Impact of Aging on Dendritic Cell Functions in humans

1.1 DC numbers and phenotype

Numerous studies have reported age-associated alterations in the functions of both CDCs and PDCs. Numbers or percentage of circulating DC and their phenotype have been measured by all groups using flow cytometry. The staining strategy utilized for identification of DC subsets is also similar (Lineage-HLADR+ CD11c+ are CDC and Lineage-HLADR+ CD123+ are PDC). Despite this, the results obtained by different groups are conflicting, Most studies find that the percentage of CDCs in the blood of the healthy aged subjects are comparable to young subjects (Agrawal et al., 2007; Jing et al., 2009) however, Della bella (Della bella et al., 2007) reported a decline in CDC percentage and absolute numbers in the blood of aged subjects. Even though all studies have been performed on healthy aged donors, only Della bella adhered stringently to SENEIUR protocol for subject selection. Variability in the donor population amongst different regions may also account for the different results. In contrast to healthy aged, a decline in CDC population has been reported in nursing home or frail aged subjects (Jing et al., 2009). There is consensus regarding MODCs numbers and all studies reported no alteration between aged and young subjects (Steger et al., 1996; Lung et al., 2000). We also observed comparable numbers of MODC between the two populations (Agrawal et al., 2007). The reports of the number of PDCs in aged subjects are conflicting; both normal and decreased numbers of PDCs have been observed (Shodell and Siegel et al, 2002; Perez-Cabeza et al., 2007; Jing et al., 2009; Canaday et al., 2010). In our own study with healthy aged, we did not observe any significant change in the frequency of CDCs or PDCs with age (Agrawal et al., 2007). Since all studies used similar methods for PDC enumeration, variability amongst the donor population seems to be logical to explain the observed differences.

There is a paucity of information regarding the status of age-associated changes in tissue DC populations which may play an important role in local pathology. A decrease in human Langerhans cell (LC) densities has been observed in the epidermis of the skin of aged subjects (Grewe, 2001). A similar decrease has also been reported in earlier studies with mice (Choi and Sauder, 1987; Sprecher et al., 1990). Bodineau et al (Bodineau et al., 2007; Bodineau et al., 2009) report an age-associated decrease in intraepithelial LC with chronic periodontitis. The dendritic processes are also smaller (reduced surface area) in the LCs from aged which may lead to reduced antigen presentation to T cells. The authors however, reported an upregulation of LAMP (Lysosomal-associated membrane protein) and CD83 (maturation marker for DCs) in the LCs from elderly. Such studies may be useful with DCs from other tissues or blood. In contrast to intraepithelial LCs, Bodineau et al observed an increased number of LCs in the upper epithelial connective tissue. Except for a study by Della bella (Della Bella et al., 2007) where an increased expression of CD83 and CD86 was observed on circulating DCs, the majority of studies (Steger et al., 1996; Lung et al., 2000; Agrawal et al., 2007) of circulating DCs or MODCs have reported comparable expression of costimulatory (CD40, CD80, CD86) and antigen presenting molecules (MHC-II) on DCs from aged and young subjects. Others (Haruna et al., 1995) have reported a decrease in HLA-DR expression on peripheral blood DCs and DCs from a strain of senescence-accelerated mouse (SAMP-1). Recently it was reported that the number of T regs were increased in lymph nodes of old mice (Zhao et al., 2007; Chiu et al., 2007). These T reg cells suppressed the expression of CD40 and CD86 co-stimulatory molecule expression in lymph node DCs and the expression could be restored to levels of young mice by inactivating Treg cells with anti-CD25 monoclonal antibodies. In summary, the majority of studies do not observe significant age-associated changes in CDC numbers and phenotype in human blood. Nevertheless, the picture may be entirely different for tissue specific DCs. PDC numbers on the contrary are generally observed to show a decrease with advancing age.

1.2 Expression of Pattern Recognition Receptors (PRR)

DCs sense and capture pathogens via the expression of pathogen recognition receptors (PRR) such as TLRs (Toll-like receptors), NLRs (NOD like receptors) and CLRs (C-type lectin receptors). Studies in aged have focused mainly on the expression and function of TLRs. Majority of investigators did not observe any age-related change in the expression of TLRs in MODCs. We also did not find any significant change in TLR4 expression in MODCs from aged (Agrawal et al., 2007). Recently, Panda et al (Panda et al., 2010) reported decreased expression of TLR1, 3 and 8 in CDCs from aged individuals; however, TLR2 levels were unchanged. They also observed reduced expression of TLR7 but normal expression of TLR9 in aged PDCs. Jing et al (Jing et al., 2009), on the other hand, reported a decreased expression of both TLR7 and TLR9 in PDCs from aged. In contrast to these studies, we did not observe any significant change in the expression of TLRs 7 and 9 in PDCs from aged subjects (unpublished observations). Most mouse studies also do not report an age-associated decline in TLR expression in CDC and PDCs (Tesar et al., 2007; Stout-Delgado et al., 2008).

Except for a study by Grollieu-Julius et al (Grolleau-Julius et al., 2008) who observed a decrease in DC-SIGN expression in bone marrow-derived DCs (BMDCs) from aged mice there is no information regarding the expression non-TLR PRRs in DCs in aging. This is an area of pivotal interest as an increasing number of non-TLR PRRs are being implicated in various pathological conditions. Furthermore, a number of new-generation vaccine adjuvants also target CLRs and NLRs (Ishii and Akira, 2007; Warshakoon et al., 2009; Ferwerda et al., 2010; Williams et al., 2010).

1.3 Phagocytosis or capture of antigens

Sensing of pathogens via PRRs leads to their capture and uptake by DCs resulting in DC maturation, activation, and subsequent T cell priming. Efficient antigen capture is thus central to generate effective immunity. Our observations demonstrate that both the phagocytosis of dextran beads and pinocytosis of the dye Lucifer Yellow are impaired in MODCs from aged subjects when compared to young (Agrawal et al., 2007). This suggests that MODCs from aged displayed reduced capacity to capture antigen via both receptor-dependent and receptor-independent mechanisms of antigen uptake. Furthermore, we identified that reduced activation of AKT kinase in the PI3-kinase signaling pathway may be responsible for the reduced antigen capture by aging DCs. Uptake and clearance of apoptotic cells by DCs plays an important role in the maintenance of peripheral self tolerance which is discussed in section (2.1). A recent study(Grolleau-Julius et al., 2006) in mice using BMDCs did not find reduction in the uptake of tumor antigens in old mice, though tumor antigen specific T cell stimulation was significantly reduced. However, these differences in mice and humans may be due to differences in the species.

1.4 Migration of DCs

Following activation, DCs migrate to T cell areas to induce effective cellular immune responses. Stimulation of immature DCs with TLR ligands results in the down-regulation of CCR6 and upregulation of CCR7, which enhances their ability to migrate from the peripheral tissues to the draining lymph node (Gunn, 2003). CCL21 and CCL19 both bind to the CCR7 receptor and are potent chemo attractants for mature DCs (Kellerman et al., 1999). Mice deficient in CCR7 are unable to mount an effective T cell immunity (Forster et al., 1999). Relatively few studies have addressed the question of migration of DCs in aging. We determined the migration of DCs from aged and young human subjects using a transwell system where the MODCs in the upper chamber migrate through a transwell of defined pore size to the lower chamber in response to a chemokine gradient (Agrawal et al., 2007). We observed that MODCs from aged subjects were impaired in their capacity to migrate in response to CCL19 and CXCL12 compared to MODCs from young, which was independent of CCR7 or CXCR4 expression since they were normally expressed. We also did not observe any significant difference in the basal level of the chemokines from MODCs in aged and young individuals. In contrast, Pietschmann et al (Pietschmann et al., 2000) observed normal trans-endothelial migration of peripheral blood myeloid-enriched lymphocyte-depleted cells in elderly subjects which could be due to the difference in the model system. Bhushan et al (Bhushan et al., 2002) also observed significantly impaired migration of LCs in response to TNF-α in elderly subjects, and aged mice. Choi and Sauder (Choi and Sauder, 1987) reported decreased LCs mobilization and the subsequent accumulation of DCs in the regional lymph nodes in aged mice in response to topical challenge with a chemical agent; however, contact hypersensitivity responses were not compromised. Linton et al (Linton et al., 2005) in a TCR transgenic mice model, have reported in vivo impaired migration of DCs from aged mice to the draining lymph nodes, which they suggested to be due to both intrinsic defect of DCs and aged microenvironment. A recent study by Toapanta et al (Toapanta et al., 2009) observed reduced CDC migration in the lungs of aged mice after infection with influenza. In summary, all studies (humans and mice) report impaired migration of aged DCs except for one study which utilized enriched DC population which may be contaminated by the presence of other cells. Since the migration of DCs to lymph nodes is pivotal to the establishment of the immune response, reduced migration may contribute to age-associated immune dysfunction.

1.5 DC Cytokine secretion

1.6 T cell priming by DCs

DCs are unique APCs because of their capacity to prime naïve T cell responses. DCs therefore function as initiators of T cell immunity. DCs can prime or tolerize T cells by virtue of their costimulatory molecule expression. DCs orchestrate the nature of T cell responses via two mechanisms: 1) via costimulation, and 2) via cytokine secretion. The sub-optimal costimulation of DCs leads to T cell anergy or generation of T regulatory cells. The expression of inhibitory costimulatory molecules such as PDL-1 (Programmed cell death 1 ligand-1), PDL-2 (Programmed cell death 1 ligand-2), ICOS L, (inducible T-cell co-stimulatory molecule), ILT-3 (Immunoglobulin-like transcript 3), PIR-B (Paired immunoglobulin-like receptor B), and DIgR2 (dendritic cell-derived immunoglobulin receptor 2) on DCs also impairs T cell priming. Cytokine secretion by DCs is dependent on the type of receptor being activated by a pathogen. For example, engagement of TLR4 and 5 on DCs induces IL-12 secretion (Agrawal et al., 2003), while engagement of TLR2/dectin induces T regulatory cells (Agrawal et al., 2006). The cytokines secreted by DCs aid in T cell priming and proliferation, and are now considered the third signal required for efficient T cell activation. In addition to priming, the nature of cytokines secreted by DCs in response to a pathogen dictates the Th0 polarization to Th1/Th2/Th17/ T regulatory.

Studies on T cell priming by DCs in humans are few, and generally report no significant difference in T cell priming by DCs in aged subjects. Steger et al (Steger et al., 1999) and Grewe et al (Grewe, 2001) reported that MODCs from young and aged subjects have similar stimulatory capacity to induce proliferation of T cell lines developed in long-term cultures. Our preliminary studies (unpublished) with MODCs show reduced proliferation of young T cells when co-cultured with aged DCs. These results may be different during an infection where decreased IFN-γ producing cells were observed in response to MODCs infected with respiratory syncytial virus (but not with influenza virus) from old individuals (Looney et al., 2002). Earlier studies in aged mice demonstrated a decrease in antigen presentation and T cell priming capacity of DCs in the lymph nodes (Donnini et al., 2002). However, data from two recent studies in mice are conflicting. A study by Tesar et al (Tesar et al., 2007) suggests an intrinsic defect in T cells during aging is responsible for the age-associated reduction in T cell functions. They observed that aged and young BMDCs stimulated with specific TLR ligands, petidoglycan (TLR 2/6), poly I:C (TLR 3), LPS (TLR 4), flagellin (TLR 5), and CpG) TLR 9) were not impaired in their capacity to upregulate co-stimulatory molecules CD40 and CD86, chemokine receptor-CCR7 or the production of pro-inflammatory cytokines IL-6 and TNF-α. Employing allogeneic and virus specific systems (LCMV), the authors determined the capacity of aged and young BMDCs to prime naïve T cells. Again they found that the capacity of both aged and young DCs to prime naïve T cells was comparable. However, when young DCs were cultured with aged and young T cells in the same model, they observed impaired T cell proliferation in aged T cells compared to young T cells suggesting that aged T cells are intrinsically defective. Wong et al (Wong et al., 2010) observed comparable levels of T cell stimulation by splenic DCs from aged and young mice. Grollaeu-julius (Grollaeu-julius et al., 2006) on the other hand reported that old bone marrow- derived DCs are less effective than young DCs in stimulating syngeneic, ova–specific CD4⁺ T-cell proliferation. They also reported a decrease in tumor regression in mice treated with the ovalbumin peptide-pulsed aged DCs than with ovalbumin peptide-pulsed young DCs. The discrepancy between the two studies may be due to differences in type of DCs; splenic verses BMDC.

PDCs are also capable of priming T cell responses. There is no information regarding their capacity to prime T cells in aging. This may be important since IFN-α secreted by PDC, which is reduced with age, is a critical factor in virus specific CD8 T cell responses.

2.1 Peripheral self Tolerance

The immune system undergoes continuous morphologic and functional changes throughout the years, characterized by a gradual decline in its function with age. Paradoxically, this decline in immune functions is associated with increased reactivity towards self or endogenous antigens, as evident by an increased frequency and increased titers of auto-antibodies with age (Franceschi et al., 1995; Lee and Weksler, 2001; Weksler and Goodhardt, 2002). Gebe et al (Gebe et al., 2006) have identified an age-dependent spontaneous loss of tolerance to two self-antigenic epitopes derived from putative diabetes-associated antigens glutamic acid decarboxylase (GAD65) and glial fibrillary acidic protein (GFAP) in humanized murine model, RIP-B7/DRB1*0404 HLA transgenic mice. They observed that diabetic and older non-diabetic mice exhibited a proliferative response to an immunodominant epitope from GAD65 and GFAP. The response to both of these self-antigens was not observed in young mice. A similar loss of oral tolerance to ova has been observed in aged mice. Faria et al (Faria et al., 1998) observed that oral tolerance to OVA could be induced in young mice at 20 weeks of age; however, oral tolerance induction was completely abolished in mice by 38 weeks of age. In general, it is believed that oral tolerance established in early age can be maintained despite aging, whereas the induction of oral tolerance to new antigens is impaired in aged mice (Faria et al., 1993; Wakabayashi et al., 1999; Kato et al., 2003; Fujihashi and Kiyono, 2009).

Most of our knowledge regarding maintenance of tolerance in aging is obtained from studies in mice; only few investigators, including ourselves, have studied tolerance in aged humans (Agrawal et al., 2009).. Apoptosis is a complex physiological programmed cell-death process, which plays an important role in certain effector functions, and immune homeostasis. One of the critical steps in apoptosis is rapid and efficient removal of apoptotic bodies and apoptotic cells by neighboring phagocytic cells, including DCs (Savill et al., 2002; Gaipl et al., 2003; Luckashenak et al., 2008; Mueller, 2010). Increasing evidence suggests that apoptosis is involved in many pathological conditions (Mevorach et al., 1998; Hermann et al., 1998; Cohen et al., 2002; Hardin, 2003; Shosan and Mevorach, 2004; Tisch R, Wang B, 2008; Ohnmacht et al., 2009). Uptake and ingestion of apoptotic cells by dendritic cells is considered to be one of the major mechanisms employed by DCs in inducing peripheral self tolerance. Under steady state, immature dendritic cells continuously sample self-antigens from apoptotic cells resulting in peripheral T cell tolerance. We have shown that DCs from aged subjects are impaired in their capacity to phagocytose antigens (section 1.3) including apoptotic cells. Apoptotic cell death and clearance of apoptotic cells and apoptotic bodies is of vital importance in developing and maintaining the normal tissue homeostasis and resolution of inflammation (Savill et al., 2002; Gaipl et al., 2003). The quick and efficient clearance of apoptotic cells is essential to prevent secondary necrosis in tissues (Sauter et al., 2000). This is of particular importance because the late phase of apoptotic cells is associated with additional proteolytic degradation of specific auto-antigens, which may release endogenous danger signals like nuclear structures clustered in apoptotic blebs (chromatin and dsDNA) and proteins such as heat-shock proteins (HSPs), resulting in maturation of dendritic cells, and T cell immunity to self and stimulating autoantibody responses (Okabe et al., 2005). Impaired clearance of apoptotic cells has been implicated in various autoimmune disorders like lupus and rheumatoid arthritis (Mevorach et al., 1998; Hermann et al., 1998; Cohen et al., 2002; Hardin, 2003; Shosan and Mevorach, 2004). To demonstrate that DCs from aged humans show increased reactivity to self antigens released during apoptosis, we purified DNA from human blood, delivered it intracellularly and compared the activation and cytokine secretion of aged and young DCs. MODCs from aged secrete increased quantities of IL-6 and IFN-α compared to MODCs from young subjects. A similar increase in cytokines was observed from aged DCs in response to late apoptotic cells. Furthermore, DNA primed DCs from aged induced increased T cell proliferation compared to young subjects. We further demonstrated that DCs from aged humans display increased reactivity to self antigens because there is increased basal level of NFκB activation which leads to increased IRF3 activation and subsequent IFN-α secretion. Increased basal level of NFκB suggests that DCs from aged exist in a semi-mature state causing increased reactivity to self antigens resulting in elevated cytokine levels even at the basal level, which contributes to loss of peripheral self tolerance and chronic inflammation observed during aging (Figure 1). Other investigators have also reported increased secretion of pro-inflammatory cytokines from unactivated CDCs of aged subjects (Della bella et al., 2007; Panda et al., 2010). Furthermore, a study by Benito et al (Benito et al., 2008) reports that renal graft from young but not from aged donors seem to induce DCs of a tolerogenic phenotype, both in aged and young recipients, suggesting that donor age may have consequences in terms of tolerance induction.

More recently we have shown that the DNA from aged subjects is more immunogenic than DNA from young, thus DNA released by late apoptotic cells in aging induce increased amounts of cytokine IFN-a from DCs (Agrawal et al., 2010). This suggests that in addition to increased reactivity of DCs to self antigens, changes in self antigens may also aid in loss of self tolerance associated with aging.

2.3 Central Tolerance

In addition to oral and peripheral tolerance, DCs have a critical role in central tolerance in the thymus. Thymic DCs play an essential role in shaping T cell-mediated immune responses by deleting self-reactive thymocytes, and by inducing regulatory T-cell (Treg) development to establish central tolerance (Klein et al., 2009). Involution of thymus results in decreased output of naïve T cells during aging. However, the effect of thymic involution on the function of thymic DCs is not well known. Earlier studies (Nabarra and Andrianarison, 1996; Varas et al., 2003) reported an age-associated decline in medullary DC numbers. Furthermore, Varas et al observed decreased expression of antigen-presenting and costimulatory molecules HLA-DR, CD54, CD86 and CD40 in thymic DCs, which was associated with an impaired function of DCs from aged to induce proliferation of alloreactive T cells as compared to their young counterparts. Others have reported that the function of thymic DCs to negatively select autoreactive T cell is impaired only in very old mice (30–36 months). More recently van Dommelen et al (van Dommelen et al., 2010) also reported decreased numbers of thymic DCs in aged mice relative to young mice. They further showed that it was possible to restore the number of aged DCs to the level of that in young when the thymus was regenerated using sex steroid ablation therapy. Thus it seems that although the numbers of thymic DCs decrease with age, it does not have a major impact on central tolerance.

Recently, Tomohisha Baba (Baba et al., 2009) and colleague have identified a unique role of thymic Sirpα⁺ dendritic cell subpopulation in the establishment of central tolerance. It would be of interest to examine changes in Sirpα expression and function of Sirpα+ DCs in relation to aging.