Hematopoietic precursors continuously colonize the thymus where they give rise mainly to T cells, but also to B and dendritic cells. The lineage relationship between these three cell types is unclear, and it remains to be determined if precursors in the thymus are multipotent, oligopotent, or lineage restricted. Resolution of this question necessitates the determination of the clonal differentiation potential of the most immature precursors in the thymus. Using a CC chemokine receptor 9–enhanced green fluorescent protein knock-in allele like a surface marker of unknown function, we identify a multipotent precursor present in bone marrow, blood, and thymus. Single cells of this precursor give rise to T, B, and dendritic cells. A more differentiated stage of this multipotent precursor in the thymus has lost the capacity to generate B but not T, dendritic, and myeloid cells. Thus, the newly identified precursor maps to the branching point of the T versus B lineage decision in the hematopoietic lineage hierarchy.

T lymphocytes that develop in the thymus are derived from a pool of self-renewing, multipotent hematopoietic stem cells (HSCs) that lodge in the bone marrow (1). T cell development in the thymus is replenished continuously by hematopoietic precursors that travel from the bone marrow via blood to the thymus because the thymus does not support precursors with the capacity for self-renewal (2). The nature of these precursors is still controversial. Many hematopoietic precursors in the bone marrow with distinct self-renewal capacities and differentiation potentials generate T cells upon adoptive transfer of irradiated hosts (e.g., HSCs [3], early lymphoid progenitors [4], LinSca-1+c-kit+ (LSK)flt3+s [5] and common lymphoid progenitors [CLPs; reference 6]), but none of these has been demonstrated to lodge in the thymus. Therefore, it remains to be determined at which level of the hematopoietic lineage hierarchy thymopoiesis branches off. Even the question if thymic precursors are multipotent, oligopotent, or lineage-restricted, and if they commit to the T cell lineage in the bone marrow or in the thymus remains controversial because adult thymic precursors only have been studied on the population level. There is good evidence that hematopoietic precursors in the thymus produce T, B (7, 8), and dendritic cells (9). Whether these cells derive from a single, oligopotent progenitor or from distinct, precommitted precursor cells is unresolved, although the existence of a T/B precursor was suggested by the predominant generation of B cells by Notch1-deficient precursor cells (10, 11). Resolution of these questions necessitates the identification of the most immature hematopoietic precursor in the thymus, and the determination of its clonal differentiation potential.

In 1991, Wu et al. (12) identified the “CD4low precursor” among adult thymocytes which was characterized further as a LinCD25CD44hi c-kithi cell by others (13). With notable exceptions (14, 15), this population is still viewed as the most immature stage of T cell development which among coreceptor CD4 and CD8 double-negative (DN) thymocytes follows the sequence DN1 (c-kit+CD44+CD25) to DN2 (c-kit+CD44+CD25+) to DN3 (c-kitCD44CD25+) to DN4 (c-kitCD44CD25; references 1619). Recently, Allman et al. showed that IL-7R-expressing cells among CD4low precursors do not contain T lineage potential and termed the remaining DN1 Linc-kithiIL-7Rneg/lo cells “early T lineage progenitors” (ETPs; reference 20). ETPs constitute 87% of CD4low precursors and are functionally indistinguishable because they contain mainly T lineage precursors and a few B and myeloid progenitors (8, 20, 21). The most immature hematopoietic precursors in the thymus are believed to be contained in the ETP population, because the only population with potent T lineage potential in the blood of adult mice carries the Lin Sca-1+ c-kit+ (LSK) surface markers that also are found on ETPs (22), but not other presumptive precursors (14, 15). Because 10,000 ETPs can be found in the thymus of an adult mouse, the ETP population is far too numerous to consist homogeneously of thymic precursors (23); the niche that contains thymus repopulating cells is believed to contain only a few hundred cells (24, 25). The large number of these cells per thymus, and the fact that ETPs in the blood cannot be separated from HSCs and other multipotent precursors that are not found in the thymus by conventional surface markers (22) has hampered the investigation of lineage relationships of hematopoietic cells in the thymus. Thus, markers that distinguish functional populations within the ETP population are called for. A recent report confirms that the only DN1 subsets that are less mature than DN2 thymocytes carry the ETP phenotype and distinguishes a DN1a and a DN1b subset by CD24 expression (26). In contrast to ETPs, the DN1a and DN1b subsets lack B potential; this suggests the independent immigration of B precursors and the most immature T lineage progenitors within separate precursor populations. Thus, the identity and the functional properties of the most immature precursors within the large ETP population are unclear.

To investigate whether mature T, B, and dendritic cells derive from a single, oligopotent progenitor or from distinct, precommitted precursor cells, we enhanced GFP (EGFP)-tagged T lineage cells by their expression of the CC chemokine receptor 9 (CCR9) that is expressed exclusively at sites of T cell development (2729). By following EGFPCCR9 expression in heterozygous CCR9-EGFP knock-in mice in which adult αβ-T cell development is indistinguishable from wild-type thymopoiesis, we now identify a thymic precursor that gives rise to T cells, B cells, and dendritic cells on the single cell level. Furthermore, we supply evidence that this progenitor maps to the branching point of the T versus B lineage decision in the hematopoietic lineage hierarchy.

Results

Regulation of EGFPCCR9 expression during the early steps of T cell development

To reinvestigate the issue of the most immature precursor in the thymus, we sought to develop a novel marker for thymic precursors. Therefore, we tagged T lineage cells by their expression of CCR9 that is expressed exclusively at sites of T cell development (2729). To this end, we generated mice in which the NH2-terminal half of the CCR9 coding region is replaced in frame by an EGFP cassette (30). Expression of the tag in the correct cell types was confirmed by FACS analysis of EGFPCCR9 in heterozygous CCR9-EGFP knock-in mice, in which adult αβ-T cell development is indistinguishable from wild-type thymopoiesis (Fig. S1). Cytoplasmic EGFP and membrane integral CCR9 can be expected to show widely divergent half-lives. Therefore, we considered EGFPCCR9 expression as a T cell precursor identifying tag, just like a surface marker of unknown function. We did not take its presence as evidence for the concomitant presence of CCR9 protein or as evidence for the involvement of CCR9 in thymus homing.

The analysis of heterozygous CCR9-EGFP knock-in embryos revealed that significant amounts of tagged cells were found only in the developing thymic anlage and the fetal liver, the site of embryonic hematopoiesis (Fig. 1 A). At embryonic day 11.5 (E11.5; the time when the first hematopoietic precursors arrive in the thymic anlage), all hematopoietic cells in the developing thymic anlage contained high levels of EGFPCCR9 (Fig. 1 B). The first detectable immigrants contained higher amounts of EGFPCCR9 than all subsequent stages of development up to the occurrence of CD44-negative DN3/4 thymocytes at E15.5. Already at E12.5, the population expressing EGFPCCR9 at the level of E11.5 precursors represents only a small fraction of all EGFPCCR9+ cells. EGFPCCR9 expression in heterozygous CCR9-EGFP knock-in mice also seems to be regulated tightly during the early steps of adult thymopoiesis. As observed during embryonic thymopoiesis, EGFPCCR9 increases in adult, immature thymocytes that progress from the CD44-positive DN2 to the CD44-negative DN3 stage (Fig. 1 C). Adult DN1 thymocytes show all levels of EGFPCCR9 expression, from low to high, which indicates phenotypic heterogeneity. EGFPCCR9 expression among the DN stages is present throughout at least at low levels.

Identification of EGFPCCR9-expressing thymic progenitors in bone marrow, blood, and thymus

Based on the embryonic analyses, we hypothesized that EGFPCCR9 could be used in adult mice as a marker for the most immature precursors in the thymus. Furthermore, we postulated that EGFPCCR9 might identify T lineage biased precursors in the peripheral blood. This was crucial since the LSK population has recently been shown to represent the only population with potent T lineage potential in the blood of adult mice (22) but conventional markers could not separate thymic precursors from cells that are absent from the thymus like self-renewing HSCs within this population. Therefore, we investigated EGFPCCR9 expressing, thymus repopulating cells in adult bone marrow, blood and thymus. In the bone marrow, we identified a LinCD25CD117+ EGFPCCR9+ population that contained all the in vivo thymus repopulating activity among EGFPCCR9+ bone marrow cells (Fig. S2). The LinCD25CD117+ EGFPCCR9+ population represents a subset of LSK and CLP precursors (Fig. S3). Apart from this population, EGFPCCR9+ bone marrow populations were found that lacked detectable T lineage potential, which indicated that EGFPCCR9 expression in the bone marrow is not restricted to thymic precursors. Comparing LinCD25CD117+EGFPCCR9+ cells in bone marrow, blood, and thymus revealed that only the CD127 (IL-7Rα)-negative to low fraction among bone marrow precursors appeared in blood and thymus (Fig. 2 A), which is consistent with the reported absence of CLPs in blood (22) and thymus (20). All other tested markers suggested phenotypic homogeneity among precursors in the different compartments. Five-color FACS analyses revealed that the CD127(IL-7Rα)-negative to low fraction of bone marrow LinCD25CD117+EGFPCCR9+ cells expresses high levels of CD117(c-kit; Fig. 2 B); this is consistent with the phenotype of the same cells in the thymus which homogeneously express high levels of CD117 (Fig. 2 A, top right). Thus, we find evidence for the presence of a LinCD25CD117hi EGFPCCR9+CD127−/lowSca-1hiCD4lowCD90.2−/lowCD44+ precursor population in heterozygous CCR9-EGFP knock-in mice that is found in bone marrow, blood, and thymus.

To find support for the idea that EGFPCCR9-tagged blood LSKs are bound for the thymus, we investigated this population in more detail. We detected only 30–60 of these precursors in the blood of one mouse—and consistent with their precursor phenotype—these cells gave rise to thymopoiesis in vivo (Fig. 2 C). Functionally, the culture of blood precursors on OP9-DL1 stroma cells revealed that LinCD25 CD117+EGFPCCR9+ blood cells, but not their EGFPCCR9− counterparts, contained T lineage precursors that gave rise to CD90.2+EGFPCCR9+ DN3/4 stage thymocytes (compare with Fig. 1 C and Fig. S4) within 12 d (Fig. 2 D). DN3/4 stage thymocytes appeared in cultures of LinCD25 CD117+EGFPCCR9− blood precursors only after 16 d of culture (unpublished data). Although both subsets contain T lineage potential, all rapidly differentiating precursors are contained in the EGFPCCR9+ population. This indicated that EGFPCCR9+ blood LSKs are biased toward T cell development relative to their EGFPCCR9− counterparts.

Analysis of the kinetics of T cell development of EGFPCCR9-expressing thymic progenitors

Our results suggest that the LinCD25CD117+EGFPCCR9+ population represents a thymus repopulating cell that travels to the thymus via the blood. If this were true, one would expect that the thymic counterpart of this population is the most immature precursor in the thymus and that cells isolated from these compartments gave rise to more mature stages of T cell development with similar kinetics. Based on our embryonic analyses we further hypothesized that among thymic LinCD25CD117hi EGFPCCR9+ cells, which correspond to the ETP population, the subset expressing the highest levels of EGFPCCR9 represented the recent immigrants while cells with lower levels of EGFPCCR9 were more mature. Indeed, thymic LinCD25CD117hi EGFPCCR9hi precursors representing ≤20% of ETPs developed in fetal thymic organ culture (FTOC) with kinetics that closely resemble those of circulating precursors from the blood (Fig. 3). In contrast, the wave of developing thymocytes that is generated by the thymic LinCD25CD117hiEGFPCCR9low population had progressed significantly beyond the DN2 stage after 8 d, which indicated a more differentiated state of this subset. Thus, the EGFPCCR9 tag allowed us to identify subsets within the ETP population that showed distinct levels of maturity. Specifically, the data demonstrate that adult thymic LinCD25CD117hiEGFPCCR9hi precursors are less mature than their EGFPlow counterparts, and that phenotypically identical precursors that show a similar state of immaturity exist in the blood.

Analysis of the differentiation potential of EGFPCCR9-expressing thymic precursors

To address the question of lineage relationship between hematopoietic cells in the thymus, we investigated the growth requirements and the differentiation potential of thymic EGFPCCR9-expressing precursors within the ETP population. Although CLPs and bone marrow precursors contained in the LSK compartment proliferated vigorously in the presence of stem cell factor (SCF), Flt3L, and IL-7 under serum-free conditions as previously reported (5, 6), numbers of thymic LinCD25CD117hiEGFPCCR9hi precursors declined rapidly over a period of 5 d of culture under these conditions (Fig. 4 A). This observation indicated that these progenitors are functionally distinct from other short-term repopulating cells that are found in the bone marrow LSK population (5). LinCD25CD117hiEGFPCCR9hi precursors did proliferate in vitro when they were sorted in pools of 2,000 cells onto layers of the bone marrow stroma cell line OP9 in the presence of SCF, Flt3L, IL-7, and IL-2 (Fig. 4 B). Under these conditions they gave rise to NK and B cells. This is in contrast to the differentiation potential of the recently described ETP subsets, DN1a and DN1b, which do not produce B cells (26) but is consistent with the reported B cell potential when unseparated ETPs were assayed (20). LinCD25CD117hi EGFPCCR9hi precursor–derived B cells consistently required 12 d to develop. In rare cases, pools of 2,000 thymic LinCD25CD117hiEGFPCCR9low cells also produced B cells; however, these cells were detected considerably earlier than in cultures of the EGFPCCR9hi counterpart, supplying further evidence for the more differentiated state of thymic LinCD25CD117hiEGFPCCR9low cells (Fig. 4 B). Limiting dilution assays revealed that only the LinCD25CD117hi EGFPCCR9hi population contained significant capacity to generate B lymphocytes. Although 1 in 32 of these cells gave rise to a colony of B cells on cytokine-supplemented OP9 layers, we did not find a single B cell colony in the cultures of 4,200 thymic LinCD25CD117hiEGFPCCR9low cells (Fig. 4 C). ETPs that contain both of these precursors are believed to consist mainly of T lineage precursors with a few B progenitors. Therefore, our observation could be explained by the presence of a few committed B cell precursors within the LinCD25CD117hiEGFPCCR9hi population. However, RT-PCR analyses of sorted LinCD25CD117hi EGFPCCR9hi cells did not detect transcripts indicative of B cell commitment, such as Pax5 or λ5 (Fig. 4 D). These findings demonstrate that only the most immature precursors in the thymus possess B cell potential, and that committed B precursors are undetectable within this population.

To our surprise, both precursors gave rise to myeloid cells. When we cultured LinCD25CD117hiEGFPCCR9hi and EGFPCCR9low cells as pools of 1,000 cells on the bone marrow stromal cell line ST2 in the presence of SCF, Flt3L, IL-7, and IL-2 we observed the growth of CD11b-positive cells for both types of precursors in nine out of nine experiments (Fig. 4 B). Cytokine-supplemented ST2 cultures from both precursors were indistinguishable in that they contained “myeloid” and “lymphoid” dendritic cells as well as immature myeloid cells, and a few mature granulocytes (Fig. 4, E and F). We then investigated the possibility that thymic precursors induced detectable granulopoiesis in the thymus but, consistent with a previous report (20), could not find significant numbers of cells with the surface markers that are characteristic for common myeloid progenitors, megakaryocyte/erythrocyte lineage-restricted progenitors, and granulocyte/macrophage lineage-restricted progenitors (31) that have been shown to give rise to all myeloid lineages (unpublished data). Therefore, we conclude that although both precursors possess granulocytic differentiation potential in vitro, it is suppressed in the thymic microenvironment in vivo. Collectively, we find that apart from T cells, thymic precursors give rise to B cells, NK cells, dendritic cells, and myeloid cells; this is consistent with most reports (8, 9, 13, 20, 21, 26, 32) but is at odds with another (33). Significantly, the presented data show that only the capacity to generate B cell development ultimately is restricted to the most immature precursor in the thymus, whereas all other lineages still can be produced effectively by a more differentiated precursor. Because the thymic LinCD25CD117hiEGFPCCR9hi progenitor gives rise to mature cells of multiple hematopoietic lineages, we termed it “thymic multipotent precursor” (TMP).

Clonal analysis of thymic multipotent progenitors

Based on the finding that immaturity among thymic precursors correlated with the most diverse differentiation capacity, we investigated the possibility that all hematopoietic lineages that develop in the thymus (namely T cells, B cells, and dendritic cells) are derived from a single precursor. The potential to give rise to granulocytic cells was not investigated further because there is no evidence for ongoing granulopoiesis in the thymus. Alternatively, immigration of committed, lineage-restricted precursors could be the prevailing mode of supply for adult thymopoiesis which would be consistent with the common assumption that immature thymic precursors contain mainly T lineage precursors and a few B and myeloid progenitors (i.e., we investigated the question if thymic precursors commit to the T cell lineage within or outside the thymus). To address this question we cultured thymic precursors on 20:1 mixtures of the OP9 bone marrow stroma cell line (34) and its derivative, the OP9-DL1 cell line (35). This mixture supports the differentiation of precursor cells into T and B cells in one and the same well (Fig. 5 A). TMPs were double-sorted to a purity of >99.9% and placed as single cells onto OP9/OP9-DL1(20:1) layers in a 96-well format. We consistently found T and B cells in 1 to 3 wells out of 95 wells to which a single TMP had been added (Fig. 5 B and Table I, A1–A3). This experiment demonstrates that the TMP population contains bipotent T/B precursors and that thymic precursors exist that commit to the T cell lineage only after thymus entry. Despite the long period of physiologically unfavorable conditions during cell isolation and double-sorting, ∼50% of the double-sorted TMPs gave rise to T cell colonies on OP9-DL1 stromal layers (Table I, B1). Strikingly, single TMPs sorted on OP9/OP9-DL1(20:1) stromal layers frequently gave rise to T cells only, whereas wells that contained only B cells were seen rarely (Table I, A1–A3). This constellation makes the possibility that the T and B cells found in one and the same well originated from two independent, lineage-restricted precursor cells highly unlikely. Single cells giving rise to T and B cells also were found among the few LinCD25CD117+EGFPCCR9+ cells in the blood, which further supported the notion of a bipotent T/B precursor colonizing the thymus (Table I, C1 and C2). Again, the low frequency of wells containing only T or only B cells ruled out the possibility that wells containing B and T cells are the result of the erroneous addition of two independent precursors to the same well. To investigate the question if the bipotent T/B progenitor found among TMPs also produced dendritic cells, we split OP9/OP9-DL1(20:1) cultures derived from single TMPs onto methylcellulose containing IL-1, IL-3, IL-6, SCF, and Flt3L in the absence of serum. After 4 to 5 d, these cultures contained small but significant numbers of dendritic cells (Fig. 5 B). This observation is consistent with our previous finding that dendritic cells can be derived in vitro from more mature stages of T cell development (Fig. 4 B), and demonstrates that a single hematopoietic precursor—which can be found among TMPs in heterozygous CCR9-EGFP knock-in mice—gives rise to all three hematopoietic lineages that develop in the thymus.

Finally, we investigated LinCD25CD117hi thymocytes of wild-type mice for bipotent T/B progenitors as the defining feature of TMPs. In a representative experiment we found two progenitors that gave rise to B and T cells among 475 singly sorted cells using the OP9/OP9-DL1(20:1) culture system (Table I, D1). Thus, bipotent thymic progenitors identified in heterozygous CCR9-EGFP knock-in mice also can be found in wild-type mice.

Discussion

The presented data demonstrate for the first time that single hematopoietic precursor cells can be isolated from the adult thymus that give rise to T, B, and dendritic cells (i.e., the three hematopoietic lineages known to develop in the thymus; references 79, 36). The fact that all three lineages can be derived from a single precursor indicates that thymic precursors exist that enter the thymus as multipotent progenitors and commit to the T cell lineage only within the thymic microenvironment. That adult thymic precursors have the capacity to give rise to hematopoietic cells of multiple lineages has long been known (8, 9, 13, 20, 21, 26, 32). These reports suggested that multipotent progenitors enter the thymus, but it had not been demonstrated that the relevant lineages in the thymus all can be derived from a single cell. Our observation that multipotent precursors are found exclusively among the most immature precursors in the thymus and in the peripheral blood supports the idea of a multipotent precursor that repopulates the thymus. These findings change the way we think of precursors in the thymus. Previously, it had been assumed that thymic precursors consist of mainly T lineage cells together with a few B and myeloid progenitors. In line with this notion, a recent report suggested the independent immigration of B precursors and the most immature T lineage progenitors within separate precursor populations (26). In contrast, our data now suggest that thymic precursors in the heterogeneous ETP population (23, 26) represent distinct levels of T cell commitment. The most immature stages are multipotent precursors with the in vitro capacity to give rise to T, B, NK, myeloid/granulocytic, and dendritic cells, and the most differentiated cells are T lineage committed—similar to cells that can be found in the DN2/3 stages.

The isolation of multipotent precursors in the thymus required the identification of the most immature precursor in the thymus within the ETP population. We identify thymic LinCD25CD117hiEGFPCCR9hi progenitors, which we term TMPs, as an ETP subset that shows nearly identical developmental kinetics as thymic precursors in the blood, and leaves little space for thymic precursors that might be even more immature. TMPs represent ≤20% of ETPs and are functionally distinct from the remaining ETPs outside of the TMP gate. Only TMPs show nearly identical developmental kinetics as their counterparts in the blood and contain significant B lineage potential. Given the fact that the CD4low precursor and the ETP population are vastly overlapping populations that are functionally indistinguishable from each other, the identification of the TMP represents a major advance in our understanding of early T cell development. We find ∼1,000 TMPs per thymus of a 4–5-wk-old mouse which is consistent with the notion that TMPs represent a small subset of the ETPs. The LinCD25CD117hi EGFPCCR9low precursor, which represents a closely related, more differentiated version of the TMP, also is contained in the ETP population. It has lost its capacity to produce B cells which indicates that the TMP maps to the branching point of the T versus B lineage decision in the hematopoietic lineage hierarchy. This more differentiated precursor still produces granulocytic cells in vitro, although granulopoiesis is not observed in vivo. This finding suggests that the T versus B cell lineage decision precedes the loss of myeloid potential in early thymic precursors.

A single thymic precursor can give rise to more than 105 thymocytes in FTOCs (37), conditions which are presumably far from optimal when compared with the in vivo situation. Thus, less than 1,000 precursors theoretically should be enough to generate the 1–2 × 108 thymocytes found in young adult mice. There is experimental evidence that the thymic niche that holds the repopulating precursors contains ∼200 cells that remained phenotypically uncharacterized (24, 25). This is close to the 1,000 TMPs that we find per thymus of a 4–5-wk-old mouse. The low number of thymic precursor cells that is required to maintain thymopoiesis may explain why the search for the thymus-repopulating precursor has proven to be the search for the proverbial needle in the haystack. Our work introduces the heterozygous CCR9-EGFP knock-in mouse as a tool to study the earliest steps of thymopoiesis. We (30) and others (38, 39) have studied adult αβ-T cell development in homozygous and heterozygous CCR9-deficient mice extensively. No evidence for a nonredundant role for CCR9 in the immigration of hematopoietic precursors into the thymus was found. In this study we used the CCR9-EGFP knock-in allele like a surface marker of unknown function to identify distinct thymus-repopulating precursors without implying any role for CCR9 in thymus homing. Only two abnormalities have been described in CCR9-deficient mice which affect adult αβ-T cell development; neither are found in the heterozygous EGFP-CCR9 knock-in mice that were used in this study. First, CCR9-deficient bone marrow repopulates the thymus inefficiently in competitive transfer experiments (39). Our own competitive transfer experiments confirm this finding, but show no significant difference for heterozygous mice (Table S1). Second, immature thymocytes do not home to the subcapsular microenvironment in CCR9-deficient thymi but do so in heterozygous CCR9-EGFP knock-in and wild-type mice (30). Furthermore, the finding that bipotent precursors give rise to T and B cell development, a capacity that characterizes the TMP in heterozygous CCR9-EGFP knock-in mice, is also found among wild-type LinCD25CD117hi thymocytes further supports the validity of the CCR9-EGFP knock-in model.

Although the characterization of thymic precursors in bone marrow and blood helped us to identify immature precursors in the thymus and to investigate their differentiation potential, many questions concerning thymus repopulation remain. Our experiments do not rule out the possibility that TMP-independent precursors exist in the thymus because we followed cells that expressed the EGFPCCR9 tag exclusively. Thus, precursors from distinct levels of the hematopoietic hierarchy may enter the thymus independently of TMPs, and these may be precommitted to one or the other lineage.

In summary, our work identifies a rare thymic precursor that gives rise to all hematopoietic lineages found in the thymus. It maps to the branching point of the T versus B lineage decision—a key developmental position in hematopoietic lineage maps. The study of this cell will add significantly to our understanding of hematopoiesis because this is the cell type that makes the T versus B cell fate decision in the context of the thymic microenvironment under physiologic conditions.

Materials And Methods

Mice.

BALB/c, C57BL/6, C57BL/6-Ly5.1, RAG2-deficient, CCR9-EGFP knock-in, and RAG2-deficient CCR9-EGFP knock-in mice were kept under specific pathogen-free conditions in the mouse facility of the Max-Planck-Institute for Immunobiology. Embryos were obtained from timed pregnant mice and the day of the vaginal plug was counted as day 0.5 of pregnancy. 3–6-wk-old mice were used for experiments unless otherwise indicated. Because the highest numbers of TMPs were recovered from 4–5-wk-old mice, such mice were used for sorting TMPs from bone marrow, blood, and thymus. Experimentation and animal care was in accordance with the guidelines of the Max-Planck-Institute for Immunobiology.

Flow cytometric analysis.

Cells were prepared and stained as described previously (40). Peripheral blood was obtained from the axilla of anesthetized mice. Bone marrow and blood cells were incubated with 1 μg mAb 2.4G2 per 106 cells (provided by M. Lamers, Max-Planck-Institute for Immunobiology, Freiburg, Germany) for 10 min on ice before staining. Dendritic cells and immature myeloid cells derived from ST2 cultures were incubated with a 1:5 dilution of normal rat serum and 1 μg 2.4G2 for 10 min on ice before staining. The following PE, PE-Cy5, PE-Cy7, APC, AlexaFluor647, or biotin-conjugated monoclonal antibodies were used (clone names given in parentheses) from BD Biosciences: anti-CD3ε (145–2C11), anti-CD4 (GK1.5), anti-CD8α (53–6.7), anti-CD8β (H35-17.2), anti-CD11b (M1/70), anti-CD11c (HL3), anti-CD25 (PC61), anti-CD44 (IM7), anti-CD45 (30-F11), anti-CD90.2 (53–2.1), anti-TCRβ; (H57–597), anti-TCRγδ (GL3), anti-NK1.1 (PK136), anti-Ter119 (Ly-76), and anti-Gr1 (RB6-8C5); from eBioscience: anti-CD117 (2B8), anti-CD127 (A7R34), anti-Sca-1 (D7); and from Caltag: anti-IgM (polyclonal). The biotin label was visualized using SA-PE-Cy5 (Invitrogen) or SA-APC-Cy7 (eBioscience) conjugates. Stainings were analyzed on a FACSCalibur or an LSRII machine (both obtained from BD Biosciences).

Cell sorting and RT-PCR analysis.

To deplete single cell suspensions of lineage marker–positive cells before cell sorting, thymocytes and bone marrow cells were incubated with unlabeled anti-CD8 (169.4.2, provided by M. Lamers) and with unlabeled mAbs directed against Ter119, B220 and Gr-1 (all BD Biosciences), respectively. Lineage marker–positive cells were depleted using goat anti–rat IgG-conjugated paramagnetic beads and MACS separation CS columns according to the manufacturer's recommendations (Miltenyi Biotec). The purified cells were stained and sorted on a high-speed FACS sorter (MoFlo; DakoCytomation) to a purity of >98%. For single cell isolation, sorted cells were sorted a second time. The secondary sort was done in single cell sort mode into 96-well plates at a flow rate of 5–20 cells/s. Purity of double-sorted cells was determined for each experiment and consistently was >99.9%. RT-PCR on RNA isolated from 10,000 sorted cells of each subset was done as previously reported (41) using primers described earlier (42, 43). All mock RT-PCR samples remained negative.

Adoptive transfer experiments.

Thymi of injected mice were analyzed 22 d after i.v. injection by FACS analysis. Intrathymic transfers of 400 cells sorted from the blood were done together with 15,000 RAG2-deficient carrier cells in a volume of 10 μl into anesthetized C57BL/6 mice that had been sublethally irradiated with 450 rad 24 h earlier. Thymi of intrathymically transferred mice were analyzed after 19 d.

Fetal thymic organ culture.

FTOCs were set up as described previously (13). Because of the low numbers of LinCD25CD117+EGFPCCR9+ cells that can be isolated from the peripheral blood, these cells were cultured in FTOCs together with 5,000 RAG2-deficient thymocytes. The rare cases in which either Balb/c (from the embryonic thymus) or RAG2-deficient carrier cells repopulated an FTOC could be identified by the EGFPCCR9 expression found in DN2- and DN3-stage thymocytes derived from EGFP-CCR9 knock-in mice. DN2 thymocytes in these mice show low-level EGFPCCR9 expression (Fig. 1 C).

Cytospin and hematologic staining.

Cells were spun onto glass slides and stained according to Pappenheim. In brief, cells were fixed in May-Grünwald solution (Sigma-Aldrich) for 2 min at RT, washed first with distilled water, and then tap water before 15 min of staining with a 1:20 dilution of Giemsa solution (Sigma-Aldrich).

In vitro cultures of thymic progenitor cells.

For the culture of thymic progenitor cells layers of OP9 (provided by M. Kondo, Duke University, Durham, NC) and OP9-DL1 (provided by J.C. Zuniga-Pflücker, University of Toronto, Toronto, Ontario, Canada) stromal cells and mixtures of OP9 and OP9-DL1 cells were plated at 2 × 103 cells per 96-well plate 48 h before progenitors were sorted. Irradiated (3000 rad) ST2 bone marrow stromal cells were seeded at 104 cells per 96-well plate. 24 h before cell sorting, the cell culture medium was replaced by RPMI 1640 containing 10% FCS, 2 mM L-glutamin, 100 U/ml penicillin, 100 μg/ml streptomycin, 1 mM sodium pyruvate, and 2 × 10−5 M β-mercaptoethanol. Thymic progenitors were sorted onto the different stromal cell layers, and progenitors were cultured in the presence of recombinant, murine IL-7, Flt3L, SCF (10 ng/ml each; R&D Systems), and 50 ng/ml recombinant, human IL-2 (Peprotech) at 37°C in a humidified chamber and 5% CO2. At the indicated time points, cultures were analyzed by FACS. For ST2 cultures, adherent hematopoietic cells were detached by incubation with PBS/0.3% BSA/5 mM EDTA for 10 min to generate single cell suspensions. LSKs (LinCD127Sca-1hiCD117hi), CLPs (LinCD127+Sca-1lowCD117low), and TMPs were cultured on methylcellulose containing IL-7 (10 ng/ml; Methocult M3630, StemCell Technologies Inc.) supplemented with recombinant SCF (100 ng/ml) and Flt3L (20 ng/ml; both obtained from R&D Systems). Differentiation of thymic progenitors toward dendritic cells was achieved by splitting OP9/OP9-DL1 (20:1) cultures at day 10 onto methylcellulose (Methocult M3231, StemCell Technologies Inc.) supplemented with recombinant, murine IL-1β (5 ng/ml), IL-3 (50 ng/ml), IL-6 (10 ng/ml), SCF (30 ng/ml), and Flt3L (30 ng/ml; all R&D Systems). Cells were analyzed by FACS after 4 d in culture.

Online supplemental material.

Fig. S1 identifies EGFP-expressing cells in heterozygous EGFP-CCR9 knock-in mice. Fig. S2 characterizes the in vivo thymus-repopulating capacity of EGFPCCR9+ populations in the bone marrow. Fig. S3 shows the fraction of EGFPCCR9-expressing cells among CLPs and LSKs. Fig. S4 demonstrates that NK1.1CD90.2+EGFPCCR9+ found in OP9-DL1 cultures represent DN3/4 stage thymocytes. Table S1 shows the results of competitive transfer experiments of wild-type versus heterozygous EGFP-CCR9 knock-in bone marrow.

Acknowledgments

We are indebted to S. Gross and C. Sainz-Rueda for excellent technical support. We thank T. Schlake for helpful discussions, B. Kanzler for blastocyst injections, A. Wuerch for FACS sorting, and T. Boehm for his support and helpful discussions on the text.

This work was supported by the Deutsche Forschungsgemeinschaft as part of the SFB620 (A7).

The authors have no conflicting financial interests.

References

1
Kondo, M., A.J. Wagers, M.G. Manz, S.S. Prohaska, D.C. Scherer, G.F. Beilhack, J.A. Shizuru, and I.L. Weissman.
2003
. Biology of hematopoietic stem cells and progenitors: implications for clinical application.
Annu. Rev. Immunol.
21
:
759
–806.
2
Frey, J.R., B. Ernst, C.D. Surh, and J. Sprent.
1992
. Thymus-grafted SCID mice show transient thymopoiesis and limited depletion of V beta 11+ T cells.
J. Exp. Med.
175
:
1067
–1071.
3
Kondo, M., D.C. Scherer, A.G. King, M.G. Manz, and I.L. Weissman.
2001
. Lymphocyte development from hematopoietic stem cells.
Curr. Opin. Genet. Dev.
11
:
520
–526.
4
Igarashi, H., S. Gregory, T. Yokota, N. Sakaguchi, and P. Kincade.
2002
. Transcription from the RAG1 locus marks the earliest lymphocyte progenitors in bone marrow.
Immunity.
17
:
117
–130.
5
Adolfsson, J., O.J. Borge, D. Bryder, K. Theilgaard-Monch, I. Astrand-Grundstrom, E. Sitnicka, Y. Sasaki, and S.E. Jacobsen.
2001
. Upregulation of Flt3 expression within the bone marrow Lin(-) Sca1(+)c-kit(+) stem cell compartment is accompanied by loss of self-renewal capacity.
Immunity.
15
:
659
–669.
6
Kondo, M., I.L. Weissman, and K. Akashi.
1997
. Identification of clonogenic common lymphoid progenitors in mouse bone marrow.
Cell.
91
:
661
–672.
7
Akashi, K., L.I. Richie, T. Miyamoto, W.H. Carr, and I.L. Weissman.
2000
. B lymphopoiesis in the thymus.
J. Immunol.
164
:
5221
–5226.
8
Wu, L., M. Antica, G.R. Johnson, R. Scollay, and K. Shortman.
1991
. Developmental potential of the earliest precursor cells from the adult mouse thymus.
J. Exp. Med.
174
:
1617
–1627.
9
Ardavin, C., L. Wu, C.L. Li, and K. Shortman.
1993
. Thymic dendritic cells and T cells develop simultaneously in the thymus from a common precursor population.
Nature.
362
:
761
–763.
10
Radtke, F., A. Wilson, G. Stark, M. Bauer, J. van Meerwijk, H.R. MacDonald, and M. Aguet.
1999
. Deficient T cell fate specification in mice with an induced inactivation of Notch1.
Immunity.
10
:
547
–558.
11
Wilson, A., H.R. MacDonald, and F. Radtke.
2001
. Notch 1-deficient common lymphoid precursors adopt a B cell fate in the thymus.
J. Exp. Med.
194
:
1003
–1012.
12
Wu, L., R. Scollay, M. Egerton, M. Pearse, G.J. Spangrude, and K. Shortman.
1991
. CD4 expressed on earliest T-lineage precursor cells in the adult murine thymus.
Nature.
349
:
71
–74.
13
Moore, T.A., and A. Zlotnik.
1995
. T-cell lineage commitment and cytokine responses of thymic progenitors.
Blood.
86
:
1850
–1860.
14
Gounari, F., I. Aifantis, C. Martin, H.J. Fehling, S. Hoeflinger, P. Leder, H. von Boehmer, and B. Reizis.
2002
. Tracing lymphopoiesis with the aid of a pTalpha-controlled reporter gene.
Nat. Immunol.
3
:
489
–496.
15
Martin, C.H., I. Aifantis, M.L. Scimone, U.H. Von Andrian, B. Reizis, H. Von Boehmer, and F. Gounari.
2003
. Efficient thymic immigration of B220(+) lymphoid-restricted bone marrow cells with T precursor potential.
Nat. Immunol.
4
:
866
–873.
16
Pearse, M., L. Wu, M. Egerton, A. Wilson, K. Shortman, and R. Scollay.
1989
. A murine early thymocyte developmental sequence is marked by transient expression of the interleukin 2 receptor.
Proc. Natl. Acad. Sci. USA.
86
:
1614
–1618.
17
Godfrey, D.I., A. Zlotnik, and T. Suda.
1992
. Phenotypic and functional characterization of c-kit expression during intrathymic T cell development.
J. Immunol.
149
:
2281
–2285.
18
Godfrey, D.I., J. Kennedy, T. Suda, and A. Zlotnik.
1993
. A developmental pathway involving four phenotypically and functionally distinct subsets of CD3−CD4−CD8− triple-negative adult mouse thymocytes defined by CD44 and CD25 expression.
J. Immunol.
150
:
4244
–4252.
19
Godfrey, D.I., J. Kennedy, P. Mombaerts, S. Tonegawa, and A. Zlotnik.
1994
. Onset of TCR-beta gene rearrangement and role of TCR-beta expression during CD3−CD4−CD8− thymocyte differentiation.
J. Immunol.
152
:
4783
–4792.
20
Allman, D., A. Sambandam, S. Kim, J.P. Miller, A. Pagan, D. Well, A. Meraz, and A. Bhandoola.
2003
. Thymopoiesis independent of common lymphoid progenitors.
Nat. Immunol.
4
:
168
–174.
21
King, A.G., M. Kondo, D.C. Scherer, and I.L. Weissman.
2002
. Lineage infidelity in myeloid cells with TCR gene rearrangement: a latent developmental potential of proT cells revealed by ectopic cytokine receptor signaling.
Proc. Natl. Acad. Sci. USA.
99
:
4508
–4513.
22
Schwarz, B.A., and A. Bhandoola.
2004
. Circulating hematopoietic progenitors with T lineage potential.
Nat. Immunol.
5
:
953
–960.
23
Bhandoola, A., A. Sambandam, D. Allman, A. Meraz, and B. Schwarz.
2003
. Early T lineage progenitors: new insights, but old questions remain.
J. Immunol.
171
:
5653
–5658.
24
Goldschneider, I., K.L. Komschlies, and D.L. Greiner.
1986
. Studies of thymocytopoiesis in rats and mice. I. Kinetics of appearance of thymocytes using a direct intrathymic adoptive transfer assay for thymocyte precursors.
J. Exp. Med.
163
:
1
–17.
25
Spangrude, G.J., and R. Scollay.
1990
. Differentiation of hematopoietic stem cells in irradiated mouse thymic lobes. Kinetics and phenotype of progeny.
J. Immunol.
145
:
3661
–3668.
26
Porritt, H.E., L.L. Rumfelt, S. Tabrizifard, T.M. Schmitt, J.C. Zuniga-Pflucker, and H.T. Petrie.
2004
. Heterogeneity among DN1 prothymocytes reveals multiple progenitors with different capacities to generate T cell and non-T cell lineages.
Immunity.
20
:
735
–745.
27
Zaballos, A., J. Gutierrez, R. Varona, C. Ardavin, and 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.
28
Youn, B.S., C.H. Kim, F.O. Smith, and H.E. Broxmeyer.
1999
. TECK, an efficacious chemoattractant for human thymocytes, uses GPR-9- 6/CCR9 as a specific receptor.
Blood.
94
:
2533
–2536.
29
Wurbel, M.A., J.M. Philippe, C. Nguyen, G. Victorero, T. Freeman, P. Wooding, A. Miazek, M.G. Mattei, M. Malissen, B.R. Jordan, et al.
2000
. The chemokine TECK is expressed by thymic and intestinal epithelial cells and attracts double- and single-positive thymocytes expressing the TECK receptor CCR9.
Eur. J. Immunol.
30
:
262
–271.
30
Benz, C., K. Heinzel, and C.C. Bleul.
2004
. Homing of immature thymocytes to the subcapsular microenvironment within the thymus is not an absolute requirement for T cell development.
Eur. J. Immunol.
34
:
3652
–3663.
31
Akashi, K., D. Traver, T. Miyamoto, and I.L. Weissman.
2000
. A clonogenic common myeloid progenitor that gives rise to all myeloid lineages.
Nature.
404
:
193
–197.
32
Saunders, D., K. Lucas, J. Ismaili, L. Wu, E. Maraskovsky, A. Dunn, and K. Shortman.
1996
. Dendritic cell development in culture from thymic precursor cells in the absence of granulocyte/macrophage colony-stimulating factor.
J. Exp. Med.
184
:
2185
–2196.
33
Balciunaite, G., R. Ceredig, and A.G. Rolink.
2005
. The earliest subpopulation of mouse thymocytes contains potent T, significant macrophage and natural killer, but no B lymphocyte potential.
Blood.
105
:
1930
–1936.
34
Nakano, T., H. Kodama, and T. Honjo.
1994
. Generation of lymphohematopoietic cells from embryonic stem cells in culture.
Science.
265
:
1098
–1101.
35
Schmitt, T.M., and J.C. Zuniga-Pflucker.
2002
. Induction of T cell development from hematopoietic progenitor cells by delta-like-1 in vitro.
Immunity.
17
:
749
–756.
36
Wu, L., D. Vremec, C. Ardavin, K. Winkel, G. Suss, H. Georgiou, E. Maraskovsky, W. Cook, and K. Shortman.
1995
. Mouse thymus dendritic cells: kinetics of development and changes in surface markers during maturation.
Eur. J. Immunol.
25
:
418
–425.
37
Kingston, R., E.J. Jenkinson, and J.J. Owen.
1985
. A single stem cell can recolonize an embryonic thymus, producing phenotypically distinct T-cell populations.
Nature.
317
:
811
–813.
38
Wurbel, M.A., M. Malissen, D. Guy-Grand, E. Meffre, M.C. Nussenzweig, M. Richelme, A. Carrier, and B. Malissen.
2001
. Mice lacking the CCR9 CC-chemokine receptor show a mild impairment of early T- and B-cell development and a reduction in T-cell receptor gammadelta(+) gut intraepithelial lymphocytes.
Blood.
98
:
2626
–2632.
39
Uehara, S., A. Grinberg, J.M. Farber, and P.E. Love.
2002
. A role for CCR9 in T lymphocyte development and migration.
J. Immunol.
168
:
2811
–2819.
40
Boehm, T., S. Scheu, K. Pfeffer, and C.C. Bleul.
2003
. Thymic medullary epithelial cell differentiation, thymocyte emigration, and the control of autoimmunity require lympho-epithelial cross talk via LTbetaR.
J. Exp. Med.
198
:
757
–769.
41
Bleul, C.C., and T. Boehm.
2000
. Chemokines define distinct microenvironments in the developing thymus.
Eur. J. Immunol.
30
:
3371
–3379.
42
Li, Y.S., R. Wasserman, K. Hayakawa, and R.R. Hardy.
1996
. Identification of the earliest B lineage stage in mouse bone marrow.
Immunity.
5
:
527
–535.
43
Anderson, M.K., G. Hernandez-Hoyos, R.A. Diamond, and 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.

Abbreviations used: CCR9, CC chemokine receptor 9; CLP, common lymphoid progenitor; DN, double negative; EGFP, enhanced GFP; ETP, early T lineage progenitor; FTOC, fetal thymic organ culture; HSC, hematopoietic stem cell; LSK, LinSca-1+c-kit+; SCF, stem cell factor; TMP, thymic multipotent precursor.

Supplementary data