A Quantitative Trait Locus on chr.4 Regulates Thymic Involution

Abstract

The mechanisms underlying age-associated thymic involution are unknown. In mice, thymic involution shows mouse strain–dependent genetic variation. Identification of the underlying genes would provide mechanistic insight into this elusive process. We previously showed that responsiveness of hematopoietic stem and progenitor cells (HSPCs) to transforming growth factor-beta 2, a positive regulator of HSPC proliferation, is regulated by a quantitative trait locus (QTL) on chr. 4, Tb2r1. Interestingly, Tgfb2^+/− mice have delayed thymic involution. Therefore, we tested the hypothesis that a QTL on chr. 4 might regulate thymic involution. Aged, but not young, B6.D2-chr.4 congenic mice, where the telomeric region of chr. 4 was introgressed from DBA/2 to C57BL/6 mice, had larger thymi, and better maintenance of early thymic precursors than C57BL/6 control mice. These observations unequivocally demonstrate that the telomeric region of chr. 4 contains a QTL, Ti1 (thymic involution 1) that regulates thymic involution, and suggest the possibility that Ti1 may be identical to Tb2r1.

INTRODUCTION

T lymphopoiesis takes place in the thymus from cells that seed this organ from the bone marrow. Before and during adolescence, the thymus begins to involute, and the production of naive T cells decreases, although some level of production of naive T cells has been detected in humans in the sixth decade of life (1,2). The mechanism and the purpose of thymic involution remains a mystery. Although thymic involution in young individuals is likely under evolutionary selection and therefore probably beneficial, further thymic involution in old individuals may be detrimental to health, and perhaps to longevity (2–4). The timing of the beginning of thymic involution, just before and during adolescence, as well as experimental evidence suggested that sex hormones or other endocrine mechanisms may play a role (5). However, at least in humans, thymic involution has been shown to begin in early childhood (6). Furthermore, recent studies have indicated however that neither sex hormones nor an age-related decrease in the activity of the growth hormone/insulin-like growth factor axis appear to play a major role in thymic involution (7,8). Although age-related changes in the epithelial component of the thymus are important in thymic involution, functional changes in the hematopoietic component are likely involved as well, as both the number and the function of the earliest T-cell progenitors in the thymus, early thymic precursors (ETPs), decline with age, suggesting an intrinsic role for ETPs in thymic involution (9).

A potentially powerful approach to unravel novel regulatory mechanisms involved in biological processes is quantitative trait analysis. Quantitative traits are traits that vary continuously across genetically different individuals and are inherited in a non-mendelian fashion because of the contribution of allelic variation at multiple loci to the phenotype. These loci are called quantitative trait loci (QTL; 10–12). Phenotypically detectable quantitative genetic variation is typically subtle as detrimental alleles that arise through random genetic variation will experience negative selection and will disappear from the population or become very rare, while highly beneficial alleles will be selected for and will increase in frequency until they become fixed in the population. Although, because of selective pressure, the phenotypic effect of individual QTLs is typically small, identification of the genes underlying these QTLs may reveal novel and critical mechanisms regulating the biological process under study (11,12). Similar to many other aspects of the hematopoietic system (13–20), thymic involution shows mouse strain–dependent quantitative genetic variation (21). Two QTLs involved in thymic involution have been mapped with a suggestive level of significance, one on chr. 9 and another on chr. 10 (21). However, for neither of these two QTLs has the map location been unequivocally confirmed. We have previously shown that the responsiveness of hematopoietic stem and progenitor cells (HSPCs) to one specific isoform of transforming growth factor-beta (TGF-β), TGF-β2, is a quantitative trait (17). In contrast to other TGF-β isoforms, TGF-β2 enhances HSPC proliferation in vitro and in vivo (17). Genetic variation in the responsiveness of HSPCs to this isoform-specific proliferative effect of TGF-β2 mapped to a locus on the telomeric region of chr.4, Tb2r1 (TGF-β2 responsiveness 1; 17,22). In addition to defects in hematopoiesis, Tgfb2^+/− mice displayed delayed thymic involution, which could be accelerated by transplantation of wild type bone marrow, suggesting an enhancing role for TGF-β2 expressed within the hematopoietic system in this process (23). As responsiveness to TGF-β2 is regulated by a QTL on chr.4, Tb2r1, these data raise the possibility that Tb2r1 may also regulate thymic involution. We show here that a QTL on chr.4, possibly Tb2r1, indeed regulates thymic involution.

METHODS

Mice

Four- to six-week-old C57BL/6J mice were purchased from the National Cancer Institute (Bethesda, MD). B6.D2-chr.4 congenic mice have been described previously (22). C57BL/6J and B6.D2-chr.4 mice were bred and aged in the animal facility of Mount Sinai School of Medicine. Males and females were housed separately, whereas genotypes were mixed. Animals were housed in a specific pathogen-free facility. Experiments and animal care were performed in accordance with the Mount Sinai Institutional Animal Care and Use Committee.

Antibodies and Flow Cytometry

Fluorescein isothiocyanate (FITC)-conjugated CD3ϵ, CD8α, B220, Ly6G/Gr1, Mac1, Ter119, TCRαβ, NK1.1, and CD62L, and phycoerythrin (PE)-CY7-conjugated IL7Rα were purchased from eBiosciences (San Diego, CA). PE-conjugated CD25, allophycocyanin (APC)-CY7-conjugated CD8, PE-CY7-conjugated CD4, FITC-conjugated CD8β and TCRγδ, APC-conjugated c-kit, and CD44 were obtained from Pharmingen (San Diego, CA). Flow cytometric analysis was performed on a three-laser LSRII or a Special Order five-laser LSRII with DiVa software (Beckon Dickinson, Mountain View, CA). Data were analyzed using FlowJo software, Tree Star, Inc., Ashland, OR.

Preparation of Hematopoietic Cells

Thymi were dissected, weighed, and minced through nylon mesh with cold Dulbecco's Modified Eagle's Medium (Cellgro, Manassas, VA) containing 2 % fetal calf serum and 100 ng/ml penicillin/streptomycin. Red blood cells (RBC) were lysed using RBC lysis (eBiosciences). Cells were counted using a hemocytometer.

Statistics

One-sample t test with an expected value of 1 was used. Data were paired per cage containing two to three B6.D2-chr.4 congenic and two or three C57BL/6 mice. The averages of all mice of the same genotype in one cage were considered one data point and normalized to the average value for C57BL/6 controls in the same cage. One-tailed tests were used as we hypothesized that thymic weight, cellularity and naive T-cell fraction would be higher in B6.D2-chr.4 congenic than in C57BL/6 mice.

RESULTS

We have mapped a QTL, Tb2r1, regulating the responsiveness of HSPCs to the proliferative effects of TGF-β2 to the telomeric region of chr.4, and demonstrated the veracity of Tb2r1 using congenic mice (22). These congenic mice were constructed by introgressing 20 cM of the telomeric region of chr.4 from DBA/2 into C57BL/6 mice by repeated backcrossing and selection of offspring where the telomeric region of chr.4 was heterozygous (24,25). Intercrossing of these mice yielded congenic mice (referred to as B6.D2-chr.4 mice hereafter) where the whole genome is of C57LB/6 origin, except for the telomeric 20 cM, which are of DBA/2 origin (Figure 1). Therefore, any difference between the phenotype of B6.D2-chr.4 and of C57BL/6 mice must be caused by one or more DBA/2 alleles in the introgressed region of chr.4. We have previously shown that in B6.D2-chr.4 mice the enhancing effect of TGF-β2 on HSPC proliferation, observed in parental C57BL/6 mice, was virtually absent. These findings unequivocally mapped Tb2r1 to the introgressed region of chr. 4 (22). In addition, and similar to Tgfb2^{+ /−} mice, B6.D2.chr.4 mice had fewer CD34⁻Flt3⁻LSK (long-term hematopoletic stem cell (HSC)), CD34⁺Flt3⁻LSK (short-term HSC), and CD34⁺Flt3⁺LSK (multipotential progenitor) cells; more slowly cycling progenitor cells; and increased resistance to a lethal dose of the cytotoxic drug 5-fluorouracil (22). Hence, B6.D2-chr.4 mice phenocopy Tgfb2^{+ /−} mice for multiple hematopoietic traits, suggesting that the gene underlying Tb2r1 regulates the HSPC compartment through regulating TGF-β2 responsiveness of HSPCs (23). In addition to the aforementioned traits, we have shown previously that Tgfb2^+/− mice also have delayed thymic involution, indicating a role for TGF-β2 in enhancing thymic involution (23). Therefore, we examined whether B6.D2-chr.4 mice would also phenocopy Tgfb2^+/− mice for this trait. To this end, we analyzed thymic involution in a cohort of aging B6.D2-chr.4 mice and C57BL/6 control mice. We hypothesized that, similar to Tgfb2^+/− mice, B6.D2-chr.4 mice would have delayed thymic involution.

Figure 1.

B6.D2-chr.4 congenic mice. Schematic representation of the genetic make-up of B6.D2-chr.4 mice. Except for the terminal region of chr. 4 (20 cM) that is DBA/2 origin, the whole genome is of C57BL/6 origin.

B6.D2-chr.4 and parental C57BL/6 mice were aged to the age of 18–26 months. As stress conditions within a cage can affect thymic involution (8), age-matched congenic and control mice were housed together in the same cage. To assess the extent of thymic involution, we measured thymic cellularity and weight. Because of large intercage variability, the average thymus size of mice of the same genotype within one cage was considered as one data point, and paired analysis was performed on a per cage basis, after normalization to the value for the C57BL/6 controls in the cage. The large intercage variability is likely explained by the fact that the age of the mouse cohort ranged from 18 to 26 months at the time of analysis, and by cage-specific stress-dependent endocrine effects, which have been described (8). Using this experimental design, it was possible to detect a quantitatively relatively subtle phenotype, which, as explained in the Introduction, is characteristic for quantitative traits. In addition, we attempted to isolate the phenotypic effect of one QTL. Each individual QTL typically contributes only a small fraction of the observed quantitative variation (10–12). Finally, as we tested the hypothesis that thymus size would be larger in B6.D2-chr.4 than in C57BL/6 mice, we used a one-tailed t test for statistical analysis. At the age of 8 weeks, there was no difference in thymic size between B6.D2-chr.4 and C57BL/6 mice (Figure 2). As expected, thymus cellularity and weight decreased steeply with age in both C57BL/6 (from 76.8 ± 9.8 to 6.1 ± 3 × 10⁶ cells and from 81.3 ± 5.5 to 36.0 ± 7.8 mg) and B6.D2-chr.4 mice (from 78.9 ± 2.8 to 9.6 ± 4.8 × 10⁶ cells and from 80.0 ± 9.1 to 40.7 ± 2.8 mg; n = 3 cages for young (8 weeks) and n = 10 cages for old mice [18 to 26 months]). When analyzed per cage, however, thymic cellularity (p = .045) and weight (p = 0.025) were significantly higher in old B6.D2-chr.4 mice compared with control C57BL/6 mice from the same cage (Figure 2). Thus, thymic involution proceeds more slowly in B6.D2-chr.4 mice than in parental C57BL/6 mice aged in the same conditions. As any difference between the phenotype of B6.D2-chr.4 and of C57BL/6 mice must be caused by allelic variation at one of the approximately 174 genes in the introgressed region of chr.4 (22), these data demonstrate that a QTL on the telomeric region of chr.4 plays a role in the regulation of thymic involution.

Figure 2.

Thymic involution in B6.D2-chr.4 mice. Thymic cellularity and weight in young (8 weeks) and old (18–26 months) B6.D2.-chr.4 congenic mice compared and normalized with C57LB/6 mice kept in the same cage (n = 3 cages for young mice, n = 10 cages for old mice [40 mice in total], *significantly different from 1, one-sample, one-tailed t test with expected value = 1).

The aged immune system is characterized by, among others, a decreased pool of naive T cells leading to a decline in the capacity of aged individuals to mount immune responses to new pathogens (2–4). This decline is believed to be due primarily to decreased de novo production of T cells by the thymus. Through peripheral expansion, memory cells fill the void in the T-cell pool caused by the decreased production of naive T cells in aged individuals (2–4,26), although this shift to a memory phenotype may in addition at least in part be antigen dependent (27). Therefore, we measured the fraction of naive CD4 and CD8 T cells in spleen and peripheral blood, hypothesizing that B6.D2-chr.4 mice would, given their larger thymus size, have a larger fraction of naive T cells in spleen and peripheral blood. As shown in Figure 3, the fraction of naive CD4 and CD8 T cells, defined by the CD62L^hiCD44^lo phenotype (28; Figure 3A) was more than twofold higher in B6.D2-chr.4 than in C57BL/6 mice kept in the same cage (Figure 3B). Except for peripheral blood CD8 cells (p = .03, one-tailed t test), these differences, though consistent, did not reach statistical significance, however. These data indicate that thymic size has an impact on the fraction of naive T cells in the periphery, but that either additional factors, such as random exposure to antigen (28), may play a role in individual mice, or that the biological variability in this measure is intrinsically large.

Figure 3.

Naive T cells in old B6.D2-chr.4 and C57BL/6 mice. (A) Representative analysis gates for naïve CD4 and CD8 T cells in spleen. Within the CD4 and CD8 fractions, naive T cells are CD62L^hiCD44^lo (B) Fraction of naive CD4 and CD8 T cells spleen (SP) and peripheral blood (PB) in B6.D2-chr.4 relative to the values in age-matched C57BL/6 mice kept in the same cage (n = 10 cages, *significantly different from 1, one-sample, one-tailed t test with expected value = 1).

Next, we analyzed T-cell development in the thymi of old and young B6.D2-chr.4 and C57BL/6 mice. T-cell development, as measured by the fractions of developing T cells (TN1 [triple negative 1], TN2, TN3, TN4, double-positive CD4⁺CD8⁺ cells and CD4⁺ and CD8⁺ single-positive cells) were similar in both mouse strains (not shown). It has been shown that number and function of ETPs decline with age, suggesting an intrinsic role for ETPs in thymic involution (9). Therefore, we compared the absolute number of ETPs. A striking difference between aged C57BL/6 and B6.D2-chr.4 mice was found in the ETP compartment, defined by the lin⁻ (lineage markers were CD8α, CD8β, TCRαβ, CD3, CD11b, NK1.1, B220, CD19, B220, Gr1, and Ter119) CD25⁻c-kit⁺Il7Rα⁻ phenotype (29; Figure 4A). In young mice, the frequency and the number of ETPs was 50-fold lower in B6.D2-chr.4 mice than in C57BL/6 mice (Figure 4B), a finding consistent with our previous observation that the number of the most primitive HSPCs was much lower in B6.D2-chr.4 mice than in C57BL/6 mice (22). However, in aged mice, ETP frequency and number were similar in B6.D2-chr.4 and C57BL/6 mice (Figure 4B). Compared with young mice, the number of ETPs declined 20-fold in C57BL/6 mice, but only twofold in B6.D2-chr.4 mice. Thus, our findings indicate that one or more QTL in the telomeric region of chr.4 affects thymic involution and suggest that this (these) QTL predominantly affect(s) the ETP compartment.

Figure 4.

Early thymic precursors in old and young B6.D2-chr.4 and C57BL/6 mice. (A) Representative examples of the analysis gates and of the frequency of ETPs in thymi from young (8 weeks) and old (18–26 months) mice. (B) Absolute number of ETPs in thymi from young (8 weeks) and old (18–26 months) mice. Data were averaged per cage, n = 3 cages for young mice, n = 10 for old mice.

DISCUSSION

Quantitative trait analysis is a powerful strategy to uncover novel regulatory mechanism that would have been difficult to discover otherwise. As genes and pathways that show quantitative genetic variation in mice often also do so in humans (12,30–34), it is possible that similar variation exists in humans and that the underlying mechanism is the same. Within the hematopoietic system, only one quantitative trait gene has been identified thus far, latexin, which regulates HSC pool size (35). This study is the first to unequivocally map a QTL regulating thymic involution. Confirmation of the location of a QTL in appropriate congenic mice is the first step in positional cloning of the underlying gene, which may eventually shed light on the mechanisms underlying thymic involution, a process that is not understood thus far but is widely believed to detrimental to organismal health (2–4).

Thymic involution is subject to quantitative genetic variation, and two QTLs have been mapped on chr. 9 and 10 (21). However, for none of these QTL, which were identified with a suggestive level of statistical significance, was the veracity confirmed in appropriate congenic mice. Here we show unequivocally that one or more QTL in the telomeric region of chr.4 play a role in thymic involution, and that the DBA/2 allele of this QTL confers a slower rate of thymic involution, at least when introgressed into the C57BL/6 background. We call this QTL Ti1 (thymic involution 1). As B6.D2-chr.4 mice phenocopy many of the hematopoietic traits of Tgfb2^+/− mice, it is possible that genetic variation in TGF-β2 responsiveness is involved in the delayed thymic involution of B6.D2-chr.4 mice. Furthermore, as Tgfb2^+/− mice also display delayed thymic involution, it stands to reason to hypothesize that the same gene underlies Ti1 and Tb2r1. If Ti1 and Tb2r1 are identical, then it is likely that Tb2r1 is involved in regulating thymic involution through its effect on TGF-β2 signaling, as Tgfb2^+/− mice have delayed thymic involution (23). It cannot be excluded, however, that distinct QTLs in the introgressed region of chr.4, which contains approximately 174 genes (23), play a role as well.

Thymic involution is considered to be primarily caused by age-related changes in the thymic epithelial component and by the aged systemic environment (1–5). However, it was recently shown that the hematopoietic component of the thymus, and in particular the earliest precursors in the thymus, the ETPs, decline in number and in function with age (9). Further evidence for a hematopoietic component in the mechanisms underlying thymic involution was also provided by our observation that increasing TGF-β2 expression in hematopoietic cells was sufficient to accelerate thymic involution in Tgfb2^+/− mice (23). Therefore, although the mechanism of this effect is unclear, Tb2r1 may predominantly affect the hematopoietic component of the thymus. The most striking difference in the hematopoietic component of the thymus between B6.D2-chr.4 and C57BL/6 mice was in the ETP fraction. In contrast to C57LB/6 mice, ETP numbers only declined to a limited extent with age in B6.D2-chr.4 mice. A similar, though less pronounced, maintenance of ETPs was observed in Tgfb2^+/− mice (23). We hypothesized (23) that the decreased cycling of HSCs in Tgfb2^+/− mice (17) leads to delayed exhaustion and better maintenance of lymphoid potential. The same may be true for B6.D2-chr.4 mice, which also have a more slowly cycling stem and progenitor compartment (22). It should be noted, however, that despite relative maintenance of ETPs in B6.D2-chr.4 mice, thymic involution was still extensive. Furthermore, at a young age, thymic size of C57BL/6 mice was similar to that of B6.D2-chr.4 mice despite the higher content in ETP in the former than in the latter. These observations suggest that other factors, such as the intrinsic characteristics of the thymic epithelial cells, the number and function of other accessory cells, and the hormonal milieu, have a major impact on thymic size and function.

We also mapped a suggestive life span QTL to this region of chr.4 (36). It has been suggested that improving thymic function in the elderly individuals will increase health, and perhaps extend life span (3,4). The aged immune system is characterized by a decreased pool of naive T cells, leading to a decline in the capacity of aged individuals to mount immune responses to new pathogens. Accumulation of senescent memory cells, likely through homeostatic proliferation, fills the void in the peripheral T-cell pool (2,3). In addition, it has been shown that in conditions of lymphopenia, naive T cells undergo homeostatic proliferation and acquire a memory phenotype (37). It is possible that this process also occurs during thymic involution. Furthermore, antigen exposure has been shown to play a role in the progressive age-related bias to a memory phenotype (27). By virtue of their constitutive production of inflammatory cytokines, which have been linked to diseases such as osteoporosis and to the general state of low-level inflammation and frailty that characterizes aging (‘inflammaging’) in humans (3), accumulation of memory cells may contribute to the aging process. The rate of thymic involution in aged individuals may therefore affect longevity, both through its effect on immune responses to neoantigens and through indirect effects on other organs systems. The observation that the frequency of memory cells in the peripheral blood of aged mice is inversely correlated with life span suggests that maintaining high levels of naive cells with age through delayed thymic involution may increase life span (38). Given our identification of a suggestive life span QTL in the telomeric region of chr.4 (36), it is tempting to speculate that Ti1 may affect longevity through its effect on thymic involution. Alternatively, it is possible that this QTL in fact affects more fundamental and general aspects of the aging process, and that delayed thymic involution in B6.D2-chr.4 mice is a biomarker of delayed aging.

In conclusion, we have unequivocally demonstrated that a QTL, Ti1, on chr.4 plays a role in the regulation of thymic involution.

FUNDING

Supported by a grant from the NIH RO1 AG016327 and NIH RO1 HL073760.