Defective early innate immune response to ectromelia virus in the draining lymph nodes of aged mice due to impaired dendritic cell accumulation

Abstract It is known that aging decreases natural resistance to viral diseases due to dysfunctional innate and adaptive immune responses, but the nature of these dysfunctions, particularly in regard to innate immunity, is not well understood. We have previously shown that C57BL/6J (B6) mice lose their natural resistance to footpad infection with ectromelia virus (ECTV) due to impaired maturation and recruitment of natural killer (NK) cells to the draining popliteal lymph node (dLN). More recently, we have also shown that in young B6 mice infected with ECTV, the recruitment of NK cells is dependent on a complex cascade whereby migratory dendritic cells (mDCs) traffic from the skin to the dLN, where they produce CCL2 and CCL7 to recruit inflammatory monocytes (iMOs). In the dLN, mDCs also upregulate NKG2D ligands to induce interferon gamma (IFN‐γ) expression by group 1 innate lymphoid cells (G1‐ILCs), mostly NK in cells but also some ILC1. In response to the IFN‐γ, the incoming uninfected iMOs secret CXCL9 to recruit the critical NK cells. Here, we show that in aged B6 mice, the trafficking of mDCs to the dLN in response to ECTV is decreased, resulting in impaired IFN‐γ expression by G1‐ILCs, reduced accumulation of iMOs, and attenuated CXCL9 production by iMOs, which likely contributes to decrease in NK cell recruitment. Together, these data indicate that defects in the mDC response to viral infection during aging result in a reduced innate immune response in the dLN and contribute to increased susceptibility to viral disease in the aged.

examining age-related changes of the innate immune system remain scarce.
Dendritic cells (DCs) are cells of the innate immune system which are well known for their role as professional antigen-presenting cells for the priming of T-cell responses and that are characterized by the expression of the integrin CD11c (Merad, Sathe, Helft, Miller, & Mortha, 2013). Migratory DCs (mDCs) are a class of DCs that have the ability to migrate from tissues to secondary lymphoid organs such as lymph nodes (LNs) (Worbs, Hammerschmidt, & Forster, 2017). mDCs are the predominant class of DCs in the skin. They constitutively migrate from the skin to the local LN and increase their migration during infection. Because they are mature, migrant mDCs can be distinguished from other DCs in LNs by high levels of expression of major histocompatibility (MHC) class II (MHC II) molecules at the cell surface (MHC II hi ) (Miller et al., 2012). Skin mDCs are best known for their ability to present antigens to T-cells in the dLN in various models of infection, immunization, and contact hypersensitivity (Bollampalli et al., 2015;Bouteau et al., 2019;Kaplan, Jenison, Saeland, Shlomchik, & Shlomchik, 2005). However, they can also play an important role at initiating the innate immune response cascade in the dLN (Wong et al., 2018).
It is known that the functionality of DCs, including mDCs, can decrease with age. For example in aged mice, defects in conventional DCs can negatively affect anti-tumor NK and T-cell responses (Grolleau-Julius, Harning, Abernathy, & Yung, 2008;Guo, Tilburgs, Wong, & Strominger, 2014a, 2014b, and mDCs of the lung migrate poorly to the dLN during respiratory syncytial virus (RSV) infection resulting in decreased T-cell responses (Zhao, Zhao, Legge, & Perlman, 2011). In humans, the numbers of various DC types are reduced with age (Della Bella et al., 2007;Gupta, 2014), and monocyte-derived dendritic cells display poor antigen uptake (Agrawal et al., 2007). Moreover, in both mice and humans, aging results in reduced numbers of Langerhans cells (a type of mDC) in the skin, and impaired migration (Cumberbatch, Dearman, & Kimber, 2002;Pilkington et al., 2018).

Ectromelia virus (ECTV) is an
Orthopoxvirus and a natural pathogen of the mouse which naturally enters the body through the skin, most frequently of the footpad. Footpad infection with ECTV causes a lethal disease known as mousepox in susceptible strains of mice such as BALB/c, but not in mousepox-resistant mice, such as young C57BL/6 (B6) (Wallace, Buller, & Morse, 1985). In both, susceptible and resistant mouse strains, ECTV spreads lymphohematogenously from the footpad to the local popliteal draining LN (dLN) and then the blood, eventually infecting the liver and spleen (Esteban & Buller, 2005;Sigal, 2016). Resistant mice survive because compared to susceptible mice, they control better the systemic spread of the virus from the dLN and also viral replication in spleen and liver.
While the dLN is largely thought of as the site where T-cell priming occurs (Hickman et al., 2008), it also serves as a site where innate immune mechanisms prevent lymphohematogenous viral dissemination. In a series of papers, we have previously shown an intricate network of collaborative innate immune responses within the dLN that lead to the control and, ultimately, resolution of ECTV infection in young B6 mice (Fang, Roscoe, & Sigal, 2010;Wong et al., 2018;Xu et al., 2015). Within this inflammatory network, skin mDCs (CD11c + MHC II hi ), play a central organizing role. Specifically, we showed that soon after ECTV infection in the footpad, CD11c + MHC II hi mDCs increase their migration from the skin of the footpad to the dLN. Once in the dLN, infected and uninfected mDCs produce a variety of inflammatory mediators. Among these, the chemokines CCL2 and CCL7 recruit inflammatory monocytes from the blood into the dLN (Wong et al., 2018). In the dLN, infected mDCs also upregulate ligands for NKG2D, such as MULT1, to predominantly induce the production of interferon gamma (IFN-γ) in NK cells and in some of the few innate lymphoid cells 1 (ILC1) already present in the dLN (Wong et al., 2018). Together, NK cells and ILC1 constitute the Group-1 Innate Lymphoid Cells (G1-ILCs), which are characterized by their ability to produce IFN-γ and their expression of NK1.1 and NKp46. The IFN-γ produced by G1-ILCs activates the uninfected newly arrived iMOs which, in response, produce the chemokine CXCL9 to recruit circulating mature NK cells into the dLN (Wong et al., 2018). These incoming NK cells have a critical role at curbing systemic virus spread from the dLN (Fang et al., 2010). Of note, once they get infected, iMOs do not produce CXCL9 but become the major producers of Type I interferon (IFN-I), which is also critical for the control of virus spread and resistance to mousepox (Jacoby, Bhatt, & Brownstein, 1989;Karupiah, Fredrickson, Holmes, Khairallah, & Buller, 1993;Xu et al., 2015).
Induction of IFN-I in iMOs requires their infection. Notably, disruption of mDC migration to the dLN impairs the recruitment of NK cells and iMOs and results in susceptibility to mousepox Wong et al., 2018).
B6 mice older than 16 months fail to recruit NK cells to the dLN and are highly susceptible to mousepox (Fang et al., 2010;Fenner, 1949;Wallace et al., 1985), mimicking the increased susceptibility to viral infections observed in the elderly. The deficient NK cell migration to the dLN in aged mice is partly intrinsic, as their numbers are decreased in the circulation and have an immature phenotype in multiple tissues when compared to NK cells in young mice (Beli et al., 2011;Nair, Fang, & Sigal, 2015;Shehata, Hoebe, & Chougnet, 2015). Yet, adoptive transfer of NK cells from young mice only partially restores resistance to mousepox in aged B6 mice (Fang et al., 2010), suggesting other mechanisms contribute to the loss of resistance. Given the critical role of mDCs, G1-ILC, and iMOs in resistance to mousepox Wong et al., 2018), we sought to investigate their role in age-related susceptibility to viral disease.

| RE SULTS
2.1 | Migratory dendritic cells are equally present in the footpad of naïve young and aged mice, but fail to accumulate in the dLN of aged mice during ECTV infection Using mice transgenic for photoactivatable green fluorescent protein, we have recently demonstrated that following footpad infection F I G U R E 1 Migratory dendritic cells are present in the naïve footpad of aged mice at similar frequency, but fail to accumulate in the dLN during ECTV infection. mDCs in the footpad of naïve young and aged mice were analyzed. (a) Representative flow cytometry plots indicating the gating strategy. (b) Number of CD45 + cells/25 mm 2 displayed as mean ± SEM of individual mice. (c) Number of mDCs (CD19 -MHCII + CD11c hi )/25 mm 2 displayed as mean ± SEM of individual mice. Data correspond to two independent experiments combined with a total of 10-15 mice/group. (d) Representative flow cytometry plots showing the gating of mDCs (CD4 -CD8 -CD11c + MHCII hi ) in the ndLN and dLN of mice young and aged mice at 48 postinoculation with 50 µg CpG. (e) Total number of mDCs as mean ± SEM in the ndLN and dLN of individual aged and young mice at 48 hr postinoculation with 50 µg CpG in the footpad. Data correspond to three independent experiments combined with a total of 6-9 mice/group. (f) Representative flow cytometry plots for the gating of mDCs (CD4 -CD8 -CD11c + MHCII hi ) in the ndLN and dLN of young and aged mice at 48 hpi with 3,000 pfu ECTV-dsRed in the footpad. (g) Total number of mDCs as mean ± SEM in the ndLN and dLN of individual aged and young mice at 24 hpi. Data correspond to two independent experiments combined with a total of 6-7 mice/group. (h) As in G but at 48 hpi. Data correspond to five independent experiments combined with a total of 25-26 mice/group. For all, *p < .05; ****p < .0001 with ECTV, CD11c + MHC II hi mDCs migrate from the footpad to the dLN to coordinate the recruitment of iMOs and NK cells to the dLN Wong et al., 2018;Xu et al., 2015). To continue elucidating the mechanisms underlying age-related susceptibility to mousepox, we used flow cytometry (gating strategy shown in (Figure 1a) to examine differences in mDCs within the skin of the footpad in naïve young and aged B6 mice.
Results showed that young and aged mice had similar numbers of hematopoietic (CD45 + ) cells ( Figure 1b) and mDCs, which in the skin are identified as (CD19 -CD11c hi MHC II + (Figure 1c).
We next focused our attention onto the dLN. It is known that mDCs constitutively migrate in low numbers from the skin to the local LNs, and increase their migration after an inflammatory stimulus (Stoitzner, Tripp, Douillard, Saeland, & Romani, 2005;. To test whether aging may affect

| Impaired early production of IFN-γ by Group 1 ILCs in response to ECTV infection in aged mice
Having found defective accumulation of mDCs in the dLN of aged mice in response to WT ECTV infection, we investigated whether downstream innate immune mechanisms in the dLN were also disrupted. quantitative PCR of reverse-transcribed RNA (RT-qPCR) of whole dLNs at 24 hpi with WT ECTV showed that Ifng mRNA was lower in the dLN of aged than young mice ( Figure 2a). Moreover, by flow cytometry, fewer G1-ILCs (NK1.1 + TCRβ -), most of which are NK cells (Wong et al., 2018), produced IFN-γ in the dLN of aged compared to young mice (Figure 2b-c). These data demonstrate very early defects in the activation of dLN G1-ILCs in aged mice. Of note, while IFN-γ produced by G1-ILCs is important for the activation of iMOs to recruit NK cells to the dLN (Wong et al., 2018), IFN-γ can also directly inhibit viral replication in infected cells (Boehm, Klamp, Groot, & Howard, 1997). Consistent with this, viral transcripts in the dLN of aged mice at 24 hpi were increased ( Figure 2d). This was likely due to the lower levels of IFN-γ, because the dLN of young mice treated with IFN-γ blocking antibody 24 hr before infection had levels of viral transcripts at 24 hpi that were similar to those in aged mice and significantly higher than in PBS-treated control young mice Notably, we found that mDCs upregulated MULT1 to similar levels in aged and young mice (Figure 2h-i) indicating that the infected mDCs of aged mice do not have defective NKG2D ligand upregulation. These data strongly suggest that the lack of IFN-γ production by G1-ILCs at 24 hpi in the dLNs of aged mice is due to the absence of mDC migration to the dLN rather than an intrinsic inability to stimulate G1-ILCs through NKG2D ligand upregulation. Yet, intrinsic defects in G1-ILCs cannot be excluded as a possible cause of IFN-γ deficiency.

| Decreased recruitment of iMOs to the dLN of aged mice in response to ECTV infection
We have shown that mDCs produce CCL2 and CCL7 to recruit iMO to the dLN (Wong et al., 2018;. Because aged mice had reduced numbers of mDCs in the dLN, we measured the expression of mRNA for these chemokines by RT-qPCR. Compared to the ndLN, Ccl2 and Ccl7 mRNAs were significantly increased in the dLN of young but not aged mice at 24 hpi ( Figure 3a,b). However, at 60 hpi, which is the peak of iMOs recruitment to the dLN in young mice (Xu et al., 2015), the relative quanti- (c) Frequencies of IFN-γ + G1-ILCs in the ndLN and dLN of young and aged mice determined by flow cytometry at 24 hpi with 3,000 pfu WT ECTV. Data correspond to the mean ± SEM of individual mice and to four independent experiments combined with a total of 14-21 mice/ group. (d) Expression of the viral gene Evm003 determined by RT-qPCR in the ndLN and dLN of aged and young mice at 24 hpi with 3,000 pfu WT ECTV. Data shown as mean ± SEM correspond to four independent experiments combined with a total of 13-18 mice/group (e) Aged and young mice were injected intraperitoneally with 100 μg mouse IFN-γ blocking monoclonal antibody 24 hr prior to infection with 3,000 pfu WT ECTV. Data show mean ± SEM relative expression of viral gene Evm003 in the ndLN and dLN of individual mice determined by RT-qPCR at 24 hpi. Data correspond to three independent experiments combined with a total of 10-13 mice/group. (f) Representative flow cytometry plots of MULT1 expression by ECTV-dsRedor ECTV-dsRed + mDCs (CD11c + MHC II hi ) in the dLN of aged and young mice at 48 hpi with 3,000 pfu ECTV-dsRed. (g) MFI of MULT1 on ECTV-dsRedor ECTV-dsRed + on total mDCs (CD11c + MHC II hi ). MFI is shown as mean ± SEM. Data correspond to three independent experiments combined with a total of 13-14 mice/group. For all, *p < .05; **p < .0; ***p < .001; ****p < .0001 that aged mice suffer a delayed rather than absent innate immune response in the dLN, and/or that aged mice require much higher virus loads than young ones to fully activate their innate immune response.

| Decreased production of CXCL9 by iMOs in the dLN of aged mice in response to ECTV
Upon recruitment to the dLN, iMOs become functionally divergent: in response to the IFN-γ produced by G1-ILCs, uninfected iMOs produce CXCL9 to recruit NK cells, while infected iMOs upregulate IFN-I (Wong et al., 2018;Xu et al., 2015). Because IFN-γ and iMOs were reduced in the dLN of aged mice, we investigated whether CXCL9 production by uninfected iMOs was impaired. To

| Defective anti-viral innate immune responses in the dLN of aged mice are independent of viral virulence and lethality
To test whether the dysfunctional anti-viral response in the dLN of aged mice was due to ECTV virulence or a general disability of aged innate immunity to rapidly respond to viral infection, we infected mice with ECTVΔ166, a mutant ECTV that lacks an IFN-I decoy receptor and is at least 10 7 -fold less virulent than WT ECTV (Xu et al., 2008).   decoys (Sigal, 2016). It is possible that the skin of aged mice produces smaller amounts of a cytokine that is targeted by the virus, which is necessary for mDC migration. In addition, ECTV encodes a MyD88 inhibitor, and its effects could be greater in aged than in young mice. Future research should look into these issues.

| D ISCUSS I ON
It has been shown that conventional aged DCs fail to activate NK cells in a tumor model, leading to poor killing of tumor cells (Guo, Tilburgs, Wong, & Strominger, 2014a, 2014b. We have previously shown that following ECTV infection of young mice, a small but F I G U R E 5 Defective anti-viral innate immune responses in the dLN of aged mice are independent of viral virulence and lethality. (a-c) Expression of mRNA for CCl2 (a), Ccl7 (b), and Ifng (c) in the dLN of aged and young mice as determined by RT-qPCR in individual mice at 24 hpi with 3,000 pfu ECTV-Δ166. Data shown as mean ± SEM of individual mice. Data correspond to two independent experiments combined with a total of 6-9 mice/group. (d) Relative expression of viral gene Evm003 was determined by RT-qPCR in the ndLN and dLN of young and aged mice at 60 hpi with 3,000 pfu ECTV Δ166. Data are shown as mean ± SEM of individual mice and correspond to three independent experiments combined with a total of 6-9 mice/group. (e) Expression of Cxcl9 mRNA was determined by RT-qPCR in the ndLN and dLN of aged and young mice at 60 hpi with 3,000 pfu ECTV-Δ166. Data shown as mean ± SEM of individual mice. Data correspond to two independent experiments combined with a total of 6-9 mice/group. (f) Total mDCs were determined in the ndLNs and dLNs of the indicated mice at 60 hpi with 3,000 pfu of ECTV-Δ166. Data shown as mean ± SEM of individual mice. Data correspond to four independent experiments combined with a total of 6-15 mice/group. (g) Total numbers of iMOs (CD11b + GR-1 + Ly6C + ) in the ndLN and dLN of aged and young mice at 60 hpi with 3,000 pfu ECTV-GFP or ECTV-Δ166. Data shown as mean ± SEM of individual mice. Data correspond to four independent experiments combined with a total of 8-18 mice/group. (h) Total number of NK cells in the ndLN and dLN of aged and young mice at 60 hpi with 3,000 pfu ECTV-GFP or ECTV-Δ166. Data shown as mean ± SEM of individual mice. Data correspond to four independent experiments combined with a total of 8-18 mice/group (i) The indicated mice were infected with 3,000 pfu ECTV-GFP or ECTV Δ166 in the footpad. Survival was monitored. Data are displayed as a combination of two independent experiments with a total of 5-10 mice/group. For all, *p < .05; **p < .0; ***p < .001; ****p < .0001 have shown intrinsic NK cell defects in aged mice (Beli et al., 2014;Fang et al., 2010;Nair et al., 2015). Notably, the decrease in mDC accumulation, IFN-γ production, iMOs recruitment, CXCL9 expression, and NK cell migration was independent of viral pathogenicity, as infection with attenuated ECTV failed to restore these processes.
Interestingly, even in the absence of this innate immune cascade, aged mice controlled attenuated ECTV.
iMOs are recruited to the dLN by CCL2 and CCL7 produced by infected mDCs. Incoming uninfected iMOs produce CXCL9 in response to IFN-γ to recruit NK cells to the dLN while infected iMOs produce anti-viral IFN-I but not CXCL9 (Wong et al., 2018;. Here we show that aged mice recruit considerably fewer iMOs to the dLN. It is known that the frequency and numbers of circulating monocytes are unchanged in aged humans but increased in aged mice (Puchta et al., 2016). Therefore, the decrease in iMO recruitment to the dLN is most likely due to reduced CCL2/7 chemotactic signals.
CXCL9 production by uninfected iMOs was significantly reduced which was most likely the result of impaired production of IFN-γ by G1-ILCs and increased frequency of infected iMOs. On the other hand, IFN-I production appears to be intact in aged mice. Indeed, the overall expression of IFN-I in the dLN of aged mice was higher than in young mice, which could be due to the increased frequency of infected iMOs, and the increase in viral loads due to deficient production of IFN-γ by G1-ILC.
As we have shown previously, NK cells fail to migrate to the dLN during ECTVΔ166 infection (Fang et al., 2010 due to unabated IFN-I production and sensing, culminating in the induction of a virus-specific CD8 T-cell response (Fang et al., 2010).
Our data suggest a delay in the first 24 hr of the immune response against ECTV. An interesting avenue to pursue would be to determine whether treating aged mice with CpG or IFN-γ prior to or during active infection could promote survival to virulent ECTV. However, the mechanisms by which these treatments could protect aged mice is unpredictable, and its elucidation would require substantive work.
For example, IFN-γ supplementation could restore iMO activation, but could also rapidly decrease viral replication and preclude or di-

| Viruses and infection
Ectromelia virus (ECTV)-Moscow strain (ATCC VR-1374), ECTV-GFP, ECTV-dsRED, and ECTVΔ166 were propagated in tissue culture as previously described (Xu et al., 2008). Mice were infected in the footpad with 3,000 plaque-forming units (pfu) ECTV as indicated. For the determination of survival, mice were monitored daily and, to avoid unnecessary suffering, mice were euthanized and counted as dead when imminent death was certain as determined by lack of activity and unresponsiveness to touch. Euthanasia was according to the 2013 edition of the AVMA Guideline for the Euthanasia of Animals. For virus titers, the entire spleen or portions of the liver were homogenized in 2.5% FBS RPMI (Corning) using a Tissue Lyser (QIAGEN). Virus titers were determined on BSC-1 cells as previously described (Xu et al., 2008).

| Cell isolation
Mice were euthanized by cervical dislocation. Single-cell suspen-

| Flow cytometry
To determine cellular responses in the LNs, intact LNs were incubated at 37°C for 1 hr in media containing 10 µg/ml brefeldin A and

| RNA preparation and RT-qPCR
Total RNA from LNs was obtained with the RNeasy Mini Kit (QIAGEN) as previously described (Rubio et al., 2013;Xu et al., 2015).

CO N FLI C T O F I NTE R E S T S
The authors declare no conflict of interests.

AUTH O R S ' CO NTR I B UTI O N S
Luis J. Sigal contributed to the conceptualization, analysis, administration, and supervision of the project, and the editing of the manuscript. Coby Stotesbury performed most of the experiments, analyzed results, and wrote the first draft of the manuscript. Eric B.