Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Lymphohematopoietic progenitors do not have a synchronized defect with age-related thymic involution

Lymphohematopoietic progenitors do not have a synchronized defect with age-related thymic involution <h1>Introduction</h1> Age-related thymic involution is believed to be one of the key drivers of immunosenescence and decline of the adaptive immune system ( Miller, 2000 ; Taub & Longo, 2005 ), because atrophic thymus reduces thymic lymphopoiesis, resulting in exhaustion of the naïve T-cell pool ( Haynes et al ., 1997 ) and contraction of the T-cell receptor repertoire. This, in turn, causes reduced responsiveness to foreign antigens, including reduction of interleukin (IL)-2 secretion in response to stimulation ( Linton et al ., 1996 ; Haynes et al ., 1997 ). Decreased thymopoiesis in the aged atrophic thymus involves both thymocytes (T-lymphocyte precursors of hematopoietic origin) and thymic stromal cells (TSC, of nonhematopoietic origin), which are primarily thymic epithelial cells (TEC). TSCs constitute the thymic microenvironment that supports thymocyte maturation; conversely, thymocytes provide growth factors that favor survival of TSCs. Because the thymus does not contain self-renewing stem cells, it semicontinuously recruits bone marrow (BM)-derived T-lymphocyte progenitors from the blood ( Foss et al ., 2001 ). The BM hematopoietic stem cells (HSC), that is, the population with LSK [T-lineage(Lin) − Sca-1 + c-kit + ] and Flt3 − phenotypes, were considered to be T-lymphocyte progenitors ( Schwarz & Bhandoola, 2004 ). Whereas, the Flt3 + LSK progenitors were considered as direct thymus-setting progenitor cells ( Schwarz et al ., 2007 ). Early thymocytes are CD3 − , CD4 − 8 − double negative (DN) cells ( Ceredig & Rolink, 2002 ) that can be divided into four developmental stages: CD44 + 25 − (DN1), CD44 + 25 + (DN2), CD44 − 25 + (DN3) and CD44 − 25 − (DN4) ( Godfrey et al ., 1993 , 1994 ). The DN1 subset is a heterogeneous population that includes multipotent cells that can produce T, B, dendritic and natural killer (NK) cells ( Shortman & Wu, 1996 ). Five subsets (DN1a, b, c, d and e) of DN1 cells were identified in the adult murine thymus ( Porritt et al ., 2004 ). Among them, early T-cell progenitors (ETP) ( Allman et al ., 2003 ) are a canonical T-cell lineage progenitor subset that consists of DN1a (Lin − DN1, CD117 hi CD24 − ) and DN1b (Lin − DN1, CD117 hi CD24 + ) cells ( Porritt et al ., 2004 ; Sambandam et al ., 2005 ). DN1a cells may comprise BM HSCs that have recently migrated to the thymus, and DN1b cells are likely derived from the DN1a population ( Porritt et al ., 2004 ). Therefore, DN1a cells are considered as early-stage ETPs, while DN1b cells are considered as late-stage ETPs. The non-ETP DN1d (CD117 − CD24 + ) and DN1e cells (CD117 − CD24 − ) can also produce T cells through a noncanonical developmental pathway, apparently without passing through the DN2 and DN3 stages ( Porritt et al ., 2004 ; Pelayo et al ., 2005 ). Currently, there are two different concepts of the cellular mechanisms responsible for age-related thymic atrophy. One holds that aging intrinsically impairs lymphohematopoietic progenitor cells (LPC), including BM HSCs ( Sudo et al ., 2000 ) and/or thymic ETPs ( Min et al ., 2004 ), which are phenotypically similar to BM HSCs and are believed to be progeny of BM HSCs ( Allman et al ., 2003 ; Schwarz & Bhandoola, 2004 ). The other maintains that aging results in dysfunction of the thymic microenvironmental cells ( Doria et al ., 1997 ; Aspinall & Andrew, 2000 , 2001 ; Geiger & Van Zant, 2002 ), which causes secondary changes in thymocytes, resulting in thymic involution. Defective LPCs could arise from an intrinsic defect in BM HSCs and/or thymic ETPs or could result from an extrinsic defect mediated by altered functional properties of stromal niches including TSCs/TECs. It is unclear which of these models accounts for aging of T-lymphocyte progenitors. In the current study, using a different experimental approach we show that aged animals did not have an intrinsic defect, that is, cell autonomous loss of ability, in T-lymphocyte progenitors. We provided a fetal thymic microenvironment to aged mice by transplanting a fetal thymus into the kidney capsules of aged mice. In this setting, T-lymphocyte progenitors from aged mice re-established normal thymic lymphopoiesis within the grafts. Conversely, intrathymic injection of ETPs from young animals into old mice did not restore normal thymopoiesis, implying that a shortage and/or intrinsic defects of ETPs do not account for age-related thymic involution. Together, our findings suggest that the underlying cause of age-related thymic involution results primarily from changes in the thymic microenvironment, which cause extrinsic rather than intrinsic defects in T-lymphocyte progenitors. <h1>Results</h1> <h2>LPCs from old mice have preserved thymic lymphopoietic potential when provided with fetal thymic grafts</h2> To establish whether aging causes intrinsic defects in LPCs that populate the thymus, we took advantage of an in vivo reconstitution approach – kidney capsule transplantation. We transplanted gestation day 14 (E14) fetal thymic lobes from CD45.1 + wild-type (wt) or CD45.1 + RAG −/– mice into the kidney capsules of aged CD45.2 + wt mice to generate a chimera composed of a young donor TEC network with LPCs from old mice. Transplantation of fetal thymus mainly replaces the thymic microenvironmental niche, without altering other physiological conditions in aged mice. As a control, CD45.1 + fetal thymic lobes were also transplanted into young CD45.2 + wt mice. We found that the dramatic decline in total thymocyte number in the native thymus of aged mice ( Fig. 1B , left subpanel) was corrected in the grafted fetal thymus ( Fig. 1B , right subpanel) by LPCs from aged host mice, because ≥ 96% of cells within the grafted thymic lobes were CD45.2 + host-type cells (data not shown). The absolute numbers of DN, double-positive (DP), and CD4 + or CD8 + single-positive (SP) thymocytes were markedly reduced in native thymi of old mice ( Fig. 1C ), but these numbers were restored to normal by CD45.2 + host cells in fetal thymic grafts ( Fig. 1D ). Similarly, the reduced absolute cell numbers of DN1–DN4 (Lin − CD4 − CD8 − CD3 − ) subpopulations in native thymi of aged mice ( Fig. 1E , lower panel) were restored to normal by CD45.2 + cells in the fetal thymic grafts ( Fig. 1F , lower panel). The partial block at the transition from the DN1 to DN2 stage, which is characteristic of the aged thymus ( Aspinall, 1997 ; Thoman, 1995 ; Aspinall & Andrew, 2000 ), was manifest as a relative increase in the percentage of DN1 cells ( Fig. 1E , upper panel), and this partial block was overcome in the fetal thymic grafts ( Fig. 1F , upper panel). The absolute number of CD45.2 + ETPs, which is significantly reduced in the native thymus of old mice, increased to normal levels in the grafted thymus ( Fig. 1G ). Therefore, we conclude that, in the presence of a fetal thymic microenvironment, LPCs from aged mice have the same functional capacity as their young counterparts to seed the thymus, to produce all thymocyte subpopulations via the canonical developmental pathway, and to support normal thymic lymphopoiesis. Young RAG −/– mice have essentially no medulla ( Fig. 2 , middle row panels), because T-cell progenitors are unable to promote TEC development ( van Ewijk et al ., 2000 ). To determine whether T-cell progenitors from aged mice could normalize the medullary TECs in RAG −/– mice, we transplanted fetal thymic lobes from CD45.1 + RAG −/– mice into kidney capsules of aged or young CD45.2 + wt mice. Under these conditions, the RAG −/– thymic medulla was regenerated to similar degrees in aged or young host mice ( Fig. 2 , bottom two rows), implying that the LPCs from old or young hosts interacted equally well with the grafted RAG −/– TECs. In combination with the finding in Fig. 1 , these results show that T-lymphocyte progenitors in aged mice retain the functional capacity to seed the fetal thymus and to generate T cells, even when the native thymus as a whole has undergone significant atrophy due to aging. Peripheral CD4 + T cells in aged animal exhibit reduced IL-2 production in response to the antigens or costimulators, while newly generated CD4 + T cells do not ( Haynes et al ., 1997 , 2005 ). In order to determine whether the grafted fetal thymus can generate newly functional peripheral T cells, we analyzed IL-2 production from splenic CD4 + T cells in the aged mice, before and 7 weeks after receiving transplants of fetal thymus, in response to 5 h of costimulation with CD3ε and CD28 antibodies ( Fig. 3 ). We found that the percentage of CD4 + IL-2 + T cells was approximately 50% greater in aged mice 7 weeks after transplantation, compared to aged mice without the treatment ( P < 0.05). Furthermore, while untreated aged mice exhibited 50% fewer CD4 + IL-2 + cells compared to young mice, transplantation of fetal thymus in aged mice rescued the age-related loss of these cells, leaving no significant difference with young controls ( Fig. 3B ). Our results imply that there are many newly produced functional CD4 + T cells derived from the young thymus grafted into the kidney capsule of aged mice. <h2>Young ETPs cannot rescue thymopoiesis in the aged thymus</h2> Because there is extensive crosstalk between TECs and early-stage thymocytes ( van Ewijk et al ., 2000 ), age-impaired ETPs are believed to cause poor conditioning of TECs; the resulting reduced proliferation of TECs may lead to thymic involution ( Min et al ., 2004 ). If the principal abnormality in thymic aging is an intrinsic defect in ETPs, provision of young ETPs to the thymus of aged mice should correct the defect. Therefore, we injected 2000–3000 sorted young wt CD45.1 + ETPs (purity > 96%) intrathymically into sublethally irradiated aged and young CD45.2 + mice. To ensure administration of adequate numbers of ETPs, we used 20- to 30-fold more cells than have been previously used for intrathymic injection ( Schwarz & Bhandoola, 2004 ). Similar injections were carried out in RAG −/– young CD45.2 + mice, whose thymus is almost as small as that of aged wt mice. Three weeks after this procedure, when the number of donor cells should peak in the host thymus ( Schwarz & Bhandoola, 2004 ), the total thymocyte number remained much lower in old than in young wt ( Fig. 4B ) and RAG −/– (data not shown) hosts, with almost no improvement in thymopoiesis in old mice compared to noninjected animals ( Fig. 4A ). CD45.1 + donor cells expanded much more effectively in the thymi of young RAG −/– (data not shown) and young wt mice than in those of aged wt mice ( Fig. 4C ). These CD45.1 + cells were clearly thymocytes, as they were virtually all DP, CD4 + or CD8 + SPs ( Fig. 4C , bottom set of dot plots). Therefore, this suggests that inadequate numbers of ETPs and/or intrinsic defects of ETPs are not the primary cause of age-related thymic involution, and again suggest that the thymic microenvironment of old mice provides inadequate signals for expansion and differentiation of ETPs within the thymus. <h2>Aged mice do not have reduced numbers of LSK cells in BM, or a substantial deficiency of early-stage ETPs in the thymus</h2> The absolute number of ETPs per thymus in the aged thymus is significantly reduced ( Fig. 1G ). However, it is unclear if decreased numbers of thymic ETPs in aged mice are due to insufficient numbers of precursors from aged BM or from defects that arise after the precursors reach the thymus. We first analyzed BM LSK cells, and found that the percentages and absolute numbers were similar in young and aged mice ( Fig. 5A,B ). Therefore, the reduced ETP numbers in the thymus of aged mice were not due to a shortage of LSK cells in BM. We next analyzed thymic ETP subsets, DN1a and DN1b cells, which represent two successive developmental stages recruited from BM-derived progenitors ( Porritt et al ., 2004 ). The absolute number of DN1b cells (CD117 hi CD24 + DN1) per thymus was significantly decreased in aged animals, while that of DN1a cells (CD117 hi CD24 − DN1) per thymus was decreased but not significantly ( Fig. 5C,D ). It seems that the numbers of all thymocyte subpopulations after DN1a, including DN2, DN3, DN4, DP and SPs, were all significantly decreased in old mice ( Fig. 1C,E , and other reports: Thoman, 1995 ; Aspinall & Andrew, 2000 ; Heng et al ., 2005 ). The relative preservation of the number of BM lymphohematopoietic progenitors and thymic early-stage ETPs in aged mice contrasts with the reduced numbers of T-lymphocyte progenitors that have resided in the thymus for longer periods. Thus, there appears to be a stage-related progressive loss in numbers of progenitors in the aged thymus, suggesting that the thymic microenvironment may result in extrinsic defects in T-cell precursors. <h1>Discussion</h1> The commonly used methods to study LPC function are BM transplantation and fetal thymic organ culture (FTOC). However, FTOC is an in vitro system that does not closely mimic physiologic conditions, whereas BM transplantation reflects physiologic conditions, but it can be used only to evaluate lymphocyte transplantation and not transplantation of a TEC network. We took advantage of an alternative approach – kidney capsule transplantation – in which both LPCs and TECs are tested in vivo . A pilot experiment using kidney capsule transplantation showed that T-cell progenitors from young and aged mice exhibited no discernible differences in their ability to colonize neonatal grafts ( Mackall & Gress, 1997 ). However, they did not evaluate the further development and function of LPCs from aged mice in detail, so that subsequent studies still questioned whether aging T-lymphocyte progenitors have intrinsic defects ( Sudo et al ., 2000 ; Min et al ., 2004 ). We demonstrated that LPCs from physiologically intact aged mice are as capable as those from young mice of producing all thymocyte subpopulations through the canonical pathway, when they are provided with a functional fetal TEC network ( Fig. 1 ). Also, LPCs from aged and young mice can equally well interact with grafted TECs from RAG −/– mice to restore the medullary architecture in the RAG −/– thymus ( Fig. 2 ). Furthermore, LPCs from aged mice can restore function of the peripheral CD4 + T cells in aged animals ( Fig. 3 ). These findings suggest that there are no intrinsic defects in T-lymphocyte progenitors of aged mice, compared to those of their young counterparts. If insufficient numbers of T-lymphocyte progenitors are the cause of age-related thymic involution, provision of adequate numbers of young HSCs and/or ETPs should correct the defect. Despite intrathymic injection of relatively large numbers of ETPs from young mice, we were unable to restore thymic lymphopoiesis in old mice to a normal level ( Fig. 4 ). These findings are consistent with those of others, who were unable to restore normal thymopoiesis and thymic architecture of aged mice despite intravenous injection of unfractionated young BM cells ( Mackall et al ., 1998 ). We also confirmed that intravenous injection of sorted young BM HSC (Lin − Sca-1 + c-kit + Flt3 − ) progenitors into old mice failed to restore thymic lymphopoiesis (Supplementary Fig. S1). Other investigators found that unfractionated BM cells from old mice showed a reduced capacity to competitively repopulate the peripheral immune system, compared to BM cells from young mice ( Sudo et al ., 2000 ), leading to the concept that T-lymphocyte progenitors from aged mice are intrinsically defective. However, the BM from old mice contains a higher proportion of myeloid precursors and a lower proportion of lymphoid precursors, compared to young mice ( Sudo et al ., 2000 ). That is why the proportion of unfractionated BM cells expressing myeloid genes is increased in aged mice ( Rossi et al ., 2005 ). Therefore, the reduced capacity of unfractionated BM from old mice to reconstitute the peripheral immune system may be due to the reduced fraction of lymphocyte progenitors, rather than any intrinsic defect in the progenitors themselves. Another evidence for an intrinsic defect in aging T-lymphocyte progenitors is based primarily on defects identified in sorted ETPs in an FTOC assay in vitro ( Min et al ., 2004 ), whereas culture of the entire DN1 population in an FTOC assay showed no age-associated defect ( Aspinall & Andrew, 2001 ). Sorted ETPs may not adequately represent the entire population of LPCs in vivo ; for example, the non-ETP DN1 subpopulations, which normally have precursors of B cells and NK cells, are not included. ETPs express c-kit (transmembrane receptor tyrosine kinase), whereas non-ETP DN1 cells do not. However, c-kit can be up-regulated in different progenitors. For example, Notch signals from stromal cells can induce B-cell precursors to up-regulate c-kit expression and develop into T cells ( Hoflinger et al ., 2004 ; Krueger et al ., 2006 ; Massa et al ., 2006 ). Therefore, non-ETP DN1 subsets may be altered to develop into T cells under certain conditions. This may explain the different results obtained using entire DN1 population ( Aspinall & Andrew, 2001 ) or using sorted ETPs ( Min et al ., 2004 ) in an FTOC assay. In addition, recently Weinberg and colleagues reported that treatment of aged mice with keratinocyte growth factor restored thymic function through improvement of the thymic epithelial microenvironment ( Min et al ., 2007 ). Their results support our current findings that T-lymphocyte progenitors in aged mice do not have intrinsic defects in their capacity to generate T cells so long as a functional thymic microenvironment is provided. The same study suggested that the ‘intrinsic’ defect found in transplanted BM cells/ETPs from aged mice may be due to TEC defects that affect transplantability of the LPCs. It has been reported that non-ETP-DN1 (c-kit − DN1) subsets, such as DN1d and DN1e cells, can generate T cells via a noncanonical pathway in vitro ( Porritt et al ., 2004 ), in which DN1 cells directly give rise to DN4 cells without passing through the intermediate DN2 and DN3 stages. This mechanism may be active in the atrophic thymus as we noted a relative increase in the DN1d and DN1e subpopulations within the aged thymus (data not shown). A disproportionate increase in T-cell development through the noncanonical pathway has been described when the thymic microenvironment is disrupted, as in the Foxn1 gene splicing (Delta) mutation that we described ( Su et al ., 2003 ) and in mice with thymocytes that overexpress GATA-3 ( David-Fung et al ., 2006 ). However, if a functional microenvironment is provided to DN1 cells, they can develop via the canonical route of DN1 to DN2, DN3, and then mature stages, as seen in our kidney capsule chimera. There is growing interest in the use of regenerative strategies to treat or prevent the decline in immune function associated with thymic insufficiency ( Chidgey & Boyd, 2006 ), coupled with a search for TEC-specific stem cells ( Bleul et al ., 2006 ) and reassessment of the function of aging LPCs ( van den Brink et al ., 2004 ). Our study provides initial evidence favoring the potential use of LPCs from aged individuals themselves as therapy for diminished thymic lymphopoiesis in the elderly, and suggests that rejuvenating therapy should focus on improving the function of TECs, rather than T-cell progenitors, by using agents such as exogenous keratinocyte growth factor ( Min et al ., 2007 ). In summary, we conclude that age-related deficient T-lymphocyte progenitors in the thymus develop extrinsic defects that result from interactions with the aged thymic microenvironment. Our findings provide a new perspective on the interactions between LPCs and the thymic epithelial microenvironment during age-related thymic involution. <h1>Experimental procedures</h1> <h2>Mice</h2> We used C57BL/6 congenic mice that expressed CD45.2 + or CD45.1 + on the cell surface. Young (1.5–2 months old) and aged (≥ 18 months old) wt mice, and young RAG −/– mice were purchased from National Institutes of Health (NIH), NCI (National Cancer Institute), NIA (National Institute on Aging) (Bethesda, MD, USA), and Jackson Laboratory (Bar Harbor, ME, USA), respectively. CD45.1 + RAG −/– mice were generated from mating of CD45.2 + RAG1 −/– with CD45.1 + wt mice. The gestation day was determined by designating the day on which the vaginal plug was found as day 0. All animal experiments were done according to the protocols approved by the Institutional Animal Care and Use Committee of the University of Texas Health Center at Tyler, in accordance with guidelines of the NIH. <h2>Kidney capsule transplantation</h2> Survival surgery was performed under sterile conditions after intraperitoneal administration of the anesthetics, ketamine (100 mg kg −1 ) and xylazine (10 mg kg −1 ) to CD45.2 + 1.5- to 2-month-old young wt and 17- to 20-month-old aged wt host mice. A small dorsolateral incision was made to expose the left kidney and a small hole was made in the kidney capsule. Thymic lobes from CD45.1 + wt or CD45.1 + RAG −/– donor mice at gestation day 14 (E14) were placed under the kidney capsule, and the incision was closed with sterile sutures. After surgery, mice were placed in a specific pathogen-free environment for indicated weeks. <h2>Staining of tissue sections with immunofluorescence and hematoxylin-eosin</h2> Sections (6 µm) from cutting temperature compound-embedded frozen thymic tissue were fixed in cold acetone and incubated with optimal dilutions of rat antikeratin-8 (Developmental Studies Hybridoma Bank, Iowa City, IA, USA) and rabbit antikeratin-5 (Covance, Berkeley, CA, USA) antibodies ( Su et al ., 2003 ). Immunoreactivity was detected with Alexa-Fluor-488-conjugated antirat-IgG (Invitrogen Molecular Probe, Carlsbad, CA, USA) and Cy3-conjugated antirabbit-IgG (Jackson ImmunoResearch Laboratory, West Grove, PA, USA), respectively. Paraffin sections (4 µm) were stained with hematoxylin-eosin (H&E). <h2>Analysis of IL-2 secretion in CD4 T cells to responsiveness of stimulation</h2> Spleen cells were freshly isolated from young and aged mice, before and 7 weeks after kidney transplantation. The erythrocytes were depleted with ACK lysing buffer (pH 7.2, 0.15 m NH 4 Cl/1.0 m KHCO 3 /0.1 m m Na 2 EDTA). The spleen cells (2 × 10 6 per well) were cultured with CD3ε and CD28 antibodies (2 µg mL −1 each) supplemented with GolgiStop (4 µg mL −1 ) for 5 h in a 48-well plate. The harvested cells were stained for surface CD4, then fixed with 1% PFA/PBS, permeabilized with 0.1% Triton-X100 in 0.1% sodium citrate, pH 7.2, then stained with intracellular IL-2. The results will be analyzed by a FACS Calibur (Becton Dickinson, San Jose, CA, USA). All antibodies and GolgiStop were purchased from BD Pharmingen (San Diego, CA, USA). <h2>Intrathymic injection of sorted ETPs</h2> Freshly isolated thymocytes from a pool of young CD45.1 + wt mice were enriched for ETPs by depletion with CD8 IgM antibody (HO2.2, ATCC) and complement (Cedarlane Laboratory Ltd., Burlington, NC, USA) as we previously described ( Su et al ., 1997 ), and then stained with PE-lineage markers, as well as PE-CD25 and PE-CD127, APC-CD117 and FITC-CD44, 2000–3000 cells of ETPs in 15 µL PBS were intrathymically injected into each aged or young control CD45.2 + mouse that received 500 rads of sublethal irradiation after 2–3 h. The methods of anesthesia and suprasternal notch surgery were as previously detailed ( Schwarz & Bhandoola, 2004 ). <h2>DN thymocyte enrichment and flow cytometric analysis</h2> For analysis and sorting of ETPs, DN cells were enriched from freshly isolated thymocytes by depletion with CD8 IgM antibody (clone HO2.2, ATCC) and complement. The lineage gate was set up with a cocktail of multiple PE-conjugated rat antibodies, as described by Porritt et al . (2004) and Schwarz & Bhandoola (2004 ). The flow cytometric strategies to identify DN1 subpopulations are provided in the relevant figure legends. For analysis of BM LSK progenitors, BM cells were collected from the femur and erythrocytes were deleted in ACK lysing buffer same as above. The cells were stained with PE-lineage markers as outlined above, and with APC-antimouse CD117 (c-kit) and FITC-antimouse Sca-1 (clone D7, eBioscience, San Diego, CA, USA). Staining of CD4 vs. CD8 and CD44 vs. CD25 for thymocyte subpopulations was performed as we previously described ( Su & Manley, 2000 ). <h2>Statistical analysis of data</h2> All data were analyzed with Prism software using a two-tailed Student's t -test. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Aging Cell Wiley

Lymphohematopoietic progenitors do not have a synchronized defect with age-related thymic involution

Download PDF
10 pages

Loading next page...
 
/lp/wiley/lymphohematopoietic-progenitors-do-not-have-a-synchronized-defect-with-WBn8ynLt5D

References (40)

Publisher
Wiley
Copyright
© 2007 The Authors Journal compilation © Blackwell Publishing Ltd/Anatomical Society of Great Britain and Ireland 2007
ISSN
1474-9718
eISSN
1474-9726
DOI
10.1111/j.1474-9726.2007.00325.x
pmid
17681038
Publisher site
See Article on Publisher Site

Abstract

<h1>Introduction</h1> Age-related thymic involution is believed to be one of the key drivers of immunosenescence and decline of the adaptive immune system ( Miller, 2000 ; Taub & Longo, 2005 ), because atrophic thymus reduces thymic lymphopoiesis, resulting in exhaustion of the naïve T-cell pool ( Haynes et al ., 1997 ) and contraction of the T-cell receptor repertoire. This, in turn, causes reduced responsiveness to foreign antigens, including reduction of interleukin (IL)-2 secretion in response to stimulation ( Linton et al ., 1996 ; Haynes et al ., 1997 ). Decreased thymopoiesis in the aged atrophic thymus involves both thymocytes (T-lymphocyte precursors of hematopoietic origin) and thymic stromal cells (TSC, of nonhematopoietic origin), which are primarily thymic epithelial cells (TEC). TSCs constitute the thymic microenvironment that supports thymocyte maturation; conversely, thymocytes provide growth factors that favor survival of TSCs. Because the thymus does not contain self-renewing stem cells, it semicontinuously recruits bone marrow (BM)-derived T-lymphocyte progenitors from the blood ( Foss et al ., 2001 ). The BM hematopoietic stem cells (HSC), that is, the population with LSK [T-lineage(Lin) − Sca-1 + c-kit + ] and Flt3 − phenotypes, were considered to be T-lymphocyte progenitors ( Schwarz & Bhandoola, 2004 ). Whereas, the Flt3 + LSK progenitors were considered as direct thymus-setting progenitor cells ( Schwarz et al ., 2007 ). Early thymocytes are CD3 − , CD4 − 8 − double negative (DN) cells ( Ceredig & Rolink, 2002 ) that can be divided into four developmental stages: CD44 + 25 − (DN1), CD44 + 25 + (DN2), CD44 − 25 + (DN3) and CD44 − 25 − (DN4) ( Godfrey et al ., 1993 , 1994 ). The DN1 subset is a heterogeneous population that includes multipotent cells that can produce T, B, dendritic and natural killer (NK) cells ( Shortman & Wu, 1996 ). Five subsets (DN1a, b, c, d and e) of DN1 cells were identified in the adult murine thymus ( Porritt et al ., 2004 ). Among them, early T-cell progenitors (ETP) ( Allman et al ., 2003 ) are a canonical T-cell lineage progenitor subset that consists of DN1a (Lin − DN1, CD117 hi CD24 − ) and DN1b (Lin − DN1, CD117 hi CD24 + ) cells ( Porritt et al ., 2004 ; Sambandam et al ., 2005 ). DN1a cells may comprise BM HSCs that have recently migrated to the thymus, and DN1b cells are likely derived from the DN1a population ( Porritt et al ., 2004 ). Therefore, DN1a cells are considered as early-stage ETPs, while DN1b cells are considered as late-stage ETPs. The non-ETP DN1d (CD117 − CD24 + ) and DN1e cells (CD117 − CD24 − ) can also produce T cells through a noncanonical developmental pathway, apparently without passing through the DN2 and DN3 stages ( Porritt et al ., 2004 ; Pelayo et al ., 2005 ). Currently, there are two different concepts of the cellular mechanisms responsible for age-related thymic atrophy. One holds that aging intrinsically impairs lymphohematopoietic progenitor cells (LPC), including BM HSCs ( Sudo et al ., 2000 ) and/or thymic ETPs ( Min et al ., 2004 ), which are phenotypically similar to BM HSCs and are believed to be progeny of BM HSCs ( Allman et al ., 2003 ; Schwarz & Bhandoola, 2004 ). The other maintains that aging results in dysfunction of the thymic microenvironmental cells ( Doria et al ., 1997 ; Aspinall & Andrew, 2000 , 2001 ; Geiger & Van Zant, 2002 ), which causes secondary changes in thymocytes, resulting in thymic involution. Defective LPCs could arise from an intrinsic defect in BM HSCs and/or thymic ETPs or could result from an extrinsic defect mediated by altered functional properties of stromal niches including TSCs/TECs. It is unclear which of these models accounts for aging of T-lymphocyte progenitors. In the current study, using a different experimental approach we show that aged animals did not have an intrinsic defect, that is, cell autonomous loss of ability, in T-lymphocyte progenitors. We provided a fetal thymic microenvironment to aged mice by transplanting a fetal thymus into the kidney capsules of aged mice. In this setting, T-lymphocyte progenitors from aged mice re-established normal thymic lymphopoiesis within the grafts. Conversely, intrathymic injection of ETPs from young animals into old mice did not restore normal thymopoiesis, implying that a shortage and/or intrinsic defects of ETPs do not account for age-related thymic involution. Together, our findings suggest that the underlying cause of age-related thymic involution results primarily from changes in the thymic microenvironment, which cause extrinsic rather than intrinsic defects in T-lymphocyte progenitors. <h1>Results</h1> <h2>LPCs from old mice have preserved thymic lymphopoietic potential when provided with fetal thymic grafts</h2> To establish whether aging causes intrinsic defects in LPCs that populate the thymus, we took advantage of an in vivo reconstitution approach – kidney capsule transplantation. We transplanted gestation day 14 (E14) fetal thymic lobes from CD45.1 + wild-type (wt) or CD45.1 + RAG −/– mice into the kidney capsules of aged CD45.2 + wt mice to generate a chimera composed of a young donor TEC network with LPCs from old mice. Transplantation of fetal thymus mainly replaces the thymic microenvironmental niche, without altering other physiological conditions in aged mice. As a control, CD45.1 + fetal thymic lobes were also transplanted into young CD45.2 + wt mice. We found that the dramatic decline in total thymocyte number in the native thymus of aged mice ( Fig. 1B , left subpanel) was corrected in the grafted fetal thymus ( Fig. 1B , right subpanel) by LPCs from aged host mice, because ≥ 96% of cells within the grafted thymic lobes were CD45.2 + host-type cells (data not shown). The absolute numbers of DN, double-positive (DP), and CD4 + or CD8 + single-positive (SP) thymocytes were markedly reduced in native thymi of old mice ( Fig. 1C ), but these numbers were restored to normal by CD45.2 + host cells in fetal thymic grafts ( Fig. 1D ). Similarly, the reduced absolute cell numbers of DN1–DN4 (Lin − CD4 − CD8 − CD3 − ) subpopulations in native thymi of aged mice ( Fig. 1E , lower panel) were restored to normal by CD45.2 + cells in the fetal thymic grafts ( Fig. 1F , lower panel). The partial block at the transition from the DN1 to DN2 stage, which is characteristic of the aged thymus ( Aspinall, 1997 ; Thoman, 1995 ; Aspinall & Andrew, 2000 ), was manifest as a relative increase in the percentage of DN1 cells ( Fig. 1E , upper panel), and this partial block was overcome in the fetal thymic grafts ( Fig. 1F , upper panel). The absolute number of CD45.2 + ETPs, which is significantly reduced in the native thymus of old mice, increased to normal levels in the grafted thymus ( Fig. 1G ). Therefore, we conclude that, in the presence of a fetal thymic microenvironment, LPCs from aged mice have the same functional capacity as their young counterparts to seed the thymus, to produce all thymocyte subpopulations via the canonical developmental pathway, and to support normal thymic lymphopoiesis. Young RAG −/– mice have essentially no medulla ( Fig. 2 , middle row panels), because T-cell progenitors are unable to promote TEC development ( van Ewijk et al ., 2000 ). To determine whether T-cell progenitors from aged mice could normalize the medullary TECs in RAG −/– mice, we transplanted fetal thymic lobes from CD45.1 + RAG −/– mice into kidney capsules of aged or young CD45.2 + wt mice. Under these conditions, the RAG −/– thymic medulla was regenerated to similar degrees in aged or young host mice ( Fig. 2 , bottom two rows), implying that the LPCs from old or young hosts interacted equally well with the grafted RAG −/– TECs. In combination with the finding in Fig. 1 , these results show that T-lymphocyte progenitors in aged mice retain the functional capacity to seed the fetal thymus and to generate T cells, even when the native thymus as a whole has undergone significant atrophy due to aging. Peripheral CD4 + T cells in aged animal exhibit reduced IL-2 production in response to the antigens or costimulators, while newly generated CD4 + T cells do not ( Haynes et al ., 1997 , 2005 ). In order to determine whether the grafted fetal thymus can generate newly functional peripheral T cells, we analyzed IL-2 production from splenic CD4 + T cells in the aged mice, before and 7 weeks after receiving transplants of fetal thymus, in response to 5 h of costimulation with CD3ε and CD28 antibodies ( Fig. 3 ). We found that the percentage of CD4 + IL-2 + T cells was approximately 50% greater in aged mice 7 weeks after transplantation, compared to aged mice without the treatment ( P < 0.05). Furthermore, while untreated aged mice exhibited 50% fewer CD4 + IL-2 + cells compared to young mice, transplantation of fetal thymus in aged mice rescued the age-related loss of these cells, leaving no significant difference with young controls ( Fig. 3B ). Our results imply that there are many newly produced functional CD4 + T cells derived from the young thymus grafted into the kidney capsule of aged mice. <h2>Young ETPs cannot rescue thymopoiesis in the aged thymus</h2> Because there is extensive crosstalk between TECs and early-stage thymocytes ( van Ewijk et al ., 2000 ), age-impaired ETPs are believed to cause poor conditioning of TECs; the resulting reduced proliferation of TECs may lead to thymic involution ( Min et al ., 2004 ). If the principal abnormality in thymic aging is an intrinsic defect in ETPs, provision of young ETPs to the thymus of aged mice should correct the defect. Therefore, we injected 2000–3000 sorted young wt CD45.1 + ETPs (purity > 96%) intrathymically into sublethally irradiated aged and young CD45.2 + mice. To ensure administration of adequate numbers of ETPs, we used 20- to 30-fold more cells than have been previously used for intrathymic injection ( Schwarz & Bhandoola, 2004 ). Similar injections were carried out in RAG −/– young CD45.2 + mice, whose thymus is almost as small as that of aged wt mice. Three weeks after this procedure, when the number of donor cells should peak in the host thymus ( Schwarz & Bhandoola, 2004 ), the total thymocyte number remained much lower in old than in young wt ( Fig. 4B ) and RAG −/– (data not shown) hosts, with almost no improvement in thymopoiesis in old mice compared to noninjected animals ( Fig. 4A ). CD45.1 + donor cells expanded much more effectively in the thymi of young RAG −/– (data not shown) and young wt mice than in those of aged wt mice ( Fig. 4C ). These CD45.1 + cells were clearly thymocytes, as they were virtually all DP, CD4 + or CD8 + SPs ( Fig. 4C , bottom set of dot plots). Therefore, this suggests that inadequate numbers of ETPs and/or intrinsic defects of ETPs are not the primary cause of age-related thymic involution, and again suggest that the thymic microenvironment of old mice provides inadequate signals for expansion and differentiation of ETPs within the thymus. <h2>Aged mice do not have reduced numbers of LSK cells in BM, or a substantial deficiency of early-stage ETPs in the thymus</h2> The absolute number of ETPs per thymus in the aged thymus is significantly reduced ( Fig. 1G ). However, it is unclear if decreased numbers of thymic ETPs in aged mice are due to insufficient numbers of precursors from aged BM or from defects that arise after the precursors reach the thymus. We first analyzed BM LSK cells, and found that the percentages and absolute numbers were similar in young and aged mice ( Fig. 5A,B ). Therefore, the reduced ETP numbers in the thymus of aged mice were not due to a shortage of LSK cells in BM. We next analyzed thymic ETP subsets, DN1a and DN1b cells, which represent two successive developmental stages recruited from BM-derived progenitors ( Porritt et al ., 2004 ). The absolute number of DN1b cells (CD117 hi CD24 + DN1) per thymus was significantly decreased in aged animals, while that of DN1a cells (CD117 hi CD24 − DN1) per thymus was decreased but not significantly ( Fig. 5C,D ). It seems that the numbers of all thymocyte subpopulations after DN1a, including DN2, DN3, DN4, DP and SPs, were all significantly decreased in old mice ( Fig. 1C,E , and other reports: Thoman, 1995 ; Aspinall & Andrew, 2000 ; Heng et al ., 2005 ). The relative preservation of the number of BM lymphohematopoietic progenitors and thymic early-stage ETPs in aged mice contrasts with the reduced numbers of T-lymphocyte progenitors that have resided in the thymus for longer periods. Thus, there appears to be a stage-related progressive loss in numbers of progenitors in the aged thymus, suggesting that the thymic microenvironment may result in extrinsic defects in T-cell precursors. <h1>Discussion</h1> The commonly used methods to study LPC function are BM transplantation and fetal thymic organ culture (FTOC). However, FTOC is an in vitro system that does not closely mimic physiologic conditions, whereas BM transplantation reflects physiologic conditions, but it can be used only to evaluate lymphocyte transplantation and not transplantation of a TEC network. We took advantage of an alternative approach – kidney capsule transplantation – in which both LPCs and TECs are tested in vivo . A pilot experiment using kidney capsule transplantation showed that T-cell progenitors from young and aged mice exhibited no discernible differences in their ability to colonize neonatal grafts ( Mackall & Gress, 1997 ). However, they did not evaluate the further development and function of LPCs from aged mice in detail, so that subsequent studies still questioned whether aging T-lymphocyte progenitors have intrinsic defects ( Sudo et al ., 2000 ; Min et al ., 2004 ). We demonstrated that LPCs from physiologically intact aged mice are as capable as those from young mice of producing all thymocyte subpopulations through the canonical pathway, when they are provided with a functional fetal TEC network ( Fig. 1 ). Also, LPCs from aged and young mice can equally well interact with grafted TECs from RAG −/– mice to restore the medullary architecture in the RAG −/– thymus ( Fig. 2 ). Furthermore, LPCs from aged mice can restore function of the peripheral CD4 + T cells in aged animals ( Fig. 3 ). These findings suggest that there are no intrinsic defects in T-lymphocyte progenitors of aged mice, compared to those of their young counterparts. If insufficient numbers of T-lymphocyte progenitors are the cause of age-related thymic involution, provision of adequate numbers of young HSCs and/or ETPs should correct the defect. Despite intrathymic injection of relatively large numbers of ETPs from young mice, we were unable to restore thymic lymphopoiesis in old mice to a normal level ( Fig. 4 ). These findings are consistent with those of others, who were unable to restore normal thymopoiesis and thymic architecture of aged mice despite intravenous injection of unfractionated young BM cells ( Mackall et al ., 1998 ). We also confirmed that intravenous injection of sorted young BM HSC (Lin − Sca-1 + c-kit + Flt3 − ) progenitors into old mice failed to restore thymic lymphopoiesis (Supplementary Fig. S1). Other investigators found that unfractionated BM cells from old mice showed a reduced capacity to competitively repopulate the peripheral immune system, compared to BM cells from young mice ( Sudo et al ., 2000 ), leading to the concept that T-lymphocyte progenitors from aged mice are intrinsically defective. However, the BM from old mice contains a higher proportion of myeloid precursors and a lower proportion of lymphoid precursors, compared to young mice ( Sudo et al ., 2000 ). That is why the proportion of unfractionated BM cells expressing myeloid genes is increased in aged mice ( Rossi et al ., 2005 ). Therefore, the reduced capacity of unfractionated BM from old mice to reconstitute the peripheral immune system may be due to the reduced fraction of lymphocyte progenitors, rather than any intrinsic defect in the progenitors themselves. Another evidence for an intrinsic defect in aging T-lymphocyte progenitors is based primarily on defects identified in sorted ETPs in an FTOC assay in vitro ( Min et al ., 2004 ), whereas culture of the entire DN1 population in an FTOC assay showed no age-associated defect ( Aspinall & Andrew, 2001 ). Sorted ETPs may not adequately represent the entire population of LPCs in vivo ; for example, the non-ETP DN1 subpopulations, which normally have precursors of B cells and NK cells, are not included. ETPs express c-kit (transmembrane receptor tyrosine kinase), whereas non-ETP DN1 cells do not. However, c-kit can be up-regulated in different progenitors. For example, Notch signals from stromal cells can induce B-cell precursors to up-regulate c-kit expression and develop into T cells ( Hoflinger et al ., 2004 ; Krueger et al ., 2006 ; Massa et al ., 2006 ). Therefore, non-ETP DN1 subsets may be altered to develop into T cells under certain conditions. This may explain the different results obtained using entire DN1 population ( Aspinall & Andrew, 2001 ) or using sorted ETPs ( Min et al ., 2004 ) in an FTOC assay. In addition, recently Weinberg and colleagues reported that treatment of aged mice with keratinocyte growth factor restored thymic function through improvement of the thymic epithelial microenvironment ( Min et al ., 2007 ). Their results support our current findings that T-lymphocyte progenitors in aged mice do not have intrinsic defects in their capacity to generate T cells so long as a functional thymic microenvironment is provided. The same study suggested that the ‘intrinsic’ defect found in transplanted BM cells/ETPs from aged mice may be due to TEC defects that affect transplantability of the LPCs. It has been reported that non-ETP-DN1 (c-kit − DN1) subsets, such as DN1d and DN1e cells, can generate T cells via a noncanonical pathway in vitro ( Porritt et al ., 2004 ), in which DN1 cells directly give rise to DN4 cells without passing through the intermediate DN2 and DN3 stages. This mechanism may be active in the atrophic thymus as we noted a relative increase in the DN1d and DN1e subpopulations within the aged thymus (data not shown). A disproportionate increase in T-cell development through the noncanonical pathway has been described when the thymic microenvironment is disrupted, as in the Foxn1 gene splicing (Delta) mutation that we described ( Su et al ., 2003 ) and in mice with thymocytes that overexpress GATA-3 ( David-Fung et al ., 2006 ). However, if a functional microenvironment is provided to DN1 cells, they can develop via the canonical route of DN1 to DN2, DN3, and then mature stages, as seen in our kidney capsule chimera. There is growing interest in the use of regenerative strategies to treat or prevent the decline in immune function associated with thymic insufficiency ( Chidgey & Boyd, 2006 ), coupled with a search for TEC-specific stem cells ( Bleul et al ., 2006 ) and reassessment of the function of aging LPCs ( van den Brink et al ., 2004 ). Our study provides initial evidence favoring the potential use of LPCs from aged individuals themselves as therapy for diminished thymic lymphopoiesis in the elderly, and suggests that rejuvenating therapy should focus on improving the function of TECs, rather than T-cell progenitors, by using agents such as exogenous keratinocyte growth factor ( Min et al ., 2007 ). In summary, we conclude that age-related deficient T-lymphocyte progenitors in the thymus develop extrinsic defects that result from interactions with the aged thymic microenvironment. Our findings provide a new perspective on the interactions between LPCs and the thymic epithelial microenvironment during age-related thymic involution. <h1>Experimental procedures</h1> <h2>Mice</h2> We used C57BL/6 congenic mice that expressed CD45.2 + or CD45.1 + on the cell surface. Young (1.5–2 months old) and aged (≥ 18 months old) wt mice, and young RAG −/– mice were purchased from National Institutes of Health (NIH), NCI (National Cancer Institute), NIA (National Institute on Aging) (Bethesda, MD, USA), and Jackson Laboratory (Bar Harbor, ME, USA), respectively. CD45.1 + RAG −/– mice were generated from mating of CD45.2 + RAG1 −/– with CD45.1 + wt mice. The gestation day was determined by designating the day on which the vaginal plug was found as day 0. All animal experiments were done according to the protocols approved by the Institutional Animal Care and Use Committee of the University of Texas Health Center at Tyler, in accordance with guidelines of the NIH. <h2>Kidney capsule transplantation</h2> Survival surgery was performed under sterile conditions after intraperitoneal administration of the anesthetics, ketamine (100 mg kg −1 ) and xylazine (10 mg kg −1 ) to CD45.2 + 1.5- to 2-month-old young wt and 17- to 20-month-old aged wt host mice. A small dorsolateral incision was made to expose the left kidney and a small hole was made in the kidney capsule. Thymic lobes from CD45.1 + wt or CD45.1 + RAG −/– donor mice at gestation day 14 (E14) were placed under the kidney capsule, and the incision was closed with sterile sutures. After surgery, mice were placed in a specific pathogen-free environment for indicated weeks. <h2>Staining of tissue sections with immunofluorescence and hematoxylin-eosin</h2> Sections (6 µm) from cutting temperature compound-embedded frozen thymic tissue were fixed in cold acetone and incubated with optimal dilutions of rat antikeratin-8 (Developmental Studies Hybridoma Bank, Iowa City, IA, USA) and rabbit antikeratin-5 (Covance, Berkeley, CA, USA) antibodies ( Su et al ., 2003 ). Immunoreactivity was detected with Alexa-Fluor-488-conjugated antirat-IgG (Invitrogen Molecular Probe, Carlsbad, CA, USA) and Cy3-conjugated antirabbit-IgG (Jackson ImmunoResearch Laboratory, West Grove, PA, USA), respectively. Paraffin sections (4 µm) were stained with hematoxylin-eosin (H&E). <h2>Analysis of IL-2 secretion in CD4 T cells to responsiveness of stimulation</h2> Spleen cells were freshly isolated from young and aged mice, before and 7 weeks after kidney transplantation. The erythrocytes were depleted with ACK lysing buffer (pH 7.2, 0.15 m NH 4 Cl/1.0 m KHCO 3 /0.1 m m Na 2 EDTA). The spleen cells (2 × 10 6 per well) were cultured with CD3ε and CD28 antibodies (2 µg mL −1 each) supplemented with GolgiStop (4 µg mL −1 ) for 5 h in a 48-well plate. The harvested cells were stained for surface CD4, then fixed with 1% PFA/PBS, permeabilized with 0.1% Triton-X100 in 0.1% sodium citrate, pH 7.2, then stained with intracellular IL-2. The results will be analyzed by a FACS Calibur (Becton Dickinson, San Jose, CA, USA). All antibodies and GolgiStop were purchased from BD Pharmingen (San Diego, CA, USA). <h2>Intrathymic injection of sorted ETPs</h2> Freshly isolated thymocytes from a pool of young CD45.1 + wt mice were enriched for ETPs by depletion with CD8 IgM antibody (HO2.2, ATCC) and complement (Cedarlane Laboratory Ltd., Burlington, NC, USA) as we previously described ( Su et al ., 1997 ), and then stained with PE-lineage markers, as well as PE-CD25 and PE-CD127, APC-CD117 and FITC-CD44, 2000–3000 cells of ETPs in 15 µL PBS were intrathymically injected into each aged or young control CD45.2 + mouse that received 500 rads of sublethal irradiation after 2–3 h. The methods of anesthesia and suprasternal notch surgery were as previously detailed ( Schwarz & Bhandoola, 2004 ). <h2>DN thymocyte enrichment and flow cytometric analysis</h2> For analysis and sorting of ETPs, DN cells were enriched from freshly isolated thymocytes by depletion with CD8 IgM antibody (clone HO2.2, ATCC) and complement. The lineage gate was set up with a cocktail of multiple PE-conjugated rat antibodies, as described by Porritt et al . (2004) and Schwarz & Bhandoola (2004 ). The flow cytometric strategies to identify DN1 subpopulations are provided in the relevant figure legends. For analysis of BM LSK progenitors, BM cells were collected from the femur and erythrocytes were deleted in ACK lysing buffer same as above. The cells were stained with PE-lineage markers as outlined above, and with APC-antimouse CD117 (c-kit) and FITC-antimouse Sca-1 (clone D7, eBioscience, San Diego, CA, USA). Staining of CD4 vs. CD8 and CD44 vs. CD25 for thymocyte subpopulations was performed as we previously described ( Su & Manley, 2000 ). <h2>Statistical analysis of data</h2> All data were analyzed with Prism software using a two-tailed Student's t -test.

Journal

Aging Cell – Wiley

Published: Oct 1, 2007

Keywords: hematopoietic stem cells; microenvironment; T-cell development; thymic aging

There are no references for this article.