TCR-self peptide:MHC interactions play a critical role in thymic positive selection, yet relatively little is known of their function in the periphery. It has been suggested that continued contact with selecting MHC molecules is necessary for long-term peripheral maintenance of naive T cells. More recent studies have also demonstrated a role for specific self peptide:MHC complexes in the homeostatic expansion of naive T cells in lymphopenic mice. Our examination of these processes revealed that, whereas self class II MHC molecules do have a modest effect on long-term survival of individual CD4+ T cells, interactions with specific TCR ligands are not required for peripheral naive CD4+ T cell maintenance. In contrast, selective engagement of TCRs by self-peptide:MHC complexes does promote proliferation of CD4+ T cells under severe lymphopenic conditions, and this division is associated with an activation marker phenotype that is different from that induced by antigenic stimulation. Importantly, however, the ability of naive T cells to divide in response to homeostatic stimuli does not appear to be stringently dependent on TCR-self peptide:MHC interactions. Therefore, these results show that the factors regulating survival and homeostatic expansion of naive T cells in the periphery are not identical. In addition, we provide evidence for a novel form of T cell proliferation that can occur independently of TCR signaling and suggest that this reflects another mechanism regulating homeostatic T cell expansion.

The TCR repertoire is shaped by positive and negative selection of immature T cell precursors in the thymus. Positive selection is mediated by low affinity TCR interactions with self peptide bound to MHC molecules and promotes the survival of self MHC-restricted thymocytes. Negative selection acts to eliminate those thymocytes bearing receptors with high affinity to self peptide:MHC complexes (1, 2). Thymocytes that survive positive and negative selection emigrate into the peripheral lymphoid system as mature naive CD4+ or CD8+ T cells. Most naive T cells appear to be long-lived, quiescent cells (3, 4, 5), but transfer of small numbers of mature T cells into immunodeficient hosts has been known for many years to result in relatively rapid T cell proliferation and functional restoration of the peripheral T cell compartment (6, 7, 8, 9). These early studies demonstrated that the size of the peripheral T cell pool is under homeostatic control; however, the precise mechanisms regulating homeostatic T cell expansion and survival were not defined.

Since then, several reports examining the proliferative response of adoptively transferred H-Y- or Ld-specific CD8+ (4, 10, 11, 12), or pigeon cytochrome c-specific CD4+ (11), TCR transgenic T cells in T cell-deficient hosts claimed that homeostatic expansion of naive peripheral T cells is dependent on Ag. Accordingly, homeostatic proliferation of naive polyclonal T cells was attributed to recognition of environmental Ags (10, 11). However, expansion of naive polyclonal wild-type T cells (13, 14), as well as naive T cells from other (but not all) TCR transgenic lines (15, 16, 17, 18, 19), in Ag-free lymphopenic mice has also been reported, demonstrating that homeostatic T cell proliferation is not necessarily a consequence of antigenic stimulation. Instead, due to the importance of self peptide:MHC complexes in regulating thymocyte positive selection, it was proposed that these weakly binding TCR ligands might control T cell survival and homeostatic proliferation in the periphery (15).

In support of this hypothesis, a number of studies have demonstrated impaired survival of mature CD4+ or CD8+ T cells in animals lacking selecting class II or class I MHC molecules in the periphery, respectively (12, 20, 21, 22, 23, 24, 25). However, the role of specific MHC bound peptides in the fate of peripheral T cells was not addressed. Furthermore, interpretation of these results is complicated by the lack of single cell analyses in individual animals under conditions where the lymphoid population being investigated included both dividing and nondividing cells.

Recent data showing impaired proliferation of adoptively transferred CD4+ or CD8+ T cells in irradiated class II or class I MHC-deficient hosts, respectively (14, 15, 17, 18, 19, 26), also support a role for selecting MHC expression in homeostatic T cell proliferation. Moreover, several groups have shown greatly diminished proliferation of naive T cells upon transfer into T cell-depleted hosts bearing a nonselecting self peptide:MHC repertoire (15, 17, 18). Therefore, these findings endorse the notion that TCR engagement by specific self peptide:MHC complexes, positively selecting ligands in particular, is essential for homeostatic T cell proliferation. Such a conclusion is favorable because the paucity or absence of certain positively selecting ligands in the peripheral lymphoid organs may explain why some, but not all, TCR transgenic T cells fail to proliferate in syngeneic T cell-deficient mice. However, this notion was challenged by another study, which found that only self peptides distinct from those involved in thymic positive selection drive homeostatic CD4+ T cell proliferation (26).

To further investigate the contribution of specific self peptide:MHC complexes to T cell survival vs homeostatic proliferation, we examined the fate of mature CD4+ T cells after adoptive transfer into irradiated or unirradiated wild-type mice, class II-deficient (Aβ0)3 mice, or mice expressing a drastically altered repertoire of class II bound peptides (H-2 M-knockout (H-2 Mα0) mice). Our analysis of individual quiescent naive CD4+ T cells suggests only a modest role for class II MHC molecules in long-term T cell survival and, unlike homeostatic T cell proliferation, shows that this process is independent of specific self peptide:MHC interactions. Furthermore, our results indicate that, whereas homeostatic T cell proliferation occurs more efficiently in the presence of a selecting peptide repertoire, specific TCR-self peptide:MHC interactions are not absolutely required for this response. Examination of activation marker expression also revealed that, during homeostatic proliferation, T cells retain a quasi naive phenotype different to that of T cells dividing in response to antigenic stimulation. Hence, we propose that alternative signals, in addition to those delivered through the TCR, are capable of regulating homeostatic T cell proliferation.

Mice used were 6–20 wk of age. H-2 Mα0, (C57BL/6 (B6) × B6.PL)F1 and TEa TCR transgenic mice (27) were bred and maintained at the University of Washington (Seattle, WA); (B6 × 129/J)F1 mice were purchased from The Jackson Laboratory (Bar Harbor, ME); B6 mice were purchased from Charles River Breeding Laboratories (Wilmington, MA); Aβ0 mice were purchased from Taconic Farms (Germantown, NY).

CD4+ and CD8+ T cells were purified from the lymph nodes and spleens of wild-type or TEa TCR transgenic mice by treatment with a mixture of anti-HSA (J11d) and anti-class II (25-9-17s and BP107) Ab supernatants followed by rabbit complement (C-SIX Diagnostics, Mequon, WI). Cells were then passed over a 20-ml Sephadex G10 (Pharmacia, Piscataway, NJ) column to remove adherent cells. In some experiments, class II+ cells were further depleted by negative panning on anti-I-Ab Ab (Y3P)-coated plates. T cell purity was tested by flow cytometry using FITC-conjugated anti-CD8 (PharMingen, San Diego, CA), PE-conjugated anti-CD4 (PharMingen), and biotin-conjugated anti-CD44 (PharMingen) followed by streptavidin-TRI-COLOR (Caltag, South San Francisco, CA), as well as PE-conjugated anti-B220 (Caltag) and biotin-conjugated anti-I-Ab (Y3P) followed by streptavidin-TRI-COLOR. Stained cells were analyzed on a Becton Dickinson FACScalibur flow cytometer using CellQuest software. The resulting population routinely showed >94% CD4+ plus CD8+, <8% CD44high, and contained ∼1% class II+ (Y3P+) cells, most of which were also B220+. These donor cells were labeled with 5,6-carboxyfluorescein diacetate succinimidyl ester (CFSE) (Molecular Probes, Eugene, OR) as described (28) and injected i.v. into host mice.

For long-term survival studies, 5–6 × 106 CFSE-labeled T cells were injected into sex-matched unirradiated mice. The presence of donor cells in peripheral blood, lymph nodes and spleen was monitored by flow cytometry for up to 12 wk using peridinin chlorophyll protein-conjugated anti-CD4 (PharMingen), APC-conjugated anti-CD8 (PharMingen), and biotin-conjugated anti-CD44 followed by streptavidin-PE (Vector Laboratories, Berlingame, CA).

For homeostatic proliferation studies, 2–3 × 106 CFSE-labeled T cells were injected into sex-matched unirradiated mice or mice that had been irradiated for 600 rad 1–2 days before injection. Then, 7, 14, 21, and 28 days later, the peripheral blood, lymph nodes, and spleen were analyzed by flow cytometry using peridinin chlorophyll protein-conjugated anti-CD4 with APC-conjugated anti-CD8 (see Figs. 2 and 3), with PE-conjugated anti-Vα2 (PharMingen) (see Fig. 4), or with PE-conjugated anti-Thy-1.1 (PharMingen) (see Fig. 5). In some studies, the phenotype of the donor T cells was also assessed using biotin-conjugated anti-CD44, anti-CD69, anti-CD25 (IL-2R α-chain), anti-CD62L, or anti-CD45RB (all PharMingen) followed by streptavidin-APC (Caltag). To test for the presence of contaminating class II+ donor APCs in the Aβ0 hosts on day 28, spleen cells were stained with a combination of PE-conjugated anti-B220 plus biotin-conjugated anti-I-Ab (Y3P) followed by streptavidin-TRI-COLOR.

FIGURE 2.

Homeostatic proliferation of naive CD4+ T cells can occur independently of a selecting peptide repertoire. A total of 2–3 × 106 CFSE-labeled wild-type (B6 × 129/J)F1 CD4+ and CD8+ T cells were injected i.v. into wild-type (a–c) or H-2 Mα0 (d–f) mice. These mice were either unirradiated (a and d), or irradiated 600 rad 2 days before injection (b, c, e, and f). Then, 7 (a, b, d, and e) or 28 (c and f) days after injection, cells from the lymph nodes, spleen, and peripheral blood of individual mice were analyzed by flow cytometry. Fig. 2 shows CFSE fluorescence of electronically gated CD4+ T cells (upper panel) or CD8+ T cells (lower panel) from the lymph node cells. Similar profiles were seen in the spleen and peripheral blood. One to three mice per group were analyzed in two similar experiments.

FIGURE 2.

Homeostatic proliferation of naive CD4+ T cells can occur independently of a selecting peptide repertoire. A total of 2–3 × 106 CFSE-labeled wild-type (B6 × 129/J)F1 CD4+ and CD8+ T cells were injected i.v. into wild-type (a–c) or H-2 Mα0 (d–f) mice. These mice were either unirradiated (a and d), or irradiated 600 rad 2 days before injection (b, c, e, and f). Then, 7 (a, b, d, and e) or 28 (c and f) days after injection, cells from the lymph nodes, spleen, and peripheral blood of individual mice were analyzed by flow cytometry. Fig. 2 shows CFSE fluorescence of electronically gated CD4+ T cells (upper panel) or CD8+ T cells (lower panel) from the lymph node cells. Similar profiles were seen in the spleen and peripheral blood. One to three mice per group were analyzed in two similar experiments.

Close modal
FIGURE 3.

Homeostatic proliferation of naive CD4+ T cells can occur independently of MHC class II expression. A total of 2–3 × 106 CFSE-labeled wild-type CD4+ and CD8+ T cells were injected i.v. into B6 (a–c) or Aβ0 (d–f) mice. These mice were either unirradiated (a and d) or irradiated 600 rad 2 days before injection (b, c, e, and f). The data shown are as described in Fig. 2. One to four mice per group were analyzed in three similar experiments.

FIGURE 3.

Homeostatic proliferation of naive CD4+ T cells can occur independently of MHC class II expression. A total of 2–3 × 106 CFSE-labeled wild-type CD4+ and CD8+ T cells were injected i.v. into B6 (a–c) or Aβ0 (d–f) mice. These mice were either unirradiated (a and d) or irradiated 600 rad 2 days before injection (b, c, e, and f). The data shown are as described in Fig. 2. One to four mice per group were analyzed in three similar experiments.

Close modal
FIGURE 4.

The phenotype of naive CD4+ T cells undergoing homeostatic proliferation is different from the phenotype of T cells dividing in response to antigenic stimulation. a, B6 mice were given 5 × 106 CFSE-labeled Eα52–68/I-Ab-specific TCR transgenic CD4+ T cells i.v., and, the next day, were injected i.p. with PBS in CFA (left) or 100 μg Eα52–68 peptide in CFA (right). Forty-two hours after immunization, spleen cells from each group of mice were analyzed by flow cytometry. b, A total of 3 × 106 of the same 106 CFSE-labeled Eα52–68/I-Ab-specific TCR transgenic CD4+ T cells were injected i.v. into unirradiated B6 (left) or B6 mice that had been irradiated 600 rad 2 days earlier (right). Twenty-eight days later, spleen cells from host mice were analyzed by flow cytometry. The data were electronically gated for CD4+ Vα2+ cells, and the expression of CFSE, CD69, CD44, IL-2Rα, CD62L, or CD45RB was examined. Similar results were obtained with lymph node cells. One to four mice per group were analyzed in two similar experiments.

FIGURE 4.

The phenotype of naive CD4+ T cells undergoing homeostatic proliferation is different from the phenotype of T cells dividing in response to antigenic stimulation. a, B6 mice were given 5 × 106 CFSE-labeled Eα52–68/I-Ab-specific TCR transgenic CD4+ T cells i.v., and, the next day, were injected i.p. with PBS in CFA (left) or 100 μg Eα52–68 peptide in CFA (right). Forty-two hours after immunization, spleen cells from each group of mice were analyzed by flow cytometry. b, A total of 3 × 106 of the same 106 CFSE-labeled Eα52–68/I-Ab-specific TCR transgenic CD4+ T cells were injected i.v. into unirradiated B6 (left) or B6 mice that had been irradiated 600 rad 2 days earlier (right). Twenty-eight days later, spleen cells from host mice were analyzed by flow cytometry. The data were electronically gated for CD4+ Vα2+ cells, and the expression of CFSE, CD69, CD44, IL-2Rα, CD62L, or CD45RB was examined. Similar results were obtained with lymph node cells. One to four mice per group were analyzed in two similar experiments.

Close modal
FIGURE 5.

CD4+ T cells undergoing homeostatic proliferation in irradiated B6 or Aβ0 mice retain a naive phenotype. B6 (a) or Aβ0 (b) mice (Thy-1.2+) were either left unirradiated (left), or irradiated 600 rad (right), and then injected i.v. 2 days later with 3 × 106 CFSE-labeled CD4+ T cells purified from the lymph nodes and spleens of (B6 × B6. PL)F1 mice (Thy-1.1+, Thy-1.2+). Twenty-eight days after injection, host spleen cells were analyzed by flow cytometry. Cells were electronically gated on CD4+ Thy-1.1+ donor cells, and the expression of CFSE, CD69, CD44, IL-2Rα, CD62L, or CD45RB was examined. Similar results were obtained with lymph node cells. One to four mice per group were analyzed in three similar experiments.

FIGURE 5.

CD4+ T cells undergoing homeostatic proliferation in irradiated B6 or Aβ0 mice retain a naive phenotype. B6 (a) or Aβ0 (b) mice (Thy-1.2+) were either left unirradiated (left), or irradiated 600 rad (right), and then injected i.v. 2 days later with 3 × 106 CFSE-labeled CD4+ T cells purified from the lymph nodes and spleens of (B6 × B6. PL)F1 mice (Thy-1.1+, Thy-1.2+). Twenty-eight days after injection, host spleen cells were analyzed by flow cytometry. Cells were electronically gated on CD4+ Thy-1.1+ donor cells, and the expression of CFSE, CD69, CD44, IL-2Rα, CD62L, or CD45RB was examined. Similar results were obtained with lymph node cells. One to four mice per group were analyzed in three similar experiments.

Close modal

B6 mice were injected i.v. with 5 × 10652–68/I-Ab-specific TCR transgenic CD4+ T cells purified from the lymph nodes of TEa mice. The next day, mice were injected i.p. with 100 μg Eα52–68 peptide emulsified in CFA (Sigma, St. Louis, MO). Control mice were given PBS emulsified in CFA. Then, 42 h after in vivo priming, the lymph node and spleen cells from individual mice were analyzed by flow cytometry, and the phenotype of the donor CD4+ T cells was assessed.

To investigate the importance of class II MHC molecules for peripheral CD4+ T cell survival, 5–6 × 106 naive CD4+ and CD8+ T cells from wild-type B6 mice were injected i.v. into unirradiated wild-type or Aβ0 mice. The CD8+ T cells cotransferred with the CD4+ T cells served as an important internal control population for survival within individual hosts. To assist detection of the donor T cells, the CD4+ and CD8+ T cells were labeled in vitro with the cytoplasmic fluorescent dye CFSE. Proliferation of the transferred T cells, if any, could also be assessed because division of CFSE-labeled cells is associated with a 2-fold reduction in CFSE fluorescence in both daughter cells (28). Survival of the undivided donor CFSE+ CD4+ and CD8+ T cells was determined by flow cytometric analysis of the peripheral blood of the host mice. Thus, this approach allowed us to determine the fate of individual quiescent cells within the donor T cell population.

As shown in Fig. 1 a, comparable survival of undivided donor CD4+ T cells relative to CD8+ T cells was observed in wild-type and Aβ0 mice for up to 42 days. However, by day 56, a slight reduction in the relative percentage of CD4+ T cells was detected in the Aβ0 mice. Impaired survival of CD4+ T cells in Aβ0 mice was also reflected in the average number of undivided donor CD4+ T cells present in the spleens of host mice on day 56: wild-type mice = 1.57 ± 0.52 × 105 (n = 5) and Aβ0 mice = 0.50 ± 0.10 × 105 (n = 4). Very little proliferation of either CD8+ or CD4+ T cells was detected over this period, and the cells retained a CD44low phenotype (data not shown). Therefore, the data imply a role for class II MHC molecules in CD4+ T cell maintenance. However, because the average life span of peripheral naive T cells in wild-type mice has been reported as 5–8 wk (4, 5), the presence of class II MHC molecules may only be necessary to prolong survival at a time when the CD4+ T cells would normally enter cell cycle or die.

FIGURE 1.

Class II MHC molecules, but not positively selecting class II-bound peptides, are required for long-term survival of naive peripheral CD4+ T cells. A total of 5–6 × 106 CFSE-labeled wild-type B6 (a) or (B6 × 129/J)F1 (b) CD4+ and CD8+ T cells were injected i.v. into B6 (open symbols) and Aβ0 (closed symbols) (a) or (B6 × 129/J)F1 (open symbols) and H-2 Mα0 (closed symbols) host mice (b), respectively. Peripheral blood was taken at regular intervals and analyzed by flow cytometry for the presence of CFSE+ CD4+ and CD8+ donor T cells. Insets show the quadrant gates used to determine the percentage of undivided CD8+ CFSE+ T cells. Similar quadrants were used for CD4 × CFSE plots. Individual data points represent the ratio of the percent undivided CFSE+ CD4+ cells to the percent undivided CFSE+ CD8+ cells in the peripheral blood of individual animals from two (a) or three (b) separate experiments (circles, triangles, and squares).

FIGURE 1.

Class II MHC molecules, but not positively selecting class II-bound peptides, are required for long-term survival of naive peripheral CD4+ T cells. A total of 5–6 × 106 CFSE-labeled wild-type B6 (a) or (B6 × 129/J)F1 (b) CD4+ and CD8+ T cells were injected i.v. into B6 (open symbols) and Aβ0 (closed symbols) (a) or (B6 × 129/J)F1 (open symbols) and H-2 Mα0 (closed symbols) host mice (b), respectively. Peripheral blood was taken at regular intervals and analyzed by flow cytometry for the presence of CFSE+ CD4+ and CD8+ donor T cells. Insets show the quadrant gates used to determine the percentage of undivided CD8+ CFSE+ T cells. Similar quadrants were used for CD4 × CFSE plots. Individual data points represent the ratio of the percent undivided CFSE+ CD4+ cells to the percent undivided CFSE+ CD8+ cells in the peripheral blood of individual animals from two (a) or three (b) separate experiments (circles, triangles, and squares).

Close modal

To test whether positively selecting self peptides were also important for long-term peripheral CD4+ T cell survival, naive CD4+ and CD8+ T cells from wild-type (B6 × 129/J)F1 mice were labeled with CFSE and injected i.v. into unirradiated wild-type or H-2 Mα0 mice on a mixed B6/129/J background. In the absence of H-2 M, the majority of class II MHC molecules are bound to CLIP peptide. Despite this, the overall cell surface class II expression in H-2 Mα0 mice is equivalent to that of wild-type mice and MHC class I presentation is normal (29, 30, 31). Importantly, the TCR repertoires positively selected in wild-type and H-2 Mα0 mice are largely nonoverlapping (27, 32, 33, 34). As shown in Fig. 1 b, the maintenance of undivided donor CD4+ T cells relative to donor CD8+ T cells over a 60-day period was similar in wild-type and H-2 Mα0 mice. In one experiment, transferred cells were monitored for up to 12 wk without any difference in the relative survival of CD4+ T cells in wild-type vs H-2 Mα0 hosts (data not shown). Analysis of splenic T cell populations was again consistent with the peripheral blood results. In two separate experiments, the average number of undivided donor CD4+ T cells in the spleen on day 60 was 2.17 ± 0.20 × 105 (n = 5) and 1.34 ± 0.86 × 105 (n = 2) for wild-type mice, and 2.04 ± 0.47 × 105 (n = 7) and 1.30 ± 0.85 × 105 (n = 2) for H-2 Mα0 mice, respectively. Thus, in contrast to class II MHC molecules, it appears that a selecting peptide repertoire is not required for peripheral maintenance of naive CD4+ T cells.

The role of a diverse peptide:MHC repertoire in homeostatic T cell expansion was investigated by injecting 2–3 × 106 CFSE-labeled CD4+ and CD8+ T cells into host mice whose lymphoid compartments were acutely ablated by sublethal irradiation (600 rad) 1–2 days earlier (Fig. 2). As expected, transfer of naive T cells from wild-type (B6 × 126/J)F1 mice into irradiated, syngeneic hosts resulted in significant cell division of donor CD4+ and CD8+ T cells in the lymph nodes (Fig. 2,b), spleen, and peripheral blood (data not shown) by day 7. However, unlike the rapidly induced proliferative response to Ag (see below), division of donor CD8+ and CD4+ T cells was not usually observed until day 3 posttransfer (data not shown). Homeostatic proliferation in irradiated wild-type hosts was monitored over 4 wk (data not shown) and, by day 28, the majority of the transferred CD4+ and CD8+ T cells had undergone cell division (Fig. 2,c). No proliferation was observed following transfer into unirradiated hosts (Fig. 2, a and d).

When the same CFSE-labeled CD4+ and CD8+ T cells were transferred into irradiated H-2 Mα0 hosts, only marginal CD4+ T cell proliferation was observed on day 7 (Fig. 2,e). This diminished proliferation did not result from an intrinsic inability of T cells to proliferate in the H-2 Mα0 mice because the control population of donor CD8+ T cells divided normally (Fig. 2,e). In agreement with other recent studies, this result appeared to support the idea that positively selecting or antagonist peptides are stringently required to drive homeostatic proliferation (15, 17, 18). However, when proliferation of adoptively transferred CD4+ T cells was examined in irradiated H-2 Mα0 mice beyond day 7, a significant amount of CD4+ T cell division was observed. By day 28, a large proportion of the donor CD4+ T cells had divided at least once in irradiated H-2 Mα0 mice, with many cells having undergone two or more cell divisions (Fig. 2,f). Similar results were generated using Eα52–68:I-Ab-specific TCR transgenic CD4+ T cells from TEa mice that are not selected on an H-2 Mα0 background (data not shown). This demonstrates that proliferation of the wild-type polyclonal CD4+ T cells in irradiated H-2 Mα0 hosts on day 28 (Fig. 2 f) was not simply limited to those cells that may have been positively selected on CLIP:I-Ab. Therefore, these latter results suggest that the presence of selecting self peptide:MHC complexes is not absolutely required to initiate homeostatic proliferation of CD4+ T cells.

To determine whether homeostatic CD4+ T cell proliferation can occur in the absence of any class II MHC-derived signal, similar experiments were performed using irradiated Aβ0 or wild-type hosts. Once again, proliferation was extensive for both CD4+ and CD8+ T cell subsets in 600-rad wild-type hosts on days 7 and 28 (Fig. 3, b and c). As seen for H-2 Mα0 mice, homeostatic expansion of CD8+ T cells occurred normally in irradiated Aβ0 mice on day 7, but donor CD4+ T cell division was barely detectable at this time (Fig. 3,e). Nevertheless, by day 28, a substantial number of the adoptively transferred CD4+ T cells had divided in the irradiated Aβ0 mice (Fig. 3 f). Similar results were generated when Eα52–68:I-Ab-specific TCR transgenic CD4+ T cells were transferred with or without CD8+ T cells into Aβ0 × RAG-I double-deficient host mice (data not shown). Hence, this response is neither an artifact of T cell depletion by sublethal irradiation, nor a bystander effect created by the presence of proliferating host T cells or cotransferred CD8+ T cells. Furthermore, the ability of naive CD4+ T cells to proliferate in the absence of class II MHC molecules implies that cell division was not the result of TCR-mediated recognition of minor histocompatibility or cross-reactive environmental Ags.

In contrast to the critical role of thymic cortical epithelial cells in positive selection, bone marrow-derived cells have been implicated in presenting self peptides to CD4+ T cells undergoing homeostatic proliferation (data not shown). In this regard, it is important to note that homeostatic proliferation of the donor CD4+ T cells in H-2 Mα0 and Aβ0 mice was not likely due to the presence of contaminating class II+ bone marrow-derived APCs in the transferred T cell population. Flow cytometric analysis of the donor population routinely showed >94% T cells with ∼1% class II+ cells, essentially all of which were B cells. No expansion of the class II+ cells was detected in irradiated Aβ0 hosts on day 28 (data not shown). To ensure that homeostatic CD4+ T cell proliferation in the irradiated Aβ0 hosts was not due to the 1% contaminating class II+ cells, CD4+ and CD8+ donor T cells were isolated from the lymph nodes of mice that had been lethally irradiated (950 rad) and reconstituted with Aβ0 bone marrow cells. No class II (Y3P)+ cells were found in the lymph nodes or peripheral blood of these donor mice 7 wk after reconstitution, but normal percentages of CD4+ T cells were detected, presumably as a result of positive selection on class II+ thymic cortical epithelial cells (data not shown). T cells from these bone marrow chimeras were labeled with CFSE and injected i.v. into irradiated wild-type or Aβ0 hosts. Then, 14 days later, one to two rounds of donor CD4+ T cell division were again observed in the lymph nodes and spleen of irradiated Aβ0 hosts (data not shown). Altogether, these results demonstrate that homeostatic naive CD4+ T cell proliferation can occur independently of class II MHC expression, albeit much less efficiently.

Division of naive CD4+ T cells in the absence of class II suggests that homeostatic proliferation can be induced by a non-TCR-derived stimulus. Therefore, the expression of cell surface markers normally associated with TCR-mediated activation was assessed for those CD4+ T cells undergoing cell division in lymphopenic hosts. These markers included the early T cell activation marker CD69, the activation/memory cell marker CD44, IL-2 receptor α-chain (IL-2Rα), the lymph node homing receptor, CD62L, and the memory cell marker, CD45RB.

Initially, we analyzed expression of these markers on T cells undergoing Ag-specific proliferation. CFSE-labeled Eα52–68:I-Ab-specific TCR transgenic CD4+ T cells transferred into unirradiated syngeneic mice were analyzed 2 days after in vivo priming with antigenic peptide in CFA (Fig. 4,a). By this time, many of the TCR transgenic CD4+ T cells had undergone three to four rounds of cell division (Fig. 4,a, right panel), whereas cells exposed to PBS in CFA remained undivided (Fig. 4,a, left panel). CD69 was highly expressed on the undivided primed T cells, illustrating the early up-regulation of this marker upon TCR engagement, and was progressively lost with each subsequent cell division. CD44 expression changed from intermediate to high following the first round of cell division. IL-2Rα expression was low on undivided cells but increased slightly on divided cells. As expected, high levels of CD62L and CD45RB were found on the undivided population of TCR transgenic CD4+ T cells, but expression of these markers progressively decreased with Ag-induced proliferation (Fig. 4 a). Similar dynamics of several of these markers in response to in vivo priming were recently reported for CFSE-labeled DO11.10 TCR transgenic T cells (35).

In contrast to the T cell phenotype characteristic of TCR-induced proliferation, no increase in CD69 expression was observed when CFSE-labeled Eα52–68:I-Ab-specific TCR transgenic CD4+ T cells from the same donor population were transferred into irradiated B6 mice and analyzed up to 28 days later (Fig. 4,b, and data not shown). IL-2Rα levels on the CFSE-labeled TCR transgenic CD4+ T cells from the 600-rad host mice were similar or only very slightly increased compared with the levels observed on donor T cells from the unirradiated hosts. Donor T cells from the irradiated mice also maintained high expression of CD62L and CD45RB as they proliferated (Fig. 4,b). The only change detected during homeostatic T cell proliferation was an increase in CD44 expression. However, unlike Ag-driven proliferation, CD44 up-regulation occurred gradually, requiring at least two to three rounds of cell division before being observed (Fig. 4 b). Thus, unlike T cells undergoing TCR-induced proliferation, the T cells seen dividing in response to homeostatic signals appear to retain an essentially naive phenotype. This confirms that the donor T cells were not dividing in response to environmental or minor histocompatibility Ags.

A similar quasi naive phenotype of dividing T cells was also observed for wild-type polyclonal donor CD4+ T cells 7, 14, and 28 days after transfer into irradiated wild-type mice (Fig. 5,a; data not shown), as well as on days 14 and 28 in irradiated Aβ0 mice (Fig. 5,b; data not shown). However, unlike the TCR transgenic donor CD4+ T cells the wild-type polyclonal donor CD4+ T cells (Thy-1.1+ × Thy-1.2+) expressed a Thy-1 Ag different from the host T cells (Thy-1.2+). This allowed us to distinguish those donor CD4+ T cells that had lost their CFSE as a result of many cell divisions from any host CD4+ T cells that were never labeled with this dye. Interestingly, when the phenotype of the CFSE-negative donor cells in the irradiated hosts was examined, the majority not only expressed high levels of CD44, but were low for CD62L and intermediate for CD45RB (Fig. 5). Such a phenotype is reminiscent of memory CD4+ T cells (36) and may be related to the finding that, after many rounds of homeostatic cell division, naive CD8+ T cells transiently acquire a memory-like phenotype (37). Alternatively or additionally, T cells in this population may be derived from a small memory T cell pool, present in the initial donor population, that divided more rapidly in response to homeostatic stimuli than the naive CD4+ T cells. Such accelerated homeostatic proliferation of memory T cells has recently been documented for CD8+ T cells and has been shown to occur independently of class I MHC expression (38). The presence of a CFSE negative donor T cell population in the unirradiated Aβ0 hosts (Fig. 5,b), but not in the unirradiated wild-type hosts (Fig. 5 a), also suggests that this population may represent an expanded memory population. That is, whereas the presence of naive CD8+ T cells has been shown to inhibit homeostatic proliferation of naive CD4+ T cells (17), naive CD8+ T cells may not be able to inhibit proliferation of memory CD4+ T cells which are generally absent in Aβ0, but not in wild-type, mice.

To address the possibility that the CFSE-negative donor T cells were derived from preexisting memory T cells in our donor population, (Thy-1.1+ × Thy-1.2+) donor T cells were sorted for CD44low and CD69low expression and then injected into unirradiated or 600-rad wild-type and Aβ0 hosts. Sorted donor T cells that had undergone two cell divisions were detected as early as day 7 in irradiated Aβ0 mice and expressed a naive phenotype (CD69low, CD44low, CD25low, CD62Lhigh, CD45RBhigh) (data not shown). Thus, whereas transfer of sorted T cells did reduce the number of CFSE-negative donor T cells in some irradiated Aβ0 mice, it did not completely abolish their appearance, especially in the spleen. In fact, this population appeared to develop more rapidly in irradiated Aβ0 mice than it did in the irradiated wild-type hosts. Similar results were also obtained using sorted CD69high and CD69low CD4 single-positive thymocytes (data not shown). Hence, it is unlikely that the CFSE-negative population simply appeared because sorting for CD44low CD69low expression on mature T cells did not exclude all memory T cells from our donor CD4+ T cell population.

These findings suggest that the CFSE-negative donor T cells detected previously in the Aβ0 mice were not all derived from a small number of contaminating memory T cells, but also represented a subset of naive T cells that divide rapidly and acquire a memory phenotype. In an attempt to view these rapidly dividing T cells at an earlier time point when they had intermediate expression of CFSE, irradiated Aβ0 mice injected with unsorted donor T cells were analyzed 3, 5, 7, and 9 days posttransfer. The CFSE-negative, memory-like population appeared as early as day 5 in both irradiated and unirradiated Aβ0 mice and increased in number thereafter (data not shown). However, the small number of these T cells together with their rapid cell division made it very difficult to detect populations of CFSE intermediate T cells.

Homeostasis of the naive peripheral T cell pool involves both long-term survival of quiescent T cells in a full lymphoid compartment, as well as the ability to expand these T cells upon exposure to a lymphopenic environment. In this study, we have distinguished between the signals that maintain quiescent T cells and those that drive homeostatic T cell proliferation by examining the contribution of both class II MHC molecules and specific peptide repertoires to the response of adoptively transferred naive CD4+ T cells in either unirradiated or sublethally irradiated hosts. Previous reports suggested that naive T cell survival is dependent on expression of selecting MHC molecules (12, 15, 17, 20, 21, 22, 23, 24, 25). The failure of mature CD4+ or CD8+ T cells to accumulate in the periphery of class II-deficient or class I-deficient mice transplanted with class II+ or class I+ thymic epithelium, respectively, has been reported by two groups (23, 25). In other studies, the disappearance of adoptively transferred naive T cells in MHC-deficient hosts was observed within 1–2 wk post transfer (12, 15, 17). In these studies, T cells from B6-background donors were transferred into knockout mice on mixed B6/129 backgrounds. Thus, the rapid loss of T cells may have been due to minor histocompatibility Ag-mediated rejection, which, from our own experience, can occur rapidly even in sublethally irradiated hosts. For other groups, loss of T cells in MHC-deficient hosts occurred gradually, requiring at least 4 wk before a substantial reduction in the peripheral T cell population was detected (20, 21, 22, 24). However, in all cases, the importance of selecting MHC molecule expression to quiescent T cell survival was not strictly examined. That is, survival was not measured under conditions representing a full lymphoid compartment and/or the population being studied included dividing cells. Moreover, unlike other reports investigating the role of specific peptide:MHC complexes in Ag-independent homeostatic proliferation of naive T cells, the nature of the signal provided by selecting MHC molecules for naive T cell survival was not addressed.

By carefully examining the fate of undivided naive CD4+ T cells in unirradiated Aβ0 hosts, we showed that T cell survival in the absence of TCR-MHC interactions was better than previously estimated. This difference may arise from the fact that several of the earlier studies assessed the presence of T cells from populations that included dividing cells. Hence, the apparent loss of these populations in MHC-deficient hosts may actually reflect their impaired ability to undergo homeostatic expansion, rather than the inability of individual T cells to survive. Nevertheless, our experiments do support previous findings that class II MHC expression is necessary to prolong the maintenance of polyclonal naive CD4+ T cells in a full lymphoid compartment. The fact that survival of naive wild-type CD4+ T cells was not impaired in unirradiated H-2 Mα0 hosts suggests that presentation of self peptides, similar to those involved in thymic positive selection, is not a necessary function of class II MHC molecules in this process. Potential interaction of CD4 with class II MHC molecules may account for the observed difference.

In contrast to CD4+ T cell survival, selective engagement of TCRs by self peptide:MHC complexes does appear to promote division of CD4+ T cells under severe lymphopenic conditions. Similar observations were reported recently by several groups for both polyclonal and monoclonal CD4+ and CD8+ T cells (14, 15, 17, 18, 19, 26). These results led to the view that homeostatic naive T cell proliferation is stringently dependent on TCR stimulation by positively selecting or low affinity peptide:MHC complexes. However, the low level of T cell division observed in the absence of specific self peptide:MHC complexes by day 7–9 in some of these studies was not commented on (15, 17, 19, 26). Here, we show that such division in mice lacking specific peptide:MHC complexes becomes quite substantial over time. Interestingly, T cell proliferation in the absence of class II MHC molecules was usually associated with the appearance of a CFSE-negative, memory-phenotype population in both irradiated and unirradiated Aβ0 hosts. This population appears to include a subset of naive T cells that divide rapidly and acquire a memory-like state. As well, these cells are probably derived from a small pool of preexisting memory T cells in the donor population. The latter alternative raises the possibility that the T cells that had undergone several divisions in the irradiated Aβ0 hosts were not naive T cells, but instead represented intermediates of the contaminating memory T cell population. However, the fact that intermediates of the CFSE-negative population could not be detected at earlier times together with the fact that these dividing donor T cells in the irradiated Aβ0 mice express a naive phenotype (Fig. 5 b), argues against this view.

In conclusion, the data suggest that different MHC-derived signals are required to facilitate naive T cell survival vs homeostatic T cell proliferation. Our results agree with the view that low affinity TCR interactions are necessary to induce efficient homeostatic proliferation. However, based on the data presented above, we also feel that a TCR-derived signal is not the only factor important for initiating naive T cell division in a lymphopenic environment. Other signals critical for regulating homeostatic T cell proliferation may include an inhibitory interaction between T cells that prevents their proliferation in a full lymphoid compartment. When such an inhibitory signal is missing in a severely lymphopenic environment, pathways allowing T cell division may be activated and can be assisted by weak signaling through the TCR by self peptide:MHC complexes. In the absence of such “TCR tickling,” cell division pathways may still be initiated, but proliferation is delayed. Additionally, there may be a third mitogenic signal, possibly delivered via the APC, which in the absence of TCR-mediated signaling, can also induce naive T cell division when the inhibitory interaction between T cells is removed. Therefore, a detailed molecular description of this important immunological phenomenon is necessary to identify the other signals regulating T cell homeostasis and will be of significance for the development of clinical treatments for lymphopenic states.

We thank D. Cho and M. Gomez for technical assistance and P. Wong and G. Barton for helpful comments.

1

S.R.M.C. is supported by the Human Frontier Science Program, and A.Y.R. is supported by the National Institutes of Health and the Howard Hughes Medical Institute.

3

Abbreviations used in this paper: Aβ0, class II-deficient; H-2 Mα0, H-2 M-knockout; CFSE, 5,6-carboxyfluorescein diacetate succinimidyl ester; B6, C57BL/6.

1
Fink, P, J., M. J. Bevan.
1995
. Positive selection of thymocytes.
Adv. Immunol.
59
:
99
2
Sebzda, E., S. Mariathasan, T. Ohteki, R. Jones, M. F. Bachmann, P. S. Ohashi.
1999
. Selection of the T cell repertoire.
Annu. Rev. Immunol.
17
:
829
3
Sprent, J., M. Schaefer, M. Hurd, C. D. Surh, Y. Ron.
1991
. Mature murine B and T cells transferred to SCID mice can survive indefinitely and many maintain a virgin phenotype.
J. Exp. Med.
174
:
717
4
von Boehmer, H., K. Hafen.
1993
. The life span of naive α/β T cells in secondary lymphoid organs.
J. Exp. Med.
177
:
891
5
Tough, D. F., J. Sprent.
1994
. Turnover of naive- and memory-phenotype T cells.
J. Exp. Med.
179
:
1127
6
Miller, R. A., O. Stutman.
1984
. T cell repopulation from functionally restricted splenic progenitors: 10,000-fold expansion documented by using limiting dilution analyses.
J. Immunol.
133
:
2925
7
Bell, E. B., S. M. Sparshott, M. T. Drayson, W. L. Ford.
1987
. The stable and permanent expansion of functional T lymphocytes in athymic nude rats after a single injection of mature T cells.
J. Immunol.
139
:
1379
8
Rocha, B., N. Dautigny, P. Pereira.
1989
. Peripheral T lymphocytes: expansion potential and homeostatic regulation of pool sizes and CD4/CD8 ratios in vivo.
Eur. J. Immunol.
19
:
905
9
Pereira, P., B. Rocha.
1991
. Post-thymic in vivo expansion of mature αβ T cells.
Int. Immunol.
3
:
1077
10
Rocha, B., H. von Boehmer.
1991
. Peripheral selection of the T cell repertoire.
Science
251
:
1225
11
Mackall, C. L., C. V. Bare, L. A. Granger, S. O. Sharrow, J. A. Titus, R. E. Gress.
1996
. Thymic-independent T cell regeneration occurs via antigen-driven expansion of peripheral T cells resulting in a repertoire that is limited in diversity and prone to skewing.
J. Immunol.
156
:
4609
12
Tanchot, C., F. A. Lemonnier, B. Perarnau, A. A. Freitas, B. Rocha.
1997
. Differential requirements for survival and proliferation of CD8 naive or memory T cells.
Science
276
:
2057
13
Bell, E. B., S. M. Sparshott.
1997
. The peripheral T-cell pool: regulation by non-antigen induced proliferation?.
Semin. Immunol.
9
:
347
14
Beutner, U., H. R. MacDonald.
1998
. TCR-MHC class II interaction is required for peripheral expansion of CD4 cells in a T cell-deficient host.
Int. Immunol.
10
:
305
15
Viret, C., F. S. Wong, C. A. Janeway, Jr.
1999
. Designing and maintaining the mature TCR repertoire: the continuum of self-peptide:self-MHC complex recognition.
Immunity
10
:
559
16
Oehen, S., K. Brduscha-Riem.
1999
. Naive cytotoxic T lymphocytes spontaneously acquire effector function in lymphocytopenic recipients: A pitfall for T cell memory studies?.
Eur. J. Immunol.
29
:
608
17
Ernst, B., D. S. Lee, J. M. Chang, J. Sprent, C. D. Surh.
1999
. The peptide ligands mediating positive selection in the thymus control T cell survival and homeostatic proliferation in the periphery.
Immunity
11
:
173
18
Goldrath, A. W., M. J. Bevan.
1999
. Low-affinity ligands for the TCR drive proliferation of mature CD8+ T cells in lymphopenic hosts.
Immunity
11
:
183
19
Kieper, W. C., S. C. Jameson.
1999
. Homeostatic expansion and phenotypic conversion of naive T cells in response to self peptide/MHC ligands.
Proc. Natl. Acad. Sci. USA
96
:
13306
20
Takeda, S., H. R. Rodewald, H. Arakawa, H. Bluethmann, T. Shimizu.
1996
. MHC class II molecules are not required for survival of newly generated CD4+ T cells, but affect their long-term life span.
Immunity
5
:
217
21
Rooke, R., C. Waltzinger, C. Benoist, D. Mathis.
1997
. Targeted complementation of MHC class II deficiency by intrathymic delivery of recombinant adenoviruses.
Immunity
7
:
123
22
Kirberg, J., A. Berns, H. von Boehmer.
1997
. Peripheral T cell survival requires continual ligation of the T cell receptor to major histocompatibility complex-encoded molecules.
J. Exp. Med.
186
:
1269
23
Nesic, D., S. Vukmanovic.
1998
. MHC class I is required for peripheral accumulation of CD8+ thymic emigrants.
J. Immunol.
160
:
3705
24
Boursalian, T. E., K. Bottomly.
1999
. Survival of naive CD4 T cells: roles of restricting versus selecting MHC class II and cytokine milieu.
J. Immunol.
162
:
3795
25
Brocker, T..
1997
. Survival of mature CD4 T lymphocytes is dependent on major histocompatibility complex class II-expressing dendritic cells.
J. Exp. Med.
186
:
1223
26
Bender, J., T. Mitchell, J. Kappler, P. Marrack.
1999
. CD4+ T cell division in irradiated mice requires peptides distinct from those responsible for thymic selection.
J. Exp. Med.
190
:
367
27
Grubin, C. E., S. Kovats, P. deRoos, A. Y. Rudensky.
1997
. Deficient positive selection of CD4 T cells in mice displaying altered repertoires of MHC class II-bound self-peptides.
Immunity
7
:
197
28
Lyons, A. B., C. R. Parish.
1994
. Determination of lymphocyte division by flow cytometry.
J. Immunol. Methods
171
:
131
29
Fung-Leung, W. P., C. D. Surh, M. Liljedahl, J. Pang, D. Leturcq, P. A. Peterson, S. R. Webb, L. Karlsson.
1996
. Antigen presentation and T cell development in H2-M-deficient mice.
Science
271
:
1278
30
Miyazaki, T., P. Wolf, S. Tourne, C. Waltzinger, A. Dierich, N. Barois, H. Ploegh, C. Benoist, D. Mathis.
1996
. Mice lacking H2-M complexes, enigmatic elements of the MHC class II peptide-loading pathway.
Cell
84
:
531
31
Martin, W. D., G. G. Hicks, S. K. Mendiratta, H. I. Leva, H. E. Ruley, L. Van Kaer.
1996
. H2-M mutant mice are defective in the peptide loading of class II molecules, antigen presentation, and T cell repertoire selection.
Cell
84
:
543
32
Tourne, S., T. Miyazaki, A. Oxenius, L. Klein, T. Fehr, B. Kyewski, C. Benoist, D. Mathis.
1997
. Selection of a broad repertoire of CD4+ T cells in H-2Ma0/0 mice.
Immunity
7
:
187
33
Surh, C. D., D. S. Lee, W. P. Fung-Leung, L. Karlsson, J. Sprent.
1997
. Thymic selection by a single MHC/peptide ligand produces a semidiverse repertoire of CD4+ T cells.
Immunity
7
:
209
34
Sant’Angelo, D. B., P. G. Waterbury, B. E. Cohen, W. D. Martin, L. Van Kaer, A. C. Hayday, C. A. Janeway, Jr.
1997
. The imprint of intrathymic self-peptides on the mature T cell receptor repertoire.
Immunity
7
:
517
35
Gudmundsdottir, H., A. D. Wells, L. A. Turka.
1999
. Dynamics and requirements of T cell clonal expansion in vivo at the single-cell level: effector function is linked to proliferative capacity.
J. Immunol.
162
:
5212
36
Sprent, J..
1997
. Immunological memory.
Curr. Opin. Immunol.
9
:
371
37
Goldrath, A. W., and M. J. Bevan. 2000. Naive T cells transiently acquire a memory-like phenotype during homeostatis-driven proliferation. J. Exp. Med. In press.
38
Murali-Krishna, K., L. L. Lau, S. Sambhara, F. Lemonnier, J. Altman, R. Ahmed.
1999
. Persistence of memory CD8 T cells in MHC class I-deficient mice.
Science
286
:
1377