Cell-intrinsic regulation of peripheral memory-phenotype T cell frequencies

Abstract

Memory T and B lymphocyte numbers are thought to be regulated by recent and cumulative microbial exposures. We report here that memory-phenotype lymphocyte frequencies in B, CD4 and CD8 T-cells in 3-monthly serial bleeds from healthy young adult humans were relatively stable over a 1-year period, while Plasmablast frequencies were not, suggesting that recent environmental exposures affected steady state levels of recently activated but not of memory lymphocyte subsets. Frequencies of memory B and CD4 T cells were not correlated, suggesting that variation in them was unlikely to be determined by cumulative antigenic exposures. Immunophenotyping of adult siblings showed high concordance in memory, but not of recently activated lymphocyte subsets. To explore the possibility of cell-intrinsic regulation of T cell memory, we screened effector memory-phenotype T cell (TEM) frequencies in common independent inbred mice strains. Using two pairs from these strains that differed predominantly in either CD4 TEM and/or CD8 TEM frequencies, we constructed bi-parental bone marrow chimeras in F1 recipient mice, and found that memory T cell frequencies in recipient mice were determined by donor genotypes. Together, these data suggest cell-autonomous determination of memory T niche size, and suggest mechanisms maintaining immune variability.

Fig 1. Comparison of intra-individual and inter-individual variance for immune subset frequency.

Box plots show comparison of intra-individual versus inter-individual variance for the immune subset frequencies indicated in each panel. Intra-individual variances indicate variance of subset frequency over 4 time points in each individual (n = 43). Inter-individual variances indicate variance of subset frequency in randomly chosen set of different individuals (n = 43). P-values obtained by bootstrapping are as indicated in the panels. Memory B cells, Naive B cells and Plasmablasts are expressed as frequency (%) of total B cells. Memory CD4 and naive CD4 frequencies are expressed as % of total CD4 T cells. Memory CD8, Naive CD8 and CD8 TEMRA frequencies are expressed as % of total CD8 T cells. Memory B cell subset was defined as CD19+CD20+CD38loCD27+CD43-. Naive B cell subset was defined as CD19+CD20+CD38loCD27-CD43-. Plasmablast subset was defined as CD19+CD20-CD38hi. For both CD4 and CD8 T cells, memory subset was defined as the sum of effector memory and central memory subsets (CD45RO+). Naive T cells were defined as CD45RO- CCR7+. TEMRA cells were defined as CD45RO-CCR7-. Box plots indicate median and interquartile ranges of variances of cell frequencies and counts in Human volunteers. Upper whisker extends till the highest value that is within 1.5 times the interquartile range from 3rd quartile. Lower whisker extends till the lowest value that is within 1.5 times the interquartile range from 1st quartile. Outliers are shown as dots.

Fig 2

Correlation between (A) memory B and memory CD4 T cell frequencies and (B) between memory CD4 and memory CD8 T cell frequencies. Memory CD4 frequency is expressed as % of total CD4 T cells, memory CD8 frequency is expressed as % of total CD8 T cells and memory B cell frequencies are expressed as % of total B cells. Each dot represents the median value of cell subset frequency of each longitudinally sampled volunteer (n = 43). Correlation coefficient (Spearman) and p-values are indicated in Table 1. Pair-wise comparison of all cell subsets is shown in S9 Fig.

Fig 3. Comparison of sibling-pair differences versus unrelated-pair differences.

Each panel indicates an immune cell subset expressed as frequency of parent lineage gate. Vertical red line represents the mean difference in sibling pairs. Density histograms represent distribution of differences in large number of unrelated-pairs. The farther the red vertical line towards the left, the greater is the significance level and indicates that sibling-pair differences are smaller than unrelated-pair differences. The p-values as calculated from the distribution by resampling methods are indicated in the plots.

Fig 4

Gating strategy for memory subsets in mouse splenic CD4 T cells (A to D) and CD8 T cells (E to H). Each plot shows representative gates from a single mouse strain as indicated. CD4 memory subsets are gated on conventional CD4+ CD25- gate. CD8 memory subsets are gated on CD8+ gate. Central memory (TCM) subset is defined as CD44hi CD62Lhi and Effector memory (TEM) subset is defined as CD44hi CD62Llo. Boxes indicate TCM and TEM gates as depicted.

Fig 5

CD4 TEM and CD8 TEM frequencies and counts (in millions) in spleen in C57BL/6J versus CBA/CaJ (A to D) and BALB/cJ versus SJL/J (E to H). Each dot represents data from one individual mouse (n > 25 per group). CD4 TEM and CD8 TEM frequencies indicate frequencies of TEM compartment (CD44hi CD62Llo) out of total conventional CD4 (CD4+CD25-) and CD8 T cells respectively. Cell counts are expressed in millions. Box plots indicate median and interquartile ranges. Upper whisker extends till the highest value that is within 1.5 times the interquartile range from 3rd quartile. Lower whisker extends till the lowest value that is within 1.5 times the interquartile range from 1st quartile. The p-values obtained by non-parametric tests are as indicated in the panels.

Fig 6

Frequencies of CD4 TEM and CD8 TEM in the spleen from donor partner in mixed bone marrow chimera of B6.SJL-CBA/CaJ pair (A, B) and SJL/J- BALB/cJ pair (C, D). Each dot represents data from one individual mouse (n > 9 per group) and individual mice data are connected by lines. The p-values are derived from paired non-parametric tests and are indicated in the panels. TEM subset is defined as CD44hi CD62Llo. CD4 TEM cell frequency is expressed as % of conventional CD4 T cells and CD8 TEM cell frequency is expressed as % of CD8 T cells.