MHC drives TCR repertoire shaping, but not maturation, in recent thymic emigrants

Abstract

After developing in the thymus, recent thymic emigrants (RTEs) enter the lymphoid periphery and undergo a maturation process as they transition into the mature naive (MN) T cell compartment. This maturation presumably shapes RTEs into a pool of T cells best fit to function robustly in the periphery without causing autoimmunity; however, the mechanism and consequences of this maturation process remain unknown. Using a transgenic mouse system that specifically labels RTEs, we tested the influence of MHC molecules, key drivers of intrathymic T cell selection and naive peripheral T cell homeostasis, in shaping the RTE pool in the lymphoid periphery. We found that the TCRs expressed by RTEs are skewed to longer CDR3 regions compared with those of MN T cells, suggesting that MHC does streamline the TCR repertoire of T cells as they transition from the RTE to the MN T cell stage. This conclusion is borne out in studies in which the representation of individual TCRs was followed as a function of time since thymic egress. Surprisingly, we found that MHC is dispensable for the phenotypic and functional maturation of RTEs.

Figure 1

The TCR repertoires of RTEs and MN T cells are globally similar. A, Naïve (CD44lo/mid) T cells from RAG2p-GFP Tg mice (black line) were gated as RTEs and MN T cells on the basis of GFP level, using GFP⁻ T cells from B6 mice (gray line) to determine the cutoff for RTEs. B, C, RTE and MN T cells were subdivided on the basis of TCR Vα and Vβ expression for bothCD4 (B) and CD8 (C) T cells. Data are averaged from 6 independent experiments analyzing a total of 5–12 mice per TCR Vα/β subunit, with error bars representing SD. *, p < .005. D, RTE and MN expression levels of TCR Vβ4 are shown for individual mice. p = .002 and p = .0001 for CD4 RTEs vs. MN T cells and CD8 RTEs vs. MN T cells, respectively, using a paired 2-tailed Student’s t-test.

Figure 2

The TCR repertoires of RTEs and MN T cells are subtly distinct. TCR CDR3 length spectratyping was performed on cDNA from sorted populations of naïve (CD44^lo/mid CD62L^hi) CD4 and CD8 RTEs and MN T cells. Representative analyses of the indicated TCR Vβ or Vβ-Jβ recombination are shown for both CD4 (A) and CD8 (B) T cells. Arrows indicate ≥ 10% CDR3 length differences between RTEs and MN T cells, approximately 2-fold the SD. Data are representative of 3 replicates from a pool of 5 mice. Error bars indicate SD from the mean of values from 3 replicates. *, p < .02.

Figure 3

The frequency of individual TCRs is modulated in RTEs. Lethally irradiated mice were reconstituted with a mixture of bone marrow from polyclonal and 2 TCR Tg donors. A, T cells from each population of cells were divided into 3 groups for both CD4 (left panel) and CD8 (right panel) chimeras: GFP^hi (young) and GFP^lo (old) RTEs, and GFP⁻ MN T cells. B, The frequencies of OT-II and SMARTA Tg (left panel) and OT-I and P14 Tg (right panel) T cells within young RTE, old RTE, and MN T cell populations was normalized to the polyclonal level at that stage. Data are representative of 3 independent sets of chimeras for a total of 6–12 individual mice.

Figure 4

RTEs mature normally following transfer to a classical MHC II-deficient environment. At the indicated time points post-transfer, CD24 and CD45RB expression by sorted CD4 RTEs transferred into I-A_β-deficient (RTEs in MHC II^−/−) or CD4- deficient (RTEs in CD4^−/−) mice were determined. Splenocytes from an unmanipulated RAG2p-GFP Tg mouse were analyzed on the same day for marker expression by CD4⁺CD44^lo/mid RTEs (Naive GFP⁺ peripheral T) and MN (Naïve GFP⁻ peripheral T) cells. Data are representative of 4 independent experiments.

Figure 5

RTEs mature normally in situ in a classical MHC II-deficient lymphoid periphery. A, B, CD24 and CD45RB expression levels by CD4⁺GFP⁺ RTEs from RAG2p-GFP Tg × K14-I-A_β Tg mice (RTEs from MHC II⁻ periphery) or from a MHC II⁺ periphery are shown. Splenocytes from an unmanipulated RAG2p-GFP Tg mouse were analyzed on the same day for marker expression by CD4⁺CD44^lo/mid RTEs (Naive GFP⁺ peripheral T) and MN (Naive GFP⁻ peripheral T) cells. Representative data are shown in A, and data in B are averaged median fluorescence intensity of RTEs split into young GFP^hi and old GFP^lo populations from 3 independent experiments analyzing a total of 6–10 mice per group, with error bars representing SD. Differences were not statistically significant (p > .05).

Figure 6

CD4 pre-RTEs mature normally following transfer to mice expressing no MHC II molecules. A, Sorted CD4⁺CD8⁻GFP⁺CD62L^hi thymocytes (pre-RTEs) were transferred into complete MHC II^−/− or CD4^−/− mice and Qa2 and CD45RB expression levels were determined on cells 8 d following transfer. Thymocytes and splenocytes from an unmanipulated RAG2p-GFP Tg mouse were analyzed on the same day for marker expression by CD4⁺CD8⁻GFP⁺CD62L^hi thymocytes (Mature SP thymocytes) and CD4⁺CD44^lo/mid MN (Naïve GFP⁻ peripheral T) cells. Data are representative of 2 experiments, analyzing a total of 4 mice per condition. B, Sorted congenic CD4⁺CD62L^hi RTEs were transferred to classical MHC II^−/− or CD4^−/− mice, and at 9 d following transfer, were CFSE-labeled and stimulated in vitro for 3 d with anti-CD3 and anti-CD28. Freshly sorted CD4⁺CD62L^hi MN T cells were concomitantly CFSE-labeled and stimulated. Labeled histograms depict CFSE dilution of stimulated (darker lines) or unstimulated (lighter lines) samples. Indicated are the percents of cells that have divided 4 or more times. Data are representative of 2 recipients per condition.