Portal de Periódicos da CAPES

Aberrant Expression of Adhesion Molecules by Sézary Cells: Functional Consequences Under Physiologic Shear Stress Conditions

2001; Elsevier BV; Volume: 116; Issue: 3 Linguagem: Inglês

10.1046/j.1523-1747.2001.01282.x

1523-1747

Sam T. Hwang, David J. Fitzhugh,

T-cell and B-cell Immunology

Although aberrations in adhesion molecule expression by lymphoma cells have been reported, the functional consequences of these changes are unclear. Herein, we report a patient with Sézary syndrome whose malignant peripheral blood T cells were TCRVβ17+. Malignant T cell adhesion molecule abnormalities included an 80% downregulation of LFA-1 compared with normal controls and no detectable expression of α4 integrin. Under shear stress conditions, malignant T cells failed to arrest on recombinant ICAM-1 in the presence of chemokines and displayed an 80% decrease in the ability to arrest on TNF-α activated dermal microvascular endothelial cells compared with normal CD4+ memory T cells. Cutaneous lymphocyte-associated antigen expression was detected in ∼25% of malignant T cells in the peripheral blood, but was substantially less than this in TCRVβ17+ T cells in the dermis. By contrast, > 95% of malignant T cells in peripheral blood expressed L-selectin (CD62L), and L-selectin ligand was detected in dermal blood vessels at affected skin sites. Compared with normal CD4+, malignant T cells attached and rolled 6-fold more efficiently on L-selectin ligand (p 95% of malignant T cells in peripheral blood expressed L-selectin (CD62L), and L-selectin ligand was detected in dermal blood vessels at affected skin sites. Compared with normal CD4+, malignant T cells attached and rolled 6-fold more efficiently on L-selectin ligand (p 50% of her body surface. A skin biopsy showed a dense dermal infiltrate consisting of large T cells with cerebriform nuclei. Minimal epidermotropism and spongiosis were present in the epidermis. An 18-fluorodeoxyglucose Positron Emission Tomography scan showed uptake in the axillary and inguinal nodes but no uptake in the skin. A lymph node (LN) biopsy from 1994 was positive for malignant T cell involvement. A complete blood count showed an elevated white count of 14.8 × 103 per mm3 consisting of 36% neutrophils, 2% lymphocytes, and 57% small-medium sized cells with high nuclear content and Sézary cell morphology. Flow cytometric analysis showed that her T cells were >93% CD45RO+, CD4+, and CD7+. Patient-derived tissue and blood were collected with consent per protocol (00-C-0068) while she had been off systemic therapy for at least 2 mo. The patient subsequently underwent experimental therapy at NIH with clearing of many skin lesions and decreases in her Sézary cell count and was unavailable for further experiments. FITC-anti-human TCRVβ-specific monoclonal antibodies (MoAb) and anti-CD29 (mIgG1) were purchased from Endogen (Boston, MA). Anti-human CD45RO (UCHL1, mIgG2a), CD45RA (HI100, mIgG2b), anti-PNAd/L-selectin ligands (MECA-79, rat IgM), anti-CLA (HECA452, rat IgM), PE-anti-CD49d (9F10, mIgG1), FITC-anti-CD11a (HI10a, mIgG1), FITC-anti-CD18 (6.7, mIgG1), and unlabeled and CyChrome-antihuman CD62L (Dreg56, mIgG1) were purchased from BD-Pharmingen (San Diego, CA). E-selectin/human Ig chimera (Erbe et al., 1992Erbe D.V. Wolitzky B.A. Presta L.G. et al.Identification of an E-selectin region critical for carbohydrate recognition and cell adhesion.J Cell Biol. 1992; 119: 215-227Crossref PubMed Scopus (170) Google Scholar) and purified GlyCAM-1 were gifts of S.D. Rosen (UCSF, San Francisco, CA). A nonfunction blocking anti-human ICAM-1 MoAb (P79) and soluble human ICAM-1 (sICAM-1) were gifts of T.K. Kishimoto (Boehringer-Ingelheim Pharmaceuticals, Ridgefield, CT). Recombinant chemokines were purchased from Peprotech (Rocky Hill, NJ). Punch biopsies of skin (4 mm) were obtained from the patient and frozen in OCT (Sakura Finetek, Torrance, CA). Acetone-fixed sections (6 μm) were blocked with phosphate-buffered saline (PBS)/2% fetal calf serum prior to staining with unlabeled antibodies (i.e., anti-CLA or anti-PNAd) at 5 μg per ml for 30 min at room temperature. After washing, a biotinylated mouse anti-rat IgM secondary MoAb followed by Cy3-strepavidin (Caltag, San Bruno, CA) along with FITC-labeled Vβ specific MoAb (1:100 dilution) was added to the sections. Fluorescence was visualized with a Nikon PCM200 confocal microscopy. Fresh peripheral blood was centrifuged over Histopaque 1007 (Sigma, St. Louis, MO), and the mononuclear fraction was cultured overnight in complete RPMI 1640 medium including 10% fetal calf serum, and antibiotics/antifungal agents (cRPMI). CD4+ CD45RO+ mTC (> 95% pure) were isolated by negative depletion with CD19, CD14, and CD8-conjugated Dynal beads (Dynal, Lake Success, NY) followed by further depletion with anti-human CD45RA/sheep anti-mouse-conjugated Dynal beads. T cells were stained with the indicated antibodies and washed in PBS before resuspension in PBS/0.1% bovine serum albumin (BSA) for flow cytometric analysis. All flow cytometric data were confirmed by a second experiment at a later time using another normal volunteer as a control. Human mTC populations were isolated as described above. Where indicated, mTC were treated with pertussis toxin (PTX, 100 ng per ml, Sigma, St. Louis, MO) for 1.5–2 h at 37°C. mTC were allowed to equilibrate at 37°C for 30 min before the flow assay and then suspended in a 12 ml syringe, which was fixed to a precision syringe pump (Harvard Apparatus, Holliston, MA). For flow assays using recombinant proteins, E-selectin/human Ig, chemokines, and soluble ICAM-1 were applied to 35 mm nontreated plastic culture dishes as described (Fitzhugh et al., 2000Fitzhugh D.J. Naik S. Caughman S.W. Hwang S.T. Cutting edge: CC Chemokine Receptor-6 (CCR6) is essential for arrest of a subset of memory T cells on activated dermal microvascular endothelial cells under physiologic flow conditions in vitro.J Immunol. 2000; 165: 6677-6681Crossref PubMed Scopus (99) Google Scholar). To test rolling of T cells on L-selectin ligand, affinity-purified murine GlyCAM-1 (Singer and Rosen, 1996Singer M.S. Rosen S.D. Purification and quantification of L-selectin reactive GlyCAM-1 from mouse serum.J Immunol Meth. 1996; 196: 153-161Crossref PubMed Scopus (24) Google Scholar) (a soluble L-selectin ligand secreted into serum by LN high endothelial cells) was used to coat plastic plates at ∼1 μg per ml in tris-buffered saline pH 9.0 overnight at 4°C. No chemokines were added and PBS/0.1% BSA was used to block nonspecific binding. A parallel plate flow chamber apparatus (Glycotech, Rockville, MD) was affixed to the tissue culture plate using a rubber gasket and a continuous vacuum. For T cell attachment experiments using endothelial cells, HDMEC were grown, stimulated with TNF-α, and used in adherence experiments as described (Fitzhugh et al., 2000Fitzhugh D.J. Naik S. Caughman S.W. Hwang S.T. Cutting edge: CC Chemokine Receptor-6 (CCR6) is essential for arrest of a subset of memory T cells on activated dermal microvascular endothelial cells under physiologic flow conditions in vitro.J Immunol. 2000; 165: 6677-6681Crossref PubMed Scopus (99) Google Scholar). In all experiments, T cells (5 × 105 cells per ml in cRPMI) were pumped into the chamber at a shear stress of 1.5 dynes per cm2 at room temperature. After allowing 5 min for interactions to establish, randomly selected fields were videotaped and subjected to digital video analysis of 10 s segments to quantify numbers of both rolling and arrested cells during the 10 s time frame. Results are expressed as the mean ratio of adherent:rolling cells (10 fields were analyzed per condition). Statistical significance (Student's t test) was calculated using Microsoft Excel. We screened the patient's peripheral blood lymphocytes for expression of specific T cell receptor β chain variable (Vβ) regions using MoAb. Given the large number of possible variations in the β chain of the TCR, < 8% of lymphocytes in peripheral blood in normal individuals should express any one Vβ epitope. Due to clonal expansion in SS, however, 59%-87% of all peripheral blood lymphocytes may express a single Vβ epitope (Heald et al., 1994Heald P. Yan S.L. Edelson R. Profound deficiency in normal circulating T cells in erythrodermic cutaneous T-cell lymphoma.Arch Derm. 1994; 130: 198-203Crossref PubMed Scopus (83) Google Scholar). Strikingly, >95% of the patient's peripheral blood lymphocytes (as defined by forward and size scatter parameters and CD3 expression) displayed the Vβ17 epitope (Figure 1a). Other Vβ epitopes tested (Vβ 3.1, 5a, 6.7, 8a, 13 (not shown), and 12 (Figure 1a)) represented <5% of the patient's lymphocytes. In order to determine whether expression of cell surface adhesion molecules were altered in this patient, we stained her lymphocytes with specific MoAb and quantified expression by flow cytometry. CLA was expressed on ∼25% of the patient's lymphocytes (Figure 1b). As nearly all the patient's lymphocytes expressed Vβ17, this also meant that approximately 25% of her malignant T cells also expressed CLA. In contrast to the partial expression seen with CLA, L-selectin was expressed by ∼95% of her malignant T cells (Figure 1b). As shown in Figure 1(c), the Sézary cells were deficient in CD18 and CD11a – the two integrin chains that make up LFA-1, an adhesion molecule thought to be critical for T cell adhesion to endothelial cells (Oppenheimer-Marks et al., 1990Oppenheimer-Marks N. Davis L.S. Lipsky P.E. Human T lymphocyte adhesion to endothelial cells and transendothelial migration.J Immunol. 1990; 145: 140-148PubMed Google Scholar). CD18 and CD11a were reduced by ∼80% when compared with CD4+ mTC from two different control donors. Moreover, the malignant T cells did not express the α4 integrin chain although β1 integrin was normally expressed on the cell surface (Figure 1c). Thus, malignant mTC in this patient demonstrated aberrant adhesion receptor expression, namely upregulation of L-selectin and downregulation of LFA-1 and α4 integrins. To address whether aberrant LFA-1 expression might lead to functional defects under physiologic flow conditions, we isolated patient and control CD4+ mTC for use in parallel plate flow chamber experiments in which the surface was coated with E-selectin, ICAM-1, and chemokines. In our experience, stromal-derived factor (SDF)-1 is the most effective chemokine inducer of arrest for mTC (DF and STH, unpublished data). The patient's lymphocytes attached and rolled on this surface but, in contrast to normal CD4+ mTC, failed to firmly arrest in the presence of SDF-1 as indicated by an arrest-to-rolling ratio that was far lower that than observed with a control donor (Figure 2c). We tested the patient's lymphocytes with two other chemokines, liver and activation-regulated chemokine (LARC)/MIP-3α, which acts via CCR6, and thymus and activation-regulated chemokine (TARC), which acts via CCR4, to determine whether this defect was specific for CXCR4. These two chemokines have been postulated to play a role in the specific homing of mTC to skin (Campbell et al., 1999Campbell J.J. Haraldsen G. Pan J. et al.The chemokine receptor CCR 4 in vascular recognition by cutaneous but not intestinal memory T cells.Nature. 1999; 400: 776-780Crossref PubMed Scopus (717) Google Scholar;Homey et al., 2000Homey B. Dieu-Nosjean M. Wiesenborn A. et al.Up-regulation of macrophage inflammatory protein-3a/CCL20 CC chemokine receptor, 6 in psoriasis.J Immunol. 2000; 164: 6621-6632Crossref PubMed Scopus (437) Google Scholar). As was the case with SDF-1, TARC (Figure 2a) and LARC (Figure 2b) were able to induce arrest in the control, but not patient, lymphocytes. In filter-based chemotaxis assays, both TARC and SDF-1 were able to strongly stimulate chemotaxis of patient lymphocytes (STH, unpublished data), suggesting that a generalized defect in the patient's chemokine receptors was not present. To determine whether this defect in firm arrest could be observed in a more physiologic system where other adhesion molecules besides ICAM-1 were available for arrest, we tested the ability of the patient's lymphocytes to roll and arrest on TNF-α-activated HDMEC. We had previously shown that in this system human memory T cell interaction was dependent upon E-selectin, CLA, and pertussis toxin (Figure 2d) as previously described (Fitzhugh et al., 2000Fitzhugh D.J. Naik S. Caughman S.W. Hwang S.T. Cutting edge: CC Chemokine Receptor-6 (CCR6) is essential for arrest of a subset of memory T cells on activated dermal microvascular endothelial cells under physiologic flow conditions in vitro.J Immunol. 2000; 165: 6677-6681Crossref PubMed Scopus (99) Google Scholar). The patient's lymphocytes rolled, but inefficiently arrested, on activated HDMEC (Figure 2d). Analysis of the mean ratio of adherent to rolling cells showed that her cells arrested only 20% as efficiently as normal memory CD4+ T cells. The patient's malignant T cells expressed 2-fold more L-selectin than L-selectin+CD4+ mTC from normal controls (Figure 1b). To determine whether the L-selectin expressed by CTCL cells was functional, these cells were applied under shear stress to plastic plates coated with an affinity-purified L-selectin ligand termed GlyCAM-1, a bonafide rolling substrate for lymphocytes (Dwir et al., 1998Dwir O. Shimron F. Chen C. Singer M.S. Rosen S.D. Alon R. GlyCAM-1 supports leukocyte rolling in flow: evidence for a greater dynamic stability of L-selectin rolling of lymphocytes than of neutrophils.Cell Adhes Commun. 1998; 6: 349-370Crossref PubMed Scopus (26) Google Scholar). CTCL cells showed a ∼6-fold increase in the number of cells rolling per field compared with CD4+ mTC at 1.5 dynes per cm2 (Figure 2e), suggesting that these cells may attach more efficiently to L-selectin ligands than normal CD4+ mTC. The specific reactivity of the malignant T cells with anti-Vβ17 MoAb provided a tool by which we could unequivocally identify malignant T cells in the skin. Using confocal fluorescence microscopy, double-staining for the mTC marker, CD45RO, revealed that nearly all of the Vβ17+ malignant T cells were of the memory phenotype (not shown). While we observed scattered CLA+ cells in the epidermis and in the papillary and reticular dermis, these CLA+ cells did not generally correspond with cells that had Vβ17 reactivity, suggesting that most of the malignant T cells were not CLA+ (data not shown). To determine which adhesion molecules were expressed by vascular endothelial cells in this patient's skin, we stained sections of skin with MECA-79 (Michie et al., 1993Michie S.A. Streeter P.R. Bolt P.A. Butcher E.C. Picker L.J. The human peripheral lymph node vascular addressin. an inducible endothelial antigen involved in lymphocyte homing.A J Path. 1993; 143: 1688-1698PubMed Google Scholar), which recognizes carbohydrate-based L-selectin ligands (i.e., PNAd) in both mouse and man. Indeed, many of the blood vessels in the deep reticular dermis stained positively for L-selectin ligands (red staining, Figure 3) and were surrounded by dense collections of Vβ17+ malignant T cells (green cells, Figure 3). Some of these vessels (white box highlighted in Figure 3) displayed morphology similar to the L-selectin ligand-rich high endothelial venules (HEV) of secondary lymphoid organs. Staining with isotype-matched rat IgM control MoAb was negative. Occasional blood vessels also stained positively for E-selectin (not shown). Sections were also stained for L-selectin, but reactivity was not observed (data not shown). A significant fraction of non-Hodgkin's lymphomas (in particular B cell lymphomas) express little or no LFA-1, but alterations in the levels of LFA-1 in SS/CTCL have not been noted in the limited number (13) of patients studied to date (Medeiros et al., 1989Medeiros L.J. Weiss L.M. Picker L.J. Clayberger C. Horning S.J. Krensky A.M. Warnke R.A. Expression of LFA-1 in non-Hodgkin's lymphoma.Cancer. 1989; 63: 255-259Crossref PubMed Scopus (44) Google Scholar;Savoia et al., 1992Savoia P. Novelli M. Fierro M.T. Cremona O. Marchisio P.C. Bernengo M.G. Expression and role of integrin receptors in Sézary syndrome.J Invest Dermatol. 1992; 99: 151-159Crossref PubMed Scopus (12) Google Scholar). Thus, we provide the first case of aberrant LFA-1 expression by malignant T cells in CTCL and demonstrate that this leads to an adhesion deficiency in vitro under shear stress conditions. How can the apparent defect in T cell adherence to activated endothelium be reconciled with the large numbers of malignant T cells in the skin and subsequent clinical severity? Although the rate of adhesion to activated endothelium was reduced in vitro to ∼20% of normal, even a relatively low efficiency of adherence may lead to clinically meaningful numbers of T cells migrating from the blood stream into skin given a sufficient rate of initial interactions with activated endothelial cells. We have noted two features in this patient that might increase initial interactions: (1) uniform expression of L-selectin on malignant T cells, and (2) abundant expression of L-selectin ligand on dermal endothelial cells. Compared with CLA expression of only ∼25% of the CTCL cells, uniform expression of L-selectin may theoretically increase initial rolling interactions with dermal endothelium by up to 4-fold and yield a higher number of adherent cells than would be found by E–selectin interactions alone. Indeed, our in vitro flow studies (Figure 2e) demonstrated a striking 6-fold increase in tethering and rolling of the CTCL cells versus normal CD4+ mTC on L-selectin ligand. Just as we had detected MECA-79 reactive vessels in the skin of this patient, L-selectin ligands have been reported to be expressed by vascular endothelial cells in chronic inflammatory dermatoses (Mackay et al., 1992Mackay C.R. Marston W. Dudler L. Altered patterns of T cell migration through lymph nodes and skin following antigen challenge.Eur J Immunol. 1992; 22: 2205-2210Crossref PubMed Scopus (123) Google Scholar;Michie et al., 1993Michie S.A. Streeter P.R. Bolt P.A. Butcher E.C. Picker L.J. The human peripheral lymph node vascular addressin. an inducible endothelial antigen involved in lymphocyte homing.A J Path. 1993; 143: 1688-1698PubMed Google Scholar) and in CTCL (Lechleitner et al., 1999Lechleitner S. Kunstfeld R. Messeritsch-Fanta C. Wolff K. Petzelbauer P. Peripheral lymph node addressins are expressed on skin endothelial cells.J Invest Dermatol. 1999; 113: 410-414Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar), where it has been proposed to influence migration of T cells. Therefore, it is possible that malignant T cells in this case use the L-selectin pathway to increase initial interactions with PNAd+ vessels in order to increase numbers of T cells that arrest and extravasate. Additionally, upregulation of L-selectin may have also contributed to LN involvement in this patient. Our ability to tag malignant T cells with a specific marker (Vβ17) has allowed us to determine that the majority of malignant T cells in the skin of our patient were not CLA+. If CLA expression were strictly required for T cell migration into skin, one might expect that nearly all T cells in skin would be CLA+ as has been reported in other skin conditions (Borowitz et al., 1993Borowitz M.J. Weidner A. Olsen E.A. Picker L.J. Abnormalities of circulating T-cell subpopulations in patients with cutaneous T-cell lymphoma: cutaneous lymphocyte-associated antigen expression on T cells correlates with extent of disease.Leukemia. 1993; 7: 859-863PubMed Google Scholar). Thus, L-selectin/L-selectin ligand interactions may play a major role in the homing of T cells to skin independent of CLA under certain conditions. These may include chronic inflammatory states since several experimental models suggest that L-selectin ligands do not appear (Mackay et al., 1992Mackay C.R. Marston W. Dudler L. Altered patterns of T cell migration through lymph nodes and skin following antigen challenge.Eur J Immunol. 1992; 22: 2205-2210Crossref PubMed Scopus (123) Google Scholar) or are not critical in the first 3 d of an immune response (Tang et al., 1997Tang M.L. Hale L.P. Steeber D.A. Tedder T.F. L-selectin is involved in lymphocyte migration to sites of inflammation in the skin. delayed rejection of allografts in L-selectin-deficient mice.J Immunol. 1997; 158: 5191-5199PubMed Google Scholar). We did not observe L-selectin expression by the patient's malignant T cells in affected skin tissue. This was not completely surprising in that L-selectin is proteolytically cleaved (shed) from the surface of leukcocytes (including lymphocytes) during rolling interactions (Walcheck et al., 1996Walcheck B. Kahn J. Fisher J.M. et al.Neutrophil rolling altered by inhibition of L-selectin shedding in vitro.Nature. 1996; 380: 720-723Crossref PubMed Scopus (256) Google Scholar) and under a variety of activating conditions (Buhrer et al., 1990Buhrer C. Berlin C. Thiele H.G. Hamann A. Lymphocyte activation and expression of the human leucocyte-endothelial cell adhesion molecule 1 (Leu-8/TQ1 antigen).Immunology. 1990; 71: 442-448PubMed Google Scholar;Palecanda et al., 1992Palecanda A. Walcheck B. Bishop D.K. Rapid activation-independent shedding of leukocyte L-selectin induced by cross-linking of the surface antigen.Eur J Immunol. 1992; 22: 1279-1286Crossref PubMed Scopus (121) Google Scholar). LFA-1 has been used as a target for the therapy of lymphoma (Zahalka et al., 1993Zahalka M.A. Okon E. Naor D. Blocking lymphoma invasiveness with a monoclonal antibody directed against the beta-chain of the leukocyte adhesion molecule (CD18).J Immunol. 1993; 150: 4466-4477PubMed Google Scholar) and T cell-mediated inflammatory skin diseases such as psoriasis (Gottlieb et al., 2000Gottlieb A. Krueger J.G. Bright R. et al.Effects of administration of a single dose of a humanized monoclonal antibody to CD11a on the immunobiology and clinical activity of psoriasis.J Am Acad Dermatol. 2000; 42: 428-435Abstract Full Text Full Text PDF PubMed Scopus (204) Google Scholar). Although these reports are promising, our data suggest that significant clinical disease can occur even when LFA-1 expression is only 20% of normal – a level at which functional impairment in vitro was clearly demonstrated in this patient. Altered expression of adhesion receptors may provide malignant T cells with the capability to migrate to tissue despite impairment of certain key elements of migration found in normal T cells. The success of future therapies directed against CD11a/CD18 may also depend upon the degree to which other integrin-dependent functions such as T cell activation or cytotoxic function are also inhibited. We wish to thank Drs Mark Udey, John Morris, John Janik, and Richard Piekarz for helpful comments, Mr Erik Gonzalez for technical assistance, and Drs S. Wright Caughman and Shubhada Naik (Department of Dermatology, Emory University School of Medicine) for kindly providing the HDMEC cells.