Macrophage Paired Immunoglobulin-Like Receptor B Deficiency Promotes Peripheral Atherosclerosis in Apolipoprotein E–Deficient Mice

Background: Peripheral atherosclerotic disease (PAD) is the narrowing or blockage of arteries that supply blood to the lower limbs. Given its complex nature, bioinformatics can help identify crucial genes involved in the progression of peripheral atherosclerosis. Materials and Methods: Raw human gene expression data for 462 PAD arterial plaque and 23 normal arterial samples were obtained from the GEO database. The data was analyzed using an integrated, multi-layer approach involving differentially-expressed gene analysis, KEGG pathway analysis, GO term enrichment analysis, weighted gene correlation network analysis, and protein-protein interaction analysis. The monocyte/macrophage-expressed leukocyte immunoglobulin-like receptor B2 (LILRB2) was strongly associated with the human PAD phenotype. To explore the role of the murine LILRB2 homologue PirB in vivo, we created a myeloid-specific PirB-knockout Apoe −/− murine model of PAD (PirB MΦKO) to analyze femoral atherosclerotic burden, plaque features of vulnerability, and monocyte recruitment to femoral atherosclerotic lesions. The phenotypes of PirB MΦKO macrophages under various stimuli were also investigated in vitro. Results: PirB MΦKO mice displayed increased femoral atherogenesis, a more vulnerable plaque phenotype, and enhanced monocyte recruitment into lesions. PirB MΦKO macrophages showed enhanced pro-inflammatory responses and a shift toward M1 over M2 polarization under interferon-γ and oxidized LDL exposure. PirB MΦKO macrophages also displayed enhanced efferocytosis and reduced lipid efflux under lipid exposure. Conclusion: Macrophage PirB reduces peripheral atherosclerotic burden, stabilizes peripheral plaque composition, and suppresses macrophage accumulation in peripheral lesions. Macrophage PirB inhibits pro-inflammatory activation, inhibits efferocytosis, and promotes lipid efflux, characteristics critical to suppressing peripheral atherogenesis.

macrophages also displayed enhanced efferocytosis and reduced lipid efflux under lipid exposure.
Conclusion: Macrophage PirB reduces peripheral atherosclerotic burden, stabilizes peripheral plaque composition, and suppresses macrophage accumulation in peripheral lesions. Macrophage PirB inhibits pro-inflammatory activation, inhibits efferocytosis, and promotes lipid efflux, characteristics critical to suppressing peripheral atherogenesis.
Keywords: atherosclerosis, PAD, LILRB2, PirB, apolipoprotein BACKGROUND The narrowing or blockage of arteries that supply blood to the lower limbs is known as peripheral atherosclerotic disease (PAD). The principal cause of PAD is the atherosclerotic occlusion of arteries supplying the affected limbs. Although the disease is mostly asymptomatic, a commonplace clinical presentation is intermittent claudication (i.e., pain on walking). More severe clinical manifestations include critical limb ischemia (CLI), which presents as pain even during rest as well as tissue loss due to ulceration or gangrene (Morley et al., 2018). PAD is estimated to affect about 13% in adults of the Western population aged 50 or above (Morley et al., 2018). Mortality due to cardiovascular disease is seen in 10-15% of patients with intermittent claudication within 5 years of diagnosis (Norgren et al., 2007). Based on this evidence, it is important to identify the pathophysiological mechanism(s) underlying PAD progression, which can provide guidance towards more effective management of PAD patients.
However, the molecular pathophysiology underlying PAD is complicated, as there are a number of pathways, proteins, and cell types involved in PAD progression (Scholz et al., 2002;Coats and Wadsworth, 2005;Kuang et al., 2008). The primary cells implicated in the development of PAD include macrophages, vascular endothelial cells (ECs), resident stem cells, platelets, vascular smooth muscle cells (SMCs), fibroblasts, and pericytes (Scholz et al., 2002;Coats and Wadsworth, 2005;Kuang et al., 2008). In an otherwise healthy individual, tissue damage due to progressive limb ischemia progresses along a continuum. Initially, the body attempts to homeostatically restore blood supply to the affected limb(s) by angiogenic and arteriogenic pathways. To further resolve limb ischemia and tissue damage, inflammatory, vascular remodeling, and apoptotic pathways are activated. However, in patients diagnosed with CLI, such compensatory mechanisms are inefficient to restore sufficient blood flow. Due to this, there is continued inadequacy in perfusion coupled with enhanced chronic inflammation, EC dysfunction, and oxidative stress. Gangrene can then present as a consequence of muscle fiber damage due to persistently high levels of oxidative stress (Hickman et al., 1994;Bhat et al., 1999;Pipinos et al., 2008b;a;Koutakis et al., 2015).
Given the complex nature of the molecular pathways at play in PAD, bioinformatics analysis of arterial gene expression datasets from PAD patients and healthy individuals can help in identifying crucial genes involved in the progression of peripheral atherosclerosis. To this end, we obtained raw gene expression data for 462 PAD arterial plaque samples and 23 normal arterial samples from the GEO database. We analyzed the data using an integrated, multi-layer approach involving differentially-expressed gene (DEG) analysis, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, Gene Ontology (GO) term enrichment analysis, weighted Gene Correlation Network Analysis (WGCNA), and protein-protein interaction (PPI) network analysis (Bindea et al., 2013;Szklarczyk et al., 2016). Based on this integrated approach, we discovered the inhibitory monocyte/macrophage-expressed receptor-leukocyte immunoglobulin-like receptor B2 (LILRB2, LIR-2, ILT-4; ENSG00000131042, human chromosomal region 19q13.4)-to be strongly associated with the PAD phenotype. We then investigated the role of the murine analogue of human LILRB2, paired immunoglobulin-like receptor B (PirB, Lilrb3, Lir-3, Gp91; ENSMUSG00000058818, mouse chr7: 3,711,409-3,720,391(-)), in a myeloid-specific PirB-null Apoe −/− murine model of PAD. The application of our integrated approach can help provide much-needed guidance into the molecular mechanisms underlying PAD and other forms of atherosclerotic disease.

METHODS
The experimental methods are fully detailed in the Supplementary Material.

DEGs in PAD and Their Functional Analysis
Application of DEG cut-off thresholds (log|fold-change (FC)| ≥ 1.5 and adjusted p < 0.05) following data processing of 13,467 common genes (Supplementary Figure S1; Supplementary File S2A) yielded a total of 680 DEGs, with 545 genes upregulated and 135 genes downregulated in PAD samples as compared with normal controls (Supplementary File S2B). Functional enrichment analysis of the 680 DEGs using gene ontology (GO, sub-ontologies: BP, CC, MF) and KEGG identified 21 enriched GO terms and 14 enriched KEGG pathways (Supplementary Figure S2; Supplementary File S3). The GO terms that were most significantly enriched included positive regulation of T-cell proliferation (GO_BP, p 2.16E-15), positive regulation of phagocytosis (GO_BP,, and tertiary granule membrane (GO_CC, p 2.28E-11). The top-ranking KEGG pathways were hematopoietic cell lineage (p 6.15E-18), rheumatoid arthritis (p 1.68E-12), and Staphylococcus aureus infection (p 5.37E-12).

WGCNA Analysis
For WGCNA analysis, we took the variance-filtered genes (≥50%) from the 485 samples that were batch-corrected using the ComBat algorithm. A soft-threshold power of 5 was chosen using the function pickSoftThreshold (Supplementary Figure S3A). Then, we generated similarity matrices based on the selected soft-threshold power; the adjacency matrix was first calculated and then converted into a topological overlap matrix (TOM) to minimize noise and spurious association (Supplementary Figure S3B). From the TOM, gene modules were identified where hierarchical clustering of the genes was done based on the TOM dissimilarity measure (Supplementary Figure S3C). Using the cutreeDynamic function, we obtained 22 significant gene modules and then assigned a unique color label to each module (note: smaller modules were merged) (Supplementary Figure S3D). After characterizing the 22 gene modules (Supplementary Figures S4, S5), we analyzed the module-trait relationship to find how each of the 22 gene modules correlated to the two phenotype traits (i.e., healthy control and PAD). From the 22 gene modules, the turquoise module (Supplementary File S2C) and pink module (Supplementary File S2D) were most characteristic of the PAD phenotype based on module-trait correlation analysis (Supplementary Figure S6). We then assessed the correlations between module membership (MM) and gene significance (GS) for PAD and found the turquoise and pink modules to show the highest correlations (0.49 and 0.48, respectively) to the PAD phenotype (Supplementary Figure S7).

PPI Network Analysis
PPI pairs were predicted for the 680 DEGs using the STRING database, with a minimum required interaction score of 0.700 (high confidence). The PPI network consisted of 444 nodes (denoting genes) and 2,398 edges (denoting interactions between genes) (Supplementary File S4). Degrees of the PPI network nodes obeyed exponential distribution (r-squared 0.849), indicating it is a scale-free network. The hub molecules of the PPI network were the nodes with the highest degrees; the top-ranking hub nodes were the MAC-1 subunit integrin-α-M (ITGAM/CD11B, degree 64), the MAC-1 subunit integrin-β-2 (ITGB2/CD18, degree 59), formyl-peptide receptor type 2 (FPR2, degree 56), protein tyrosine phosphatase receptor-type C (PTPRC/CD45, degree 51), and spleen tyrosine kinase (SYK, degree 50) (Supplementary File S4). Notably, four of the top five PPIderived hub nodes overlapped with the WGCNA-derived turquoise module (ITGAM/CD11B, ITGB2/CD18, PTPRC/ CD45, and SYK), suggesting some convergence between gene-level correlations and protein-level interactions in PAD. The other significant interactions from the PPI network were CDK1 interacting with PLK1 and FOS, LPAR3 interacting with FOS and SAA1, and FOS further interacting with RLN3 (relaxin 3). These proteins and interactions likely play key roles in the PAD phenotype.
The MCODE plug-in in Cytoscape was used for identifying the significant protein clusters within the PPI network. A total of 14 clusters were obtained; however, only four clusters with a score greater than ten were considered for analysis (Supplementary Figure S8). Cluster 1 (score 28.00) consisted of 28 nodes and 378 edges, cluster 2 (score 19.24) consisted of 22 nodes and 202 edges, cluster 3 (score 13.69), had 13 nodes and 89 edges, and cluster 4 (score 13.47) had 16 nodes and 101 edges (Supplementary Figure S8). As the top-ranking PPI hub nodes (the MAC-1 subunits ITGAM and ITGB2) were both located in cluster 2 and the leukocyte integrin MAC-1 has a well-recognized role in atherosclerosis (Aziz et al., 2017;Ramos et al., 2018), we chose to focus on MAC-1 protein interactors within cluster 2 to identify key regulators in PAD ( Figure 1A; Supplementary File S4). Based on our ranking analysis of MAC-1 protein interactors within cluster 2, we identified the inhibitory monocyte/macrophage receptor LILRB2 as a key potential regulator of MAC-1 and chose this protein for further analysis.

Macrophage PirB Knockout Increases Peripheral Atherogenesis and Plaque Vulnerability In Vivo
After 2 weeks of cuff-induced atherogenesis in HFD-fed mice, the mice were assessed for femoral atherosclerotic lesions. A significant increase was observed in the Oil Red O positive lesion area in the femoral arteries in PirB MΦKO mice ( Figure 1B). The effect was confirmed in both sexes indicating that PirB deficiency's effects on peripheral atherosclerosis is not sex-specific ( Figure 1B). Consistent with the increased atherogenesis, we also observed significant increases in femoral plaque sizes and intima/ media ratios in PirB MΦKO mice ( Figures 1C-E); however, we did not observe significant differences in overall luminal stenosis between PirB flox and PirB MΦKO mice ( Figure 1F). As expected, we noted increases in Mac3-positive cell (macrophage) content ( Figure 1G) in PirB MΦKO plaques. PirB MΦKO plaques also showed lower α-SMA-positive cell (SMC) content ( Figure 1H), more thinning in SMCpositive fibrous caps ( Figure 1H), and reduced collagen content by Masson's trichrome stain ( Figure 1I). Since there was a reduction in collagen content, we wanted to assess if this could be due to increased matrix metalloproteinase (Mmp) activity. Significant increases in protein levels of Mmp-1, Mmp-2, Mmp-8, Mmp-9, Mmp-12, Mmp-13, and Mmp-14 were observed in femoral plaques from PirB MΦKO mice (( Figure 1J). Validating our findings, we also observed increased expression of Mmp-1, Mmp-2,  in PirB MΦKO peritoneal macrophages in vitro ( Figure 1K).
Since our results in PirB MΦKO indicated enhanced peripheral plaque vulnerability, we evaluated the other indicators of femoral artery plaque vulnerability. Femoral artery cross-sections were stained with Carstairs' (Carstairs, 1965) and Verhoeff-Van Gieson stain (Kozaki et al., 2002) to identify disruptions in fibrous cap, intra-plaque hemorrhage, deposition of fibrin, or medial elastin breaks (Supplementary Figure S10). If any of these features were observed in three consecutive sections, we considered it a positive sign of plaque vulnerability. PirB MΦKO mice showed enhanced plaque vulnerability with higher percentages of intraplaque hemorrhage (11/
As nuclear factor-κB (NFκB) activation enhances inflammatory cytokine and chemokine production, we analyzed NFκB DNA binding activity. We observed a significantly greater NFκB DNA binding activity in PirB MΦKO macrophages that was further augmented by IFNγ ( Figure 2E). Further assessing the participation of NFκB, we exposed PirB MΦKO cells to the NFκB inhibitors BMS-345541 and parthenolide. Both BMS-345541 and parthenolide totally eliminated the increases in IFNγ-driven Il-6 production ( Figures 2F,G) and IFNγ-driven Tnfα production ( Figures  2H,I). This evidence supports the role of NFκB-dependent cytokine production in the pro-inflammatory response of PirB MΦKO macrophages to IFNγ.  Figure S11A-C).
Frontiers in Cell and Developmental Biology | www.frontiersin.org March 2022 | Volume 9 | Article 783954 6 deficiency favors a shift toward M1 polarization through a Shp1dependent mechanism.

Macrophage PirB Knockout Enhances Macrophage Efferocytosis In Vitro
Efferocytosis is the phagocytic clearance of apoptotic cells and is a crucial anti-inflammatory process performed by   Figures S12A-E).
We next evaluated the effect of PirB deficiency on macrophage efferocytosis of CMFDA-labeled apoptotic thymocytes. In PirB flox macrophages, M1 activation by IFNγ+LPS leads to repression of efferocytosis activity as measured by short-term CMFDA+ apoptotic thymocyte uptake (Supplementary Figure S12F) as well as phagocytic indices from longer-term single-feeding and double-feeding experiments (Supplementary Figures S12G,H). In contrast, M2 activation by IL-4 had no influence on these efferocytosis parameters ( Supplementary Figures S12F-H). In PirB flox macrophages, oxLDL did not show any noticeable effect on these efferocytosis parameters. However, oxLDL enhanced efferocytosis activity in PirB MΦKO macrophages ( Supplementary Figures S12F-H). Thus, PirB deficiency appears to promote efferocytosis activity under oxLDLtreated conditions.

Macrophage PirB Knockout Downregulates Cholesterol Efflux In Vitro
Macrophage accumulation of modified LDLs produces atherogenic foam cells (Orekhov, 2018). Here, we tested for peritoneal macrophages were either nonactivated (NA) or exposed to oxLDL and IFNγ and LPS for M1 activation or IL-4 for M2 activation for 24 h and exposed to (A) acetylated LDL (acLDL) or (B) oxidized LDL (oxLDL) for 48 h. Oil Red O staining was quantified as the relative fluorescent unit (RFU) per cell. (C,D) Macrophages were exposed to acLDL and 3H-cholesterol for 24 h; (C) apolipoprotein AI (ApoAI)-or (D) high-density lipoprotein (HDL)-dependent 3H-cholesterol efflux were analyzed for 24 h. (E) Macrophages were exposed to acLDL for 24 h followed by Western blotting analysis of Abca1, Abcg1, and Srb1 protein expression. (F) AcLDL-exposed macrophages were treated with GW3965 (1 µM) or T0901317 (1 µM) for 48 h and assessed for Abcg1 expression. (G-J) Adherent PirB MΦKO macrophages that had been transfected with WT PirB or a Myctagged PirB ITIM domain tyrosine phosphorylation site mutant (PirB 3Y-F ) were exposed to acLDL for 24 h. (G) Immunoprecipitation (IP) with an anti-Jak1 antibody (or IgG control) and standard immunoblotting (IB) followed by densitometric analysis of Shp1-Jak1 binding and Stat1 phosphorylation.   Figures 5A,B), while M2 cells accumulated higher lipid content relative to NA cells following exposure to acLDL ( Figure 5A). Notably, PirB knockout did not affect lipid accumulation, regardless of NA/M1/M2 status or lipid exposure ( Figures 5A,B). Cholesterol efflux also plays a role in macrophagic lipid incorporation, we tested activated macrophages for cholesterol efflux activity. PirB knockout downregulated apoAI-dependent efflux in M1 macrophages ( Figure 5C) and HDL-dependent efflux in NA, M1, and M2 macrophages ( Figure 5D).

Macrophage PirB Knockout Enhances Monocyte Recruitment, Macrophage Efferocytosis, and Inflammatory Marker Expression in Femoral Plaques In Vivo
As we had observed an upsurge in Mac3-positive plaques in PirB MΦKO mice, we wanted to assess several key indicators of macrophage activity in atherosclerotic lesions, namely: 1) recruitment of monocytes, 2) intra-lesion macrophage proliferation, 3) intra-lesion macrophage apoptosis, and 4) inflammatory marker expression (Robbins et al., 2013;Randolph, 2014). For evaluating the recruitment of monocytes into femoral lesions ( Figure 6A), we administered vehicle or clodronate followed by labeling of circulating monocytes by injecting red microspheres into murine tail veins. As these microspheres cannot pass through the interstitial space, all red-fluorescent/DAPI-positive cells detected within plaques by flow cytometry were considered recruited and infiltrated from the circulation. There were no detectable redfluorescent cells (i.e., Ly6C lo monocytes) with vehicle administration in either PirB flox or PirB MΦKO plaques (Higashi et al., 2016). However, due to administration of clodronate, redfluorescent labeling was applied to both circulating Ly6C lo monocytes and Ly6C hi monocytes (Higashi et al., 2016) ( Figure 6B). Following clodronate administration, a significantly higher number of red-fluorescent cells were detected in PirB MΦKO plaques ( Figure 6C). This indicates that PirB deficiency promotes pro-inflammatory Ly6C hi monocyte infiltration into femoral lesions (Swirski et al., 2007).
CD11b expression on leukocytes regulates leukocyte adhesion and rolling on the endothelium, an early step in atherogenesis (Woollard and Geissmann, 2010). To evaluate the role of PirB on monocyte adhesion and rolling on the endothelial luminal surface in vivo, CD11b + Ly6C + inflammatory monocytes were identified in the mesenteric circulation using intravital microscopy. There was a significant rise in CD11b + Ly6C + monocyte adhesion ( Figure 6D) and rolling ( Figure 6E) on the endothelial luminal surface of PirB MΦKO mice. Given that PirB MΦKO does not impact circulating total monocyte or Ly6C hi monocyte counts (Supplementary Table S2), this combined evidence supports that PirB knockout enhances recruitment of CD11b + Ly6C + monocytes into lesions.
To validate our in vitro findings, we finally evaluated intra-plaque macrophage proliferation and apoptosis as well as inflammatory marker expression within the macrophage-rich areas in the femoral plaques (Kelman, 1997;Ozen and Ittmann, 2005;Brizova et al., 2010). To dissect out the macrophage-rich areas, we used the SC-101447 monoclonal antibody against mouse macrophages (Foryst-Ludwig et al., 2010;Kuhn et al., 2013). These macrophage-rich areas were subjected to qPCR, which confirmed equivalent levels of Cd68, Ki67, and Pcna expression in PirB MΦKO lesions as compared with PirB flox lesions ( Figure 6F). Although PirB MΦKO produced no significant differences in macrophage apoptosis (Mac3 + TUNEL + ) counts, it did increase the ratio of macrophage-associated TUNEL + (Mac3 + TUNEL + ) cells-to-free TUNEL + cells ( Figure 6G), an indicator of efferocytosis (Yurdagul Jr et al., 2020). Accordingly, we also found enhanced staining of the efferocytosis marker Mertk in PirB MΦKO lesions as compared with PirB flox lesions ( Figure 6H). With respect to inflammatory markers, we observed enhanced Il6, Tnf, and Ccl2 levels within the macrophage-rich areas of PirB MΦKO lesions as compared with those of PirB flox lesions by qPCR ( Figure 6I). These results show that PirB knockout increases recruitment of monocytes to lesions as well as intra-lesional efferocytosis and inflammation but does not significantly affect intra-lesion macrophage proliferation or apoptosis.

DISCUSSION
PAD is a common disease affecting patient quality of life. Although the epidemiology is known, the pathogenesis mechanism is not well elucidated. Based on an integrated bioinformatics approach, here we discovered the inhibitory monocyte/macrophage receptor LILRB2 to be strongly associated with the PAD phenotype. Although a number of reports have indicated LILRB2 expression in human macrophages (Chen et al., 2018;Van Dalen et al., 2019), its exact role with respect to molecular mechanisms underlying peripheral atherosclerosis remain elusive. To understand the role of LILRB2 in PAD, we analyzed the role of its close murine homologue PirB by generating myeloid-specific PirB knockout (PirB MΦKO ). In our myeloid-specific PirB-null Apoe −/− murine model of PAD, PirB deficiency increased macrophage recruitment to atherosclerotic plaques and increased atherosclerotic burden. Accordingly, we identified that PirB deficiency plays a key role in promoting the inflammatory response and inhibiting cholesterol efflux in macrophages in vitro. The present work suggests that macrophage PirB deficiency is pro-inflammatory and leads to an increase in peripheral atherosclerotic burden in mice.
Upon infiltration into a target tissue and exposure to stimuli, macrophages become activated either along the classic M1 pathway (producing a pro-inflammatory phenotype) or along the alternate M2 pathway (producing a less inflammatory, phagocytic phenotype) (Vergadi et al., 2017). Previous work in hypercholesterolemic mouse models suggest a phenotypic shift favoring M2 over M1 activation reduces atherosclerotic burden (Babaev et al., 2014;Harmon et al., 2014;Wolfs et al., 2014). Recent findings also report the existence of a continuum of macrophage polarization with M1 and M2 being extreme ends of the continuum (Mosser and Edwards, 2008;Martinez and Gordon, 2014;Xue et al., 2014). There has been mixed findings regarding the effects of PirB knockout on macrophage polarization, with some favoring M1 over M2 polarization (Ma et al., 2011;Karo-Atar et al., 2013) and some favoring the opposite (Kondo et al., 2013). Here, we found that macrophage PirB knockout upregulated pro-inflammatory M1 marker levels in a Shp1-dependent manner along with downregulating M2 marker levels. Consistently, PirB MΦKO robustly enhanced macrophage production of the proinflammatory cytokines Il-1α, Tnfα, and Il-6. Moreover, oxLDL and its related oxidized lipid moieties promote M1 polarization (Wiesner et al., 2010;Van Tits et al., 2011;Qin et al., 2014) and induce a unique Mox activation status characterized by upregulation of Hmox1 and Txnrd1 (Kadl et al., 2010). Here, we found that PirB MΦKO promoted Mox activation upon oxLDL exposure. Taken together, PirB deficiency skews macrophages toward a proatherogenic state characterized by enhanced pro-inflammatory M1 polarization as well as Mox activation.
We employed the femoral cuff-induced model of PAD on PirB MΦKO and PirB flox mice to induce peripheral atherosclerosis in vivo. Post-mortem histological evaluations of femoral artery sections by Oil Red O staining indicated increased femoral atherosclerotic burden in PirB MΦKO mice, indicating that monocyte/macrophage PirB deficiency enhances peripheral atherogenesis in vivo. PirB MΦKO mice also displayed increased levels of plaque macrophage burden. We determined that enhanced macrophage adhesion/rolling on the endothelial luminal surface (as opposed to changes in intra-lesion macrophage proliferation or apoptosis) was primarily responsible for this phenomenon. Consistently, PirB deficiency has been previously shown to produce excessive macrophage adhesion from enhanced integrin signaling (Pereira et al., 2004).
Medial SMCs are responsible for producing the majority of ACTA + SMCs within plaque-stabilizing fibrous caps as well as secreting plaque-stabilizing extracellular matrix components, such as collagen (Misra and Fisher, 2021). Medial SMCs retain their proliferative capacity within plaques, and cues within the plaque microenvironment can regulate plaque SMC proliferation and clonal expansion (Liu and Gomez, 2019). Notably, myeloidderived plaque macrophages have been shown to inhibit plaque SMC polyclonality (Liu and Gomez, 2019). Consistent with this model, PirB MΦKO mice also displayed increased levels of macrophage burden, medial elastin breaks, and intra-plaque hemorrhage coupled with decreased SMC content, enhanced thinning of SMC-positive fibrous caps, and reduced collagen content, suggesting that myeloid PirB deficiency enhances plaque vulnerability in vivo (Chen et al., 2016). However, we did not analyze the mechanism(s) by which monocyte/ macrophage PirB deficiency affects plaque SMC phenotype. Future studies should employ macrophage-SMC co-culture studies with PirB-null macrophages to thoroughly investigate this question.

CONCLUSION
The importance of LILRB2 in human atherosclerosis is increasingly evident with reports linking the LILRB2 ligand Frontiers in Cell and Developmental Biology | www.frontiersin.org March 2022 | Volume 9 | Article 783954 ANGPTL2 to cardiovascular disease (Gellen et al., 2016;Tian et al., 2016;Tian et al., 2018). In this study, we identify the murine homologue of LILRB2 --PirB --as a key regulator in PAD. We show that macrophage PirB reduces peripheral atherosclerotic burden, stabilizes peripheral plaque composition, and suppresses macrophage accumulation in peripheral lesions. Our findings also demonstrate that macrophage PirB inhibits pro-inflammatory activation, inhibits efferocytosis, and promotes lipid efflux, characteristics critical to suppressing peripheral atherogenesis. Further pre-clinical studies will be needed to ascertain the potential of PirB/LILRB2-based therapeutic strategies against PAD.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material; further inquiries can be directed to the corresponding author.

ETHICS STATEMENT
All experiments were carried out with protocols approved in advance by the Ethics Committee of First People's Hospital of Yunnan Province (Kunming, China).