Lymphotoxin β Receptor Controls T Cell Progenitor Entry to the Thymus

The recruitment of lymphoid progenitors to the thymus is essential to sustain T cell production throughout life. Importantly, it also limits T lineage regeneration following bone marrow transplantation, and so contributes to the secondary immunodeficiency that is caused by delayed immune reconstitution. Despite this significance, the mechanisms that control thymus colonization are poorly understood. In this study, we show that in both the steady-state and after bone marrow transplant, lymphotoxin β receptor (LTβR) controls entry of T cell progenitors to the thymus. We show that this requirement maps to thymic stroma, further underlining the key importance of this TNFR superfamily member in regulation of thymic microenvironments. Importantly, analysis of the requirement for LTβR in relationship to known regulators of thymus seeding suggests that it acts independently of its regulation of thymus-homing chemokines. Rather, we show that LTβR differentially regulates intrathymic expression of adhesion molecules known to play a role in T cell progenitor entry to the thymus. Finally, Ab-mediated in vivo LTβR stimulation following bone marrow transplant enhances initial thymus recovery and boosts donor-derived T cell numbers, which correlates with increased adhesion molecule expression by thymic stroma. Collectively, we reveal a novel link between LTβR and thymic stromal cells in thymus colonization, and highlight its potential as an immunotherapeutic target to boost T cell reconstitution after transplantation.

I n the thymus, immature lymphoid progenitors undergo a complex differentiation program that biases thymocyte development toward the generation of self-tolerant MHCrestricted T cells (1). Importantly, the hemopoietic progenitors that colonize the thymus are generated in extrathymic sites, and so T cell development depends on thymic colonization by migrant progenitors (2,3). As the thymus does not contain hemopoietic stem cells with long-term self-renewal capacity, there is an ongoing requirement for this recruitment process, and this is important for several reasons. First, it creates successive waves of thymopoiesis to maintain long-term T cell production (4,5). Second, it establishes competition for intrathymic niches that limits the selfrenewal of intrathymic progenitors (6)(7)(8). Importantly, absence of competition manifests as T cell acute lymphoblastic leukemia, indicating that thymus seeding is part of an intrathymic tumor suppression mechanism that requires constant replacement of the immature thymocyte pool (9).
Although lymphoid progenitors are known to enter the adult thymus via blood vessels at the corticomedullary junction (10), their rarity means that their exact nature remains unclear (11)(12)(13). However, insight into the mechanisms that control thymus colonization can be obtained by studying the frequency and requirements of CD4 2 CD8 2 CD44 + CD25 2 CD117 + thymocytes that represent the earliest thymic progenitors (ETP) in the adult mouse thymus (13)(14)(15)(16). Thus, thymus entry is recognized as a multistep process involving chemokines, adhesion molecules, and growth factors produced by thymic microenvironments. For example, thymic endothelial cells express VCAM-1, ICAM-1, and P-selectin (17)(18)(19) to enable the attachment of blood-borne lymphoid progenitors. Significantly, Ab blockade of VCAM-1/ICAM-1 impairs lymphoid progenitor entry to the thymus (20), whereas mice deficient in either P-selectin or its receptor PSGL-1 have fewer ETP and an increased availability of intrathymic niches (18). ETP express the chemokine receptors CXCR4, CCR7, and CCR9 (21)(22)(23)(24), and the chemokines CCL19, CCL21, CCL25, and CXCL12 are all products of thymic stroma (21,25,26). Significantly, disruption of these molecules either individually or in combination results in impaired thymus seeding (22,23,27,28). Importantly, however, although these studies emphasize the importance of the thymic microenvironment in the recruitment of lymphoid progenitors to the thymus, this process is still poorly understood and relatively few of its regulators are known.
The importance of thymus seeding is further emphasized by its regulation of immune system recovery that follows ablative therapy and bone marrow (BM) transplant (BMT), where limited thymus entry of donor progenitors slows down T cell reconstitution in comparison with other blood cell lineages (29,30). Indeed, intrathymic progenitor niches are not saturated until at least 10 wk after BMT (29), suggesting that delayed T cell reconstitution is linked to inefficient thymus seeding. Interestingly, although PSGL-1 has been identified as an important regulator of thymus seeding after BMT (29), the cellular and molecular mechanisms that limit T cell recovery after transplant, and how they relate to the requirements of steady-state T cell development, remain poorly understood.
In this study, we show that mice lacking lymphotoxin b receptor (LTbR) demonstrate a dramatic reduction in the frequency of ETP, and that increased compensatory intrathymic progenitor proliferation accounts for their normal thymocyte numbers. Importantly, thymus transplant and BM chimera experiments show the requirement for LTbR maps to thymic stromal cells. We also show that LTbR differentially regulates thymic stromal expression of VCAM-1 and ICAM-1 but not P-selectin, which collectively represent adhesion molecules previously linked to thymus entry. Finally, we show that thymic recovery after BMT also requires LTbR, and that agonistic anti-LTbR treatment enhances donor-derived T cell reconstitution. Collectively, our findings identify a novel regulatory axis of T cell progenitor entry to the thymus, and they extend our understanding of the importance of LTbR in the functional control of thymic stromal microenvironments.

Mice
Adult wild-type (WT) C57BL/6 and congenic CD45.1 + C57BL/6 mice, and Ltbr 2/2 (31) and plt/plt (32) mice on a C57BL/6 background were used at 8-12 wk of age. All mice were housed at the Biomedical Services Unit at the University of Birmingham in accordance with local and national Home Office regulations.

BrdU incorporation
Mice were injected i.p. with 1.5 mg of BrdU (Sigma-Aldrich) and tissue was harvested 3 h later. Staining for BrdU was performed using the BrdU flow kit (BD Biosciences), according to the manufacturer's instructions.

Generation of BM chimeras
Recipient mice were lethally irradiated (23 500 rad) and reconstituted i.v. with 5 3 10 6 T cell-depleted adult BM preparations from CD45 congenically marked mice, as indicated. Depletion of T cells was performed using anti CD3-PE and anti-PE microbeads (Miltenyi Biotec) according to the manufacturer's instructions. Mice were sacrificed at the indicated time points, and tissues were analyzed by flow cytometry. In some experiments, mice received 100 mg of agonistic anti-LTbR (35) or isotype control on days 1, 3, 5, 7, and 9, and in these experiments tissues were harvested for analysis at day 10 or day 28.

Thymus transplantation
Embryonic day 15 thymuses, organ cultured for 5 d in 1.35 mM 2deoxyguanosine, were transplanted under the kidney capsule of WT mice and harvested after 6-8 wk (34).

Statistical analysis
Data were analyzed using an unpaired t test. A p value ,0.05 was considered significant. Statistical analysis was performed using GraphPad Prism. Data are represented as mean 6 SEM.

LTbR controls ETP frequency
Given the importance of the TNFR superfamily (TNFRSF) in the organization and development of functionally competent thymic stromal microenvironments, we screened a panel of mutant mice for evidence of impaired thymus colonization. Although no obvious alterations were found in Tnfsf1 2/2 and Tnfrsf11b 2/2 mice, we found that Tnfrsf3 2/2 mice (described here as Ltbr 2/2 mice) had reduced numbers of downstream early T cell progenitors, including those at the double-negative (DN) 1-3 stages of development (Fig. 1A). Importantly, we also saw a significant reduction in both the proportion and absolute number of lineage 2 CD44 + CD25 2 CD117 + ETP in Ltbr 2/2 mice (Fig. 1B). Analysis of cellular proliferation following BrdU injection showed a similar frequency of BrdU + ETP in WT and Ltbr 2/2 mice (Fig. 1C), arguing against the notion that the ETP reduction in Ltbr 2/2 mice was due to alterations in their ability to proliferate intrathymically. Despite these changes, total thymocyte numbers in WT and Ltbr 2/2 mice were comparable, including numbers of CD4 2 CD8 + immature single-positive and CD4 + CD8 + cells (Fig. 1D), suggesting that increased thymocyte expansion during early stages of T cell development may restore normal thymocyte cellularity. Indeed, analysis of BrdU incorporation showed an increased frequency of BrdU + DN3 thymocytes in Ltbr 2/2 mice compared with WT mice (Fig. 1E). Taken together, these findings show that LTbR controls the frequency of ETP in the adult, suggesting a role in the regulation of lymphoid progenitor entry to the thymus. They also indicate that ETP reduction in Ltbr 2/2 mice is compensated by enhanced DN3 stage thymocyte proliferation that restores thymus cellularity to normal levels.

Stromal cell expression of LTbR controls thymus entry of T cell progenitors
Although LTbR regulates the development and function of thymic stromal microenvironments (25,37), it can also directly influence hemopoietic cells (38). To investigate the LTbR-expressing cellular compartment that regulates progenitor entry to the thymus, we first established a series of reciprocal BM chimeras to confine LTbR expression to either stromal cells or hemopoietic cells. Initially, WT mice at 8 wk of age were lethally irradiated and injected i.v. with congenically marked T cell-depleted BM obtained from either WT or Ltbr 2/2 donor mice. Additionally, Ltbr 2/2 adult recipients were reconstituted with congenic T-depleted BM preparations from either WT or Ltbr 2/2 donors. Mice were harvested after 8 wk, and analysis of thymocyte development from donor-derived progenitors was performed. In chimeras using WT hosts, similar ETP proportions and numbers were generated from both WT and Ltbr 2/2 BM, and no significant alterations in early stage T cell development were observed ( Fig. 2A, 2B). In contrast, a dramatic reduction in ETP frequency and proportion was observed when Ltbr 2/2 hosts were reconstituted with WT donor BM (Fig. 2C, 2D). Alterations in early thymocyte development, including a reduction in DN1-3 T cell progenitor compartments, were also noted in WT → Ltbr 2/2 chimeras (Fig. 2D). Next, to directly assess whether LTbR expression by thymic stroma is involved in progenitor entry to the thymus, we transplanted alymphoid WT or Ltbr 2/2 thymus lobes under the kidney capsule of WT mice. In this setting, LTbR is exclusively absent from stromal cells in the transplanted thymus. Im-portantly, flow cytometric analysis 6-8 wk after grafting showed a significant reduction in the proportion and absolute number of ETP in Ltbr 2/2 grafts compared with WT (Fig. 2E). Collectively, these results indicate that LTbR expression by thymic stroma is important for lymphoid progenitors to enter the thymus, a finding consistent with the cell-intrinsic requirement for LTbR in the development and function of thymic stromal microenvironments (25,37,39).

LTbR differentially controls known regulators of thymus seeding
To investigate how LTbR influences T cell progenitor entry to the thymus, we compared the expression of known regulators of thymus seeding in purified thymic stromal subsets from WT and Ltbr 2/2 mice. Ccl25 and Cxcl12 mRNA levels were not altered in Ltbr 2/2 mice (Fig. 3A), suggesting that the requirement for LTbR is not explained by its regulation of ligand availability for the chemokine receptors CCR9 and CXCR4. Similarly, we saw no substantial alteration in Kitl expression, an important growth factor for immature T cell progenitors (40,41). Interestingly, we also found no differences in Ccl21 mRNA levels in WT and Ltbr 2/2 TEC (Fig. 3A). Importantly, however, note that this observation is not incompatible with an earlier study showing that CCL21 + medullary TEC (mTEC) are present, but at a reduced frequency in Ltbr 2/2 mice (25), and may simply reflect differences in methods (qPCR/flow cytometry) used to measure CCL21 expression. Interestingly, however, we did see decreased expression of Ccl19 mRNA in both TEC and thymic mesenchyme from Ltbr 2/2 mice (Fig. 3A), suggesting that LTbR could play a role in the recruitment of T cell progenitors via regulation of intrathymic CCR7 ligand availability. To investigate this directly, we examined ETP frequency and early T cell development in plt/plt mice that lack expression of CCL19 and CCL21 (32). Interestingly, whereas our analysis of plt/plt mice showed alterations in DN1-4 T cell progenitor frequencies that are consistent with an earlier report (24), we saw no differences in ETP numbers in WT and plt/plt thymuses (Fig. 3B). This agrees with other studies showing that the major impact of CCR7 deficiency on ETP requires the combined absence of CCR9 (22,23,27). Thus, mice lacking either LTbR or CCR7 ligands do not share alterations in ETP frequency, suggesting that control of CCL19 and CCL21 expression by LTbR does not explain its requirement during thymus seeding.
In addition to chemokines, thymus entry requires attachment of lymphoid progenitors to stromal cells expressing the adhesion molecules P-selectin, VCAM-1, and ICAM-1 (17,19,20). We found that Selp mRNA levels were comparable in WT and Ltbr 2/2 thymic endothelium (Fig. 3C). In contrast, levels of VCAM-1 and ICAM-1 were altered. Specifically, both Ltbr 2/2 thymic endothelium and mesenchyme had significantly reduced levels of VCAM-1/ICAM-1 (Fig. 3C). Interestingly, levels of expression on WT and Ltbr 2/2 TEC were comparable (Fig. 3C). Collectively, our findings demonstrate a differential requirement for LTbR in the control of adhesion molecule expression by TEC and non-TEC stroma, and they show that within the latter, adhesion molecules known to influence thymus entry can be subdivided into LTbR-dependent (VCAM-1/ICAM-1) and LTbR-independent (P-selectin) groups.

LTbR mediates thymus recovery after BMT
T cell progenitor recruitment to the thymus influences T cell reconstitution and thymus recovery following ablative therapy and BMT (29). We next examined the possible role of LTbR in this context and focused on early events in thymus reconstitution (42)(43)(44). Importantly, other studies have shown that donor-derived ETP are not detectable in irradiated mice at time points shortly after BMT, with a clearly defined ETP population not apparent until after 3 wk following transplant (45). Thus, to assess the role of LTbR in early phases of thymus recovery, we determined the frequency and number of donor-derived congenically marked thymocytes 13 d after the transplantation of WT BM into lethally irradiated WT and Ltbr 2/2 mice. Importantly, at early stages of thymus recovery, note that the thymus is dominated by thymocytes of host origin that survive and expand after irradiation (Fig. 4A, 4B) (43,44). Furthermore, although 2-3 3 10 7 CD45.1 + donorderived thymocytes were recovered from the thymuses of irradiated WT hosts, a frequency that is in line with other studies (46), we saw a significant 3-to 4-fold reduction in the number of CD45.1 + donor-derived thymocytes recovered from Ltbr 2/2 hosts (Fig. 4A-C). Interestingly, the pattern of development (Fig. 4A-C) and frequency of BrdU + cells (Fig. 4D) in donor-derived thymocytes was comparable in WT and Ltbr 2/2 hosts, indicating that this difference was not due to an inability of donor progenitors to undergo proliferation and differentiation in the irradiated Ltbr 2/2 host thymus. Rather, the reduced frequency of donor-derived thymocytes in the thymus of Ltbr 2/2 hosts suggests that as in the steady-state, progenitor entry to the thymus after BMT involves LTbR.

In vivo agonistic anti-LTbR treatment enhances thymus recovery after BMT
Given our findings on thymus reconstitution in Ltbr 2/2 mice, we next investigated whether exogenous LTbR stimulation may be a potential therapeutic means to boost thymus recovery after BMT. WT mice were lethally irradiated and then reconstituted with congenic CD45.1 + T cell-depleted BM preparations. The day after BMT, mice then received either 100 mg of agonistic anti-LTbR (35) or an isotype control Ab, every other day until day 9 (Fig. 5A). Tissues were harvested from chimeric mice on days 10 and 28 after BMT, and donor-derived thymocytes and T cells were analyzed by flow cytometry. Again, at this early day 10 time point (Fig. 5B), the dominant population of cells in the thymus were host-derived thymocytes that survived irradiation. Importantly, however, although ETP cannot be detected at this early time point (45), we found that a population of donor-derived thymocytes was detectable, representing initial donor engraftment of the host thymus (Fig. 5B). Strikingly, 10 d after transplant, the proportion and absolute number of donor-derived CD45.1 + thymocytes was significantly increased in mice receiving agonistic LTbR compared with isotype control (Fig. 5B, 5C), and cells showed a normal pattern of progression through DN thymocyte stages at this early posttransplant time point (Fig. 5B, 5C). Furthermore, analysis of chimeras 28 d after transplant showed a significant increase in donor-derived peripheral T cell numbers in the spleens of mice receiving agonistic anti-LTbR (Fig. 5D). Interestingly, this effect of anti-LTbR was specific, as residual host-derived splenic T cell numbers were not affected by Ab treatment (Fig. 5D). Thus, our data suggest that agonistic anti-LTbR treatment enhances the recovery of thymopoiesis by increasing the frequency of donor-derived progenitors in the thymus of irradiated recipient mice, and this leads to an increase in donor-derived T cells in the periphery.
Finally, given our data suggesting that LTbR may influence thymus seeding by regulating expression of VCAM-1 and ICAM-1, we analyzed the impact of anti-LTbR treatment on their levels in thymic stroma after BMT. Lethally irradiated mice were reconstituted with T cell-depleted BM and subjected to anti-LTbR/isotype control treatment, and thymuses were harvested after 10 d. Following digestion, CD45 2 EpCAM1 + TEC, CD45 2 CD31 + endothelium, and CD45 2 podoplanin + mesenchymal stromal cells were analyzed by flow cytometry. Interestingly, in mice receiving anti-LTbR treatment, both thymic endothelium and mesenchymal cells showed increased levels of both VCAM-1 and ICAM-1 whereas levels on TEC were not altered (Fig. 5E). Collectively, these observations show that in vivo stimulation of LTbR boosts thymic recovery after BMT, and this correlates with enhanced expression of adhesion molecules known to facilitate thymic entry of T cell progenitors in non-TEC stroma.

Discussion
The absence of an intrathymic hemopoietic stem cell pool means that to sustain T cell production throughout life, the thymus must continuously import lymphoid progenitors from extrathymic sites. Additionally, the entry of donor-derived lymphoid progenitors to the thymus represents a rate-limiting step in re-establishing T cellmediated immunity that follows ablative therapy and BMT. Despite this importance, relatively few regulators of T cell progenitor entry to the thymus are known. Collectively, the work described in the present study shows that the TNFRSF member LTbR plays a key role in the regulation of thymus seeding. We also show LTbR is involved in thymus entry in both the steady-state and during the early phases of thymus recovery that take place after BMT. Regarding the latter, manipulation of the LTbR axis using agonistic Abs significantly improved donor-derived thymopoiesis and boosted T cell recovery after BMT, suggesting that LTbR stimu-lation is part of a common mechanism that controls thymus entry in both the steady-state and during immune reconstitution.
Our finding that the requirement for LTbR maps to thymic stromal cells is significant, as it extends our understanding of its importance in the regulation of thymus function. For example, identification of a role for LTbR in thymic entry complements work demonstrating its importance for the thymic egress of mature thymocytes (37). Interestingly, another study has shown that mature thymocytes and T cell progenitors are present within the same perivascular spaces that surround intrathymic blood vessels (47). Taken together, these findings raise the possibility that the limited entry of T cell progenitors to the Ltbr 2/2 thymus is caused by an accumulation of mature thymocytes at a common site of thymic exit and entry. Further experiments are required to examine the possible relationship between these processes and the mechanisms that regulate them.
Although LTbR has been shown to influence the frequency of CCL21 + mTEC (25), our findings, including normal ETP frequency in plt/plt mice, suggest that its involvement in thymus seeding is not simply explained by its regulation of CCR7 ligands. However, whether the requirement for LTbR shown in this study is linked to the noted decrease in availability of CCL21-expressing mTEC lo cells (25) is currently not clear. Importantly, the positive impact of LTbR stimulation on thymic reconstitution shown in this study is observed shortly after BMT, a time point at which thymus seeding is transiently independent of CCR7 and CCR9 (29), further suggesting that LTbR stimulation augments thymic reconstitution via mechanisms other than chemokine availability. As indicated from studies on peripheral lymphoid tissues (48), another way in which LTbR may influence thymus entry is through its ability to regulate the expression of adhesion molecules by endothelium and/or mesenchyme. This would fit well with our demonstration of reduced levels of VCAM-1/ICAM-1 in steady-state Ltbr 2/2 mice, and their enhanced expression in both thymic endothelium and mesenchyme following in vivo anti-LTbR treatment. It is also supported by previous studies in which Ab-mediated blockade of VCAM-1/ ICAM-1 inhibited thymic entry of transferred lymphoid progenitors (20). Interestingly, unlike VCAM-1 and ICAM-1, we found that LTbR did not control P-selectin expression, a finding that highlights its differential ability to influence expression of adhesion molecules linked to thymus entry. Taken together, these data raise the possibility that LTbR regulates thymic entry through control of integrin-mediated firm adhesion and transendothelial migration, rather than initial phases of selectin-mediated rolling and chemokine-driven activation/migration. Alternatively, the requirement for LTbR may relate to its importance in the regulation of medullary microenvironments. Thus, altered mTEC development and organization in Ltbr 2/2 mice (25, 37) may limit the ability of the thymus to attract migrant lymphoid progenitors. However, it is interesting that whereas thymus medulla disorganization is also evident in plt/plt mice (24), we found that their ETP frequency was not altered.
Finally, the identity of the LTbR ligands and the cells they are expressed by that influence thymus seeding are not known. Relevant to this, earlier studies have shown the difficulty in correlating known roles for LTbR in thymus development and function with the availability of the LTbR ligands lymphotoxin and LIGHT. For example, whether defects in thymus organization caused by absence of LTbR are mirrored in mice lacking LTbR ligands, either individually or in combination, is not certain (37,49). However, as thymic expression of LTbR ligands has been mapped to a variety of hemopoeitic cells (37,50), we suggest that the requirement for LTbR during thymus entry represents a further example of how thymic crosstalk regulates the TNFRSF-mediated control of thymus function. In summary, our study identifies a new FIGURE 4. Initial thymic reconstitution after BMT is controlled by LTbR. Lethally irradiated WT (A) and Ltbr 2/2 (B) mice were reconstituted with T-depleted congenically marked WT BM cells and harvested after 13 d. Thymic reconstitution was determined by calculating the intrathymic frequency of CD45.1 + donor cells, and bar charts in (C) show numbers of total donor thymocytes and percentages of donor-derived double-positive and DN thymocytes. n $ 13 from five independent experiments; representative FACS plots are shown. (D) Analysis of BrdU incorporation in WT donor-derived CD45.1 + thymocytes from WT and Ltbr 2/2 hosts (n = 6 from three independent experiments). ****p , 0.0001. role for LTbR in the control of thymus function, where it acts as a regulator of the earliest phases of T cell development by influencing the intrathymic availability of the earliest thymocyte progenitors. That this role extends to thymic recovery after BMT suggests the potential of LTbR as an immunotherapeutic target to boost T cell recovery and restore essential immune system functioning following ablative therapy.