Skip main navigation

ApoA-I Nanotherapy Rescues Postischemic Vascular Maladaptation by Modulating Endothelial Cell and Macrophage Phenotypes in Type 2 Diabetic Mice

Originally publishedhttps://doi.org/10.1161/ATVBAHA.122.318196Arteriosclerosis, Thrombosis, and Vascular Biology. 2023;43:e46–e61

Abstract

Background:

Diabetes is a major risk factor for peripheral arterial disease. Clinical and preclinical studies suggest an impaired collateral remodeling and angiogenesis in response to atherosclerotic arterial occlusion in diabetic conditions, although the underlying mechanisms are poorly understood.

Objective:

To clarify the cellular and molecular mechanisms underlying impaired postischemic adaptive vascular responses and to evaluate rHDL (reconstituted HDL)-ApoA-I nanotherapy to rescue the defect in type 2 diabetic mouse model of hindlimb ischemia.

Methods and Results:

Hindlimb ischemia was induced by unilateral femoral artery ligation. Collateral and capillary parameters together with blood flow recovery were analyzed from normoxic adductor and ischemic gastrocnemius muscles, respectively, at day 3 and 7 post-ligation. In response to femoral artery ligation, collateral lumen area was significantly reduced in normoxic adductor muscles. Distally, ischemic gastrocnemius muscles displayed impaired perfusion recovery and angiogenesis paralleled with persistent inflammation. Muscle-specific mRNA sequencing revealed differential expression of genes critical for smooth muscle proliferation and sprouting angiogenesis in normoxic adductor and ischemic gastrocnemius, respectively, at day 7 post-ligation. Genes typical for macrophage (Mϕ) subsets were differentially expressed across both muscle types. Cell-specific gene expression, flow cytometry, and immunohistochemistry revealed persistent IFN-I response gene upregulation in arterial endothelial cells, ECs and Mϕs from T2DM mice associated with impaired collateral remodeling, angiogenesis and perfusion recovery. Furthermore, rHDL nanotherapy rescued impaired collateral remodeling and angiogenesis through dampening EC and Mϕ inflammation in T2DM mice.

Conclusions:

Our results suggest that an impaired collateral remodeling and sprouting angiogenesis in T2DM mice is associated with persistent IFN-I response in ECs and Mϕs. Dampening persistent inflammation and skewing ECs and Mϕ phenotype toward less inflammatory ones using rHDL nanotherapy may serve as a potential therapeutic target for T2DM peripheral arterial disease.

Highlights

  • Type 2 diabetes significantly impairs postischemic perfusion recovery, vascular growth, and regeneration.

  • Persistent endothelial and macrophage IFN-I response is critically involved in impaired collateral remodeling and sprouting angiogenesis.

  • rHDL (reconstituted HDL; ApoA-I-PS) nanotherapy could dampen IFN-I response and improve collateral remodeling, sprouting angiogenesis, and perfusion recovery in type 2 diabetes.

Type 2 diabetes (T2DM) is associated with an increased incidence of morbidity and mortality from atherosclerotic disease including coronary heart disease and peripheral arterial disease (PAD).1,2 Clinical studies suggest a strong association between diabetes and the development of PAD. In diabetic patients, for every 1% increase in Hemoglobin A1c, there is a corresponding 26% increase in the risk of PAD.3 PAD in diabetic patients is more aggressive than in non-diabetics. In addition, the need for major amputations is 10 to 20 times higher in diabetic patients than in non-diabetics,4,5 and diabetes is associated with half of the lower limb amputations worldwide.6 Diabetic patients with PAD have 3- to 4-fold increased mortality compared with healthy individuals.7

Collateral remodeling and angiogenesis act in concert to reestablish blood flow to the distal tissues during atherosclerotic arterial occlusion.8 Collateral growth involves shear stress induced extensive remodeling of preexisting low flow interarteriolar connections that run parallel to the occluded main artery. An increase in shear stress leads to multiple events including the expression of various adhesion molecules and chemoattractants leading to infiltration of monocytes, which in turn secrete several cytokines and growth factors. These events are followed by EC and smooth muscle cell proliferation, migration and synthesis of ECM (extracellular matrix) ultimately resulting in outward remodeling of preexisting collaterals into functional conductance arteries. Thus, collateral remodeling involves complex interactions between mechanical, cellular, and immune components.9–12 On the other hand, angiogenesis involves capillary sprouting from preexisting vessels through EC proliferation and migration, triggered mainly by ischemia. Impairment in collateral remodeling and angiogenesis can thus have severe consequences on the skeletal muscle function after arterial occlusion.

T2DM is a complex disease with hyperglycemia being a key causal factor in the development of vascular complications by causing divergent cellular dysfunction.13 Clinical and preclinical studies have consistently demonstrated an impaired collateral remodeling and angiogenesis in response to arterial occlusion in presence of T1DM and T2DM.2,14–16 Furthermore, the outcomes of therapeutic revascularization procedures using gene and cell-based approaches are poor in patients with PAD as seen in phase II and III clinical trials and many patients progress toward major amputation.17 Since collateral remodeling and angiogenesis require complex interplay between numerous cell types, understanding molecular mechanisms involved in these processes and their dysregulation in presence of T2DM is critical to design novel therapies.18,19

HDL (high-density lipoprotein) has been previously shown to have vasculoprotective and anti-inflammatory effect and shown to reduce atherosclerotic burden in clinical and preclinical studies.20 Individuals with diabetes are shown to have lower levels of HDL cholesterol and ApoA-I and a clear evidence of reduced HDL function including reduced cholesterol efflux capacity and anti-inflammatory function.21 Thus, increasing functional HDL represents a potential approach to improve vascular function and decrease Mϕ inflammation in diabetic conditions. Recently, ApoA-I, a major protein component of HDL was shown to promote atherosclerosis regression in diabetic mice by reducing monocytosis and monocyte recruitment to plaques, decreased plaque content of atherogenic neutrophil extracellular traps, enrichment in M2-like Mϕs, and improved regression.22 Present study addressed mechanisms underlying impaired collateral remodeling and angiogenic responses after femoral artery ligation (FAL) and evaluated the therapeutic effect of rHDL (reconstituted HDL) nanoparticles to rescue these defects in a well-established mouse model of T2DM.23

Methods

Data and supporting materials have been provided with the published article. RNA Sequencing (RNA-seq) data has been deposited in NCBI Gene Expression Omnibus (See Foot Note).

Mouse Model

Mice with β-cell specific over-expression of insulin-like growth factor-2 in atherosclerotic background (IGF-II [insulin-like growth factor II]/ LDLR−/−ApoB100/100) with T2DM features23 were used in the study with C57BL/6J mice (RRID: IMSR_JAX:000664) fed with a regular chow-diet (R36, Lactamin) serving as controls. Male mice aged between 16 to 20 weeks were used in the study. All animal experiments were approved by the Experimental Animal Committee, University of Eastern Finland.

Mouse Hindlimb Ischemia

Mice were anesthetized with 1.5% to 2% isoflurane/air mixture during the surgical procedure. Hair was removed from the operation site and cleaned with sterile water. Single incision was made just above medial thigh and superficial femoral artery was separated from femoral vein and nerve. HLI was induced by unilateral ligation of femoral artery distal to the origin of the profunda femoral artery. The skin was closed with interrupted silk sutures as described previously.24 Mice were given analgesic (Carprofen, 5 mg/kg, Rimadyl, Pfizer, Inc, NY) before and after the operation (for 3 days). This methodology allows to study both collateral remodeling (in normoxic adductor muscles) and angiogenesis (in ischemic gastrocnemius and tibialis muscles).16

Contrast Enhanced Ultrasound Imaging of Mouse Hindlimb Blood Flow

Hindlimb perfusion was evaluated at pre-operation, post-operation, day 2, 4, and 6 after FAL. Briefly, mice were anesthetized with 1.5% to 2% isoflurane/air mixture and hindlimb blood flow was imaged using VEVO ultrasound imaging system 2100 (RRID:SCR_015816) following tail vein injection of gas-filled micro-bubbles (Vevo MicroMarker Non-Targeted Contrast Agent, Visualsonics, Canada). Imaging acquisition was performed with MS-250 Scan head using following parameters: Transmit frequency-18MHZ, Power-4%, Gate—4, Beam Width-Wide, Acquisition contrast gain-35dB, 2D gain-18dB, Frame rate-15, Depth-20.00 mm, Width-23.04 mm, Sensitivity-1, Line Density-High, Persistence—Off, ECG/Resp Gate—Off/Off, Extended buffer—On. Quantitation of blood flow was reported as a ratio between ischemic and contralateral nonischemic muscles.

rHDL-ApoA-I Nanoparticle Preparation and Quality Control

1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-L-serine (sodium salt) (POPS) was obtained from Avanti polar lipids (Catalog no. 840034C, Alabaster, AL). ApoA-I isolated from human plasma was obtained from Athens Research & Technology (Catalog no. 16-16-120101, GA). POPS Liposomes loaded with ApoA-I were prepared by hydration of thin films (POPS: ApoA-I molar ratio of 100:1) with 1 mg/ml ApoA-I and Tris-buffered saline (TBS) solution, followed by stirring and sonication. To release unloaded ApoA-I, sonicated ApoA-I loaded POPS liposomes were subsequently dialyzed in a 1 ml, ready-to-use dialysis tube (MWCO at 50000 Da, Spectrum Labs) in TBS with 3 buffer changes over 2 days at +4C. To ensure quality control, total ApoA-I loading amount was analyzed by BCA assay. For morphometric quality control, liposome size and distribution were analyzed with Nanosight NS300 (RRID:SCR_020310). The average hydrodynamic particle size was determined from 3 independent runs. SD (PDI) was obtained to evaluate the size distribution.

Ischemic Muscle Mϕ and EC Isolations

Ischemic muscle Mϕs and ECs were isolated using a combination of magnetic assisted cell sorting and flow Cytometry as described previously.16

Mϕ and EC RNA-seq Library Preparation and Sequencing

Total RNA was isolated from fluorescence-activated cell sorting (FACS) sorted ECs from adductor muscles, EC and Mϕs from ischemic muscles using Arcturus PicoPure RNA isolation Kit (KIT0204, Applied biosystems) as described previously.24 Lexogen QuantSeq 3’mRNA-Seq Library Prep Kit-FWD (catalog no. 15, Lexogen, Vienna, Austria) was used to prepare 3′ libraries. For the 3’-LEXO libraries, indices from the first 2 columns of the i7 Index Plate for QuantSeq/SENSE for Illumina adapters 7001–7096 (cat #044, Lexogen, Vienna, Austria) were used, and 11 cycles of library amplification were performed. Libraries were eluted in 22 μL of the kit’s Elution Buffer. The double stranded DNA concentrations were quantified using a 5200 Fragment Analyzer (Agilent Technologies, RRID:SCR_013575), which gave similar concentrations for each sample that ranged from 1.7 to 4.3 ng/μL. The molar concentration of cDNA molecules in the individual 3’-LEXO libraries was calculated from the double stranded DNA concentration and the region average size. Aliquots containing an equal number of nmoles of cDNA molecules from each library were pooled to give a pooled library with a concentration of 10 nM cDNA molecules. After library preparation, libraries were pooled and sequenced using single-end sequencing with 75 bp reads on an Illumina HiSeq4000 instrument (Illumina HiSeq 3000/HiSeq 4000 System, RRID:SCR_016386, San Diego, CA).

RNA-Seq Read Processing and Differential Expression Analysis

RNA-Seq reads were first trimmed to remove low-quality bases, poly(A) and Illumina adapter using cutadapt (RRID:SCR_011841, version 2.825). Subsequently, the reads were processed using the nf-core RNA-Seq pipeline (version 1.4.2;26) with the Mus musculus genome (mm10) assembly and the STAR aligner (RRID:SCR_004463) for read mapping and the Ensembl GRCm38 release 93 gene annotations for counting reads in transcripts. The following gene biotypes were retained in the gene expression matrix: protein coding, lincRNA and antisense. Subsequent analyses steps were performed separately for ECs and Mϕs. Genes with very low expression levels were filtered out with the filterByExpr function of the edgeR Bioconductor R package (RRID:SCR_012802, version 3.24.327) using minimum count 5 and minimum total count 15. Differential expression analysis was carried out using DESeq2 Bioconductor R package (RRID:SCR_000154, version 1.22.228) with the independent filtering option disabled. The False Discovery Rate (FDR) procedure was used to correct P for multiple testing.

Cell State Deconvolution of Bulk RNA-Seq Profiles Using Single-Cell RNA-Seq as Reference

Previously published29 scRNA-Seq datasets of mouse myocardial infarction generated using the 10x Genomics platform were obtained from the EMBL-EBI ArrayExpress database (https://www.ebi.ac.uk/biostudies/arrayexpress/studies/E-MTAB-7376). Cells from the different experimental conditions (3 or 7 days following myocardial infarction injury or sham treatment) were processed together. Samples were processed with the recommended scRNA-Seq workflow of Seurat version 3, followed by the standard multi-sample integration workflow of the same package.30 After unsupervised clustering (using the standard Seurat v3 clustering workflow with default parameters), any clusters representing endothelial, or Mϕ cells were identified (using the original authors’ markers) and selected for further analysis. One cell type at a time, the cells were reprocessed with Seurat 3 to obtain new Uniform Manifold Approximation and Projection (UMAP) projections to serve as the references of transcriptional state space for the Cell Population Mapping method (version 0.1.5).31 Bulk RNA-Seq expression profiles were natural log-transformed, averaged by condition, and Cell Population Mapping was run with default parameters to assess relative changes in the abundance of transcriptional states (or cell subtypes) between the conditions profiled by bulk RNA-Seq.

Phenotypic Analysis of Muscle Mϕs

Flow Cytometry was performed as described previously.16 Adductor and Ischemic gastrocnemius muscles were minced and enzymatically dissociated using a cocktail containing 450 U/mL Collagenase I, 125 U/mL Collagenase XI, 60 U/mL DNAseI, and 60 U/mL hyaluronidase (Sigma Aldrich) for 1 hour at 37 °C. The cells were then counted and stained for Mϕ antigens after initially blocking with rat anti-mouse CD16/32 mAb (BD Biosciences Catalog no. 553141, RRID: AB_394656) for 10 minutes at 4˚C. To identify M1 and M2-like Mϕs, we performed staining for Cd45, F4/80, CD86, and MRC1. Fluorescence minus 1 controls were used to analyze the specificity of the stainings. FACS was performed on FACS AriaIII (BD Biosciences), and data were analyzed with FCS express 6 (RRID:SCR_016431).

Plasma Analysis

Blood glucose (Catalog no. 81692, Crystal chem., Inc, IL) and LDL-cholesterol (Cat# 79980, Crystal chem Inc, IL) was analyzed according to manufacturer’s instructions. Plasma HDL-C levels were analyzed using an ELISA kit according to manufacturer’s instructions (Cusabio, Cat#CSB-E12874m). Plasma cytokines were analyzed using BD cytometric bead array mouse inflammation kit (BD Biosciences Cat# 552364, RRID: AB_2868960) using FACS Aria III (BD Biosciences) and calculated using FCAP Array v2.0.2 (Soft Flow Hungary, Ltd).

Statistical Analysis

Results were expressed as mean±SEM. Comparisons among 3 or more groups were tested using 1-way ANOVA followed by Bonferroni posthoc tests. If data did not pass normality test, Mann-Whitney U test was used for 2 group comparisons. Normalization and equality of all variables was evaluated with Kolmogorov-Simonov and F-tests, respectively. For data passed with both tests, differences between 2 groups were determined using Student t test. Comparison of the time course of hindlimb perfusion was performed using a 2-way repeated-measures ANOVA with Sidak multiple comparisons test. All statistical analyses were performed using GraphPad Prism 9 (RRID:SCR_002798). P<0.05 was considered statistically significant.

Detailed description of all materials and methods is available in the Supplemental Material.

Results

RNA Sequencing Reveals Insights Into Impaired Collateral Remodeling and Modulation by rHDL in T2DM Mice

In order to detect transcriptional changes associated with collateral remodeling, we performed bulk RNA seq on normoxic adductor muscles collected at day 7 post-ligation and detected mRNA expression of 29682 Refseq genes from each sample. The expression of each gene was quantified from sequencing read counts. The scatter plot of differentially expressed genes in T2DM adductor muscles compared with controls is shown in Figure S1A. Among the 29 682 expressed genes, we found 614 differentially expressed genes (P<0.05, Reads per kilobase per million mapped reads [RPKM] cutoff of >1 and fold change of >1.5) in T2DM mice compared to controls. Among them, 242 genes were upregulated, and 372 genes were downregulated as shown in a volcano plot in Figure S1B. Ingenuity pathway analysis of differentially expressed genes suggested their enrichment to biological pathways relevant to monocyte and Mϕ subsets (Figure S1C). Interestingly, Ingenuity network analysis suggested enrichment to smooth muscle cell proliferation (Figure S1D).

To understand the transcriptional changes in AECs during outward collateral remodeling, we FACS sorted AECs at day 3 and 7 post-FAL and performed RNA-seq. Differential gene expression analysis showed an upregulation of Apln, Plaur, Sirt7, Pdgfra, Thbs1, and Hyal2 at day 3, while upregulation of Plxnd1, Dll4, Cxcr4, Spry4, Aplnr, and Kdr at day 7 in control mice. While, AECs from T2DM mice displayed a persistent activation of pro-inflammatory/type I interferon (IFN-I) response genes, including Irf1, Irf7, Irf8, Ifit1, 2, 3 and Stat1, 2 and 3, and downregulation of anti-oxidant genes (Day 3—Gstk, Gstm1, Gstm2, Gstt1, Gclm) and vasoprotective genes (Day 7—Apln, Robo2, Aplnr, Vegfb, Vegfc, Sirt6) compared with controls (Figure 1B, PBS group). Double immunostaining for pericyte/smooth muscle marker (α-SMA) and EC marker (CD31) showed a reduced collateral arteriole lumen area in normoxic adductor muscles of T2DM mice at day 7 post-ligation compared with controls (Figure 1C). Additionally, remodeling collaterals from T2DM mice showed a significantly decreased smooth muscle cell proliferation (Ki-67+) compared with controls at day 7 post-ligation (Figure 1D). To test whether ApoA-I treatment can rescue genes critical for collateral remodeling, reconstituted ApoA-I nanoparticles with phosphatidylserine core (rHDL) were administered intravenously starting 2 days post-HLI with doses every 2 days until sacrification (Figure 1A). rHDL treatment significantly reduced circulating levels of inflammatory cytokines TNF-α and IFN-γ at day 7 post-FAL while not altering glucose and lipid levels (Table). Differential gene expression analysis of AECs from rHDL treated mice showed an upregulation of Ccr2, Cx3cr1, Trem2, CD74, and Cxcl10 (at day3) and Pdgfra, CD74, Ccl4, Gbp3, 4, 5, 7 (at day 7) compared with PBS groups in control mice. Interestingly, rHDL downregulated pro-inflammatory genes including, Tnf, IL-1b, Myd88, S100a9 (at day 3) and Osm, Ace, Nfkbia, IL17ra, Serpine1 (at day 7), while upregulating Gstk1, Gsto1, Gstm2 (at day 3) and Apln, Sirt6, Vegfb, Nrp1, Efna1, and Plxnc1 (at day 7) compared with PBS in T2DM mice (Figure 1B, rHDL group). Functionally, rHDL treatment rescued impaired collateral remodeling in T2DM mice as shown by an increased arteriole lumen area and smooth muscle cell proliferation activity in collaterals at day 7 FAL (Figure 1C and 1D).

Table. Metabolic Parameters and Plasma Cytokine Levels

TreatmentControlT2DMP Value
Glucose, mmol/LPBS9.76±2.0014.51±4.290.04
rHDL9.45±1.3513.08±3.090.06
LDL cholesterol, mmol/LPBS2.55±0.3740.88±8.020.01
rHDL2.24±0.3930.14±3.350.001
IL-6, pg/mlPBSND6.05±167
rHDL6.68±1.268.76±0.14
IL-10, pg/mlPBS25.48±4.9234.26±5.830.31
rHDL31.01±2.2733.39±6.750.72
TNF-α, pg/mlPBSND12.65±3.65
rHDLND2.68±0.31
IFN-γ, pg/mlPBSND42.37±10.80
rHDLND7.22±3.18

IFN indicates interferon; LDL, low-density lipoprotein; ND, not detectable; rHDL, reconstituted HDL; and T2DM, type 2 diabetes.

Figure 1.

Figure 1. rHDL (reconstituted HDL) rescues impaired collateral remodeling in normoxic adductor muscles of type 2 diabetes (T2DM) mice. A, Study outline showing treatment regimen and sampling scheme. B, Volcano plots showing differential gene expression patterns among genotypes and treatment groups. C, CD31 and α-SMA double staining of remodeling collaterals and corresponding quantitation of lumen area. Scale bar; 20 µm. D, Quantitation of collateral arteriole parameters at day 3 and 7 post-ligation. E, Double immunofluorescence staining for α-SMA and Ki67 demonstrating proliferation of cells in lumen, smooth muscle layer and peri-collateral space in remodeling collaterals. Scale bar; 20 µm. F, Quantitation of Ki67+ SMCs in adductor muscles at day 3 and 7 post-ligation (n=4–6 per group per time point). Control; white bars, T2DM; gray bars. One-way ANOVA followed by Bonferroni posthoc test (D, F).

rHDL Favorably Modifies Altered Pericollateral Mϕs in Normoxic Adductor Muscles of T2DM Mice

Bulk RNA-Seq analysis of normoxic adductor muscles showed a significant upregulation of M1-Mϕ specific genes involved in inflammation including Ccl7, Ccl8, Nfkbib, Nfkb2, Hpgd, whereas downregulation of M2-Mϕ specific genes including MafB, Baiap2, CCN3, Wnt5a, Adamts5, and Prkcd in T2DM mice compared with controls (Table S1). To follow-up on the bulk RNA-seq based differential expression of genes typical for Mϕs, we performed flow cytometry analysis using M1-Mϕ (CD86) and M2-Mϕ (CD206) specific markers at 3 and 7 days post-FAL (Figure 2A and 2B). Quantification revealed an increased CD86+ M1-Mϕs and a reduction in CD206+ M2-Mϕs in normoxic adductor muscle of T2DM mice at day 7 post-FAL (Figure 2C). Immunofluorescence double stainings revealed accumulation of Mϕs preferentially around remodeling collaterals (Figure 2D). Furthermore, double staining with Mϕ marker (F4/80) together with M2-Mϕ marker CD206 (MRC1) showed a decreased number of CD206+ M2-Mϕs in T2DM mice compared to controls at day 7 post-FAL (Figure 2E). Collectively, these results suggest an increased M1-Mϕs and a reduced number of M2-Mϕs in peri-collateral space in normoxic adductor muscles of T2DM mice. rHDL treatment following FAL increased CD206+ M2-Mϕs and decreased CD86+ M1-Mϕs as evidenced by FACS analysis. Immunofluorescence double stainings further supported this observation by revealing an increase in CD206+ M2-Mϕs around remodeling collaterals in rHDL treated groups compared with PBS. Collectively, these findings suggest that rHDL could rescue defective collateral remodeling in T2DM mice through selective modulation of AEC and peri-collateral Mϕ phenotypes.

Figure 2.

Figure 2. rHDL (reconstituted HDL) favorably modifies an altered peri-collateral Mϕ phenotypes in normoxic adductor muscles of type 2 diabetes (T2DM) mice. A–C, Flow-cytometry analysis and quantification of CD45+, F4/80+ cells, M2, and M1-Mϕs in adductor muscles at day 3 and 7 post-ligation. Control; white bars, T2DM; gray bars. D, Double immunofluorescence staining with Mϕ marker (F4/80) and smooth muscle cell maker (α-SMA) showing the presence of Mϕs in peri-collateral space at day 3 and 7 post-ligation. Scale bar; 20 µm. E, Double immunostainings with Mϕ marker F4/80 and M2-Mϕ marker CD206 (MRC1) showing the presence of CD206+ M2-Mϕs (arrow heads) in peri-collateral space in adductor muscles at day 3 and 7 post-ligation (n=4–6 per group per time point). Scale bar; 20 µm. One-way ANOVA followed by Bonferroni posthoc test (C).

rHDL Rescues an Impaired Perfusion Recovery and Angiogenesis in Ischemic Muscles of T2DM Mice

Bulk RNA seq analysis of ischemic gastrocnemius muscles detected mRNA expression of 29 682 Refseq genes. The scatter plot of differentially expressed genes in ischemic muscles of T2DM mice compared with controls is shown in Figure S2A. Among the 29 682 genes expressed in ischemic skeletal muscles, we found 471 differentially expressed genes (P<0.05, RPKM cut off of >1 and fold change of >1.5) in ischemic muscles of T2DM mice compared with controls. Among them, 241 genes were upregulated, and 230 genes were downregulated as shown in the volcano plot in Figure S2B. Ingenuity pathway analysis of the differentially expressed genes suggested a significant enrichment to Mϕ subsets (Figure S2C). Interestingly, ingenuity network analysis suggested enrichment to “neovascularization of organ and cell death of ECs” (Figure S2D and S2E).

Contrast enhanced ultrasound analysis showed a significantly decreased perfusion recovery in ischemic muscles of T2DM mice compared with controls at day 4 and 6 post-ligation (Figure 3A). Angiogenesis as revealed by EC marker (CD31) staining (Figure 3C) showed a significantly decreased capillary to myocyte ratio at day 7 post-FAL. Similarly, capillary lumen area was significantly reduced in ischemic muscles of T2DM mice compared with controls at day 3 and 7 post-FAL (Figure 3B through 3D). Since sprouting angiogenesis crucially depends on proliferation of EC, we performed double staining with EC marker (CD31) together with cell proliferation marker (Ki-67) in ischemic muscles. Quantification of Ki67+ ECs suggested a significant reduction in EC proliferation in ischemic muscles of T2DM compared with controls at day 7 post-FAL (Figure 3E and 3F). Collectively, these results suggest that ischemic muscles of T2DM mice display impaired perfusion recovery in parallel with impaired angiogenesis compared with controls. To test whether ApoA-I treatment can improve perfusion recovery, rescue revascularization, and muscle regeneration, rHDL nanoparticles were administered intravenously starting 2 days post-FAL with doses repeated every 2 days until sacrification. ELISA for mouse HDL-C (Cusabio, Cat # CSB-E12874m) on plasma samples collected at day 3 and 7 after ischemia and rHDL (human ApoA-I-POPS) treatment showed no major changes across groups or treatments (data not shown). Perfusion analysis revealed a marked improvement followed by improved revascularization and muscle regeneration responses across all disease backgrounds (Figure 3A). This was paralled with an increased capillaries per myocyte, mean capillary area and EC proliferation in rHDL-treated groups (Figure 3B through 3F). Overall, these findings suggest that rHDL could enhance postischemic perfusion, revacularization, and muscle repair responses under T2DM conditions.

Figure 3.

Figure 3. rHDL (reconstituted HDL) rescues an impaired post-ischemic tissue perfusion recovery and angiogenesis in type 2 diabetes (T2DM) mice. A, Contrast enhanced ultrasound based quantification of perfusion expressed as a ratio of right (ischemic) to left (non-ischemic) limbs across treatment groups (n=5–7 per group per time point). B and C, Quantitation of capillary per myocyte and capillary inner area. D, Muscle regeneration (HE staining) and angiogenesis (EC specific CD31 staining) responses in ischemic muscles at 3- and 7-days post-ligation. Scale bar; 50 µm. E, Double immunostainings with CD31 and Ki-67 showing proliferating capillary ECs in ischemic muscles of at day 7 post-ligation. F, Quantification of Ki67+ capillaries at day 7 post-ligation (n=4–6 per group per time point). Control; white bars, T2DM; gray bars. Scale bar; 20 µm. Two-way repeated measures ANOVA with Sidak multiple comparison test (A), 1-way ANOVA followed by Bonferroni posthoc test (B, C, and F).

rHDL Dampens Persistent Postischemic Endothelial Inflammation in T2DM

Bulk RNA-seq analysis of ischemic muscles suggested a downregulation of endothelial tip cell-enriched genes including Nrarp, Esm1, Igfbp3, Vegf-c, Apln, and Angpt2 in T2DM compared with control mice at day 7 post-ligation (Table S2, Figure S2F). FACS sorting of ECs and Mϕs from ischemic muscles, followed by qRT-PCR analysis further suggested a significant reduction in tip cell enriched genes specifically in ECs of T2DM ischemic muscles compared with controls (Figure S2G).

To further understand EC heterogeneity during ischemia, we FACS purified ECs from ischemic muscles at day 3 and 7 and performed RNA-seq analysis. Purified EC-specific RNA-Seq followed by unsupervised cell deconvolution analysis revealed 7 major EC clusters with distinct gene marker expressions (Figure 4A and 4B; Table S3). Distinct cell types dominated the ischemic muscles in a time and genotype dependent manner. Early phase (day 3) following FAL was dominated by proliferating EC types, whereas later regenerating phase (day 7) was dominated by mixture of cycling and inflammatory EC types in control mice. On the contrary, both phases were dominated by inflammatory (IFN-I activated) EC types in ischemic muscles of T2DM compared with controls at both time points. A persistent endothelial inflammatory signature was noticeable with enhanced IFN-I response gene expression in T2DM (Cluster 5—ifi203, ifi204, ifi35, ifi44, ifi47, ifih1, ifit1, ifit2, ifit3, ifit3b, ifitm3, igtp, iigp1, irf7, irf9, dhx58, irgm1, irgm2, isg15, isg20, lgals3bp, oas2, stat1, stat2) compared with control mice. Contrarily, Angiogenic/Immature EC (Cluster 2—Adm, Apln) and proliferating EC (Cluster 7—Apln, Pcna) were reduced in T2DM (Figure 4C). Further validation by double immunostainings with EC marker (CD31) together with IFN-γ revealed an increased IFN-γ+ EC cells. While, EC tip cell marker (apelin) together with EC marker (CD31) showed a decreased number of double positive cells in ischemic muscles of T2DM mice compared with control mice at day 7 post-FAL consistent with gene expression analysis (Figure 4D and 4E). (rHDL)-ApoA-I nanoparticle treatment skewed the EC types to proliferating phenotype in both controls and T2DM mice compared with saline treatment. Additionally, (rHDL)-ApoA-I nanoparticle treatment completely dampened IFN-I+ EC types in T2DM mice (Figure 4C). Immunofluorescence double stainings further revealed the reduction of IFN-I+ EC types and skewing toward Apelin+ EC types in ischemic muscles of controls and T2DM. Gene ontology (GO) biological process analysis of rHDL upregulated genes (fold change >1.5, P<0.05) at day 3 showed enrichment to aerobic electron transport chain and mitochondrial ATP synthesis, while downregulated genes showed enrichment to protein deubiquitination and regulation of programmed necrotic cell death. rHDL upregulated genes at day 7 showed enrichment to vascular transport and regulation of cardiac conduction. While, downregulated genes showed enrichment to regulation of hydrolase activity and cellular response to glucocorticoid stimulus (Figure S8). Collectively, these findings suggest that rHDL nanoparticle treatment could rescue impaired angiogenesis in T2DM mice through selective modulation of EC phenotypes under ischemic conditions.

Figure 4.

Figure 4. Single-cell deconvolution reveals postischemic tissue EC heterogeneity, temporal shifts and modulation by rHDL (reconstituted HDL) treatment. A, Purified EC-specific Bulk RNA-Seq followed by Unsupervised cell deconvolution analysis showing major clusters. Cell Cluster Key—EC2: angiogenic/immature, EC3: stressed/activated, EC5: interferon activated, EC7: proliferating. B, Top 10 gene markers discriminating major EC clusters. C, Uniform Manifold Approximation and Projection (UMAP) projections showing time, genotype and treatment dependent shifts in EC populations. D, Double immunofluorescence staining of EC marker (CD31) and inflammatory cytokine (IFN-g+) (arrows) in ischemic muscles at day 7 post-ligation. E, Double immunofluorescence staining of EC marker (CD31) and tip cell marker (Apelin; arrows) in ischemic muscles at day 7 post-ligation (n=3–5 per group per time point). Scale bar; 20 µm.

Persistent Mϕ Inflammation With a Strong Postischemic IFN-I Responsive Gene Signature in T2DM

Bulk RNA-Seq of ischemic muscles, followed by qRT-PCR analysis showed an upregulation of genes typical of M1-Mϕs including Lcn2, Ccin, Serpina3f, Gapdh, Tagap, Fabp3, Il18bp, Mid1, Bst2, Ifi44, and Cd74. Conversely, genes typical of M2-Mϕs including Hmox1, Ccl2, Ccl7, Arg1, Il1rl1, Mmp9, Ccr1, Egr2, Dhcr24, Ch25h, Cyp51, Fos, Lyve1, Sqle, and Loxl2 were significantly downregulated in ischemic muscles of T2DM mice compared with control mice at day 7 post-ligation (Table S2, Figure S2H). To track Mϕ specificity of these differentially expressed genes, we FACS sorted Mϕs and ECs at day 7 post-ligation from ischemic muscles. qRT-PCR analysis of genes typical for M2-Mϕs Arg1, Mmp9, Ccl2, Hmox1 and Angpt2 showed significant downregulation in Mϕs of T2DM compared with control mice (Figure S2I).

In order to get a deeper understanding of Mϕ heterogeneity, we performed RNA-Seq on FACS purified Mϕs. Bulk RNA-Seq followed by cell deconvolution analysis revealed 12 major Mϕ clusters with distinct gene marker expressions (Figure 5A and 5B; Table S4). Distinct cell types dominated the ischemic muscles in a time- and genotype-dependent manner. Early phase (day 3) was dominated by anti-inflammatory phenotype, whereas later regenerating phase (day 7) was dominated by mixture of reparative and IFN-I+ Mϕs in control mice. On the contrary, both phases were dominated by inflammatory (IFN-I+) Mϕs in ischemic muscles of T2DM compared with controls (Figure 5C, PBS group). Thus, consistent with Bulk RNA-Seq, cell-specific RNA-Seq followed-by single cell deconvolution revealed a persistent Mϕs inflammatory signature with enhanced IFN-I response gene expression (Cluster 4—ifi203, ifi204, ifi27l2a, ifi35, ifi47, ifih1, ifit3b, ifitm3, igtp, irf1, irf7, il15, irgm1, dhx58, isg15, Stat1) in T2DM compared with controls, whereas reduced phagocytic/anti-inflammatory M2-Mϕ gene signature (Cluster 0/1—Arg1, Hmox1, spp1, adssl1, anxa3, clec4d, itgb2, lgals1, lgals3; Figure S7B and S7C). (rHDL)-ApoA-I nanoparticle treatment significantly dampened IFN-I+ Mϕ gene expression (Cluster 4—Ifi47, Ifi205, Ifi213, Ifi208, Ifi203, Ifi209, Irgm1 Ifi207, Ifih1, Irf1, Stat1), while skewing them toward anti-inflammatory Mϕ gene signature (Cluster 0—Arg1, Hmox1, spp1) compared with PBS groups in T2DM (Figure 5C rHDL group, Figure S7G and S7H). List of Top 10 GO biological processes enriched for genes (fold change >1.5, P<0.05) rescued by rHDL treatment at day 3 and day 7 further suggest a down modulation of pro-inflammatory/IFN-I response, while activating proangiogenic/anti-inflammatory pathways as shown in Figure 5D through 5G.

Figure 5.

Figure 5. Single-cell deconvolution reveals postischemic tissue Mϕ heterogeneity, temporal shifts and modulation by rHDL (reconstituted HDL) treatment. A, Purified Mϕ specific Bulk RNA-Seq followed by unsupervised cell deconvolution analysis showing major Mϕ clusters. Cell Cluster Key-MP2: chemokine/inflammatory, DC-like, MP3: monocyte-like, MP4: interferon activated, MP0, MP1: phagocytic/anti-inflammatory, MP8: resident-like, MP9: cycling. B, Top 10 gene markers discriminating major Mϕ clusters. C, Uniform manifold approximation and projection (UMAP) projections showing time, genotype and treatment dependent shifts in Mϕ populations. D, List of Top 10 gene ontology (GO) biological processes enriched for genes (FC >1.5, P<0.05) rescued by rHDL treatment in Mϕs at day 3 and day 7 (n=3–5 per group per time point). Enricher GO biological processes, Bonferroni-corrected for P<0.05.

rHDL Rescues Defective Angiogenesis by Favorably Modifying Mϕ Phenotypes in Ischemic Muscles

To follow-up on single cell deconvolution analysis, we performed FACS based Mϕ phenotypic analysis using M1-Mϕ (CD86) and M2-Mϕ (CD206) specific markers at 3 and 7 days after PBS and rHDL treatments. Flow cytometry analysis showed a significant increase in inflammatory CD86+ M1-Mϕ and a decrease in anti-inflammatory CD206+ M2-Mϕ populations in PBS treated T2DM mice (Figure 6A and 6C, PBS group). Furthermore, staining with M2-Mϕ marker (Arginase1) showed decreased number of Arg1+ M2-Mϕs, whereas staining with M1-Mϕ marker (IFN-γ) showed increased number of IFN-γ + M1-Mϕs in ischemic muscles of T2DM mice compared with control mice at day 7 post-ligation (Figure 6D and 6E, PBS group). Collectively, ischemic muscle and cell specific gene expression together with flow cytometry and immunohistochemistry suggested a persistent M1-Mϕ presence and a significant reduction in M2-Mϕs in ischemic muscles of T2DM mice compared with controls. Interestingly, rHDL treatment markedly decreased M1-Mϕs and increased M2-Ms in a time dependent manner in both groups. Importantly rHDL treatment could normalize Mϕ phenotypes mimicking control groups (Figure 6B and 6C). In addition, immunofluorescence double stainings using M1-Mϕ (IFN-γ) and M2-Mϕ (Arg1) specific markers in combination with Mϕ marker (F4/80) revealed patterns consistent with RNA-seq and FACS observations with the reduction in IFN-g+ Mϕs and skewing towards anti-inflammatory CD206+ M2-Mϕ phenotype in ischemic muscles T2DM by (rHDL)-ApoA-I nanoparticle treatment (Figure 6D and 6E). Collectively, these results suggest a persistent Mϕ inflammation paralleled with impaired tissue revascularization and perfusion recovery in ischemic muscles of T2DM mice. Importantly, rHDL treatment could rescue these defects through selective modulation of Mϕ phenotypes under ischemic conditions.

Figure 6.

Figure 6. rHDL (reconstituted HDL) rescue defective angiogenesis by favorably modifying ischemic muscle Mϕ phenotypes. A–C, FACS-based Mϕ phenotypic analysis using M1-Mϕ (CD86) and M2-Mϕ (CD206) specific markers and corresponding quantitation at 3 and 7 days after PBS and rHDL treatments. D and E, Immunofluorescence double stainings using M2-Mϕ (Arg1) and M1-Mϕ (IFN-γ) specific markers in combination with pan-Mϕ marker F4/80 showing Mϕ phenotypes in ischemic muscles in PBS and rHDL treatment groups (n=3–5 per group per time point). Scale bar; 20 µm. One-way ANOVA followed by Bonferroni posthoc test (C).

Discussion

In our previous study, we have shown an involvement of an altered DNA methylation in promoting M1-Mϕ polarization in ischemic hindlimbs of hyperlipidemic and T2DM mice, associated with impaired angiogenesis.16 In the current study, we report an impaired collateral remodeling and sprouting angiogenesis in response to FAL in T2DM mice, consistent with previous studies.14–16 Importantly, for the first time we addressed temporal changes in whole muscle and cell specific transcriptome during both collateral remodeling and sprouting angiogenesis, which is seldom addressed in mouse models of peripheral arterial disease. Furthermore, we evaluated ApoA-I nanotherapy to rescue postischemic vascular growth and remodeling defects in T2DM conditions. The novel finding from our study is the persistent IFN-I response gene upregulation in AECs, ECs, and Mϕs from T2DM mice associated with impaired collateral remodeling, angiogenesis and perfusion recovery. Indeed, type-I interferon (IFN-β) was shown to be enhanced in patients with insufficient coronary collateral artery development and demonstrated to inhibit arteriogenesis in mice. Furthermore, blocking IFN-β was shown to stimulate smooth muscle cell proliferation and arteriogenesis.32,33 IFN has also been shown to suppress angiogenesis indirectly via the regulation of immune cell production of angiogenic factors, such as VEGFs, in part by direct suppression of the proliferation of vascular cells and also via induction of anti-angiogenic chemokines.34,35 IFN induced inflammatory M1-Mϕ are shown to inhibit, whereas anti-inflammatory M2-Mϕ were shown to promote angiogenesis through fusion of vascular anastomosis.36,37

A persistent muscle inflammation was also reported in recent studies with T1DM and T2DM mouse models of HLI. Using T1DM mouse models, López-Díez et al showed a reduced expression of Ccl2, Cxcl2, Cxcl5, Angpt2, and Arg1 after hindlimb ischemia, associated with increased inflammation and impaired angiogenesis and perfusion recovery. Interestingly, deletion of gene encoding receptor for advanced glycation end products (RAGE) restored adaptive inflammation, angiogenesis and blood flow recovery in T1DM mice.38 Recently, Leung et al showed more CD4+ Th1 cells and fewer Tregs in T2DM patients with PAD. CD4 coreceptor blockade led to reduced vascular inflammation, promoted EC proliferation, and enhanced vascular function via Treg derived apelin mediated sprouting angiogenesis in Leprdb/db Mice.39 Ganta et al also demonstrated a persistent M1-Mϕ activation as causing an impaired angiogenesis using muscle-biopsies from PAD patients and murine models of HLI. Mechanistically, Mϕ-specific antiangiogenic VEGF165b was shown to inhibit VEGFR-1 in an autocrine fashion promoting M1-Mϕ activation and subsequent impairment of angiogenesis.40 Collectively, our present study along with these recent findings support a notion that a persistent inflammatory EC/M1-Mϕ activation and/or a reduced activation of anti-inflammatory M2-Mϕs contribute to impaired collateral remodeling and angiogenesis in T2DM conditions.

Previous studies have shown vasoprotective and anti-inflammatory effects of HDL in clinical and preclinical studies of atherosclerosis. HDL improves endothelial function by activating eNOS (endothelial nitric oxide synthase).41 HDL binds to endothelial scavenger receptor-B1 (SR-B1) and stimulates eNOS-mediated nitric oxide (NO) production through ApoA-I binding.42 HDL was shown to induce endothelial tube formation in vitro.43 HDL promotes the transport of excess cholesterol from Mϕs in peripheral tissues to the liver, that is, cholesterol efflux capacity.44 HDL was also shown to suppress IFN-I response and inhibit human M1-Mϕ polarization, while promoting M2-Mϕ polarization.45–47 In order to rescue T2DM induced defects in collateral remodeling and sprouting angiogenesis, we administered reconstituted phosphatidylserine (PS) nanoparticles embedded with ApoA-I (PS-rHDL) intravenously starting 2 days post-ischemia with doses every 2 days until sacrification. rHDL treatment improved collateral remodeling defects in adductor muscles by dampening AEC and Mϕ inflammation as evidenced by gene expression, immunostainings and FACS analysis. This was paralleled with increased peri-collateral accumulation of M2-Mϕs and smooth muscle cell/EC proliferation. rHDL treatment also led to a marked improvement in ischemic muscle perfusion followed by revascularization and muscle regeneration responses. Strikingly, rHDL could dampen persistent IFN-I gene expression in ECs, Mϕ and shift the phenotypes mimicking control mice.

Recent studies further support the beneficial role of rHDL in perfusion recovery and angiogenesis. In subjects with CLI, plasma concentrations of apoA-I, apoB (apolipoprotein B), and apoM were significantly lower than in control individuals, but only apoA-I was independently associated to CLI.48 Reconstituted HDL (rHDL) (containing 0.2 mg apoA-I in rHDL/body suspended in 0.3 mL PBS) injected intravenously twice per week, starting 1 week before surgery were shown to augment blood flow recovery and angiogenesis in mouse model of HLI via enhanced incorporation of bone marrow–derived cells in new vessels. rHDL was shown to promote differentiation of endothelial progenitor cells via PI3K/Akt pathway, whereas the beneficial effect of rHDL was abrogated in eNOS-deficient mice.49 ApoA-I (40 mg/kg) intravenous administrations every second day following surgery were shown to improve blood flow recovery in the ischemic limbs at Day 7. Consistent with these findings, apoA-I increased capillary density in the gastrocnemius muscle of ischemic hindlimbs. Furthermore, apoA-I increased hindlimb mRNA levels of 2 pro-angiogenic factors that play a key role in ischemia: VEGF and chemokine CXCL12 compared with PBS controls. ApoA-I infusions had no effect on plasma lipid profiles.50 In a streptozotocin-induced T1DM model, rHDL (200 μg/mouse) treatment every second day following surgery were shown to promote blood flow recovery and increased capillary density in the gastrocnemius muscle of ischemic hindlimbs. Mechanistically, these favourable outcomes were attributed at least in part, to enhanced posttranslational HIF-1α modulation and nuclear translocation, increased VEGF-A/VEGFR2 production and signaling, and augmented eNOS activity.51 HDL-mimicking nanodiscs carrying peptides were shown to enhance angiogenesis in diabetic HLI. rHDL was also shown to promote wound repair and blood flow recovery in response to ischemia in aged mice.52,53 Collectively, our findings together with recent reports suggest that ApoA-I nanotherapy is a potential therapeutic approach that can rescue post-ischemic vascular growth and remodeling defects in T2DM conditions by favorably modifying arterial and capillary EC and Mϕ phenotypes via dampening IFN-I gene expression signature.

Conclusions

Our results suggest that a persistent EC and Mϕ inflammation and decreased anti-inflammatory M2-Mϕ activation are involved in both impaired collateral remodeling and angiogenesis in T2DM mouse model of HLI. Dampening persistent inflammation and skewing ECs and Mϕ phenotypes towards less inflammatory ones using rHDL nanotherapy may serve as a potential therapeutic target for T2DM PAD.

Limitations

Our study has a few limitations. Our findings are in concordance with several recent studies on potential anti-inflammatory and vasculoprotective effects of ApoA-I. Since our rHDL formulations used POPS as a backbone, we cannot rule out the additive anti-inflammatory and vasculoprotective benefit afforded by POPS. Only male mice were used in the study, limiting the generalizability of the findings. Additionally, our single-cell deconvolution approach is not a substitute for a standard single-cell RNA-Seq analysis, which could provide high resolution information on cellular heterogeneity. Further studies using spatial transcriptomics will be needed to understand the spatial-temporal changes in divergent cell populations involved in postischemic tissue recovery.

Article Information

Acknowledgments

Authors would like to thank technicians from National Laboratory Animal Center for expert care of animals during the study.

Supplemental Material

Expanded Materials and Methods

Figures S1–S8

Major Resources Table

Tables S1–S4

Supplemental References

Nonstandard Abbreviations and Acronyms

ApoB

apolipoprotein B

ECM

extracellular matrix

eNOS

endothelial nitric oxide synthase

FACS

fluorescence-activated cell sorting

FAL

femoral artery ligation

HDL

high-density lipoprotein

IGF-II

insulin-like growth factor II

PAD

peripheral arterial disease

rHDL

reconstituted HDL

T2DM

type 2 diabetes

Disclosures None.

Footnotes

Supplemental Material is available at https://www.ahajournals.org/doi/suppl/10.1161/ATVBAHA.122.318196.

For Sources of Funding and Disclosures, see page e60.

Correspondence to: Seppo Ylä-Herttuala, MD, PhD, Department of Biotechnology and Molecular Medicine, A.I. Virtanen Institute, University of Eastern Finland, FIN-70211 Kuopio, Finland. Email

References

  • 1. Grundy SM, Benjamin IJ, Burke GL, Chait A, Eckel RH, Howard BV, Mitch W, Smith SC, Sowers JR. Diabetes and cardiovascular disease: A statement for healthcare professionals from the american heart association.Circulation. 1999; 100:1134–1146. doi: 10.1161/01.cir.100.10.1134LinkGoogle Scholar
  • 2. Abaci A, Oguzhan A, Kahraman S, Eryol NK, Unal S, Arinc H, Ergin A. Effect of diabetes mellitus on formation of coronary collateral vessels.Circulation. 1999; 99:2239–2242. doi: 10.1161/01.cir.99.17.2239LinkGoogle Scholar
  • 3. Selvin E, Marinopoulos S, Berkenblit G, Rami T, Brancati FL, Powe NR, Golden SH. Meta-analysis: Glycosylated hemoglobin and cardiovascular disease in diabetes mellitus.Ann Intern Med. 2004; 141:421–431. doi: 10.7326/0003-4819-141-6-200409210-00007CrossrefMedlineGoogle Scholar
  • 4. Van Houtum WH, Lavery LA. Outcomes associated with diabetes-related amputations in the netherlands and in the state of california, USA.J Intern Med. 1996; 240:227–231. doi: 10.1046/j.1365-2796.1996.42868000.xCrossrefMedlineGoogle Scholar
  • 5. Wrobel JS, Mayfield JA, Reiber GE. Geographic variation of lower-extremity major amputation in individuals with and without diabetes in the medicare population.Diabetes Care. 2001; 24:860–864. doi: 10.2337/diacare.24.5.860CrossrefMedlineGoogle Scholar
  • 6. Global Lower Extremity Amputation Study Group.Epidemiology of lower extremity amputation in centres in europe, north america and east asia. the global lower extremity amputation study group.Br J Surg. 2000; 87:328–337. doi: 10.1046/j.1365-2168.2000.01344.xCrossrefMedlineGoogle Scholar
  • 7. Beks PJ, Mackaay AJ, de Neeling JN, de Vries H, Bouter LM, Heine RJ. Peripheral arterial disease in relation to glycaemic level in an elderly caucasian population: The hoorn study.Diabetologia. 1995; 38:86–96. doi: 10.1007/BF02369357CrossrefMedlineGoogle Scholar
  • 8. Scholz D, Cai WJ, Schaper W. Arteriogenesis, a new concept of vascular adaptation in occlusive disease.Angiogenesis. 2001; 4:247–257. doi: 10.1023/a:1016094004084CrossrefMedlineGoogle Scholar
  • 9. Deindl E, Buschmann I, Hoefer IE, Podzuweit T, Boengler K, Vogel S, van Royen N, Fernandez B, Schaper W. Role of ischemia and of hypoxia-inducible genes in arteriogenesis after femoral artery occlusion in the rabbit.Circ Res. 2001; 89:779–786. doi: 10.1161/hh2101.098613LinkGoogle Scholar
  • 10. Wahlberg E. Angiogenesis and arteriogenesis in limb ischemia.J Vasc Surg. 2003; 38:198–203. doi: 10.1016/s0741-5214(03)00151-4CrossrefMedlineGoogle Scholar
  • 11. Heil M, Eitenmuller I, Schmitz-Rixen T, Schaper W. Arteriogenesis versus angiogenesis: Similarities and differences.J Cell Mol Med. 2006; 10:45–55. doi: 10.1111/j.1582-4934.2006.tb00290.xCrossrefMedlineGoogle Scholar
  • 12. Heil M, Schaper W. Influence of mechanical, cellular, and molecular factors on collateral artery growth (arteriogenesis).Circ Res. 2004; 95:449–458. doi: 10.1161/01.RES.0000141145.78900.44LinkGoogle Scholar
  • 13. Sheetz MJ, King GL. Molecular understanding of hyperglycemia’s adverse effects for diabetic complications.JAMA. 2002; 288:2579–2588. doi: 10.1001/jama.288.20.2579CrossrefMedlineGoogle Scholar
  • 14. Shen Y, Ding FH, Dai Y, Wang XQ, Zhang RY, Lu L, Shen WF. Reduced coronary collateralization in type 2 diabetic patients with chronic total occlusion.Cardiovasc Diabetol. 2018; 17:26. doi: 10.1186/s12933-018-0671-6CrossrefMedlineGoogle Scholar
  • 15. van Golde JM, Ruiter MS, Schaper NC, Voo S, Waltenberger J, Backes WH, Post MJ, Huijberts MS. Impaired collateral recruitment and outward remodeling in experimental diabetes.Diabetes. 2008; 57:2818–2823. doi: 10.2337/db08-0229CrossrefMedlineGoogle Scholar
  • 16. Babu M, Durga Devi T, Makinen P, Kaikkonen M, Lesch HP, Junttila S, Laiho A, Ghimire B, Gyenesei A, Yla-Herttuala S. Differential promoter methylation of macrophage genes is associated with impaired vascular growth in ischemic muscles of hyperlipidemic and type 2 diabetic mice: genome-wide promoter methylation study.Circ Res. 2015; 117:289–299. doi: 10.1161/CIRCRESAHA.115.306424LinkGoogle Scholar
  • 17. Gupta R, Tongers J, Losordo DW. Human studies of angiogenic gene therapy.Circ Res. 2009; 105:724–736. doi: 10.1161/CIRCRESAHA.109.200386LinkGoogle Scholar
  • 18. Carmeliet P. Mechanisms of angiogenesis and arteriogenesis.Nat Med. 2000; 6:389–395. doi: 10.1038/74651CrossrefMedlineGoogle Scholar
  • 19. Semenza GL. Vasculogenesis, angiogenesis, and arteriogenesis: Mechanisms of blood vessel formation and remodeling.J Cell Biochem. 2007; 102:840–847. doi: 10.1002/jcb.21523CrossrefMedlineGoogle Scholar
  • 20. Riwanto M, Landmesser U. High density lipoproteins and endothelial functions: Mechanistic insights and alterations in cardiovascular disease.J Lipid Res. 2013; 54:3227–3243. doi: 10.1194/jlr.R037762CrossrefMedlineGoogle Scholar
  • 21. Lemmers RFH, van Hoek M, Lieverse AG, Verhoeven AJM, Sijbrands EJG, Mulder MT. The anti-inflammatory function of high-density lipoprotein in type II diabetes: A systematic review.J Clin Lipidol. 2017; 11:712–724.e5. doi: 10.1016/j.jacl.2017.03.013CrossrefMedlineGoogle Scholar
  • 22. Barrett TJ, Distel E, Murphy AJ, Hu J, Garshick MS, Ogando Y, Liu J, Vaisar T, Heinecke JW, Berger JS, et al. Apolipoprotein AI) promotes atherosclerosis regression in diabetic mice by suppressing myelopoiesis and plaque inflammation.Circulation. 2019; 140:1170–1184. doi: 10.1161/CIRCULATIONAHA.119.039476LinkGoogle Scholar
  • 23. Heinonen SE, Leppanen P, Kholova I, Lumivuori H, Hakkinen SK, Bosch F, Laakso M, Yla-Herttuala S. Increased atherosclerotic lesion calcification in a novel mouse model combining insulin resistance, hyperglycemia, and hypercholesterolemia.Circ Res. 2007; 101:1058–1067. doi: 10.1161/CIRCRESAHA.107.154401LinkGoogle Scholar
  • 24. Limbourg A, Korff T, Napp LC, Schaper W, Drexler H, Limbourg FP. Evaluation of postnatal arteriogenesis and angiogenesis in a mouse model of hind-limb ischemia.Nat Protoc. 2009; 4:1737–1746. doi: 10.1038/nprot.2009.185CrossrefMedlineGoogle Scholar
  • 25. Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads.EMBnet.journal. 2011; 17:10–12. Accessed November 10, 2020. http://journal.embnet.org/index.php/embnetjournal/article/view/200. doi: https://doi.org/10.14806/ej.17.1.200Google Scholar
  • 26. Ewels PA, Peltzer A, Fillinger S, Patel H, Alneberg J, Wilm A, Garcia MU, Di Tommaso P, Nahnsen S. The nf-core framework for community-curated bioinformatics pipelines.Nat Biotechnol. 2020; 38:276–278. doi: 10.1038/s41587-020-0439-xCrossrefMedlineGoogle Scholar
  • 27. Robinson MD, McCarthy DJ, Smyth GK. edgeR: A bioconductor package for differential expression analysis of digital gene expression data.Bioinformatics. 2010; 26:139–140. doi: 10.1093/bioinformatics/btp616CrossrefMedlineGoogle Scholar
  • 28. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2.Genome Biol. 2014; 15:550. doi: 10.1186/s13059-014-0550-8CrossrefMedlineGoogle Scholar
  • 29. Farbehi N, Patrick R, Dorison A, Xaymardan M, Janbandhu V, Wystub-Lis K, Ho JW, Nordon RE, Harvey RP. Single-cell expression profiling reveals dynamic flux of cardiac stromal, vascular and immune cells in health and injury.Elife. 2019; 8:1. doi: 0.7554/eLife.43882CrossrefGoogle Scholar
  • 30. Stuart T, Butler A, Hoffman P, Hafemeister C, Papalexi E, Mauck WM, Hao Y, Stoeckius M, Smibert P, et al. Comprehensive integration of single-cell data.Cell. 2019; 177:1888–1902.e21. doi: 10.1016/j.cell.2019.05.031CrossrefMedlineGoogle Scholar
  • 31. Frishberg A, Peshes-Yaloz N, Cohn O, Rosentul D, Steuerman Y, Valadarsky L, Yankovitz G, Mandelboim M, Iraqi FA, Amit I, et al. Cell composition analysis of bulk genomics using single-cell data.Nat Methods. 2019; 16:327–332. doi: 10.1038/s41592-019-0355-5CrossrefMedlineGoogle Scholar
  • 32. Schirmer SH, Fledderus JO, Bot PT, Moerland PD, Hoefer IE, Baan J, Henriques JP, van der Schaaf RJ, Vis MM, et al. Interferon-beta signaling is enhanced in patients with insufficient coronary collateral artery development and inhibits arteriogenesis in mice.Circ Res. 2008; 102:1286–1294. doi: 10.1161/CIRCRESAHA.108.171827LinkGoogle Scholar
  • 33. Schirmer SH, Bot PT, Fledderus JO, van der Laan AM, Volger OL, Laufs U, Böhm M, de Vries CJ, Horrevoets AJ, Piek JJ, et al. Blocking interferon {beta} stimulates vascular smooth muscle cell proliferation and arteriogenesis.J Biol Chem. 2010; 285:34677–34685. doi: 10.1074/jbc.M110.164350CrossrefMedlineGoogle Scholar
  • 34. Zheng H, Qian J, Carbone CJ, Leu NA, Baker DP, Fuchs SY. Vascular endothelial growth factor-induced elimination of the type 1 interferon receptor is required for efficient angiogenesis.Blood. 2011; 118:4003–4006. doi: 10.1182/blood-2011-06-359745CrossrefMedlineGoogle Scholar
  • 35. Indraccolo S, Pfeffer U, Minuzzo S, Esposito G, Roni V, Mandruzzato S, Ferrari N, Anfosso L, Dell’Eva R, Noonan DM, et al. Identification of genes selectively regulated by IFNs in endothelial cells.J Immunol. 2007; 178:1122–1135. doi: 10.4049/jimmunol.178.2.1122CrossrefMedlineGoogle Scholar
  • 36. Spiller KL, Anfang RR, Spiller KJ, Ng J, Nakazawa KR, Daulton JW, Vunjak-Novakovic G. The role of macrophage phenotype in vascularization of tissue engineering scaffolds.Biomaterials. 2014; 35:4477–4488. doi: 10.1016/j.biomaterials.2014.02.012CrossrefMedlineGoogle Scholar
  • 37. Zajac E, Schweighofer B, Kupriyanova TA, Juncker-Jensen A, Minder P, Quigley JP, Deryugina EI. Angiogenic capacity of M1- and M2-polarized macrophages is determined by the levels of TIMP-1 complexed with their secreted proMMP-9.Blood. 2013; 122:4054–4067. doi: 10.1182/blood-2013-05-501494CrossrefMedlineGoogle Scholar
  • 38. Lopez-Diez R, Shen X, Daffu G, Khursheed M, Hu J, Song F, Rosario R, Xu Y, Li Q, Xi X, et al. Ager deletion enhances ischemic muscle inflammation, angiogenesis, and blood flow recovery in diabetic mice.Arterioscler Thromb Vasc Biol. 2017; 37:1536–1547. doi: 10.1161/ATVBAHA.117.309714LinkGoogle Scholar
  • 39. Leung OM, Li J, Li X, Chan VW, Yang KY, Ku M, Ji L, Sun H, Waldmann H, Tian XY, et al. Regulatory T cells promote apelin-mediated sprouting angiogenesis in type 2 diabetes.Cell Rep. 2018; 24:1610–1626. doi: 10.1016/j.celrep.2018.07.019CrossrefMedlineGoogle Scholar
  • 40. Ganta VC, Choi M, Farber CR, Annex BH. Anti-angiogenic VEGF165b regulates macrophage-polarization via S100A8/S100A9 in peripheral artery disease.Circulation. 2019; 139:226–242. doi: 10.1161/CIRCULATIONAHA.118.034165LinkGoogle Scholar
  • 41. Nofer JR, van der Giet M, Tölle M, Wolinska I, von Wnuck Lipinski K, Baba HA, Tietge UJ, Gödecke A, Ishii I, Kleuser B, et al. HDL induces NO-dependent vasorelaxation via the lysophospholipid receptor S1P3.J Clin Invest. 2004; 113:569–581. doi: 10.1172/JCI18004CrossrefMedlineGoogle Scholar
  • 42. Yuhanna IS, Zhu Y, Cox BE, Hahner LD, Osborne-Lawrence S, Lu P, Marcel YL, Anderson RG, Mendelsohn ME, Hobbs HH, et al. High-density lipoprotein binding to scavenger receptor-BI activates endothelial nitric oxide synthase.Nat Med. 2001; 7:853–857. doi: 10.1038/89986CrossrefMedlineGoogle Scholar
  • 43. Miura S, Fujino M, Matsuo Y, Kawamura A, Tanigawa H, Nishikawa H, Saku K. High density lipoprotein-induced angiogenesis requires the activation of ras/MAP kinase in human coronary artery endothelial cells.Arterioscler Thromb Vasc Biol. 2003; 23:802–808. doi: 10.1161/01.ATV.0000066134.79956.58LinkGoogle Scholar
  • 44. Rosenson RS, Brewer HB, Davidson WS, Fayad ZA, Fuster V, Goldstein J, Hellerstein M, Jiang XC, Phillips MC, et al. Cholesterol efflux and atheroprotection: Advancing the concept of reverse cholesterol transport.Circulation. 2012; 125:1905–1919. doi: 10.1161/CIRCULATIONAHA.111.066589LinkGoogle Scholar
  • 45. Sanson M, Distel E, Fisher EA. HDL induces the expression of the M2 macrophage markers arginase 1 and fizz-1 in a STAT6-dependent process.PLoS One. 2013; 8:e74676. doi: 10.1371/journal.pone.0074676CrossrefMedlineGoogle Scholar
  • 46. Lee MK, Moore XL, Fu Y, Al-Sharea A, Dragoljevic D, Fernandez-Rojo MA, Parton R, Sviridov D, Murphy AJ, Chin-Dusting JP. High-density lipoprotein inhibits human M1 macrophage polarization through redistribution of caveolin-1.Br J Pharmacol. 2016; 173:741–751. doi: 10.1111/bph.13319CrossrefMedlineGoogle Scholar
  • 47. Suzuki M, Pritchard DK, Becker L, Hoofnagle AN, Tanimura N, Bammler TK, Beyer RP, Bumgarner R, Vaisar T, de Beer MC, et al. High-density lipoprotein suppresses the type I interferon response, a family of potent antiviral immunoregulators, in macrophages challenged with lipopolysaccharide.Circulation. 2010; 122:1919–1927. doi: 10.1161/CIRCULATIONAHA.110.961193LinkGoogle Scholar
  • 48. Ahnström J, Gottsäter A, Lindblad B, Dahlbäck B. Plasma concentrations of apolipoproteins A-I, B, and M in patients with critical limb ischemia.Clin Biochem. 2010; 43:599–603. doi: 10.1016/j.clinbiochem.2010.01.001CrossrefMedlineGoogle Scholar
  • 49. Sumi M, Sata M, Miura S, Rye KA, Toya N, Kanaoka Y, Yanaga K, Ohki T, Saku K, Nagai R. Reconstituted high-density lipoprotein stimulates differentiation of endothelial progenitor cells and enhances ischemia-induced angiogenesis.Arterioscler Thromb Vasc Biol. 2007; 27:813–818. doi: 10.1161/01.ATV.0000259299.38843.64LinkGoogle Scholar
  • 50. Prosser HC, Tan JT, Dunn LL, Patel S, Vanags LZ, Bao S, Ng MK, Bursill CA. Multifunctional regulation of angiogenesis by high-density lipoproteins.Cardiovasc Res. 2014; 101:145–154. doi: 10.1093/cvr/cvt234CrossrefMedlineGoogle Scholar
  • 51. Tan JT, Prosser HC, Dunn LL, Vanags LZ, Ridiandries A, Tsatralis T, Lecce L, Clayton ZE, Yuen SC, Robertson S, et al. High-density lipoproteins rescue diabetes-impaired angiogenesis via scavenger receptor class B type I.Diabetes. 2016; 65:3091–3103. doi: 10.2337/db15-1668CrossrefMedlineGoogle Scholar
  • 52. Park HJ, Kuai R, Jeon EJ, Seo Y, Jung Y, Moon JJ, Schwendeman A, Cho SW. High-density lipoprotein-mimicking nanodiscs carrying peptide for enhanced therapeutic angiogenesis in diabetic hindlimb ischemia.Biomaterials. 2018; 161:69–80. doi: 10.1016/j.biomaterials.2018.01.027CrossrefMedlineGoogle Scholar
  • 53. Tsatralis T, Ridiandries A, Robertson S, Vanags LZ, Lam YT, Tan JT, Ng MK, Bursill CA. Reconstituted high-density lipoproteins promote wound repair and blood flow recovery in response to ischemia in aged mice.Lipids Health Dis. 2016; 15:150. doi: 10.1186/s12944-016-0322-4.CrossrefMedlineGoogle Scholar