Published online before print April 1, 2009
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Laboratory of Neuroendocrine-Immunology,
Division of Experimental Obesity, Pennington Biomedical Research Center, Louisiana State University System, Baton Rouge, Louisiana, USA;
Huffington Center on Aging, Baylor College of Medicine, Houston, Texas, USA; and
Radiology Department, University of Michigan, Ann Arbor, Michigan, USA
1 Correspondence: Laboratory of Neuroendocrine-Immunology, Pennington Biomedical Research Center, Louisiana State University System, 6400 Perkins Rd., Baton Rouge, LA 70808, USA. E-mail: Vishwa.dixit{at}pbrc.edu
ABSTRACT
The adipocytes are the predominant cell types that constitute the bulk of the thymic microenvironment by the fifth decade of life in healthy humans. An age-related increase in thymic adiposity is associated with reduced thymopoiesis and compromised immune surveillance in the elderly. However, the mechanisms regulating the generation of intrathymic adipocytes during aging remain to be elucidated. Here, we report that the CD45– thymic stromal cells (TSCs) are amenable to adipogenesis. We identified that the Wnt inhibitor axin is expressed in the lymphoid as well as stromal cells of the thymus with increased expression in CD45– TSCs of older mice. Knockdown of axin by RNA interference in CD45– primary TSCs led to a marked reduction in adipogenesis with significantly lower expression of adipogenic transcripts peroxisome proliferator-activated receptor 
2 (PPAR), adipocyte fatty acid-binding protein (aP2), and perilipin. Age-related elevated axin expression was increased specifically in thymic fibroblasts and medullary thymic epithelial cells (TECs) but not in the cortical TEC or CD45+ cells. Consistent with a role of axin in promoting thymic adipogenesis, axin expression was also colocalized with lipid-expressing adipogenic cells in aging thymus. The prolongevity intervention, caloric restriction (CR), prevented the age-related increase in axin and the adipogenic cell in the thymus together with increase in thymic output. We have recently demonstrated that CR induces ghrelin, which can partially reverse thymic involution. Here, we show that axin expression is not affected by ablation of ghrelin receptors in aging mice, suggesting a ghrelin-independent mechanism for regulation of axin. Our data are consistent with the hypothesis that blocking the specific proadipogenic signals in the thymus may complement the present approaches to rejuvenate thymic function during aging.
Key Words: aging • Wnt • adipogenesis • GH • adipocyte • dietary restriction • obesity
INTRODUCTION
Deterioration of immune function with progressive aging is linked in large part to the reduced ability of thymus to produce naïve T cells [1 ]. The thymic involution process can be detected post-puberty and is associated with changes in the production of sex-steroid hormones [2 ]. It is presently unclear why thymic involution is initiated early on in life, but this maladaptive response has severe consequence for the immune responsiveness with progressive age [3 ]. Apart from circulating neuroendocrine factors regulating thymic function [4 ], the alteration in the thymic microenvironment [5 6 7 ], changes in intrathymic growth factors [8 , 9 ], loss of lymphoid progenitors [10 ], and epithelial cell populations [11 ] are considered to play important roles in reduced thymic output with age.
The unique three-dimensional thymic meshwork is comprised of the cortex and medulla, composed mainly of distinct developing T cell subsets and thymic stromal cells (TSCs), which are a diverse group of cells including cortical thymic epithelial cells (cTECs), medullary TECs (mTECs), mesenchymal cells, fibroblasts, pericytes, endothelial cells, and adipocytes [12 ]. One of the hallmark changes in the thymic microenvironment with aging is a progressive increase in adipocytes in the thymic parenchyma, septa, cortex, and medulla with replacement of epithelial and T-lymphopoietic thymic zones into adipose tissue [3 , 13 14 15 ]. The mechanisms responsible for the origin of adipocytes in thymus and their progressive deposition, such that they replace virtually all of the thymopoietic microenvironment with age, remain to be elucidated.
The Wnt pathway has been implicated in lineage-commitment and cell-fate regulation during development and aging [16 , 17 ]. Axin is an integral component of the Wnt pathway and a product of the mouse-fused locus, which is known to be a negative regulator of Wnt signaling [18 ]. Axin is ubiquitously expressed during embryogenesis, and consistent with a integral role of axin in Wnt signaling machinery, the embryos lacking axin die at Embryonic Day 9.5 as a result of severe abnormalities in neural tube formation, forebrain development, and axis duplication [19 ]. Wnts are a family of paracrine and autocrine factors known to play a critical role in cell growth and differentiation, tissue morphogenesis, and cancer [20 ]. It appears that a central function of axin is to promote the phosphorylation of β-catenin by glycogen synthase kinase 3 (GSK-3) [21 ]. According to the prevailing view of canonical Wnt signaling, in the absence of Wnt signals, GSK-3 phosphorylates axin, leading to the ubiquitination and degradation of cytosolic β-catenin [22 ]. In the presence of Wnt signals, GSK-3 activity is inhibited, causing hypophosphorylation of axin and preventing β-catenin from the degradation, allowing its accumulation and translocation to the nucleus to initiate transcription of target genes [21 ]. In contrast, noncanonical Wnt signaling functions in a β-catenin-independent manner via extracellular Wnt ligands and their interactions with frizzled receptors [23 ]. These interactions cause activation of specific G proteins and phospholipase C and turnover and release of intracellular calcium ions [22 ]. The canonical [24 ] and noncanonical Wnt pathway [25 ] has been shown to play an integral role in T cell development in the thymus [26 , 27 ].
We have demonstrated recently that ghrelin, a stomach-derived anabolic peptide, can promote thymopoiesis during aging [4 ]. Interestingly, during the course of these studies, we observed that the ghrelin infusions reduced the neutral lipid-expressing adipocytes in the aging thymus, and ghrelin receptor (GHSR) knockout (KO) mice developed excess thymic adipocytes and exhibited reduced thymic output [4 ]. Here, we investigated the mechanism responsible for thymic involution with focus on specific adipogenic regulators that may provide inductive signals required for adipocyte development in the thymus. The focus of many of the current studies about Wnt proteins in thymocyte development and hematopoietic stem cells has led to identification of the critical function of these proteins in the thymus [28 ]. Considering a vital role played by Wnt inhibitors in aging and cellular differentiation, we hypothesized that the axin may regulate adipocyte development within the thymus and contribute to age-related thymic involution. Using in vitro and in vivo approaches, we provide evidence that the axin regulates adipogenic programming of the TSC, and an age-related increase in the thymic axin and reduced T cell output can be prevented by prolongevity intervention, caloric restriction (CR). Our data suggest that increased expression of the axin in the aging thymus may be one of the molecular triggers involved in regulating the process of age-related thymic adiposity.
MATERIALS AND METHODS
Thymic magnetic resonance imaging (MRI)
MRI examinations were performed on healthy subjects between 40 and 45 years of age (n=12) using Institutional Review Board-approved protocols with informed consent. The imaging was performed with subjects’ head first, supine in a 3T full-body scanner (Signa, General Electric Medical Systems, Milwaukee, WI, USA). A body-surface receive-only coil and a full-body transmit coil were used to scan at the level of tracheal bifurcation. T1-weighted spin echo axial images were performed using the following parameters: repetition time/echo time (TR/TE), 500/10 ms; slice thickness, 4 mm; field of view (FOV), 32 x 32 cm; matrix, 320 x 320 (resolution, 1x1 mm); 10 slices; and slice gap, 1 mm. T2-weighted fast-spin echo axial images were done using the following parameters: TR/TE, 3000/80 ms; echo train, 8; slice thickness, 4 mm; FOV, 32 x 32 cm; matrix, 320 x 320; 10 slices; and slice gap, 1 mm. Total acquisition times for the T1-weighted and T2-weighted images were 
5 min and 2 min, respectively; total examination time was 7 min. To minimize image artifacts and increase thymus contrast, modification of both sequences may be necessary at the time of examination.
Mice
The female CR (n=20) mice and ad libitum (AL)-fed (n=20) 3-, 10-, 12-, and 18-month-old C57/B6 mice were purchased from a National Institute on Aging aging rodent colony (Harlan Sprague-Dawley, Indianapolis, IN, USA). CR was initiated at 14 weeks of age at 10% restriction, increased to 25% restriction at 15 weeks, and to 40% restriction at 16 weeks, where it is maintained throughout the life of the animal. Mice were maintained on the CR diet during transit and sacrificed after 1 day’s rest. We also used 3- to 4-month-old female C57BL/6 mice as controls. The mice were maintained under pathogen-free conditions in the animal facility at the Pennington Biomedical Research Center (PBRC; Baton Rouge, LA, USA) using protocols approved by the PBRC Animal Care and Use Committee. The GHSR mice have been described previously [4
]. The GHSR null mice have been backcrossed to C57B6 background for at least eight generations. A total of six to eight age-matched female wild-type (WT) and GHSR null mice was used for these experiments.
TSC culture and transfections
Thymic lobes were minced into small fragments and treated for 1 h at 37ºC with an enzymatic mixture containing 1 mg/ml collagenase (Sigma Chemical Co., St. Louis, MO, USA), 0.2 mg/ml DNAse type I (Sigma Chemical Co.), and 1 mg/ml trypsin EDTA (Gibco, Grand Island, NY, USA) in PBS. The CD45 cells were depleted using magnetic bead-based positive selection as described previously [2
, 11
]. The CD45-negative fractions were resuspended, and 1 x 106 cells/well were seeded in six-well or 24-well culture plates and cultured for 14 days in DMEM nutrient F12 (DMEM/F12; with 15 mM HEPES, NaHCO3, L-glutamine, Gibco), supplemented with 3 µg/ml insulin, 20 ng/ml epidermal growth factor (EGF), 100 U/ml penicillin-streptomycin, and 20% FBS. Cultures were maintained at 37ºC and 5% CO2. After cells reached 70% confluence, the TSCs were transfected by control and axin-specific small interfering (si)RNA using a lipofectamine-based kit based on the manufacturer’s recommendation (Dharmacon, Chicago, IL, USA). Adipogenesis was induced by changing the medium to DMEM/F12 containing 10% FBS, 0.5 mM 3-isobutylmethylxanthine, 1 µM dexamethasone, and 1.7 µM insulin [multiple daily injections (MDI)].
Flow cytometry
The thymic fragments were digested enzymatically, and cells were labeled with CD45, anti-MHCII, Ly51, platelet-derived growth factor receptor 
(PDGF-R; eBioscience, San Diego, CA, USA), MTS15, and isotype control IgG with appropriate secondary-conjugated antibodies. For intracellular axin staining, cells were fixed by paraformaldehyde and permeabilized using saponin. After blocking for nonspecific-binding sites, the cells were stained using secondary Alexa Fluor 488 (Invitrogen, Carlsbad, CA, USA) and allophycocyanin (APC)-conjugated anti-CD45 antibody. FACS sorting of TSC was performed on FACSAria as described previously [2
]. The FACS analysis was performed on FACSCalibur, and data were analyzed by post-collection compensation using FlowJo (Tree Star Inc., Ashland, OR, USA) software.
Immunofluorescence staining
The whole thymi obtained from mice were stored in 2% sucrose in PBS overnight at 4ºC. The tissues were embedded in OCT compound and frozen on a cryo-microtome specimen holder. Thick frozen sections (0.5–0.7 µm) were cut in a cryo-microtome, and at least three serial sections were used for each staining. Tissue sections were then fixed with cold ethanol and 4% buffered paraformaldehyde, and then nonspecific binding sites were blocked with protein-blocking buffer. After overnight incubations with unconjugated rabbit polyclonal anti-human axin (H-98, Santa Cruz Biotechnology, Santa Cruz, CA, USA), unconjugated rat antibody to ERTR7 (HM1086, Cell Sciences, Canton, MA, USA), or unconjugated rabbit antibody to fibroblast-specific protein-1 (FSP-1; ab27957, Abcam, Cambridge, MA, USA), the sections were labeled with Alexa Fluor 594-conjugated polyclonal chicken anti-rabbit IgG secondary antibody or Alexa Fluor 488-conjugated polyclonal donkey anti-rat IgG (Invitrogen). In addition, for the visualization of lipid droplet and adipocyte populations, frozen thymic sections and cultured cells were fixed with 4% buffered paraformaldehyde and then stained with LipidTOX Green (Invitrogen) for 20 min. Negative controls, as obtained by occulting the primary antibody or by using an unrelated IgG, displayed no specific labeling. Fluorescence mounting solution (Vector, Burlingame, CA, USA) was applied to slides and observed with a Zeiss Axioplan 2 microsocpe.
Western blot analysis
Thymi were lysed in radioimmune precipitation assay buffer supplemented with a protease and phosphatase inhibitor mixture (Sigma Chemical Co.), and protein concentrations of cell lysates were determined. The lysates were diluted with sample buffer, separated on 4–20% Tris-HCl/SDS-polyacrylamide gels (Novex, Wadsworth, OH, USA), and electrophoretically transferred to nitrocellulose membranes (Invitrogen). The blots were then incubated with rabbit polyclonal anti-human axin (H-98, Santa Cruz Biotechnology) antibody, rabbit polyclonal anti-mouse β-catenin or goat polyclonal anti-mouse Wnt-10b (N-19, Santa Cruz Biotechnology), and actin-specific mouse mAb (Sigma Chemical Co.). Immune complexes were visualized by incubation with specific secondary antibodies conjugated to Alexa Fluor 680 or IRDye® 800CW-conjugated affinity-purified anti-mouse IgG (Rockland Immnunochemicals, Gilbertsville, PA, USA), and then membranes were imaged for fluorescence on an Odyssey infrared imaging system (LI-COR Biosciences, Lincoln, NE, USA). The quantification of Western blots was performed using ImageJ software from the National Institutes of Health (NIH; http://rsb.info.nih.gov/ij/index.html).
Real-time RT-PCR
Total RNA was prepared with RNAzol (Iso-Tex Diagnostics, Friendswood, TX, USA). The cDNA synthesis and real-time RT-PCR were performed as described previously [4
]. Real-time RT-PCR analyses were done in duplicate on the ABI PRISM 7900 sequence detector TaqMan system with the SYBR Green PCR kit as instructed by the manufacturer (Applied Biosystems, Foster City, CA, USA). The following primer pairs were used: GAPDH forward (5'-TTGATGGCAACAATCTCCAC-3') and reverse (5'-CGTCCCGTAGACAAAATGGT-3'), axin forward (5'-CCCCCATACAGGATCCGTAA-3') and reverse (5'-GGTACCCGCCCATTGACTT-3'), peroxisome proliferator-activated receptor (PPAR) forward (5'-ACAAGACTACCCTTTACTGAAATTACCAT-3') and reverse (5'-TGCGAGTGGTCTTCCATCAC-3'), PPAR2 forward (5'-GCCTATGAGCACTTCACAAGAAATT-3') and reverse (5'-TGCGAGTGGTCTTCCATCAC-3'), adipocyte fatty acid-binding protein (aP2) forward (5'-GCGTGGAATTCGATGAAATCA-3') and reverse (5'-CCCGCCATCTAGGGTTATGA-3'), perilipin forward (5'-GACACCACCTGCATGGCT-3') and reverse (5'-TGAAGCAGGGCCACTCTC-3'), Naked 1 (Nkd1) forward (5'-GGGAAACTTCACTCGAAGCC-3') and reverse (5'-ACTCCTCGATGCCTTTCCTT-3'), preadipocyte factor 1 (Pref-1) forward (5'-TTCGGCCACAGCACCTATG-3') and reverse (5'-ACATTGTCAGCCTCGCAGAA-3'), C/EBP homologous protein (CHOP) forward (5'-CAGCGACAGAGCCAGAATAA-3') and reverse (5'-GACCAGGTTCTGCTTTCAGG-3'), Kruppel-like factor 6 (KLF6) forward (5'-CACGAAACGGGCTACTTCTC-3') and reverse (5'-ACACGTAGCAGGGCTCACTC-3'), and IL-7 forward (5'-GGGAGTGATTATGGGTGGTGAG-3') and reverse (5'-TGCGGGAGGTGGGTGTAG-3').
Quantification of signal joint TCR excision circles (sjTRECs)
CD4+ T subsets were isolated from splenocytes using the mouse CD4+ T cell positive section kit (Invitrogen). The sorted cells were lysed in 100 mg/L proteinase K (Sigma Chemical Co.) for 1 h at 56°C followed by 10 min at 95°C. The amount of TRECs in 5 x 106 cells was determined by real-time quantitative PCR using the ABI PRISM 7900 sequence detector TaqMan system (Applied Biosystems). The PCR performed with m
Rec- and 
J-specific primers and mRec-J fluorescent probe as described previously [4
]. The standard curves for murine were generated by using Rec and J TREC PCR product cloned into a pCR-XL-TOPO plasmid, a generous gift from Dr. Gregory Sempowski, Duke University Medical Center, Durham, NC, USA.
Statistical analysis
The values given in this article are presented as mean ± SEM. All analyses were performed by ANOVA followed by a Fisher’s least significant difference test using Sigma Stat software. P < 0.05 was viewed as statistically significant.
RESULTS
Thymic adiposity in healthy humans
The histological analysis of human thymus shows a marked increase in adipocytes with increasing age within the thymic microenvironment, suggesting a potentially prominent role played by fat cells in the age-related thymic involution process [13
, 14
]. Given that the lymphoid and adipose tissue can be distinguished from each other as a result of differences in water content and proton relaxation, we sought to investigate the phenomenon of adipocyte accumulation in 40- to 45-year-old healthy human (n=12) thymi using MRI. Two image modalities were used for thymus volume measurements in this study: T1-weighting and T2-weighting. These modalities provide ample tissue contrast and require short acquisition times. Given that the T1 relaxation times for water is much longer than that of fat, MRI can analyze the presence of lipid within the thymus.
As evident from the images in Figure 1A , our thymic MRI protocol is sensitive to detect the changes in the thymic volumes and fat infiltration. The region of interest (ROI) depicting the thymus is highlighted in red. The images were acquired at the level of tracheal bifurcation and show trapezoidal-shaped thymus located anterior to the sternum and posterior to the ascending aorta. The T1 of fat is much shorter than water, resulting in less saturation and more signal from fat in a T1-weighted sequence (Fig. 1B) . A hyperintense region within the thymus is attributable to fat accumulation, and thymic remnants are visible as a pale area within the ROI, which was selected based on the criteria published previously [29 ]. Analysis of cryosections of 18-month-old murine thymi demonstrated the presence of lipid-expressing cells; interestingly, we noted that the Wnt regulator axin was colocalized with adipogenic cells in the aging thymus (Fig. 1C) .
 View larger version (56K): [in a new window] |
Figure 1. Adipocyte accumulation in the thymic microenvironment. (A) The region of interest (ROI) depicting the thymus from a representative 45-year-old healthy male (n=12) is highlighted. The fat in the thymus appears as the hyperintense region highlighted with a red line within the ROI. The white outline region shows the lymphoid tissue within the thymic space. (B) The T1 and T2 spin echo axial MRI sequences above the tracheal bifurcation area. The thymus is highlighted with yellow arrows. (C) The immunoflourescent staining showing colocalization of axin (red) in lipid-expressing cells (LipidTOX, green) in the thymus of a representative 18-month-old mouse (n=4). DAPI, 4',6-Diamidino-2-phenylindole.
|
+MTS15+). The real-time PCR analysis in sorted TSC populations revealed no changes in axin expression in cTECs. Interestingly, an age-related increase in axin expression in mTEC and TF cells was detected (Fig. 2C)
.
 View larger version (23K): [in a new window] |
Figure 2. Axin expression in TSC populations. (A and B) The thymi from young (n=6) and 12-month-old (n=6) mice were digested enzymatically and labeled with Axin-AF-488 and CD45-APC. CD45–Axin+ cells (in red box) are increased significantly (P<0.05) in older mice, CD45–Axin– (blue box) decrease with age, and no change in CD45+Axin+ (green box) is detected. All data are expressed as mean ± SEM. (C) The CD45 cells were depleted from thymic digests using magnetic bead-based positive selection. The CD45– cells from six thymi, each from 2- and 12-month-old mice were pooled and used for sorting the mTEC (CD45–MHCII+Ly51–), cTEC (CD45–MHC11+Ly61+), and fibroblasts (MTS15+PDGF-R+). The comparative threshold values from triplicate real-time PCR analyses were collapsed together, and data are expressed as relative axin mRNA expression. SSC-A, Side-scatter-A.
|
 View larger version (43K): [in a new window] |
Figure 3. Axin promotes adipogenesis of TSCs. (A) Real-time PCR analysis of axin in control (Ctl) and axin siRNA-transfected primary TSC cells. (B) Forty-eight hours post-siRNA transfections, the TSCs were treated with an adipogenic MDI cocktail. The control cells differentiated efficiently into adipocytes 28 h post-MDI exposure, and axin-deficient TSC displayed reduced adipogenic transformation. (C) The real-time PCR analysis of proadipogenic transcripts PPAR, PPAR2, aP2, and perilipin (Plin) in control and axin siRNA-transfected cells. The experiments were performed in duplicate wells and repeated thrice. All of the data are expressed as mean ± SEM.
|
1 and 2 isoforms of PPAR, generated as a result of alternative splicing and promoter use, are induced upon adipogenesis and are required for the maintenance of the differentiated state of the adipocyte [32
, 33
]. In addition, aP2 and perilipin are required for adipogenesis [31
, 34
]. We found that axin-deficient TSCs displayed reduced adipogenesis and expressed significantly lower levels of PPAR, PPAR2, aP2, as well as perilipin (Fig. 3C)
. These findings demonstrate that axin regulates the adipogenesis of TSC in vitro and age-related increase in axin expression in TSC, and fibroblast may contribute to thymic adipogenesis.
Age-related increase in axin expression in TF and lipid-bearing cells is blocked by CR
The histological hallmark of thymic involution is loss of thymocyte and adipocyte accumulation in the thymic space that eventually replaces the lymphoid and stromal cell compartment with adipose tissue [13
, 14
]. Axin has been demonstrated previously to play an important role in promoting adipogenesis of the cultured 3T3-L1 fibroblast cell line [34
]. Given that Wnt signaling has a prominent role in aging and regulates cell-fate determination [35
36
37
] and our initial findings that axin is colocalized in adipogenic cells in thymus and expressed in MTS15+PDGF-R+ TF prompted us to study the axin localization further within the aging thymic microenvironment. Furthermore, we hypothesized that prolongevity intervention, CR, a state of chronic negative energy balance, will prevent axin-induced adipogenic changes in aging thymus. Compared to young mice (3 months), the thymi of 18-month-old animals showed a large increase in axin-expressing and lipid-bearing cells (Fig. 4
). Furthermore, axin was found to be co-expressed in greater frequency with neutral lipid-containing cells in thymus. Interestingly, the thymi from CR animals displayed axin and lipid expression patterns similar to a 3-month-old mouse (Fig. 4)
, suggesting that lower thymic axin expression is associated with reduced ectopic adipocyte accumulation in thymic space.
 View larger version (45K): [in a new window] |
Figure 4. Increased axin expression in the thymus is associated with thymic adiposity. The neutral lipid-expressing cells in thymic cryosections (n=4/age group) were stained for neutral lipid with LipidTOX (green) and anti-rabbit Axin antibody conjugated with Alexa Fluor 594 (red). The axin colocalizes with lipid-expressing cells in the thymus, as well as nonadipogenic thymic cells in aging mice.
|
 View larger version (31K): [in a new window] |
Figure 5. Age-related increase in axin-expressing TF is blocked by CR. (A) TF were stained with ERTR7, followed by specific secondary antibody labeled with Alexa Fluor 488 (green), and axin was visualized by Alexa Fluor 594 (red)-conjugated antibody. The elevated axin expression in 12-month-old thymus colocalized with ERTR7. Representative images (original, 200x) of thymic sections of four mice in each group are shown. (B) Axin also partially colocalizes with FSP-1-expressing TF. The negative isotype IgG control antibody displayed no specific staining in the negative control cryosections of young and old mice.
|
β emigrate from the thymus, and as TRECs are stable and do not replicate, their presence in T cells is used to distinguish the RTEs from pre-existing, naïve T cells. It should, however, be noted that analysis of TRECs as an absolute measure of thymopoiesis is imperfect, as long-lived, naïve T cell populations can still carry TREC episomal DNA circles. In addition, half of the cells that have undergone a division after rearrangement can no longer be identified as RTEs [41
]. However, mathematical models have demonstrated that decreased thymic output is reflected sufficiently in reduction of TREC levels [41
]. Indeed, the TRECs decline by fivefold from birth to 35 years of age and up to 50-fold by age 65 [41
], and in the absence of precise surface markers to identify RTEs, TREC measurement remains a surrogate marker of thymopoiesis. Consistent with these data, we observed that aging led to a significantly reduced number of TRECs (Fig. 6C)
. Interestingly, chronic CR in 10-month-old mice blocked the age-related decline in thymopoiesis and maintained the TREC counts to the levels observed in young 4-month AL-fed animals (Fig. 6C)
. Collectively, these data suggest that an inhibition of axin expression in the thymus is associated with increased thymic output.
 View larger version (31K): [in a new window] |
Figure 6. Age-related increase in axin expression is prevented by chronic CR. (A) Quantitation of axin expression by Western blot of total thymic protein lysates displayed a marked increase in axin protein expression with age, and CR maintains the thymic axin to the levels observed in young animals. (B) The real-time PCR analysis of thymic axin mRNA in young and old mice. The age-matched CR mouse thymi revealed significantly (P<0.05) reduced axin expression in 12- and 18-month thymi (n=6/group). The data are expressed as mean ± SEM. (C) The splenic CD4 cells were isolated to prepare the DNA. Compared with AL-fed 4-month-old mice, the 10-month-old animals displayed a significant reduction in the sjTRECs, suggesting a reduced thymic cell output. The 10-month-old mice on a CR diet displayed a significant increase in the sjTRECs. A total of eight mice/group was used for TREC real-time PCR analysis. The data are expressed as mean ± SEM.
|
 View larger version (25K): [in a new window] |
Figure 7. Regulation of Wnt and adipogenic genes during aging and by CR. (A) Western blot analysis of β-catenin and Wnt10b in the thymus did not show any significant changes. (B) Thymic mRNA expression of Wnt antagonist Nkd1 in mice is not affected by age (n=6). (C) The real-time PCR analysis of adipogenic transcripts in the thymus shows that aging leads to a significant (P<0.05) reduction in Pref-1 without affecting CHOP and KLF6 mRNA expression. Compared with 3-month-old mice, no differences in IL-7 mRNA were observed in 12-month AL- and CR-fed mice.
|
CR induced reduction in axin expression is not mediated by ghrelin signaling
It is well-recognized that rodents on a CR diet display enhanced food-seeking behavior in an effort to make up for the reduction in caloric intake [3
]. Ghrelin is a potent stomach-derived acylated peptide that induces food intake [4
]. We have recently demonstrated that in 10- to 12-month-old mice, the CR diet causes elevation in gastric ghrelin production, resulting in high circulating levels of this orexigenic peptide [47
]. In addition, our previous studies demonstrate that ghrelin infusion in aging mice reduces lipid-positive cells in the thymus, and GHSR null mice have increased adipocyte deposition in the thymus and display accelerated age-related thymic involution [4
]. Therefore, we next asked the question whether elevated axin expression in GHSR-deficient mice contributes to thymic involution and investigated if ghrelin-receptor (GHSR) mediated signaling is involved in regulating thymic axin expression during aging. The GHSR knockout and control wild type (WT) mice were aged for 10 months, and axin expression was studied in the thymus and primary TSCs of these mice. Compared with age-matched WT animals, we did not observe any change in axin-expressing cells or thymic axin mRNA expression in the 10-month-old GHSR–/– mice (Fig. 8
). In addition, the TSCs from WT and GHSR–/– mice showed no difference in the axin mRNA expression (Fig. 8B)
. These data suggest that regulation of axin during aging is likely mediated via a ghrelin-independent mechanism, and thymic adipogenesis in ghrelin null animals may be mediated by distinct adipogenic pathways.
 View larger version (35K): [in a new window] |
Figure 8. Ghrelin signaling does not regulate age-related changes in thymic axin expression. (A) The 12-month-old GHSR KO and WT mice thymic cryosections were labeled with axin (10x original magnifications), and nuclei were counterstained with DAPI. (B) The axin mRNA expression in the thymus and TSCs is not altered in 12-month-old GHSR null mice.
|
According to the prevailing view, the mechanisms responsible for reduced, naïve T cell production from aging thymus involve defects in lymphoid progenitors [10 , 48 ], TECs [5 , 49 ], thymocyte subsets, loss of intrathymic growth factors [9 ], and alteration in neuroendocrine [3 , 4 ] and sex-steroid hormones [2 ]. Considering that the thymus undergoes almost complete "lipomatous atrophy" with age [13 ], we studied the changes in the thymic microenvironment and the role of the Wnt regulator axin in the aging thymus. We identified that axin regulates adipogenesis of TSCs and axin is markedly elevated in the aging thymus. These results led us to test whether the anti-aging intervention CR regulates Wnt signals and adipogenic progression in aging thymus. Our data suggest that CR inhibits adipocyte accumulation, increases thymic output, and inhibits axin expression in aging thymus.
In contrast to a young thymus, where thymocytes are the major contributors to the thymic environment, the situation is reversed in aging, where the adipocytes constitute the bulk of the microenvironment. Despite these dramatic changes, the aging thymus retains limited capacity for generating naïve T cells until late in life [50 ], suggesting that the improvement of thymic function is achievable. We have demonstrated recently that thymic regeneration in aged mice supplemented with exogenous ghrelin leads to reduction in lipid-expressing preadipocyte cells [4 ]. Furthermore, the GHSR KO mice express a greater number of adipocytes in the thymus and have reduced thymic output [4 ]. These data further suggest that the presence of thymic adipocytes is negatively related to the ability of thymus to produce naïve T cells. It is known that the reduction in thymocyte numbers precedes the development of adipocytes during aging. Whether the generation of thymic adipocytes with progressive aging is cause or consequence of thymic involution is still unclear. However, our findings provide clear evidence that primary thymic stromal elements can undergo adipogenic transformation, a process that can be attenuated by specific down-regulation of the axin. Given that TSCs are required for various aspects of T cell development, it is plausible that adipogenic transformation of these cells with age may lead to reduced thymopoiesis (Fig. 9 ).
 View larger version (17K): [in a new window] |
Figure 9. Hypothetical model showing a role for axin in regulating thymic adiposity and involution. Aging leads to an increase in TF and a decrease in TEC, altering the TSC compositions. Increased expression of axin in TF induces adipogenesis, which may contribute to age-related accumulation of adipocytes in thymic space. CR maintains the thymic microenvironment by blocking the age-related increase in TF, reduced axin expression, and decreased thymic adiposity.
|
The CR-associated negative energy balance is the only preventive strategy that prolongs a healthy lifespan and attenuates age-related deficits in T cell output in primate models [54 , 55 ]. The age-related increase in the Wnt inhibitor, axin, was blocked by CR and was associated with increased thymopoiesis. Given the broad systemic effects of CR, it is possible that extrinsic factors altered by CR, such as ghrelin, insulin-like growth factor 1, or sex steroids, may affect the thymic axin expression indirectly. As GHSR null mice do not have altered thymic axin expression, it is likely that CR-induced ghrelin release does not mediate axin expression. However, our findings suggest that reduction of axin expression in aging is associated with improved thymic output. The axin null mice are embryonically lethal, so a TSC-specific deletion of axin and subsequent aging of these genetic models may provide direct causal evidence for a role of this gene in thymic involution and thymoadipogenesis. Furthermore, considering that negative energy balance elicited by CR affects multiple pathways, the increased thymopoiesis in CR mice is not solely a result of down-regulation of axin expression. However, given our findings that axin knockdown in TSCs prevents their differentiation into adipocytes, together with previous studies that axin overexpression reduces thymic cellularity [52 ], this suggests that increased axin with aging may promote thymic adipogenesis and involution (Fig. 9) .
The neuroendocrine hormones that signal the state of negative energy balance induced by CR are believed to play an important role in mediating some of the complex prolongevity effects of CR [3
, 49
]. Ghrelin is induced post-CR and ghrelin infusions in old mice can partially reverse thymic involution [3
, 4
]. Similar to CR, ghrelin also inhibits inflammation [56
, 57
], although it remains unclear whether ghrelin mediates the proimmune effects of CR. Interestingly, ablation of ghrelin signaling leads to an increase in proadipogenic regulators, such as PPAR, aP2-4 (FABP4/aP2), CD36, and perililipin in the TSCs [58
]. However, the aged GHSR null mice with accelerated thymic involution and thymic adiposity [4
, 58
] do not show an increase in axin expression in the thymus or TSCs, suggesting a distinct and specific regulatory mechanism.
These data suggest that ghrelin signaling via GHSR does not regulate the axin in the aging thymus. In addition, this also suggests that deficiency of ghrelin-mediated signaling and consequent thymic adiposity are under regulation of alternate regulatory pathways. Collectively, our data are consistent with the hypothesis that axin may play an important role in promoting the age-related thymic involution by increasing the adipocyte accumulation in the thymus. Thus, in addition to the present focus on regenerating the aging thymus by targeting lymphoid progenitors and TSCs, our findings suggest that cell-specific blocking of an age-related increase in proadipogenic regulators may significantly complement ongoing efforts to rejuvenate thymic function.
ACKNOWLEDGEMENTS
The research work was supported in part by the Coypu Foundation and a grant from the NIH (AG031797) to V. D. D. The present work used the facilities of the Genomics and Cell Biology & Bioimaging (CBB) Core facilities supported by NIH Grant 1 P20 RR02/1945 and CBB Core Facility of the Pennington Center of Biomedical Research Excellence (NIH P20 RR-021945) and Clinical Nutrition Research Unit (NIH P30 DK072476). We thank Chiaki Nakata, Rachel Ohlmeyer, and Wubing Ye for technical assistance, Dr. Donald Ingram for thoughtful discussions, Dr. Roy Smith for providing the GHSR null mice, Dr. Gregory Sempowski for the mouse sjTREC plasmid, Dr. Richard Boyd for the generous gift of the MTS15 antibody, and Dr. Ann Chidgey for advice with TSC sorting procedures. We also thank Dr. Ping Zhang and Ms. Constance Porretta at Louisiana State University (LSU)-Health Science Center and Ms. Marilyn Dietrich at LSU School of Veterinary Medicine for FACS sorting.
Received October 14, 2008; revised January 10, 2009; accepted February 2, 2009.
REFERENCES
2, a lipid-activated transcription factor Cell 79,1147-1156[CrossRef][Medline] and β TCR excision circles in naive T cells J. Immunol. 177,4391-4401-like expression necessary for inducing T cell development J. Immunol. 175,4858-4865
B activation in human endothelial cells Circulation 109,2221-2226This article has been cited by other articles:
 |
|
 |
|
  H. Yang, Y.-H. Youm, B. Vandanmagsar, J. Rood, K. G. Kumar, A. A. Butler, and V. D. Dixit Obesity accelerates thymic aging Blood, October 29, 2009; 114(18): 3803 - 3812. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Yang, Y.-H. Youm, and V. D. Dixit Inhibition of Thymic Adipogenesis by Caloric Restriction Is Coupled with Reduction in Age-Related Thymic Involution J. Immunol., September 1, 2009; 183(5): 3040 - 3052. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
