Plasmacytoid dendritic cells (pDCs) are innate sensors that produce IFN-α in response to viral infections. Determining how aging alters the cellular and molecular function of these cells may provide an explanation of increased susceptibility of older people to viral infections. Hence, we examined whether aging critically impairs pDC function during infection with HSV-2, a viral pathogen that activates TLR9. We found that impaired IFN-α production by aged murine pDCs led to impaired viral clearance with aging. Upon TLR9 activation, aged pDCs displayed defective up-regulation of IFN-regulatory factor 7, a key adaptor in the type I IFN pathway, as compared with younger counterparts. Aged pDCs had more oxidative stress, and reducing oxidative stress in aged pDCs partly recovered the age-induced IFN-α defect during TLR9 activation. In sum, aging impairs the type I IFN pathway in pDCs, and this alteration may contribute to the increased susceptibility of older people to certain viral infections.
Clinical studies have demonstrated that aging is associated with an increased susceptibility to pathogens, in particular viruses, implying that aging negatively impacts immunity. Consequently, older people exhibit an increased incidence of morbidity and mortality from infections (1). According to previous work, aging impairs adaptive T cell function (2, 3, 4, 5); however, the effect of aging on the innate immune system is less clear. Previous studies reported conflicting results about whether cellular components of the innate system, such as conventional dendritic cells (DC),3 macrophages, or NK cells are altered with age (4, 6, 7). Because innate immunity is the first line of defense against pathogens, understanding how aging modifies this defense is imperative, given that this information could be critical for the development of therapies to augment immunity in older people.
Plasmacytoid DCs (pDC) are key cellular components of the innate immune system that secrete high levels of type I IFNs, such as IFN-α, and IL-12 in response to certain bacteria (8) and viruses (9, 10, 11, 12). In contrast to conventional myeloid DCs, which express many TLRs, pDCs express TLR7 and TLR9 (13, 14). The production of type I IFNs and IL-12 occurs via separate pathways; the adaptor IFN regulatory factor (IRF)-7 is essential for type I IFN production, whereas IL-12 production is dependent on NF-κB translocation (15). Activation of TLR7 and TLR9 receptors within the endosomes of pDCs is critical for the recognition of ssRNA via TLR7 (16, 17, 18, 19) and unmethylated CpG sequences via TLR9 (14, 20, 21, 22). The high levels of type 1 IFNs produced by TLR-activated pDCs (21, 23) activate cellular host innate defense mechanisms and initiate adaptive immunity (8, 12, 24, 25). Hence, pDCs are critical for host defense against many viruses.
Prior studies indicate that aging impairs TLR immune responses in macrophages in vitro (26, 27), whereas TLR immune responses are preserved in conventional DCs (28). Regarding pDCs and aging, one clinical study reported an association between increased age and reduced circulating pDC numbers (29). Additionally, decreased IFN-α responses were observed during ex vivo challenge of PBMCs and purified pDCs from people over the age of 65 years when compared with young counterparts (30, 31); however, whether altered pDC function with aging impacts the in vivo response to viral infection is not known. Furthermore, the underlying molecular defects in pDCs that occur with aging have not been examined. One of the leading theories of age-induced cellular damage is the accumulation of damaging reactive oxygen species (ROS; Refs. 32 and 33), but whether increased ROS levels are important for any age-induced phenotype in pDCs remains unclear.
In this study, we examined the effect of aging on pDC function and any resultant effects on host defense to viral infection. We used a HSV-2 infection as an experimental murine model to activate TLR9 in pDCs (21). We demonstrated that aging impaired systemic IFN-α production during viral infection and subsequently resulted in impaired viral clearance, a phenotype that was reversed by the adoptive transfer of young pDCs into aged recipients. Gene and protein analysis demonstrated that aged pDCs exhibited an impaired ability to up-regulate IRF-7 and PI3K during either TLR9 or IFN-αβ receptor activation. These findings were associated with increased oxidative stress during TLR9 activation in aged pDCs as compared with young cells. Importantly, reduction of age-induced oxidative stress led to augmented IFN-α production in TLR9-activated aged pDCs. In conclusion, aging leads to increased oxidative stress and decreased up-regulation of IRF-7 within pDCs, and these alterations may be one of the mechanisms by which aging leads to defective host defense to viral infection.
Materials and Methods
Specific pathogen-free young (2–4 mo of age) and aged (18–20 mo of age) CBA (H2k) or C57BL/6 (H2b) calorie-restricted (CR) or ad libitum-fed (AL) mice were acquired from the National Institute on Aging rodent facility (Baltimore, MD). CR mice were fed 40% of a National Institutes of Health-31 diet from 3 mo of age. Upon arrival, mice were continued on their respective diets (AL or CR). No animals were used in the study if they had evidence of skin lesions, weight loss, or lymphadenopathy. This study was approved by the Yale University Institutional Animal Care and Use Committee.
HSV-2 and murine CMV (MCMV) viral propagation and preparation
HSV-2 (strain 186TKΔKpn; Ref. 21), which was generously provided by Dr. Akiko Iwasaki (Yale University, New Haven, CT), was propagated in Vero cells as previously described (21, 34). Virus was purified by centrifugation and was resuspended in PBS before use. The virus was administered to mice via i.v. (tail vein) injection at 1 × 107 PFU/mouse. In experiments comparing young vs aged mice, viral dosages were adjusted to body weight. Bone marrow purified pDCs were infected with HSV-2 at 1 × 103 PFU or PBS as a control, and IFN-α levels in supernatants were measured 24 h after infection. HSV-2 viremia in livers and spleens was measured by plaque assay as previously described (21, 34). MCMV Smith strain was generously provided by Anthony Van Den Pol (Yale University). The virus was administered to mice via i.p. injection of 2 × 105 PFU/mouse as previously described (24). MCMV viremia in livers and spleens was measured with a plaque assay as previously described (35).
Reagents, in vitro culture, and adoptive transfer of bone marrow cells
The CpG-A sequence used for in vitro and in vivo pDC stimulation (CpG-oligodeoxynucleoide (ODN) 5′-ctattggaaaaCGttcttCGgggcG-3′; Ref. 36) was synthesized by MWG Biotech. For in vitro and in vivo administration, various concentrations of the CpG-A sequence were diluted in PBS and then added dropwise to 1,2-dioleoyl-3-(trimethyammonium)propane (DOTAP) methosulfate (1 μg of CpG-A DNA per 5 μg of DOTAP; Sigma-Aldrich) to form a liposome-DNA complex. For in vitro experiments, CpG-A (10 μg/ml or indicated dose) was added to 3–5 × 105 cells in a total volume of 200 μl/well for 18 h. For in vivo experiments, CpG-A (5 μg/mg body weight) were administered via i.v. (tail vein) injection. RNA40 was administered at 50 μg/ml for in vitro experiments (Invitrogen). In addition, 1 μM N-acetylcysteine (NAC) was added to cultures where indicated, as previously described (37, 38). Anti-PDCA-1 Ab (500 μg/mouse; Miltenyi Biotec) or control Ab was administered via i.p. injection 12 h before HSV-2 administration. For adoptive transfer experiments, 5 × 106 bone marrow cells or 1 × 104 purified pDCs were transferred via i.v. (tail vein) injection. Murine rIFN-α (PBL Laboratories) was used in vitro at a dose of 10 U/well.
Primary bone marrow cell isolation and purification
Bone marrow cells were prepared from the femurs and tibias of mice. The pDCs were purified by positive selection by incubating cells with magnetic microbeads that were coupled to a PDCA-1 mAb (Miltenyi Biotec) and then passing the suspension through a MACS LS column according to the manufacturer’s instructions (Miltenyi Biotec). We routinely obtained a purity of >90% PDCA-1+CD11clow cells according to flow cytometry analysis. A similar procedure with microbeads coupled to PDCA-1 mAb was used to exclude cells except that depletion columns were used (Miltenyi Biotec). Flow cytometric analysis demonstrated that the resulting solution contained <0.3% pDCs.
Serum IFN-α (PBL Biomedical Laboratories) and IL-12p40 (R&D Systems) levels were assessed by ELISA according to the manufacturer’s instructions.
IRF-7 activity ELISA
Nuclear extracts were purified via a nuclear extraction kit (Active Motif), and IRF-7 activity was measured by ELISA using the TransAM IRF-7 kit, according to the manufacturer’s instructions (Active Motif). Results are expressed as optical density per microgram of nuclear extract.
RNA samples were extracted using the Qiagen RNAeasy mini kit (Qiagen) and quantified by OD260/OD280 readings. Superscript III reverse transcriptase (Invitrogen) was used to create cDNA from mRNA according to the manufacturer’s instructions. The RT-PCR primer sets were designed with Primer Express software (Applied Biosystems). RT-PCR was performed in a final volume of 25 μl containing complementary DNA from 20 ng of reverse-transcribed total RNA, 150 nM forward and reverse primers, and SYBR Green universal PCR master mix (Applied Biosystems). PCR was performed in 96-well plates with the MJ Research detection system (MJ Research). All reactions were performed in triplicate. The melting curve and the dilution curve standards were analyzed to identify primer sets and conditions yielding specific products with 100% amplification efficiency. The primer sequences were as follows: IRF-7: forward primer, acacttcctcatggacctgg; reverse primer, ttcccacttcccattctgag; and PI3K: forward primer, tgtcagatgaggaggctgtg; reverse primer, gggtcaaatcccctttcatt. Relative levels of mRNA were calculated by the comparative cycle threshold method (User Bulletin No. 2; Applied Biosystems). The 18S RNA mRNA levels were used as the invariant control for each sample. Gene expression analysis was performed after 18 h of stimulation, given that this time point exhibited the greatest differences in IFN-α between young and aged pDCs (Fig. 1 D).
Cells were washed, and cytosolic and nuclear protein fractions were isolated using the NE-PER Nuclear and Cytoplasmic Extraction kit (Pierce Biotechnology) according to the manufacturer’s protocol. Protein concentrations were determined using the Bio-Rad protein assay reagents (Bio-Rad Laboratories) with BSA as the standard. Equal amounts of total cytosolic or nuclear protein (10 μg/lane) were separated on 10% bisacrylamide gels by SDS-PAGE and electrophoretically transferred to nitrocellulose in NuPage buffer (Invitrogen). Immunodetection was performed using primary anti-IRF-7, anti-IκB kinase (IKK)α mAbs (Santa Cruz Biologicals) or rabbit polyclonal anti-mouse Abs to the phosphorylated site of serine 124 of AKT (denoted as phosphorylated AKT; Abcam) and secondary HRP-conjugated anti-goat Ig. The ECL Western Blotting Analysis System was used for detection of the signal. For IRF-7 activation, detectable protein levels were detected 1 h post-CpG-A activation, which peaked after 4 h and were maintained up to 18 h (data not shown). Eighteen hours after CpG-A activation was chosen for assessment of differences in IRF-7 levels between young and aged pDCs, because this time point demonstrated the largest IFN-α differences between young and aged pDCs (Fig. 1 D).
Cells were stained with PE-conjugated rat anti-mPDCA-1 (Miltenyi Biotec) and FITC-conjugated anti-CD11c (BD Biosciences). In agreement with prior work (39), we confirmed that anti-mPDCA-1 identified CD11clow/intermediateB220high cells (data not shown). For all staining, incubation was performed for 30 min at 4°C, and a second incubation step with streptavidin PerCP was performed for 15 min at 4°C. Isotype control Abs were used in every experiment. For staining with dihydroethidium (DHE), which is rapidly oxidized by ROS into its fluorescent derivative (40), cells were incubated for 15 min at 37°C in culture medium containing a 1/100 dilution of a 10 μM solution of DHE in DMSO (Invitrogen). Fluorescence data were acquired with a FACSCalibur flow cytometer and analyzed with FlowJo software (Treestar).
Statistical analysis was performed using GraphPad Prism Software. Consistent results were noted in repeat experiments, allowing pooling of data and subsequent comparisons of means by a two-way Student t test. Statistical significance is indicated as a p value of <0.05.
Aging impairs the ability of pDCs to produce IFN-α in response to in vitro CpG-A activation
Previous work has demonstrated that oligonucleotides containing type-A CpG sequences (CpG-A) induce the production of IFN-α in a TLR9-dependent manner (36). We first examined whether aging altered the above response by isolating pDCs from the bone marrow of young (2–4 mo of age) and aged (18–20 mo of age) mice and by stimulating these cells with CpG-A in vitro. The pDCs isolated from the bone marrow of aged mice produced significantly less IFN-α than did the younger mice post-CpG-A administration (Fig. 1,A). In contrast to the reduced IFN-α response with aging, pDCs from aged and young mice produced similar levels of IL-12p40 following in vitro CpG-A treatment (Fig. 1,B). Decreased IFN-α production by aged pDCs remained evident after treatment with different doses of CpG-A and at different time points following CpG-A stimulation (Fig. 1, C and D).
The similar IL-12p40 responses induced by CpG-A treatment in aged and young pDCs implied that aged pDCs expressed TLR9 at levels similar to young pDCs, and these levels were confirmed via flow cytometric analysis (data not shown). We also performed cell viability and apoptosis experiments to examine whether CpG-A-induced toxicity was altered with aging. Eighteen hours of activation with CpG-A induced similar levels of cell death and apoptosis in aged and young pDCs (data not shown). Finally, the reduced IFN-α response by aged pDCs was obtained in two different genetic backgrounds (data not shown), CBA (H2k) and C57BL/6 (H2b), demonstrating that defective IFN-α production with aging was not strain specific. In sum, these data demonstrate that aging impairs the ability of pDCs to produce IFN-α but not IL-12p40 in response to in vitro TLR9 activation with CpG-A.
Systemic IFN-α levels are impaired in aged hosts in response to in vivo CpG-A administration
To broaden the implications of our in vitro results, we examined the impact of aging on the response of pDCs to CpG-A stimulation in vivo. Hence, young and aged C57BL/6 mice received CpG-A i.v., and serum IFN-α was measured. Using pDC-depleting mAbs, we demonstrated that administration of CpG-A to young mice increased IFN-α serum levels in a pDC-dependent manner, whereas pDCs were not essential for IL-12p40 production (data not shown). TLR9 specificity of CpG-A treatment was confirmed in TLR9-deficient mice that were unresponsive to such treatment (data not shown). In agreement with our in vitro experiments described above, aged mice exhibited significantly decreased IFN-α levels but maintained IL-12p40 responses as compared with young mice following in vivo CpG-A administration (Fig. 1, E and F).
Aged mice manifest decreased IFN-α production in response to HSV-2 infection
We next examined the impact of aging on pDC function in response to viral infection with HSV-2, a pathogen that activates TLR9 (21, 34). Systemic HSV-2 infection in young mice induced IFN-α production and viral clearance in a pDC-dependent manner (Fig. 2, A and B) in agreement with a prior study (34), whereas IL-12p40 production was pDC independent (data not shown). Subsequently, young and aged C57BL/6 mice were infected with HSV-2, and serum IFN-α production was measured. Again, the specificity of TLR9 for HSV-2 was confirmed by infecting TLR9-deficient mice, which did not produce IFN-α (data not shown). Purified pDCs from aged mice manifested an impaired IFN-α response during in vitro HSV-2 infection as compared with young pDCs (Fig. 2,C). Importantly, aged mice exhibited decreased serum IFN-α levels during HSV-2 infection as compared with young infected mice (Fig. 2 D). Thus, these results demonstrate that aging impairs the IFN-α response to HSV-2 infection.
Next, we examined the effect of aging on viral clearance of HSV-2 by measuring viral titers in spleens and livers isolated from infected young and aged mice. At 96 h post-HSV-2 infection, aged hosts demonstrated a significantly increased viral load in comparison with young counterparts (Fig. 2 E). These data demonstrate that aged hosts failed to clear HSV-2 infection as effectively as young hosts.
To broaden the implications of our findings, we also employed an alternative viral pathogen, MCMV, which relies on pDCs for host defense (41, 42). Because this virus has been recently shown to depend on dual signaling of TLR7 and TLR9 (43), we first demonstrated that aged pDCs produced reduced levels of IFN-α in response to RNA40, ssRNA that has been shown to stimulate TLR7 (Fig. 3,A) (44, 45). We next infected young and aged mice with MCMV and measured IFN-α levels and viral clearance. IFN-α responses and viral clearance were reduced in aged MCMV-infected mice as compared with young mice (Fig. 3, B and C). Therefore, in response to a viral pathogen that is known to activate pDCs, IFN-α responses were impaired in aged mice.
Aged mice display similar numbers of pDCs in the spleen, bone marrow, and blood compared with younger mice
Because prior clinical reports suggested that aging may lead to a reduction in pDC numbers (29), we next assessed whether aged mice had altered absolute numbers of pDCs in the spleen, bone marrow, and peripheral blood as this decrease may account for the reduced IFN-α during viral infection in aged mice. However, we did not find evidence that aging led to reduced numbers of pDCs within these compartments (Fig. 4). Aged mice did not exhibit reduced spleen or bone marrow cellularity as compared with young mice (data not shown.) Thus, in agreement with our in vitro data, the diminished IFN-α response to in vivo viral infection with aging is likely due to a defect in pDC IFN-α production on a per cell basis.
Adoptive transfer of young pDCs into aged mice augments IFN-α during HSV-2 viral infection and increases viral clearance
We next examined whether the reduced IFN-α response and impaired clearance of HSV-2 in aged hosts could be recovered by adoptive transfer of young cells. In preliminary experiments, we established the efficacy of augmenting IFN-α responses by adoptively transferring young wild-type bone marrow cells into TLR9-deficient hosts. Specifically, i.v. administration of CpG-A 72 h posttransfer of bone marrow cells resulted in detectable serum IFN-α levels in the TLR9-deficient mice that received transfer of wild-type cells when compared with control TLR9-deficient mice that received transfer of TLR9-deficient cells (data not shown). Hence, aged mice received either young whole bone marrow cells or young bone marrow cells that were depleted of pDCs. The aged hosts that received young whole bone marrow cells produced a superior IFN-α response and a reduction in viral load when compared with aged hosts that received young pDC-depleted bone marrow cells (Fig. 5, A and B). We next examined whether the presence of young pDCs transferred to aged hosts were sufficient to elevate IFN-α responses and lead to increased viral clearance. Indeed, transfer of young but not aged pDCs to aged hosts led to elevated IFN-α responses and reduced viral load during HSV-2 infection (Fig. 5, C and D). As young pDCs were functional in an aged environment, our results indicate that aging does not induce factors that suppress pDC responses to HSV-2 infection.
IRF-7 up-regulation is impaired with aging in either response to TLR9 activation or IFN-α stimulation
TLR9-mediated IFN-α induction in pDCs is dependent on the adaptor protein MyD88 and subsequent assembly of a signaling complex that up-regulates IRF-7 (46, 47). Indeed, IRF-7 is essential for IFN-α production during infection with HSV-1, a virus that also activates TLR9 (46). IRF-7 up-regulation also occurs in response to stimulation via the IFN-α/β receptor. Importantly, after TLR9 activation, type I IFN production in pDCs is sustained by a positive feedback loop in which IFN-α activates the IFN-α/β receptor to maintain IRF-7 expression (46).
Thus, we examined whether aging altered signaling in adaptors downstream of either the TLR9 receptor, the IFNα/β receptor, or IRF-7, an adaptor common to both pathways. The fact that IL-12p40 production was preserved in aged pDCs (Fig. 1,B) indicated that adaptors downstream of TLR9 that were independent of the IFN-α/β receptor were likely intact during aging. Indeed, upon TLR9 activation via CpG-A, aged pDCs were able to up-regulate MyD88, TNFR- associated factor 6, IL-1R-associated kinase (IRAK) 2, IRAK4, IRAKM, and IKKα similarly to young pDCs (data not shown and Fig. 6 B; IKKα). We next determined whether IFN-α/β receptor expression was reduced in aged pDCs at rest and after TLR activation. The expression of this receptor was preserved in aged pDCs under these conditions, according to flow cytometric analysis and RT-PCR (data not shown). Furthermore, gene expression analysis indicated that aged pDCs exhibit similar expression of signal adaptors downstream of the IFNα/β receptor, including Stat1, Stat2, Jak1, Tyk2, and IKKε, as compared with young cells at rest and during activation with rIFN-α (data not shown). Thus, no evidence indicating that aging impaired the basal expression and up-regulation of adaptors downstream of either the TLR9 or the IFN-α/β receptor but upstream of IRF-7 was found.
As a result, we next examined whether aging impairs IRF-7 up-regulation in response to either TLR9 or IFN-α activation. Neither IFN-α treatment nor TLR9 activation via CpG-A induced IRF-7 gene or protein up-regulation in aged pDCs, whereas either condition induced IRF-7 up-regulation in young pDCs (Fig. 6, A and B). Furthermore, combined treatment with IFN-α and CpG-A also failed to up-regulate IRF-7 in aged pDCs, whereas this combination effectively induced IRF-7 up-regulation in young pDCs (Fig. 6, A and B). These data demonstrate that aging impairs IRF-7 up-regulation in response to either TLR9 or IFN-αβ receptor activation.
Recent evidence has indicated that PI3K is critical for the nuclear translocation of IRF-7 by pDCs (48). Furthermore, a recent study found that aging impairs PI3K signaling in human myeloid DCs (49). In our study, PI3K up-regulation was impaired in aged pDCs relative to young pDCs during activation with either CpG-A or IFN-α (Fig. 6,C). In response to IFN-α treatment, aged pDCs down-regulated PI3K, whereas IFN-α was up-regulated in young pDCs (Fig. 6,C). Young pDCs were able to translocate IRF-7 to the nucleus during either CpG-A or IFN-α treatment, whereas aged pDCs failed to do so (Fig. 6,D). This result was expected given the fact that aged pDCs also failed to up-regulate cytosolic IRF-7 during these experimental conditions. Furthermore, nuclear IRF-7 transcriptional activity was reduced in aged pDCs after CpG-A activation as compared with that in young pDCs (Fig. 6 E). In sum, aging impairs PI3K up-regulation and IRF-7 up-regulation during TLR9 or IFN-α/β receptor activation.
Aging increases ROS levels in pDCs, and reducing ROS in aged pDCs augments IFN-α production in response to TLR9 activation
One of the leading theories of age-induced cellular damage is the accumulation of damaging ROS (32, 33). Therefore, we assessed whether aged pDCs exhibit elevated levels of ROS as compared with young pDCs. Indeed, aged pDCs displayed increased amounts of ROS at rest and during TLR9 activation as compared with young pDCs (Fig. 7, A and B). To determine whether increased ROS in aged pDCs influenced the type I IFN response during TLR9 activation, aged pDCs were pretreated with the antioxidant NAC, which has been shown to reduce ROS (37, 38, 50). NAC treatment reduced ROS and increased IFN-α production by TLR9-activated aged pDCs as compared with aged pDCs activated with TLR9 in the absence of NAC (Fig. 7, C and D). Young pDCs that were pretreated with NAC also exhibited an increased IFN-α response compared with control treated cells (Fig. 7 D). These data indicate that reducing ROS in aged pDCs augments IFN-α production during TLR9 activation.
We next used CR as a model to reduce oxidative stress during the lifespan of mice. CR has been shown to decrease ROS generation and increase ROS clearance pathways, and this model has been established to reduce age-induced oxidative stress in yeast, Caenorhabditis elegans, Drosophila, and rodents (33). Furthermore, CR leads to an increased lifespan in all of these organisms. Hence, we compared TLR9 responses in pDCs purified from aged CR with AL-fed aged and AL-fed young mice. First, we demonstrated that pDCs from aged CR mice manifested reduced oxidative stress at rest and during TLR9 stimulation compared with pDCs from aged AL-fed controls (Fig. 8, A and B). The pDCs purified from aged CR mice produced more IFN-α than pDCs from aged AL-fed mice in response to in vitro CpG-A stimulation or HSV-2 infection (Fig. 8, C and D); however, the responses from pDCs from aged CR mice remained inferior to the responses of pDCs from AL-fed young mice (Fig. 8, C and D). These results demonstrate that CR reduced age-induced ROS in pDCs and partially recovered the defective IFN-α response to TLR9 activation in aged pDCs.
Next, we examined whether pDCs purified from aged CR mice manifested altered PI3K and IRF-7 up-regulation as compared with pDCs harvested from aged AL-fed mice following either CpG-A or IFN-α/β receptor activation. The pDCs from aged CR mice exhibited superior IRF-7 up-regulation and IRF-7 activity compared with pDCs from aged AL-fed mice during TLR9 activation with CpG-A, although the response was not quite as robust as for pDCs from young AL-fed controls (Fig. 8, E and F). Furthermore, PI3K up-regulation was elevated in pDCs from aged CR mice relative to pDCs from aged AL-fed mice under these conditions. In fact, PI3K up-regulation by pDCS from aged CR mice was similar to pDCs from young AL-fed mice (Fig. 8,G). Additionally, pDCs from aged CR mice displayed superior phosphorylation of AKT, which signals downstream of PI3K, during CpG-A activation as compared with pDCs from aged AL-fed mice and exhibited a similar response to that of pDCs from young AL-fed mice (Fig. 8 H). In sum, our results indicate that the age-induced impairment in the up-regulation of PI3K and IRF-7 in pDCs during TLR9 activation is, at least partly, recovered by CR.
In our study, we first examined the impact of aging on pDC function and whether any age-induced defects within pDCs impaired host defense to viral infection. Using CpG-A or HSV-2 virus, which both activate pDCs via TLR9, we found that aged pDCs exhibit an impaired ability to produce IFN-α, but not IL-12p40, as compared with pDCs from young animals. During HSV-2 infection, defective pDC function was responsible for an impaired ability of aged mice to clear the virus. Thus, our study provides evidence that defective type I IFN responses by aging pDCs impairs host defense to viral infection in these experimental models.
Gene and protein analyses were used to uncover the age-related defects within pDCs during TLR9 stimulation. Aged pDCs displayed an impaired ability to up-regulate IRF-7 and PI3K during either TLR9 or IFN-α/β receptor activation. We did not find evidence that adaptors upstream of IRF-7 that were unique to either the TLR9 or IFN-α/β receptor pathways were altered by aging under our experimental conditions. Therefore, our study has demonstrated that aging impairs the up-regulation of PI3K and IRF-7 within pDCs during TLR9 or IFN-α/β receptor activation.
We also noted that aged pDCs exhibited increased oxidative stress both at rest and during TLR9 activation compared with young pDCs (Fig. 7). This result is in agreement with prior work indicating that aging leads to accumulation of damaging ROS, which augment oxidative stress and subsequently impair cellular functions (33). Reducing ROS by pretreating aged pDCs with NAC increased TLR9-induced IFN-α responses by these cells. Furthermore, pDCs from aged CR mice showed a partly recovered phenotype. In fact, these pDCs had reduced ROS compared with pDCs from AL-fed age-matched controls (Fig. 8), and these pDCs demonstrated elevated IFN-α levels and IRF-7 up-regulation and activity during TLR9 activation as compared with pDCs from aged AL-fed controls (Fig. 8). Prior work indicates that CR may mediate its effects by reducing ROS generation or increasing the clearance of ROS, in part, by improving mitochondrial function (33). Our data support these findings in that pDCs from aged CR mice exhibited reduced ROS as compared with pDCs from aged AL-fed controls (Fig. 8, A and B). Thus, age-induced elevations in ROS levels may lead to IRF-7 protein degradation; however, the possibility that CR may be mediating its effects independently of ROS reduction cannot be excluded by this study. In addition, perhaps other age-induced cellular perturbations besides elevated ROS levels may be important for the impaired type I IFN response by aged pDCs, given that NAC treatment reduced ROS levels in aged pDCs but did not recover the response to the same level noted in non-NAC-treated, CpG-A-activated young pDCs (Fig. 7 D). Future studies will be required to determine the mechanism of susceptibility of the type I IFN signaling pathway to the damaging effects of ROS generation during aging. Overall, our study suggests that increased oxidative stress during aging leads to impaired type I IFN function in pDCs in response to TLR9 activation.
IFN-α produced by pDCs is essential for host defense against certain viruses (8, 11). Importantly, reduced IFN-α levels result in increased mortality during viral infections (51, 52). Furthermore, the early release of IFN-α during viral infection leads to direct inhibition of viral replication in addition to activation of NK cells, CTLs, and macrophages to eliminate virally infected cells (51, 52, 53). Here, we demonstrated that adoptive transfer of young pDCs into aged hosts augmented IFN-α responses and improved viral clearance during HSV-2 infection. These results suggest that augmenting the IFN-α response in aged hosts may be a potential therapeutic approach to protect aged hosts from viral infections.
Aged pDCs exhibited impaired responses to TLR7 activation in vitro (Fig. 3). We found that during infection with MCMV, a virus that requires pDCs for host defense (41) and has recently been shown to activate both TLR7 and TLR9 (43), aging also impaired IFN-α responses and viral clearance (Fig. 3). Additionally, a prior report demonstrated that aged mice produced reduced type I IFN response after stimulation with poly(IC), a substance that activates TLR3 (54), although this study did not focus on pDCs. Overall, our study indicates that aging impairs both TLR7 and TLR9 responses in pDCs.
A prior clinical report indicated that pDCs purified from the peripheral blood of older people yielded impaired IFN-α responses as compared with cells from younger control subjects (31). This result agrees with our study and suggests that the findings of our report may be translatable to humans. Thus, confirmation of our results in human cells will be critical.
Our study clearly demonstrates that aged pDCs exhibit a defective response to TLR9 activation. Other studies have indicated that aging also impairs the function of other cellular components of the immune system, including T cells, B cells, macrophages, and NK cells. In our study, however, viral clearance was augmented in aged hosts by the adoptive transfer of young bone marrow cells or young pDCs but not with pDC-depleted young bone marrow. These data imply that pDCs are both necessary and sufficient for the adoptive transfer to alter the phenotype in our experimental model. Recently, innate cytosolic receptors of viral infection, such as the RNA helicase system, have been identified (55). Future investigation will be required to determine whether aging alters this pathway of innate viral recognition and to integrate the many alterations that occur in the immune response with aging. Such avenues of investigation will have important implications for the design of potential vaccines to augment immune function with aging.
In conclusion, our study indicates that impaired IFN-α responses within pDCs lead to defective host defense to viral infections. Importantly, aging impaired the up-regulation of IRF-7, a key signal adaptor in the type I IFN signaling pathway, in response to TLR9 activation. This fundamental information helps to clarify the complex interaction between aging and host immunity and also provides a potential explanation for the increased susceptibility to viral infections observed in older people.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported in part by National Institutes of Health Grants AG028082 and AG026772 (Paul B. Beeson Award in Aging Research) (to D.R.G.). H.W.S.-D. is supported by National Institutes of Health Grant T32AG019134. W.E.W. is supported by National Institutes of Health Grant 5T32HL007778.
Abbreviations used in this paper: DC, dendritic cell; AL, ad libitum; CR, calorie restriction; DHE, dihydroethidium; IRF, interferon regulatory factor; NAC, N-acetylcysteine; pDC, plasmacytoid DC; ROS, reactive oxygen species; DOTAP, 1,2-dioleoyl-3-(trimethyammonium)propane; IKK, IκB kinase; ODN, oligodeoxynucleotide; MCMV, murine CMV; IRAK, IL-1R-associated kinase.