Ageing combines CD4 T cell lymphopenia in secondary lymphoid organs and T cell accumulation in gut associated lymphoid tissue

C57BL/6 animals were analysed at three intervals of ages: 2 to 6 months (young), 10 to 14 months (middle-aged) and 22 to 26 months (old). We first determined the absolute number of CD45+ TCRβ+ CD4+ (CD4 T cells) or CD45+ TCRβ+ CD8α+ (CD8 T cells) lymphocytes recovered from secondary lymphoid organs (spleen and lymph nodes) at different ages (Figure 1A). We observed a significant decrease in the number of CD4 T cells in lymphoid organs in “middle-aged” and “old” mice compared to “young” animals. No significant decay was further detected between 10–14 months and 22–26 months old mice. In contrast, we did not detect variations in CD8 T cell numbers whatever the age interval considered. We thus confirmed a differential behaviour of CD4 and CD8 T cells depending on the age [15, 34‐36]. Ageing is commonly associated to a shift of naïve and effector/memory T cell subsets proportion towards effector/memory T cells [25, 29, 37, 38]. We thus evaluated whether the proportion of naïve and effector/memory cell pools were equally affected in CD4 and CD8 T cell subsets during ageing. Analysing L-selectin (CD62L), a lymphoid homing molecule, and CD44 expression, we identified naïve and effector memory T cell subsets in secondary lymphoid organs, as exemplified on splenic T cells in Figure 1B. Analyses on CD4 T cells were performed on FOXP3- cells to exclude regulatory T cells (FOXP3+) that may exhibit different homeostasis compared to the so-called conventional FOXP3- CD4 T cells. Increasing frequencies of both effector/memory CD4 and CD8 T cells were already detectable when comparing middle-aged to young animals (Figure 1C). However the amplitude of skewing differed when considering CD4 and CD8 T cell subsets: effector/memory CD4 T shift was progressively developing from young to old animals, whereas effector/memory CD8 T shift was mildly increased in middle-aged animals, and became more intense from middle-aged to old animals. Thus, whereas CD4 and CD8 T cell numbers are differently affected during ageing, increasing proportion of effector/memory T cells are detected in both CD4 and CD8 T cells. We further investigated this apparent discrepancy by analysing naïve and effector/memory T cell numbers. Naïve CD4 T cell numbers were significantly different between young and middle-aged mice whereas naïve CD8 T cell numbers remain statistically stable in the same interval (Figure 2A). However, both naive CD4 and CD8 T cell numbers were significantly decreased in old mice compared to young animals. We thus observed a different impact of ageing on CD4 and CD8 T cells among naïve T cell compartment: naïve CD4 T cell compartment being affected more rapidly than naïve CD8 T cell pool. Reduction in naïve CD4 T cell numbers in lymphoid organs essentially corroborated the decay observed in total CD4 T cells. However, the drastic decay observed in naïve CD8 T cells in old mice was not reflected in total CD8 T cell numbers. We next considered the evolution of effector/memory T cell numbers: the number of effector/memory CD4 T cells (TCRβ+ CD4+ FOXP3- CD62L- CD44high cells) significantly increased in middle-aged and old mice, albeit with low amplitude, compared to young animals (Figure 2B). Evolution of regulatory FOXP3+ CD4+ T cells (Treg) numbers exhibited a similar profile than effector/memory CD4 T cells numbers (not shown). Regarding effector/memory CD8 T cell numbers (TCRβ+ CD8α+ non CD62L+CD44- cells) recovered from lymphoid organs, no significant difference was detected between young and middle-aged animals. However, a significant increase was detected when comparing old animals to young or middle-aged groups (Figure 2B). As a control for ageing, we finally determined thymocyte numbers in the three groups of age. As expected, reduction in thymocyte numbers, a clear hallmark of T cell ageing, was significantly detected in middle-aged mice and further aggravated in old mice (Figure 2C).

Collectively, analysing naïve and effector/memory absolute numbers provided interesting insights on the shift of naïve T cells towards effector/memory T cells during ageing. We observed that physiological ageing is not equally affecting CD4 and CD8 T cell pools. Total CD4 T cell decay reflected massive reduction of naïve CD4 T cells occurring in middle-aged animals combined to a mild increase of effector/memory CD4 T cells in old animals. A different timeline emerged when considering CD8 T cell compartment: naïve and effector/memory CD8 T cells numbers were essentially not affected in middle-aged animals in contrast to older animals who exhibited clear naïve CD8 T cell decay and increase in effector/memory CD8 T cells.

Because some contradictions emerged from data on T cell numbers recovered from lymph nodes and/or spleen [14, 39], we next ascertain whether differential behaviour of CD4 and CD8 T cells was homogenous in all secondary lymphoid organs. When considering separately spleen, mesenteric lymph nodes and superficial lymph nodes (i.e. axillary, brachial and inguinal lymph nodes), CD4 T cell decay was detected in all organs when comparing middle-aged or old mice to young animals (Figure 3A left). However, the amplitude differed: CD4 T cells from superficial lymph nodes appeared more affected than those in mesenteric lymph nodes and spleen. Because total CD8 T cell numbers were essentially preserved in pooled secondary lymphoid organs analysis, we were not expecting a major difference in secondary lymphoid organs considered individually. As expected, numbers of CD8 T cells recovered in the spleen and mesenteric lymph node were essentially not affected, as mice grew older. However, superficial lymph nodes exhibited a different profile revealing a significant decay in the numbers of CD8 (Figure 3A right). In conclusion, T cell distribution was gradually affected depending on the lymphoid organs considered: splenic cells appeared mildly affected; mesenteric lymph nodes exhibited partial T cell lymphopenia; T cell lymphopenia was more marked in superficial lymph nodes. To directly compare T cell decay in each secondary lymphoid structure considered, we presented the percentage of residual CD4 and CD8 T cells in middle-aged and old animals compared to young mice (Figure 3B). The percentage of residual CD4 T cells at 10–14 months (middle-age) was significantly lower in the superficial lymph nodes compared to mesenteric lymph nodes and spleen (p < 0.01 and <0.001 respectively). CD4 T cell decay in mesenteric lymph nodes resembled those observed in spleen. Stronger decay in superficial lymph nodes compared to mesenteric lymph nodes or spleen was also found in 22–26 months old mice (p < 0.05 and <0.001 respectively). Similarly, the specific impact of ageing in superficial lymph nodes was detected among CD8 T cells, although detectable solely in old animals (p < 0.05 and p < 0.001 when comparing to mLN and SP respectively). We thus observed that age related T cell lymphopenia is exacerbated in superficial lymph nodes, whereas T cells in mesenteric lymph nodes and spleen are better preserved.

The obvious dichotomy between superficial and mesenteric lymph nodes led us to further investigate gut associated lymphoid structure. Because CD4 T cells in the gut have been essentially located in the lamina propria, we isolated T lymphocytes from colonic lamina propria (cLP), small intestine lamina propria (siLP) and associated lymphoid structures: Peyer’s patches. The gating strategy is shown in Figure 4A. We studied the dynamics of T cell numbers in these structures depending on the age. In Peyer’s patches, CD4 T cell numbers increased with age but absolute CD8 T cell numbers remained constant through ageing. In siLP, both CD4 and CD8 absolute numbers progressively increased from young to middle-aged or old mice (Figure 4B). In cLP, we observed a different dynamic: increase in CD4 and CD8 T cell numbers was predominantly observed between 10–14 months and 22–26 months (Figure 4B). We performed a similar set of experiment in a different strain background using B6/SJL F1 animals. Comparing “young” and “intermediate” B6/SJL F1 animals, we detected decreased thymocytes (157.9 +/−13.5 10⁶ versus 15.1 +/−5.6 10⁶ thymocytes in young and middle-aged animals respectively; <0.0001). Regarding gut associated lymphoid tissue (GALT), an increase in CD4 T cell numbers in Peyer’s patches was detected (0.35 +/−0.09 10⁶ versus 0.85 +/−0.15 10⁶ CD4 T cells in young and middle-aged animals respectively; p 0.0128) and preserved CD4 T cell counts were observed in cLP and siLP. Thus we confirmed in a different strain model that T cell lymphopenia developing in lymphoid organs was not recapitulated in GALT and even associated to CD4 T cell accumulation at some sites.

Simultaneous accumulation of CD4 T cell at mucosal sites and the development of T cell lymphopenia at lymphoid sites raised the question of a direct association between these two phenomena during ageing. Indeed, a strong inverse association was observed between CD4 T cell numbers in colonic lamina propria (cLP) sites and lymphoid sites (Figure 4C). To note, the association was not detected when considering small intestine lamina propria (siLP) or Peyer’s patches (PP) presumably due to the limited variation in numbers observed in these organs. Indeed, ageing induced a 6-fold increase in CD4 T cells in cLP but a 2-fold increase in siLP and PP.

To evaluate whether such accumulation in GALT is reflecting local proliferation or, instead, recruitment to the site, we determined Ki67 expression in CD4 T cells recovered from GALT. We could not detect any increase in Ki67 expression in any of the three gut associated lymphoid structures during ageing. A significant reduction of Ki67 positive CD4 T cells was even detected in all three organs at old ages. Reduced proliferation was already detectable at middle age in siLP and cLP, suggesting local proliferation was not responsible for such accumulation (Figure 5A). As a control, Ki67 expression was also assessed in secondary lymphoid organs, which exhibited significant increase in proliferative fraction (data not shown). Because GALT has been described as a favourable environment for regulatory T cells (Treg) induction, we finally ascertain that CD4 T cell accumulation was not directly reflecting specific accumulation or production of FOXP3 expressing Treg. No or mild increase in Treg percentages was detected in lamina propria (siLP or cLP) and Peyer’s patches respectively. We also examined the percentages of γ-IFN, IL-4 and IL-17 producing CD4 T cells in secondary lymphoid organs and GALT. Because no deliberate immune response was induced in these animals, percentages of cytokine producing CD4 T cells were relatively low (< 5% for IL-4 and IL-17; < 10% for γ-IFN) confirming that CD4 T cell accumulation in GALT was not related to on-going immune responses. We detected increased percentages of γ-IFN and IL-4 producing CD4 T cells in secondary lymphoid organs as previously described but not among GALT CD4 T cells recovered from old animals (data not shown).

Collectively, our data suggest that CD4 T cell increase observed in the GALT reflect progressive recruitment and accumulation of CD4 T cells from lymphoid sites rather than peripheral expansion of gut resident CD4 T cells.

We next questioned whether increase in CD4 T cell numbers observed in gut associated lymphoid tissue (GALT) was a common feature of all tertiary lymphoid organs and even of non-lymphoid tissues in middle-aged animals. We thus analysed two additional sites: the mucosa associated lymphoid tissues (MALT) in the lungs and a non-lymphoid tissue: the liver which has been described to comprise important T cell numbers, especially double negative CD4- CD8- T cells [40, 41]. To note, T cell recovery in young animals confirmed that GALT represented a major site of CD4 T cell accumulation: CD4 T cell recovery in lungs and liver were approximately 15-fold lower than numbers recovered from small intestine lamina propria (Figure 6A). In the lungs, CD4 T cells were preserved at middle age, whereas CD8 T cell absolute numbers increased, thus demonstrating that lungs MALT differed from GALT. In the liver, no difference among CD4 or CD8 T cell absolute numbers was detected between middle-aged and young mice. Collectively, we demonstrated that CD4 T cell accumulation was essentially restricted to the gut associated lymphoid structures.

Sex of animals and sex hormones may differentially regulate CD4 homeostasis [42]. To ascertain whether CD4 T cell redistribution in cLP occurs equally in male and female animals, we analysed CD4 T cell distribution based on the sex of the animals. As shown in Figure 6B, sex had only minor influence on CD4 T cell lymphopenia at lymphoid sites and CD4 T cell accumulation at mucosal sites. Comparing young, intermediate and old animals, we observed an equal decay of thymocytes in male and female animals. In pooled secondary lymphoid organs, CD4 T cells decay was observed in both groups, although significantly more exacerbated in female animals (p 0.0156 when comparing old groups). Conversely, CD4 T cell numbers recovered in cLP were equally increased in male and female animals. These data demonstrate that CD4 T cell accumulation in gut associated lymphoid tissue associated to CD4 T cell lymphopenia during ageing occurs both in male and female animals, although the amplitude of CD4 T cell depletion in lymphoid organs appeared to be more drastic among female mice.

6 to 8 weeks old C57BL/6 mice were purchased from Janvier Laboratories, and mated in our facilities. Mice were sacrificed at 2 to 6 months of age, 10 to 14 months of age and at 22 to 26 months old respectively named “young”, “middle-aged” and “old” animals. Mice were maintained under pathogen free conditions at the central animal facility of Paris-Sud Faculty of Medicine. As control, young (2 months, n = 6) and middle aged B6-SJL F1 animals (11 months, n = 6) were also analysed. All protocols were conducted in compliance with French and European animal welfare regulations (agreement B-94-043-12 and license 94–440, delivered by the French veterinary authorities) and validated by the Ethics Committee of University Paris-Sud (CEEA27).

Cells counts were performed in duplicates after addition of Trypan blue dye using Malassez haemocytometer cell.

Extracellular staining was preceded by incubation with purified anti-CD16/32 antibodies (FcγRII/III block, 2.4G2) (eBioscience) to block non-specific staining. Cells were stained with FITC-, PE-, PerCP-Cy5.5-, PE-Cy5-, PE-Cy7-, APC-, and APC-H7- labelled appropriate antibodies including: TCRβ (H57-597, eBioscience); CD45 (30-F11, eBioscience and BD Biosciences); CD44 (IM7, eBioscience); CD8α (53–6.7, eBioscience); CD62L (MEL-14, eBioscience); CD4 (GK1.5, BD Biosciences and Miltenyi Biotec); or appropriate isotype Abs. Intranuclear FOXP3 staining was performed using eBioscience PE- or APC-conjugated FOXP3 staining buffer set (FJK-16 s). Intranuclear Ki-67 staining was performed using BD Pharmingen Ki-67 (B56) on permeabilized cells. Six-colour flow cytometry was performed with a FACSCanto cytometer (BD Biosciences). Data files were analysed using FlowJo software (Tree star Inc.).