Int. J. Cancer: 71, 178–182 (1997) r 1997 Wiley-Liss, Inc. Publication of the International Union Against Cancer Publication de l’Union Internationale Contre le Cancer IDENTIFICATION OF AN ENRICHED CD41 CD8a11 CD8b1 T-CELL SUBSET AMONG TUMOR-INFILTRATING LYMPHOCYTES IN HUMAN RENAL CELL CARCINOMA Ludwig E. VAN DEN HOVE1, Stefaan W. VAN GOOL1,2, Hein VAN POPPEL3, Luc BAERT3, Lieve COOREVITS1, Boudewijn VAN DAMME4, Paola DAL CIN5, Herman VAN DEN BERGHE5 and Jan L. CEUPPENS1* 1Laboratory of Experimental Immunology, Faculty of Medicine, Catholic University of Leuven, Leuven, Belgium 2Division of Pediatrics, Faculty of Medicine, Catholic University of Leuven, Leuven, Belgium 3Division of Urology, Faculty of Medicine, Catholic University of Leuven, Leuven, Belgium 4Division of Pathology, Faculty of Medicine, Catholic University of Leuven, Leuven, Belgium 5Center for Human Genetics, Faculty of Medicine, Catholic University of Leuven, Leuven, Belgium Three-color immunofluorescence flow-cytometric analysis of freshly isolated tumor-infiltrating lymphocytes (TIL) from patients with primary renal cell carcinoma (RCC) revealed a unique, not previously described TIL subset with a CD31 CD41 CD8a11 CD8b1 phenotype. This subset represented at least 5% of CD31 TIL in 15 of 21 patients with clear cell RCC, whereas it was not or only marginally represented in patients with papillary RCC or sarcomatoid RCC. In one-third of the patients with clear cell RCC, more than 20% of CD31 TIL and in one patient more than half of the CD31 TIL displayed this phenotype. The occurrence of this subset was not associated with pathological stage, tumor diameter, nuclear grade, cytogenetic abnormalities or vascular invasion in this patient cohort. When present, the CD31 CD41 CD8a11 CD8b1 subset was detected in similar proportions in tumor tissue and tumor capsula, and it was also detected in adjacent non-tumoral renal tissue, albeit in much lower proportions. Despite strong cell surface expression of various activation markers (CD69, CD54 and HLA-DR), CD31 CD41 CD8a11 CD8b1 cells displayed no ex vivo cytolytic activity in an anti-CD3-redirected cytotoxicity assay. In contrast with CD31 CD41 CD82 cells from the same tumor sample, they were markedly deficient in IL-2Ra up-regulation following antiCD3 triggering. The possibility that these cells represent either anergic cells or a highly specialized effector population with a discrete, as yet undescribed function is discussed. Int. J. Cancer 71:178–182, 1997. r 1997 Wiley-Liss, Inc. Renal cell carcinoma (RCC) is one example of human cancer in which the immune system is thought to play a role in the control of malignant cell growth. The basis for this belief originated from observations that metastases may regress following nephrectomy, which has been interpreted to be the consequence of an immune response (reviewed in de Riese et al., 1991). Furthermore, about 15–20% of patients with advanced RCC respond to systemic interleukin-2 (IL-2) administration, and durable (.60 months) responses with complete disappearance of all measurable disease have been achieved in some patients with metastatic disease (Whittington and Faulds, 1993). Finally, RCC tumors are infiltrated by a large mononuclear infiltrate mainly composed of T lymphocytes (tumor-infiltrating lymphocytes, TIL), which have been considered to be part of the host immune response against the tumor and to be enriched in T cells with tumor specificity (Mitropoulos et al., 1994; Gaudin et al., 1995). The observations of defects in CD3-z-chain expression in CD31 TIL of some patients (Finke et al., 1993; Farace et al., 1994; Tartour et al., 1995), the defective expression of IL-2Ra (CD25) on in vitro stimulated CD31 TIL of some patients (Alexander et al., 1993), the heterogeneity in TIL proliferative responses (Alexander et al., 1993; Tartour et al., 1995), the failure to grow TIL from some tumor samples (Whiteside and Parmiani, 1994) and the variable results with TIL therapy in clinical trials (Pierce et al., 1995) suggest that at least some TIL subsets are functionally impaired. In the present report, we describe a unique, previously uncharacterized CD31 TIL subset in human RCC that expresses CD4, CD8a and CD8b chains and that is markedly deficient in IL-2Ra expression following stimulation. MATERIAL AND METHODS Tissue specimens Peripheral blood and tumoral (tumor stroma and tumor capsula) and adjacent non-tumoral renal tissues were obtained from 27 patients during surgery for a primary RCC (Table I). Tumor specimens were examined for pathological tumor and nodal stage (TNM classification; Spiessl et al., 1992), largest tumor diameter, nuclear grade (I–IV according to Fuhrman et al., 1982), histologic cell type (clear cell, papillary, granular cell, chromophobe cell, sarcomatoid or collecting duct according to Murphy et al., 1994) and presence or absence of vascular invasion. The latter was demonstrated to be the single most significant prognostic factor after radical surgery (Van Poppel et al., 1995). Thymic tissue was obtained from a 29-week-old fetus during autopsy. Isolation of lymphocytes from renal and thymic tissue and from peripheral blood Single-cell suspensions from renal and thymic tissue specimens were prepared by mincing the tissues with scissors and scalpel into tiny pieces, which were passed through a metal mesh. Enrichment for viable lymphocytes was done by centrifugation of the singlecell suspensions or peripheral blood over a Lymphoprep gradient (Nycomed, Oslo, Norway). For purification of CD41 and CD81 TIL subsets and peripheral blood CD81 T cells, magnetic immunoselection on isolated lymphocytes with anti-CD4- and anti-CD8coated Dynabeads (Dynal, Oslo, Norway), respectively, was used. Beads were applied with a bead-to-target ratio of 3:1. Immunofluorescence studies Three-color analyses were performed by direct and combined indirect/direct immunofluorescence with FITC (fluorescein isothiocyanate)-, PE (phycoerythrin)- and PerCP (peridinin chlorophyll protein)-conjugated and unconjugated monoclonal antibodies (MAbs). PerCP-, PE- or FITC-labeled mouse MAbs anti-CD3, anti-CD4, anti-CD8a, anti-CD25, anti-CD28, anti-CD45, antiCD45RA, anti-CD54, anti-CD56, anti-CD57, anti-CD62L, antiCD69, anti-HLA-DR, anti-T-cell receptor a/b, anti-T-cell receptor g/d, the appropriate mouse immunoglobulin isotype controls and rat-anti-mouse IgG1 were all from Becton Dickinson (Mountain Contract grant sponsor: Nationaal Fonds voor Wetenschappelijk Onderzoek, contract grant number G.3123.94; Contract grant sponsor: Algemene Spaar, Lijfrentekas, Brussels, Belgium. *Correspondence to: Laboratory of Experimental Immunology, University Hospital Gasthuisberg, Herestraat 49, 3000 Leuven, Belgium. Fax: 32/16346035. Received 28 September 1996; accepted 27 December 1996 ‘‘DOUBLE-POSITIVE’’ TUMOR-INFILTRATING LYMPHOCYTES TABLE I – RELATIVE NUMBER OF Patient1 Sex2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 CD31CD41CD8a11 % CD31CD41CD8a11 lymphocytes 179 TIL AND HISTOPATHOLOGIC, CYTOGENETIC AND CLINICAL FINDINGS IN 27 PATIENTS WITH PRIMARY RCC pTN stage3 Maximal tumor diameter (cm) Nuclear grade4 Histologic cell type5 Venous invasion Follow-up6 (months) Clear Clear Clear Clear Clear Clear Clear Clear Clear Clear Clear Clear Clear Clear Clear Clear Clear Papillary No No No No Yes No No Yes Yes Yes No No No Yes No Yes No No NED (11) NED (18) NED (16) NED (18) NED (5) NED (16) NED (18) NED (13) NED (20) NED (14) NED (8) NED (8) NED (8) AMD (18) NED (12) NED (18) AMD (16) NED (16) Papillary No NED (18) Sarcomatoid Clear Clear Papillary Yes No No No CRD (14) NED (15) NED (16) NED (18) II II Clear Papillary Yes No NED (21) NED (18) IV III Clear Papillary Yes No CRD (4) NED (7) Within CD31 TIL Within total TIL M F M M M M M M M M M F M M F M M M 54 32 32 30 25 24 22 22 16 14 13 6 6 5 5 3 3 2 31 30 27 25 20 17 14 13 12 10 7 4 4 3 1 2 2 2 pT2N0 pT2N0 pT2N0 pT2N0 pT2N0 pT2N0 pT2N0 pT2N0 pT3bN0 pT3bN0 pT2N0 pT3aN0 pT3aN0 pT2N0 pT2N0 pT3bN0 pT3aN1 pT2N0 3.5 6 6.5 6 8 3 4 10 10 7.5 6 7.5 5 8 9 13.5 5 6 I III IV II III II III III II II II III III III III III III II 19 M 2 1 pT2N0 5 II 20 21 22 23 M M F M 2 1 1 1 1 1 1 1 pT2Nx pT2N0 pT2N0 pT2N0 4 10.5 3 5 III III II III 24 25 M M 0 0 0 0 pT3aN1 pT2N0 8 5 26 27 F F 0 0 0 0 pT4N1 pT2N0 14 3.5 Tumor karyotype7 [number of cells] 46,XY  46,XX  46,XY  Failure 45,X,2Y,del(3)(p21p25)  50,XY,17,112,119,120  Failure Failure Failure Failure 46,XY  Not assessed 46,XY  Failure Not assessed Failure 46,XY  47,XY,17/46,X,2Y,17/46,XY [2/2/18] 48,X,2Y,17,17,der(11)t(3; 11)(q11;p15),117  Failure 46,XY  47,XX,17/46,XX [5/15] 90–103,XXY,2Y,21,der(9; 12)(q10; q10)32,1mar1,1mar2/ 46,XY [25/5] Not assessed 47,X,2Y,12,17,17,28,der(13;15) (q10;q10),215,der(15; 17)(q10;q10),116  46,XX,del(3)(p14p23)  46,XX  1The 27 patients were ordered according to the relative number of CD31CD41CD8a11 lymphocytes within total CD31 TIL.–2M, male; F, female.–3Pathological tumor and node stages were determined on the basis of criteria stipulated by the UICC (Spiessl et al., 1992).–4Nuclear grade was determined on the basis of criteria according to Fuhrman et al. (1982).–5Histologic cell type was classified according to the method of the AFIP (Murphy et al., 1994).–6AMD, alive with metastatic disease; CRD, cancer-related death; NED, no evidence of disease.–7Tumor karyotype was determined by conventional cytogenetic analysis on G-banded metaphase cells obtained from short-term culture. View, CA). Normal mouse serum was from DAKO (Glostrup, Denmark). Anti-CD1a was from Immuno Quality Products (Groningen, The Netherlands). In the case of direct immunofluorescence, 5 3 105 cells were incubated with the MAb for 15 min, red cells were lysed with FACS lysing solution (Becton Dickinson) and cells were fixed in a 0.5% paraformaldehyde/PBS solution. In the case of combined indirect/direct immunofluorescence, 5 3 105 cells were first incubated with unconjugated anti-CD8b (IgG1; Immunotech, Marseille, France) for 15 min. After 2 washings, cells were incubated with rat-anti-mouse-IgG1-PE for 15 min, washed twice and incubated with normal mouse serum to block the free antigen-binding sites of the rat-anti-mouse antibodies. After 15 min incubation, anti-CD8a-FITC and anti-CD4-PerCP MAbs were added, after which the procedure of direct immunofluorescence was followed. Three-color immunofluorescence analyses were performed on a FACScan flow cytometer (Becton Dickinson). Functional assays Ex vivo cytotoxic activity of the cells was tested in a 4 hr target cell lysis assay immediately after isolation, using murine P815 mastocytoma cells as (NK-resistant) targets and OKT3 (2 µg/ml) as anti-CD3 bridging MAb, as previously described (van Gool et al., 1993). This redirected cytotoxicity assay evaluates all T-cell cytotoxic activity, irrespective of antigen specificity (Leeuwenberg et al., 1985). For T-cell stimulation, 2.5 3 105 immunomagnetically purified CD41 TIL were cultured in 1 ml complete medium in 24-well flat-bottomed tissue culture plates in the presence of anti-CD3 MAb UCHT1. This anti-CD3 MAb was shown to have a strong stimulatory capacity in the absence of accessory signals (Verwilghen et al., 1991). Culture plates were coated by incubating the wells with UCHT1 (5 µg/ml) in 300 µl PBS for 4 hr at 37°C and then washed 3 times for 3 min with PBS. After 48 hr of stimulation, cells were collected for immunophenotyping. RESULTS CD31 CD41 CD8a11 T cells accumulate in the tumor tissue of patients with RCC Freshly isolated TIL from RCC were studied for CD3, CD4 and CD8a expression by 3-color flow-cytometric analysis. As shown in Figure 1 for 3 representative patients, an aberrant CD41 CD8a11 T-cell population could clearly be identified in some tumor tissues (Fig. 1a,b). The CD41 CD8a11 T-cell population is different from the CD31 CD41 CD8adim population, which could also be detected in many of the tumor samples (3.7% 6 3.2% of total TIL; .5% of total TIL in 8 of the 27 tumor samples) but which could clearly be separated in the dot plot from the CD41 CD8a11 T cells (Fig. 1b). For comparison, thymocytes which contain abundant CD4 and CD8a double-positive cells are also shown in Figure 1. Table I gives an overview of all patients with RCC who have been studied. In approximately one-third (8/27) of the patients, the CD41 CD8a11 T-cell subset represented more than 20% of CD31 TIL and in one patient more than half of CD31 TIL. Significant 180 VAN DEN HOVE ET AL. FIGURE 1 – CD4 and CD8a cell surface antigen expression on freshly isolated TIL of 3 patients with clear cell RCC (a–c) and on freshly isolated thymocytes (d). TIL were isolated from RCC tumor tissue stained with anti-CD4-FITC, anti-CD8a-PE and anti-CD3PerCP; and analyzed on a FACS. FACS dot plots show the results of CD4 and CD8a expression for the gated CD31 population. Thymocytes were isolated from the thymus of a 29-week-old fetus and analyzed similarly. proportions ($5%) of CD31 CD41 CD8a11 TIL were detected in 15 of 21 patients with clear cell RCC, whereas this subset was only marginally present (,3%) in all 5 patients with papillary RCC and in the one patient with sarcomatoid RCC, suggesting a possible relation with the histological tumor type. Furthermore, the occurrence of this subset was not clearly associated with pathological stage, tumor diameter, nuclear grade, cytogenetic abnormalities or vascular invasion in this patient cohort (Table I). The CD41 CD8a11 T-cell subset was detected in similar proportions in tumor tissue (6.5% 6 8.8% of total TIL) and tumor capsula (6.2% 6 7.9%) and in the adjacent normal renal tissue, albeit in a lower proportion (1.7% 6 2.3%). CD41 CD8a11 T cells were detectable in the peripheral blood of some of the 27 patients as a tiny population (0.4% 6 0.4% of total lymphocytes), which did not correlate with the presence of its counterpart in the tumor (not shown). CD31 CD41 CD8a11 TIL co-express CD8b but not CD1a and strongly express early and late activation antigens To further investigate the potential origin and in vivo significance of the CD31 CD41 CD8a11 TIL, additional 3-color immunofluorescence analysis was performed on TIL from 3 patients containing high proportions of CD41 CD8a11 T cells. This revealed that the CD41 CD8a11 TIL subset expressed the T-cell receptor a/b (Fig. 2b) but not the T-cell receptor g/d (not shown); CD8b (Fig. 2d); and strongly expressed CD69 (mean 6 SD 5 97% 6 2%), CD54 (99% 6 1%), HLA-DR (99% 6 1%) and CD28 (76% 6 10%) as well. Only a minor fraction of the cells were positive for CD45RA (15% 6 12%), CD25 (9% 6 7%), CD62L (18% 6 17%), CD (16 1 56) (25% 6 9%), CD57 (33% 6 12%) and CD1a (6% 6 2%). CD31 CD41 CD8a11 CD8b1 TIL of patients with RCC display no ex vivo cytolytic activity in an anti-CD3-redirected cytotoxicity assay against murine P815 cells To examine the function of the CD31 CD41 CD8a11 CD8b1 TIL subset, we first assessed the ex vivo cytolytic activity of FIGURE 2 – Identification of CD31 CD41 CD8a11 CD8b1 cells among TIL of RCC patient 5 by 3-color immunofluorescence. Freshly isolated TIL were stained with anti-T-cell receptor a/b-FITC, antiCD8a-PE and anti-CD4-PerCP (a, b) or anti-CD8a-FITC, antiCD8b-PE and anti-CD4-PerCP (c, d) and analyzed on a FACS. FACS fluorescence dot plots (a, c) show the expression levels (fluorescence intensities on a 4-log scale) of CD4/CD8a cell surface antigens. CD41 CD8a11 cells were gated and analyzed for expression of T-cell receptor a/b (b) and CD8b (d) in the 3rd color. FACS histograms (b, d) show the relative cell number vs. the fluorescence intensity (on a 4-log scale). Dotted line shows staining of control mouse IgG. immunomagnetically purified CD31 CD8a1 TIL of 2 patients with high proportions of CD31 CD41 CD8a11 CD8b1 TIL in 51Crrelease assays and compared this with the ex vivo cytolytic activity of immunomagnetically purified CD31 CD8a1 T cells from the peripheral blood of 3 healthy donors. FcgR-expressing mouse mastocytoma P815 cells and anti-CD3 MAb were used to detect cytotoxic activity regardless of antigen specificity of the cytotoxic effector cells (%SRP815). As can be seen in Figure 3, the CD31 CD8a11 TIL of both patients failed to lyse P815 cells in the anti-CD3-redirected cytotoxicity assay, whereas the freshly isolated peripheral blood CD31 CD8a1 T cells of 3 healthy donors demonstrated a strong cytolytic activity in this assay. No significant LAK activity (evaluated by 51Cr-release in the absence of anti-CD3 MAb) was detected in any of the experiments (data not shown). Anti-CD3-stimulated CD31 CD41 CD8a11 CD8b1 TIL demonstrate a defective IL-2Ra up-regulation when compared with CD31 CD41 CD8a2 TIL Subsequently, we investigated the up-regulation of the IL-2Ra chain (IL-2Ra, CD25) on anti-CD3-triggered CD31 CD41 CD8a11 CD8b1 and CD31 CD41 CD82 TIL. As shown in Figure 4, CD31 CD41 CD8a11 CD8b1 TIL demonstrated a strongly reduced capacity to up-regulate IL-2Ra when compared with CD31 CD41 CD8a2 TIL. In both experiments, we also noted that the expression level (mean fluorescence intensity) of IL-2Ra among the positive cells was higher on CD31 CD41 CD8a2 (mean 6 SD 5 1,143 6 304) than on CD31 CD41 CD8a11 CD8b1 (305 6 65) TIL (Fig. 4b,c). The addition of IL-2 (10 U/ml) to the anti-CD3stimulated CD41 TIL did not significantly enhance IL-2Ra expression on any subset (data not shown). ‘‘DOUBLE-POSITIVE’’ TUMOR-INFILTRATING LYMPHOCYTES FIGURE 3 – Ex vivo cytotoxic activity of CD31 CD8a1 T cells isolated from tumor tissue of 2 patients with primary RCC and from peripheral blood of 3 healthy donors. CD31 CD8a1 T cells were purified with Dynabeads (.95% pure). Purified CD31 CD8a1 TIL of patients 1 and 2 contained 55% and 35% of CD31 CD41 CD8a11 CD8b1 T cells, respectively. Murine P815 cells and anti-CD3 MAbs were used to detect cytotoxic activity regardless of antigen specificity (%SRP815). Results are expressed as mean of triplicate experiments. FIGURE 4 – Induction of IL-2Ra (CD25) on CD41 TIL of RCC patient 3. CD41 TIL were isolated from RCC tumor tissue and stimulated for 48 hr with immobilized anti-CD3. Cells were then stained with anti-CD4-FITC, anti-CD25-PE and anti-CD8a-PerCP and analyzed on a FACS. The FACS fluorescence dot plot shows the expression (fluorescence intensities on a 4-log scale) of CD4 and CD8a cell surface antigens (a). The CD41 CD8a11 and CD41 CD8a2 populations were further analyzed for CD25 expression. FACS histograms (b, c) show the relative cell number vs. the fluorescence intensity of CD25 staining (on a 4-log scale). Dotted line shows staining with control mouse IgG. Similar results were obtained with CD41 TIL of another patient. DISCUSSION We report the identification of an enriched CD41 CD8a11 CD8b1 T-cell subset in the tumor tissue of patients with RCC. In nearly one-third (8/27) of the patients, this subset comprised more than 20% of total CD31 TIL, and in one patient more than half of CD31 TIL displayed this phenotype. The presence of a substantial population of CD41 CD8a11 CD8b1 T cells outside of the thymus was surprising since peripheral T cells with this phenotype are very rarely detected. Persistent expansions of CD41 CD8a11 T cells in the peripheral 181 blood are extremely rare (Tonutti et al., 1994). A minor subset (3–10%) of double-positive CD31 CD41 CD8a1 intra-epithelial lymphocytes has been described in the small intestine of normal mice (Mosley et al., 1990; Lefrançois, 1991). Relevant to this, peripheral CD41 CD82 T cells have been shown to co-express the CD8a chain when they migrate into the small intestinal epithelium (Reimann and Rudolphi, 1995), and CD41 CD82 T-cell clones have been shown to co-express CD8a when cultured in the presence of IL-4 (Paliard et al., 1988; Hori et al., 1991). The CD41 CD8a11 CD8b1 TIL subset is, however, not an intratumoral homologue of the intra-epithelial double-positive T cells or the IL-4-stimulated T-cell clones because the latter do not express the CD8b chain (Lefrançois, 1991; Hori et al., 1991). Up-regulation of CD8a chains on CD41 cells, however, may explain the emergence of a distinct CD31 CD41 CD8adim CD8b2 TIL subset observed in some patients (Fig. 1b,c). The local production of IL-4 in RCC as supported by some reports (Schoof et al., 1993; Maeurer et al., 1995) may explain the induction of these CD8adim-positive CD41 TIL. The CD41 CD8a11 CD8b1 T cells in RCC tumor tissue may have several origins. First, they may derive from peripheral blood double-positive cells (which are probably immature thymic emigrants). In the thymus, CD31 CD41 CD8a1 cells represent intermediates in T-cell differentiation, which are derived from CD42 CD82 precursors and which further differentiate into mature CD31 CD41 CD8a2 or CD31 CD42 CD8a1 T cells (Kisielow and von Boehmer, 1995). However, this TIL subset did not express the CD1a cell surface marker of its counterpart in the thymus and displayed a different CD8a expression pattern from that of double-positive thymocytes (compare b with d in Fig. 1). If these cells are thymic emigrants, we have to assume that they lost CD1a expression upon migration into the tumor. A second (more likely) possibility is that the CD31 CD41 CD8a11 CD8b1 TIL arise in situ by further differentiation of mature single-positive CD41 T cells, which up-regulate CD8a and ultimately also express CD8b. Third, CD42 CD81 single-positive T cells which acquire CD4 expression also have been described (e.g., following infection with human lymphotropic herpes virus 6 [HHV-6]; Lusso et al., 1991). The possibility that the CD31 CD41 CD8a11 CD8b1 TIL subset originated from CD31 CD42 CD8a11 CD8b1 cells which upregulated CD4 thus should also be explored. Irrespective of their differentiation pathway, the aberrant T-cell phenotype most likely results from interaction with the tumor micro-environment, with either epithelial cell surface molecules or secreted molecules. With respect to their function, the expression of both CD8 isoforms together with CD4 suggests a functional capacity to interact with both MHC class I and class II molecules, broadening their immunological competence and contributing to high-affinity attachment to target tissues upon recognition of foreign antigen by the T-cell receptor. However, in contrast to freshly collected peripheral blood CD31 CD8a1 T cells from controls, freshly collected CD31 CD8a1 TIL containing high proportions of CD31 CD41 CD8a11 CD8b1 TIL (55% and 35%, respectively) demonstrated no anti-CD3-redirected cytotoxic activity against P815 cells. This strongly suggests that the CD31 CD41 CD8a11 CD8b1 TIL may not fulfill a cytotoxic effector function in vivo. Furthermore, CD31 CD41 CD8a11 CD8b1 TIL demonstrated a markedly high expression of CD69, CD54 and HLA-DR activation markers but were nearly negative for CD25. This combination of early and late activation markers together with a low expression of CD25 is indicative of a particular in vivo activation state that cannot be mimicked by in vitro activation of T cells via the T-cell receptor/ CD3 complex (Iannone et al., 1994). Furthermore, in contrast with CD31 CD41 CD8a2 TIL from the same tumor sample, CD31 CD41 CD8a11 CD8b1 TIL were incapable of expressing IL-2Ra in response to anti-CD3 triggering, suggesting a decreased capacity for IL-2-dependent clonal expansion. A marked defect in IL-2Ra expression of some TIL samples was reported in the study of Alexander et al. (1993), and our results suggest that the enriched 182 VAN DEN HOVE ET AL. CD31 CD41 CD8a11 CD8b1 TIL subset may importantly contribute. Finally, we have preliminary data which suggest that these cells also have a lower capacity to produce cytokines. Defective cytotoxic activity and lack of IL-2Ra expression upon anti-CD3 triggering collectively raise the possibility that CD31 CD41 CD8a11 CD8b1 TIL represent T cells in an anergic state, as has been suggested to occur in the tumor micro-environment as a consequence of incomplete stimulation by tumor cells lacking B7-co-stimulatory molecules (for review, see Chen et al., 1993). Alternatively, we cannot exclude the possibility that this subset represents a highly specialized effector population with a discrete, as yet unidentified function. Our characterization of true ‘‘double-positive’’ T lymphocytes among RCC TIL expressing CD4 and both CD8 chains may stimulate further research to reveal their origin, antigenic specificity and potential in vivo significance. An elucidation of their exact function may provide a better understanding of host-tumor interactions at tumor sites and may be of important clinical interest regarding the potential use of TIL in adoptive cellular immunotherapy. ACKNOWLEDGEMENTS The authors thank Dr. P. Beverley, Imperial Research Cancer Fund, London, UK, for generously providing the anti-CD3 MAb UCHT1, and Mrs. V. Boets, Miss K. Brugmans and Miss M. Adé for providing skilled technical assistance in the preparation of renal tissue samples. This work was supported by a grant from the Nationaal Fonds voor Wetenschappelijk Onderzoek (G.3123.94) and by a grant for cancer research from the Algemene Spaar—en Lijfrentekas, Brussels, Belgium. This text presents research results of the Belgian program on Interuniversity Poles of Attraction initiated by the Belgian State Prime Minister’s Office, Science Policy Programming. The scientific responsibility is assumed by its authors. S.V.G. is a postdoctoral researcher of the Fonds voor Wetenschappelijk Onderzoek Vlaanderen, Brussels. REFERENCES ALEXANDER, J.P., KUDOH, S., MELSOP, K.A., HAMILTON, T.A., EDINGER, M.G., TUBBS, R.R., SICA, D., TUASON, L., KLEIN, E., BUKOWSKI, R.M. and FINKE, J.H., T-cells infiltrating renal cell carcinoma display a poor proliferative response even though they can produce interleukin 2 and express interleukin 2 receptors. Cancer Res., 53, 1380–1387 (1993). CHEN, L., LINSLEY, P.S. and HELLSTRÖM, K.E., Costimulation of T cells for tumor immunity. Immunol. Today, 14, 482–486 (1993). DE RIESE, W., GOLDENBERG, K., ALLHOFF, E., STIEF, C., SCHLICK, R., LIEDKE, S. and JONAS, U., Metastatic renal cell carcinoma: spontaneous regression, long-term survival and late recurrence. Int. Urol. Nephrol., 23, 13–25 (1991). FARACE, F., ANGEVIN, E., VANDERPLANCKE, J., ESCUDIER, B. and TRIEBEL, F., The decreased expression of CD3 z chains in cancer patients is not reversed by IL-2 administration. Int. J. Cancer, 59, 752–755 (1994). FINKE, J.H., ZEA, A.H., STANLEY, J., LONGO, D.L., MIZOGUCHI, H., TUBBS, R.R., WILTROUT, R.H., O’SHEA, J.J., KUDOH, S., KLEIN, E., BUKOWSKI, R.M. and OCHOA, A.C., Loss of T-cell receptor zeta chain and p56lck in T-cells infiltrating human renal cell carcinoma. Cancer Res., 53, 5613–5616 (1993). FUHRMAN, S.A., LASKY, L.C. and LIMAS, C., Prognostic significance of morphologic parameters in renal cell carcinoma. Amer. J. surg. Pathol., 6, 655–663 (1982). GAUDIN, C., DIETRIC, P.-Y., ROBACHE, S., GUILLARD, M., ESCUDIER, B., TERRIER LACOMBE M.-J., KUMAR, A., TRIEBEL, F. and CAIGNARD, A., In vivo local expansion of clonal T cell subpopulations in renal cell carcinoma. Cancer Res., 55, 685–690 (1995). HORI, T., PALIARD, X., DE WAAL MALEFIJT, R., RANES, M. and SPITS, H., Comparative analysis of CD8 expressed on mature CD41 CD81 T cell clones cultured with IL-4 and that on CD81 T cell clones: implication for functional significance of CD8b. Int. Immunol., 3, 731–741 (1991). IANNONE, F., CORRIGALL, V.M., KINGSLEY, G.H. and PANAYI, G.S., Evidence for the continuous recruitment and activation of T cells into the joints of patients with rheumatoid arthritis. Europ. J. Immunol., 24, 2706–2713 (1994). KISIELOW, P. and VON BOEHMER, H., Development and selection of T cells: facts and puzzles. Adv. Immunol., 58, 87–209 (1995). LEEUWENBERG, J.F.M., SPITS, H., TAX, W.J.M. and CAPEL, P.J.A., Induction of nonspecific cytotoxicity by monoclonal anti-T3 antibodies. J. Immunol., 134, 3770–3775 (1985). LEFRANÇOIS, L., Phenotypic complexity of intraepithelial lymphocytes of the small intestine. J. Immunol., 147, 1746–1751 (1991). LUSSO, P., DE MARIA, A., MAINATI, M., LORI, F., DE ROCCO, S.E., BUSCIER, M. and GALLO, R.C., Induction of CD4 and susceptibility to HIV infection in human CD81 T lymphocytes by human herpes virus 6. Nature (Lond.), 349, 533–535 (1991). MAEURER, M.J., MARTIN, D.M., CASTELLI, C., ELDER, E., LEDER, G., STORKUS, W.J. and LOTZE, M.T., Host immune response in renal cancer: interleukin-4 (IL-4) and IL-10 mRNA are frequently detected in freshly collected tumor-infiltrating lymphocytes. Cancer Immunol. Immunother., 41, 111–121 (1995). MITROPOULOS, D., KOOI, S., RODRIGUEZ-VILLANEUVA, J. and PLATSOUCAS, C.D., Characterization of fresh (uncultured) tumour-infiltrating lymphocytes (TIL) and TIL-derived T cell lines from patients with renal cell carcinoma. Clin. exp. Immunol., 97, 321–327 (1994). MOSLEY, R.L., STYRE, D. and KLEIN, J.R., CD41 CD81 murine intestinal intraepithelial lymphocytes. Int. Immunol., 2, 361–365 (1990). MURPHY, W.M., BECKWITH, J.B. and FARROW, G.M., Atlas of tumor pathology: tumors of the kidney, bladder, and related urinary structures, Armed Forces Institute of Pathology, Washington, D.C. (1994). PALIARD, X., DE WAAL MALEFIJT, R. and DE VRIES, J.E., Interleukin-4 mediates CD8 induction on human CD41 T-cell clones. Nature (Lond.), 335, 642–644 (1988). PIERCE, W.C., BELLDEGRUN, A. and FIGLIN, R.A., Cellular therapy: scientific rationale and clinical results in the treatment of metastatic renal-cell carcinoma. Semin. Oncol., 22, 74–80 (1995). REIMANN, J. and RUDOLPHI, A., Co-expression of CD8a in CD41 T cell receptor ab1 T cells migrating into the murine small intestine epithelial layer. Europ. J. Immunol., 25, 1580–1588 (1995). SCHOOF, D.D., TERASHIMA, Y., PEOPLES, G.E., GOEDEGEBUURE, P.S., ANDREWS, J.V.R., RICHIE, J.P. and EBERLEIN, T.J., CD41 T cell clones isolated from human renal cell carcinoma possess the functional characteristics of Th2 helper cells. Cell. Immunol., 150, 114–123 (1993). SPIESSL, B., BEAHRS, O.H., HERMANEK, P., HUTTER, R.V.P., SCHEIBE, O., SOBIN, L.H. and WAGNER, G. (eds.), UICC-TNM atlas, pp. 261–269, Springer-Verlag, Berlin (1992). TARTOUR, E., LATOUR, S., MATHIOT, C., THIOUNN, N., MOSSERI, V., JOYEUX, I., DUBOIS D’ENGHIEN, C., LEE, R., DEBRÉ, B. and FRIDMAN, W.H., Variable expression of CD3-z chain in tumor-infiltrating lymphocytes (TIL) derived from renal-cell carcinoma: relationship with TIL phenotype and function. Int. J. Cancer, 63, 205–212 (1995). TONUTTI, E., SALA, P., FERUGLIO, C., YIN, Z. and COLOMBATTI, A., Phenotypic heterogeneity of persistent expansions of CD41 CD81 T cells. Clin. Immunol. Immunopathol., 73, 312–320 (1994). VAN GOOL, S.W., DE BOER, M. and CEUPPENS, J.L., CD28 ligation by monoclonal antibodies or B7/BB1 provides an accessory signal for the cyclosporin A-resistant generation of cytotoxic T cell activity. J. Immunol., 150, 3254–3263 (1993). VAN POPPEL, H., BOEL, K., VAN DAMME, B. and BAERT, L., Microscopic vascular invasion: the most significant prognostic factor in renal cell carcinoma [Abstract]. J. Urol., 153 (Suppl.), 255A (1995). VERWILGHEN, J., BAROJA, M.L., VAN VAECK, F., VAN DAMME, J. and CEUPPENS, J.L., Differences in the stimulating capacity of immobilized anti-CD3 monoclonal antibodies: variable dependence on interleukin-1 as a helper signal for T-cell activation. Immunology, 72, 269–276 (1991). WHITESIDE, T.L. and PARMIANI, G., Tumor-infiltrating lymphocytes: their phenotype, functions and clinical use. Cancer Immunol. Immunother., 39, 15–21 (1994). WHITTINGTON, R. and FAULDS, D., Interleukin-2. A review of its pharmacological properties and therapeutic use in patients with cancer. Drugs, 46, 446–514 (1993).