Differential expression and regulation of CD6 on T-cell subsets revealed by monoclonal antibody (MAb) CH11.

A monoclonal antibody (MAb), CH11, was developed by immunizing mice with CD4+ gammadelta T-cell receptor (TCR)+ cells. It recognized an antigen expressed in the surface membrane of T-cell lines, but not of U937, lymphoblastoid B cells (LBC), K562, Raji or Daudi cells, indicating selectivity for the T-cell lineage. In addition, it labelled 70-80% of normal peripheral blood mononuclear cells (PBMC), with high expression on the erythrocyte rosetting (E+) fraction, and low/absent expression on E- cells. However, CD4+ T cells expressed higher levels of reactivity than CD8+ or gammadelta+ T-cell receptor (TCR)+ lymphocytes in PB. Furthermore, in 7 of 10 individuals tested, 7.34+/-3.88% of unselected PBMC were CH11- CD3+ and were relatively enriched in CD8+ and in gammadelta TCR+-cells. In addition, thymic gammadelta T cells, and gammadelta lymphoproliferations from two patients were nonreactive or weakly reactive with the MAb. Activation of E+ cells with phorbol-12-myristate-13-acetate (PMA) enhanced CH11 expression uniformly, whereas activation with phytohemagglutinin (PHA) selectively down-regulated expression of the antigen on the CD8+ subset. In Western blots performed in nonreducing (NR) conditions, MAb CH11 detected a 100 kDa molecule in PBMC and Jurkat T-cell lysates. Preincubation of T cells with MAb CH11 specifically abrogated their subsequent reactivity with MAb to CD6, suggesting that MAb CH11 is recognizing an epitope of CD6. Given its function as a receptor for ligands on thymic epithelium, activated leukocytes and synoviocytes, this newly defined heterogeneity of expression and regulation of the CD6 molecule on subsets of T cells may help determine their functional repertoire in vivo.

Specific inhibition of glucocorticoid-induced thymocyte apoptosis by substance P.

Glucocorticoids (GC) are strong inducers of thymocyte apoptosis. In the present study we looked into the possibility that the neuropeptide substance P (SP) might serve as an antagonist to GC-induced apoptosis. Indeed, SP inhibited hydrocortisone (HC)-induced apoptosis of CD4+CD8+ thymocytes in mice, both in vivo and in vitro. It also inhibited HC-induced apoptosis in the T cell hybridoma line 2B4.11, which is sensitive to GC. The inhibitory effect was complete if SP was given with HC or within 1 h after it; partial inhibitory effect could be seen at 2 h and no effect at 3 h. The presence of the SP antagonist nullified SP effect. The effect was specific to both components of the system (i.e., HC as apoptosis inducer and SP as its inhibitor), as judged from comparison to three other apoptosis-inducing means (irradiation, thymic epithelial cells, or retinoic acid), and to two other neuropeptides (somatostatin and vasoactive intestinal peptide). SP/HC antagonism was further demonstrated in two relevant molecular events: 1) HC augmented GC receptor production in our cell system and this was inhibited by SP; and 2) HC reduced the expression of the transcription factor NF-kappaB, SP increased it and when both were present, SP effect dominated. On the other hand, the level of IkappaB (NF-kappaB inhibitory molecule) was decreased by SP, preserved at a relatively high level with HC, and when both SP and HC were present, SP effect dominated. The intensity of SP effect, both in vivo and in vitro, its specificity, its inhibition by SP antagonist, as well as the previously documented presence of SP and its receptor in the thymus suggest that SP might be a physiological antagonist of the potent thymocyte apoptosis induced by GC.

Targeting the vulnerable: a review of the necessity and feasibility of targeting vulnerable households.

This article examines whether it is possible to target vulnerable households within a geographically defined area. It looks first at the justification for targeting and then reviews recent practical experience in actually trying to reach vulnerable groups. As complex emergencies increasingly last longer, strategies to target vulnerable households are common in the protracted phase of the emergency. While this is often necessary because of a decline in resources, it is not always justified by an improvement in nutritional status or food security of the beneficiary population. Common target groups are the poor and the malnourished, but in complex emergencies these are not always the most vulnerable. Moreover, recent practical experience has shown considerable difficulties in targeting the poor. Methods to target the poor rely on community-based relief committees, whose priorities are not necessarily the same as those of external agencies. This paper gives examples of such targeted assistance programmes in Kenya, south Sudan and Tanzania. The paper concludes that situations where targeting vulnerable households is justified and feasible are extremely limited. It is suggested that if targeting has to be done because of scarce resources, this should be done on a geographical basis and on the basis of nutritional status. Case-study material shows that it is essential to understand the political determinants of vulnerability and to design methods that will reach the most vulnerable.

p94fer facilitates cellular recovery of gamma irradiated pre-T cells.

p94fer is a ubiquitous, nuclear and cytoplasmic tyrosine kinase, whose accumulation has been demonstrated in all mammalian cell lines analysed. In the present work, the p94fer expression profile was determined in cell lines which were not tested before. While being present in several hematopoietic and non hematopoietic cell lines including thymic stromal cells, the p94fer kinase could not be detected in pre-T and T cell lines. p94fer was also absent in pre-B line, but accumulated in these cells upon their induced development to antibody producing cells. This is in agreement with the absence of p94fer in primary thymic and splenic T lymphocytes and its induced accumulation in stimulated B cells. Relatively high p94fer levels were detected in primary thymic and splenic stromal cells. Ectopic expression of p94fer in pre-T cells slightly affected their cell cycle profile but it did not affect their apoptotic death which was induced by ionizing radiation. However, p94fer facilitated dramatically, the cellular recovery of gamma irradiated pre-T cells which have escaped the apoptotic death. The enhanced recovery of the irradiated, p94fer expressing pre-T cells, resulted most probably from their increased survival, rather than from a prominent change in their proliferation rate. The absence of p94fer from pre-B and pre-T cells, may thus contribute to the relative sensitivity of these cells to ionizing radiation and to their dependence on the functioning of other nuclear tyrosine kinasese.

Analysis of thymic stromal cell subpopulations grown in vitro on extracellular matrix in defined medium--V. Proliferation regulating activities in supernatants of human thymic epithelial cell cultures.

In previous reports in this series, we described the growth conditions, morphology and supernatant activities of human thymic epithelial cell cultures. The human thymic epithelial cell supernatant (HTES) contained IL-6, G-CSF activities and exhibited a strong enhancing effect on thymocyte proliferative response to mitogens, which was identified as IL-6 related. The responding thymocyte population was apparently identified as PNA, mature T cells. In order to simplify further analysis of HTES activities, we selected to use a well-defined mature murine T cell clone which has a Th2 phenotype (8-5 clone). HTES induced 8-5 cell proliferation without the presence of antigen, antigen presenting cells (APC) or mitogens. This enhancing effect of HTES was completely blocked with anti hIL-6 antibody but could not be reproduced by rhIL-6 alone. Hence, IL-6 is a necessary but insufficient factor in mediating this effect. HTES induced proliferation was accompanied by endogenous IL-4 secretion from 8-5 cells. Furthermore, the proliferation was blocked by anti mIL-4 antibody, implicating IL-4 as an autocrine growth factor in this system. HTES increased also the expression of IL-2 receptor. In addition, rhIL-2 and rmIL-4 each has a synergistic effect on the proliferative response of 8-5 cells to HTES. A similar synergistic activity was demonstrated when rhIL-6 was used instead of HTES, suggesting that IL-6 regulates some HTES activities. Our findings indicate that HTES activities, of which IL-6 is only part, are mediated via the induction of autocrine growth factors and by the regulation of T cell growth factor receptor expression.

Ubiquitous expression of a cloned murine thymopoietin cDNA.

Thymopoietin (TP) was originally isolated as a 5-kD 49-aa protein from bovine thymus and was subsequently observed to affect T-cell differentiation and function. We report here the molecular cloning of a murine TP cDNA. The 2,514 bp fragment contains a 630 bp open reading frame that encodes for 210 aa, highly homologous to the first 220 aa of the human TP beta and TP gamma isoforms and to bovine TP. Southern blot analysis of genomic DNA revealed that in mouse, calf and human TP is encoded by a single genomic locus. Recently, it was found that one of the TP isoforms designated TP beta is a homologous protein to the lamina-associated polypeptide 2 (LAP2), which is thought to play an important role in the regulation of nuclear architecture by binding lamin B1 and chromosomes in a manner regulated by phosphorylation during mitosis. In this study we report the TP expression at the transcription level in 17 murine and rat tissues of different origins and 18 lymphoid and nonlymphoid cell lines. The assessment of TP mRNA expression by S1-nuclease protection assays and in situ hybridizations revealed that its expression is not exclusive to thymus, but rather ubiquitous, higher in lymphatic tissues, but also in other cells characterized with a high rate of proliferation. TP was also shown to be expressed in athymic and old animals, lacking a functional thymus gland. Further in situ hybridization studies revealed that within the thymus, the highest levels of TP mRNA are noted at the cortex. These results suggest a possible role for TP in proliferation and cell cycle control.

Food aid and nutritional programmes during the Rwandan emergency.

In contrast to several other recent emergencies, the response of the international relief community to the Rwandan emergency appears largely to have prevented widespread malnutrition and related mortality. While it is true that aspects of the response in the food and nutrition sector were in various ways open to criticism and may have contributed to unnecessarily high levels of wasting in some camps at various points in time, the appalling excesses of famine witnessed in other recent African crises was not revisited during this emergency. Indeed, the main factors contributing to mortality and morbidity during the Rwandan emergency were violence and epidemics rather than lack of food and nutritional support.

Induction of apoptosis and p53 expression in immature thymocytes by direct interaction with thymic epithelial cells.

Apoptosis of normal thymocytes was shown to be triggered by several mechanisms (e.g. glucocorticoids, gamma-irradiation). In the present study the authors report on thymocyte apoptosis that is induced by thymic epithelial cells. The thymocytes undergo a massive apoptotic death within 24 h of cocultivation with thymic epithelial cell monolayers derived from primary cultures (PTEC) or from a thymic epithelial cell line (TEC). Non-thymic monolayers were inactive. Apoptosis induction in this experimental model requires direct contact between the thymocytes and the thymic epithelial monolayer and can be blocked by anti-CD2 and anti-LFA-1 antibodies. The immature CD3-/+dull CD4+CD8+ thymocytes were the cells which undergo apoptosis. The fact that the authors are dealing with a massive apoptotic process of immature cells in the absence of exogenous antigen suggests that it involves the nonselected thymocytes. The apoptotic pathway selected by thymocytes following their culturing on TEC involves p53 expression. Indeed it was found that TEC-induced apoptosis, led to the accumulation of p53 protein that preceded the step of DNA fragmentation in freshly isolated thymocytes as well as in a glucocorticoid resistant thymoma cell line. Since glucocorticoid-induced thymocyte apoptosis is p53-independent, glucocorticoids are conceivably not involved in TEC-induced thymocyte death. The in vitro experimental model presented here may reflect the physiological sequence of events leading to thymocyte death in the thymus.

The characterization and localization of the mouse thymopoietin/lamina-associated polypeptide 2 gene and its alternatively spliced products.

Thymopoietins (Tmpos) are a group of ubiquitously expressed nuclear proteins, with sequence homology to lamina-associated polypeptide 2 (LAP2). Here we report the isolation and characterization of seven mouse Tmpo mRNA transcripts named Tmpo alpha, beta, beta', gamma, epsilon, delta, and zeta. The alpha, beta, and gamma Tmpo cDNA clones are the mouse homologs of the previously characterized human alpha, beta, and gamma TMPOs, respectively, whereas Tmpo epsilon, delta, and zeta are novel cDNAs. Additionally, the mouse Tmpo gene was cloned and characterized. It is a single-copy gene organized in 10 exons spanning approximately 22 kb, which encodes all of the described Tmpo cDNA sequences, located in the central region of mouse chromosome 10. The almost identical genomic organization between the human and mouse genes, and the novel alternatively spliced mouse transcripts, led us to reanalyze the human TMPO gene. The human beta-specific domain was found to be encoded by 3 exons designated 6a, 6b, and 6c and not by a single exon as described previously. These findings suggest that there may be more human transcripts than currently recognized. The possible involvement of the new growing family of Tmpo proteins in nuclear architecture and cell cycle control is discussed.

The role of thymocytes in regulating thymic epithelial cell growth and function.

The basic tenet underlying the present work and supported by recent studies is that there is a dialogue between developing thymocytes and thymic stromal cells. One direction in this dialogue, i.e. thymic stromal cell role in shaping thymocyte maturation, has been extensively studied. The other direction, thymocyte effect on stromal cell development and function, started to emerge only recently on the basis of in vivo observations in SCID and knockout mice. An in vitro approach to the analysis of this interaction may add substantial insight into the process, as demonstrated by the present work. We made use of a culture system of either murine thymic epithelial cells (TEC line) cultured alone or cocultured with thymocytes. Unstimulated thymocytes or their supernatant caused 40-80% inhibition of TEC cell proliferation, as measured by 3H-thymidine incorporation. Cell cycle analysis by flow cytometry indicated that this inhibition can be attributed to reduction in G2/M phase cell number pari passu with an increase in Go/G1 cell number. This inhibitory effect was found to be partially mediated by TGF-beta produced by thymocytes. On the other hand, thymocytes augmented IL-6 production by TEC cells in coculture, an effect which could not be reproduced by thymocyte culture supernatant and was not inhibited by thymocyte pretreatment with formaldehyde or emetine. Furthermore, antibodies against thymocyte adhesion molecules (CD2, LFA-1) blocked the thymocyte-induced IL-6 secretion. IL-6 was found to be an autocrine growth factor of TEC in culture, since a combination of anti IL-6 and anti IL-6 receptor antibodies caused 70% inhibition of TEC proliferation and addition of exogenous recombinant IL-6 doubled the rate of proliferation. These results suggest that thymocytes regulate thymic epithelial cell growth by a complex set of inhibitory and enhancing signals mediated through either soluble factors or direct contact. The ultimate effect is dependent on the balance between different signals and may be different in different microenvironmental settings in vivo. In coculture in vitro the dominant effect was growth inhibition of the epithelial cells by thymocytes.

Demonstration of thymopoietin transcripts in different hematopoietic cell lines.

We have cloned and characterized the human thymopoietin (TP) coding region and studied the mRNA expression of this gene in different hematopoietic cell lines. The 150-bp PCR fragment that encodes the 49-amino-acid human TP peptide was isolated from genomic placental DNA. Its colinearity with the cDNA sequence suggests lack of introns within the coding region. TP mRNA expression was demonstrated in lymphocytes from all the differentiation stages investigated, as well as in a myeloid cell line (K-562). These findings suggest a further expansion of the proposed TP functions.

Analysis of thymic stromal cell subpopulations grown in vitro on extracellular matrix in defined medium. IV. Cytokines secreted by human thymic epithelial cells in culture and their activities on murine thymocytes and bone marrow cells.

In previous reports we described our approach to the cultivation of murine and human thymic epithelial cells in primary cultures, using defined, serum-free growth factor-supplemented medium and extracellular matrix-coated culture plates. The cells in these cultures displayed high metabolic activity and their supernatant was highly active on thymocytes. In the study reported here we analysed cytokine activities in the supernatant of human thymic epithelial cell cultures (HTES), by using the respective cytokine-dependent cell lines and by neutralization with specific monoclonal antibodies. Three cytokine activities were detected--interleukin-6 (IL-6), granulocyte colony-stimulating factor (G-CSF) and macrophage (M)-CSF. Other cytokine activities tested for [IL-1, IL-2, IL-7, interferon (IFN) and tumour necrosis factor (TNF)] were negative. The effect of HTES on concanavalin A (Con A)-induced proliferation of murine thymocytes could be completely abolished by anti-IL-6 antibodies, but not by antibodies to CSF, whereas enhancement of bone marrow cell proliferation by HTES was partially inhibited by either anti-G-CSF or anti-M-CSF antibodies and completely inhibited by both antibodies, but not at all by anti-IL-6. We can thus distinguish between thymocyte-related cytokines (IL-6) and bone marrow (myeloid/monocyte) related ones (G-CSF, M-CSF) in HTES.

Molecular cloning of the bovine thymopoietin gene and its expression in different calf tissues: evidence for a predominant expression in thymocytes.

Thymopoietin (TP), a 49 amino acid peptide, is regarded as a thymic hormone, secreted specifically by some epithelial cells in the thymic stroma and exerting a multitude of effects on maturation and function of T lineage cells. As part of our study on the molecular biology of TP, we isolated cDNA clone coding for a bovine TP precursor and used it as a probe to analyze the presence of mRNA coding for TP in different tissues. The cDNA clone reported here is 1.1 kb long and contains an open reading frame (ORF) of 741 bp which corresponds to 247 amino acids. The 147 bp coding for the entire bovine TP are at the 5' end of the ORF. A DNA fragment coding for amino acids 1-42 of bovine TP was used as a probe to look for hybridizable RNA sequences, extracted from various calf tissues, by the S1 nuclease protection method. Our results indicate that the TP gene is expressed predominantly in lymphatic tissues. Lymphatic tissues with the highest levels observed were thymocytes and not thymic stroma. Lower, but still significant, amounts were present in tonsils, neck lymph nodes, and small intestine (probably because of its lymphatic part--the Peyer's patches), whereas cultured thymic stromal cells, spleen tissue and peripheral blood mononuclear cells displayed a low level of TP mRNA. The TP gene expression in all other (non-lymphatic) tissues tested, was weak, barely detectable or virtually absent. However, the cerebellum could be singled out with relatively strong expression of TP mRNA.(ABSTRACT TRUNCATED AT 250 WORDS)

Analysis of thymic stromal cell subpopulations grown in vitro on extracellular matrix in defined medium. III. Growth conditions of human thymic epithelial cells and immunomodulatory activities in their culture supernatant.

We report here on a new approach to the cultivation of human thymic epithelial (HTE) cells, which apparently allows more faithful preservation of cell function. This approach, previously developed by us for mouse thymic epithelial (MTE) cells, is based on the use of culture plates coated with extracellular matrix (ECM), and on the use of serum-free, growth factor-supplemented medium. The nutritional requirements of HTE and MTE are somewhat different. Although both are critically dependent on ECM and insulin, they differ in their dependency on other growth factors: selenium and transferrin are much more important for HTE cells, whereas epidermal growth factor and hydrocortisone play a more essential role in MTE cultures. The epithelial nature of the cultured cells is indicated by positive staining with anti-keratin antibodies and by the presence of desmosomes and tonofilaments. The ultrastructural appearance of the cells further suggests high metabolic and secretory activities, not usually found in corresponding cell lines. The culture supernatant (CS) of HTE cells exhibited a strong enhancing effect on thymocyte response to Con A stimulation, as measured by cell proliferation and lymphokine production. The effect was observed on both human and mouse thymocytes, but was much stronger in the homologous combination. Thymic factors tested in parallel did not have such a differential effect. The dose-effect relationships were in the form of a bell-shaped curve, with fivefold enhancement of response at the peak and a measurable effect even with 1:1000 dilution, when human thymocytes were used. The responding thymocytes were those which do not bind peanut agglutinin and are resistant to hydrocortisone. The culture system described here may have advantages for the in vitro study of thymic stromal cell function.

Analysis of thymic stromal cell subpopulations grown in vitro on extracellular matrix in defined medium. I. Growth conditions and morphology of murine thymic epithelial and mesenchymal cells.

We report here the successful selective cultivation of murine thymic mesenchymal reticular cells (MTMC) and murine thymic epithelial cells (MTEC) grown on extracellular matrix in the presence of defined medium. The selective growth of these two cell types was based on 1) conditions of tissue disruption and 2) differential growth requirements. Both cell types were dependent on transferrin, high density lipoproteins, insulin, hydrocortisone, and epidermal growth factor, whereas MTMC was dependent also on selenium and 3,5,3'-triiodothyronine. The elimination of single factors or extracellular matrix resulted in specific and different changes in the growth pattern of each cell subpopulation. Cells of both types exhibited the ultrastructural features of high metabolic activity. The epithelial nature of MTEC cultures was defined by bundles of tonofilaments and desmosomes and by positive staining to keratins and negative to vimentin. In addition MTEC were positively stained with mAb to thymic medullary epithelial cells and by Ulex europeus agglutinin, and were able to form Hassall's corpuscles, suggesting their medullary origin. MTEC were also H-2 and Ia positive. In contrast MTMC were positive for vimentin and periodic acid-Schiff, low positive for H-2, and negative for keratin and Ia. Both cells did not contain nonspecific esterase, nor did they phagocytize latex beads. With the use of all these criteria we classified MTEC as epithelial cells from the medullary compartment of the thymus and MTMC as reticular cells of mesenchymal origin.

Analysis of thymic stromal cell subpopulations grown in vitro on extracellular matrix in defined medium. II. Cytokine activities in murine thymic epithelial and mesenchymal cell culture supernatants.

Two morphologically distinct primary cultures of murine thymic stroma were established and found to be of epithelial (MTEC) and mesenchymal (MTMC) origin. These cultures were generated by selective conditions of tissue disruption and were maintained on extracellular matrix in defined medium. Culture supernatants (CS) from these cultures (EC-CS and MC-CS respectively), were tested for cytokine production and for effects on thymocyte maturation. Both supernatants displayed the activities of IL-3 and of granulocyte/macrophage-CSF and not of IL-1, -2, -4, or IFN. In addition they were found to be mitogenic to murine thymocytes in a "spontaneous" [3H]TdR incorporation assay. The two supernatants differed, however, in their effect on Con A stimulation. EC-CS had a strong enhancing effect, both when used for preincubation (18 h) before Con A stimulation or when present simultaneously with it. MC-CS had a small inconsistent effect under these conditions. Also EC-CS enhanced IL-2 and IL-3 production by thymocytes. The responsive thymocyte subpopulation was the one that does not bind peanut agglutinin. CS of an established thymic epithelial cell line displayed only part of these activities at a considerably lower level. CS from primary kidney cell culture was completely devoid of activity. The results suggest that primary thymic stromal cell cultures, cultivated under the defined conditions described here, may better preserve physiologic secretory activities, and probably also other cell functions, compared with established cell lines. Furthermore, the results are compatible with the hypothesis that the soluble factors, secreted by thymic stromal cells, are active on either very early or late stages of thymic differentiation, whereas the main intrathymic stages of differentiation are conceivable dependent primarily on direct contact with stromal cells.

Augmentation of tumor cell immunogenicity by viruses--an approach to specific immunotherapy of cancer.

Several viruses have been evaluated as potential agents for cancer treatment using either their oncolytic properties, in order to lyse cancer cells, or their potential augmenting effects on the immune response to tumors. However, the direct oncolytic effect was found to be limited in time, in scope and in specificity, whereas the use of viral oncolysates to augment antitumor immunity was shown to be better than tumor cell homogenates or extracts but inferior to noninfected intact tumor cells, attesting for the importance of membrane architecture in preserving immunogenicity of tumor specific surface antigens. In order to get the maximum benefit from this approach we selected a nonlytic virus-tumor cell combination, using Newcastle disease virus as a nonpathogenic virus, to treat the experimental tumor model, Lewis lung carcinoma (3LL) in mice. The virus effectively infected 3LL cells without any cytopathic effect. The infected cells induced strong antitumor immunity, as judged by the appearance of immune cells in the spleen (Winn test and lymphocytotoxicity) and by the resistance to challenge with the 3LL cells after immunization. The antitumor immunity was superior to that obtained with intact noninfected tumor cells. We also designed a treatment protocol using the same virus-tumor cell preparation to treat mice after tumor inoculation. This treatment resulted in cure of 40% of the animals.(ABSTRACT TRUNCATED AT 250 WORDS)

Contact-mediated maturational effects of thymic stromal cells on murine thymocytes in culture.

The effect of direct contact between thymic stromal cells and thymocytes on differentiation markers and functions of the latter was studied. The thymic stromal cells included two epithelial and one fibroblast cell lines, previously described. Murine thymocytes were incubated with confluent monolayers of these cells or their supernatants for 24 hr, using monolayers of non-thymic cells and their supernatants as controls. Then, the thymocytes were tested for changes in expression of several surface antigens [Thy-1, Lyt-1, Lyt-2, L3T4, IL-2-receptor (IL-2R)], spontaneous [3H]thymidine incorporation (STI), lectin-induced proliferative response (PR) and lymphokine (IL-2 and IL-3) production. All three thymic stromal cell lines reduced the expression of Thy-1, Lyt-1 and Lyt-2 significantly. The expression of L3T4 was also reduced in some of the experiments, while IL-2R was not expressed by the thymocytes, neither before nor after the co-culture. The thymic stromal cell lines also increased the spontaneous [3H]thymidine incorporation and lymphokine production by the thymocytes and inhibited their proliferative response to lectins. Under the same experimental conditions, the culture supernatants of the thymic stromal cells and the non-thymic cells did not have any effect on the thymocytes, either when collected and used separately or when used in a co-culture system which allowed thymocyte contact with the medium but not with the stromal cells (Transwell system). These results suggest a specific effect of thymic stromal cells, epithelial as well as fibroblasts, on thymocyte maturation. The effect is mediated by direct cell contact and not by secreted factors.