LTβR controls thymic portal endothelial cells for haematopoietic progenitor cell homing and T-cell regeneration

Abstract

Continuous thymic homing of haematopoietic progenitor cells (HPCs) via the blood is critical for normal T-cell development. However, the nature and the differentiation programme of specialized thymic endothelial cells (ECs) controlling this process remain poorly understood. Here using conditional gene-deficient mice, we find that lymphotoxin beta receptor (LTβR) directly controls thymic ECs to guide HPC homing. Interestingly, T-cell deficiency or conditional ablation of T-cell-engaged LTβR signalling results in a defect in thymic HPC homing, suggesting the feedback regulation of thymic progenitor homing by thymic products. Furthermore, we identify and characterize a special thymic portal EC population with features that guide HPC homing. LTβR is essential for the differentiation and homeostasis of these thymic portal ECs. Finally, we show that LTβR is required for T-cell regeneration on irradiation-induced thymic injury. Together, these results uncover a cellular and molecular pathway that governs thymic EC differentiation for HPC homing.

Figure 1. Endothelial LTβR is required for thymic progenitor cell homing.

(a,b) Flow cytometric analysis of ETPs (Lin⁻CD25⁻CD44⁺c-Kit⁺) in Ltbr^−/− and control mice. (a) Representative dot plots are shown, gated on Lin⁻CD25⁻ thymic cells (the same below). (b) The graphs display the statistical analysis of the frequency and number of ETPs among total thymocytes. (c,d) Flow cytometric analysis of ETPs (Lin⁻CD25⁻CD44⁺c-Kit⁺) in Ltbr^fl/fl Tek^Cre and control mice. (c) Representative dot plots are shown. (d) The graphs display the statistical analysis of the frequency and number of ETPs among total thymocytes. (e,f) Short-term thymic homing assay in Ltbr^−/− (e), Ltbr^fl/fl Tek^Cre (f) and littermate control mice. The frequency (e) and number (f) of donor-derived lineage-negative cells among total thymocytes were analysed by flow cytometry. The data are representative of at least two independent experiments with three or more mice per group in each experiment. Error bars represent s.e.m. Asterisks mark statistically significant difference (*P<0.05, **P<0.01 and ***P<0.001 determined by two-tailed unpaired Student's t-test).

Figure 2. LT and LIGHT coordinate to control thymic ETP population and progenitor homing.

(a,b) Flow cytometric analysis of ETPs (Lin⁻CD25⁻CD44⁺c-Kit⁺) in Lta^+/− and Lta^−/− mice. (a) Representative dot plots are shown. (b) The graphs display the statistical analysis of the frequency and number of ETPs among total thymocytes. (c,d) Flow cytometric analysis of ETPs (Lin⁻CD25⁻CD44⁺c-Kit⁺) in Light^+/− and Light^−/−mice. (c) Representative dot plots are shown. (d) The graphs display the statistical analysis of the frequency and number of ETPs among total thymocytes. (e,f) Flow cytometric analysis of ETPs (Lin⁻CD25⁻CD44⁺c-Kit⁺) in Lta^−/− Light^−/− and littermate control mice. (e) Representative dot plots are shown. (f) The graphs display the statistical analysis of the frequency and number of ETPs among total thymocytes. (g) Short-term thymic homing assay in Lta^−/− Light^−/− and littermate control mice. The frequency and number of donor-derived lineage-negative cells among total thymocytes were analysed by flow cytometry. The data are representative of at least two independent experiments with three or more mice per group in each experiment. Error bars represent s.e.m. Asterisks mark statistically significant difference (*P<0.05, **P<0.01 and ***P<0.001 determined by two-tailed unpaired Student's t-test).

Figure 3. T cells deliver LTβR signalling to control thymic ETP population and progenitor homing.

(a,b) Flow cytometric analysis of ETPs (Lin⁻CD25⁻CD44⁺c-Kit⁺) in WT and Tcra^−/− mice. (a) Representative dot plots are shown. (b) The graphs display the statistical analysis of the frequency and number of ETPs among total thymocytes. (c) Short-term thymic homing assay in WT and Tcra^−/− mice. The frequency and number of donor-derived lineage-negative cells among total thymocytes were analysed by flow cytometry. (d,e) Flow cytometric analysis of ETPs (Lin⁻CD25⁻CD44⁺c-Kit⁺) in WT and Tcra^−/−, and Lta^−/− and Tcra^−/− mixed bone marrow chimeric mice. (d) Representative dot plots are shown. (e) The graphs display the statistical analysis of the frequency and number of ETPs among total thymocytes. (f) Short-term thymic homing assay in WT and Tcra^−/−, and Lta^−/− and Tcra^−/− mixed bone marrow chimeric mice. The frequency and number of donor-derived lineage-negative cells among total thymocytes were analysed by flow cytometry. (g,h) Flow cytometric analysis of ETPs (Lin⁻CD25⁻CD44⁺c-Kit⁺) in WT and Tcra^−/−, and Lta^−/−Light^−/− and Tcra^−/− mixed bone marrow chimeric mice. (g) Representative dot plots are shown. (h) The graphs display the statistical analysis of the frequency and number of ETPs among total thymocytes. (i) Short-term thymic homing assay in WT and Tcra^−/−, and Lta^−/−Light^−/− and Tcra^−/− mixed bone marrow chimeric mice. The frequency and number of donor-derived lineage-negative cells among total thymocytes were analysed by flow cytometry. The data are representative of at least two independent experiments with three or more mice per group in each experiment. Error bars represent s.e.m. Asterisks mark statistically significant difference (*P<0.05, **P<0.01 and ***P<0.001 determined by two-tailed unpaired Student's t-test).

Figure 4. Ly6C⁻Selp⁺ thymic ECs are the specialized population associated with perivascular spaces and immigrating progenitor cells.

(a) Flow cytometric analysis of P-selectin and Ly6C expression on thymic ECs from WT mice, gated on the CD45⁻EpCAM⁻CD31⁺ cell population, as shown in Supplementary Fig. 7a. (b) Immunofluorescence analysis of Ly6C (green) expression on CD31⁺ (red) thymic ECs associated with perivascular spaces, which were defined with double-basement membrane staining of collagen IV (cyan). Scale bars, 20 μm. (c) Statistical analysis of the distribution of Ly6C⁻CD31⁺ thymic ECs in PVS and non-PVS areas. (d) Statistical analysis of the percentages of PVS or non-PVS associated with Ly6C⁻CD31⁺ thymic ECs. (e) Representative images showing the location of thymic seeding progenitor cells (CD45.2⁺, circled red) closer to Ly6C⁻ vessels. Scale bars, 20 μm. The data are representatives of at least three independent experiments (a,b,e); for statistical analysis (c,d), >40 fields (× 100) containing at least six non-PVS- or PVS-associated vessels from six WT mice were included for analysis. Error bars represent s.e.m. Asterisks mark statistically significant difference (***P<0.001 determined by two-tailed unpaired Student's t-test).

Figure 5. LTβR signalling and T cells are required for the development and homeostasis of Ly6C⁻Selp⁺ thymic ECs.

(a,b) Flow cytometric analysis of the thymic EC subsets in Ltbr^−/− and littermate control mice. The ECs were gated on the CD45⁻EpCAM⁻CD31⁺ population, as shown in Supplementary Fig. 7a. (a) Representative dot plots are shown. (b) The graph displays the statistical analysis of the frequency of each endothelial subset. (c–e) Six-day-old neonatal WT mice were treated with LTβR agonistic antibody (clone 9B10) or control antibody twice with an interval of 5 days between treatments; the thymic EC subsets were analysed by flow cytometry. (c) Representative dot plots are shown. The graph displays the statistical analysis of the frequency of each endothelial subset (d) and the ratio of Ly6C⁻Selp⁺ and Ly6C⁻Selp⁺ thymic ECs (e). (f,g) Adult WT mice (4–6 weeks old) were treated with LTβR-hIgG or hIgG once a week for 4 weeks; the thymic EC subsets were analysed by flow cytometry. (f) Representative dot plots are shown. (g) The graph displays the statistical analysis of the frequency of each endothelial subset. (h,i) Flow cytometric analysis of the thymic EC subsets in WT and Tcra^−/− mice. The thymic ECs were gated on the CD45⁻EpCAM⁻CD31⁺ population. (h) Representative dot plots are shown. (i) The graph displays the statistical analysis of the frequency of each endothelial subset. The data are representative of at least two independent experiments with three or more mice per group in each experiment. Error bars represent s.e.m. Asterisks mark statistically significant difference (*P<0.05, **P<0.01 and ***P<0.001 determined by two-tailed unpaired Student's t-test).

Figure 6. LTβR is required for thymic regeneration upon injury.

(a) Short-term thymic homing assay in WT and Ltbr^−/− mice on SL-TBI. The frequency of donor-derived lineage-negative cells was analysed by flow cytometry. (b) The total thymic cellularity of WT and Ltbr^−/− mice 4 weeks after thymic injury induced by SL-TBI. (c) The number of developing thymocyte subsets in WT and Ltbr^−/− mice 4 weeks after thymic injury induced by SL-TBI. The data are representative of at least two independent experiments with three or more mice per group in each experiment. Error bars represent s.e.m. Asterisks mark statistically significant difference (*P<0.05, **P<0.01 and ***P<0.001 determined by two-tailed unpaired Student's t-test).