Coronary artery bypass grafting (CABG) is still one of the most common surgical procedures aimed at improving blood circulation in atherosclerotic coronary arteries [1]. The predominant reason for early acute coronary syndromes is graft failure. Thus, better understanding of the activated molecular mechanisms occurring soon after surgery may further allow a reduction of the rate of these adverse cardiovascular events. There are a number of blood vessels which can be used as grafts in this revascularization procedure. Nevertheless, the internal thoracic artery (ITA) and the saphenous vein (SV) are the most commonly applied as aortocoronary conduits.
The early occlusion of ITA grafts is observed in approximately 5% of patients, whereas SV transplants occlude within a one year period after surgery in 10–15% of CABG patients [2]. Thus, venous conduits are commonly used in older patients. Despite advances in perioperative management and better understanding of the vessel wall histological architecture, current knowledge regarding molecular pathways’ activation and their possible mechanisms in the perioperative period of CABG procedure is highly limited. There are several molecular pathways indicating increased inflammatory status, haemostasis activation, as well as increased oxidative stress, which seem to play a pivotal role in graft patency. Surgical stress accounts for more protracted and marked molecular pathway perturbations overall. Our previous molecular studies on ITA and SV grafts have shown that there are significant discrepancies between these vessels in the level of expression of genes involved in osteogenic changes in these vessels. According to our findings, we have suggested that venous grafts may be more exposed to atherosclerotic changes [3].
Bettering understanding of the activated molecular mechanisms occurring in vessels soon after surgery may allow a further reduction of the rate of adverse cardiovascular events. Therefore, our study aimed to investigate the transcriptomic profile of genes characterizing both ITA and SV. Employing the microarray approach, we analyzed differences at the molecular level between both blood vessels that often serve as aortocoronary grafts. We decided to additionally describe “cellular response to interferon-gamma”, “inflammatory response”, “interferon-gamma-mediated signaling pathway”, “response to interferon-gamma” and “positive regulation of inflammatory response” gene ontology biological process (GO BP) terms, because inflammatory effects can affect the usefulness of all the blood vessels used as conduits in the CABG procedure.
In most patients, the left ITA was used to bypass the left anterior descending coronary artery (LAD). The other target coronary arteries were usually revascularized with SV grafts.
All surgeries were performed through median sternotomy. SV grafts were obtained through a full-length thigh incision over its course [4]. Pivotal points of the procedure included minimal manipulation of the graft (“no-touch” technique), avoiding extensive dilation of the conduits, using low-intensity electrocautery and the control of the branches with stainless-steel vascular clips. In all cases, the distal part of the obtained SV segment (at least 15–20 mm in length) was saved for further laboratory studies.
ITA conduits were harvested as pedicled, together with satellite veins and endothoracic fascia from the 2nd to 6th intercostal space. The distal end of the ITA segment was divided at the level of its bifurcation. After heparinization, ITA conduits were clipped distally, injected with 10 mL of a papaverine solution (1 mg/mL), and allowed to pharmacologically dilate. Immediately before anastomosis of the distal end of ITA to the recipient coronary artery, a 10-mm segment of the conduit was harvested for further molecular and histological tests.
The sets of the vessel samples, both SV and ITA, were immediately snap-frozen in liquid nitrogen and stored at −80 °C until RNA isolation. Another set of samples was directed for histochemical examination. Transcriptome screening analysis was performed on 18 SV and 20 ITA samples.
Our experiment employed 38 GeneChip® HGU219 (Affymetrix, Santa Clara, CA, USA) microarrays to simultaneously examine thousands of transcripts for each of the analyzed samples. In the first step, the total RNA (500 ng) from each pooled sample was subjected to two rounds of sense cDNA amplification (Ambion® WT Expression Kit, provided by Ambion, Austin, TX, USA). The synthesis of cRNA was performed by in vitro transcription (16 h, 40 °C). Then, cRNA was purified and re-transcribed into cDNA. Subsequently, cDNA samples were used for biotin labeling and fragmentation using an Affymetrix GeneChip® WT Terminal Labeling and Hybridization kit (Affymetrix). Next, the biotin-labeled samples were loaded onto and hybridized to the Affymetrix® Human Genome U219 Array Strip. Hybridization was conducted at 48 °C for 20 h, employing an AccuBlock™ Digital Dry Bath (Labnet International, Inc., Edison, NJ, USA) hybridization oven. Then, microarrays were washed and stained, according to technical protocol, using an Affymetrix GeneAtlas™ Fluidics Station (Affymetrix, Santa Clara, CA, USA). The strips were scanned using an Affymetrix GeneAtlas™ Imaging Station (Affymetrix, Santa Clara, CA, USA). The scans of the microarrays were saved on hard drives as *.CEL files for downstream data analysis.
Quality control (QC) studies were performed using the Affymetrix GeneAtlas™ Instrument Control Software 2.0.0.460 (Affymetrix, Santa Clara, CA, USA), according to the manufacturer’s standards. The generated *.CEL files were subjected to further analysis performed using the R statistical language and Bioconductor package with the relevant Bioconductor libraries. To correct the background, normalize, and summarize the results, we used the robust multiarray averaging (RMA) algorithm. Assigned biological annotations were obtained from the “pd.ragene.2.1.st” library and employed for the mapping of normalized gene expression values with their symbols, gene names, and Entrez IDs, allowing generation of a complex gene data table. To determine the statistical significance of the analyzed genes, moderated t-statistics from the empirical Bayes method were performed. The obtained p-values were corrected for multiple comparisons using Benjamini and Hochberg’s false discovery rate and described as adjusted p-values. The selection of significantly altered genes was based on a p-value beneath 0.05 and an expression higher than two-fold. The differentially expressed gene list (separated for upregulated and downregulated genes) was uploaded to the DAVID Bioinformatics Resources 6.8 software (Database for Annotation, Visualization and Integrated Discovery) [5], where the significantly upregulated Gene Ontology (GO) terms were extracted. The selection of significantly altered GO terms was based on a p-value (Benajamini) < 0.05 and the volume of at least five genes.
To further investigate the chosen gene sets, we investigated their mutual relations with the GOplot package [6]. Subsequently, sets of differentially expressed genes from selected GO BP terms were applied to the STRING10 software (Search Tool for the Retrieval of Interacting Genes/Proteins) for interactions prediction. STRING is a huge database containing information on protein/gene interactions, including experimental data, computational prediction methods, and public text collections.
The research related to human use has been complied with all the relevant national regulations, institutional policies and in accordance the tenets of the Helsinki Declaration, and has been approved by the authors’ institutional review board or equivalent committee. Bioethical Committee approval no. 1201/08, approved on 18/12/2008.
We used Human Genome U219 Array Strip for the microarray gene expression analysis of internal thoracic artery (ITA) and the saphenous vein (SV). This method allowed us to study the gene expression of 49,308 transcripts. We selected genes with more than 2- fold changes and corrected p-values less than 0.05 for downstream analysis. A total of 1170 differentially expressed genes (DEGs) were identified according to the above criteria. We started the microarray gene expression analysis with subjecting the list of DEGs to DAVID software, which showed that the genes can be assigned to many gene ontology groups (GO BP terms). This paper focused on the genes involved in inflammatory response. The DAVID software indicated the following GO BP terms, which cover the above processes: “cellular response to interferon-gamma”, “inflammatory response”, “interferon-gamma-mediated signaling pathway”, “response to interferon-gamma” and “positive regulation of inflammatory response”. The 44 genes involved in those processes were clustered using hierarchical clustering and presented as heatmaps (
It is worth mentioning that 29 genes were downregulated while 15 genes were upregulated. The 10 most significantly upregulated and downregulated genes, their symbols, fold changes and corrected p-values are shown in
The 10 most significantly upregulated and 10 most significantly downregulated genes involved in inflammatory response
Gene symbol | Gene name | Fold change | Adj. p.val |
---|---|---|---|
TNFRSF11B | tumor necrosis factor receptor superfamily, member 11b | 4.74 | <0.01 |
CCL4L1 | chemokine (C-C motif) ligand 4—like 1 | 4.53 | <0.01 |
SELE | selectin E | 3.76 | 0.01 |
PTGER3 | prostaglandin E receptor 3 (subtype EP3) | 3.53 | <0.01 |
CCL8 | chemokine (C-C motif) ligand 8 | 3.23 | <0.01 |
JAK2 | Janus kinase 2 | 3.18 | <0.01 |
TAC1 | tachykinin, precursor 1 | 3.17 | <0.01 |
S100A9 | S100 calcium binding protein A9 | 3.15 | 0.02 |
SPP1 | secreted phosphoprotein 1 | 2.97 | <0.01 |
TGM2 | transglutaminase 2 (C polypeptide, protein-glutamine-gamma-glutamyltransferase) | 2.58 | <0.01 |
CCL14 | chemokine (C-C motif) ligand 14 | -2.43 | <0.01 |
BMP6 | bone morphogenetic protein 6 | -2.45 | <0.01 |
CFH | complement factor H | -2.63 | <0.01 |
CD44 | CD44 molecule (Indian blood group) | -2.73 | <0.01 |
C7 | complement component 7 | -2.78 | <0.01 |
TRIL | TLR4 interactor with leucine-rich repeats | -2.95 | <0.01 |
PDE2A | phosphodiesterase 2A, cGMP-stimulated | -2.98 | <0.01 |
NLRC5 | NLR family, CARD domain containing 5 | -3.16 | <0.01 |
CNR1 | cannabinoid receptor 1 (brain) | -3.27 | <0.01 |
P2RX1 | purinergic receptor P2X, ligand-gated ion channel, 1 | -6.48 | <0.01 |
In the next part of analysis, we focused on the z-scores, which tell us whether the biological process is more likely to be decreased (negative value) or increased (positive value). The z-scores were presented as bar plot (
In the next section, we checked the interaction between selected ontological groups. One of the most visually appealing way of presenting such interaction is dendrogram (
In the gene ontology database, single genes may belong to many ontological terms. For this reason, we used plots with visualization of logFC values and relationship between genes and selected GO BP terms (
In the next part of analysis, we focused on the interaction between proteins encoded by DEGs belonging to studied GO BP terms. Firstly, we used STRING software for the interaction prediction (
Finally, we used ReactomeFIViz app for investigation of functional interactions between proteins encoded by DEGs belonging to selected GO BP terms. Among the most significantly enriched functional interaction networks were FI networks for “Immune response” and “Interferon-gamma-mediated signaling pathway” (
While CABG has been performed for decades, questions remain regarding the detection and characterization of potential markers that will be valuable to evaluate the conduits patency used in this procedure. The predominant cause of early acute coronary syndromes (ACS) is graft failure. Thus, better understanding of the activated molecular mechanisms occurring soon after surgery may allow a further reduction of the rate of these adverse cardiovascular events. The importance of inflammation and inflammatory pathways in atherosclerotic disease and ACS is well established. Nevertheless, not all molecular mechanisms underlying the inflammatory process appear to be clear. Thus, the present study aimed to compare the level of expression of genes involved in inflammatory processes in both ITA and SV conduits and describe potential molecular factors for the evaluation of ITA and SV segment quality.
Employing the expressive microarray approach to analyze the transcriptome of both venous and arterial grafts, five GO BP terms have been selected: “cellular response to interferon-gamma”, “inflammatory response”, “interferon-gamma-mediated signaling pathway”, “response to interferon-gamma” and “positive regulation of inflammatory response”. Among all selected ontological groups genes involved in the formation and maintenance of inflammatory processes can be distinguished. Overall, the genes presented on heatmaps showed differential expression patterns in both analyzed vessels.
Our results indicate the highest fold change of TNFRSF11B transcript levels from all differentially expressed genes analyzed in this study. Our data indicates higher TNFRSF11B transcript expression levels in ITA. Tumor necrosis factor receptor superfamily member 11b (TNFRSF11B, other name: osteoprotegerin (OPG)) is a glycoprotein that acts as a cytokine of the tumor necrosis factor (TNF) [7]. In mice model without expression of OPG authors have shown increased calcification of the aortic media, particularly when given a high dose of phosphate or vitamin D3. These arteries are also the sites of endogenous OPG expression in normal arteries, raising the possibility of a protective role of OPG. This protective role of OPG is particularly important for preventing vascular calcification occurring secondary to administration of warfarin and vitamin D3 [8]. A potential protective effect of OPG against dangerous rupture in advanced abdominal aortic calcification development has been described [9]. Interestingly, in human studies, a significant association between levels of circulating OPG and vascular calcification has been shown. The presence of abdominal aortic calcification, a known risk factor in the development of abdominal aortic aneurysms, was weakly associated with the progression of abdominal aortic aneurysms [10]. Other studies suggested that a pathological increase of serum
Transcript levels of C-C chemokine ligand 4-like 1 (
It is also worth noting that our results indicate a very significant participation of the human leukocyte antigen (HLA) gene complex encoding the major histocompatibility complex (MHC) proteins. We found 4 of the HLA Class I members (
The most significantly downregulated genes was purinergic receptor P2X 1 (
Our transcriptomic analysis of two conduits most commonly applied in coronary artery bypass grafting procedure, the internal thoracic artery (ITA) and the saphenous vein (SV), showed potential molecular markers of formation and maintenance of inflammatory changes. We hope that an analysis of the expression of genes involved in inflammatory response may help to identify patients at high risk of grafts occlusion.