was found in one patient (B.F.) in the deeper layer of the infiltrate. Most Iymphocytes made contact with dendritic cells in the plane of the section (Table 1). The percentage of lymphocytes with a nuclear contour index higher than 6.5 was slightly greater among those not in contact with dendritic cells than among those in contact with dendritic cells (Table 2). This difference was not evaluated statistically because of the very small number of lymphocytes not in contact with dendritic cells. The dermal infiltrate was mainly arranged around blood capillaries and venules. Cells with nuclear and cytoplasmic features resembling dendritic cells, oval in shape and with microvilli on their surface, were also found inside blood vessels (Figure 4). zyxwv zy Dendritic cells in Stzary syndrome 31 zyxwvutsrq zyxwvutsr Figure 3. Peripheral cytoplasm of a dermal dendritic cell of the patient S.V., with multiple contact areas with a lymphocyte (Ly). Many vesicles, two of which are coated (arrow-heads), are close to, or open on, the plasma membrane of the dendritic cell. Besides electron dense bodies and rough endoplasmic reticulum, bundles of intermediate filaments (arrows) are seen in the cytoplasm of the dendritic cell. X30000 zyxwvutsr Figure 4. This cell, which was within a dermal blood vessel, shows the ultrastructural features of a dendritic cell with some microvilli on its surface. ~ 1 2 0 0 0 32 zyxwvutsrq zyxwvu zyxwvutsr zyxwvutsr P. Romagnoli et al. Table 2. Lymphocyte and dendritic cell contact; relationship to nuclear contour index ~~ Percentage of lymphocytes contacting dendritic cells Percentage of lymphocytes not contacting dendritic cells Nuclear contour index Nuclear contour index <5 5 to 6.5 >6.5 <5 5 to 6.5 >6.5 S.V. S.J. B.B. B.F. 21.6 25.8 18.8 28.6 45.1 32.3 56.2 33.3 33.3 41.9 25.0 38.1 33.3 25.0 12.5 20.0 22.2 37.5 25.0 28.9 44.5 31.5 62.5 51.1 Mean 23.7 41.7 34.6 22.7 28.4 48.9 Discussion This paper provides evidence, for the first time supported by quantitative analysis of ultrathin sections, that dendritic cells are interspersed with lymphocytes in the dermal infiltrate of SS and that most infiltrating lymphocytes, either with or without cerebriform nuclei, are in contact with dendritic cells. The ultrastructural features of the cell surface nucleus and cytoplasm, as well as the contacts with the infiltrating lymphocytes, indicate that the dendritic cells found in the dermis of our patients are very similar to the dendritic cells found in the T-cell areas of lymph nodes and spleen and to those purified from blood and lymph (Nussenzweig & Steinman 1980, Balfour et al. 1981, Hoefsmith et al. 1982, Van Voorhis et al. 1982, Knight et al. 1983). The shape of the dendritic cells, however, appeared much more irregular in the dermis than in the blood. This finding agrees with what is known about dendritic cells in other conditions (Hoefsmith et al. 1982). Their conspicuous cytoskeleton, especially the great number of intermediate filaments, probably plays a major role in maintaining their dendritic shape in tissues. This is probably correlated with the function of these cells, allowing them to form a reticular framework for the lymphocytes. It is highly probable that clusters of lymphocytes not in contact with dendritic cells is a rare event in SS. In fact, only one lymphocyte cluster without dendritic cells was found among our patients (in Patient B.F.). In this patient too, however, dendritic cells were numerous (19% of the infiltrating cells) and were in contact with many lymphocytes (61% of the lymphocytes contacted a dendritic cell in the plane of the section). Most lymphocytes contacting dendritic cells had nuclear contour indexes higher than 5 , similar to those lymphocytes which were not seen in contact with a dendritic cell in the plane of the section. Therefore, on a morphological basis, the lymphocytes contacting dendritic cells had the same characteristics as those not in contact. Morever, 75% of the lymphocytes seen in an ultrathin section were in contact with dendritic cells. If we estimate that 1-2% of the whole cell surface is seen in the one plane then this suggests that probably all of the infiltrating zy zyxw zy zyxwvut Dendritic cells in Skzary syndrome 33 lymphocytes are in contact with at least one dendritic cell, thus supporting the concept of a functional interaction between these two cell types. This hypothesis is also supported by other morphological findings. Multiple areas of contact were often seen between each lymphocyte and a dendritic cell. Morever , vesicles on the dendritic cell plasmalemma were found almost exclusively on the cell surfaces facing lymphocytes. This finding suggests that exocytosis or endocytosis of molecules occurs allowing communication between dendritic cells and lymphocytes. At least some of these vesicles, especially the coated ones, may be involved in receptor-mediated endocytosis of messenger molecules adhering to the cell coat (Steinman et al. 1983). Coexistence of and contacts between lymphocytes and dendritic cells have been found in MF and in SS (Goos 1976, Van der Loo et al. 1979, Iwahara & Hashimoto 1984), as well as in inflammatory conditions, such as lichen planus (Giannotti et al. 1983) and contact dermatitis (De Panfilis et al. 1983). Therefore participation of this accessory cell type is probably required for the homing of lymphocytes in the dermis. Our data indicate that the presence of Birbeck granules may not be mandatory for dendritic cells to play this role in the dermis, since Birbeck granules were not detected in our material. Nevertheless, since serial sections were not studied, we cannot exclude the possibility that some dendritic cells did contain a small number of Birbeck granules. SCzary syndrome and MF are known to represent the most widely recognized members of the cutaneous T-cell lymphomas. From a clinical standpoint, SS is characterized by generalized exfoliative erythroderma, lymphadenopathy and peripheral blood involvement, whereas MF is characterized by multiple circumscribed erythematous lesions in its early stages and by plaques and tumours in the later stages (Slater 1984). A dermo-epidermal infiltrate, made up of lymphocytes with cerebriform nuclei (Lutzner cells) and dendritic cells, is characteristically found in both diseases. In SS, moreover, lymphocytes with cerebriform nuclei are present also in the peripheral blood; therefore, SS is generally considered to represent a leukaemic stage of MF (Slater 1984). The present study indicates that the dendritic cells found in the dermal infiltrates are not similar in SS and MF. In MF, some dendritic cells contain Birbeck granules (Jimbow et al. 1982), react with the OKT6 monoclonal antibody (Chu et al. 1982) and are thus identifiable as Langerhans cells (De Panfilis 1983); others do not contain Birbeck granules and are identified as interdigitating reticulum cells (Goos 1976, Goos, Kaiserling & Lennert 1976). Conversely, the dendritic cells in our cases of SS did not contain Birbeck granules, and only a very few cells reacting with OKT6 monoclonal antibody were found in the dermal infiltrate of one patient (S.V., data not shown). Therefore the dendritic cells in our cases resemble interdigitating reticulum cells rather than Langerhans cells. The significance of these differences remains unclear; the functional meaning of the Birbeck granules and the OKT6 defined antigen remains as yet unknown. However, these morphological and immunological features may prove useful as an additional diagnostic aid in doubtful cases. The presence in our patients of cells with ultrastructural features of dendritic cells in the blood vessels of the dermis suggests that these cells circulate through zyxwv 3 34 zyxwvutsrq P. Romagnoli et al. the skin, as has been shown in other conditions (Hoefsmith et af. 1982, Silberberg 1973). Their accumulation in the dermis, together with lymphocytes, suggests that possibly chemotactic factors are secreted by the cells infiltrating the skin. There is evidence that dendritic cells and lymphocytes cooperate in attracting and trapping each other in the dermis (Goos 1976, Edelson 1980, Chu et al. 1984). It is not known, however, whether the dermal infiltrate begins with dendritic cells or with lymphocytes. The skin biopsies of our patients were from fully developed lesions and were taken many years after the onset of symptoms. The close and constant association of dendritic cells with lymphocytes suggests that both cell types are necessary to develop and maintain the infiltrate, and lead to the hypothesis that factors affecting either of these cell types may influence the whole course of the disease. Coexistence of, and intercellular contacts between, dendritic cells and lymphocytes have been detected in the dermal infiltrates of a variety of non-neoplastic diseases of certain or probable immunological origin such as contact dermatitis and lichen planus. Similar contacts occur also in the paracortical T-dependent areas of lymph nodes and in the thymic medulla (Janossy et al. 1980). These similarities support the hypothesis that SS is a disease which results from chronic immune stimulation (Tan et al. 1974, Slater 1984). In other sites, following chronic immune stimulation, the sequence of a diffuse lymphocytic infiltration with eventual onset of a true lymphoma is well known (Isaacson & Wright 1978, Otto et af. 1981). In the case of SS, a similar mechanism may also operate. Some authors have suggested that a chronic retroviral infection of dendritic cells may play a pathogenetic role in MF and SS, leading to chronic stimulation of T-lymphocytes. This hypothesis is supported by the ultrastructural finding of C-type virus-like particles in the epidermal dendritic cells of SS and MF (Van der Loo et al. 1979, MacKie 1981, Slater et al. 1983, Fullbrandt et al. 1983). However, reverse transcriptase activity has been assessed only in one patient with MF (Van der Loo et af. 1979); hybridization studies were not performed. The lOOnm particles with a central core that we found in dendritic cells in each of our patients resemble those previously identified as C-type virus-like (Van der Loo et al. 1979, Fullbrandt et al. 1983). However, on the basis of electron microscopy alone, we are reluctant to interpret these particles as viral; they may be primary lysosomes (Fullbrandt et al. 1983). Furthermore we did not detect integrated HTLV-1 retrovirus DNA-specific sequences in patient B .B. using a sensitive hybridization assay. However, in view of the finding of HTLV-like particles in MF and the possibility that there may be a large number of such viruses, they have still to be considered as a possible cause of SCzary syndrome. zyxwvu zyxwv zyxwvuts Acknowledgements The authors thank Professor R.Bondi (Institute of Pathological Anatomy and Histology, University of Florence, Italy), for his help in histological diagnosis; the technicians of the Institute of Histology and General Embryology (University of zy zy zyxwvu Dendritic cells in Skzary syndrome 35 Florence, Italy), for their skilful assistance; and Dr M.Manzari (Institute of General Pathology, University of Rome ‘La Sapienza’, Italy), for performing restriction analysis. References zyxwv zyxwv BALFOUR B.M., DREXHAGE H.A., KAMPERDIJK E.W.A. & HOEFSMITH E.C.M. (1981) Antigen presenting cells, including Langerhans cells, veiled cells and interdigitating cells. In Microenvironments in Haemopoietic and Lymphoid Differentiation, Ciba Foundation Symposium 84, pp. 281-301. Pitman Medical, London CHUA , , BERGER C.L., KUNGP. & EDELSON R.L. (1982) In situ identification of Langerhans cells in the dermal infiltrate of cutaneous T-cell lymphoma. Journal of the American Academy of Dermatology 6, 350-354 CHU A , , PATIERSON J., BERGERc., VODERHEID E. & EDELSONR. (1984) In situ study of T-cell subpopulations in cutaneous T-cell lymphoma. Cancer 54,2414-2422 DE PANFILISG., GIANNOTTI B., MANARA G.C., LOMBARDO-DOMENICI L., FERRARI C. & CAPPUGI P. (1983) Macrophage-T lymphocyte relationships in man’s contact allergic reactions. British Journal of Dermatology 109, 183-189 DEPANFILIS G. (1983) ATPase, antigens detected by monoclonal antibodies, ultrastructural characters: which is the best marker for Langerhans cells? Archives of Dermatological Research 275,407-411 EDELSON R.L. (1980) Round cells of the epidermis: clues from studies on neoplastic lymphocytes of cutaneous T-cell lymphoma. Journal of Investigative Dermatology 75,95-102 FATTOROSSI A . , DELPRETE G.F., GIANNOTTI B., MORETTI S. & PALERMO A. (1984) Heterogeneity of Sezary cells. Multiparameter approach in two cases. In Immunodermatology, pp. 119-121, ed. D. MacDonald, Butterworths, London FULLBRANDT U., MEISSNERK., LONINGTH. & JANNERM. (1983) A second look at intraepithelial Langerhans cells in mycosis fungoides and related disorders. Ultrastructural study with special reference to Langerhans granules and virus-like particles. Virchows Archives A. Pathological Anatomy and Histology 402,4740 GIANNOT~ B., DE PANFILISG., MANARA G.C. & ALLEGXAF. (1983) Macrophage-T lymphocyte interaction in lichen planus. Archives of Dermarological Research 275, 35-40 Goos M. (1976) T-cell specific microenvironment in mycosis fungoides. Archives of Dermatological Research 255,215-218 Goos M., KAISERLING E. & LENNERT K. (1976) Mycosis fungoides: model for T-lymphocyte homing to the skin? British Journal of Dermatology 94,221-222 HOEFSMITH E.C.M., DUIJVESTIJN A.M. & KAMPERDIJK E.W.A. (1982) Relation betweenLangerhans cells, veiled cells, and interdigitating cells. Immunobiology 161, 255-265 ISAACSON,P. & WRIGHTD.H. (1978) Intestinal lymphoma associated with malabsorption. Lancet 14, 67-70. IWAHARA K. & HASHIMOTO K. (1984) T-cell subsets and nuclear contour index of skin-infiltrating T-cells in cutaneous T-cell lymphoma. Cancer 54,44&446 JANOSSY G., TIDMAN N., SELBYW.S., THOMAS J.A., GRANGER S., KUNGP.C. & GOLDSTEIN G. (1980) Human T lymphocytes of inducer and suppressor type occupy different microenvironments. Nature 487,81-84 JIMBOW K., CHIBAM. & HORIKISHI T. (1982) Electron microscope identification of Langerhans cells in the dermal infiltrates of mycosis fungoides. Journal of Investigative Dermatology 78, 102-107 KNIGHTS.C., MERTINJ., STACKPOOLE A . & CLARKJ. (1983) Induction of immune responses in vivo with small numbers of veiled (dendritic) cells. Proceedings of the National Academy of Sciences, USA SO, 6032-6035 LUETJ.H. (1961) Improvements in epoxy resin embedding methods. Journal of Biophysical and Biochemical Cytology 9 (Suppl.), 409-414 zyxwvutsr zyxwvut 36 zyxwvutsrq zyxwvut zyxwvut z P. Rornagnoli et al. zyxwvutsr zyxw MACKIER.M. (1981) Initial event in mycosis fungoides of the skin is viral infection of epidermal Langerhans cells. Lancet 8,283-285 MANZARI V., FAZIO V.M., MARTINOTTI S., GRADILONE A., COLLALTI E., FATIOROSSIA , , STOPPACIARO A , , VERANIP. & F u n L. (1984) Human T-cell leukaemidlymphoma virus (HTLV-I) DNA: detection in Italy in a lymphoma and in a Kaposi sarcoma patient. International Journal Cancer 34, 891-892 NUSSENZWEIG M.C. & STEINMAN R.M. (1980) Contribution of dendritic cells to stimulation of the murine syngenic mixed leukocyte reaction. Journal of Experimental Medicine 151, 11961212 OTTO H.F., BETTMANN I . , WELTZIENJ.V. & GEBBERSJ.O. (1981) Primary intestinal lymphomas. Virchows Archives A . Pathological Anatomy and Histology 391, 9-31 ROMACNANI S . , DELPRETEG.F., MAGCIE . , CADONIA., CORTEG., GHEZZIG., MORETTI S. & Ricci M. (1982) Phenotipic and functional characterization of a Sezary cell. Journal of Clinical Immunology 2,343-349 G., PHILLIPS T.M. & LEWISM.G. (1979) Ia antigens on indeterminate cells of the epidermis. ROWDEN Immunoelectronmicroscopic studies of surface antigens. British Journal of Dermatology 100, 531-542 SILBERBERG I . (1973) Apposition of mononuclear cells to Langerhans cells in contact allergic reactions. Acta Dermato-Venereologica 53, 1-12 SLATER D., BLEEHEN S . , ROONEY N. &r HAMED A . (1983). Type C retrovirus-like particles in mycosis fungoides. British Journal of Dermatology 109, 120-121 SLATER D.N. (1984) Lyrnphoproliferative conditions of the skin. In Recent Advances in Histopathology, vol. 12, pp. 83-1 10, eds P.P.Anthony & R.N.M.MacSween, Churchill Livingstone, Edinburgh STEINMAN R.M., MELLMAN I.S., MULLER W.A. & COHNZ.A. (1983) Endocytosis and the recycling of plasma membrane. Journal of Cell Biology 96, 1-27 STOLZW., SCHMOECKEL C., BURGG. & BRAUN-FALCO 0. (1983) Circulating Stzary cells in the diagnosis of SCzary syndrome (Quantitative and morphometric analysis). Journal of Investigative Dermatology 81, 314-319 TANR.S., BUTTERWORTH C.M., MCLAUGHIN H., MALKAS. & SAMMAN P.D. (1974) Mycosis fungoides, a disease of antigen persistence. British Journal of Dermatology 91, 607-616 VANDER Loo E.M., VANMUIJENG.N.P., VANVLOTENW.A., BEENSW., SCHEFFER E. & MEIJER C.J.L.M. (1979) C-type virus-like particles specifically localized in Langerhans cells and related cells of skin and lymph nodes of patients with mycosis fungoides and SCzary syndrome. Virchows Archiv B. Cell Pathology 31, 193-203 VANVOORHIS W.C., HAIRL., STEINMAN R.M. & KAPLAN G. (1982) Human dendritic cells. Enrichment and characterization from peripheral blood. Journal of Experimental Medicine 155, 1172-1 187 VAN VOORHIS W.C., WITMERM.D. & STEINMAN R.M. (1983) The phenotype of dendritic cells and macrophages. Federation Proccedings 42, 31 14-31 18 WEIBEL E.R. (1979) Stereological Methods Vol. I , Academic Press, London