β-Selection-induced proliferation is required for αβ T cell differentiation

Abstract

Proliferation and differentiation are tightly coordinated to produce an appropriate number of differentiated cells and often exhibit an antagonistic relationship. Developing T cells, which arise in the thymus from a minute number of bone-marrow-derived progenitors, undergo a major expansion upon pre-T cell receptor (TCR) expression. The burst of proliferation coincides with differentiation toward the αβ T cell lineage-but the two processes were previously thought to be independent from one another, although both were driven by signaling from pre-TCR and Notch receptors. Here we report that proliferation at this step was not only absolutely required for differentiation but also that its ectopic activation was sufficient to substantially rescue differentiation in the absence of Notch signaling. Consistently, pharmacological inhibition of the cell cycle machinery also blocked differentiation in vivo. Thus the proliferation step is strictly required prior to differentiation of immature thymocytes.

Figure 1. αβ T cell differentiation is associated with proliferation in WT and genetically manipulated mice

A. WT DN3a cells were sorted, labeled with CFSE and placed in OP9-DL1 co-culture. CFSE dilution, CD4 and CD8 expression was analyzed 4 days later. CFSE dilution vs CD4 and CD8 expression (summed electronically) (left) and expression of CD4 and CD8 individually on divided and undivided cells (center and right) is shown. B. DN3a and DN3b cells from WT and Myc^fl/fl Lck-Cre were treated and analyzed as in A. C. CellTrace Violet-labeled Ccnd3^−/− and Ccnd3^+/- DN3a cells were placed in OP9-DL1 co-cultures for 4 (top panels) or 8 (bottom panels) days. Dilution of CellTrace Violet, and expression of CD4 and CD8 was assessed. Numbers above CellTrace Violet dilution plots represent the number of divisions. Note the dulling of CellTrace Violet from day 4 to day 8. Representative results of two independent experiments are shown. See also Figure S1.

Figure 2. Proliferation is required for αβ T cell differentiation

CellTrace violet-labeled Vav-Bcl2 transgenic DN3a cells were cultured on OP9-DL1 monolayers in the presence or absence of 1μM PD0332991 cdk inhibitor for two days; undivided undifferentiated cells were resorted and cultured for additional three days with or without the inhibitor. Expression of CD4 and CD8 (A, B) and CellTrace violet dilution (A) was analyzed at both time points. C, D. CellTrace violet-labeled Vav-Bcl2 transgenic DN3a cells were cultured in the presence or absence of the indicated inhibitors (for concentrations see Supplemental Experimental Procedures). On day 4, CD4 and CD8 upregulation (C) and CellTrace violet dilution (D) were analyzed. See also Figure S2.

Figure 3. Proliferation dependence of the αβ T cell program is not limited to CD4 and CD8 coreceptor upregulation and holds true in vivo

A, B CellTrace violet-labeled Vav-Bcl2 transgenic DN3a cells were cultured on OP9-DL1 monolayers in the presence or absence of 1μM PD0332991 or 1 mM thymidine for 3 days, stained for surface expression of CD25 and intracellular expression of TCRβ and RORγt. RORγt (A) and CD25 (B) expression is plotted against CellTrace violet dilution (top) or is compared in TCRβ⁺ and TCRβ⁻ cells (bottom). C. WT mice were treated with PD0332991 or vehicle for 5 days as described in Supplemental Experimental Procedures. Expression of CD4 and CD8 on Lin⁻ (Gr1⁻ CD11b⁻CD11c⁻CD19⁻Ter119⁻NK1.1⁻) thymocytes (left), surface expression of TCRβ on Lin⁻CD4⁺CD8⁺ cells (center) and CD44 vs CD25 expression on DN Lin⁻TCRβ⁻ thymocytes (right) are shown. See also Figure S3.

Figure 4. Several rounds of cell division are required for efficient progression to the DP stage

Vav-Bcl2 transgenic DN3a cells were sorted, labeled with CellTrace violet and placed in OP9-DL1 co-culture. On day 3 individual CellTrace violet peaks corresponding to different number of cell divisions were sorted and cells were placed into secondary OP9-DL1 co-culture with or without 1 mM thymidine to block further divisions. Expression of CD4 and CD8 was assessed after two days of secondary culture. Representative FACS plots (A) and frequency of DP cells in the cultures (B) is shown. As the cell cycle block was not always complete, letting some cells undergo one more division, additional gating on cells that did not divide in secondary cultures was applied for cultures in the presence of thymidine. Representative results of three independent experiments are shown.

Figure 5. Ectopic activation of proliferation is insufficient to compensate for the lack of pre-TCR signaling

A, B. DN3 cells generated in OP9-DL1 co-cultures from Vav-Bcl2 transgenic Rag2^−/− HSCs were infected with retroviruses encoding IRES-GFP, Ccnd3-IRES-GFP, proliferator-IRES-GFP, or TCRβ-IRES-GFP, sorted, labeled with CellTrace Violet and plated again on OP9-DL1 monolayers. Six days later CellTrace violet dilution (A, gated on GFP⁺ cells) as well as CD4 and CD8 upregulation (B) was assessed. Representative results of four independent experiments are shown. C. Differentiation of Vav-Bcl2 Tg, Rag-sufficient cells treated as in B is shown. See also Figure S4.

Figure 6. Ectopic activation of proliferation rescues DP progression in the absence of Notch signaling

A. DN3 cells generated in OP9-DL1 co-cultures from Vav-Bcl2 transgenic HSCs were infected with retroviruses encoding IRES-GFP or proliferator-IRES-GFP, sorted, labeled with CellTrace Violet and plated again on OP9-DL1 or control OP9 monolayers. Six days later cells were harvested, stained for surface CD45, CD4 and CD8 expression, fixed, permeabilized and stained for intracellular TCRβ. CellTrace violet dilution (right) as well as CD4 and CD8 upregulation (left) were analyzed in CD45⁺IC TCRβ⁺GFP⁺ cells. Representative results of five independent experiments in which proliferator-infected cells divided three times or more are shown. B, C. Cells were generated, infected with Myc-IRES-GFP or empty vector, cultured and analyzed as in A. Myc-infected cells were cultured in presence or absence of 1μM PD0332991. CellTrace violet dilution (B, right), CD4 and CD8 upregulation (B, left) were analyzed in CD45⁺IC TCRβ⁺GFP^hi cells. C shows GFP vs CellTrace violet and GFP vs CD8 plots for Myc-infected CD45⁺IC TCRβ⁺ cells. Representative results of three independent experiments. See also Figure S5.

Figure 7. Tcf3^−/− DN3a cells can progress to the DP stage without proliferation

A. WT or Tcf3^−/− DN3a cells were labeled with CellTrace violet and placed in OP9-DL1 cocultures in the presence or absence of 1μM PD0332991. CellTrace violet dilution as well as CD4 and CD8 expression was analyzed on day 3. B. Cell cycle status of ex vivo WT or Tcf3^−/− DN3a and DN3b thymocytes was analyzed by DAPI staining. C. WT or Tcf3^−/− DN3a cells (from bm chimeras, as described in the Supplemental Experimental Procedures) were labeled with CellTrace violet, placed on OP9-DL1 or control OP9 monolayers and analyzed as in A. See also Figure S6.