Functional and ontogenetic analysis of murine CD45Rhi and CD45Rlo CD4+ T cells

CD4+ murine T cell clones, TH1 and TH2, can be distinguished by both functional responses and by their patterns of lymphokine secretion. Recently, a mAb, 23G2, which reacts with a subset of CD45 molecules (CD45R), has been reported to bind differentially to clones of TH1 and TH2 cells. In the present study, normal splenic T cells were analyzed for differences in 23G2-reactivity and were separated into two populations based on their density of CD45R (CD45Rhi and CD45Rlo). The CD45Rhi cells secrete more IL-2 than IL-4 after stimulation in vitro; the reverse is true for the CD45Rlo cells. Because neither population secretes only IL-2 or IL-4, we were unable to classify cells as TH1 or TH2. In vivo and in vitro analyses of the CD45Rhi and CD45Rlo cells suggest a lineage relationship between the two subsets that correlates with the degree of Ag exposure and the state of maturation of the mice. In newborn mice and mice raised under sterile conditions, splenic CD4+ T cells are predominantly CD45Rhi. Under conditions of increased antigenic exposure and maturation of the mice, CD45Rlo cells develop; after long term priming in vivo, the majority of specific Ag-reactive cells are CD45Rlo. Adoptive transfer studies using BALB/c nu/nu recipients demonstrate that CD45Rhi cells become CD45Rlo cells and that the recall response (IgG) to specific Ag is mediated by CD45Rlo cells. Taken together, these data indicate that the level of expression of CD45R on CD4+ T cells distinguishes virgin (CD45Rhi) from primed/memory (CD45Rlo) T cells in normal mice.     

Activation requirements for CD4+ T cells differing in CD45R expression.

Murine CD4+ T cells can be subdivided into naive and memory T cells based on surface phenotype, on recall response to Ag, and on differences in activation requirements. Furthermore, several studies have shown that two signals are required for CD4+ T cell activation; one signal is provided by occupancy of the TCR and the other signal is provided by the APC. In this report, analysis of naive and memory CD4 T cells, separated on the basis of CD45 isoform expression, has shown that their requirements for two signals differ. Activation of memory CD4 T cells to proliferate and secrete IL-2/IL-4 only required occupancy of the TCR complex, whereas activation of naive CD4 T cells required an APC-derived signal as well. Moreover, the signal induced by anti-CD3 antibodies differs from the signal provided by anti-V beta cross-linking of the TCR because both antibodies activate memory CD4 T cells but only anti-CD3 activates naive CD4 T cells. Together these data suggest that the consequence of stimulation through the TCR/CD3 signal complex differs between memory and naive CD4 T cells.     

CD45 isoform expression on human neonatal T cells: expression and turnover of CD45 isoforms on neonatal versus adult T cells after activation.

Neonatal T cells are phenotypically similar to "naive" T cells from adult donors in the CD45 isoform expression. Despite the phenotypic similarity, large differences were found between neonatal and adult T cells when T cells were activated. After activation with PHA, adult CD45RA+ T cells began to express CD45RO and no loss of CD45RA expression had yet occurred at Day 3 post-stimulation. Three days after activation, CD45RA+ neonatal T cells also coexpressed CD45RO; however, in contrast to adult T cells, a marked loss of CD45RA was observed. We analyzed the rapid loss of CD45RA found in neonatal T cells. The de novo synthesis of CD45 isoforms in neonatal T cells was essentially the same as that in the adult T cells. Turnover of the CD45RA was very rapid in both resting adult and neonatal T cells. After activation with PHA, the turnover of CD45RA on adult T cells was decreased significantly, while the turnover of CD45RA on neonatal T cells was not changed after activation. Therefore, the regulation of CD45 isoform expression not only involves switches in alternative splicing, but also involves different regulation of turnover of these isoforms from the cell membrane.     

The regulation of CD5 expression in murine T cells

Background  

CD5 is a pan-T cell surface marker that is also present on a subset of B cells, B-1a cells.Functional and developmental subsets of T cells express characteristic CD5 levels that vary over roughly a 30-fold range. Previous investigators have cloned a 1.7 Kb fragment containing the CD5 promoter and showed that it can confer similar lymphocyte-specific expression pattern as observed for endogenous CD5 expression.     

Distribution and ontogeny of CD2 expression by murine T cells

We have raised a polyclonal antiserum to murine CD2 by immunization of a rabbit with a synthetic peptide corresponding to a hydrophilic sequence in the extracellular domain of the murine CD2 gene. The antiserum immunoprecipitates a 55 kDa protein, consistent with the size predicted by the cDNA sequence. Flow microfluorometric analysis of a panel of T cell tumors and clones demonstrated concordance of reactivity of intact cells with the anti-CD2 serum and the presence of CD2 mRNA. Surprisingly, although splenic T cells were found to uniformly express high levels of CD2, several of a panel of functional T cell clones were found to lack CD2 expression. This suggests that the clones lost CD2 upon in vitro cultivation, and may not be required for activation or maintenance in culture. Adult thymocytes exhibited heterogeneous expression of CD2. The majority of CD4-8- thymocytes expressed low levels and CD4+8+ thymocytes intermediate levels, whereas all CD4+8- and the majority of CD4-8+ thymocytes expressed high levels of CD2. Multiparameter analysis of CD2 expression and that of CD3, CD5, JIId, and IL-2R p55 chain showed that expression of CD2 correlates with the maturational state of thymocytes. Finally, analysis of fetal thymuses from timed pregnancies revealed that expression of CD2 is preceded by that of IL-2R p55 chain.     

CD45 expression by B cells. Expression of different CD45 isoforms by subpopulations of activated B cells.

To determine the effect of distinct activation stimuli on CD45 expression by B cells, we have examined the expression of CD45 molecules on murine B cells stimulated with LPS or the Th cell cytokine IL-5. Analysis of CD45 by flow cytometry revealed that unstimulated and stimulated B cells expressed homogeneous amounts of total CD45 but that stimulation with IL-5 resulted in a CD44hi, hyaluronate-adherent subpopulation of activated B cells that expressed a markedly altered pattern of expression of exon-specific CD45R or B220 determinants. The predominant CD45 immunoprecipitated from either unstimulated or LPS-stimulated B cells was of the high molecular mass form (approximately 220 kDa) usually associated with B cells. In contrast, the CD45 proteins immunoprecipitated from the hyaluronate-adherent subpopulation of IL-5-activated B cells were predominantly lower m.w. forms. PCR analysis of amplified CD45 cDNA also showed distinct expression profiles characteristic of each B cell population. The highest molecular size PCR product, corresponding to expression of all three variably expressed CD45 exons (A, B, and C) was prominent in resting B cells and in LPS-activated B cells but was selectively reduced in hyaluronate-adherent IL-5-activated B cells, where lower molecular size PCR products predominated, corresponding to expression of one or two of the variable exons. In contrast, LPS-activated B cells expressed reduced levels of these one- or two-exon forms. In addition, all B cell populations expressed a lower m.w. PCR product corresponding in size to the product expected when exons A, B, and C are spliced out of CD45 mRNA. Thus, analysis of alternative splicing of CD45 mRNA, as well as cell surface expression of CD45 provides a novel parameter for analysis of B cell activation by different stimuli.     

Reversal of CD45R isoform switching in CD8+ T cells.

Both CD4+ and CD8+ T cells express either CD45RA or CD45R0 isoform of CD45R in an exclusive way. Recent reports have shown that CD45RA+ T cells lose CD45RA and gain CD45R0 upon activation. This switching has been suggested to be irreversible although more recently, examples of reversal of CD45R isotype switching in CD4+ T cells have been reported. We report here that freshly isolated unprimed CD8+ T cells, when activated with PHA, temporarily lose CD45RA but reexpress an intermediate level of CD45RA 2-3 weeks after activation with PHA. This reversal seems to take place much more slowly in unprimed CD4+ T cells: the majority of CD4+ T cells that had lost CD45RA and gained CD45R0 remained CD45RA-CD45R0+ in 3 weeks after the stimulation. Also, long-term CD8+ CD45RA+ T cell lines stimulated with PHA or OKT3 showed even more rapid recovery of CD45RA while PPD-specific CD4+ T cell clones retained the original CD45R0 phenotype 3 weeks after stimulation with PPD or PHA.     

Developmental sequence of T200 antigen modifications in murine T cells

The T200 glycoproteins of T cells were analyzed at different stages of T cell development. Immunoprecipitation and analysis by sodium dodecyl sulfate-polyacrylamide gel electrophoresis revealed that Lyt-2-L3T4-, and Lyt-2+L3T4+ thymocytes had similar T200 proteins, whereas Lyt-2+L3T4- and Lyt-2-L3T4+ thymocytes expressed a distinct set of T200 molecules. This result indicated a molecular switch in regulation of T200 protein expression upon differentiation of thymocytes to mature phenotype T cells. Further modifications were evident when the T200 proteins of peripheral T cell subsets were examined. In particular L3T4+ T cells expressed T200 proteins of m.w. 220,000, 200,000, and 175,000, whereas Lyt-2+ lymph node T cells expressed an additional T200 protein of m.w. 235,000. Antigenic differences in the T200 glyco-proteins of peripheral L3T4+ and Lyt-2+ T cells were also detected. The anti-B220 monoclonal antibody, 14.8, reacted with lymph node Lyt-2+ T cells but did not react with lymph node L3T4+T cells or with Lyt-2+L3T4- thymocytes. This finding demonstrated a lineage-specific modification of the T200 protein of Lyt-2+ T cells that occurred after exit of these cells from the thymus into peripheral lymphoid organs. This modification apparently occurred on the m.w. 235,000 and 220,000 proteins since these species were precipitated by 14.8, whereas the others were not. In vitro growth and activation also resulted in further T200 antigen alterations. The monoclonal antibody, RA3, which reacts with the B220 antigen of B cells but, unlike 14.8, does not react with any peripheral T cells, showed significant reactivity with Lyt-2+ cytotoxic T cell (CTL) clones but not with L3T4+ T helper cell clones. CTL clones were also 14.8+ but T helper cell clones were not. Immunoprecipitation by 14.8 and RA3 of T200 proteins from CTL clones yielded a single protein of m.w. 240,000 that co-migrated with the B cell form of T200. Overall, the results indicate the presence of developmentally regulated mechanisms that control T200 glycoprotein expression during T cell differentiation in the thymus and in peripheral lymphoid organs.     

Activation of T cells through a T cell-specific epitope of CD45.

The 180- and 190-kDa isoforms of CD45 are preferentially expressed on the helper inducer (memory) subset of CD4 cells. In order to generate monoclonal antibodies against the extracellular domains of these isoforms and determine whether they could regulate the function and activation of these cells, we developed a mAb, anti-4H2D, by immunizing Balb/c mice with an isogenic mouse pre-B cell line expressing the human 190-kDa CD45 isoform. Anti-4H2D reacts with approximately 60% of T cells, 70% of CD4 cells, and 60% of CD8 cells. The CD4 cell population defined by this mAb corresponds functionally and phenotypically to that defined by the CD45RO+CD29+ subset. Western blotting demonstrated that anti-4H2D reacts primarily with the 190-kDa isoform of CD45 and to a minor extent, the 205- and 180-kDa CD45 isoforms. Interestingly, this mAb reacted with only a subpopulation of mature thymocytes and peripheral T cells, despite the fact that the 190-kDa CD45 isoform, as well as CD45RO and CD29, is more widely distributed on cells of hematopoietic origin. The 4H2D epitope was neuraminidase sensitive, indicating that anti-4H2D reacts with a carbohydrate epitope which is present on only a subset of the T cells containing the 190-kDa CD45 isoform epitopes. Functional studies showed that soluble anti-4H2D augmented T cell proliferation induced by the CD2 and CD3 pathways, and treatment of T cells with this mAb up-regulated [Ca2+]i flux induced by both anti-CD2 and anti-CD3 mAbs. These results suggest that the 190-kDa CD45 isoform on human CD4 cells is heterogeneous and that the 190-kDa isoform recognized by anti-4H2D regulates the function and activation of CD4 helper T cells.     

Limiting dilution analysis of CD45Rhi and CD45Rlo T cells: further evidence that CD45Rlo cells are memory cells

Murine CD4+ T cells can be separated into two distinct populations on the basis of their levels of expression of the CD45RB antigen (CD45Rhi and CD45Rlo). Murine CD45Rlo cells arise from CD45Rhi cells after antigenic exposure and provide antigen-specific help to B cells in a secondary immune response. In the present study, the ability of CD45Rhi and CD45Rlo cells to proliferate in response to either soluble antigen or allogeneic cells was examined by limiting dilution analysis. CD45Rhi cells were the major responding cells in unprimed animals; priming caused a large increase in the frequencies of responding CD45Rlo cells and this increase was evident 11 months later. These results further support the notion that CD4+ CD45Rlo cells are long-term memory cells.     

Regulation of the CD2 alternate pathway of T cell activation by CD3. Evidence for heterologous desensitization

Normal resting T cells were stimulated through the alternate CD2 pathway. A CD3 mAb VIT3 completely blocked their proliferative response. The time interval for 50% inhibition lasted for 24 h after the onset of CD2 stimulation. Mitogen-activated cloned long term cultured T cells could also be stimulated via CD2. This proliferative response was again inhibitable by VIT3, indicating that CD3 regulates the CD2 pathway not only in resting cells, but also in lymphocytes actively involved in an Ir. T cells were further loaded with Quin2 and their free cytoplasmic Ca2+ levels were monitored in response to CD3 and CD2 stimulation. Antibodies directed against both surface R triggered a rapid elevation of Ca2+ levels. Both responses were abrogated when the cells had been treated overnight with VIT3. The free cytoplasmic Ca2+ levels of VIT3-pretreated cells, however, were not higher than those of control cells. These results point to a functional interaction between CD3 and CD2 possibly at the level of signal transducing proteins. Finally, cholera toxin was found to inhibit the Ca2+ response in Jurkat T cells. Both the CD3 and CD2 stimulation were sensitive to cholera toxin, indicating that a GTP-binding protein may be involved in signal transduction for both surface structures.     

CD4+CD45RA+ and CD4+CD45RA- T cell subsets in man maintain distinct function and CD45RA expression persists on a subpopulation of CD45RA+ cells after activation with Con A.

We have previously shown that Con A-induced suppressor T cells belong to the CD45RA+ subset. After unseparated T cells are activated with Con A, CD45RA expression increases to a maximum (Day 2), and then decreases significantly, but does not disappear entirely (Day 9), while CD29 expression increases steadily. In the present study, we examined the fate of these cell surface molecules on isolated CD4+CD45RA+ and CD4+CD45RA- cells following activation with Con A, and their relationship to the regulatory functions of these subsets. After activation of CD4+CD45RA+ cells with Con A, CD45RO and CD29 antigen expression rapidly increases (greater than 90%). While CD45RA expression is downregulated, approximately 40% of the cells continue to express low-density CD45RA in a stable fashion through Day 21. Despite these phenotypic changes, cells originally CD45RA+ continue to suppress IgG synthesis and provide only minimal B cell help. Furthermore, when cells originally CD45RA+ were sorted on the basis of continued presence, or loss of CD45RA antigen 14 days after activation, both populations demonstrated potent suppression and minimal help. In contrast, after activation with Con A, CD4+CD45A- cells maintain stable phenotype and provide significant help and minimal suppression. Immunoprecipitation of the CD45RA antigen from Day 14 activated CD4+CD45RA+ cells confirms the continued presence of the 205-kDa isoform, but reveals a significant decrease in the 220-kDa isoform. These results suggest that after activation with Con A, cells originally CD45RA+ remain functionally distinct from cells originally CD45RA-, and that CD45RA antigen persists on a subpopulation of CD45RA+ cells after activation with Con A.     

Human virus-specific CD8+ CTL clones revert from CD45ROhigh to CD45RAhigh in vivo: CD45RAhighCD8+ T cells comprise both naive and memory cells.

It has been generally believed that human CD8+ memory cells are principally found within the CD45ROhigh population. There are high frequencies of CD8+ memory CTL specific for the human CMV tegument phosphoprotein pp65 in PBMC of long-term virus carriers; the large population of memory CTL specific for a given pp65 peptide contains individual CTL clones that have greatly expanded. In this study, we found high frequencies of pp65 peptide-specific memory CTL precursors in the CD45ROhighCD45RA- population, but also appreciable frequencies in the CD45RAhigh subpopulation. Because the majority of CD8+ T cells in PBMC are CD45RAhigh, more of the total pp65-specific memory CTL pool is within the CD45RAhigh than in the CD45ROhigh compartment. Using clonotypic oligonucleotide probes to quantify the size of individual pp65-specific CTL clones in vivo, we found the CD45RAhigh population contributed 6- to 10-fold more than the CD45ROhigh population to the total virus-specific clone size in CD8+ cells. During primary CMV infection, an individual virus-specific CTL clone was initially CD45ROhigh, but after resolution of infection this clone was detected in both the CD45ROhigh and the CD45RAhigh populations. We conclude that CD45RA+ human CD8+ T cells do not solely comprise naive cells, but contain a very significant proportion of memory cells, which can revert from the CD45ROhigh to CD45RAhigh phenotype in vivo.