IL-12α KO and control wild type (WT) C57/6J mice of both genders were used for TAC or sham surgery. As the female mice generally gain less body weight as compared to the male mice, we used male mice at the age of ~7 to 9 weeks and female mice at the age of ~8 to 10 weeks for TAC or sham surgery. The TAC surgery was performed using a 27G needle to create aortic constriction in male mice as previously described (36). A 28G needle was used for the TAC procedure in female mice, as female mice are generally resistant to TAC-induced HF development as compared to male mice under the same aortic constriction. Following the final cardiac functional test at 8 weeks after TAC or sham surgery, cardiac and pulmonary tissues were collected for further histological, immune-histological, and flow cytometry studies. The mice were housed in a temperature-controlled environment with 12 hours light/dark cycles. This study was approved by the Animal Care and Use Committee at the University of Mississippi Medical Center.

Echocardiographic images were obtained using a Visualsonics Vevo 2100 imaging system as previously described (36). The heart rate, ejection fraction, fractional shortening, end-systolic diameter, and end-diastolic diameter were determined using echocardiography.

Cardiac and pulmonary samples were fixed in 10% formaldehyde, embedded in paraffin blocks, and sections of 10 µm were sliced. Sirius red and Fast green staining (Chondrex, Inc.) was done to quantify fibrosis in the heart and lung tissues. LV cardiomyocyte size was measured by staining the sliced sections with FITC-conjugated wheat germ agglutinin (AF488, Invitrogen). The cross-sectional area of at least 100 LV cardiomyocytes were measured and averaged to get mean cardiomyocyte size. Tissue fibrosis and cardiomyocyte cross-sectional area were quantified by using ImageJ Software. Infiltrating leukocytes and macrophages in the heart and lung tissues were stained with CD45 (R&D systems, AF114) and Mac2 antibody (R&D systems, AF1197), respectively. CD45 and Mac2 were visualized by using a secondary Alexa Flour-555 conjugated antibody. Pulmonary vessels were stained with smooth muscle actin (SMA) (Proteintech, 14395-1-AP) and Alexa Flour-594 conjugated Isolectin B4 (IB4) (Thermo Fisher Scientific, 121413). SMA was visualized by using a secondary Alexa Flour-555 conjugated antibody. Then, the tissue sections were mounted with a mounting media containing 4’, 6’-diamidino-2-phenylindole (DAPI) to visualize the nucleus. All of the histological images were taken using the Mantra Quantitative Pathology Imaging System (Perkin Elmer) and the immuno-histological images were analyzed by inForm software version 2.2.1 (Perkin Elmer) to quantify infiltrating leukocytes and macrophages in the tissues.

LV tissue samples were ground in liquid nitrogen and sonicated in PBS with 1X phosphatase and protease inhibitors. The protein content in the sample was quantified using the BCA protein assay kit (ThermoFisher Scientific, 23225). Then, 40µg protein was loaded on each well and separated using SDS-PAGE and transferred into methanol-activated PVDF membrane. The membrane was then blocked with 5% non-fat milk in TBST for 1 hour. Then, the membrane was incubated with primary antibodies for cardiac β-myosin heavy chain (Abcam, ab50967 at 1:1000 dilution), atrial natriuretic peptide (Thermo Fisher Scientific, PA5-29559 at 1:1000 dilution), and GAPDH (Cell Signaling Technologies Inc., CST 97166 at 1:1000 dilution) overnight at 4°C. Next day, the membrane was washed with TBST three times, 10 minutes each. The membrane was then incubated with HRP-conjugated secondary antibodies (Abcam ab97064, ab97030 at 1:4000 dilution in 5% non-fat milk in TBST) for 1 hour at room temperature. Then, the membrane was washed with TBST three times, 10 minutes each to remove unbound secondary antibodies. The protein band was detected using an iBright FL1500 instrument (Thermo Fisher Scientific) and the bands were quantified using ImageJ analysis software from the National Institutes of Health.

The heart and lung tissues were ground in liquid nitrogen and sonicated in PBS with 1X phosphatase and protease inhibitors. The tissue lysate was centrifuged at 15000 rcf for 15 minutes to get the supernatant. The concentration of cytokines in the supernatant was measured by LEGENDplex flow-based 13-plex mouse inflammation panel (BioLegend, 740446).

Cells were isolated from lung and flow cytometry was performed as previously described (8). Briefly, lung tissue was minced into small pieces and digested in 5 ml collagenase digestion buffer (HBSS without Ca⁺⁺/Mg⁺⁺ (Life Technologies Corporation), 1mg/dL collagenase D (Roche Diagnostics, Germany)) at 37°C for 30 minutes using a cell dissociator (Miltynyi Biotec) and the cells were passed through a 100 µm cell strainer. Then, the RBC was lysed using 2 ml of RBC lysis buffer (Life Technologies Corporation). The remaining cells were counted using an automated cell counter (Bio-Rad Technologies, Inc.). Dead cells were stained with Zombie aqua dye (Biolegend). The cell suspensions were pre-incubated with anti-mouse CD16/32 antibody to prevent non-specific binding of antibodies to FcRγ and then stained with multi-staining antibodies ( Supplementary Table 1 ). The samples were subjected to FACS BD LSRII analysis (BD Biosciences). Data were analyzed using FlowJo_V10 (FlowJo, OR) software. Gating strategies are presented in Supplementary Figures 1 and 2 .

Data were presented as mean ± SEM. A two-way ANOVA followed by a Bonferroni correction post-hoc test was used to test the differences between more than 2 groups. All pair-wise p-values are two sided. The null hypothesis was rejected at p<0.05.

Since echocardiograms showed that LV ejection fraction and LV fractional shortening were similar in WT and IL-12α KO male and female mice under control conditions ( Figures 1A, B ; Supplementary Figure 3 ). As compared with the corresponding sham group, TAC caused significant decreases of LV ejection fraction and fractional shortening in both WT and IL-12α KO male mice ( Figures 1A, B ). Interestingly, TAC-induced LV dysfunction was significantly attenuated in IL-12α KO male mice as compared with corresponding WT mice ( Figures 1A, B ). TAC-induced increases of LV end-systolic diameter and/or end-diastolic diameter were comparable in WT and IL-12α KO male mice ( Figures 1C, D ). Heart rates were comparable between WT and IL-12α KO male mice in both sham and TAC groups ( Supplementary Table 2 ).

Consistent with the observation in male mice, IL-12α KO also significantly attenuated TAC-induced LV dysfunction in female mice as evidenced by significantly less reduction of LV ejection fraction, less reduction in LV fractional shortening, and a smaller increase in the LV end-systolic diameter ( Supplementary Figure 3 ).

Western blots demonstrated that TAC-induced expression of LV beta-myosin heavy chain (β-MHC) and atrial natriuretic peptide (ANP), two commonly used biomarkers for pathological cardiac hypertrophy and the consequent heart failure, was significantly attenuated in IL-12α KO male mice ( Figures 1E–G ; Supplementary Figure 4 ).

LV weight, LA weight, lung weight, RV weight, total heart weight, and their ratios to tibial length or body weight were similar in WT and IL-12α KO male and female mice under control conditions ( Figures 1H–O ; Supplementary Tables 2, 3 , Supplementary Figure 3 ). Body weight and tibial length were comparable in WT and IL-12α KO mice either under control condition or after TAC for both genders ( Supplementary Tables 2, 3 ). TAC-induced increases of LV weight, lung weight, and their ratios to tibial length or body weight were significantly attenuated in both IL-12α KO male and female mice as compared with corresponding WT mice of the same gender ( Figures 1H–M ; Supplementary Tables 2, 3 , Supplementary Figure 3 ). TAC-induced increases of RV weight and its ratios to tibial length or body weight were only significantly attenuated in male IL-12α KO mice as compared with corresponding WT mice ( Figures 1N, O ; Supplementary Table 2 ). These data indicate that IL-12α KO effectively attenuated TAC-induced LV hypertrophy, and the consequent increases of lung weight in mice of both genders.

Since IL-12α KO significantly attenuated TAC-induced LV hypertrophy and dysfunction in both male and female mice, we further determined TAC-induced LV cardiomyocyte hypertrophy using FITC-conjugated wheat germ agglutinin (WGA) staining in male mice. As compared with corresponding WT mice, IL-12α KO significantly attenuated TAC-induced increase in LV cardiomyocyte size ( Figures 2A, B ).

Since IL-12α KO significantly attenuated TAC-induced LV hypertrophy and dysfunction in both male and female mice, we further determined TAC-induced LV fibrosis using Sirius red/Fast green staining in male mice. As compared with corresponding WT mice, IL-12α KO significantly attenuated TAC-induced increase in LV fibrosis ( Figures 2C, D ). In addition, LV leukocyte and macrophage infiltrations were significantly increased in WT mice after TAC, while TAC-induced LV leukocyte and macrophage infiltrations were significantly attenuated in IL-12α KO mice ( Figures 2E–H ).

Monocyte chemoattractant protein 1 (MCP1), also commonly referred to as chemokine ligand 2 (CCL2), exerts an important role in regulating the inflammatory response by enhancing the recruitment of monocytes, memory T cells, and dendritic cells to the sites of injury or infection. We found that TAC caused a significant increase of LV MCP1 protein expression in WT mice but not in IL-12α KO mice ( Supplementary Figure 5 ). In addition, TAC caused significant increases of LV IL-1α and IL-27 in WT but not in IL-12α KO mice, while TAC did not affect LV IFNβ and IL-6 in both WT and KO mice ( Supplementary Figure 5 ). TAC caused similar decrease of LV IL-17A in WT and KO mice ( Supplementary Figure 5 ).

Since our previous studies demonstrated that HF could cause severe lung inflammation and remodeling, and lung inflammation and remodeling could further promote HF progression (24), we also determined lung leukocyte infiltration, fibrosis, and vessel remodeling in male mice. As anticipated, TAC also caused significant accumulation of pulmonary CD45⁺ leukocytes and the macrophage subset in WT mice as compared to control sham mice, while the TAC-induced accumulation of pulmonary CD45⁺ leukocytes and macrophages was significantly attenuated in IL-12α KO mice ( Figures 3A–D ). TAC also caused significant increase of lung fibrosis in WT mice as compared to control sham mice, while TAC-induced lung fibrosis was significantly attenuated in IL-12α KO mice as compared with corresponding WT mice ( Figures 3E, F ). Moreover, the number of muscularized pulmonary vessels were significantly increased in WT mice after TAC, while TAC-induced pulmonary vessel remodeling was significantly attenuated in IL-12α KO mice ( Figures 3G, H ).

Furthermore, pulmonary MCP-1 was also significantly increased in WT mice after TAC, while TAC-induced increase of pulmonary MCP1 was significantly attenuated in IL-12α KO mice ( Figure 3I ). Moreover, the concentrations of pro-inflammatory cytokines such as interleukin-6 (IL-6), interleukin-1α (IL-1α), interleukin-1β (IL-1β), granulocyte macrophage colony-stimulating factor (GM-CSF), interleukin-27 (IL-27), tumor necrosis factor α (TNFα), and interferon β (IFNβ) were significantly increased in WT mice after TAC, while TAC-induced increases in these pro-inflammatory cytokines were abolished in the lung of IL-12α KO mice ( Figures 3J–L , Supplementary figure 6 ). TAC did not significantly affect lung interleukin-17A (IL-17A) expression in both WT and KO mice ( Supplementary Figure 6 ).

Since IL-12α KO significantly attenuated TAC-induced lung leukocyte accumulation ( Figures 3A, B ), we further determined the alterations of the major lung leukocyte subsets in these mice. As we previously found that CD11c⁺ antigen presenting cells (APCs) play an important role in TAC-induced cardiac hypertrophy and dysfunction (21), we determined the pulmonary dendritic cells in WT and IL-12α KO mice ( Figure 4 ). As most lung alveolar macrophages are CD11c^high cells, to exclude the influence of alveolar macrophages, lung CD11c^highF4/80^- cells were gated as dendritic cells ( Figure 4A ). We found that the percentages of CD11c^highF4/80^- dendritic cells within CD45⁺ cells were significantly increased in WT mice after TAC, while IL-12α KO significantly attenuated TAC-induced increase of dendritic cells ( Figure 4B ). As increased MHCII expression in dendritic cells generally indicates dendritic cell activation, the percentages of MHCII^highCD11c^highF4/80^- cells were further determined ( Figures 4C, G ). The percentage of MHCII^highCD11c^highF4/80^- cells within CD45⁺ cells was significantly increased in WT mice after TAC, while IL-12α KO significantly abolished TAC-induced increase of the percentage of MHCII^highCD11c^highF4/80^- cells ( Figure 4E ). In addition, the average MHCII expression in CD11c^highF4/80^- cells, as shown by the GEO mean, was significantly increased in WT mice after TAC, while IL-12α KO significantly attenuated the TAC-induced increase of MHCII expression in CD11c^highF4/80^- cells ( Figure 4F ).

Lung macrophages exert important roles in regulating the lung inflammatory response. While the alterations of lung macrophages during viral or bacterial infections have been well documented, the alterations of lung macrophage subsets and their activation status after HF development have not been previously described. Thus, we determined lung macrophage subsets and their activation status in WT and IL-12α KO mice after HF development.

We found that the percentage of F4/80⁺ macrophages within CD45⁺ cells were also significantly increased in WT mice after TAC, while this increase in the percentage of F4/80⁺ macrophages within CD45⁺ cells was significantly attenuated in IL-12α KO mice after TAC ( Figures 5A, B ). IL-12α KO did not affect the percentage of lung macrophages within CD45⁺ leukocytes under control conditions ( Figures 5A, B ).

Since MHCII expression in macrophages reflects their antigen-presenting capacity and polarization, macrophage MHCII expression and the percentage of MHCII^highF4/80⁺ cells were determined ( Figures 5C–H ). Interestingly, the percentages of pulmonary MHCII^highF4/80⁺ macrophages within F4/80⁺ cells or CD45⁺ cells were significantly increased in WT mice after TAC, while IL-12α KO abolished the TAC-induced increase of the percentage of MHCII^highF4/80⁺ ( Figures 5D, E ). TAC-induced increase of MHCII expression in pulmonary F4/80⁺ cells was also significantly attenuated in the 12α KO mice ( Figure 5F ).

Since lung tissues contain various macrophage subsets, and the composition of lung macrophages contribute to pulmonary inflammation, we further gated the lung F4/80⁺ macrophages according to the expression of CD11c, CD11b, MHCII, and Ly6C, a group of cell surface makers commonly used for the classification of pulmonary macrophage subsets (37–40). Accordingly, F4/80⁺ macrophages were further gated to identify the alveolar macrophages (CD11c^highCD11b^lowF4/80⁺) ( Figure 6 ), Ly6C^low interstitial macrophages (Ly6C^lowCD11c^lowCD11b^highF4/80⁺) ( Figure 7 ), and the Ly6C^high monocyte derived pro-inflammatory interstitial macrophages (Ly6c^highCD11c^lowCD11b^highF4/80⁺) ( Figure 8 ), as well as the corresponding subsets according to their MHCII expression.

As shown in Figure 6 , the percentages of pulmonary alveolar macrophages (CD11c^highCD11b^lowF4/80⁺) were not significantly increased in either WT or IL-12α KO mice after TAC, but the percentage of pulmonary alveolar macrophages within CD45⁺ leukocytes in IL-12α KO mice was significantly reduced as compared with WT mice after TAC ( Figures 6A, B ). Interestingly, the percentages of MHCII^highCD11c^highCD11b^lowF4/80⁺ cells within alveolar macrophages subset, F4/80⁺ cells, or within CD45⁺ leukocytes were all significantly increased in WT mice after TAC, while the above changes were largely abolished in IL-12α KO mice after TAC ( Figures 6C–F, H ). Furthermore, the average MHCII protein expression of pulmonary alveolar macrophages was significantly increased in WT mice after TAC, while IL-12α KO abolished the TAC-induced increase of MHCII protein expression in alveolar macrophages ( Figure 6G ).

As presented in Figure 7 , TAC caused a significant increase of the percentage of Ly6c^low interstitial macrophages within CD45⁺ leukocytes in WT mice after TAC, while IL-12α KO significantly attenuated the TAC-induced increase of lung Ly6c^low interstitial macrophages ( Figures 7A, B ). Furthermore, the percentages of MHCII^highLy6c^low interstitial macrophages within Ly6c^low interstitial macrophages subset, F4/80⁺ cells, or within CD45⁺ leukocytes were all significantly increased in WT mice after TAC, while the above changes were completely abolished in IL-12α KO mice after TAC ( Figures 7C–F, H ). In addition, the average MHCII protein expression of Ly6C^low interstitial macrophages was significantly increased in WT mice after TAC, while IL-12α KO abolished this TAC-induced increase of MHCII protein expression ( Figure 7G ).

Lung Ly6C^highCD11c^lowCD11b^highF4/80⁺ cells are generally recognized as Ly6C^high monocyte-derived classical pro-inflammatory interstitial macrophages. As presented in Figure 8 , we found that TAC did not affect the percentage of Ly6C^high interstitial macrophages in both WT and IL-12α KO mice ( Figures 8A, B ). Although it is generally reported that lung Ly6c^highCD11c^lowCD11b^highF4/80⁺ macrophages have low MHCII expression or no MHCII expression under basal conditions, while MHCII expression in Ly6c^highCD11c^lowCD11b^highF4/80⁺ interstitial macrophages is relatively lower as compared to the MHCII expression in lung alveolar macrophages (CD11c^highCD11b^lowF4/80⁺) ( Figure 6 ), interestingly, MHCII expression in Ly6c^high interstitial macrophages was significantly increased in WT mice after TAC ( Figures 8C–H ). IL-12α KO significantly attenuated TAC-induced increase of MHCII expression in Ly6c^high interstitial macrophages ( Figures 8C–H ).

We also determined the lung neutrophils (CD11b^highLy6G⁺), and monocytes (F4/80^-Ly6C^high), two important leukocyte subsets in regulating lung inflammatory response. The percentage of neutrophils within CD45⁺ cells was not significantly changed in KO mice under control conditions and in WT and KO mice after TAC ( Supplementary Figure 7 ). While TAC did not affect the percentages of lung monocytes within CD45⁺ in both wild type and KO mice, the percentages of lung monocytes were significantly lower in IL-12α KO mice after TAC as compared with WT mice after TAC ( Supplementary Figure 8 ).

Since our previous studies showed that TAC resulted in increased cardiac and lung T cell activation, and T cell activation exerts an important role in HF development (3), we determined the activation of CD3⁺, CD4⁺, and CD8⁺ T cells in the lung of WT and IL-12α KO mice ( Figure 9 ). As anticipated, TAC resulted in significantly increased pulmonary CD3⁺ effector memory T cells (CD44⁺CD62L^-CD3⁺), increased CD3⁺ central memory T cells (CD44⁺CD62L⁺CD3⁺), and decreased of CD3⁺ naïve T cells (CD44^-CD62L⁺CD3⁺) cells in WT but not in IL-12α KO mice, and TAC resulted in a significantly smaller increase of pulmonary CD44⁺CD62L^-CD3⁺ T cells and a smaller decrease of CD44^-CD62L⁺CD3⁺ T cells in IL-12α KO mice as compared with corresponding WT mice ( Figures 9A–F ). The percentage of CD44^-CD62L^-CD3⁺ T cells was significantly decreased in WT mice but not in IL-12α KO mice after TAC (data not shown).

The percentages of effector memory CD4⁺ T cells (CD44⁺CD62L^-CD4⁺) and effector memory CD8⁺ T cells (CD44⁺CD62L^-CD8⁺) were also significantly increased in WT after TAC, while the percentage of naïve CD4⁺ T cells (CD44^-CD62L⁺CD4⁺) were significantly decreased in WT mice after TAC ( Figures 9G, H, J–N, Q, R ). TAC-induced increases of pulmonary effector memory CD44⁺CD62L^-CD4⁺ and CD44⁺CD62L^-CD8⁺ T cells were significantly attenuated in IL-12α KO mice ( Figures 9G, H, J–N, Q, R ). The percentages of central memory CD44⁺CD62L⁺CD4⁺ and CD44⁺CD62L⁺CD8⁺ T cells were unaffected by TAC in both WT and IL-12α KO group ( Figures 9G, I, M, O ). The percentages of CD44^-CD62L^-CD4⁺ and CD44^-CD62L^-CD8⁺ T cells were also unaffected by TAC in both WT and IL-12α KO group (data not shown).

Furthermore, the percentage of lung CD4⁺ within CD3⁺ T cells in the lung was significantly decreased in WT mice after TAC, and this change was abolished in IL-12α KO after TAC ( Supplementary Figure 9 ). The percentage of lung CD8⁺ within CD3⁺ T cells was not significantly changed in WT and IL-12α KO mice after TAC ( Supplementary Figure 9 ). The percentages of CD3⁺ and CD4⁺ T cells in the CD45⁺ leukocyte population were significantly decreased in WT mice after TAC, while these changes were abolished in IL-12α KO after TAC ( Supplementary Figure 9 ). The reduced percentage of CD3⁺ and CD4⁺ T cells within CD45⁺ leukocytes in WT mice after TAC appears to be the outcome of an increased percentage of lung macrophages within CD45⁺ leukocytes in these mice ( Figure 5B ).