Mice were anesthetized using isoflurane and then whole blood was collected from retro-orbital into EDTA-coated MiniCollect tubes (Greiner, USA). For flow cytometry, 50 μL of whole blood was incubated with BD Fc block™ (CD16/32 clone, BD Bioscience, 553141) for 15 min on ice followed by surface marker staining with CD45-BV510 (Biolegend, 103138), CD11b-PE (Biolegend, 101208), Ly6C-eFluor450 (Thermofisher, 48-5932-82), Ly6G-BV785 (Biolegend, 127645), CD115-APC (Biolegend, 135509), CCR2-FITC (Biolegend, 150607) for 30 min. Red blood cells were then lysed with FACS Lysing Solution (BD Bioscience, 349202) in 1:20 ratio for 20 min at room temperature followed by 2% PFA for 30 min on ice. AccuCheck counting beads (Thermofisher, PCB100) were added to samples before data acquisition to enable the computation of absolute cell counts in PBMC, as per the manufacturer’s instructions. Fluorescence was detected using BD LSR Fortessa™ X-20 (BD Bioscience, San Jose, CA, USA) and/or Cytek® Aurora (Cytek®, Fremont, CA, USA). The gating strategy is noted in Supplementary Figure 1A . Briefly, neutrophils were gated as CD45⁺CD11b⁺Ly6G⁺ and monocytes were gated as CD45⁺CD11b⁺CD115⁺. Then, monocyte subpopulations were further distinguished by differential expression of Ly6C. We confirmed that the Ly6C^hi subset expressed high levels of CCR2, whereas Ly6C^interand Ly6C^low subsets had low expression of CCR2.

Peritoneal fluid was collected as described previously (23). Briefly, 3 ml of ice-cold phosphate-buffered saline (PBS) containing 2% fetal bovine serum (FBS) was injected into the peritoneum. After gentle shaking for 1 min, peritoneal fluid was collected with an 18G syringe needle. The samples were then centrifuged at 500 g for 5 mins at 4°C. To lyse red blood cells, ACK lysing buffer (Fisher scientific, 50-101-9080) was added to the cell pellets and incubated for 5 min. Then, cells were collected by centrifugation at 500 g for 10 min at 4°C followed by washing twice with PBS. Golgi plug (BD Bioscience) was added to the cells for 4 hours followed by antibody staining as described in the Flow cytometry staining and cell analysis section below. The surface staining markers were, CD45-eFluor450 (eBioscience, 50-112-9409), Ly6G-BV785 (Biolegend, 127645), CD11b-APC-Cy7 (Biolegend, 101226), F4/80-PE-Cy7 (eBioscience, 25-4801-82), CD38-PE (BD Bioscience, 553764), and CD206-Alexa Fluor 488 (Biolegend, 141710). Additionally, to assess functional outcome more directly, intracellular antibodies were used to test cytokine expression in specific cells. TNFα-PE-Dazzle594 (Biolegend, 506345), IL-1β-PE (Thermofisher, 12-7114-80), iNOS-Alexa Fluor 488 (Thermofisher, 53-5920-80) and Arg1-Alexa Fluor700 (Thermofisher, 56-3697-82) were used to stain cells for 30 min on ice. Cells were then washed with the perm wash buffer. Data were acquired using a BD LSR Fortessa™ X-20 (BD Bioscience, San Jose, CA, USA) and/or Cytek® Aurora (Cytek®, Fremont, CA, USA). The flow cytometry gating strategy is detailed in Supplementary Figure 2 . Briefly, peritoneal neutrophils were defined as CD45⁺CD11b⁺Ly6G⁺. The intracellular markers (TNFα, IL-1β, and iNOS) were evaluated from neutrophil population (CD45⁺CD11b⁺Ly6G⁺). Peritoneal macrophages were defined as CD45⁺CD11b⁺Ly6G^-F4/80⁺. The surface markers (CD38 and CD206) and intracellular markers (TNFα, IL-1β, iNOS, and Arg1) were evaluated from macrophage population (CD45⁺CD11b⁺Ly6G^-F4/80⁺).

Stromal vascular fraction was collected from epididymal adipose tissue as described previously (24). In brief, approximately 1 g of epididymal fat was minced using scissors and incubated in digestion buffer (1 mg/ml collagenase type I in RPMI medium) for 30 min at 37°C in a shaking water bath. The samples were filtered with 70 μm cell strainers to remove undigested tissue. After centrifuging at 1,700 g for 3 min, the cell pellets were collected and incubated with 4 ml of ACK lysing buffer (Fisher scientific, 50-101-9080) on ice for 10 min. The samples were then centrifuged at 1,700 g for 3 mins and the pellet of stromal vascular fraction (SVF) was re-suspended in 1 ml of PBS with 2% FBS. The antibody staining was described in the Flow cytometry staining and cell analysis section below. Surface markers were stained using CD45-BV510 (Biolegend, 557235), Ly6G-AlexaFluor 700 (Biolegend, 127622), CD11b-APC-cy7 (Biolegend, 101226), F4/80-PE-cy7 (eBioscience, 25-4801-82), and CX3CR1-BV421 (Biolegend, 149023) for 30 min on ice. The cells were washed with FACS buffer (1xPBS and 2% FBS) and then stained with 7AAD (BD Bioscience) before data collection. The data was acquired using MoFlo Astrios EQ (Beckman Coulter Life Sciences, Indianapolis, IN, USA) and/or Cytek® Aurora (Cytek®, Fremont, CA, USA). The flow gating strategy and representative plots are shown in Supplementary Figure 3 . Neutrophils were gated as CD45⁺CD11b⁺Ly6G⁺. Infiltrating macrophages were defined as CD45⁺CD11b⁺Ly6G^-F4/80^hi and resident macrophages were defined as CD45⁺CD11b⁺Ly6G^-F4/80^low. The surface markers (CD38 and CD206) were evaluated from the macrophage population (CD45⁺CD11b⁺Ly6G^- F4/80⁺).

The gallbladder was carefully removed first, and then the liver was dissected. A portion of liver tissue was digested into single cell suspension using the Liver Dissociation Kit for mouse (Miltenyi Biotec, Auburn, CA, USA, 130-105-807) following the manufacturer’s instructions. Briefly, liver was rinsed with DMEM then transferred into MACS C tubes containing a dissociation enzyme mix, which was pre-heated for 30 min at 37°C. The samples were then digested using GentleMACS™ Octo Dissociator with heaters (Miltenyi Biotec, Auburn, CA, USA). The cell suspensions were filtered with 100 μm strainers, then centrifuged at 300 g for 10 min. Resuspended cells were stained with antibodies as described in the Flow cytometry staining and cell analysis section below. The antibodies used for cell surface staining were CD45-PerCP (BD Bioscience, 557235), Ly6G-BV785 (Biolegend, 127645), F4/80-PE-Cy7 (eBioscience, 25-4801-82), CD11b-APC-Cy7(Biolegend, 101226), CD38-PE (BD Bioscience, 553764), and CD206-BV650 (Biolegend, 141723). The intracellular staining antibodies were TNFα-Alexa Fluor 700 (BD Bioscience, 558000), IL-1β-PE (Thermofisher, 12-7114-80), and iNOS-Alexa Fluor 488 (Thermofisher, 53-5920-80). The data was acquired using Cytek® Aurora (Cytek®, Fremont, CA, USA). The flow gating strategy and representative plots are shown in Supplementary Figure 4 . Neutrophils were selected as CD45⁺CD11b⁺Ly6G⁺. The intracellular markers (TNFα, IL-1β, and iNOS) were evaluated from neutrophil population (CD45⁺CD11b⁺Ly6G⁺). Infiltrating macrophages were defined as CD45⁺CD11b^hiLy6G^-F4/80^low and resident macrophages were defined as CD45⁺CD11b^hiLy6G^- F4/80^hi. The surface markers (CD38 and CD206) and intracellular markers (TNFα, IL-1β, and iNOS) were evaluated from macrophage population (CD45⁺CD11b⁺Ly6G^-F4/80⁺).

The colon was first washed with ice-cold phosphate-buffered saline, then cut longitudinally in the middle. Half colon was digested to obtain a single-cell suspension using the Lamina Propria Dissociation Kit for mouse (Miltenyi Biotec, Auburn, CA, USA, 130-097-410) according to the manufacturer’s instructions. In brief, the colon samples were washed with 1x HBSS without Ca²⁺ and Mg²⁺ containing 10 mM HEPES, 5mM EDTA and 5% FBS, and then the samples were digested by the enzyme mix from the Lamina Propria Dissociation kit (Miltenyi Biotec, Auburn, CA, USA) using GentleMACS™ Octo Dissociator with heaters (Miltenyi Biotec, Auburn, CA, USA). The cell suspensions were filtered with 100 μm strainers. The single-cell suspensions were then incubated with Golgi plug (BD Bioscience) for 4 hours then stained with fluorescence-conjugated antibodies of targeted markers for flow cytometry as described in the Flow cytometry staining and cell analysis section below. Surface marker antibodies used were CD45-FITC (eBioscience, 11-0451-82), CD11b-APC-cy7 (Biolegend, 101226), Ly6G-BV785 (Biolegend, 127645), Ly6C-eFluor450 (eBioscience,48-5932-82), MHCII-Alexa Fluor700 (Biolegend, 107622), CD64-Alexa Fluor647 (Biolegend, 139322), and CD206-PerCP-cy5.5 (Biolegend, 141716). Intracellular markers used were TNFα-PE-Dazzle594 (Biolegend, 506345) and IL-1β-PE (Thermofisher, 12-7114-80). The data were acquired using MoFlo Astrios EQ (Beckman Coulter Life Sciences, Indianapolis, IN, USA) and/or Cytek® Aurora (Cytek®, Fremont, CA, USA). The flow gating strategy and representative plots are shown in Supplementary Figure 5 . Briefly, neutrophils were identified as CD45⁺CD11b⁺Ly6G⁺. The intracellular markers (TNFα and IL-1β) were evaluated from neutrophil population (CD45⁺CD11b⁺Ly6G⁺). The monocyte/macrophage population was selected as CD45⁺CD11b⁺Ly6G^-CD64⁺. The surface markers (Ly6C and CD206) and intracellular markers (TNFα and IL-1β) were evaluated from monocytes/macrophage population (CD45⁺CD11b⁺Ly6G^-CD64⁺).

Bone marrow cells were collected from mouse femurs and tibias as previously described (25). 4×10⁶ cells were seeded in a 10 cm plate with differentiation medium (RPMI 1640 medium supplemented with 10 ng/mL macrophage colony-stimulating factor (M-CSF) (Peprotech, 315-02-250UG), 10% fetal bovine serum, 100U/ml penicillin, and 100 µg/ml streptomycin). The cells were incubated in a humidified incubator at 37°C with 5% CO₂ for 7 days. On day 7, 2×10⁵ cells/well were seeded to 96 well plates and treated with LPS 100 ng/mL for either 4 or 24 hours, and then incubated with Golgi Plug™ (BD Bioscience) for 4 hours to enhance the detectability of cytokine-producing cells in subsequent immunofluorescent staining. The cells were subsequently stained as described in the Flow cytometry staining and cell analysis section below. The surface staining antibodies were CD11b-eFluor450 (eBioscience, 48-0112-82), F4/80-PE-Cy7 (eBioscience, 25-4801-82), and CD38-PE (BD Bioscience, 553764) for 30 min on ice. The intracellular staining antibodies used were TNFα-Alexa Fluor700 (BD Bioscience, 558000), IL-1β-PE (Thermofisher, 12-7114-80), and iNOS-Alexa Fluor 488 (Thermofisher, 53-5920-80). The data were acquired by BD LSR Fortessa™ X-20 (BD Bioscience, San Jose, CA, USA). The flow gating strategy and representative plots are shown in Supplementary Figure 6 . BMDM were defined as CD11b⁺F4/80⁺. The intracellular markers (TNFα, IL-1β and iNOS) were evaluated from the macrophage population (CD11b⁺F4/80⁺).

Monocytes in circulation migrate to sites of inflammation and then differentiate into macrophages and dendritic cells to refill the pool of tissue immune cells (26). Circulating monocytes in mice are categorized into three subsets based on the surface expression of Ly6C; Ly6C^hi, Ly6C^inter and Ly6C^low (27). Ly6C^hi cells are pro-inflammatory monocytes which rapidly migrate into inflamed tissue driven by the expression of CC-chemokine receptor 2 (CCR2). Ly6C^inter cells are pro-inflammatory monocytes which are differentiated from Ly6C^hi. Ly6C^low subsets express low levels of CCR2, patrolling the vascular endothelium and engaging in tissue repair (27). To investigate neutrophils and monocytes in the circulation, peripheral blood of young and aged mice was investigated using flow cytometry. Aging did not change total neutrophil count and percentage, but significantly increased monocyte percentage in peripheral blood ( Figure 1A ). As expected, aging increased all three subsets of monocytes, pro-inflammatory Ly6C^hi, Ly6C^inter, and patrolling Ly6C^low compared to peripheral blood from young mice ( Figure 1A ). In summary, both pro-inflammatory and patrolling monocytes were increased in aging peripheral blood.

The peritoneal fluid has abundant immune cells, predominantly with mature macrophages and neutrophils, and plays an important role in modulating immune response of the host (28–30). The total neutrophil percentage was not changed with aging, but the percentage of TNFα⁺ neutrophils showed a trend of increase in the aged mice ( Figure 1B ). While IL-1β⁺ neutrophils remained unchanged, iNOS⁺ neutrophils significantly decreased in aged mice ( Figure 1B ). TNFα⁺ macrophages showed a trend of increase, whereas macrophages with other pro-inflammatory markers of CD38, IL-1β, and iNOS did not differ between young and aged mice ( Figure 1B ). The anti-inflammatory CD206⁺ macrophages also did not change with age ( Figure 1B ). Collectively, the aging effect was quite mild in neutrophils and macrophages of peritoneal fluid.

We previously showed that aging increases adiposity and inflammation in adipose tissues, which further exacerbates insulin resistance in aging (31, 32). To further elucidate whether aging re-programs innate immune cells of neutrophils, monocytes, and macrophages in epididymal fat, we studied stromal vascular fraction (SVF) of epididymal fat in young and aged mice. While aging did not alter the percentage of total neutrophils and macrophages, the percentage of infiltrating macrophages showed a trend of increase in aging ( Figure 2A ). In addition, CD38⁺ pro-inflammatory macrophages were significantly elevated in aging. While the percentage of CD206⁺ anti-inflammatory macrophages also showed a trend of increase in aging, that trend was highly varied in aged mice ( Figure 2A ). Collectively, aging increased pro-inflammatory myeloid cells, mainly by increasing infiltrating macrophages and CD38⁺ pro-inflammatory macrophages in aging epididymal fat.

Fatty liver diseases, including non-alcoholic fatty liver disease (NAFLD) and nonalcoholic steatohepatitis (NASH), are more prevalent in aging (33–35). Here we examined whether innate immune cells in liver are altered in aging. Aging did not change the total percentage of neutrophils and pro-inflammatory TNFα⁺, IL-1β⁺, and iNOS⁺ neutrophils ( Figure 2B ). However, aging significantly elevated the total percentage of macrophages, and the pro-inflammatory CD38⁺, TNFα⁺, and IL-1β⁺ macrophages were significantly higher in the aging liver. Notably, although the infiltrating macrophages were significantly increased in aging liver, the resident macrophages (Kupffer cells) did not change ( Figure 2B ). Collectively, aging primarily affects macrophages, exhibiting the infiltration of pro-inflammatory macrophages in the liver.

The gastrointestinal tract has the largest pool of immune cells in the body due to the greatest amount of antigen exposure from diet and microbiome (36). To investigate if innate immune populations are altered in the aging colon, we used flow cytometry to assess innate immune cells of young and aged mice. The percentage of the total colonic neutrophils was similar between young and aged mice. However, aging significantly increased IL-1β⁺ pro-inflammatory neutrophils in colon compared to young mice colon, whereas aging did not significantly alter TNFα⁺ neutrophils ( Figure 2C ). Macrophage total percentage did not change in the aging colon ( Figure 2C ). However, a robust change of macrophage subpopulations was detected. Ly6C^hi infiltrating monocytes and macrophages were significantly increased in aging ( Figure 2C ). Also, IL-1β⁺ pro-inflammatory macrophages were significantly elevated in aged mice colon, whereas TNFα⁺ macrophages remained similar in aging ( Figure 2C ). Furthermore, aging colon showed significantly decreased CD206⁺ anti-inflammatory macrophages ( Figure 2C ). Collectively, aging significantly enhances pro-inflammatory neutrophils and macrophages, and reduces anti-inflammatory macrophages in the colon.

In peripheral blood, young mice showed significantly increased neutrophils upon LPS exposure ( Figure 4A ). Aged mice also exhibited increased neutrophils upon LPS treatment, whereas the response was more robust than in young mice ( Figure 4A ). Interestingly, total monocyte percentage was significantly reduced in both young and old mice in response to LPS ( Figure 4B ). Furthermore, the monocyte subsets of Ly6C^hi, Ly6C^inter and Ly6C^low were all decreased by LPS in both young and aged mice ( Figure 4C ). Of note, old mice showed a more pronounced reduction of Ly6C^hi, Ly6C^inter and Ly6C^low monocytes as compared to that of young mice ( Figure 4C ). Collectively, peripheral blood neutrophils were increased, but total monocytes and monocyte subsets were decreased in both young and aged mice. On the other hand, aged mice showed more robust change of neutrophils and monocyte subsets in response to LPS as compared to young mice.

In the peritoneal fluid, young mice showed significantly elevated neutrophil percentage in response to LPS. The aged mice also showed a trend of increase in neutrophil percentage by LPS treatment ( Figure 5A ). Interestingly, the secretion of pro-inflammatory cytokines in neutrophils was the opposite between young and aged mice in response to LPS ( Figure 5A ): In young mice, TNFα⁺ and iNOS⁺ neutrophils were increased in peritoneal fluid; counterintuitively, aged mice exhibited decreased TNFα⁺ and iNOS⁺ neutrophils upon LPS exposure. Next, we found that the young mice displayed a trend of decrease in total macrophage percentage in response to LPS, whereas aged mice exhibited a significantly lower macrophage percentage under LPS treatment ( Figure 5B ). On the other hand, TNFα and iNOS expressing pro-inflammatory macrophages were significantly increased in both young and aged mice ( Figure 5B ). Of note, aged mice showed a less robust increase of TNFα⁺ macrophages than did young mice ( Figure 5B ). On the contrary, aged mice had a more significant increase of iNOS⁺ macrophages compared to young mice under LPS ( Figure 5B ). The response of pro-inflammatory CD38⁺ macrophages in LPS was also different in young and aged mice. Young mice showed increased CD38⁺ pro-inflammatory macrophages in response to LPS, while aged mice showed significantly decreased CD38⁺ macrophages ( Figure 5B ). In addition to pro-inflammatory markers, anti-inflammatory markers Arg1 and CD206 were also tested in macrophages. Under LPS exposure, young mice showed merely a trend of decreased Arg1⁺ macrophages, whereas aged mice showed significantly decreased Arg1⁺ macrophages ( Figure 5C ). Similarly, young mice did not show an altered CD206⁺ macrophage percentage, whereas aged mice showed a significantly reduced CD206⁺ macrophage percentage ( Figure 5C ). Collectively, aged mice showed an impaired response of neutrophils and macrophages to LPS. The macrophage subsets in aged mice showed differential responses to LPS depending on cytokine production.

In epididymal fat of young mice, LPS treatment did not change total myeloid cells whereas aged mice showed significantly increased myeloid cells ( Figure 6A ). The percentage of neutrophils was significantly increased following LPS stimulation in both groups ( Figure 6B ); however, the response was less pronounced in the aged mice compared to young mice ( Figure 6B ). On the other hand, macrophages in aged epididymal fat were more responsive to LPS than macrophages in epididymal fat from young mice. Total macrophages in young mice were not altered under LPS exposure, whereas the total macrophages in aged mice exhibited a significantly increased in epididymal fat ( Figure 6C ). Similarly, the macrophage subsets showed significant changes in aged mice but not in young mice ( Figure 6C ): While young mice did not show an altered macrophage subpopulation in response to LPS, aged mice revealed significantly increased CD38⁺ pro-inflammatory macrophages. The CD206⁺ anti-inflammatory macrophage population was not changed in either young or aged mice in response to LPS treatment ( Figure 6C ). Furthermore, in aged mice, there was significantly elevated macrophage infiltration and reduced resident macrophages upon LPS treatment ( Figure 6D ). Collectively, aging decreased neutrophil response but increased macrophage responses to inflammatory stimuli in epididymal fat.

Liver neutrophils were significantly elevated in both young and old mice in response to LPS ( Figure 7A ). Young mice showed significantly increased TNFα⁺ and IL-1β⁺ neutrophils in response to LPS, whereas these cells in livers of aged mice did not respond to LPS ( Figure 7A ). Old mice presented a robust change of pro-inflammatory macrophages compared to young mice in response to LPS ( Figure 7B ). CD38⁺ macrophages in young mice did not change, whereas old mice showed significantly elevated CD38⁺ macrophages upon LPS challenge ( Figure 7B ). In young mice, TNFα⁺ macrophages did not respond to LPS; however, this cell type was significantly reduced in old mice ( Figure 7B ). The IL-1β⁺ and iNOS⁺ macrophages were significantly increased in both young and old mice, but aged mice showed a more robust increase in those macrophages ( Figure 7B ). Interestingly, young mice showed increased infiltrating macrophages in response to LPS exposure, whereas aged mice did not show a change in macrophage infiltration ( Figure 7C ). The resident macrophages were not significantly altered in both young and aged mice following LPS treatment ( Figure 7C ). Collectively, liver neutrophils responded similarly between young and old mice, whereas pro-inflammatory macrophages from the livers of aged mice showed a more potent response to LPS.

Our data above support a significant aging effect of increased pro-inflammatory innate immune cells in the colon. Interestingly, colonic innate immune cells from both young and aged mice displayed relatively mild responsiveness to LPS challenge. Total neutrophil percentage was significantly higher in both young and aged mice colons, but aged mice colons showed a smaller increase of neutrophils in response to LPS as compared to young mice ( Figure 8A ). TNFα⁺ neutrophils were steady in both young and aged mice upon LPS ( Figure 8A ). While in young mice, IL-1β⁺ neutrophils remained non-responsive to LPS, aged mice showed significantly escalated IL-1β⁺ neutrophils in LPS exposure ( Figure 8A ). Total macrophage population from young mouse colons was increased by LPS, whereas the percentage of the total macrophages in aged mouse colons did not change ( Figure 8B ). Also, young mice showed a trend of increased Ly6C^hi macrophages, whereas aged mice showed a trend of decreased Ly6C^hi macrophages; however, neither reached statistical significance ( Figure 8B ). LPS did not alter TNFα⁺ and IL-1β⁺ macrophages in both young and aged mice ( Figure 8B ). Overall, total neutrophils and macrophages increased in response to LPS. While pro-inflammatory neutrophils were increased significantly with LPS stimulation in aged mice, pro-inflammatory macrophages barely responded to LPS in either young or aged colons.

The collective data of the age-associated alterations of immune cells in each tissue are summarized in Figure 9 . Also, responses of the immune cells to LPS in each tissue are summarized in Table 1 .