Migrant μ+δ+ and static μ+δ− B lymphocyte subsets

Immunoglobulin isotype expression in isolated lymph node (LN), spleen and blood lymphocyte suspensions was assessed in rats. The proportion of μ⁺δ^? B cells in spleen (34%) was approximately twice that in blood and LN. Immunohistological examination of spleens showed the cells of the marginal zones to be predominantly μ⁺δ^?. On the other hand, μ⁺δ⁺ B cells were mainly confined to the follicles in both spleen and LN. These follicles had a minor μ⁺δ^? component. The migratory properties of B cells with these two phenotypes were assessed by depleting lymphocytes migrating through the white pulp of rat spleen. This was achieved by placing a ³²P-impregnated β-emitting polythene strip over one half of the spleen. Examination of the nonirradiated half of the spleen, LN and peripheral blood after 12 days irradiation showed selective loss of δ⁺ B cells. The μ⁺δ^? cells of the splenic marginal zone were numerically unaltered. There was also a substantial residual μ⁺δ^? B cell presence in the small lymphocyte compartment of follicles of LN and spleen in depleted animals. In addition, the blood selectively retained a μ⁺δ^? B cell component. This was not derived from the spleen, as μ⁺δ^? blood B cell numbers were sustained even where both halves of the spleen were irradiated. It is concluded that: (a) the major static B cell component of spleen and LN is μ⁺δ^?, (b) that most if not all δ⁺ B cells repeatedly migrate through the spleen and (c) there is a blood-born μ⁺δ^? component which is resistant to depletion by splenic irradiation.     

CD5− CD8αβ intestinal intraepithelial lymphocytes (IEL) are induced to express CD5 upon antigen-specific activation: CD5− and CD5+ CD8αβ IEL do not represent separate T cell lineages

We followed αβ T cell receptor (TCR) usage in subsets of gut intraepithelial lymphocytes (IEL) in major histocompatibility complex class I-restricted αβ TCR-transgenic (tg) mice. The proportion of tg αβ TCR⁺ CD8αβ IEL is reduced compared with CD8⁺ splenocytes of the same animal, particularly under conventional conditions of maintenance. Further fractionation of CD8αβ IEL according to the expression level of surface CD5 revealed that in conventionally housed animals tg TCR⁺ CD5^? CD8αβ IEL are as frequent as in specific pathogen-free (SPF) mice, whereas tg TCR⁺ CD5^int or, even more pronounced, tg TCR⁺ CD5^hi CD8αβ IEL are greatly diminished when compared with mice kept under SPF conditions. Upon antigen-specific stimulation of CD5^? CD8αβ IEL in vitro, CD5 surface expression is up-regulated on a large fraction of cells within 48 h. Up-regulation of CD5 surface expression is further enhanced by the presence of the anti-αIEL monoclonal antibody 2E7. This clearly demonstrates that CD5^?, and CD5⁺ CD8αβ IEL cannot be considered as separate T cell lineages.     

Pro‐inflammatory self‐reactive T cells are found within murine TCR‐αβ+CD4−CD8−PD‐1+ cells

TCR‐αβ⁺ double negative (DN) T cells (CD3⁺TCR‐αβ⁺CD4^?CD8^?NK1.1^?CD49b^?) represent a minor heterogeneous population in healthy humans and mice. These cells have been ascribed pro‐inflammatory and regulatory capacities and are known to expand during the course of several autoimmune diseases. Importantly, previous studies have shown that self‐reactive CD8⁺ T cells become DN after activation by self‐antigens, suggesting that self‐reactive T cells may exist within the DN T‐cell population. Here, we demonstrate that programmed cell death 1 (PD‐1) expression in unmanipulated mice identifies a subset of DN T cells with expression of activation‐associated markers and a phenotype that strongly suggests they are derived from self‐reactive CD8⁺ cells. We also found that, within DN T cells, the PD‐1⁺ subset generates the majority of pro‐inflammatory cytokines. Finally, using a TCR‐activation reporter mouse (Nur77‐GFP), we confirmed that in the steady‐state PD‐1⁺ DN T cells engage endogenous antigens in healthy mice. In conclusion, we provide evidence that indicates that the PD‐1⁺ fraction of DN T cells represents self‐reactive cells.     

CD8α+ dendritic cells prime TCR‐peptide‐reactive regulatory CD4+FOXP3− T cells

CD4⁺ T cells with immune regulatory function can be either FOXP3⁺ or FOXP3^?. We have previously shown that priming of naturally occurring TCR‐peptide‐reactive CD4⁺FOXP3^? Treg specifically controls Vβ8.2⁺CD4⁺ T cells mediating EAE. However, the mechanism by which these Treg are primed to recognize their cognate antigenic determinant, which is derived from the TCRVβ8.2‐chain, is not known. In this study we show that APC derived from splenocytes of naïve mice are able to stimulate cloned CD4⁺ Treg in the absence of exogenous antigen, and their stimulation capacity is augmented during EAE. Among the APC populations, DC were the most efficient in stimulating the Treg. Stimulation of CD4⁺ Treg was dependent upon processing and presentation of TCR peptides from ingested Vβ8.2TCR⁺CD4⁺ T cells. Additionally, DC pulsed with TCR peptide or apoptotic Vβ8.2⁺ T cells were able to prime Treg in vivo and mediate protection from disease in a CD8‐dependent fashion. These data highlight a novel mechanism for the priming of CD4⁺ Treg by CD8α⁺ DC and suggest a pathway that can be exploited to prime antigen‐specific regulation of T‐cell‐mediated inflammatory disease.     

Progenies of fetal thymocytes are the major source of CD4−CD8+ αα intestinal intraepithelial lymphocytes early in ontogeny

Present literature supports the view of an extrathymic origin for the subset of intestinal intraepithelial lymphocytes (IEL) that express the CD4^?CD8⁺ αα phenotype. This subset would include virtually all T cell receptor (TCR) γδ IEL and a portion of TCR αβ IEL. However, these reports do not exclude the possibility that some CD4^?CD8⁺ αα IEL are actually thymically derived. To clarify this issue, we examined the IEL day 3 neonatally thymectomized (NTX) mice. NTX resulted in as much as 80 % reduction in total TCR γδ IEL and in a nearly complete elimination of TCR αβ CD4^?CD8⁺ αα IEL early in ontogeny (3-to 5-week-old mice). The thymus dependency of TCR γδ IEL and TCR αβ CD4^?CD8⁺ IEL was less prominent in older mice (7- to 10-week-old mice), as the total number of these IEL increased in NTX mice, but still remained severalfold less than that in euthymic mice. Furthermore, we demonstrate, by grafting the fetal thymus of CBF1 (H-2^b/d) mice under the kidney capsule of congenitally nude athymic mice of BALB/c background (H-2^d), that a substantial number of TCR γδ IEL and TCR αβ CD4^?CD8⁺ αα IEL can be thymically derived (H-2^b+). In contrast, but consistent with our NTX data, grafting of adult thymi into nude mice generated virtually no TCR γδ IEL and relatively less TCR αβ CD4^?CD8⁺ αα IEL than did the grafting of fetal thymi. These results suggest that the thymus is the major source of TCR γδ and TCR αβ CD4^?CD8⁺ αα IEL early in ontogeny, but that the extrathymic pathway is probably the major source of these IEL later in ontogeny. A reassessment of the theory that most CD4^?CD8⁺ IEL are extrathymically derived is needed.     

Increased number of cytotoxic T cells within CD4+8− T cells in β2-microglobulin,major histocompatibility complex class I-deficient mice

Targeted disruption of β₂-microgobulin gene results in deficient major histocompatibility complex class I expression and failure to develop CD4^?8⁺ T cells. Despite this, β₂M^?/? mice reject skin grafts and cope with most viral infections tested. We asked whether CD4⁺8^? cytotoxic T cells could play a role in compensating for the defect in CD4^?8⁺ cytotoxic T cell function. We found that the cytotoxic activity against class II⁺ targets is significantly higher among CD4⁺8^? T cells of β₂M^?/? than among those of β₂M^+/+ mice. In the limiting dilution experiment, we showed that the precursor frequency for the cytotoxic, CD4⁺8^?, class II-specific T cells is at least fivefold higher in β₂M ^?/?than in β₂M^+/+ mice. These results suggest that CD4+8^? cytotoxic T cells could play a major role in carrying out cytotoxic function in β₂M^?/? mice.     

A human CD4 transgene rescues CD4−CD8+ cells in β2-microglobulin-deficient mice

The specificity of the αβ T cell receptor for class I or class II major histocompatibility complex (MHC) molecules determines whether a mature T cell will be of the CD4^?CD8⁺ or CD4⁺CD8^? phenotype, respectively. We show here that a human CD4 transgene can rescue a significant fraction of CD4^?CD8⁺ T cells in β2-microglobulin-deficient mice. Cells with this phenotype could be induced to become potent killers of targets expressing allogeneic MHC antigens, indicating that lineage commitment can precede the rescue of developing cells by the T cell receptor for antigen and the CD4 coreceptor.     

Expansion of the CD4−, CD8− γδ T cell subset in the spleens of mice during non-lethal blood-stage malaria

Splenic γδ T cells (CD4^?, CD8^?) increased more that 10-fold upon resolution of either Plasmodium chabaudi adami or P. c. chabaudi infections in C57BL/6 mice compared to controls. Similarly, a 10- to 20-fold expansion of the γδ T cell population was observed in β₂-microglobulin deficient (β²-m^0.0) mice that had resolved P. c. adami, P. c. chabaudi or P. yoelii yoelii infections. In contrast, increases in the number of splenic αβ T cells in these infected mice were only two to three-fold indicating a differential expansion of the γδ T cell subset during malaria. Because nucleated cells of β₂-m^0/0 mice lack surface expression of major histocompatibility complex class I and class Ib glycoproteins, our findings suggest that antigen presentation by these glycoproteins is not necessary for the increasing number of γδ T cells. Our observation that after resolution of P. c. adami malaria, C57BL/6 mice depleted of CD8⁺ cells by monoclonal antibody treatment had lower numbers of γδ T. cells than untreated controls suggests that the demonstrated lack of CD8⁺ cells in β₂-m^0/0 mice does not contribute to the expansion of the γδ T cell population during non-lethal malaria.     

Preparative nonlytic separation of Lyt2+ and Lyt2− T lymphocytes,functional analyses of the separated cells and demonstration of synergy in graft-vs.-host reaction of Lyt2+ and Lyt2− cells

A convenient, preparative scale, nonlytic separation of mouse T lymphocytes into Lyt2.2⁺ and Lyt2.2^? populations is reported. Immunoglobulin-negative (Ig^?) spleen cells, Ig^? lymph node cells, and peanut lectin-unagglutinated (PNA^?) thymocytes were incubated under sterile conditions at 0°C with monoclonal mouse antibody to the Lyt2.2 T cell differentiation antigen. The antibody-treated cells were washed and placed in polystyrene tissue culture dishes that had been precoated with antibody to mouse Ig. Nonadherent populations were depleted of Lyt2.2⁺ cells and were essentially devoid of cytotoxic T lymphocyte precursors (CTLp), but contained helper activity for in vivo T-dependent IgM, IgG and IgA antibody formation. Adherent cell populations were enriched for Lyt2.2⁺ cells and for CTLp. The graft-vs.-host activity of the separated, adherent (Lyt2.2⁺) and nonadherent (Lyt2.2^?) cells in the Simonsen spleen assay in neonatal (C57BL/6 × BALB/c)F₁ mice was less than that of unfractionated cells, but the activity of remixed Lyt2.2⁺ plus Lyt2.2^? cells was higher than the sum of the contributions of these cells tested separately, and equal to that of the unfractionated cells. PNA^? thymocytes were also separated into Lyt2.2⁺ and Lyt2.2^? populations by fluorescence-activated cell sorting. Nonlytic separation allows the recovery of the Lyt1⁺2⁺ population, which is lost in cytotoxic elimination experiments. Under the conditions described for the plate separation, the purity of the separated cells and recovery of activity approaches that of cells separated by sorting. Therefore, the plate separation offers a convenient alternative to fluorescence-activated cell sorting when large numbers (i.e. up to 5 × 10⁷ positively selected cells) are needed, as in studies of in vivo cell-mediated immune reactions.     

The nature of the signals regulating CD8 T cell proliferative responses to CD8α+ or CD8α− dendritic cells

The CD8α^?-expressing dendritic cells (DC) of mouse spleen have been shown to be poor inducers of interleukin (IL)-2 production by CD8 T cells when compared to the CD8^? DC. As a consequence, CD8 T cells give a more prolonged proliferative response to CD8^? DC than to CD8⁺ DC. The possible mechanisms underlying these functional differences in DC subtype have been investigated. Inadequate co-stimulation did not underlie the poor T cell response to allogeneic CD8⁺ DC. Equivalent levels of B7-1 (CD80) and B7-2 (CD86) were found on the two DC subtypes and co-stimulator assays did not reveal any functional differences between them. Although CD8⁺ DC were found to die more rapidly in culture than CD8^? DC, this did not explain their reduced stimulatory ability. Neither prolonging DC survival in culture nor renewing the stimulator cells by repeated addition of freshly isolated DC had any significant effect on the T cell responses. Furthermore, later addition to the cultures of DC of the opposite type to the initiating DC did not reverse or eliminate the differential response to the initiating DC. The role of DC-derived soluble factors was examined by addition to the cultures of supernatants derived from freshly isolated or stimulated DC of the opposite type. This neither enhanced the poor stimulatory capacity of CD8⁺ DC nor inhibited the stimulation by CD8^? DC. Furthermore, addition of a series of cytokines that might have been produced by the DC did not eliminate the differences in T cell proliferation. Only the addition to the cultures of the growth factors IL-2 and IL-4 overcame the stimulatory difference between the two DC populations, confirming that the difference in T cell proliferative responses was a consequence of differences in induced cytokine production. The difference in the response of CD8 T cells to CD8⁺ and CD8^? DC is therefore determined by direct DC-T cell contact during the earliest stages of the culture and involves an undetermined and possibly new signaling system.     

Co-expression of CD8α in CD4+ T cell receptor αβ+ T cells migrating into the murine small intestine epithelial layer

We investigated the surface phenotype of CD3⁺CD4⁺ T cell receptor (TCR) αβ⁺ T cells repopulating the intestinal lymphoid tissues of C.B-17 scidlscid (severe-combined immunodeficient; scid) (H-2^d, L^d+) mice. CD4⁺ CD8^? T cells were cell sorter-purified from various secondary and tertiary lymphoid organs of congenic C.B-17 +/+ (H-2^d, L^d+) or semi-syngeneic dm2 (H-2^d, L^d?) immunocompetent donor mice. After transfer of 10⁵ cells into young scid mice, a mucosa-homing, memory CD44^hi CD45RB^lo CD4⁺ T cell population was selectively engrafted. Large numbers of single-positive (SP) CD3⁺ CD2⁺ CD28⁺ CD4⁺ CD^8? T cells that expressed the α₄ integrin chain CD49d were found in the spleen, the mesenteric lymph nodes, the peritoneal cavity and the gut lamina propria of transplanted scid mice. Unexpectedly, large populations of donor-type doublepositive (DP) CD4⁺ CD8α⁺ CD8β^? T cells with high expression of the CD3/TCR complex appeared in the epithelial layer of the small intestine of transplanted scid mice. In contrast to SP CD4⁺ T cells, the intraepithelial DP T cells showed low expression of the CD2 and the CD28 co-stimulator molecules, and of the α₄ integrin chain CD49d, but expressed high levels of the α_IEL integrin chain CD103. The TCR-Vβ repertoire of DP but not SP intraepithelial CD4⁺ T cells was biased towards usage of the Vβ6 and Vβ8 viable domains. Highly purified populations of SP and DP CD4⁺ T cell populations from the small intestine epithelial layer of transplanted scid mice had different abilities to repopulate secondary scid recipient mice: SP CD4⁺ T cells repopulated various lymphoid tissues of the immunodeficient host, while intraepithelial DP CD4⁺ T cells did not. Hence, a subset of CD3⁺ CD4⁺ TCRαβ⁺ T cells apparently undergoes striking phenotypic changes when it enters the microenvironment of the small intestine epithelial layer.     

CD3−4−8− Thymocyte Precursors with Interleukin-2 Receptors Differentiate Phenotypically in Coculture with Thymic Stromal Cells

CD3^?4^?8^? interleukin-2 receptor positive (IL-2R⁺) thymocyte precursors from adult mice were cocultured with thymic stromal cells from syngeneic adult mice. The IL-2R⁺CD3^?4^?8^? thymocytes were obtained by positive panning of IL-2R⁺ cells followed by either sorting or negative panning of triple negative cells, and they were cocultured with primary or secondary cultures of heterogeneous thymic stromal cells. Phenotypic maturation of these precursor cells was extremely rapid. Within 2½ days significant numbers of CD4⁺8⁺ and CD3⁺4⁺8^? cell populations developed, the latter expressing the αβ T-cell receptor (αβ-TCR). Thus heterogeneous stromal cell cultures support the development of IL-2R⁺ precursors and with these methods it will now be possible to isolate the particular stromal cells involved at each stromal-dependent step.     

CD3−CD8+ intestinal intraepithelial lymphocytes (IEL) and the extrathymic development of IEL

Present evidence suggests that a majority of murine CD3⁺ intraepithelial intestinal lymphocytes (IEL) are extrathymically derived T cells and that these extrathymically derived IEL phenotypically express the CD8 homodimer (CD8αα). Recently, CD3^? IEL have been reported to express the recombination activating gene (RAG-1), suggesting that precursors to extrathymically derived CD3⁺CD8⁺αα IEL exist on the intestinal epithelium. To study in detail whether these CD3^? IEL can develop into CD3⁺CD8⁺αα IEL, we analyzed the CD3^? IEL subset and found that it can be separated into two subsets, namely CD3^?CD8^? and CD3^?CD8⁺ IEL. We show that (1) CD3^?CD8^? IEL are mostly small, non-granular and phenotypically Pgp-1⁺ IL-2R⁺ B220^?, while CD3^?CD8⁺ IEL are mostly large, granular and phenotypically Pgp-1^? IL-2R⁺ B220⁺, (2) CD3^?-CD8⁺ IEL express the RAG-1 gene, and (3) CD3^?CD8^?, CD3^?CD8⁺ and CD3⁺CD8⁺αα IEL, respectively, appear sequentially in normal ontogeny and in bone marrow-reconstituted thymectomized radiation chimeras. In the latter, virtually all CD3⁺CD8⁺αα IEL expressed the γδ T cell receptor (TCR), but not the αβ TCR. From this and what is presently known about T cell development, we propose that CD3^?CD8⁺ IEL are an intermediate in extrathymic IEL development and that the development of extrathymically derived IEL occurs at the intestinal epithelium from CD3^?CD8^? to CD3^?CD8⁺ to CD3⁺(γδ TCR)CD8⁺αα.     

Characterization of CD4−CD8− T cell receptor αβ+ T cells appearing in the subarachnoid space of rats with autoimmune encephalomyelitis

Inflammation of the central nervous system (CNS) in experimental autoimmune encephalomyelitis (EAE) starts in the subarachnoid space (SAS) and spreads later to the adjacent CNS parenchyma. To characterize the nature of lesion-forming T cells in situ in more detail, T cells were isolated from the SAS and their surface phenotype and the nucleotide sequence of the junctional region of the T cell receptor (TCR) was determined and compared with those of the lymph node (LN) and spinal cord (SC) T cells. Characteristically, more than 70% of SAS TCR αβ⁺ T cells isolated at the early stage of EAE lacked both CD4 and CD8 molecules, whereas those from LN and SC were either CD4⁺ or CD8⁺. Analysis of nucleotide sequences of the junctional region of TCR revealed that T cells bearing a sequence identical to that for encephalitogenic T cell clones were found in both SAS and SC. Furthermore, purified CD4^?CD8^? T cells expressed CD4 molecules after culture. At the same time, these T cells acquired reactivity to myelin basic protein and induced passive EAE in naive animals after adoptive transfer. Our results suggest that CD4^?CD8^? T cells in the SAS are precursors of lesion-forming T cells in the SC and that phenotype switching takes place during the process of T cell infiltration into the CNS parenchyma. The double-negative nature of these T cells may explain an escape of encephalitogenic T cells from negative selection in T cell differentiation.     

Expression of pro-inflammatory cytokines by TCRαβ+ T and TCRγδ+ T cells in an experimental model of colitis

An inflammatory bowel disease (IBD) comparable to human ulcerative colitis is induced upon transfer of T cell-depleted wild-type (F1) bone marrow into syngeneic T cell-deficient (tgε26) mice (F1 → tgε26). Previously we have shown that activated CD4⁺ T cells predominate in transplanted tgε26 mice, and adoptive transfer experiments verified the potential of these cells to cause disease in immunodeficient recipient mice. Using flow cytometry for the detection of intracellular cytokine expression, we demonstrate in the present study that large numbers of CD4⁺ and CD8⁺ TCRαβ⁺ T cells from the intraepithelial region and lamina propria of the colon of diseased, but not from disease-free mice, produced interferon-γ (IFN-γ) and tumor necrosis factor-α (TNF-α). Large numbers of T cells from peripheral lymphoid tissues of these animals also expressed IFN-α and TNF-α, but few expressed interleukin-4, demonstrating g strong bias towards Th1-type T cell responses in these animals. TCRγδ⁺ T cells, typically minor constituents of the inflammatory infiltrate of the colon in F1 → tgε26 mice, also expressed IFN-γ at a high frequency upon CD3 stimulation. In light of these findings we examined the potential involvement of TCRγδ⁺ T cells by testing their ability to induce colitis in tgε26 mice. We report here that tgε26 mice transplanted with T cell-depleted bone marrow from TCRα^null and TCRβ^null animals developed IBD. Furthermore, disease in these mice correlated with the development of peripheral and colonic TCRαδ⁺ T cells capable of IFN-γ production. These results suggest that IFN-γ may be a common mediator of IBD utilized by pathogenic T cells of distinct phenotype.     

Control of Schistosoma mansoni egg‐induced inflammation by IL‐4‐responsive CD4+CD25−CD103+Foxp3− cells is IL‐10‐dependent

Host protection to helminth infection requires IL‐4 receptor α chain (IL‐4Rα) signalling and the establishment of finely regulated Th2 responses. In the current study, the role of IL‐4Rα‐responsive T cells in Schistosoma mansoni egg‐induced inflammation was investigated. Egg‐induced inflammation in IL‐4Rα‐responsive BALB/c mice was accompanied with Th2‐biased responses, whereas T‐cell‐specific IL‐4Rα‐deficient BALB/c mice (iLck^creIl4ra^?/lox) developed Th1‐biased responses with heightened inflammation. The proportion of Foxp3⁺ Treg in the draining LN of control mice did not correlate with the control of inflammation and was reduced in comparison to T‐cell‐specific IL‐4Rα‐deficient mice. This was due to IL‐4‐mediated inhibition of CD4⁺Foxp3⁺ Treg conversion, demonstrated in adoptively transferred Rag2^?/? mice. Interestingly, reduced footpad swelling in Il4ra^?/lox mice was associated with the induction of IL‐4 and IL‐10‐secreting CD4⁺CD25^?CD103⁺Foxp3^? cells, confirmed in S. mansoni infection studies. Transfer of IL‐4Rα‐responsive CD4⁺CD25^?CD103⁺ cells, but not CD4⁺CD25^high or CD4⁺CD25^?CD103^? cells, controlled inflammation in iLck^creIl4ra^?/lox mice. The control of inflammation depended on IL‐10, as transferred CD4⁺CD25^?CD103⁺ cells from IL‐10‐deficient mice were not able to effectively downregulate inflammation. Together, these results demonstrate that IL‐4 signalling in T cells inhibits Foxp3⁺ Treg in vivo and promotes CD4⁺CD25^?CD103⁺Foxp3^? cells that control S. mansoni egg‐induced inflammation via IL‐10.     

Encephalitogenic,myelin basic protein-specific T cells from naive rat thymus: preferential use of the T cell receptor gene Vβ8.2 and expression of the CD4− CD8− phenotype

Using a primary limiting dilution approach to generate T cell lines, we compared myelin basic protein (MBP)-specific T cell clones from naive unprimed Lewis rat thymuses with the corresponding T cell repertoire of primed rats. We found that in the naive thymus repertoire MBP-specific, encephalitogenic T cell clones preferentially use T cell receptor Vβ8.2 genes, along with CDR3 sequences typical for the primed Lewis anti-MBP response. In contrast to T cells from primed immune organs, which all display the CD4⁺ CD8^? phenotype, the majority of naive thymus-derived T cell clones expressed reduced levels of the CD4 co-receptor. Some clones were completely CD4^?CD8^?, while others included CD4^? CD8^? subpopulations along with CD4⁺CD8^? T cells. In the one mixed population examined in detail, the CD4^?CD8^? and CD4⁺CD8^? T cell subpopulations used a T cell receptor with identical β chain sequence. The data suggest that in the Lewis rat the biased T cell receptor gene usage by encephalitogenic T cells is a property of the natural thymic T cell repertoire, possibly as a consequence of positive selection. The unusually low expression of CD4 in the major histocompatibility complex class II-restricted autoreactive T cells could be related to their escape from negative selection within the thymus.