Homeostatic proliferation of naive T cells transferred to T cell-deficient syngeneic mice is driven by low-affinity self-MHC/peptide ligands and the cytokine IL-7. In addition to homeostatic proliferation, a subset of naive T cells undergoes massive proliferation in chronically immunodeficient hosts, but not in irradiated normal hosts. Such rapid T cell proliferation occurs largely independent of homeostatic factors, because it was apparent in the absence of IL-7 and in T cell-sufficient hosts devoid of functional T cell immunity. Strikingly, immunodeficient mice raised under germfree conditions supported only slow homeostatic proliferation, but not the marked T cell proliferation observed in conventionally raised immunodeficient mice. Thus, polyclonal naive T cell expansion in T cell-deficient hosts can be driven predominantly by either self-Ags or foreign Ags depending on the host’s previous state of T cell immunocompetency.

Homeostatic mechanisms control the overall size and composition of the mature T cell pool in the peripheral lymphoid and nonlymphoid tissues (1, 2). Recent work has established that homeostasis of naive T cells is regulated by signals derived from contact with self components, namely, self-MHC/peptide ligands and the cytokine IL-7 (2). The main role of the homeostatic factors is to keep cells alive, because depriving cells of contact with these factors leads to gradual T cell death (2). In addition, homeostatic factors induce T cells to undergo intermittent homeostatic proliferation, which under normal T cell (T)5-sufficient conditions is evident for memory cells, but not for most naive T cells (2). Naive T cells are known to undergo homeostatic proliferation after transfer to syngeneic T-depleted hosts, presumably because of increased availability of MHC ligands and/or IL-7 (2, 3, 4).

Recent studies have shown that homeostatic proliferation of naive T cells in T-depleted hosts is driven by MHC molecules loaded with self rather than foreign peptides. In support of this idea, many lines of TCR transgenic cells in RAG-deficient backgrounds were found to undergo efficient homeostatic proliferation in lymphopenic hosts in the absence of the agonist peptides (2, 5). Moreover, studies with mice engineered to express MHC molecules loaded with a single species of self peptides have demonstrated that homeostatic proliferation of naive T cells is most efficiently driven by self peptides that have initially induced positive selection of the T cells in the thymus (2, 5). Despite these findings, T cell proliferation in lymphopenic hosts might be driven in part by foreign Ags. Thus, exposure to environmental Ags could enhance the rate of homeostatic proliferation through the activation of APC and/or by induction of inflammatory cytokines that augment the action of IL-7. Alternatively, foreign Ags could act as a direct TCR stimulus. Indeed, before work with TCR transgenic cells and “single-peptide” mice as hosts, T cell expansion in lymphopenic hosts was thought to be driven largely by environmental Ags rather than self-ligands (1, 6).

One situation where foreign Ags may participate in T cell expansion is in hosts that are constitutively immunodeficient, such as in nude, SCID, and RAG-deficient mice. Being immunodeficient, these mice are highly susceptible to chronic infection and thus may present a spectrum of microbial Ags to adoptively transferred T cells. In support of this idea, we show that, unlike TCR transgenic T cells, naive T cells from normal mice proliferate extensively in immunodeficient RAG-deficient and SCID mice, but only quite slowly when SCID mice are maintained in a germfree (GF) environment.

The origins of C57BL/6 (B6), B6.PL, B6.CD45.1, B6.IL-7, and TCR transgenic lines OT-I, OT-II, 2C, and P14, all in a B6 background, were described (4, 5). B6.RAG-1 mice and BALB/c.JH mice were obtained from The Jackson Laboratory. Bone marrow (BM) chimeras in RAG-1 mice were generated by injecting 5 × 106 B6 BM cells into irradiated (400 cGy) RAG-1 mice. BALB/c.Thy-1.1 mice were generated by crossing once with B6.PL mice and then backcrossing 10 times with BALB/c mice. Conventionally reared and GF C.B17 SCID mice, maintained sterilely in Trexler plastic isolators, were housed at the University of Pennsylvania (Philadelphia, PA). GF SCID mice of 6–8 wk of age were colonized with Schaedler’s Escherichia coli (O81:H21) (7) (provided by Dr. R. Schaedler, Philadelphia, PA), and their fecal smears were analyzed 2–3 wk later as described (8), just before use in experiments.

Whole lymph nodes (LN) and purified CD4 and CD8 T cells were CFSE labeled, i.v. injected, and detected in host tissues as described (5). For intracellular cytokine staining, donor T cells was stimulated with anti-CD3 mAb (eBioscience) in the presence of GolgiPlug (BD Pharmingen) for 5 h and then stained with Cy5-conjugated anti-Thy-1.1, PE-conjugated anti-CD4 (eBioscience), and FITC-conjugated anti-IFN-γ (eBioscience) or anti-TNF-α (eBioscience) using the Cytofix/Cytoperm kit (BD Pharmingen).

In the course of studying homeostatic proliferation in various types of T-depleted hosts, we have previously observed that polyclonal T cells proliferated at different rates depending on how the hosts were rendered T-deficient (5). In normal immunocompetent hosts that were manipulated to be acutely T-deficient, e.g., by irradiation or by treatment with anti-T cell mAbs, a significant proportion of Thy-1-congeneic donor polyclonal CD4 and CD8 T cells proliferated slowly, dividing every 24–48 h (Fig. 1,A and Ref. 5). The results were different in hosts that were chronically T-deficient as a result of a genetic failure to generate mature T cells, e.g., RAG-1 and TCRα mice. In these hosts, some cells underwent one to four rounds of cells division typical of homeostatic proliferation (Fig. 1,A and Ref. 5). However, other cells divided extensively (more than eight divisions) and led to a marked increase in cell recoveries. Thus, the total recoveries of donor T cells at 1 wk postinjection were usually 5- to 10-fold greater in RAG-1 hosts than in irradiated normal B6 hosts (see Fig. 1 and also Ref. 5).

FIGURE 1.

Naive polyclonal T cells undergo different types of proliferation depending on how the hosts were rendered lymphopenic. A, Proliferation of polyclonal and OT-I transgenic T cells in irradiated normal vs RAG-1 hosts. Aliquots (2 × 106/mouse) of CFSE-labeled LN cells from B6.PL and OT-I.Thy-1.1 mice were injected into irradiated (600 cGy) B6 and B6.RAG-1 mice, and the donor cells in host LN and spleen were analyzed 7 days later. Shown are CFSE profiles of donor T cells detected in host LN by double staining for CD8 and Thy-1.1; donor CD4 cells are identified as Thy-1.1+CD8 cells. The respective average recoveries of B6.PL CD4 and CD8 cells/mouse were 0.05 and 0.06 × 106 cells for B6, and 0.6 and 0.2 × 106 cells for RAG-1 hosts. The average recoveries of OT-I cells from B6 and RAG-1 hosts were 0.1 × 106 and 0.08 × 106 cells/mouse, respectively. B, Proliferation of naive phenotype polyclonal T cells in irradiated normal vs RAG-1 hosts. Aliquots of CFSE-labeled unsorted or CD44low sorted B6.PL LN T cells were injected into the indicated hosts, and the donor cells were analyzed 7 days later as described in A. The data are representative of two to three separate experiments comprised of two tothree mice per group.

FIGURE 1.

Naive polyclonal T cells undergo different types of proliferation depending on how the hosts were rendered lymphopenic. A, Proliferation of polyclonal and OT-I transgenic T cells in irradiated normal vs RAG-1 hosts. Aliquots (2 × 106/mouse) of CFSE-labeled LN cells from B6.PL and OT-I.Thy-1.1 mice were injected into irradiated (600 cGy) B6 and B6.RAG-1 mice, and the donor cells in host LN and spleen were analyzed 7 days later. Shown are CFSE profiles of donor T cells detected in host LN by double staining for CD8 and Thy-1.1; donor CD4 cells are identified as Thy-1.1+CD8 cells. The respective average recoveries of B6.PL CD4 and CD8 cells/mouse were 0.05 and 0.06 × 106 cells for B6, and 0.6 and 0.2 × 106 cells for RAG-1 hosts. The average recoveries of OT-I cells from B6 and RAG-1 hosts were 0.1 × 106 and 0.08 × 106 cells/mouse, respectively. B, Proliferation of naive phenotype polyclonal T cells in irradiated normal vs RAG-1 hosts. Aliquots of CFSE-labeled unsorted or CD44low sorted B6.PL LN T cells were injected into the indicated hosts, and the donor cells were analyzed 7 days later as described in A. The data are representative of two to three separate experiments comprised of two tothree mice per group.

Close modal

Notably, the marked donor T cell expansion in RAG-1 hosts did not apply to TCR transgenic T cells. Thus, with transfer of transgenic OT-I, P14, and 2C cells, the kinetics of proliferation and the total donor cell recoveries in RAG-1 vs irradiated B6 hosts were similar (Fig. 1,A and not shown). Because nearly all transgenic T cells display a naive phenotype, it seemed possible that the rapidly dividing population of polyclonal B6 cells in RAG-1 hosts was derived from pre-existing memory-phenotype T cells, which undergo a faster rate of homeostatic proliferation than naive T cells (9). However, this idea is unlikely, because the CFSE profiles and recoveries of donor T cells in RAG-1 hosts were similar whether the mice were injected with unseparated T cells or with sorted naive phenotype (CD44low) T cells (Fig. 1 B). Especially for CD8 cells, proliferation was somewhat slower for sorted CD44low cells than unseparated T cells, presumably reflecting the absence of CD44high cells (which proliferate more rapidly than CD44low cells; Ref. 9).

To seek further information on the fast-dividing cells, a group of RAG-1 hosts was injected with small numbers of CFSE-labeled B6.PL LN cells and the hosts were analyzed on a daily basis. As shown in Fig. 2,A, most of the donor T cells underwent less than two rounds of cell division during the first 3 days. Starting around day 4, however, a population of CFSE donor cells became detectable, indicative of extensive division. This cell population expanded massively over the following days, and overshadowed the slowly dividing cells. Analysis of cell recoveries on days 4–7 indicated that the CFSE cells divided every 6–8 h, i.e., at a rate typical of T cells dividing in response to foreign Ags. Further supporting the idea that foreign Ags induced this expansion, the fast-dividing T cells in RAG-1 hosts exhibited features of effector cells. Thus, nearly all donor T cells obtained from RAG-1 hosts on day 8 were CD44high, with about half of the cells expressing a high level of CD25 (data for CD4 cells shown in Fig. 2,B). In addition, a significant proportion (20–30%) of these cells were found to synthesize either or both IFN-γ and TNF-α upon 5-h in vitro stimulation with anti-CD3 mAb (Fig. 2,B). By contrast, T cells undergoing slow homeostatic proliferation in irradiated RAG+ littermate hosts showed only slight up-regulation of CD44 and CD25 and minimally acquired the capacity to synthesize IFN-γ and TNF-α (Fig. 2 B).

FIGURE 2.

Characterization of polyclonal T cell proliferation in RAG-1 hosts. A, Kinetics of T cell proliferation in RAG-1 hosts. A group of RAG-1 hosts was injected with aliquots of CFSE-labeled B6.PL LN cells, and some of the mice were analyzed later on the indicated days as described in Fig. 1. Shown are CFSE profiles of donor T cells detected in host spleen. B, T cells expanded in RAG-1 hosts acquire phenotype and characteristics of effector cells. Aliquots of B6.PL LN cells were injected into RAG-1+/− and RAG-1 mice and analyzed 8 days later for expression of activation markers and the ability to synthesize cytokines. Shown are expression of CD44 and CD25 and intracellular IFN-γ and TNF-α staining on donor T cells in host spleens detected by triple staining as described in Materials and Methods.

FIGURE 2.

Characterization of polyclonal T cell proliferation in RAG-1 hosts. A, Kinetics of T cell proliferation in RAG-1 hosts. A group of RAG-1 hosts was injected with aliquots of CFSE-labeled B6.PL LN cells, and some of the mice were analyzed later on the indicated days as described in Fig. 1. Shown are CFSE profiles of donor T cells detected in host spleen. B, T cells expanded in RAG-1 hosts acquire phenotype and characteristics of effector cells. Aliquots of B6.PL LN cells were injected into RAG-1+/− and RAG-1 mice and analyzed 8 days later for expression of activation markers and the ability to synthesize cytokines. Shown are expression of CD44 and CD25 and intracellular IFN-γ and TNF-α staining on donor T cells in host spleens detected by triple staining as described in Materials and Methods.

Close modal

One possible source of non-self-Ags in RAG-1 (and TCRα) hosts is the neomycin-resistant gene inserted into the genome to disrupt the host genes. However, these Ags are unlikely to be the main stimulus for the rapidly dividing T cells, because similar rapid proliferation occurs in nude and SCID hosts, which possess mutations that have arisen spontaneously (see Figs. 3 and 5). Moreover, T cells from RAG-1+/− mice, which are tolerant to the products of the neomycin-resistant gene, behaved identically to B6 T cells upon transfer into B6 and RAG-1 hosts. Thus, most RAG-1+/− donor T cells proliferated slowly in irradiated B6 hosts, whereas both fast- and slow-dividing populations of cells were evident in RAG-1−/− hosts (not shown).

FIGURE 3.

Strong proliferation of polyclonal T cells in RAG-1 hosts is dependent on chronic immunodeficiency from the absence of T, but not B cells. A, Reconstitution with normal BM cells abolishes the capacity of RAG-1 mice to strongly activate donor T cells. Aliquots of CFSE-labeled B6.PL LN cells were injected into B6 BM → RAG-1 chimeras with or without prior exposure to irradiation and analyzed 7 days later; RAG-1 and irradiated B6 mice were also used as controls. Shown are CFSE profiles of donor T cells in host LN. B, Rapid T cell proliferation occurs in T-sufficient immunodeficient mice. Aliquots of CFSE-labeled B6.PL LN cells were injected into irradiated OT-I.RAG-1 and OT-II.RAG-1 mice and analyzed 7 days later; irradiated RAG-1 and B6 mice were also used as controls. Shown are CFSE profiles of donor T cells in host LN. C, Rapid T cell proliferation does not occur in B cell-deficient mice. Aliquots of CFSE-labeled BALB/c.Thy-1.1 LN cells were injected into irradiated BALB.JH mice and analyzed 8 days later; SCID and irradiated BALB/c mice were also used as controls. Shown are CFSE profiles of donor T cells in host LN. Data are representative of two to three experiments.

FIGURE 3.

Strong proliferation of polyclonal T cells in RAG-1 hosts is dependent on chronic immunodeficiency from the absence of T, but not B cells. A, Reconstitution with normal BM cells abolishes the capacity of RAG-1 mice to strongly activate donor T cells. Aliquots of CFSE-labeled B6.PL LN cells were injected into B6 BM → RAG-1 chimeras with or without prior exposure to irradiation and analyzed 7 days later; RAG-1 and irradiated B6 mice were also used as controls. Shown are CFSE profiles of donor T cells in host LN. B, Rapid T cell proliferation occurs in T-sufficient immunodeficient mice. Aliquots of CFSE-labeled B6.PL LN cells were injected into irradiated OT-I.RAG-1 and OT-II.RAG-1 mice and analyzed 7 days later; irradiated RAG-1 and B6 mice were also used as controls. Shown are CFSE profiles of donor T cells in host LN. C, Rapid T cell proliferation does not occur in B cell-deficient mice. Aliquots of CFSE-labeled BALB/c.Thy-1.1 LN cells were injected into irradiated BALB.JH mice and analyzed 8 days later; SCID and irradiated BALB/c mice were also used as controls. Shown are CFSE profiles of donor T cells in host LN. Data are representative of two to three experiments.

Close modal
FIGURE 5.

Proliferation of polyclonal T cells in GF and gnotobiotic SCID mice. A, Proliferation of T cells in GF SCID mice. Aliquots (5 × 106/mouse) of CFSE-labeled LN cells from BALB/c.Thy-1.1 mice were injected into groups of GF SCID mice and analyzed 7 days later; conventional SCID and irradiated BALB/c mice were used as controls. Shown are CFSE profiles of donor T cells in host LN. The respective average recoveries of donor CD4 and CD8 cells/mouse were 0.14 and 0.04 × 106 for BALB/c, 2.7 and 1.7 × 106 for conventional SCID, and 0.17 and 0.03 × 106 for GF SCID. B, Proliferation of T cells in gnotobiotic SCID mice. Aliquots of CFSE-labeled BALB/c.Thy-1.1 LN cells were injected into groups of GF SCID mice reconstituted with Schaedler’s E. coli (SEC) and analyzed 7 days later; conventional and GF SCID mice were used as controls. The respective average recoveries of donor CD4 and CD8 cells/mouse were 0.6 and 0.4 × 106 for conventional SCID, 0.1 and 0.03 × 106 for GF SCID, and 0.05 and 0.02 × 106 for SCID with SEC hosts. Shown are CFSE profiles of donor T cells in host LN. Each group consisted of four to five mice analyzed individually. Two other experiments showed similar results.

FIGURE 5.

Proliferation of polyclonal T cells in GF and gnotobiotic SCID mice. A, Proliferation of T cells in GF SCID mice. Aliquots (5 × 106/mouse) of CFSE-labeled LN cells from BALB/c.Thy-1.1 mice were injected into groups of GF SCID mice and analyzed 7 days later; conventional SCID and irradiated BALB/c mice were used as controls. Shown are CFSE profiles of donor T cells in host LN. The respective average recoveries of donor CD4 and CD8 cells/mouse were 0.14 and 0.04 × 106 for BALB/c, 2.7 and 1.7 × 106 for conventional SCID, and 0.17 and 0.03 × 106 for GF SCID. B, Proliferation of T cells in gnotobiotic SCID mice. Aliquots of CFSE-labeled BALB/c.Thy-1.1 LN cells were injected into groups of GF SCID mice reconstituted with Schaedler’s E. coli (SEC) and analyzed 7 days later; conventional and GF SCID mice were used as controls. The respective average recoveries of donor CD4 and CD8 cells/mouse were 0.6 and 0.4 × 106 for conventional SCID, 0.1 and 0.03 × 106 for GF SCID, and 0.05 and 0.02 × 106 for SCID with SEC hosts. Shown are CFSE profiles of donor T cells in host LN. Each group consisted of four to five mice analyzed individually. Two other experiments showed similar results.

Close modal

Because RAG-1, TCRα, SCID, and nude mice are severely immunodeficient, the question arises whether the strong donor T cell proliferation in these hosts is driven by Ags derived from microbes in the environment that have invaded the hosts. According to this idea, restoration of normal immunity in these mice should control these microbes and thus abrogate the ability of these mice to induce strong proliferation of donor T cells. To test this idea, RAG-1 mice were reconstituted with normal B6 BM cells and rested for 2 mo before injection of CFSE-labeled B6.PL T cells. As shown in Fig. 3,A, donor T cells injected into BM-reconstituted RAG-1 chimeras remained in interphase, presumably due to the T-sufficient state of the hosts. Even when the chimeras were first exposed to irradiation, the donor B6.PL T cells underwent only a few rounds of slow cell division, without any sign of the prominent rapidly dividing population apparent in un-reconstituted RAG-1 hosts (Fig. 3 A). Indeed, the pace of donor T cell proliferation in irradiated chimeras was slower than that observed in irradiated B6 hosts, possibly because the additional dose of irradiation diminished the ability of the chimeras to produce homeostatic factors, such as IL-7.

The above findings raise the question whether the failure of the BM-reconstituted RAG-1 hosts to support rapid proliferation of the donor polyclonal T cells is due to T-repletion per se, rather than T-dependent elimination of infection. To address this question, we transferred CFSE-labeled B6.PL LN cells into mice that have near-normal numbers of T cells, but are nevertheless immunodeficient, i.e., TCR transgenic mice bred to a RAG background. The following two OVA-specific TCR transgenic lines were used: the CD8+ OT-I transgenic line, which has near-normal numbers of T cells, and CD4+ OT-II transgenic mice, which possess 5- to 10-fold lower numbers of T cells than normal mice (data not shown). Notably, when OT-I.RAG-1 and OT-II.RAG-1 hosts were irradiated and used as hosts, polyclonal T cell proliferation was as rapid in these hosts as in nontransgenic RAG-1 hosts (Fig. 3 B). Indeed, equivalent rapid proliferation of donor T cells occurred when OT-I.RAG-1 and OT-II.RAG-1 hosts were not irradiated (not shown). These findings indicate that strong T cell proliferation in immunodeficient hosts is dependent on the hosts being immunodeficient rather than T cell deficient.

Because rapid polyclonal T cell proliferation is observed in T-sufficient OT-I.RAG-1 mice, it suggests that such proliferation is not dependent on the hosts possessing elevated levels of IL-7. To determine whether IL-7 is required for inducing rapid polyclonal T cell proliferation in immunodeficient hosts, CFSE-labeled B6.PL LN cells were injected into RAG-1 mice bred to be further deficient in IL-7. As shown in Fig. 4, rapid proliferation of donor CD4 and CD8 cells was observed in IL-7RAG-1 hosts, although the relative size of the expanded T cell populations was about half of that observed in IL-7+RAG-1 hosts (see Fig. 4). Although some T cells may require IL-7, it appears IL-7 is not essential for most T cells to undergo rapid proliferation in immunodeficient hosts, but promotes accumulation of the activated cells.

FIGURE 4.

IL-7 is not critical for rapid T cell proliferation in immunodeficient hosts. Aliquots (4 × 106/mouse) of CFSE-labeled B6.PL LN cells were injected into RAG-1 and IL-7RAG-1 mice and analyzed 8 days later; IL-7 and irradiated B6 mice were also used as controls. Shown are CFSE profiles of donor T cells in host spleen. The respective average recoveries of B6.PL CD4 and CD8 cells/mouse were 1.2 and 0.5 × 106 cells for RAG-1 hosts, and 0.7 and 0.3 × 106 cells for IL-7RAG-1 hosts. Data are representative of two experiments.

FIGURE 4.

IL-7 is not critical for rapid T cell proliferation in immunodeficient hosts. Aliquots (4 × 106/mouse) of CFSE-labeled B6.PL LN cells were injected into RAG-1 and IL-7RAG-1 mice and analyzed 8 days later; IL-7 and irradiated B6 mice were also used as controls. Shown are CFSE profiles of donor T cells in host spleen. The respective average recoveries of B6.PL CD4 and CD8 cells/mouse were 1.2 and 0.5 × 106 cells for RAG-1 hosts, and 0.7 and 0.3 × 106 cells for IL-7RAG-1 hosts. Data are representative of two experiments.

Close modal

As mentioned earlier, a rapid proliferation of donor T cells is also apparent in nude and TCRα mice (5), i.e., mice that lack T cells, but contain B cells. To assess the role of B cells, CFSE-labeled BALB/c.Thy-1.1 LN cells were injected into irradiated BALB/c.JH mice. Because of a deletion in the Ig H chain J region (JH), these mice are deficient in B cells, but possess normal numbers of polyclonal T cells (10). As shown in Fig. 3 C, the kinetics of donor T cell proliferation in irradiated BALB/c JH mice were similar to that seen in irradiated normal BALB/c hosts and, especially for CD8 cells, much slower than in control BALB/c (CB-17) SCID mice. These findings suggest that fast proliferation of polyclonal T cells in immunodeficient hosts is dependent on T cell deficiency and largely independent of B cell deficiency.

To seek direct evidence on whether fast proliferation is driven by foreign Ags, we used SCID mice raised under GF conditions as hosts. Groups of GF SCID mice were injected with CFSE-labeled BALB/c.Thy-1.1 T cells, and the proliferation of the donor cells was measured 7 days later. As controls, groups of irradiated BALB/c mice and SCID mice raised under conventional environments were also injected with the same aliquot of polyclonal T cells. The key finding was that most of the polyclonal T cells injected into GF SCID mice underwent only a few rounds of cell division and failed to give rise to the prominent population of rapidly proliferating cells apparent in conventionally reared SCID mice (Fig. 5,A). Indeed, kinetics of proliferation of donor T cells in GF SCID mice was slow as in irradiated BALB/c mice and markedly different to the rapid proliferation seen in conventionally reared SCID mice (Fig. 5 A). The GF hosts were confirmed to be in a sterile condition by testing for the absence of intestinal bacteria at the end of the experiment.

Because conventionally reared immunodeficient mice are raised under specific pathogen-free conditions and remain healthy for several months, these mice are unlikely to harbor pathogenic microbes. Thus, the major source of foreign Ags in immunodeficient mice is likely to be nonpathogenic microflora present in the digestive system. To determine whether Ags from any enteric bacteria can act as the stimulus for rapid T cell proliferation, GF SCID mice were reconstituted with Schaedler’s E. coli (7), a Gram-negative bacteria, and tested for their ability to support proliferation of BALB/c.Thy-1.1 LN cells. As shown in Fig. 5,B, unlike the control SCID mice that were reared under conventional conditions, the gnotobiotic SCID mice reconstituted with only one species of E. coli failed to support a significant population of fast-dividing cells: indeed, the donor cell CFSE profiles in the gnotobiotic hosts mirrored that observed in unmanipulated GF SCID hosts (Fig. 5 B).

If Ags from enteric bacteria do indeed drive rapid T cell proliferation in immunodeficient hosts, the above finding with gnotobiotic mice indicates that the presence of specific strains of bacteria is required. In pondering this issue, valuable clues can be obtained when one considers that adoptive transfer of naive CD4 cells into syngeneic SCID or RAG-deficient mice is widely used to induce inflammatory bowel disease (IBD) (11). Here, severe colitis is thought to reflect unregulated activation of naive donor CD4 cells into IFN-γ- and TNF-α-secreting effector cells that infiltrate the colon (12). Indeed, it is most likely that the fast-proliferating cells observed in immunodeficient hosts are the initial responders that differentiate into the pathologic cells. The key relevant finding from the IBD studies is that IBD cannot be induced in SCID mice raised under GF conditions and that only three species of commensal bacteria have been shown to be associated with the induction of IBD in different animal models (8, 13, 14), with Helicobacter muridarum, and possibly Helicobacter hepaticus, identified as the provocateur of IBD in SCID mice injected with naive T cells (8, 15). Accordingly, it is likely that rapid donor T cell proliferation in immunodeficient hosts will be observed only upon reconstitution of GF SCID mice with these strains of bacteria.

Collectively, the above findings indicate that the Ags driving the bulk of the T cell expansion in syngeneic lymphopenic hosts differ greatly depending on the host’s microflora content and recent state of T cell immunocompetency. In conventional immunocompetent hosts manipulated to be lymphopenic, low-affinity self-ligands induce the bulk of T cell expansion, and the pace of proliferation is thus slow. We refer to such proliferation as true homeostatic proliferation because it is evident even in GF SCID hosts. On the contrary, in severely immunodeficient hosts, non-self-ligands, presumably Ags from microflora, become the dominant force and induce rapid proliferation of a fraction of T cells. The difference in the nature of the Ags in these two situations also appears to explain why slow homeostatic proliferation in the former case occurs largely independent of signals through the costimulatory molecules, such as CD28, whereas the rapid proliferation in the latter situations is abolished by blocking the costimulatory signals (16, 17, 18).

The interesting question why Ags from the enteric bacteria become overtly immunogenic in T-deficient mutant mice is currently unknown. Although there are many possibilities, we favor the idea that mutant mice lack a specific population(s) of T cells that is somehow involved in regulating presentation of normal enteric bacterial Ags by APC, a process which is known to occur at low levels under normal conditions (19). Accordingly, the residual T cells present in acutely T-depleted normal mice may continue to regulate presentation of enteric Ags. Currently, the identity of the T cells and how these cells regulate presentation of enteric Ags are unclear.

Finally, because of proliferation induced by host microflora Ags, results from experiments on homeostatic proliferation using polyclonal cells and immunodeficient hosts should be interpreted with caution. In this regard, the original studies that documented homeostatic expansion was performed by transferring polyclonal T cells into syngeneic immunodeficient nude hosts (20). The results from these studies were interpreted to indicate that mature T cells have a vast potential to undergo homeostatic expansion, in the range of 10,000- to 100,000-fold expansion over a few months. However, because much of the proliferation in these hosts was likely to be driven by foreign Ags, the potential of naive T cells to undergo genuine slow homeostatic proliferation is likely to be much smaller. Moreover, strong proliferation of polyclonal T cells in TCR transgenic mice was recently interpreted to reflect self-ligand-driven homeostatic proliferation from lack of clonal competition from naive or memory T cells (21, 22). Although homeostatic proliferation can occur from the absence of clonal competition to self-ligands (3, 4), our current findings argue against this being the cause of polyclonal T cell proliferation in RAG-deficient hosts.

The authors have no financial conflict of interest.

We thank Dr. N. Klinman for facilitating the collaborations, Dr. J. Sprent for critically reading the manuscript, and D. Kim, K. Ayala, E. LeRoy, B. Bondi-Boyd, and J. Kuhns for various support.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported by U.S. Public Health Service Grants AI41079, AI45809, and AG20186 (to C.D.S.) and AI37108 (to J.J.C.), and the Crohn’s and Colitis Foundation of America (to J.J.C. and H.-Q.J.). W.C.K. was supported by U.S. Public Health Service Institute National Research Service Award AI07244.

5

Abbreviations used in this paper: T, T cell; GF, germfree; BM, bone marrow; LN, lymph node; IBD, inflammatory bowel disease.

1
Tanchot, C., M. M. Rosado, F. Agenes, A. A. Freitas, B. Rocha.
1997
. Lymphocyte homeostasis.
Semin. Immunol.
9
:
331
.
2
Jameson, S. C..
2002
. Maintaining the norm: T-cell homeostasis.
Nat. Rev. Immunol.
2
:
547
.
3
Moses, C. T., K. M. Thorstenson, S. C. Jameson, A. Khoruts.
2003
. Competition for self ligands restrains homeostatic proliferation of naive CD4 T cells.
Proc. Natl. Acad. Sci. USA
100
:
1185
.
4
Kieper, W. C., J. T. Burghardt, C. D. Surh.
2004
. A role for TCR affinity in regulating naive T cell homeostasis.
J. Immunol.
172
:
40
.
5
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
.
6
Rocha, B., H. von Boehmer.
1991
. Peripheral selection of the T cell repertoire.
Science
251
:
1225
.
7
Schaedler, R. W., R. Dubs, R. Costello.
1965
. Association of germfree mice with bacteria isolated from normal mice.
J. Exp. Med.
122
:
77
.
8
Jiang, H. Q., N. Kushnir, M. C. Thurnheer, N. A. Bos, J. J. Cebra.
2002
. Monoassociation of SCID mice with Helicobacter muridarum, but not four other enterics, provokes IBD upon receipt of T cells.
Gastroenterology
122
:
1346
.
9
Tan, J. T., B. Ernst, W. C. Kieper, E. LeRoy, J. Sprent, C. D. Surh.
2002
. Interleukin (IL)-15 and IL-7 jointly regulate homeostatic proliferation of memory phenotype CD8+ cells but are not required for memory phenotype CD4+ cells.
J. Exp. Med.
195
:
1523
.
10
Jakobovits, A., G. J. Vergara, J. L. Kennedy, J. F. Hales, R. P. McGuinness, D. E. Casentini-Borocz, D. G. Brenner, G. R. Otten.
1993
. Analysis of homozygous mutant chimeric mice: deletion of the immunoglobulin heavy-chain joining region blocks B-cell development and antibody production.
Proc. Natl. Acad. Sci. USA
90
:
2551
.
11
Powrie, F., M. W. Leach, S. Mauze, L. B. Caddle, R. L. Coffman.
1993
. Phenotypically distinct subsets of CD4+ T cells induce or protect from chronic intestinal inflammation in C.B-17 scid mice.
Int. Immunol.
5
:
1461
.
12
Bouma, G., W. Strober.
2003
. The immunological and genetic basis of inflammatory bowel disease.
Nat. Rev. Immunol.
3
:
521
.
13
Rath, H. C., M. Schultz, R. Freitag, L. A. Dieleman, F. Li, H. J. Linde, J. Scholmerich, R. B. Sartor.
2001
. Different subsets of enteric bacteria induce and perpetuate experimental colitis in rats and mice.
Infect. Immun.
69
:
2277
.
14
Balish, E., T. Warner.
2002
. Enterococcus faecalis induces inflammatory bowel disease in interleukin-10 knockout mice.
Am. J. Pathol.
160
:
2253
.
15
Cahill, R. J., C. J. Foltz, J. G. Fox, C. A. Dangler, F. Powrie, D. B. Schauer.
1997
. Inflammatory bowel disease: an immunity-mediated condition triggered by bacterial infection with Helicobacter hepaticus.
Infect. Immun.
65
:
3126
.
16
Prlic, M., B. R. Blazar, A. Khoruts, T. Zell, S. C. Jameson.
2001
. Homeostatic expansion occurs independently of costimulatory signals.
J. Immunol.
167
:
5664
.
17
Gudmundsdottir, H., L. A. Turka.
2001
. A closer look at homeostatic proliferation of CD4+ T cells: costimulatory requirements and role in memory formation.
J. Immunol.
167
:
3699
.
18
Hagen, K. A., C. T. Moses, E. F. Drasler, K. M. Podetz-Pedersen, S. C. Jameson, A. Khoruts.
2004
. A role for CD28 in lymphopenia-induced proliferation of CD4 T cells.
J. Immunol.
173
:
3909
.
19
Mowat, A. M..
2003
. Anatomical basis of tolerance and immunity to intestinal antigens.
Nat. Rev. Immunol.
3
:
331
.
20
Rocha, B., N. Dautigny, P. Pereira.
1989
. Peripheral T lymphocytes: expansion potential and homeostatic regulation of pool sizes and CD4/8 ratios in vivo.
Eur. J. Immunol.
19
:
905
.
21
Troy, A. E., H. Shen.
2003
. Cutting edge: homeostatic proliferation of peripheral T lymphocytes is regulated by clonal competition.
J. Immunol.
170
:
672
.
22
Min, B., G. Foucras, M. Meier-Schellersheim, W. E. Paul.
2004
. Spontaneous proliferation, a response of naive CD4 T cells determined by the diversity of the memory cell repertoire.
Proc. Natl. Acad. Sci. USA
101
:
3874
.