We identified committed T cell progenitors (CTPs) in the mouse bone marrow that have not rearranged the TCRβ gene; express a variety of genes associated with commitment to the T cell lineage, including GATA-3, T cell-specific factor-1, Cβ, and Id2; and show a surface marker pattern (CD44+CD25CD24+CD5) that is similar to the earliest T cell progenitors in the thymus. More mature committed intermediate progenitors in the marrow have rearranged the TCR gene loci, express Vα and Vβ genes as well as CD3ε, but do not express surface TCR or CD3 receptors. CTPs, but not progenitors from the thymus, reconstituted the αβ T cells in the lymphoid tissues of athymic nu/nu mice. These reconstituted T cells vigorously secreted IFN-γ after stimulation in vitro, and protected the mice against lethal infection with murine CMV. In conclusion, CTPs in wild-type bone marrow can generate functional T cells via an extrathymic pathway in athymic nu/nu mice.

We have previously identified a rare population (∼1:2000 cells) of bone marrow resident progenitors expressing the Thy-1.2highTCRαβNK1.1B220Mac-1CD2CD44+CD16+ phenotype that are capable of generating T cells in vitro (1) and in vivo (2). Because these progenitors only developed into T cells and no other lineages, they were called committed T cell progenitors (CTPs).4 CTPs have not rearranged the TCR Vβ genes, and express mRNA for RAG-1, RAG-2, and preTα (1, 2). In vitro, the CTPs developed into mature CD4+ and CD8+ TCRαβ+ single-positive T cells after 48 h of culture via a Thy-1.2highLinCD2+ intermediate progenitor (CIP) stage and a CD4+CD8+ TCRαβ+ double-positive T cell stage (1). Rare cells with the CIP phenotype are found in the fresh bone marrow of adult mice, and can rapidly generate mature TCRαβ+ T cells also in vitro (1).

Phenotypic CTPs can be found in the bone marrow of athymic nu/nu mice, but these CTPs are not functional due to abnormalities in the nu/nu bone marrow microenvironment that result in defective differentiation from hemopoietic stem cells (3). In contrast, hemopoietic stem cells from wild-type hosts differentiated into functional CTPs in the bone marrow of irradiated thymectomized wild-type hosts within a few weeks after adoptive transfer (3). These extrathymically derived CTPs were able to develop into mature TCRαβ+ T cells after transfer into additional irradiated wild-type hosts. The latter experiments demonstrated that CTPs are prethymic (thymus-independent) progenitors (3). Studies of the normal human bone marrow have identified early progenitors with characteristics shared by CTPs and/or CIPs, and are committed to the T cell lineage as judged by the expression of T cell lineage-specific genes, active rearrangement of TCR gene segments, and expression of pre-TCRα receptors on the cell surface associated with intracellular CD3 (4). Although the mouse CTPs were able to generate mature T cells in vitro in the absence of thymic stromal cells, we did not determine whether the CTPs can generate T cells in vivo in athymic mice nor whether such extrathymic T cells would be functional with a full Vβ repertoire.

The current study compares T cell lineage surface receptors, gene expression, and ability of the CTPs and the CIPs in the bone marrow, and T cell progenitors in the thymus to generate mature T cells in athymic nu/nu mice. Key surface receptor differences between the CTPs and CIPs include the down-regulation of CD16 with the concomitant up-regulation of CD5 and CD122 by the CIPs. The more mature CIPs also down-regulate expression of pre-Tα, RAG-1, and RAG-2 genes, and up-regulate expression of CD3ε and TCR Vβ8 genes in association with the TCRβ gene rearrangement. Although these patterns are similar to progenitor maturation in the thymus, key differences were observed as well. In addition, we showed that limiting dilutions of wild-type mouse CTPs transferred to irradiated nu/nu hosts can generate CD4+ and CD8+TCRαβ+ T cells with a full Vβ receptor repertoire in the absence of the thymus. Progenitors in the thymus failed to generate mature T cells in the nu/nu hosts. Furthermore, we determined that extrathymically derived T cells from CTPs are functional and can protect against viral infection.

Congenic wild-type C57BL/6 CD45.1 mice (Department of Comparative Medicine, Stanford University), nu/nu C57BL/6 CD45.2 mice (Taconic Farms), and RAG-1−/−C57BL/6 CD45.2 mice (Taconic Farms) were used at 8–12 wk old. All experiments were performed in compliance with National Institutes of Health and institutional guidelines, and were approved by the Stanford Administrative Panel on Laboratory Animal Care.

Fluorochrome-conjugated anti-CD2 (RM2-5), CD3ε (145-2C11), CD4 (RM4-5), CD11b (Mac-1, M1/70.15), CD16/CD32 (2.4G2), CD25 (IL-2Rα, 7D4), CD45R (B220, RA3-6B2), CD45.1 (Ly-5.2, A20.1.7), CD45.2 (Ly-5.1, AL1-4A2), CD122 (IL-2Rβ, TM-β1), CD127 (IL-7Rα, B12-1), GR-1 (RB6-8C3), NK1.1 (PK136), TCRαβ (H57-597), and TCRγδ (GL3) mAbs were obtained from BD Pharmingen. Fluorochrome-conjugated CD4 (CT-CD4), CD8α (CT-CD8), CD25 (PC61-5.3), CD90.2 (Thy-1.2, 5a-8), rat IgG2b, and hamster IgG1 isotype controls were obtained from Caltag Laboratories. The following Abs were conjugated as previously described and directed to (3): CD3ε (KT31.1), CD4 (GK1.5), CD5 (53-7.3), CD8α (53-6.7), CD11b (Mac-1, M1/70), CD24 (heat stable Ag, M1/69), CD44 (IM781), CD45R (B220, RA3-6B2), CD45.1 (Ly-5.2, A20.1.7), CD45.2 (Ly-5.1, AL1-4A2), CD90.2 (Thy-1.2, 5a-8), CD117 (c-Kit, 2B8), GR-1 (RB6-8C3), NK1.1 (136TC), and Sca-1 (E13-161).

Single cell suspensions of bone marrow, spleen, and thymus cells were obtained, as previously described (1, 2). All single cell suspension samples were pretreated with purified anti-CD16/32 at saturation to block FcγRIIIA/B receptors. Following CD16/32 block, the samples were incubated with various combinations of fluorochrome- or biotin-conjugated Abs at saturation for 15–20 min on ice. In the case of biotin-conjugated reagents, counterstaining for 5 min with fluorochrome-conjugated streptavidin was performed. After staining, cells were washed and resuspended in fresh staining medium containing PBS without calcium or magnesium (BioWhittaker) with 1% FBS (HyClone) before sorting or analysis. Stained samples were resuspended in staining medium with propidium iodide at 0.5 μg/ml (Sigma-Aldrich).

Eight-color FACS analysis and sorting were done using a highly modified dual laser (488-nm argon and 599-nm dye lasers) FACS III (BD Biosciences) with four-decade logarithmic amplifiers (5) and a Moflo console (DakoCytomation). Four-color sorting was done using the FACSVantage (BD Biosciences), while one- to three-color sorting was done using the FACStar (BD Biosciences). Data were analyzed using Flowjo software (Tree Star) as either histograms or two-parameter 5% probability plots.

The isolation of CTPs has been previously described (1). In brief, bone marrow of adult mice was enriched by incubation with biotin-conjugated anti-Thy-1.2 mAb (5a-8; Caltag Laboratories), incubated further with streptavidin-conjugated immunomagnetic beads, and positively selected on MACS-LS magnetic separation columns (Miltenyi Biotec), according to the manufacturer’s specifications. For molecular genetics and extended surface phenotype studies, the positively selected marrow was stained for both T cell lineage (CD3) and non-T cell lineage (B220, Mac-1, NK1.1) markers. The Thy-1.2highLinCD2 (CTP) and Thy-1.2highLinCD2+ (CIP) populations were sorted thereafter, wherein Lin cells were depleted of both T cell and non-T cell lineage markers. For in vitro studies, Thy-1.2-enriched marrow was stained for both T cell lineage (TCRαβ) and non-T cell lineage markers and subsequently sorted for Thy-1.2highLinCD2 cells.

Total RNA was extracted from sorted cells using the RNeasy Mini Kit (Qiagen). For RNA isolation, between 0.3 and 5 × 105 cells were used. RNA was then reversed transcribed using random hexamer primers, followed by PCR amplification. Optimal conditions for PCR amplification for β-actin message, to be used as an internal standard, were established by titration of the number of amplification cycles using primers specific for β-actin, followed by densitometry analysis to measure ethidium bromide luminescence of PCR products. PCR products were identified on agarose gels with ethidium bromide staining. RAG-1, RAG-2, and IL-7Rα primers were used for RT-PCR, as previously described (6). Additional primers used for RT-PCR were designed using Oligo4.03 (National Biosciences) with the sequences found in GenBank (7) under the listed accession number: CD3ε (M23376), 5′-CACCCTGCTACTCCTTGTTC-3′ and 5′-TTACAGAATCCATCCCAGAG-3′; GATA-3 (X55123), 5′-GGGGCCTCTGTCCGTTTACC-3′ and 5′-CCTTCTGTGCTGGATCGTGC-3′; T cell-specific factor-1 (TCF-1) (X61385), 5′-CTACTCTGCCTTCAATCTGC-3′ and 5′-TTGATGACTGGCTTCTTAGC-3′; Vβ8 (AE000664), 5′-TACAAGGCCTCCAGACCAAG-3′; pre-Tα.1 (U16958), 5′-GGCTCCACCCATCACACTGC-3′ and 5′-CCATTTACAAGAGGCAGATCAC3′; pre-Tα.2, 5′-TGCTGGTGGTTTGCCTGGTC-3′ and 5′-GGGAGCAGTAGTGTCCAGCATC-3′; TCR Cβ (AE000665), 5′-AGGCTACCCTCGTGTGCTTG-3′ and 5′-TGACCAGCACAGCATATAGG-3′; β-actin (X03672), 5′-TGGGTCAGAAGGACTCCTATG-3′ and 5′-ACCAGACAGCACTGTGTTGGC-3′; CCR9 (AJ132336), 5′-TGCTGATCTGCTCTTTCTTG-3′ and 5′-GTGCTTGGATGACTTCTTGG-3′; CXCR4 (U59760), 5′-CTAGGATGAGGATGACTGTCG-3′ and 5′-ACCTCTACAGCAGCGTTCTC-3′; Aiolos (AF001293), 5′-ATCGAAGCAGTGCCGCTTCTCACC-3′ and 5′-GTGTGCGGGTTATCCTGCATTAGC-3′; PU.1 (M38252), 5′-TGTGCTTCCCTTATCAAACC-3′ and 5′-TCTTCACCTCGCCTGTCTTG-3′; Id2 (AF077860), 5′-GCGAGTGCACATAAAAGACG-3′ and 5′-ACGATAGTGGGATGCGAGTC-3′; Pax5 (NM_008782), 5′-ACGGCCACTCCCAGATGTAG-3′ and 5′-TCTTGCGTTTGTTGGTGTCG-3′; Notch-1 (Z11886), 5′-TATGCCAGGTTATGAAGGTG-3′ and 5′-TTCAGACTCCTTGCATACGC-3′; E2a (BC018260), 5′-CAGTGAGCGGAATGCCTATG-3′ and 5′-TGGATACAGCTGGGAGCGAG-3′. PCR used a single set of primers, except for pre-Tα analysis, which used two nested primer pairs (pre-Tα.1 and pre-Tα.2).

Total RNA was extracted from 3–6 × 104 sorted cells using an RNeasy Mini kit (Qiagen). RNA was reverse transcribed using random hexamer primers, and amplified by PCR using the following forward and reverse primers: 5′-TGGGTCAGAAGGACTCCTATG-3′ and 5′-ACCAGACAGCACTGTGTTGGC-3′ for β-actin; 5′-TACAAGGCCTCCAGACCAAG-3′ and 5′-AGGATTGTGCCAGAAGGTAG-3′ for the Vβ8-Cβ2 rearrangement; and 5′-GTGTCATCATGTCCCGTCTC-3′ and 5′-CAACCTTCCTCACAAATGTG-3′ for the Vα8-Cα rearrangement.

Flat-bottom 96-well plates (Corning Glass) were coated for 90 min with 10 mg/ml purified anti-TNP mAb (A19-3, hamster IgG1, isotype control) or a 10 mg/ml combination of anti-CD3 mAb and CD28 mAb (clones 2C11 and 37.51, respectively; both hamster IgG1) diluted in RPMI 1640 (BioWhittaker) in the 5% CO2 humidified incubator at 37°C. Following incubation, the coated wells were washed three times in complete RPMI 1640 medium.

Whole lymph node cells (1 × 105) from C57BL/6 wild-type, nu/nu, or CTP-reconstituted nu/nu mice were used for cytokine stimulation in triplicate in a final volume of 0.2 ml/well. These cells were incubated for 48 h in the absence or presence of plate-bound anti-TNP Ab or plate-bound anti-CD3/CD28 Abs, or treated with a combination of 20 ng/ml PMA (Sigma-Aldrich) and 1 μM ionomycin (Calbiochem). Following culture, the supernatants were harvested and frozen at −80°C. ELISA tests were performed following the manufacturer’s specifications using Cytoscreen IL-2, IL-4, IL-10, and IFN-γ mouse ELISA kits (BioSource International).

Nu/nu mice reconstituted with 500 CTP were prepared, as described above. As controls, nu/nu mice were reconstituted with 2 × 106 RAG-1−/− (CD45.2) bone marrow cells alone or together with 1 × 107 whole lymph node cells from CD45.1 C57BL/6 mice. Stocks of recombinant-type MCMV strain RM427+ (8) were grown and titered on NIH 3T3 cells (American Type Culture Collection CRL1658), as described (9). The titer of all stocks was confirmed at the time when mice were inoculated. All mice were infected i.p. with 106 PFU of MCMV 3 wk posttransplant. Mice were monitored for weight loss and signs of MCMV infection. Moribund mice were sacrificed, and the kidneys, liver, lungs, lymph nodes, salivary glands, and spleen were harvested for histological analysis.

Briefly, male BALB/c host mice were given a single dose of lethal whole body irradiation (950 cGy) from a 200 kV (20 mAb) source (Philips Medical Systems) at a rate of 72.5 cGy/min. Within 24 h after irradiation, the host mice were transplanted i.v. with 1.5 × 106 T cell-depleted bone marrow and graded numbers of whole lymph node cells from C57BL/6 mice, as described previously (10). Treatment with neomycin and penicillin was started concomitantly with the transplantation and terminated between 3 and 4 wk after transplantation. The mice were weighed weekly and monitored daily for survival and signs of GVHD, including kyphosis (hunched back), alopecia (hair loss), skin lesions, diarrhea, and facial swelling. Mean body weight and survival were determined 100 days after transplantation. Chimerism, as evidenced by the presence of H-2Kb+ cells in the spleen, was confirmed in all hosts by flow cytometry.

Multicolor analysis of surface markers on CTPs and CIPs was performed to look for patterns reported to be present on other early progenitor cells, such as common lymphoid progenitors (CLPs), and hemopoietic stem cells (HSCs) in the marrow, and thymic and fetal blood progenitors (11, 12, 13, 14, 15). One-color analyses of key markers were compared after bone marrow cells enriched with anti-Thy-1.2 immunomagnetic beads were gated for CD3B220Mac-1NK1.1(Lin) cells, as shown in Fig. 1,A. These cells were gated further for Thy-1.2high cells, and the gated cells were analyzed for expression of CD2. Two discrete populations of Thy-1.2highCD2 and Thy-1.2highCD2+ cells (CTPs and CIPs, respectively) were observed, as shown in Fig. 1,B. As previously shown (1), the gated CTPs expressed higher levels of CD16 than the CIPs (Fig. 1,D). In contrast, the CIPs expressed increased levels of CD44, CD5, and CD122 (IL-2Rβ) as compared with the CTPs (Fig. 1, C, G, and K, respectively). Fig. 1,E shows that the large majority of both CTPs and CIPs express little or no CD25 (IL-2Rα). However, there is a small subset of CTPs that is CD25+. The CTPs and CIPs expressed intermediate levels of CD24 (HSA) (Fig. 1 F). Taken together, the data indicate that the CTPs have a CD44+CD25CD16medium/highCD24+CD5 phenotype similar to the most undifferentiated T cell progenitors described previously in the adult thymus, and in the fetal thymus and blood (11, 13, 14). In contrast to the CTPs, the earliest progenitors in the thymus do not express high levels of Thy-1 (16). With the exception of CD25, the CIPs have the CD44highCD16mediumCD24+CD5+ phenotype, characteristic of T cell progenitors further along the T cell developmental pathway (11, 13, 14).

FIGURE 1.

Comprehensive analysis of CTPs and CIPs for expression of surface markers and genes found in pro-T cells, CLP, HSC, and DN progenitors in the thymus. Bone marrow cells were analyzed for lineage markers (A), and gated Lin cells were gated for Thy-1.2high cells and then analyzed for Thy-1.2 vs CD2 in B. Cells in boxes in B were gated for CTPs and CIPs. CTPs (open histograms) and CIPs (shaded histograms) were studied for expression of CD44 (C), CD16 (D), CD25 (E), CD24 (F), CD5 (G), Sca-1 (H), c-Kit (I), CD127 (J), and CD122 (K) using a fluorochrome not used for gating. Gated CIPs and CTPs were also studied for expression of CD3, CD4, and CD8. Gated CD3+, CD4+, and CD8+ cells from the normal spleen are shown for comparison in L, CD3; M, CD4; and N, CD8. Thick lines show spleen cells, thin lines show CIPs, and dashed lines show CTPs. Data shown display one of three representative experiments for surface staining. Total RNA from sorted CTPs and CIPs was analyzed for expression of genes associated with T cell differentiation by RT-PCR in O. cDNA was prepared and then amplified by PCR using gene-specific primer pairs, as indicated. Sorted splenic TCRαβ cells, and CD4CD8 DN thymocytes from C57BL/6 mice were used as controls.

FIGURE 1.

Comprehensive analysis of CTPs and CIPs for expression of surface markers and genes found in pro-T cells, CLP, HSC, and DN progenitors in the thymus. Bone marrow cells were analyzed for lineage markers (A), and gated Lin cells were gated for Thy-1.2high cells and then analyzed for Thy-1.2 vs CD2 in B. Cells in boxes in B were gated for CTPs and CIPs. CTPs (open histograms) and CIPs (shaded histograms) were studied for expression of CD44 (C), CD16 (D), CD25 (E), CD24 (F), CD5 (G), Sca-1 (H), c-Kit (I), CD127 (J), and CD122 (K) using a fluorochrome not used for gating. Gated CIPs and CTPs were also studied for expression of CD3, CD4, and CD8. Gated CD3+, CD4+, and CD8+ cells from the normal spleen are shown for comparison in L, CD3; M, CD4; and N, CD8. Thick lines show spleen cells, thin lines show CIPs, and dashed lines show CTPs. Data shown display one of three representative experiments for surface staining. Total RNA from sorted CTPs and CIPs was analyzed for expression of genes associated with T cell differentiation by RT-PCR in O. cDNA was prepared and then amplified by PCR using gene-specific primer pairs, as indicated. Sorted splenic TCRαβ cells, and CD4CD8 DN thymocytes from C57BL/6 mice were used as controls.

Close modal

We evaluated the expression of Sca-1, c-Kit, and IL-7Rα (CD127) on the CTPs and CIPs. The CTPs are predominantly Sca-1 (Fig. 1,H), c-Kit+ (Fig. 1,I), and IL-7Rα (Fig. 1,J). However, a small subset of CTPs is Sca-1low and IL-7Rαlow. In contrast, the CIPs are Sca-1low (Fig. 1,H), c-Kit+ (Fig. 1,I), and IL-7Rα−/low (Fig. 1 J). Background staining with isotype-matched irrelevant mAbs showed a peak channel of fluorescence that approximated channel 1 (data not shown), and overlapped with negative staining for CD25, Sca-1, CD5, CD127, and CD122 on CTPs.

In additional experiments, the expression of CD3, CD4, and CD8 on CTPs and CIPs was compared with that of mature splenic T cells. Fig. 1, L, M, and N, shows one-color analyses for the intensity of staining of each marker on the gated cell populations. For mature T cells, spleen cells were stained for CD3, CD4, or CD8 vs light scatter, and the intensity of staining for each positive population was determined. Whereas the mature T cells showed bright staining for CD3, CD4, and CD8, the intensity staining of the latter markers on CIPs and CTPs was dull and similar to background (Fig. 1, L, M, and N).

We analyzed CTPs and CIPs for expression of T cell lineage genes by RT-PCR analysis using bone marrow cells enriched with anti-Thy-1 beads and then sorted for Thy-1.2highCD3B220Mac-1NK1.1 (Thy-1.2highLin) cells that were either CD2+ or CD2. Sorted CD4CD8 double-negative (DN) thymocytes, used as a control population, showed intense RT-PCR product bands for RAG-1, RAG-2, pre-Tα, and PU.1 genes (Fig. 1 O). The control cells confirmed previous studies that showed RAG-1, RAG-2, pre-Tα, and PU.1 gene expression in the DN population (17, 18, 19, 20). In contrast, the sorted CIPs and control sorted splenic TCRαβ T cells showed either undetectable or faint bands for these genes. The sorted CIPs had intense bands for both the TCR Vβ8 and CD3ε genes, as did the sorted splenic control T cells and DN thymocytes. These genes are expected to be expressed in the latter control cells because the pre-TCR complex that contains both a rearranged TCRβ chain and CD3ε is present in the CD44CD25+ DN thymus subpopulation (11). However, only the CIPs and sorted T cells had intense bands for Vα8 indicative of TCRα chain gene rearrangement and expression. The CTPs expressed little or no RT-PCR product for either the rearranged TCRβ, TCRα, or CD3ε genes. This is consistent with the absence of a rearranged TCR Vβ8 gene segment. Furthermore, both CTP and CIP cells expressed very low levels of RNA for Notch-1 reported to be involved in T lineage commitment (21), whereas we could readily find PCR products for this receptor in sorted T cells and DN thymocyte populations.

Both CTPs and CIPs expressed RNA for the IL-7Rα gene and for the T cell-specific transcription factor genes GATA-3 and TCF-1 (22). Moreover, the CTPs expressed RNA for TCR Cβ, indicative of sterile transcripts made during the earliest stages of T cell development in extrathymic microenvironments (23). All the populations tested had RNA for Aiolos and E2a genes, expressed only in lymphoid cells, as well as low, but detectable levels of Id-2 RNA, which is restricted to NK and T cell development (24, 25, 26, 27).

Finally, both populations were tested for expression of RNA of the chemokine receptors CCR9 and CXCR4. CCR9 is expressed on both thymocytes and mature T cells (28), and is the receptor for the thymus-expressed chemokine (29). CXCR4 is the receptor for stromal cell-derived factor-1, a chemokine produced in multiple tissues, including the thymus and bone marrow (30). Both the CTP and CIP populations showed intense bands for these chemokine receptors (Fig. 1 O).

Our previous studies showed that CTPs generated mature T cells in the lymphoid tissues of irradiated euthymic adoptive hosts (2, 3). In the current study, CTPs were examined for their capacity to generate T cells in athymic hosts. CTPs obtained from the marrow of wild-type CD45.1 C57BL/6 mice were sorted (Fig. 2), and 250 were injected i.v. into groups of five CD45.2/C57BL/6 athymic nu/nu hosts that were coinjected with 1 × 106 CD45.2 RAG-1−/− marrow cells. The host lymph node, spleen, and bone marrow cells were harvested and analyzed 1, 2, 4, and 8 wk later. Two-color analysis of cells stained for TCRαβ vs CD45.1 was performed at each time point. At 1 wk, CD45.1+ cells accounted for <1% of cells in all three tissues (data not shown). However, at 2 wk, TCRαβ+ T cells derived from the CTPs accounted for 31.2% of lymph node cells (Fig. 2). The percentage increased to 49.5% at 8 wk. In control animals injected with RAG-1−/− marrow cells alone, only 0.02% of lymph node cells were TCRαβ+ T cells at 8 wk. Background control staining was up to 0.1%. Control nu/nu mice had ∼0.1–0.8% of TCRαβ+ T cells in the lymphoid tissues before irradiation and cell transfer (bottom panels, Fig. 2).

FIGURE 2.

Isolation of CTPs from wild-type (WT) bone marrow (BM), and flow cytometric assay of their ability to reconstitute mature T cells in the lymphoid tissues of irradiated nu/nu hosts. Far left upper panel, Shows staining of immunomagnetic bead-enriched C57BL/6/CD45.1 marrow cells for Thy-1.2 vs lineage-specific (Lin) markers, including B220, Mac-1, TCRαβ, and NK1.1. Box encloses Thy-1.2highLincells, and percentage enclosed in box is shown. The latter cells were gated, and staining with anti-Thy-1.2 vs anti-CD2 is shown in adjacent panel. CTPs enclosed in Thy-1.2highCD2 box were sorted, and 250 cells were injected into lethally irradiated (950 cGy) nu/nu C57BL/6 CD45.2 mice along with 1 × 106 whole bone marrow cells from RAG-2−/− C57BL/6 CD45.2 donors. Right panels, Show staining of lymphoid tissues of nu/nu hosts for TCRαβ vs donor CTP marker (CD45.1) at serial time points. Control hosts given 1 × 106 RAG-2−/− marrow cells without CTP (NO CTP) or 1 × 106 RAG-1−/− marrow cells with 250 sorted TCRαβ+ T cells from wild-type marrow (BMTC) are shown also. The staining of untreated nu/nu C57BL/6 CD45.2 mice for TCRαβ vs forward light scatter is shown in the three bottom panels, and boxes enclose T cells. In additional experiments, CD45.1 wild-type thymocytes were stained for CD4 vs CD8, and the CD4CD8 cells are shown in the box in the far left lower panel. The cells in the box were gated, and analyzed for CD44 vs CD25. CD44highCD25 cells (DNSI thymocytes) enclosed in the box in the adjacent panel were sorted, and 250 sorted cells were injected with RAG-2−/− marrow cells into nu/nu hosts. Outgrowth of T cells is shown in panels at 4 wk (DNSI).

FIGURE 2.

Isolation of CTPs from wild-type (WT) bone marrow (BM), and flow cytometric assay of their ability to reconstitute mature T cells in the lymphoid tissues of irradiated nu/nu hosts. Far left upper panel, Shows staining of immunomagnetic bead-enriched C57BL/6/CD45.1 marrow cells for Thy-1.2 vs lineage-specific (Lin) markers, including B220, Mac-1, TCRαβ, and NK1.1. Box encloses Thy-1.2highLincells, and percentage enclosed in box is shown. The latter cells were gated, and staining with anti-Thy-1.2 vs anti-CD2 is shown in adjacent panel. CTPs enclosed in Thy-1.2highCD2 box were sorted, and 250 cells were injected into lethally irradiated (950 cGy) nu/nu C57BL/6 CD45.2 mice along with 1 × 106 whole bone marrow cells from RAG-2−/− C57BL/6 CD45.2 donors. Right panels, Show staining of lymphoid tissues of nu/nu hosts for TCRαβ vs donor CTP marker (CD45.1) at serial time points. Control hosts given 1 × 106 RAG-2−/− marrow cells without CTP (NO CTP) or 1 × 106 RAG-1−/− marrow cells with 250 sorted TCRαβ+ T cells from wild-type marrow (BMTC) are shown also. The staining of untreated nu/nu C57BL/6 CD45.2 mice for TCRαβ vs forward light scatter is shown in the three bottom panels, and boxes enclose T cells. In additional experiments, CD45.1 wild-type thymocytes were stained for CD4 vs CD8, and the CD4CD8 cells are shown in the box in the far left lower panel. The cells in the box were gated, and analyzed for CD44 vs CD25. CD44highCD25 cells (DNSI thymocytes) enclosed in the box in the adjacent panel were sorted, and 250 sorted cells were injected with RAG-2−/− marrow cells into nu/nu hosts. Outgrowth of T cells is shown in panels at 4 wk (DNSI).

Close modal

Spleen cells of the nu/nu hosts injected with CTPs contained 11.5% of TCRαβ+ CD45.1+ cells at 2 wk, and the percentage rose to 39.8% at 8 wk. In the absence of CTPs, the host spleen cells contained 0.01% of the CD45.1+ T cells. A similar pattern was observed in the bone marrow, although the percentage of T cells was much lower than in the lymph nodes and spleen, as expected. It was possible that CD45.1+ T cell outgrowth from injected CTPs were generated from mature TCRαβ+ T cells that contaminated the purified CTPs. Accordingly, control irradiated athymic hosts were injected with RAG-1−/− marrow and 250 sorted bone marrow CD45.1 TCRαβ+ T cells (BMTC) instead of CD45.1 CTPs from the Thy-1.2-enriched donor bone marrow. Hosts that received the sorted T cells had <1% TCRαβ+ T cells in the lymph nodes, spleen, and bone marrow at 8 wk (Fig. 2). In additional experiments, we tested the ability of 250 sorted CD4CD8CD44highCD25 cells from the CD45.1 wild-type thymus (stage I DN thymocytes) to reconstitute the irradiated nu/nu hosts. Hosts that received the CD44highCD25 thymic progenitors had <1% TCRαβ+ T cells in the lymphoid tissues at 4 wk (Fig. 2).

The kinetics of the increase in the absolute number of CTP-derived T cells in the spleen, lymph nodes, and bone marrow are shown in Table I. At 2 wk after injection of 250 CTPs, a mean of 1.3 × 106 CD45.1+ T cells was harvested from the spleen, 0.54 × 106 from cervical and axillary lymph nodes, and 0.21 × 106 from the marrow cells from both femurs and tibias. The absolute numbers in the spleen and bone marrow increased ∼2-fold from 2 to 8 wk, and those in lymph nodes declined slightly (Table I). The total number of CD45.1+ T cells in all three tissues (∼3 × 106) indicates an expansion of ∼10,000-fold as compared with that of the injected CTPs (250 cells). As compared with untreated wild-type C57BL/6 mice, the absolute number was not reduced in the bone marrow, but was reduced 10-fold in the spleen and lymph nodes (data not shown). This suggests that the intact thymus pathway is still required to achieve wild-type levels of TCRαβ T cells in the periphery.

Table I.

Absolute number of donor-derived T cells in lymphoid tissues of adoptive hosts (mean ± SE × 106)

Time after Cell Injectiona (weeks)Host Tissues Analyzedb
Bone MarrowcSpleenLymph Noded
WT-CTP 0.21 ± 0.06 1.3 ± 0.7 0.54 ± 0.09 
WT-CTP 0.8 ± 0.1 1.0 ± 0.2 0.33 ± 0.08 
WT-CTP 0.40 ± 0.06 2.1 ± 0.8 0.38 ± 0.02 
WT-BMTC <0.001 <0.001 <0.001 
nu/nu-CTP <0.001 <0.001 <0.001 
WT-HSC <0.001 <0.001 <0.001 
WT-DN thymus cells <0.001 <0.001 <0.001 
WT-DNSI <0.001 <0.001 <0.001 
Time after Cell Injectiona (weeks)Host Tissues Analyzedb
Bone MarrowcSpleenLymph Noded
WT-CTP 0.21 ± 0.06 1.3 ± 0.7 0.54 ± 0.09 
WT-CTP 0.8 ± 0.1 1.0 ± 0.2 0.33 ± 0.08 
WT-CTP 0.40 ± 0.06 2.1 ± 0.8 0.38 ± 0.02 
WT-BMTC <0.001 <0.001 <0.001 
nu/nu-CTP <0.001 <0.001 <0.001 
WT-HSC <0.001 <0.001 <0.001 
WT-DN thymus cells <0.001 <0.001 <0.001 
WT-DNSI <0.001 <0.001 <0.001 
a

A total of 250 CTPs, BMTC, CD44highCD25 thymocytes (DNSI) or hemopoietic stem cells (HSC), or 1 × 104 CD4CD8 thymocytes (DNTC) from wild-type (WT) or nu/nu C57BL/6 donors injected along with 1 ×106 bone marrow cells from RAG-1−/− C57BL/6 mice sharing CD45 marker of host.

b

Lethally irradiated C57BL/6 nude host tissues stained for donor-type CD45 markers (n = 4).

c

Bone marrow cells harvested from two femurs and tibias from each host.

d

Cells harvested from cervical and axillary lymph nodes.

When 250 CTPs from the bone marrow of nu/nu C57BL/6 mice were used instead of wild-type CTPs to reconstitute the T cells of irradiated congenic nu/nu hosts, then the absolute numbers of CTP-derived T cells at 8 wk were <0.001 × 106 (below limit of deletion) in each tissue (Table I). In addition, the injection of 1 × 103 sorted Thy-1lowSca-1+c-Kit+ HSCs from the bone marrow of wild-type C57BL/6 donors into irradiated congenic nu/nu hosts resulted in minimal T cell reconstitution as compared with that with 250 wild-type CTPs (Table I). The latter result was expected, because wild-type stem cells fail to develop into functional CTPs in the nu/nu bone marrow after adoptive transfer (3). The injection of 250 sorted TCRαβ+ T cells (BMTC) from the wild-type bone marrow resulted in <0.001 × 106 T cells in the lymphoid tissues of the nu/nu hosts at 8 wk (Table I). When the dose of sorted T cells was increased to 1 × 104 cells, then ∼0.1 × 106 donor T cells were found in the spleen at 8 wk as compared with the 2.1 × 106 splenic T cells generated by 40-fold fewer CTPs (data not shown). When 250 sorted CD4CD8CD44highCD25 DN stage I (DNSI) thymocytes were used instead of CTPs, then the number of T cells that developed from the thymic progenitors was <0.001 × 106 in each tissue (Table I). Injection of 1 × 104 sorted CD4CD8 thymus cells (DN thymus cells) resulted in a similar lack of T cell outgrowth (Table I).

To determine whether CD45.1+ cells derived from the injected CTP were restricted to the T cell lineage, host spleen cells harvested at 4 wk were stained for a variety of markers that identify B cells (B220), macrophages (Mac-1), granulocytes (Gr-1), and NK cells (NK1.1) in addition to T cells (TCRαβ) and T cell-associated markers (CD4 and CD8). Fig. 3,A shows that ∼17% of host spleen cells contained CD45.1+ TCRαβ+ cells, and <1% of the spleen cells were CD45.1+ cells expressing the non-T cell lineage markers (B220, Gr-1, Mac-1, NK1.1). A small population of CD45.1 TCRαβlow/− and CD45.1 CD4CD8 cells was observed in the spleen (Fig. 3,B, top panels). To further characterize these cells, CD45.1-gated cells were analyzed for Thy-1.2 vs CD2 and TCRαβ+ vs CD2 markers. Fig. 3,B shows that 99.6% of the gated CD45.1 cells expressed both Thy-1.2 and CD2 markers, and ∼12% of these fail to express high levels of TCRαβ or any of the other markers. Thus, the latter cells derived from the injected CTP express the Thy-1+CD2+Lin phenotype present on intermediate T cell progenitors that are rapidly generated by CTPs in vitro (1), and are present also in the wild-type bone marrow (Fig. 1 B).

FIGURE 3.

Sorted wild-type Thy-1.2highCD2Lin bone marrow cells reconstitute CD4+ and CD8+ T cells in nu/nu hosts. A, Staining of irradiated nu/nu host spleen cells 4 wk after injection of 1 × 106 RAG-1−/− CD45.2 marrow cells and 250 wild-type CD45.1 CTPs for T cell (TCRαβ or combined CD4 and CD8), B cell (B220), granulocyte (Gr-1), macrophage (Mac-1), and NK cell (NK1.1) vs CD45.1 markers. B, Gated CD45.1 cells from the spleen at 4 wk were stained for TCRαβ vs CD2 and CD4 vs CD8 in two top panels. The gated CD45.1 cells were stained for Thy-1.2 vs CD2 in the lower left panel. Staining for CD4 vs CD8 on CD45.1-gated spleen cells at 8 wk is shown in lower right panel.

FIGURE 3.

Sorted wild-type Thy-1.2highCD2Lin bone marrow cells reconstitute CD4+ and CD8+ T cells in nu/nu hosts. A, Staining of irradiated nu/nu host spleen cells 4 wk after injection of 1 × 106 RAG-1−/− CD45.2 marrow cells and 250 wild-type CD45.1 CTPs for T cell (TCRαβ or combined CD4 and CD8), B cell (B220), granulocyte (Gr-1), macrophage (Mac-1), and NK cell (NK1.1) vs CD45.1 markers. B, Gated CD45.1 cells from the spleen at 4 wk were stained for TCRαβ vs CD2 and CD4 vs CD8 in two top panels. The gated CD45.1 cells were stained for Thy-1.2 vs CD2 in the lower left panel. Staining for CD4 vs CD8 on CD45.1-gated spleen cells at 8 wk is shown in lower right panel.

Close modal

The expression of CD4 vs CD8 markers on gated CD45.1 TCRαβ+ cells in the host spleen at 4 and 8 wk is also shown in Fig. 3,B. Approximately 65–68% of the gated cells were single-positive CD4+ cells, and ∼26–28% were single-positive CD8+ cells. The latter cells were almost all CD8αβ+ cells, as determined by staining for CD8α vs CD8β (data not shown). A small population (2.8%) of gated cells was CD4+CD8+ double-positive cells at 4 wk, and a discrete population of the latter cells was not observed at 8 wk. Some CD4CD8 T cells were observed at both time points (Fig. 3 B). Thus, almost all cells derived from the injected CTPs were members of the T cell lineage, and among the latter cells almost all were single-positive CD4+ or CD8+ T cells.

To provide an estimate of the precursor frequency of functional progenitors, we injected 100, 150, or 250 CTPs into groups of 10 irradiated nu/nu hosts, and observed the fraction without T cell reconstitution. After injection, 0 of 10, 6 of 10, and 10 of 10 hosts, respectively, showed reconstitution of the lymph nodes 4 wk later. This suggests that the precursor frequency is in the range of 1 precursor in ∼100–250 purified CTPs. Wild-type CD45.1 CTPs generated a normal repertoire of Vβ receptors near limit dilution (250 CTPs injected), as judged by the percentage of Vβ3+, Vβ6+, and Vβ8+ (2.8, 7.0, and 19.4%, respectively) among lymph node T (Thy-1.2+) cells in reconstituted hosts at 4 wk (Fig. 4 A). These percentages were similar to those reported among TCRαβ+ T cells in the normal C57BL/6 spleen (31).

FIGURE 4.

Vβ receptor expression of donor T cell reconstitution of lethally irradiated nu/nu hosts with wild-type CTPs. A, Representative staining of lymph node cells for Thy-1.2 vs Vβ3, Vβ6, or Vβ8 receptors is shown for a host given 250 CTPs. B, RT-PCR analysis of mRNA for Vβ8 and Vα8 gene expression using specific primer pairs that hybridized to reverse-transcribed cDNA obtained from control untreated nu/nu lymph node cells (lane 1), sorted wild-type donor Thy-1.2highLinCD2 CTPs (lane 2), sorted donor-type CD45.1+ TCRαβ+ T cells from lymph nodes of irradiated nu/nu hosts reconstituted with 250 wild-type CTPs (lane 3), and sorted TCRαβ+ T cells from the lymph nodes of untreated wild-type C57BL/6 mice (lane 4). PCR with no cDNA template added is shown in lane 5, and amplification of β-actin T cell cDNA is shown for all samples to ensure adequacy of RNA yields and amplifications.

FIGURE 4.

Vβ receptor expression of donor T cell reconstitution of lethally irradiated nu/nu hosts with wild-type CTPs. A, Representative staining of lymph node cells for Thy-1.2 vs Vβ3, Vβ6, or Vβ8 receptors is shown for a host given 250 CTPs. B, RT-PCR analysis of mRNA for Vβ8 and Vα8 gene expression using specific primer pairs that hybridized to reverse-transcribed cDNA obtained from control untreated nu/nu lymph node cells (lane 1), sorted wild-type donor Thy-1.2highLinCD2 CTPs (lane 2), sorted donor-type CD45.1+ TCRαβ+ T cells from lymph nodes of irradiated nu/nu hosts reconstituted with 250 wild-type CTPs (lane 3), and sorted TCRαβ+ T cells from the lymph nodes of untreated wild-type C57BL/6 mice (lane 4). PCR with no cDNA template added is shown in lane 5, and amplification of β-actin T cell cDNA is shown for all samples to ensure adequacy of RNA yields and amplifications.

Close modal

We performed RT-PCR analysis on mRNA from the sorted CTPs from wild-type C57BL/6 bone marrow to look for expression of Vβ8 and Vα8 genes, and failed to detect an appropriate size cDNA fragment (Fig. 4,B). However, the sorted CD45.1+ TCRαβ+ T cells derived from CTPs injected into the congenic nu/nu hosts showed expression of both genes (Fig. 4,B). Sorted lymph node T cells from wild-type mice showed expression of both genes also, but an equivalent number of whole nu/nu lymph node cells failed to provide a detectable signal (Fig. 4,B). Samples from all cell sources showed a dense band for β-actin after amplification (Fig. 4 B).

The results using adoptive nu/nu C57BL/6 congenic hosts indicated that the transferred CTPs generate mature CD4+ and CD8+ T cells with a full Vβ receptor repertoire. To study negative selection, we injected 500 CTPs from the bone marrow of Thy-1.1 BALB/c mice donors along with 1 × 106 Thy-1.2 BALB/c RAG-2−/− bone marrow cells into lethally irradiated Thy-1.2 nu/nu BALB/c hosts. We compared the Vβ repertoires of sorted CTP-derived T cells (Thy-1.1+) from the spleens of BALB/c nu/nu hosts 4 wk after the adoptive transfer with that of wild-type BALB/c and C57BL/6 T cells.

The percentages of Vβ receptor-expressing cells in the spleen among gated Thy-1.1+ cells were determined, and the ratios of each Vβ to that of the predominant nondeleted Vβ8+ cells were compared. As shown in Table II, statistically significant reductions of these Vβn:Vβ8 ratios (p < 0.05; Student’s t test of independent means) were observed for Vβ3, Vβ5, Vβ11, and Vβ12 receptors that are expected to be deleted in BALB/c as compared with wild-type C57BL/6 mice (Table II). The Vβ receptor deletion pattern observed in the CTP-reconstituted nu/nu BALB/c-adoptive hosts was similar to that observed in the wild-type BALB/c mice (Table II). Thus, negative selection of the CTP-derived T cells occurred extrathymically in vivo.

Table II.

Analysis of Vβ receptor expression among T cells in the spleen of BALB/c and C57BL/6 wild-type control and CTP-reconstituted nu/nu hosts

Source of Spleen CellsRatio of the Percentages of Vβn+:Vβ8+ T Cellsa
Vβ2Vβ3Vβ4Vβ5.1/5.2Vβ6Vβ9Vβ11Vβ12
BALB/c wild-type controlb 0.26 ± 0.07 0.08 ± 0.01 0.31 ± 0.10 0.09 ± 0.00 0.54 ± 0.1 0.10 ± 0.02 0.07 ± 0.01 0.02 ± 0.00 
C57BL/6 wild-type controlb 0.28 ± 0.01 0.17 ± 0.02 0.28 ± 0.04 0.44 ± 0.09 0.40 ± 0.05 0.10 ± 0.03 0.39 ± 0.04 0.29 ± 0.07 
BALB/c nu/nu host-CTP reconstituted 0.25 ± 0.02 0.04 ± 0.00 0.35 ± 0.05 0.10 ± 0.02 0.6 ± 0.09 0.15 ± 0.01 0.01 ± 0.00 0.02 ± 0.01 
Source of Spleen CellsRatio of the Percentages of Vβn+:Vβ8+ T Cellsa
Vβ2Vβ3Vβ4Vβ5.1/5.2Vβ6Vβ9Vβ11Vβ12
BALB/c wild-type controlb 0.26 ± 0.07 0.08 ± 0.01 0.31 ± 0.10 0.09 ± 0.00 0.54 ± 0.1 0.10 ± 0.02 0.07 ± 0.01 0.02 ± 0.00 
C57BL/6 wild-type controlb 0.28 ± 0.01 0.17 ± 0.02 0.28 ± 0.04 0.44 ± 0.09 0.40 ± 0.05 0.10 ± 0.03 0.39 ± 0.04 0.29 ± 0.07 
BALB/c nu/nu host-CTP reconstituted 0.25 ± 0.02 0.04 ± 0.00 0.35 ± 0.05 0.10 ± 0.02 0.6 ± 0.09 0.15 ± 0.01 0.01 ± 0.00 0.02 ± 0.01 
a

Mean ± SE of ratios of Vβn+:Vβ8+ T cells in the spleen at 4 wk. Vβ T cells in reconstituted spleen shown in bold.

b

There were five mice in control groups, and three mice in CTP-reconstituted groups.

To determine whether the T cells in the lymph nodes of the CTP-reconstituted nu/nu mice were functional, in vitro assays of cytokine secretion were performed in which 1 × 105 lymph node cells from the latter mice were harvested 8 wk after reconstitution with 250 CTPs and compared with an equal number of lymph node cells obtained from untreated C57BL/6 wild-type mice and untreated C57BL/6 nu/nu mice. Cells were stimulated with a combination of anti-CD3 and anti-CD28 mAbs bound to the surface of 96-well plates, and control cultures contained an irrelevant (anti-TNP) mAb or no mAb. Supernatants were collected at 48 h and cytokines were measured by ELISA. Fig. 5, A and B, shows that wild-type lymph node cells secreted ∼1400 pg/ml IL-2 and 400 pg/ml IFN-γ. Although the CTP-reconstituted nu/nu lymph node cells secreted ∼10-fold less IL-2, they secreted similar levels of IFN-γ (∼350 pg/ml). Control cultures with irrelevant mAb or no mAb secreted no detectable IL-2 or IFN-γ, and lymph node cells from untreated nu/nu mice failed to secrete the cytokines in response to anti-CD3/CD28 stimulation. Supernatants were also assayed for the presence of IL-4 and IL-10, and all culture supernatants, including those with wild-type lymph node cells, contained <100 pg/ml cytokine after stimulation with anti-CD3/CD28 mAb (data not shown).

FIGURE 5.

Extrathymically derived T cells secrete IFN-γ, protect against MCMV infection, but do not induce GVHD. A and B, Cytokine analysis after stimulation with plate-bound anti-CD3/CD28 mAb. A total of 1 × 104 unfractionated lymph node cells from C57BL/6 wild-type, C57BL/6 nu/nu, and CTP-transplanted C57BL/6 nu/nu mice (4 wk posttransplant) was cultured for 48 h in medium (left bars) with or without stimulation by plate-bound anti-TNP mAb (middle bars) and plate-bound anti-CD3/CD28 mAb (right bars). ELISA was performed to determine the levels of IL-2 (A) or IFN-γ (B). Mean and SD from six independent experiments are shown. Asterisks show no cytokine was detected. C, Lethally irradiated (950 cGy) CD45.2 C57BL/6 nu/nu mice were injected with 2 × 106 CD45.2 RAG-1−/− bone marrow cells, alone or together with 500 CD45.1 CTP, or 107 CD45.1 C57BL/6 wild-type lymph node cells. Three weeks posttransplant, the mice were challenged with 106 PFU of MCMV (clone RM427+). Host mice were monitored for signs of MCMV infection and survival over a period of 56 days; RAG-1−/− bone marrow alone () (n = 5), RAG-1−/− bone marrow and 500 CD45.1 CTPs (- - -) (n = 10), and RAG-1−/− bone marrow and 107 wild-type CD45.1 lymph node cells (▮ ▮ ▮ ▮ ▮ ▮) (n = 5). One experiment of two with similar outcomes is shown. D and E, Unfractionated lymph node cells from C57BL/6 nu/nu, or CTP-reconstituted C57BL/6 nu/nu mice (4 wk posttransplant), or from C57BL/6 wild-type mice, together with a constant number (1.5 × 106) of T cell-depleted bone marrow cells from C57BL/6 wild-type mice were injected i.v. into lethally irradiated (950 cGy) BALB/c hosts. Control hosts received marrow without lymph node cells. D, Host survival over a 100-day period is shown for groups of 10 mice; no lymph node cells (— —), 2 × 105nu/nu lymph node cells (- - - -), 2 × 105 CTP-reconstituted nu/nu lymph node cells (- ▪ - ▪), 0.5 × 105 wild-type lymph node cells (▮ ▮ ▮ ▮ ▮ ▮), and 2 × 105 wild-type lymph node cells (). E, Host body weight; no lymph node cells (-▪-), 2 × 105nu/nu lymph node cells (-□-), 2 × 105 CTP-reconstituted nu/nu lymph nodes (-•-), 0.5 × 105 wild-type lymph node cells (-▴-), and 2 × 105 wild-type lymph node cells (-○-). Each point represents the mean and SD.

FIGURE 5.

Extrathymically derived T cells secrete IFN-γ, protect against MCMV infection, but do not induce GVHD. A and B, Cytokine analysis after stimulation with plate-bound anti-CD3/CD28 mAb. A total of 1 × 104 unfractionated lymph node cells from C57BL/6 wild-type, C57BL/6 nu/nu, and CTP-transplanted C57BL/6 nu/nu mice (4 wk posttransplant) was cultured for 48 h in medium (left bars) with or without stimulation by plate-bound anti-TNP mAb (middle bars) and plate-bound anti-CD3/CD28 mAb (right bars). ELISA was performed to determine the levels of IL-2 (A) or IFN-γ (B). Mean and SD from six independent experiments are shown. Asterisks show no cytokine was detected. C, Lethally irradiated (950 cGy) CD45.2 C57BL/6 nu/nu mice were injected with 2 × 106 CD45.2 RAG-1−/− bone marrow cells, alone or together with 500 CD45.1 CTP, or 107 CD45.1 C57BL/6 wild-type lymph node cells. Three weeks posttransplant, the mice were challenged with 106 PFU of MCMV (clone RM427+). Host mice were monitored for signs of MCMV infection and survival over a period of 56 days; RAG-1−/− bone marrow alone () (n = 5), RAG-1−/− bone marrow and 500 CD45.1 CTPs (- - -) (n = 10), and RAG-1−/− bone marrow and 107 wild-type CD45.1 lymph node cells (▮ ▮ ▮ ▮ ▮ ▮) (n = 5). One experiment of two with similar outcomes is shown. D and E, Unfractionated lymph node cells from C57BL/6 nu/nu, or CTP-reconstituted C57BL/6 nu/nu mice (4 wk posttransplant), or from C57BL/6 wild-type mice, together with a constant number (1.5 × 106) of T cell-depleted bone marrow cells from C57BL/6 wild-type mice were injected i.v. into lethally irradiated (950 cGy) BALB/c hosts. Control hosts received marrow without lymph node cells. D, Host survival over a 100-day period is shown for groups of 10 mice; no lymph node cells (— —), 2 × 105nu/nu lymph node cells (- - - -), 2 × 105 CTP-reconstituted nu/nu lymph node cells (- ▪ - ▪), 0.5 × 105 wild-type lymph node cells (▮ ▮ ▮ ▮ ▮ ▮), and 2 × 105 wild-type lymph node cells (). E, Host body weight; no lymph node cells (-▪-), 2 × 105nu/nu lymph node cells (-□-), 2 × 105 CTP-reconstituted nu/nu lymph nodes (-•-), 0.5 × 105 wild-type lymph node cells (-▴-), and 2 × 105 wild-type lymph node cells (-○-). Each point represents the mean and SD.

Close modal

In additional experiments, CTP-reconstituted nu/nu mice were challenged with MCMV to determine whether the extrathymically derived T cells afforded protection against a lethal viral infection. The irradiated CD45.2 nu/nu hosts were reconstituted with 2 × 106 bone marrow cells from CD45.2 RAG-1−/− mice and 500 CTPs from CD45.1 wild-type mice. Control mice received either RAG-1−/− bone marrow cells alone or RAG-1−/− bone marrow cells and 1 × 107 lymph node cells from CD45.1 wild-type mice. Three weeks after the injection of the cells, the three groups of mice were challenged with 1 × 106 PFU of MCMV, and mice were monitored for weight loss, clinical signs, and survival thereafter. As shown in Fig. 5,C, mice reconstituted with RAG-1−/− bone marrow cells alone all died by 24 days after the MCMV challenge. These mice developed diffuse abdominal swelling (ascites) ∼1 wk before they died. The mice reconstituted with marrow cells and wild-type CTPs remained healthy after the MCMV injection, and all survived without weight loss until day 56 when they were sacrificed for harvesting of tissues (Fig. 5 C). At that time, ∼50% of cells in the lymph nodes were TCRαβ+ T cells derived from CTPs (CD45.1). The group given marrow and wild-type lymph node cells remained healthy also until sacrifice at day 56.

Histopathological analysis of tissue sections of the liver stained with H&E was obtained either at the time of death in the group given marrow cells alone or at the time of sacrifice for the two other groups. Multiple patchy areas of liver parenchymal cell necrosis associated with inflammatory cell infiltrates were observed in the group given marrow cells alone, and no evidence of necrosis or inflammation was observed in the group given marrow and wild-type lymph node cells (data not shown). Although some areas of necrosis and inflammation were observed in the livers from mice given marrow cells and CTPs, the extent was reduced as compared with that of mice given marrow cells alone.

Protection against MCMV is predominantly mediated by CD8+ T cells (32). We also tested the functional capacity of lymph node cells from CTP-reconstituted C57BL/6 nu/nu mice to induce lethal GVHD in BALB/c mice, because tissue injury in this disease is mediated predominantly by CD4+ T cells (10). Groups of lethally irradiated (950 cGy) BALB/c mice were given injections of T cell-depleted wild-type C57BL/6 bone marrow cells alone or with 0.5 × 105 or 2 × 105 lymph node cells from CTP-reconstituted nu/nu mice, untreated nu/nu mice, or untreated wild-type mice (Fig. 5, D and E). Whereas the injection of 0.5 and 2 × 105 wild-type lymph node cells caused diarrhea, weight loss, and the death of 20 and 100% of the allogeneic hosts, respectively, during a 100-day observation period, all mice in the groups given either no lymph node cells, or 2 × 105 CTP-reconstituted nu/nu lymph node cells, or 2 × 105nu/nu lymph node cells remained healthy and survived at least 100 days (Fig. 5, D and E). Thus, the CTP-reconstituted lymph node cells had little capacity to induce lethal GVHD even when the absolute number of lymph node CD4+ T cells injected was at least as high as that of untreated wild-type CD4+ T cells (2 × 105nu/nu reconstituted vs 0.5 × 105 wild-type lymph node cells injected, respectively).

Our data show that the CTPs in the normal marrow have the CD16+CD44highCD25CD24+CD5 phenotype and the unrearranged TCR Vβ gene pattern of immature T cell precursors that are similar to stage I DN progenitors in the thymus (11, 14). CTPs rapidly generate CIPs with rearranged Vβ genes in vitro (1). In the thymus, rearrangement of the TCRβ locus is mirrored by the down-regulation of CD16 with the concomitant up-regulation of CD2 (14) and CD5 (13). The current phenotypic analysis shows that decreased expression of CD16 and increased expression of CD2 and CD5 are associated with rearrangement of Vβ genes in the CIPs. In the thymus, up-regulation of CD25 and down-regulation of CD44 expression indicate progression through the T cell developmental pathway (11). Unlike DN thymocytes, the CIPs do not show either up-regulation of CD25 nor down-regulation of CD44. Thus, they are not similar to stages II, III, or IV of thymic T cell development, as judged by surface receptor expression. Both CTPs and CIPs expressed transcripts for GATA-3 and TCF-1, transcription factors exclusively expressed on cells of the T cell lineage (22, 33).

Although we could not detect RNA for CD3ε or TCR Vβ8, in the CTPs, the CIPs expressed RNA for both. However, the CTP TCR gene loci was transcriptionally active, as evidenced by the expression of sterile TCR Cβ RNA, but showed no V-J rearrangements. Both CTPs and CIPs expressed RNA for IL-7Rα, Aiolos, an Ikaros-related transcription factor only expressed on lymphoid cells (24, 25), and for Id2 (26), a transcription factor involved in fate decision between T vs NK cells. The lack of Notch1 expression in CTPs and CIPs and presence in thymocytes may reflect differences in the kinetics of expression during thymic vs extrathymic T cell development and commitment. Together, the data indicate that CTPs are early committed progenitors, as judged by both surface markers and the expression of a variety of T cell-specific genes.

After adoptive transfer, wild-type CTPs reconstituted mature TCRαβ T cells in the lymph nodes, spleen, and bone marrow of adoptive nu/nu hosts at ∼2 wk after cell transfer, as judged by staining for the TCRαβ and the congenic CD45 markers. We could not determine the kinetics of expansion of progenitors vs mature T cells at 1 wk, because there were too few CTP-derived cells to analyze. HSCs from the bone marrow of wild-type donors and CTPs from nu/nu donors were unable to reconstitute the lymphoid tissues. This was expected, because HSCs generate only functionally defective CTPs in the nu/nu bone marrow (3). Reconstitution with wild-type CTPs indicates that the nu/nu lymphoid tissues can support the extrathymic maturation of T cells after the CTP stage of development has been achieved. Control infusions of mature TCRαβ+ T cells from the wild-type bone marrow showed that expansion of TCRαβ+ T cells that contaminate the sorted CTPs cannot account for the reconstitution of mature T cells in the irradiated nu/nu lymphoid tissues. Lack of homeostatic expansion of bone marrow TCRαβ+ T cells was expected, because these T cells produce markedly reduced levels of IL-2 as compared with peripheral T cells (10). In contrast to the CTPs, CD4CD8CD44highCD25 stage I progenitors in the thymus failed to reconstitute nu/nu hosts despite their similarities in surface phenotype and T cell lineage gene expression.

We found that the injection of 250 CTPs containing one to two precursors generated a normal distribution of Vβ receptors among reconstituted splenic T cells when compared with the distribution among splenic T cells in untreated wild-type BALB/c mice. In addition, the marked reduction in the percentage of T cells expressing Vβ3, Vβ5, Vβ11, and Vβ12 as compared with that of C57BL/6 wild-type mice showed that the reconstituted T cells had been appropriately negatively selected. Our previous in vitro studies suggested that the bone marrow stroma can perform the role of positive and negative selection during extrathymic T cell maturation, in the same way that the thymic stroma performs this function during intrathymic T cell maturation (31).

Functional tests of lymph node T cells from CTP-reconstituted nu/nu hosts showed that the cells vigorously secreted IFN-γ and but little IL-2 after in vitro stimulation with anti-CD3 and anti-CD28 mAbs. The reconstituted hosts were afforded protection against lethal MCMV infection. Protection against MCMV infection has been reported to be mediated predominantly by CD8+ T cells (32). The results suggest that the CTP extrathymically derived CD8+ T cells are immunocompetent. These results are consistent with a recent report that CLP (12) can generate CD8+ T cells in irradiated adoptive hosts that contribute to the protection against lethal MCMV infection (34). Interestingly, the protective CD8+ T cells appear to be generated extrathymically, because the percentage of CLP-derived CD8+ T cells in the spleen and protection against MCMV infection were similar in the thymectomized as compared with euthymic allogeneic hosts (34). CLP-protected hosts had reduced viral loads. The meager secretion of IL-2 and inability of the CTP-derived lymph node cells to mediate GVHD in irradiated BALB/c hosts in the current study indicated that the CTP-derived CD4+ T cells were defective as compared with wild-type CD4+ T cells.

In conclusion, we show that CTPs and CIPs are committed T cell lineage progenitors that express T cell-specific genes and gene rearrangements previously reported in the thymus. The data provide evidence that commitment to the T cell lineage can occur in the bone marrow as well as in the thymus, and that these marrow progenitors can generate mature T cells via an extrathymic pathway in athymic nu/nu mice.

We thank Dr. Leonore Herzenberg for her helpful discussions.

The authors have no financial conflict of interest.

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 study was supported in part by National Institutes of Health Grants AI-43013, HL-57443, and CA-92225.

4

Abbreviations used in this paper: CTP, committed T cell progenitor; BMTC, bone marrow T cell; CIP, committed intermediate progenitor; CLP, common lymphoid progenitor; DN, double negative; GVHD, graft-vs-host disease; HSC, hemopoietic stem cell; MCMV, murine CMV; TCF, T cell-specific factor; TNP, trinitrophenol; DNSI, DN stage I.

1
Dejbakhsh-Jones, S., S. Strober.
1999
. Identification of an early T cell progenitor for a pathway of T cell maturation in the bone marrow.
Proc. Natl. Acad. Sci. USA
96
:
14493
.-14498.
2
Dejbakhsh-Jones, S., M. E. Garcia-Ojeda, D. Chatterjea-Matthes, D. Zeng, S. Strober.
2001
. Clonable progenitors committed to the T lymphocyte lineage in the mouse bone marrow; use of an extrathymic pathway.
Proc. Natl. Acad. Sci. USA
98
:
7455
.-7460.
3
Chatterjea-Matthes, D., M. E. Garcia-Ojeda, S. Dejbakhsh-Jones, L. Jerabek, M. G. Manz, I. L. Weissman, S. Strober.
2003
. Early defect in prethymic bone marrow T cell progenitors in athymic nu/nu mice.
J. Immunol.
171
:
1207
.-1215.
4
Klein, F., N. Feldhahn, S. Lee, H. Wang, F. Ciuffi, M. Von Elstermann, M. L. Toribio, H. Sauer, M. Wartenberg, V. S. Barath, et al
2003
. T lymphoid differentiation in human bone marrow.
Proc. Natl. Acad. Sci. USA
100
:
6747
.-6752.
5
Parks, D. R., L. A. Herzenberg.
1984
. Fluorescence-activated cell sorting: theory, experimental optimization, and applications in lymphoid cell biology.
Methods Enzymol.
108
:
197
.-241.
6
Taubenberger, J. K., A. H. Reid, D. Izon, S. A. Boehme.
1996
. Development and characterization of v-myc/v-raf-transformed murine fetal thymocyte cell lines.
Cell. Immunol.
171
:
41
.-47.
7
Benson, D. A., M. S. Boguski, D. J. Lipman, J. Ostell, B. F. Ouellette, B. A. Rapp, D. L. Wheeler.
1999
. GenBank.
Nucleic Acids Res.
27
:
12
.-17.
8
Saederup, N., Y. C. Lin, D. J. Dairaghi, T. J. Schall, E. S. Mocarski.
1999
. Cytomegalovirus-encoded β chemokine promotes monocyte-associated viremia in the host.
Proc. Natl. Acad. Sci. USA
96
:
10881
.-10886.
9
Manning, W. C., C. A. Stoddart, L. A. Lagenaur, G. B. Abenes, E. S. Mocarski.
1992
. Cytomegalovirus determinant of replication in salivary glands.
J. Virol.
66
:
3794
.-3802.
10
Zeng, D., P. Hoffmann, F. Lan, P. Huie, J. Higgins, S. Strober.
2002
. Unique patterns of surface receptors, cytokine secretion, and immune functions distinguish T cells in the bone marrow from those in the periphery: impact on allogeneic bone marrow transplantation.
Blood
99
:
1449
.-1457.
11
Godfrey, D. I., J. Kennedy, T. Suda, A. Zlotnik.
1993
. A developmental pathway involving four phenotypically and functionally distinct subsets of CD3CD4CD8 triple-negative adult mouse thymocytes defined by CD44 and CD25 expression.
J. Immunol.
150
:
4244
.-4252.
12
Kondo, M., I. L. Weissman, K. Akashi.
1997
. Identification of clonogenic common lymphoid progenitors in mouse bone marrow.
Cell
91
:
661
.-672.
13
Pearse, M., P. Gallagher, A. Wilson, L. Wu, N. Fisicaro, J. F. Miller, R. Scollay, K. Shortman.
1988
. Molecular characterization of T-cell antigen receptor expression by subsets of CD4CD8 murine thymocytes.
Proc. Natl. Acad. Sci. USA
85
:
6082
.-6086.
14
Rodewald, H. R., K. Awad, P. Moingeon, L. D’Adamio, D. Rabinowitz, Y. Shinkai, F. W. Alt, E. L. Reinherz.
1993
. FcγRII/III and CD2 expression mark distinct subpopulations of immature CD4CD8 murine thymocytes: in vivo developmental kinetics and T cell receptor β chain rearrangement status.
J. Exp. Med.
177
:
1079
.-1092.
15
Leclercq, G., V. Debacker, M. de Smedt, J. Plum.
1996
. Differential effects of interleukin-15 and interleukin-2 on differentiation of bipotential T/natural killer progenitor cells.
J. Exp. Med.
184
:
325
.-336.
16
Matsuzaki, Y., J. I. Gyotoku, M. Ogawa, S. Nishikawa, Y. Katsura, G. Gachelin, H. Nakauchi.
1993
. Characterization of c-kit positive intrathymic stem cells that are restricted to lymphoid differentiation.
J. Exp. Med.
178
:
1283
.-1292.
17
Bruno, L., B. Rocha, A. Rolink, H. von Boehmer, H. R. Rodewald.
1995
. Intra- and extra-thymic expression of the pre-T cell receptor α gene.
Eur. J. Immunol.
25
:
1877
.-1882.
18
Hozumi, K., A. Kobori, T. Sato, H. Nozaki, S. Nishikawa, T. Nishimura, S. Habu.
1994
. Pro-T cells in fetal thymus express c-kit and RAG-2 but do not rearrange the gene encoding the T cell receptor β chain.
Eur. J. Immunol.
24
:
1339
.-1344.
19
Anderson, M. K., G. Hernandez-Hoyos, R. A. Diamond, E. V. Rothenberg.
1999
. Precise developmental regulation of Ets family transcription factors during specification and commitment to the T cell lineage.
Development
126
:
3131
.-3148.
20
Spain, L. M., A. Guerriero, S. Kunjibettu, E. W. Scott.
1999
. T cell development in PU.1-deficient mice.
J. Immunol.
163
:
2681
.-2687.
21
Allman, D., F. G. Karnell, J. A. Punt, S. Bakkour, L. Xu, P. Myung, G. A. Koretzky, J. C. Pui, J. C. Aster, W. S. Pear.
2001
. Separation of Notch1 promoted lineage commitment and expansion/transformation in developing T cells.
J. Exp. Med.
194
:
99
.-106.
22
Hattori, N., H. Kawamoto, S. Fujimoto, K. Kuno, Y. Katsura.
1996
. Involvement of transcription factors TCF-1 and GATA-3 in the initiation of the earliest step of T cell development in the thymus.
J. Exp. Med.
184
:
1137
.-1147.
23
Soloff, R. S., T. G. Wang, L. Lybarger, D. Dempsey, R. Chervenak.
1995
. Transcription of the TCR-β locus initiates in adult murine bone marrow.
J. Immunol.
154
:
3888
.-3901.
24
Morgan, B., L. Sun, N. Avitahl, K. Andrikopoulos, T. Ikeda, E. Gonzales, P. Wu, S. Neben, K. Georgopoulos.
1997
. Aiolos, a lymphoid restricted transcription factor that interacts with Ikaros to regulate lymphocyte differentiation.
EMBO J.
16
:
2004
.-2013.
25
Georgopoulos, K., S. Winandy, N. Avitahl.
1997
. The role of the Ikaros gene in lymphocyte development and homeostasis.
Annu. Rev. Immunol.
15
:
155
.-176.
26
Quong, M. W., W. J. Romanow, C. Murre.
2002
. E protein function in lymphocyte development.
Annu. Rev. Immunol.
20
:
301
.-322.
27
Engel, I., C. Johns, G. Bain, R. R. Rivera, C. Murre.
2001
. Early thymocyte development is regulated by modulation of E2A protein activity.
J. Exp. Med.
194
:
733
.-745.
28
Zaballos, A., J. Gutierrez, R. Varona, C. Ardavin, G. Marquez.
1999
. Cutting edge: identification of the orphan chemokine receptor GPR-9-6 as CCR9, the receptor for the chemokine TECK.
J. Immunol.
162
:
5671
.-5675.
29
Norment, A. M., L. Y. Bogatzki, B. N. Gantner, M. J. Bevan.
2000
. Murine CCR9, a chemokine receptor for thymus-expressed chemokine that is up-regulated following pre-TCR signaling.
J. Immunol.
164
:
639
.-648.
30
Kim, C. H., L. M. Pelus, J. R. White, H. E. Broxmeyer.
1998
. Differential chemotactic behavior of developing T cells in response to thymic chemokines.
Blood
91
:
4434
.-4443.
31
Garcia-Ojeda, M. E., S. Dejbakhsh-Jones, I. L. Weissman, S. Strober.
1998
. An alternate pathway for T cell development supported by the bone marrow microenvironment: recapitulation of thymic maturation.
J. Exp. Med.
187
:
1813
.-1823.
32
Reddehase, M. J., W. Mutter, K. Munch, H. J. Buhring, U. H. Koszinowski.
1987
. CD8-positive T lymphocytes specific for murine cytomegalovirus immediate-early antigens mediate protective immunity.
J. Virol.
61
:
3102
.-3108.
33
Ting, C. N., M. C. Olson, K. P. Barton, J. M. Leiden.
1996
. Transcription factor GATA-3 is required for development of the T-cell lineage.
Nature
384
:
474
.-478.
34
Arber, C., A. BitMansour, T. E. Sparer, J. P. Higgins, E. S. Mocarski, I. L. Weissman, J. A. Shizuru, J. M. Brown.
2003
. Common lymphoid progenitors rapidly engraft and protect against lethal murine cytomegalovirus infection after hematopoietic stem cell transplantation.
Blood
102
:
421
.-428.