Association of hypoxia and mitochondrial damage associated molecular patterns in the pathogenesis of vein graft failure: A PILOT STUDY

Abstract

Coronary artery bypass grafting (CABG) is the standard treatment modality in revascularization of the myocardium. However, the graft failure remains the major complication following CABG procedure. Involvement of mitochondrial damage-associated molecular patterns (mt-DAMPs) in the pathogenesis of vein-graft failure is largely unknown. Here, we investigated the expression of major protein-mt-DAMPs, cytochrome-C (Cyt-C), heat shock protein-60 (Hsp-60), mitochondrial transcription factor A (mtTFA), in the occluded graft and associated tissues, including distal left anterior descending (LAD), LAD adjacent to anastomosis, and left internal mammary artery (LIMA) in the microswine CABG model. The protein expression of peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC-1α) was significantly decreased in the graft and LIMA, whereas the protein expression of hypoxia inducible factor-1 alpha (HIF-1α) and Cyt-C was decreased and that of mtTFA and Hsp60 was increased in all tissues compared to controls. There was no significant difference in the protein expression of citrate synthase, complex-1, and mitochondrial pyruvate dehydrogenase in the graft and associated tissues compared to control. Hypoxia in cultured smooth muscle cells (SMCs) significantly upregulated all mitochondrial biomarkers and mt-DAMPs compared to normoxia. The increased reactive oxygen species (ROS) content and compromised membrane integrity in the hypoxic SMCs correlated well with increased mt-DAMPs in the graft and associated tissues, suggesting a possible role of mt-DAMPs in the pathogenesis of graft failure. These findings suggest that the pathological signals elicited by mt-DAMPs could reveal targets for better therapeutic approaches and diagnostic strategies in the management of CABG graft failure.

Brief Commentary

Patency of the vein graft decreases over time following the CABG leading to graft failure resulting in the recurrence of MI and associated complications. The elevated levels of major mt-DAMPs in the occluded CABG vein grafts and adjacent vessels suggest an association of mt-DAMPs in the underlying pathology of vein graft failure where the hypoxic insults trigger the release of mt-DAMPs from vascular SMCs. Better understanding of the pathological signaling elicited by mt-DAMPs and their regulation in SMCs open immense translational avenues in the development of management/diagnostic strategies of CABG graft failure.

INTRODUCTION

Coronary artery bypass grafting (CABG) remains the ultimate treatment modality in the revascularization of the myocardium which has significantly increased the survival of patients with multiple coronary artery diseases (CAD). However, the patency of the graft decreases over time leading to graft failure which in turn is associated with myocardial infarction (MI), recurrent angina, repeated coronary revascularization procedures and even death¹. Saphenous vein grafts (SVGs), the widely used conduit for CABG in human, have the failure rate more than 25% within 12–18 months of the CABG procedure and the 10-year failure rate is 50–60%². Both intrinsic and extrinsic factors contribute to SVG failure, and include the time of vein harvest, quality of the vein, pre-existing vascular pathology, trauma and/or endothelial damage during the harvest, technical insufficiencies during the surgical procedure and damage to the vasa vasorum and nervi vasorum during the harvest³. Importantly, ischemic insult and decreased levels of nitric oxide (NO) lead to hyperplasia of smooth muscle cells (SMCs) affecting the hemodynamics and physiology of the graft and subsequent failure³.

Overall pathology of the vein graft failure represents complete occlusion of the lumen in the grafted conduit, preventing the blood flow through the graft to the site of revascularization in the heart. However, pathological events and the underlying mechanisms differ during the course of time which include acute thrombosis (within 30 days) followed by intimal hyperplasia (within 1-year) and atherosclerosis (beyond 1-year)⁴. Recently, it has been established that the degradation fragments (generated as a result of hypoxia/ischemia) of extracellular matrix (ECM), such as hyaluronic acid, proteoglycans and fibronectin, act as damage associated molecular patterns (DAMPs) which function as endogenous ligands for toll-like receptors (TLRs) triggering the sterile inflammatory events in the surviving tissue and results in accelerated graft remodeling⁵. In addition, the hypoxic insult and associated ischemia/reperfusion increase the pool of DAMPs and enhance oxidative stress in the vein graft which in turn result in vascular damage and upregulation of pro-inflammatory cytokines and subsequent remodeling⁶. In addition, the increased pro-inflammatory pool created by the pro-atherogenic environment aggravates the oxidative/ischemic damage. These events occur within the first day of the CABG procedure, mark the initiation of the pathogenesis of graft failure and pave the ways for increased influx of inflammatory cells to the graft vessel sustaining inflammation⁷.

Mitochondrial dysfunction in the vessel graft due to hypoxic/ischemic episodes disturbs the vascular homeostasis resulting in SMC hyperplasia⁸. Various stimuli including hypoxia alter the mitochondrial function and de-differentiate SMCs from contractile phenotype to proliferative synthetic phenotype, which in turn results in intimal hyperplasia⁹. However, the molecular signaling that triggers the mitochondrial dysfunction, sterile inflammation and subsequent vein graft failure has not been explored. Moreover, the pathological signaling elicited by various mitochondrial-DAMPs (mt-DAMPs) released from the mitochondria of injured/stressed cells under hypoxic/ischemic episodes accelerate inflammation in diverse tissue types¹⁰. The mt-DAMPs, including mtDNA, succinate, n-formyl peptide, cytochrome-C (Cyt-C), heat shock protein-60 (Hsp-60), mitochondrial transcription factor A (mtTFA) and ATP released during tissue damage, induce innate or adaptive immune responses via the activation of cell surface receptors especially the Nod-like receptors (NLRs)¹⁰. However, the involvement/association of mt-DAMPs in the pathogenesis of vein graft failure following CABG procedures and the effect of hypoxic insults in mt-DAMP release are largely unknown regarding the graft and vessels adjacent to or associated with anastomosis. Generally, the overall pool of mt-DAMPs in the vascular lesion are contributed by multiple cell types including immune cells, endothelial cells and SMCs; however, the major contributors for intimal hyperplasia are SMCs⁹. Therefore, the present study focuses on the association of major mt-DAMPs in occluded vein grafts and adjacent vessels followed by CABG procedure and to investigate the mt-DAMP release by the cultured SMCs under hypoxic conditions. This study utilized the vein grafts and associated vessels from a novel Yucatan microswine CABG model and the isolated arterial SMCs.

METHODOLOGY

Yucatan microswine and treatment

The research protocol of the study was approved by the Institutional Animal Care and Use Committee of Creighton University. Three female Yucatan microswine (Sus scorfa, Sinclair bioresources) (n=3) that survived the CABG procedure were included in this study. The animal procedures strictly followed the NIH and OLAW guidelines throughout the study and extreme care was taken to minimize pain and distress to the animals. The animals (8–10 months of age, weighing 60–80 pound) were acclimatized for one week, maintained in 12/12 light-dark cycle, fed with high cholesterol high fat diet twice daily for 6–12 months prior to the procedure. The complete blood count, comprehensive metabolic panel (CMP) and blood lipid profile were regularly monitored to confirm hypercholesterolemia. The animals were given uninterrupted access to water throughout the study and the hypercholesterolemic swine were subjected to CABG surgery.

CABG surgery and vein graft harvest

Superficial epigastric vein (SEV) was used as the graft and the CABG was performed by anastomosing the SEV proximally to the left internal mammary artery (LIMA) and distally to the to the diagonal branch of left anterior descending (LAD) artery. The surgical procedures were typically similar to human CABG and the procedures were performed by skilled cardiac surgeons. The vascular system was accessed through femoral artery using ultrasound guided percutaneous puncture with a follow up monitor using EKG, echocardiography, and coronary angiography. Briefly, the hypercholesterolemic Yucatan microswine were administered with prophylactic antiplatelet (aspirin 350 mg) for 2 days prior to the surgery and were fasted overnight. The animals were anesthetized with acepromazine and ketamine and were maintained under isoflurane (1–4%) in oxygen throughout the surgery. The antibiotic (cephazolin 3 mg/kg IM), analgesic (Carprofen 2.2 mg/kg IM) and antiarrhythmic (amiodarone 5–10 mg/kg IV) drugs were administered as prophylactic medication. The animals were maintained in mechanical ventilation (tidal volume of 10–15 ml/kg and respiratory rate of 12–20 breath/min), covered with circulating warm water blankets to prevent hypothermia and the arterial and central venous pressures were monitored respectively via femoral arterial and venous lines. The animals were closely monitored and were administered with medications accordingly; whenever required using the intravenous line. SEV graft (10–15 cm) was harvested aseptically guided with doppler ultrasound and was stored in sterile heparinized saline solution at room temperature. The heart was exposed by median sternotomy followed by the incision of pericardium. The SEV graft was anastomosed to the LIMA proximally (end-to-side fashion) and the LAD distally (end-to-side anastomosis) using 7–0 prolene suture. The patency of the SEV graft after anastomoses was confirmed by the blood flow and the absence of the signs of myocardial infarction. Following the CABG procedure, sternotomy incision was closed using sterile stainless-steel suture (size 5–7), the chest drainage system was inserted, and the skin was sutured with subcuticular vicryl suture and sealed with skin glue. The chest drainage was removed when the animals regained the hemodynamic stability and long acting local anesthetic, liposomal bupivacaine (0.2 ml/kg), was locally administered to the sternal wound. Post-surgery, the animals were closely observed for arterial blood pressure, SpO₂, EKG, urine and chest drain output and were administered with inotropic and antiarrhythmic drugs, and injectable coronary vasodilator, nitroglycerine. Upon complete recovery from anesthesia the animals were extubated, and the arterial and venous lines were removed. The animals were supplemented with oxygen via a face mask and non-narcotic analgesic, carprofen, was administered. The animals were regularly monitored for their general health until the wounds were healed and were under antiplatelet medication until sacrificed. Two out of six animals died during the procedure and the survived animals were the candidates for further studies.

Post-CABG microswine were continued to be fed with high cholesterol-high fat diet, sacrificed after 6 months of CABG procedure (n=3) and the tissues including the SEV graft, LIMA, LAD and the LAD adjacent to anastomoses (LAD-AA) were harvested for analysis. The harvested tissues were formalin fixed at room temperature for 24 h, embedded in paraffin and sections of 5μm thickness were taken onto microscopic slides for histology and immunofluorescence analyses. SEV harvested freshly from female atherosclerotic pigs (n=3) reared under same dietary plan as in CABG pigs were used as control for comparison. The control pigs (8 months old) were fed with high fat diet twice daily for 4–8 months prior to SEV harvest and were monitored similarly as mentioned above; however, were excluded from CABG procedure.

Histology

The tissue sections were deparaffinized and pentachrome staining was performed to examine the tissue morphology, ECM components and matrix organization following our previously reported protocol¹¹. The stained slides were mounted using xylene based mounting media and imaged using a fluorescent slide scanner system (VS120-S6-W, Olympus) at 20x magnification and the taken images were converted to TIF format using OlyVIA Desktop software.

Immunofluorescence

The protein expression of biomarkers to assess mitochondrial health status and mt-DAMP molecules in the harvested vessel tissue sections were analyzed by assessing the immunopositivity following our previously reported protocol¹². Briefly, the deparaffinized sections were rehydrated, subjected to antigen retrieval by heating in HIER buffer (Heat Induced Antigen Retrieval) (TA-135-HBM) at 95°C for 20 min followed by blocking (0.25 % Triton X-100 and 5% horse serum in PBS) at room temperature for 90 min. Primary antibodies were used in a dilution of 1:200, and included the antibodies against the mitochondrial biomarkers including citrate synthase (CS) (ab-96600), complex-1 (ab109798), mitochondrial pyruvate dehydrogenase (mt-PDH) (ab92696), PGC-1α (sc-518025), the hypoxia response mediator hypoxia inducible factor-1 alpha (total) (HIF-1α) (ab1), and mt-DAMPs including Cyt-C (ab110325), Hsp-60 (ab46798), and mtTFA (ab176558). The fluorochrome-conjugated secondary antibodies with a dilution of 1:400 were used to bind the primary antibodies. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) (H-1200) and imaged using the fluorescent slide scanner system (VS120-S6-W, Olympus). Similarly, the SEV controls were examined and imaged using the fluorescent slide scanner system (Thunder, Leica). Two-to-four images (depending on the size of tissue) acquired randomly from each specimen were used for the quantification of mean fluorescence intensity (MFI) using ImageJ software. The number of nuclei was counted from each image using ‘Analyze Particle’ mode by setting standard threshold in the software. The MFI was normalized to 100 cells using the formula (MFI/Number of cells in the field) x 100, and the results are expressed as mean fluorescence intensity/100 cells (MFI/100 cells). The average MFI/100 cells from 2–4 images of each specimen was calculated which in turn was used for statistical analysis. A negative control was run in parallel without the primary antibody to detect background fluorescence and to set the exposure time.

SMC isolation, culture and induction of hypoxia

SMCs were isolated from the common carotid arteries of the atherosclerotic female microswine using collagenase digestion method following the established protocol from our laboratory. The SMCs were characterized by the immunopositivity to α-SMA¹³. The cells were maintained in SMC growth media (Millipore Sigma) supplemented with 10% fetal bovine serum and antibiotics at 37°C and 5% CO₂ in a humidified incubator. The SMCs isolated from different microswine were pooled (passage 2–5) and stored under liquid nitrogen. The cells were split at 70–80% confluency and the cells in 2–5 passages were used in the study and all experiments were conducted in quadruplicate. The SMCs cultures were subjected to hypoxia by incubating overnight in a hypoxia chamber by passing N₂ gas to maintain 2% pO₂.

Determination of mitochondrial oxidative stress

The status of ROS and oxidative stress in the hypoxia-challenged SMCs was examined by determining mitochondrial superoxide using MitoSOX Red (M36008, Invitrogen) following our previously published protocol¹⁴. Briefly, the SMCs were grown under hypoxia in 4-well chamber slides and were treated with 5μM MitoSOX Red reagent for 10 min in DMEM (serum free) and the live cell imaging was performed using a fluorescent slide scanner (VS120-S6-W, Olympus). The cells grown under normoxia served as the control. The experiments were run in quadruplicate and the MFI values were normalized to 100 cells and compared with the control.

Mitochondrial pore transition assay

The mitochondrial pore integrity was assessed using Image-iT LIVE mitochondrial transition pore assay kit (I35103, Invitrogen) following our previously reported protocol¹⁴. Briefly, the SMCs grown in chamber slides under hypoxia were washed with serum-free DMEM and incubated in 200μl labeling solution (1μl each of 1mM calcein AM, 200μM MitoTracker Red, 1mM Hoechst 33342, and 1M CoCl₂ in 1ml serum free DMEM) at 37°C for 15 min. After incubation, the cells were washed with serum-free DMEM and immediately imaged under fluorescence slide scanner (VS120-S6-W, Olympus). The normoxic SMCs were used as control. The experiments were run in quadruplicate and the MFI values were normalized to 100 cells.

Protein expression of mitochondrial parameters under hypoxia

The SMCs were grown under hypoxia (as mentioned above) in chamber slides, fixed with 10% formalin for 20 min and immunofluorescence of mitochondrial markers, including citrate synthase, complex-1, mt-PDH, PGC-1α, the hypoxia response mediator HIF-1α, and mt-DAMPs including Cyt-C, Hsp-60, and mtTFA, was performed¹⁴. The dilution of 1:400 and 1:500 were used for primary and the fluorochrome-conjugated secondary antibodies, respectively. The cells grown under normoxic conditions served as control and the experiments were done in triplicates. The expression of each genes was represented as MFI/100 cells.

Statistical analysis

The results of immunofluorescence intensity in the microswine (n=3) tissues and cell culture experiments (n=4) were expressed as mean ± SEM. The nuclei were counted using ImageJ software with ‘Analyze particle’ mode and the threshold values for the count were maintained between 80–100 for all specimen. The mean for each specimen was the average MFI of 2–4 images randomly acquired from different fields in the experimental tissues and the chamber slides of cultured SMCs. The images that displayed background fluorescence due to the presence of sutures were exempted from the analysis. The statistical significance for in vivo experiments was determined by one way-ANOVA using Dunnett’s multiple comparison test and for in vitro experiments were determined by unpaired ‘t’ test for non-parametric data using Mann Whitney test employing GraphPad Prism software. The p<0.05 values were considered significant in all experiments.

RESULTS

CABG model

The CABG procedure was successfully completed in three Yucatan microswine by anastomosing SEV graft from LIMA to LAD (Fig. 1A). Post-CABG animals were recovered and returned to normal cage activities within 7–10 days with complete healing of the thoracotomy wound. The animals were continued on high calorie diet until sacrificed 6 months after the CABG procedure. The graft was found to be completely occluded in all three animals and was covered with fibroadipose tissue which offered challenge to trace the entire graft (Fig. 1B). The portion of the SEV graft, intact LAD distal to anastomoses, LAD adjacent to the anastomoses (LAD-AA) and the LIMA distal to anastomoses were harvested from each animal for further analysis. The SEV harvested freshly from the female pigs provided with same atherosclerosis-inducing dietary regimen, and similar timeline and without CABG surgery served as the control for comparison.

Fig. 1:

(A) SEV graft anastomosed from LIMA to LAD by CABG procedure showing the blood flow through the graft conduit. (B) The post-CABG after 6 months showing the graft covered by fibroadipose tissue and occlusion.

Pentachrome staining

Pentachrome staining displayed considerable difference in ECM components between graft and other three vessels; LAD, LAD-AA and LIMA (Fig. 2). The images were acquired randomly from the similar microscopic fields/locations of each vessel and were assessed qualitatively based on the histomorphology. Increased ECM disorganization was evident in the graft tissues whereas LAD, LAD-AA and LIMA showed minimal alterations in ECM orientation. The decreased cell density (evident from the decreased violet staining of nucleus by hematoxylin) was observed in the graft tissues compared with other groups. The graft tissues exhibited highly fibrous collagenous tissue with diffused and/or disoriented elastic fibers (yellow color) than the other three groups. On the other hand, LAD, LAD-AA and LIMA displayed organized collagen and elastic fibers (black stain). The collagen content (yellow color) was minimal in LIMA and LAD than LAD-AA, whereas the elastic fiber content was higher in LAD-AA when compared with LIMA and LAD. The mucin (blue-green color) content was negligible in the graft tissue upon comparison with LAD, LAD-AA and LIMA. In contrast, LAD and LAD-AA exhibited increased mucin content compared to LIMA. The foam cells were completely absent in the graft, minimal in LAD and prevalent in LAD-AA and LIMA. The staining for smooth muscle (pinkish red stain) component was negligible in the graft tissue while predominated in the other three groups. The lumen was completely occluded in the graft tissue whereas the intimal hyperplasia was evident in LAD, LAD-AA and LIMA.

Fig. 2: The histomorphology evaluations Movat’s Pentachrome staining:

The evaluations were performed in four vessels LAD, LAD-AA, graft and LIMA. The yellow arrows signify smooth muscle tissue; the black arrows point collagen fibers; green arrows show bundles of elastic fibers; red arrows specify mucin content; and the blue arrows point foam cells. The images were acquired in 40x magnification.

Protein expression of mitochondrial mediators and mt-DAMPs

The protein expression of mitochondrial biogenesis biomarker PGC1α was significantly decreased in the graft tissues (P=0.0080) and LIMA (P=0.0230) when compared to SEV control. Similarly, the MFI for the protein expression of PGC1α in LAD (P=0.2640) and LAD-AA (P=0.0726) was decreased; however, was statistically not significant when compared to SEV (Figs. 3A, 3B). The protein expression of the mitochondrial biomarker complex-1 decreased non-significantly in the graft (P=0.1365), LAD-AA (P=0.5204), LIMA (P=0.4874) and LAD (P=0.0843) than the SEV control (Figs. 3C, 3D). MFI for the expression of the hypoxia responsive mediator HIF-1α (total) was significantly higher in SEV control when compared with LIMA (P<0.0001), LAD (P<0.0001), LAD-AA (P<0.0001) and the graft (P<0.0001) (Figs. 4A, 4B). Similar level of protein expression of mtPDH was observed in LAD-AA (P=0.9614) whereas the increase in protein levels of mtPDH in LAD (P=0.4596), graft (P>0.9999), and LIMA (P=0.5157) were statistically not significant when compared with the SEV control (Figs. 4A, 4C). The protein expression for CS was increased in all four vessels; however, was statistically not significant (P=0.0967, P=0.0511, P=0.8255, and P=0.2717 for LAD, LAD-AA, graft and LIMA, respectively) compared to SEV control (Figs. 5A, 5B). The expression of PGC1α and complex-1 were mainly confined to the intimal layer in LAD and LIMA and throughout the tissues in the graft and LAD-AA whereas mtPDH and CS were expressed throughout in all five experimental groups (Figs. 3A, 3C, 4A and 5A).

Fig. 3:

Immunofluorescence analysis for the expression of PGC1α (panel A) and complex-1 (panel C) showing altered expression in the vessel groups. Images in the top row are histological sections of LAD, followed by LAD-AA (second row) and graft tissue (third row) and the images in the bottom rows are histological sections of LIMA. Images in the left column show nuclear staining with DAPI; the images in the middle column show expression of PGC1α/complex1 while the images in the right column show overlay of PGC1α/complex1 staining with DAPI. Images were acquired at 20x. The images in (B) and (D) panels show the quantification of gene expression. The intensity of gene expression obtained through immunofluorescence was acquired and the mean fluorescence intensity normalized to 100 cells (MFI/100 cells) was quantified from each vessel specimen. The graphs represent MFI mean values with standard error. The statistical significance based on the theoretical mean of each group is represented in the figure (NS – non-significant, **P<0.01 and * P<0.05).

Fig. 4:

Immunofluorescence analysis for the expression of HIF-1α and mtPDH (panel A) showing the altered expression in the vessel groups. Images in the top row are histological sections of LAD, followed by LAD-AA (second row) and graft tissue (third row) and the images in the bottom rows are histological sections of LIMA. Images in the left column show nuclear staining with DAPI; the images in the middle column show expression of HIF-1α/mtPDH while the images in the right column show overlay of HIF-1α/mtPDH staining with DAPI. Images were acquired at 20x magnification. The images in (B) and (C) panels show the quantification of gene expression. The intensity of gene expression obtained through immunofluorescence was acquired and the mean fluorescence intensity normalized to 100 cells (MFI/100 cells) was quantified from each vessel specimen. The graphs represent MFI mean values with standard error. The statistical significance based on the theoretical mean of each group is represented in the figure (NS – non-significant, and **** P<0.0001).

Fig. 5:

Immunofluorescence analysis for the expression of citrate synthase (CS; panel A) and mtTFA (panel C) showing the altered expression in the vessel groups. Images in the top row are histological sections of LAD, followed by LAD-AA (second row) and graft tissue (third row) and the images in the bottom rows are histological sections of LIMA. Images in the left column show nuclear staining with DAPI; the images in the middle column show expression of CS/mtTFA while the images in the right column show overlay of CS/mtTFA staining with DAPI. Images were acquired at 20x magnification. The images in (B) and (D) panels show the quantification of gene expression. The intensity of gene expression obtained through immunofluorescence was acquired and the mean fluorescence intensity normalized to 100 cells (MFI/100 cells) was quantified from each vessel specimen. The graphs represent MFI mean values with standard error. The statistical significance based on the theoretical mean of each group is represented in the figure (NS – non-significant, and * P<0.05).

The mt-DAMP molecules displayed considerable alterations at the protein level compared to the control. The fluorescence intensity (MFI) results were normalized to 100 cells and the MFI/100 cells for the protein expression of mtTFA was significantly increased in the graft (P=0.0227) and LAD (P=0.0083) whereas the increase was not statistically significant in LAD-AA (P=0.0525) and LIMA (P=0.2216) with respect to the SEV control (Figs. 5C, 5D). The protein expression of Cyt-C was significantly higher in the SEV control compared to LAD (P=0.0093), LAD-AA (P=0.0355), graft (P=0.0473) and LIMA (P=0.0126) (Figs. 6A, 6B). Similar level of protein expression of Hsp60 was observed in all four tissues (P=0.1758, P=0.1390, P=0.4168, and P=0.7967 for LAD, LAD-AA, graft and LIMA, respectively) with respect to the SEV control (Figs. 6A, 6C). The protein expression of mt-DAMPs was distributed uniformly throughout LAD, LAD-AA, graft and LIMA when compared to the control (Figs. 5C, 6A).

Fig. 6:

Immunofluorescence analysis for the expression of Cyt-C and Hsp60 (panel A) showing the altered expression in the vessel groups. Images in the top row are histological sections of LAD, followed by LAD-AA (second row) and graft tissue (third row) and the images in the bottom rows are histological sections of LIMA. Images in the left column show nuclear staining with DAPI; the images in the middle column show expression of Cyt-C/Hsp60 while the images in the right column show overlay of Cyt-C/Hsp60 staining with DAPI. Images were acquired at 20x. The images in (B) and (C) panels show the quantification of gene expression. The intensity of gene expression obtained through immunofluorescence was acquired and the mean fluorescence intensity normalized to 100 cells (MFI/100 cells) was quantified from each vessel specimen. The graphs represent MFI mean values with standard error. The statistical significance based on the theoretical mean of each group is represented in the figure (NS – non-significant, * P<0.05 and ** P<0.01).

Mitochondrial superoxide

The oxidative environment resulting due to hypoxia in the SMCs was determined by assessing the mitochondrial superoxide using MitoSox Red assay. The hypoxic SMCs exhibited significantly drastic increase (2282.91% increase, P<0.0001 for MFI) in mitochondrial superoxide when compared to the SMCs grown under normoxic conditions. The level of mitochondrial superoxide was negligible in the normoxic SMCs (Figs. 7A, 7B).

Fig. 7:

(A) Determination of mitochondrial superoxide using MitoSox showing increased superoxide in hypoxic SMCs compared with normoxic control. Images in the left column show nuclear staining with DAPI; the images in the middle column show expression of superoxide while the images in the right column show overlay of MitoSox with DAPI. Images were acquired at 20x magnification using CCD camera attached to the Olympus microscope. (B) The image shows quantification of the mitochondrial superoxide and the mean fluorescence intensity (MFI) was normalized to 100 cells. The graphs represent the values of mean MFI/100 cells with standard error. The statistical significance of each hypoxic SMCs vs normoxic SMCs are represented in the figure (*** P<0.001). (C) Determination of mitochondrial pore transition using Image-iT LIVE mitochondrial transition pore assay Kit in hypoxic vs normoxic SMCs. Images in the left column show nuclear staining with DAPI; the images in the middle columns show Calcein production and mitochondrial content while the images in the right column show overlay with DAPI. Images were acquired at 20x magnification using CCD camera attached to the Olympus microscope. (D) The image shows quantification of the mitochondrial superoxide and the mean fluorescence intensity (MFI) was normalized to 100 cells. The graphs represent the values of mean MFI/100 cells with standard error. The statistical significance of each hypoxic SMCs vs normoxic SMCs are represented in the figure (n=4; NS – non-significant, * P<0.05, and **** P<0.0001).

Mitochondrial membrane integrity

Hypoxia induced significant alterations in the mitochondrial membrane integrity of SMCs. The MFI corresponding to the MitoTracker Red was employed to assess the mitochondrial density which was significantly higher (73.56% increase, P<0.0001 for MFI) in the normoxic SMCs when compared with the hypoxic cells suggesting decreased density of mitochondria in hypoxic cells. However, the ionomycin-treated controls showed significant decrease (89.37% decrease, P=0.0164 for MFI) in the MFI for MitoTracker Red in the normoxic SMCs than the hypoxic cells (Figs. 7C, 7D). The intensity of the calcein fluorescence was significantly increased (89.30% increase, P=0.0032) in normoxic SMCs than the hypoxic cells. The ionomycin-treated cells displayed 42.47% increase in calcein level in the normoxic SMCs; however, was statistically not significant (P=0.0879) (Figs. 7C, 7D). Since the healthy mitochondria are impermeable to CoCl₂ present in the reagent to quench the calcein fluorescence, the decreased MFI for calcein in the hypoxic SMCs signifies the compromised mitochondrial membrane integrity.

Effect of hypoxia on mitochondrial mediators and mt-DAMPs in SMCs

The protein expression of mitochondrial mediators and mt-DAMPs displayed considerable alterations in the isolated SMCs grown under hypoxia (Figs. 8A, 8B, 8C). Significant upregulation of PGC1α (201.88% increase, P<0.0001 for MFI), citrate synthase (201.00% increase, P=0.0001 for MFI), mt-PDH (639.98% increase, P<0.0001 for MFI), HIF-1α (811.50% increase, P<0.0001 for MFI), Hsp60 (387.45% increase, P<0.0001 for MFI), Cyt-C (342.96% increase, P<0.0001 for MFI) and mt-TFA (303.26% increase, P=0.0001 for MFI) were observed in hypoxic cells when compared to normoxic SMCs. The MFI for mitochondrial protein complex-1 was decreased in normoxic cells than the hypoxic SMCs (11.94% decrease, P=0.4730 for MFI); however, was statistically not significant (Figs. 8A, 8B, 8C).

Fig. 8:

Immunofluorescence analysis for the protein expression of mitochondrial mediators and mt-DAMPs by SMCs cultured under normoxic (panel A) and hypoxic (panel B) conditions: Images in the top row show nuclear staining with DAPI; the images in the middle row show expression of mitochondrial biomarkers and the images in the bottom row show overlay of mitochondrial biomarkers with DAPI. Images were acquired at 20x magnification using CCD camera attached to the Olympus microscope. (C) The image shows quantification of the mitochondrial superoxide and the mean fluorescence intensity (MFI) was normalized to 100 cells. The graphs represent the values of mean MFI/100 cells with standard error. The statistical significance of each hypoxic SMCs vs normoxic SMCs are represented in the figure (NS – non-significant, *** P<0.01 and **** P<0.0001).

DISCUSSION

The effectiveness of surgical revascularization following CABG procedure is diminished due to increased incidence of vein graft failure¹⁵. Even though the poor patency and increased rate of graft failure are significant, the research findings unveiling the underlying molecular pathology of graft failure has not been extensively studied. In addition, the patients with prior CABG history are at a higher risk for future cardiac events and the alterations in their dietary regimens has significant implications in sustaining the patency of the graft¹⁵. The lack of classical definition in the pathology of the graft failure and the unavailability of the pathological status of the vessels adjacent to the distal and proximal anastomoses hurdle the elucidation of the molecular pathogenesis underlying CABG graft failure. The focus of this pilot study was to understand the association of the major protein-mt-DAMPs such as mt-TFA, Hsp60 and Cyt-C in the pathogenesis of vein graft failure and the involvement of the vessels adjacent to proximal (LIMA) and distal (LAD) anastomoses in the high-calorie fed Yucatan microswine CABG model.

The interposition of vein grafts to the arterial system manifest the venous wall to undergo mechanical stress resulting in a series of biological events leading to arterialization of the graft¹⁶. The histological alterations including intimal hyperplasia, venous wall thickening, SMC migration and fibrous tissue deposition in the intimal layer are prevalent in graft failure¹⁷. In our study, the graft was completely occluded and covered by fibroadipose adhesions and microscopically it displayed severe ECM disorganization with extensive collagen deposition and minimal muscle infiltration as displayed by the pentachrome staining. The occlusion and graft fibroadipose adhesions offered serious hurdles to trace the graft post-operatively. In addition, we were interested to assess the grade of immunopathology and the implications of graft failure in the vessels associated with the anastomoses (LAD, LAD-AA and LIMA). Freshly harvested SEVs (from same anatomical location following the same surgical procedures) from atherosclerotic pigs were used as the control for comparison as it represents a viable/healthy vein graft. Moreover, the SEV control serves to understand the pathological features and biomarker expression underlying the arterialization of vein grafts. The decreased cell density and the elongated nuclear morphology in the graft suggest the presence of mature fibroblasts which represent the terminal phase of fibrosis/scaring¹⁸. On the other hand, the associated vessels (LAD, LAD-AA and LIMA) retained their histomorphology; however, displayed the pathological features mentioned above¹⁷. These observations suggest that the excessive fibrosis marks the terminal stage pathology for graft occlusion, and we speculate that the sustenance of the classical pathology of atheroma formation and the alterations in the mechanical properties of the venous wall result in permanent fibrosis. Also, the pathological features exhibited by the associated vessels (LAD, LAD-AA and LIMA) have implications in accelerating the fibrotic changes in the luminal side of the graft which could be aggravated by the altered hemodynamics of the vessel system.

The graft vessel undergoes several episodes of severe hypoxic insults during pre- and peri-surgical procedures prior to the revascularization. And, the transient loss of perfusion, sustained hypoxia and subsequent ischemic episodes exacerbate the inflammatory events in the vein graft¹⁹. Hence, we assessed the status of hypoxia by determining the expression of the major hypoxic mediator, HIF-1α, in the tissue specimen and investigated the effect of acute hypoxia under cultured SMCs as SMCs are intimately associated with intimal hyperplasia and graft occlusion⁸. Generally, the level of HIF-1α reflects the hypoxic status of the vessel and the assessment of co-expression of other hypoxic biomarkers is relevant and warrants further research. Hence, the present study relied on HIF-1α and SMCs. In addition, the increased level of various DAMP molecules following the hypoxic insult results in the exacerbation of sterile inflammatory pathways in several tissue types²⁰. The cells of the vascular tissues respond to hypoxia via HIF-1α that upregulates growth factors including TGF-β which in turn triggers collagen synthesis and deposition²¹. However, the sustenance of pro-inflammatory signals impairs the TGF-β signaling resulting in the continued deposition of collagen in the luminal side of the graft vessel¹⁹. The decreased level of HIF-1α in the graft and the associated vessels suggests that vessel tissues were recovered/adapted to the hypoxic episodes as the increased expression of HIF-1α was evident in the cultured hypoxic-SMCs. In addition, the study assessed the total HIF-1α; the implications of nuclear fraction and the ratio of nuclear to cytoplasmic fractions of HIF-1α in graft failure warrant further investigation. Also, potential for the persistence of the pathological triggers (nuclear, cytoplasmic and/or both) initiated as a result of hypoxia is extremely high where the identification and signaling of such triggers warrant further investigation However, the increased expression of HIF-1α in the SEV control is tightly associated with the persistence of atherosclerosis²².

The level of the mitochondrial biogenesis marker PGC1α has been reported to be decreased during the maturation phase of fibrosis²³ suggesting that the significant drop of PGC1α in the graft tissue reflects the maturation of the fibrotic tissue. However, the PGC1α was mainly confined to the intimal layer in the LAD, LAD-AA and LIMA and was minimal in the medial layer. Also, PGC1α signals the proliferative functions in the endothelial cells and inhibits ROS-mediated migration of SMCs in the medial layer; however, sustained ROS and oxidative stress inhibit the cellular level of PGC1α²⁴. These findings suggest that the decreased level of PGC1α in SMCs could be a major trigger in the initiation of hyperplasia. Also, the increased PGC1α expression in the hypoxic SMCs with drastic ROS content suggests that the level of PGC1α expression triggered by hypoxia may be insufficient to elicit an anti-atherogenic response. Furthermore, the decreased levels of PGC1α in the graft and associated tissues suggest the increased inflammatory status of these tissues which in turn could be due to the cumulative effects of vein graft pathology and atherosclerosis. However, the SEV control is an intact/healthy tissue despite the overall pro-atherogenic milieu resulting in the increased PGC1α compared to the graft and associated vessels²⁵. In addition, the molecular signaling and the pathological significance of PGC1α in the graft failure are largely unknown which warrant further investigation.

Transient hypoxia upregulates the mitochondrial biomarkers via PGC1α signaling; however, the persistent hypoxia downregulates PGC1α and subsequently affects mitochondrial function²⁶. The non-significant alterations in the level of mitochondrial proteins including citrate synthase, complex-1 and mt-PDH in the graft and the associated vessels suggest similarities in the status of mitochondrial health which reflects either the persistence of hypoxic episodes (chronic hypoxia) or the tissue recovery following the hypoxic insults. Also, the chronic tissue hypoxia due to persistent hyperlipidemia resulting from high calorie diet in our experimental animals results in the sustained activation of pro-inflammatory milieu which aggravates the pathology²⁷. However, the cultured hypoxic SMCs drastically increased the expression of these mitochondrial biomarkers reflecting the initial adaptions elicited by the cells to withstand the hypoxic insult by increasing the mitochondrial activity¹⁴. Interestingly, the decreased signal of the MitoTracker Red in the hypoxic cells reflected the compromised mitochondrial integrity owing to the increased oxidative stress as reflected by the drastic increase in mitochondrial superoxide. Importantly, the hypoxic insult and subsequent oxidative stress impair the overall mitochondrial health, dynamics, and calcium homeostasis and promotes apoptosis^28,29. Taken together, these findings suggest the persistence of mitochondrial pathology following hypoxic transient/chronic insults which in turn trigger inflammatory cascades via the generation and release of mt-DAMPs due to the compromised mitochondrial membrane integrity ultimately leading to graft failure³⁰. Interestingly, the upregulation of PGC1α in the hypoxic SMCs did not correlate with the mitochondrial density as revealed by the MitoTracker Red. The hypoxia-triggered ROS generation and subsequent upregulation of HIF-1α signaling hurdle the mitochondrial biogenesis as the increased ROS content stabilizes HIF-1α which in turn inhibits the activity of PGC1α³¹. In addition, the molecular signaling underlying the hypoxia-mt-DAMPs axis in the pathology of vein graft failure warrants further investigation.

The endosymbiotic origin of mitochondria reflects the possibilities of mitochondrial biomolecules to act as DAMPs which trigger innate immune responses³². The mitochondrial dysfunction disrupts the mitochondrial membrane integrity leading to the release of the contents; several of which including mt-DNA, ATP, Cyt-C, Hsp60 and mt-TFA act as potential DAMPs to activate multiple sterile inflammatory pathways³³. Since hypoxia is the major trigger for mitochondrial dysfunction, it is reasonable to correlate the hypoxia associated with the vessels involved in the CABG procedure and the pathology of the graft failure. The increased level of the mt-DAMPs such as Hsp60 and mt-TFA in the graft and associated tissues and in the cultured hypoxic-SMCs signify the association of the mt-DAMPs with the progression of vascular pathology; atherogenesis and graft fibrosis/failure in our case. To our knowledge, this is the first report studying the association of mt-DAMPs in the pathogenesis of CABG graft failure; however, the mt-DAMP-mediated molecular signaling and the underlying pathological mechanism warrant further research.

The Cyt-C is an electron carrier hemoprotein located at the inner mitochondrial membrane, where the extra mitochondrial Cyt-C acts as mt-DAMP resulting in apoptosis and triggers inflammation³⁴. Interestingly, Cyt-C exhibits strong association in the pro-atherogenic phase of CVDs which signifies increased expression in the SEV control³⁵. We and others reported the sterile inflammatory potential of the nuclear DAMP protein, the HMGB1 (high mobility group box-1), in the pathogenesis of multiple clinical conditions¹¹. Interestingly, mt-TFA is the structural and functional homologue of HMGB1, associated with mt-DNA, which is released acutely from damaged cells to the cytosol and to ECM where it acts as a potential mt-DAMP and exhibits immune response similar to HMGB1³⁶. Hsp60 functions as a mitochondrial chaperone; however, the extracellular Hsp60 facilitates immune response and has been considered as a strong mt-DAMP³⁷. In general, the mt-DAMPs trigger the sterile inflammatory signaling via TLR4 receptors leading to the activation of NLRP3 inflammasome resulting in the upregulation of the pro-inflammatory cytokines IL-1β and IL-18³². However, the information regarding the exact signaling and regulatory mechanisms associated with mt-DAMPs especially in the pathology of CABG graft failure are not available which warrants further extensive investigations.

Strengths and weaknesses

Overall, in this pilot study we successfully established a CABG model in high cholesterol high fat diet-fed atherosclerotic Yucatan microswine to understand the molecular pathology underlying the graft failure and to investigate the pathological status of the arteries associated with the graft especially LAD, LAD-AA and the LIMA. To our knowledge, this is the first study to report the association of mt-DAMPs in the pathogenesis of graft failure that correlated with the effect of hypoxia in aggravating the pathological features. However, the study warrants further investigation to unveil the molecular signaling of mt-DAMPs in the pathogenesis of graft failure which could reveal multiple molecular targets to open novel translational avenues in the management of graft failure. The major limitation of the study is the absence of no-treatment control animals/specimen for comparison as this pilot study intended to minimize the number of animal use and relied on the SEV controls harvested from atherosclerotic microswine. Importantly, the SEV controls were aimed to simulate the vein graft employed clinically for CABG as the patients display the pathology of vascular atheroma. Usually, the clinical vein graft displays a pro-atherogenic microenvironment as most patients present the history of hyperlipidemia. Hence, the SEV control used form the atherosclerotic pigs simulates the vascular status of the clinically used vein grafts. The tracing of the entire graft post-sacrifice was challenging due to the excessive fibroadipose tissue adhesions. Also, the relatively smaller size of the affected blood vessels prevented the analysis of mRNA transcripts and Western blot analysis which made us to rely mainly on histology/immunofluorescence to examine the protein expression. In addition, the study investigated only three major protein-mt-DAMPs, including Cyt-C, Hsp60 and mt-TFA in the tissue specimen owing to the unavailability of suitable/specific antibodies; however, the influence of other mt-DAMPs such as mt-DNA, succinate, N-formyl peptide, and ATP needs to be studied in detail. Furthermore, the study focused to establish the association between mt-DAMPs, hypoxia and CABG graft failure; however, the underlying mechanisms require further investigations both in vivo and in vitro using appropriate pharmacological agents and other approaches. Also, the limited sample population resulted in increased variability among the experimental groups which hurdled to define the expression status of mt-DAMPs in the graft and associated vessels.

The findings from our study showed increased level of mt-DAMPs in the graft and associated tissues, and the in vivo findings correlated with hypoxia using cultured SMCs. The increased ROS level and compromised mitochondrial membrane integrity in the hypoxia challenged SMCs (as observed in our experiments) facilitate the extracellular release of these mt-DAMPs which in turn aggravate the pathology of graft failure. Understanding of the pathological signaling elicited by mt-DAMPs would reveal immense translational potential for the development of management/diagnosis strategies for the CABG graft failure.