Aging affects epidermal Langerhans cell development and function and alters their miRNA gene expression profile.

Immunosenescence is a result of progressive decline in immune system function with advancing age. Epidermal Langerhans cells (LCs), belonging to the dendritic cell (DC) family, act as sentinels to play key roles in the skin immune responses. However, it has not been fully elucidated how aging affects development and function of LCs. Here, we systemically analyzed LC development and function during the aging process in C57BL/6J mice, and performed global microRNA (miRNA) gene expression profiles in aged and young LCs. We found that the frequency and maturation of epidermal LCs were significantly reduced in aged mice starting at 12 months of age, while the Langerin expression and ability to phagocytose Dextran in aged LCs were increased compared to LCs from < 6 month old mice. The migration of LCs to draining lymph nodes was comparable between aged and young mice. Functionally, aged LCs were impaired in their capacity to induce OVA-specific CD4+ and CD8+ T cell proliferation. Furthermore, the expression of miRNAs in aged epidermal LCs showed a distinct profile compared to young LCs. Most interestingly, aging-regulated miRNAs potentially target TGF-β-dependent and non- TGF-β-dependent signal pathways related to LCs. Overall, our data suggests that aging affects LCs development and function, and that age-regulated miRNAs may contribute to the LC developmental and functional changes in aging.

INTRODUCTION neutrophils, natural killer (NK) cells, NKT, and macrophages, which are all important first lines of defense against bacterial and parasitic infections, are defective in aged mice and human [3][4][5]. DCs are professional antigen-presenting cells (APCs) that are important in the initial priming of naïve T cells and act as a bridge to link innate immunity and adaptive immunity. Several studies indicate that the number of DCs does not change with age [6,7]. However, Schodel et al. found a decrease in circulating plasmocytoid DCs numbers but observed no change in the number of myloid DCs [8]. There is also some controversy regarding the capacity of aged DCs to stimulated T cells [8,9]. Epidermal LCs are skin-resident DCs with a life cycle distinct from other types of DCs. They play a key role in the development of cutaneous immune responses. Previous studies have speculated that agerelated changes in epidermal LCs frequency and/or migration patterns results in impaired immunosurveillance in human [10]. However, the precise effects of aging on LCs are still not well understood.
Aging is a multifactorial process where deterioration of organism functions is driven by distinct genetically encoded genes. To better understand the genetic component of aging, many studies have addressed the distinct gene and protein expression profiles in different aging cells. microRNAs (miRNAs), small non-coding RNAs, have recently emerged as key regulators of gene expression through mRNA degradation, and/or translation inhibition [11]. The ribonuclease III enzyme Dicer is required for the processing of mature miRNAs. Therefore, deletion of Dicer provides a good genetic tool to test the relevance of miRNAs in mammalian development. Using a mouse model with the tissuespecific deletion of Dicer, studies from our group and others indicate that lack of miRNAs by Dicer-deletion dramatically interrupts T and B cell development and function [12][13][14]. Interestingly, deletion of Dicer specifically in DCs only affects the homeostasis and function of LCs, but not other types of DCs. This suggests that miRNAs are required for normal LCs development and function [15]. Importantly, miRNAs are involved in aging-mediated T cell and macrophage defects [16,17], and play an important role in aging process and aging-related diseases [18][19][20]. However, it is unclear whether aging affects miRNA gene expression in LCs, which may contribute to agingmediated LC development and function.
In the current study, we systemically investigated the development and function of LCs during aging using C57BL/6J mice and performed miRNA expression profiles of LCs between aged and young mice. Our results demonstrate that the frequency and maturation of epidermal LCs were reduced in a stepwise fashion in aged mice, starting at 12 months old. Aged LCs exhibited a reduced ability to stimulate T cell proliferation. However, aged LCs migration toward skin draining lymph nodes was not altered compared to young LCs. Most important of all, genome-wide assessment of miRNA expression in epidermal LCs revealed a distinct aging-specific miRNA gene expression profile. Taken together, our findings further confirm the defects of LCs development and function, and suggest the involvement of miRNAs in aged LCs.

Decreased frequencies of epidermal LCs in aged mice
The density of LCs in the epidermis is known to decrease with age in mice [21]. However, the kinetics of LCs reduction with age has not been analyzed yet. To determine the effect of age on epidermal LCs homeostasis, we examined the frequency of skin LCs at different time points. The frequency of epidermal LCs (Langerin + CD45.2 + ) was 1.60 ± 0.09% in < 6 months old mice and was significantly reduced to 1.32 ± 0.02% in 12 months old mice and 1.13±0.10% in 18 months old mice. There was no significant difference between 12 months and 18 months old mice ( Figure 1 A and B). In consistence with reduced epidermal LCs, the proportion of Langerin + EpCAM + cells in the gated MHCII + CD8cells of LN, which represents the migrant epidermal LCs, decreased 2-3-fold in 12 and 18 months old mice compared with 6 months old mice (Figure 1 C). Interestingly, the frequency of dendritic epidermal T cells (DETC) was comparable between 6 and 12 month old mice but significantly decreased at 18 months ( Figure 1 A and B). A previous study pointed out that the frequency of conventional DCs was reduced in 20-22 months of age mice [22]. Thus, the frequency of epidermal LCs is reduced earlier compared to other immune cells during the aging process. In subsequent studies, we used mice more than 12 months of age as aged mice for our LCs studies.
The markedly reduced frequency of LCs in aged mice could be explained by either a decrease in the LC proliferation, increased LC apoptosis, or by a decrease in the number of LCs progenitors. To study the first possibility, we preformed anti-Ki67 antibody staining as a way of measuring the kinetics of LCs turnover. We found no difference in the lifespan of LCs in aged and young mice (  www.impactaging.com

Epidermal LCs from aged mice exhibit less ability to up-regulate MHCII and co-stimulatory molecules during in vitro maturation
The initiation, execution, and sustainment of T cell immunity by DCs require cell-cell contact mediated by costimulatory molecules on the DCs [23]. Immature LCs in the epidermis mature in response to various stimuli, and then migrate to draining LNs to prime T cells [24]. Hence, we investigated whether aging can affect epidermal LCs maturation in vitro. Epidermal cells derived from aged and young mice were cultured for 48h in vitro, and then stained with anti-CD80, CD86 and MHC-II, and assessed by FACS. As shown in Fig    www.impactaging.com

Aging LCs have an increased ability to phagocytize antigen but do not display altered allergen-induced migration
LCs act as sentinels for contact with various environmental antigens. They capture and process exogenous antigens and migrate to the skin LNs where they activate T lymphocytes. Therefore we investigated whether aging can affect LC phagocytosis and migration. Freshly isolated LCs from aged and young mice were incubated in vitro, for 45 minutes at 37℃ or 4℃ (control) with Dextran-FITC and then stained with anti-MHCII and anti-CD45.2 mAb. As shown in Figure  3, LCs from aged mice had increased phagocytic capacity compared to that from young mice based on the ratio of FITC-positive (Figure 3 A) or mean fluorescence intensity expression levels (Figure 3 B). This is the additional example of hyper-functions associated with aging as discussed recently [25,26]. Previous studies [27,28] suggested that Langerin is the main molecular component of Birbeck granules and plays a substantial role in antigen uptake. Thus, we further investigated if aging affects Langerin expression. Interestingly, consistent with increased antigen uptake, we found increased Langerin expression in aged LCs (Figure 3 C).
To test whether or not the migratory capability of epidermal LCs was affected in vivo by aging, FITC was applied to the belly of young and aged mice as a tracer. After 24h, LCs from inguinal LN were harvested and the frequency of FITC + LCs was analyzed. Recent studies revealed the presence of at least 3 distinct skin DC subsets in LN based on the differential expression of CD8, EpCAM and Langerin [29], with migrant epidermal LCs being Langerin + EpCAM + in the gated MHCII + CD8population. As shown in Figure 3 D, the frequencies of Langerin + FITC + LCs were comparable between aged and young mice. Together, these data indicated that phagocytic capacity increased with aging, however, the migration ability of aged LCs did not alter compared to young mice.

Aging impaired LCs-mediated T cell proliferation
LCs play a key role in mediating antigen-specfic T cell proliferation. Upregulation of MHCII and costimulatory molecules on LCs are required for T cell activation [30]. Given the in vitro downregulation of MHCII, CD80 and CD86 on aged LCs, we expected that aged LCs would have a reduced ability to active T cells. To test this, epidermal LCs were first pulsed with OVA protein and sorted LCs were then cultured with OVA-specific CD8 + OT-I or CD4 + OT-II T cells for 72h. As shown in Figure 4, aged LCs-mediated CD8 + OT-I cell and CD4 + OT-II T cell proliferation were dramatically reduced compared to young LCs, Thus, our data further suggest the defective LC-mediated T cell activation in aged mice.

Distinct miRNA expression profiles in aged epidermal LCs
A recent study showed that deletion of miRNA precursor-processing enzyme Dicer in DC precursors interrupted the homeostasis and function of LCs. This suggested a key role of miRNAs in LC development and function [15]. Interestingly, miRNAs have also been shown to regulate the aging processes in different tissue and cells [31]. To test if the miRNAs are potentially involved in aged LCs development and function, the global miRNA expression profiles of sorted LCs from aged and young mice ( Figure 5 A) were performed by TaqMan qPCR-based arrays ( Figure 5 B). Using global normalization [32], we have identified 53 age-regulated   Aging regulated miRNAs potentially target TGF-βdependent and non-TGF-β-dependent pathways that control LC development and function during aging. To explore the potential targets and pathways that are regulated by aging-associated miRNAs in LCs, we used the Target Scan Mouse 6.2 database to predict miRNA targets related to LC homeostasis and function ( Table  1). As shown in Table 1, Gfi1, IRF8, RunX3, C/EBP, TGFβRII, TGFβ1, AhR, RANKL, Langerin, and CSF1R are shown to regulate the homeostasis, development and function of epidermal LCs [27,[33][34][35][36][37][38][39][40], and are predictive targets by age-associated miRNAs. A hypothetical model of age-dependent miRNAs regulating LCs development and function is shown in Figure 6.

DISCUSSIONS
DCs are critical immune players involved in the induction of immunity and tolerance. Age-associated changes in DC development and function significantly compromise the immune system and directly impact both adaptive and innate immunity in the elderly [41]. LCs, part of the DC family, form a network within the epidermis where they play key roles in immunosurveillance for cutaneous malignancy, infection, allergy and autoimmune diseases [42]. We have demonstrated here that, in steady-state, the frequency of epidermal LCs is significantly reduced starting at 12 months old mice compared with that of young mice. This observation is in consistence with previous reports in mice [21]. Interestingly, LC survival and proliferation in the epidermis were not affected by aging. This finding suggests that the decreased LC density found in aged mice may result from a deficiency in bone marrow-derived LC progenitors, or that LC progenitors are less responsive to the chemokine and cytokine signals that are known to be required for the recruitment of LC progenitors into the skin [21,43].

Effective migration of LCs into lymph nodes is necessary for LCs to induce T cell-mediated immune responses.
Previous studies reported that migration of LCs in response to TNFa is significantly impaired in elderly subjects, but responses stimulated by similar exposure to IL-1β are unimpaired in the elderly [10,44]. In contrast, Ogden et al. found that migration of LCs in response to a chemokine ligand, CCL19, was equivalent in both these groups [45]. Our data indicate that the migration of FITC-bearing LCs to skin draining LN was comparable between aged and young mice. However, the total LCs in draining LN was significantly decreased during aging (Figure 1 C), which could be related to reduced epidermal LCs. Interestingly, the ability of respiratory DCs to migrate to dLNs is compromised in aged mice with a decline in migration occurring as early as 6 months of age [9]. This suggests that different DC subsets may respond to aging differently.
The functional abilities of aged DCs in mice have primarily been generated using in vitro differentiated BMDCs. In some studies, BMDCs from aged mice had impaired maturation, antigen-presenting function, and reduced ability to stimulate T cell response [46,47]. However, others have shown that the T cell priming ability of aged BMDCs was equivalent to their younger counterparts [6]. Whether the results generated from BMDCs can be extrapolated to LC functions in aged mice remains unclear. In our study, we found that LCs from aged mice appear to have a lower ability to upregulate MHCII and co-stimulatory molecules including CD80 and CD86 compared to LCs from young mice after in vitro maturation. These molecules are known to play an important role in T cell priming. Thus, as expected, our data show that LCs from old mice are less effective in activating MHC class I-restricted CD8 + T cells as well as MHC class II-restricted CD4 + T cells. This finding is consistent with previous findings in other aging DC subsets such as respiratory DCs and mucosal DCs [9,48]. Reduced LC frequency in aged mice combined with their functionally defective of priming T cell immunity could contribute to age-related skin diseases.
Aging is a multifactorial process where deterioration of an organisms function is driven by distinct genetically encoded systems [49]. In recent years, more and more evidence is emerging that miRNAs are key players in the regulation of a wide range of biological functions in a variety of mammalian cell types, including immune cells [12,50,51], and that miRNAs have participated in the complex networks of cellular senescene and contribute to aging [31,[52][53][54] . The deletion of Dicer, an enzyme required for the processing of mature and functional miRNAs, provides a genetic test for the relevance of miRNAs in mammalian development. We (unpublished data) and others [15] recently found that Dicer deletion interrupted the homeostasis and function of LCs but not other DCs subsets, suggesting a specific role of miRNAs in LCs development and function. However, most individual miRNA effects on LCs development and function are unknown. We recently found miR-150 may regulate LC-mediated crosspresentation [55,56]. Here, we identified specific agerelated miRNA gene expression profiles in LCs, which could contribute to aged LCs homeostasis and function. TGF-β is well known for its essential role in LCs development and maturation [37,57]. The importance of TGF-β signaling in LCs lineage decisions has been demonstrated in RunX3 knockout (KO) and Id2KO mice. Deletion of these two signaling molecules downstream of the TGFβR abolished LC development [35,58]. Reciprocal roles for C/EBP and PU.1 in LC development have also been identified [36,59].
www.impactaging.com miRNAs miR-449 and miR-9 potentially target TGFβ1, TGFβRI, TGFβRII, RunX3 and C/EBP, which are involved in TGF-β signaling ( Figure 6). Thus, upregulated miR-449 and miR-9 in aged LCs could downregulate the TGF-β signaling pathway and block LC development. In addition, non-TGF-β signaling via CSF1/CSFR and IRF8 may also play an important role in LC homeostasis and development. IRFKO and CSFRKO mice both exhibit deficiencies in the frequency of LCs ( Figure 6) [34,40]. Interestingly, miR-709 and miR-449 also target IFR8 and CSF1R. Thus, upregualated miR-709 and miR-449 in aging LCs may downregulate the expression of IFR and CSFR, causing a deficiency in LCs development in aging mice. Recent reports indicated that Gfi1 plays a key role in LC homeostasis and Gfi1KO mice increased the frequency of LCs [33]. Gfi1 is a potential target of miR-200. Downregulated miR-200 in aged LCs may upregulate its target, Gfi1, which could then arrest LCs commitment. The RANKL/RANK system has been shown to play a key role in LCs homeostasis and maturation, and RANKL-deficient mice displayed a marked reduction of epidermal LCs [39,60]. Interestingly, RANKL/RANK are putative targets of miR-9 or miR-20. Valladeau et al. previously reported that Langerin was the main molecular component of Birbeck granules, and that Langerin plays an important role in antigen uptake [27,28]. Our miRNA array data showed that the expression level of miR-200 that potentially targets Langerin, was downregualted in aged LCs. This may account for the high expression of Langerin and the enhanced antigen-uptake ability in aging LCs. Overall, aging-regulated miRNAs could impair aged LCs development, maturation, and function by targeting different signaling pathways. To our knowledge, this is the first report to demonstrate that aging regulates miRNA gene expression in LCs.
In summary, our results demonstrate that the number and function of epidermal LCs are changed during aging development. Our comprehensive miRNA profile analyses on epidermal LCs uncover specific miRNA signatures in aged LCs, which could target LC-related signaling pathways and underlie impaired age-related LCs developmental and functional defects. Future work is needed to determine the role of specific ageassociated miRNAs and their targets in LCs development and function, which will not only help for fully understanding aging LC biology, but also for LCsrelated therapeutic strategies.

METHODS
Animals. Young (< 6 months) and aged (> 12 months) C57BL/6J mice, as well as OVA-specific MHCI-restricted and MHCII-restricted T cell receptor(TCR)transgenic C57BL/6J mice (OT-I and OT-II) were purchased from Jackson Laboratory(Bar Harbor, ME), and were bred and housed in a specific pathogen-free barrier unit. Handing of mice and experimental procedures were in accordance with requirements of the Institutional Animal Care and Use Committee.
Cell suspension preparations and Flow cytometry. Whole body skin from young (< 6 months) and aged (>12 months) mice was floated on dispase (2.5mg/ml) for 50 min in a 37℃ incubator. The epidermis was separated and then digested with typsin (0.05%) and Dnase (0.1mg/ml) for 15 min in a shaking bath. Cells were filtered through a 70µm strainer to obtain single cell suspensions. Real-time RT-PCR for miRNAs expression. The expression levels of miRNAs were examined using the Applied Biosystems TaqMan MicroRNA Assay kit (Applied Biosystems, Foster, CA) according to the manufacturer's instructions. The kit uses gene-specific stem-loop reverse transcription primers and TaqMan probes to detect mature miRNA transcripts. PCR amplification was carried out on the Applied Biosystems 7900 Real-time PCR system. The assay was run in triplicate for each case to allow for assessment of technical variability. SnoRNA135 was used as endogenous control. Relative quantitations using the ∆∆Ct method in aged versus young mice were carried out and fold changes were calculated by 2 −∆∆Ct for each miRNA gene.
Statistical analysis. Statistical analysis was performed with Prism5.0 (Graphpad Software). The two-tailed Student t test was used. Differences were considered statistically significant when values of p < 0.05 were obtained.