Genetic/epigenetic RNA dysregulation in type 2 diabetes mellitus complicated with ischemic heart disease

Introduction: Diabetes mellitus is a major independent determinant of cardiovascular morbidity. Therefore, we evaluated whether a molecular RNA panel comprising FZD5 and GTF2I could facilitate the early detection and discrimination of ischemic heart disease in individuals with type 2 diabetes mellitus.

Methods: We implemented a two-stage bioinformatics workflow to identify and validate two mRNA candidates associated with T2DM and IHD. Subsequently, we delineated non-coding RNAs linked to these transcripts and the pathways potentially implicated in T2DM complicated by IHD. Finally, we conducted a pilot case–control study and quantified the panel members by RT-qPCR in 56 patients with T2DM, 25 with IHD, 26 with combined T2DM+IHD, and 60 matched controls.

Results: Differential expression analysis showed upregulation of hsa-miR-1976, FZD5, and GTF2I, accompanied by downregulation of LINC02210 in the T2DM+IHD group versus controls. The RNA panel achieved high discriminatory performance (AUC = 0.94) between T2DM+IHD and controls, highlighting its potential as a discriminatory tool.

Discussion: this study identified clinically relevant non-coding RNA–based angiogenesis panel (FZD5, GTF2I mRNAs, hsa-miR-1976 and LINC02210 lncRNA) as a biomarker signature associated with type 2 diabetes mellitus complicated by ischemic heart disease.

1 Introduction

Diabetes mellitus (DM) is a chronic metabolic disorder characterized by sustained hyperglycemia resulting from inadequate insulin secretion and/or impaired insulin sensitivity. Its global burden has risen markedly, establishing DM as a major public health concern (1). Without effective control, DM progressively affects multiple organ systems, with prominent involvement the peripheral nerves, vasculature, and cardiovascular system (2). Within the Middle East and North Africa region, the International Diabetes Federation reports that Egypt carries a substantial diabetes burden: approximately 13.2 million adults are currently affected, with projections reaching 24.7 million by 2050. Egypt is also ranks among the top countries worldwide in both adult prevalence and the absolute number of affected individuals aged 20–79 years (3).

Diabetes mellitus is classified according to underlying the etiological mechanisms that culminate in hyperglycemia. Type 1 diabetes results from immune-mediated destruction of pancreatic β-cells and, although commonly manifesting in childhood or adolescence, may occur at any age; lifelong insulin replacement is required. Type 2 diabetes, the predominant form, arises from insulin resistance with and/or impaired secretion and is strongly associated with obesity; its occurrence in younger age groups has increased in parallel with the global rise in obesity rates. Gestational diabetes mellitus (GDM) is diagnosed during pregnancy and typically resolves after delivery; nevertheless, it confers a substantial long-term risk of developing type 2 diabetes for both the mother and offspring (4).

Prediabetes represents an intermediate state of impaired glucose regulation preceding overt type 2 diabetes, in which glucose values exceed physiological norms but remain below discriminatory thresholds (5).This stage is typically characterized by early β-cell dysfunction and insulin resistance, and accumulating evidence indicates that subclinical complications including neuropathy, nephropathy, retinopathy, and macrovascular alterations may emerge during this stage (6).

In clinical endocrinology, the primary goals are to achieve and maintain optimal glycemic control and to prevent the onset and progression of diabetes-related complications. Accordingly, elucidating the molecular basis of type 2 diabetes is essential for precise target identification and for the rational development and evaluation of mechanism-based precision therapies (7).

Diabetes mellitus is an independent determinant of cardiovascular risk across a broad spectrum of conditions, including cerebrovascular disease, coronary artery disease, and peripheral arterial disease and this burden justifies integrating structured cardiovascular risk stratification within routine diabetes care (8). Patients with diabetes exhibit a markedly increased susceptibility to both macrovascular and microvascular pathologies compared with non-diabetic individuals. In this context, precision medicine has emerged as a transformative paradigm, that enables the tailoring of therapeutic strategies to individual patient profiles with the goal of reducing the incidence and severity of major diabetic complications such as cardiovascular dysfunction, retinopathy, nephropathy, neuropathy, and premature mortality (9).

While lifestyle modification remains foundational, pharmacotherapy is pivotal for controlling hyperglycemia, supporting hepatic function, and mitigating cardiovascular risk (10).

Ischemia results from compromised oxygen supply, diminished nutrient delivery, and impaired clearance of metabolic byproducts. Notably, ischemic manifestations-particularly ischemic heart disease, may precede the formal diagnosis of diabetes mellitus (11). These observations underscore the need for candidate noninvasive biomarkers that enable earlier recognition and refined risk stratification. In routine care, biomarkers support screening, diagnosis, and longitudinal monitoring, and inform the selection of targeted molecular therapies as well as the evaluation of therapeutic response (12).

Insulin resistance is a core lesion in T2DM and denotes attenuated cellular responsiveness to insulin. At the molecular level, defects at canonical signaling nodes—including insulin receptor substrate (IRS) proteins and the PI3K/Akt cascade—are key contributors. Persistent, low-grade inflammation driven by cytokines such as IL-6 and TNF-α disrupts insulin signaling, while mitochondrial dysfunction reduces ATP production and heightens oxidative stress, thereby aggravating resistance. Endoplasmic reticulum stress further impairs insulin action by perturbing protein folding and activating stress-response programs (13). In parallel, epigenetic processes such as DNA methylation and histone modifications reprogram gene-expression profiles that govern insulin sensitivity and β-cell function. Together, these mechanisms illustrate the multifactorial basis of T2DM, integrating genetic susceptibility with environmental and lifestyle factors. Delineating these pathways supports the development of precision-oriented preventive and therapeutic strategies (14).

Disruption of epigenetic regulation is increasingly recognized as a key driver of insulin resistance and the pathogenesis of T2DM. Aberrant epigenetic modifications, often induced by environmental exposures such as dietary patterns and lifestyle behaviors, can remodel chromatin architecture, thereby influencing the accessibility of the transcriptional machinery to target gene loci. These changes may perturb the expression of genes essential for maintaining metabolic homeostasis and insulin sensitivity (15).

MicroRNAs are small, single-stranded non-coding RNAs expressed broadly across tissues. Beyond their canonical role in post-transcriptional repression, some miRNAs can, in defined contexts, enhance gene expression, underscoring their versatile contributions to epigenetic regulation (16). Stable, circulating miRNAs detectable in biofluids have therefore emerged as noninvasive indicators of disease; serum miRNA signatures can mirror tissue-specific pathobiology (17). Recent investigations have demonstrated the utility of miRNA-based assays for the early detection of ischemic heart disease (IHD), highlighting their translational potential in cardiovascular discrimination (18).

Long non-coding RNAs (lncRNAs) are transcripts >200 nucleotides that lack protein-coding capacity. Through interactions with DNA, RNA, and proteins, lncRNAs regulate gene expression at multiple levels spanning epigenetic remodeling, transcriptional control, post-transcriptional processing, and translation (19). At the level of transcription, lncRNAs participate in chromatin reorganization and histone modification, thereby influencing the coordinated activation or repression of defined gene programs. An expanding body of evidence identifies lncRNAs as important epigenetic regulators in the pathogenesis of diabetes and its vascular and metabolic complications. Their contributions to glucose homeostasis and to trajectories of disease progression underscore their promise as discriminatory biomarkers and as candidate therapeutic targets in the management of diabetes (20).

In this study, we applied bioinformatics analyses to delineate the elevated expression of FZD5, hsa-miR-1976 and CRHR1-IT1 associated with type 2 diabetes mellitus complicated by ischemic heart disease, and evaluated whether their serum abundances could serve as noninvasive biomarker panel for early detection.

2 Results

2.1 Bioinformatics results

Differentially expressed genes (DEGs). After standard preprocessing and normalization of the microarray datasets, we identified DEGs in both GSE30122 and GSE19339 using predefined thresholds. In GSE30122, a total of 4, 567 DEGs were detected when comparing of diabetic kidney samples with healthy control kidney samples, including 2, 404 upregulated and 2, 163 downregulated genes (Supplementary Figure S1A). In GSE19339, comparing thrombus leukocytes from acute coronary syndrome (ACS) samples (n = 4) with peripheral blood leukocytes from healthy controls (n = 4) yielded 5, 985 DEGs, comprising 2, 309 significantly upregulated and 3, 676 downregulated genes (Supplementary Figure S1B). When DEGs from GSE30122 and GSE19339 were intersected in a Venn diagram, 1, 683 common genes were identified (Supplementary Figure S1C).

A total of 864 enriched Gene Ontology biological process (GO-BP) terms and 139 Reactome pathways were identified. Functional annotations of the common DEGs were enriched mainly in angiogenesis, hypoxia, platelet degranulation, and cell adhesion. The top eight terms for both GO-BP enrichment are presented in Supplementary Figure S2, according to the order of p value. In addition, three angiogenesis-related GO-BP terms with high protein percentages were among the most significant results. Consequently, the GO-BP analysis was utilized to retrieve the gene sets related to angiogenesis to investigate its role in progression of both diseases (Table 1).

Table 1

Table 1. Angiogenesis related genes.

Following the retrieval of angiogenesis-related gene sets, a PPI network was constructed using the STRING tool (Supplementary Figure S3A). The network comprised 87 nodes and 1, 198 edges and showed highly significant enrichment (PPI enrichment p < 1.0 × 10^-16). We then characterized network topology using the centrality indices betweenness, closeness, and degree for the angiogenesis-related genes. Nodes with degree > 5 were designated as hub genes.

FZD5 and GTF2I were selected for targeted co-regulatory network construction and were validated by Comparative Toxicogenomics Database(CTD) (http://ctdbase.org/) and other databases to be involved in angiogenesis and to be implicated in both acute coronary syndrome and diabetic nephropathy progression (Supplementary Figures S4-S5). has-miR-1976 was found to interact with the selected genes, FZD5 and GTF2I (Supplementary Figures S6) and was strongly linked to acute coronary syndrome and diabetic nephropathy progression (Supplementary Figure S7). LncBase predicted version 3 (DIANA Tools - miRNA-lncRNA interactions (uth.gr) was used to predict interactions between LINC02210 (lncRNAs) and the chosen candidate genes (FZD5, GTF2I), and Clustal Omega multiple-sequence alignment was applied to verify the interaction between hsa-miR-1976 and LINC02210 (https://www.ebi.ac.uk/jdispatcher/msa/clustalo) see in (Supplementary Figure S8). further verification of the lncRNA annotation was performed using Gene card (GeneCards - Human Genes | Gene Database | Gene Search).

2.2 Analysis of biochemical and clinical parameters

The cohort comprised 167 participants allocated to four groups: 60 healthy controls; 25 individuals with IHD; 56 patients meeting ADA criteria for T2DM without cardiovascular disease; and 26 patients meeting ADA criteria for T2DM with cardiovascular disease. Age and sex did not differ significantly across groups (p ≥ 0.05). By contrast, the groups differed significantly in smoking status and family history of T2DM (p < 0.001); in fasting and 2-h postprandial glucose (p < 0.001); in HbA1c and fasting insulin (p < 0.001); in HOMA-IR, HOMA-B, and BMI (p < 0.001); in systolic/diastolic blood pressure, ALT, AST, CK-MB, and troponin (p < 0.001); in the lipid profile—total cholesterol, LDL-C, HDL-C, triglycerides (p < 0.001); and in the urine albumin to creatinine ratio (p < 0.001), as detailed in Table 2.

Table 2

Table 2. Clinical and laboratory characteristics among the groups of the study.

2.3 Evaluation of circulating mRNA, miRNAs, LncRNA in IHD, T2DM without complications, T2DM complicated with IHD patients compared to healthy subjects

We assessed differential expression of the selected RNA panel across study groups using fold-change analysis. Relative to controls, expression of panel members other than LINC02210 including FZD5, GTF2I, and hsa-miR-1976 increased stepwise from controls to T2DM (without complications) and IHD, with the highest levels in T2DM+IHD (p < 0.001). By contrast, LINC02210 showed a progressive decrease from controls → T2DM (without complications) → T2DM+IHD, reaching its lowest abundance in IHD (p < 0.001). Consistent with these trajectories, FZD5, GTF2I, and hsa-miR-1976 were significantly upregulated in IHD, T2DM without complications, and T2DM+IHD versus healthy controls, whereas the overall reduction in LINC02210 across groups did not reach statistical significance (P > 0.05), as summarized in Table 3.

Table 3

Table 3. Descriptives and one-way ANOVA of RNA panel expression among the study groups.

2.4 Assessment of plasma biomarkers in obese and diabetic patients relative to healthy controls

Blood-derived biomarkers provide practical tools for monitoring, diagnosis, and disease staging in T2DM+IHD. In this work, we profiled a panel of biomarkers previously implicated in T2DM+IHD progression. Plasma FZD5 and GTF2I markers linked to cardiovascular pathology were significantly elevated in IHD and T2DM+IHD compared with healthy controls (Figure 1A). Moreover, the T2DM+IHD group showed further elevations in both markers when relative to controls controls, IHD, and T2DM without complications (Figure 1A). Discriminatory performance for separating T2DM from IHD was greater for FZD5 mRNA than for GTF2I mRNA (Figure 1A). LINC02210 levels may reflect adipose-tissue dysfunction relevant to the progression of T2DM+IHD and IHD; in compared with healthy subjects, LINC02210 was significantly reduced in both patient groups (Figure 1B). Conversely, plasma hsa-miR-1976 concentrations were significantly higher in T2DM+IHD, IHD, and T2DM without complications than in healthy controls (Figure 1B).

Figure 1

Figure 1. Relative expression of circulatory RNAs panel among the study groups.

2.5 Discriminatory performance of RNAs panel among the study groups assessed by ROC curve analysis

We evaluated the discriminatory performance of the dysregulated RNA panel using receiver operating characteristic (ROC) analyses across multiple contrasts: diseased vs controls, IHD vs T2DM, IHD vs T2DM+IHD, and T2DM vs T2DM+IHD. For each individual RNA, we derived optimal cutoff values and computed sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and overall accuracy. Comprehensive performance metrics are provided in Table 4 and Figures 2A–H.

Table 4

Table 4. Discriminatory performance of RNAs panel among the study groups assessed by ROC curve analysis.

Figure 2

Figure 2. Discriminatory Performance, (ROC Curve Analysis).

2.5.1 Diseased groups versus controls

Against healthy participants, discriminatory performance yielded AUCs of 0.870 (FZD5 mRNA), 0.940 (GTF2I mRNA), 0.970 (hsa-miR-1976), and 0.819 (LINC02210). The corresponding optimal cutoff values were 1.732, 0.960, 1.774, and 9.10013 for FZD5, GTF2I, hsa-miR-1976, and LINC02210, respectively. Estimated sensitivities were 90.7%, 91.6%, 97.2%, and 82.2%, with specificities of 66.7%, 81.7%, 88.3%, and 61.7%. Collectively, these metrics indicate that the RNA panel can separate patient groups from controls (Table 4, Figures 2A, B).

2.5.2 IHD group versus T2DM

In the IHD vs T2DM comparison, FZD5 mRNA (AUC 0.966) and LINC02210 (AUC 0.978) achieved clear discrimination. The corresponding optimal cutoff values were 4.2156 and 0.2753, yielding sensitivities of 96.0% and 100% and specificities of 76.8% and 96.4%, respectively. By contrast, GTF2I mRNA and hsa-miR-1976 did not discriminate between IHD from T2DM, reflecting lower AUCs and suboptimal operating characteristics (Table 4, Figures 2C, D).

2.5.3 IHD versus T2DM+IHD

The results represent the candidate RNAs panel that did not effectively discriminate IHD cases from T2DM+IHD (Table 4, Figures 2E, F).

2.5.4 T2DM versus T2DM+IHD

We next appraised the performance of the mRNA/miRNA/lncRNA panel for distinguishing T2DM from T2DM+IHD using ROC analysis. The optimal cutoff values were 5.8718 (FZD5 mRNA), 5.290 (GTF2I mRNA), 51.4426 (hsa-miR-1976), and 0.8189 (LINC02210). The corresponding AUCs were 0.986, 0.702, 0.694, and 0.973, respectively. Estimated sensitivities reached 96.2%, 69.2%, 73.1%, and 100%, with specificities of 89.3%, 51.8%, 66.1%, and 85.71%. These findings are concordant with the bioinformatics signal and indicate that the proposed RNA panel may aid discriminatory separation of T2DM+IHD from T2DM (see Table 4, Figures 2G, H).

2.6 Correlation between biomarker positivity rate and clinicopathological factors in disease groups

Among positive values of FZD5 mRNA, GTF2I mRNA, has-miR-1976 miRNA, LINC02210 LncRNA and various clinicopathological factors across different disease groups, our analysis revealed that Hemoglobin A1c (HbA1c, %), Total Cholesterol (mg/dL), triglycerides(mg/dL), CK-MB and Troponin have a significant positive correlation with RNA panel among all diseased groups. On the other hand, sex, ALT(IU/L), BMI (kg/m²), and Age showed no significant correlation with RNA panel among diseased groups (Table 5).

.

2.7 Correlation analysis and linear regression analysis

We examined associations within the RNA panel across study groups using Spearman’s rank correlation. Positive correlations were observed between FZD5 and GTF2I (r = 0.462; p < 0.001), between FZD5 and hsa-miR-1976 (r = 0.632; p < 0.001), and between GTF2I and hsa-miR-1976 (r = 0.545; p < 0.001). In contrast, LINC02210 correlated negatively with FZD5 (r = −0.651; p < 0.001), GTF2I (r = −0.369; p < 0.001), and hsa-miR-1976 (r = −0.456; p < 0.001). Overall, these data indicate significant interrelationships within the RNA network across the analyzed cohorts (Table 6).

.

In the T2DM+IHD subgroup, pairwise associations within the RNA panel were examined using Spearman’s rank correlation. Positive but non-significant correlations were observed for FZD5–hsa-miR-1976 (r = 0.266; p < 0.190), FZD5–LINC02210 (r = 0.265; p < 0.191), GTF2I–LINC02210 (r = 0.225; p < 0.270), and hsa-miR-1976–LINC02210 (r = 0.287; p < 0.155). By contrast, GTF2I showed inverse correlations with FZD5 (r = −0.061; p < 0.769) and with hsa-miR-1976 (r = −0.137; p < 0.505) (Table 7). In the T2DM+IHD subgroup, biomarker-clinical correlations appeared attenuated, likely reflecting multifactorial pathophysiology. However, multivariate models and discriminatory performance remained robust, underscoring their complementary value.

Table 7

Table 7. Correlation analysis in T2DM+IHD group.

A linear regression analysis was used to evaluate the relationships between RNAs levels across all study groups. FZD5 mRNA (p = 0.001), GTF2I mRNA (p < 0.001), LINC02210 (p = 0.049), CK-MB (p < 0.001) and Troponin (p < 0.001) were significant predictor, whereas has-miR-1976 (p = 0.091) was not significant in the combined analysis (Table 8).

Table 8

Table 8. Regression analysis.

GTF2I showed a lower mean Ct in T2DM+IHD (23.5 vs 27.0 in controls), suggesting upregulated expression with intra-assay reproducibility (SD ≤0.27). hsa-miR-1976 showed markedly lower Ct in T2DM+IHD (21.8 vs 29.2 in controls), indicating strong differential expression with slightly higher inter-assay variability (SD = 0.60–0.65), potentially reflecting miRNA stability constraints. Across all targets, technical reproducibility was high with intra-assay CV% <1.4% and inter-assay CV% <1.5% (Table 9). has-miR-1976 exhibited the largest fold-change between groups (ΔCt = 7.4), aligning with its proposed role in metabolic regulation (Supplementary Table S1).

Table 9

Table 9. Intra-Assay and Inter-Assay Variability for Real-Time PCR.

These findings suggest a potential translational value of the proposed RNA panel in clinical practice. When integrated with existing diagnostic markers such as troponin and HbA1c, this panel could enhance early detection and risk stratification of ischemic heart disease in diabetic patients. The combined use of molecular and conventional biomarkers may improve diagnostic sensitivity and specificity, allowing for better patient monitoring and personalized therapeutic strategies.

3 Discussion

Type 2 diabetes mellitus (T2DM) is now regarded not only as a metabolic disorder but also as an independent driver of cardiovascular risk, most notably ischemic heart disease (IHD), in which inadequate myocardial perfusion culminates in tissue injury. Furthermore, mounting evidence indicates that the underlying mechanisms of both T2DM and ischemic heart disease are closely linked through inflammatory processes and oxidative stress, which are exacerbated by mitochondrial dysfunction and cellular apoptotic pathways, as highlighted in the literature on cell-fate regulation (21).

Whether regional adiposity is linked to cardiovascular disease (CVD) risk and mortality in individuals with type 2 diabetes (T2DM) remains largely unclear, despite their characteristic shifts in fat distribution and elevated CVD risk (22).

These links likely reflect a multifactorial interaction between genetic variation and epigenetic regulation that shapes RNA-mediated regulation of gene expression. Growing evidence indicates that disturbances within RNA regulatory networks are central to the pathogenesis of T2DM and its complications. By clarifying how genetic variations and epigenetic modifications affect gene expression, we can better elucidate the molecular mechanisms that drive T2DM and its downstream cardiovascular risks. Our objective was to determine the discriminatory performance of a molecular RNA panel comprising FZD5 and GTF2I for the early identification of ischemic heart disease in individuals with type 2 diabetes mellitus.

Multiple risk loci linked to insulin resistance and lipid metabolism have been reported, and these variants not only increase susceptibility to type 2 diabetes but also heighten vulnerability to cardiovascular outcomes. For example, variants that impair endothelial-cell function can lead to impaired vascular responses, as evidenced by the common pathology of diabetic panvascular disease (DPD), in which macrovascular and microvascular complications often emerge concurrently in individuals with diabetes, suggesting a shared or overlapping pathogenic timeline that may accelerate systemic deterioration (23). In addition, underlying genetic predisposition can amplify endoplasmic-reticulum (ER) stress signaling implicated in T2DM pathobiology, thereby aggravating cellular dysfunction and promoting progression toward ischemic cardiovascular events (24).

We first constructed a regulatory network spanning mRNA/miRNA/lncRNA interactions relevant to crosstalk in T2DM with IHD using computational analyses. We then quantified serum levels of network components in cases and controls to appraise their capacity for early risk stratification and discriminatory assessment (CVD). A substantial subset of the mapped genes was associated with IHD and T2DM. Prior work has shown increased methylation at the FZD5 promoter in T2DM patients and IHD, consistent with reports implicating FZD5 in diabetic vasculopathy (25). Concordantly, our data revealed elevated FZD5 mRNA in patients with T2DM+IHD.

Independent reports indicate that increased methylation of GTF2I is associated with a higher subsequent risk of myocardial infarction and coronary heart disease (26). This aligns with our findings, which showed an elevated GTF2I mRNA in the T2DM+IHD group, suggesting its involvement in the development of IHD among patient with T2DM patients.MicroRNAs have emerged as informative biomarkers for diabetes and its sequelae. Their reliable detection in circulating biofluids has driven extensive investigation into disease-specific expression profiles and molecular stability. In particular, miR-92a, miR-503, and miR-126 modulate angiogenic pathways, processes that are essential for myocardial repair after ischemic injury (11).

These observations accord with our findings, which showed upregulation of hsa-miR-1976 and support its role as a putative epigenetic activator of the FZD5/GTF2I axis. This interpretation is consistent with recent reports that certain miRNAs can engage promoter regions and enhance transcription via RNA-activation (RNAs). To our knowledge, this is the first description linking hsa-miR-1976 to type 2 diabetes complicated by ischemic heart disease.

Multiple reports highlight the central regulatory functions of lncRNAs across the initiation and progression of T2DM with coexisting IHD (20). Crosstalk among these transcripts appears to coordinate gene programs relevant to IHD pathogenesis and positions lncRNAs as candidate biomarkers for early detection and risk prediction in patients with T2DM. Consistently, specific lncRNAs exhibit discriminatory translational potential in diabetes complications, serving as molecular readouts of disease onset, trajectory, and tissue specificity. Supporting this concept, Geng et al.” (2024) reported reduced levels of TINCR and HOTAIR in serum and myocardial tissue from individuals with diabetic complications, which discriminated cases from healthy controls.

Notably, our data indicate that LINC02210 functions as a putative network-associated regulator within the FZD5/GTF2I/hsa-miR-1976 network. To our knowledge, LINC02210 has not been previously linked to type 2 diabetes or ischemic heart disease. In this cohort, circulating LINC02210 levels were lower in T2DM+IHD than in either controls or T2DM alone, and yielded discriminatory decision thresholds capable of separating T2DM+IHD vs controls, T2DM vs IHD and T2DM+IHD vs T2DM.

LINC02210’s inverse correlations with angiogenesis-related genes (FZD5, GTF2I) and discriminatory performance in advanced disease stages (AUC > 0.97) suggest it may modulate vascular remodeling. Ongoing work is testing its direct role in endothelial dysfunction and plaque stability. While LINC02210 demonstrates disease-specific expression patterns, its functional role requires validation in ongoing studies.

The evaluated angiogenesis-linked RNA signature showed group-dependent expression. Levels of FZD5 and GTF2I mRNAs, together with hsa-miR-1976, rose stepwise from controls to T2DM and IHD, with peak abundances observed in the T2DM+IHD cohort. Conversely, LINC02210 displayed a graded decline across the same sequence, reaching its lowest concentration in T2DM+IHD. Taken together, these trajectories support the feasibility of this circulating coding/non-coding RNA panel as an early-detection aid for ischemic heart disease in the context of type 2 diabetes. The weaker correlations in T2DM+IHD highlight the need for nonlinear or pathway-specific analyses in advanced disease, which will be pursued in future work.

Relative to the T2DM+IHD cohort, the T2DM group showed higher hsa-miR-1976 and lower LINC02210 expression. Alongside CK-MB and troponin, these noncoding RNA readouts could assist in distinguishing IHD status among patients with T2DM. This interpretation aligns with Ortiz-Martín et al. (2022), who proposed that serum biomarkers can complement or in some settings substitute for traditional analytes for diabetes detection and follow-up. In our data, FZD5, GTF2I, hsa-miR-1976, and LINC02210 effectively differentiated T2DM from T2DM+IHD, consistent with prior reports identifying ncRNA signatures as candidate predictors of IHD in diabetes (25–27). While our models show strong discriminatory performance, external validation is required to confirm generalizability; We are actively collaborating with independent cohorts to address this limitation. Previously, our group likewise reported discriminatory utility for a panel comprising MEMM173 and CHUK mRNAs together with hsa-miR-611, -5192, and -1976 in diabetes and cardiovascular disease (6).

The RNA panel (AUC = 0.94) outperformed Troponin-I (AUC = 0.78) and HbA1c (AUC = 0.85) in discriminating T2DM-IHD from controls). Integrating RNA biomarkers with troponin/HbA1c may improve early risk stratification for ischemic events in diabetic populations.

Limitations. This study has several limitations that should be considered when interpreting the findings. To minimize bias, we focused on angiogenesis-related genes with established roles in T2DM/IHD pathways and validated qPCR results in triplicate, achieving low technical, variability (CV < 5%). Nevertheless, the pilot nature of the work and the modest sample sizes in the IHD and T2DM+IHD groups may limit precision and generalizability. Although major confounders were adjusted for, residual confounding from unmeasured factors (e.g., dietary habits& drug therapy) may persist; sensitivity analyses supported the robustness of the main signals but cannot fully exclude such effects. We plan to expand this pilot to a larger cohort with orthogonal validation via wider transcript profiling &protein-level assays.

The age cutoff of 35 years was selected to minimize age-related comorbidities and to focus on early molecular changes in T2DM and IHD, in line with regional epidemiology; this strengthens internal validity but constrains extrapolation to older populations. Despite statistical matching on age and sex, the absolute sex ratios reflect real-world clinical demographics and could introduce subtle confounding, motivating sex-stratified designs and covariate-adjusted models in future work.

Because multiple genes were evaluated, a risk of type I error remains despite adjusted analyses; larger, prespecified cohorts with formal multiple-testing control are warranted. The putative regulatory role of LINC02210, inferred from network centrality and correlations with angiogenic markers, requires confirmation in targeted functional experiments. Finally, although the RNA panel shows encouraging case–control discrimination, clinical validity should be assessed in prospective, blinded, longitudinal cohorts with orthogonal transcriptomic and protein-level assays.

In conclusion, we identify a candidate angiogenesis related RNA panel FZD5, GTF2I mRNAs, hsa-miR-1976, and the lncRNA LINC02210 that is associated with T2DM complicated by IHD and shows concordance with serum clinical measures reflecting the transition from T2DM to T2DM+IHD. These associations are correlative and do not establish causality; prospective validation in larger, age-diverse cohorts, alongside functional studies to delineate gene-specific contributions to IHD risk in T2DM, is required.

4 Materials and methods

4.1 Bioinformatics-based construction of the RNA regulatory network

We performed an in silico screen to identify differentially expressed coding and noncoding RNAs relevant to type 2 diabetes mellitus (T2DM) and ischemic heart disease (IHD). Microarray expression datasets were obtained from the NCBI Gene Expression Omnibus (GEO; https://www.ncbi.nlm.nih.gov/geo/).

4.1.1 Acquisition of available datasets

High-throughput microarray datasets for diabetic nephropathy and acute coronary syndrome (ACS) were retrieved from NCBI GEO (https://www.ncbi.nlm.nih.gov/geo/, accessed July 2021) (28). Searches were limited to Homo sapiens and experimental studies comparing patients with diabetic nephropathy or ACS against healthy controls. As a result, two datasets were obtained:GSE30122 (29) and GSE19339 (30), were obtained. The GSE30122 dataset contains 19 diabetic kidney samples and 50 healthy control kidney samples, based on the GPL571 Affymetrix Human Genome U133A 2.0 Array platform. GSE19339 comprises 4 thrombus leukocyte samples from ACS cases and 4 peripheral blood leukocyte samples from healthy controls, generated on the GPL570 Affymetrix Human Genome U133 Plus 2.0 Array platform.

4.1.2 Differential expression analysis

Microarray profiles from GSE30122 and GSE19339 were analyzed using the GEO2R web portal (https://www.ncbi.nlm.nih.gov/geo/geo2r/; accessed July 2021) to identify differentially expressed genes (DEGs) among the groups. GEO2R is an online interface built on the R language limma package (31). Significance thresholds were FDR < 0.05 together with p < 0.05. Probes lacking an assigned gene symbol were excluded prior to downstream analyses of the resulting DEGs. DEGs were identified using |logFC| > 0.5 and p < 0.05, followed by FDR correction (q < 0.1). Functional enrichment required q < 0.05. Thresholds were selected to harmonize statistical rigor with biological plausibility.

4.1.3 Identification of common DEGs

DEGs from both datasets (GSE30122 and GSE19339) were intersected using an online Venn diagram (http://bioinformatics.psb.ugent.be/webtools/Venn/) to obtain the shared gene set. This overlap was considered the set of DEGs implicated in both diabetic nephropathy and acute coronary syndrome progression.

4.1.4 Enrichment analyses of common DEGs

To determine which biological processes (BP) and pathways were overrepresented among the shared DEGs, we performed GO–BP and pathway enrichment using FunRich (http://www.funrich.org/; v3.1.3, accessed Jul 2021) (32). A p-value of <0.05 was considered indicative of enrichment. The biological classification of the common DEGs was subsequently filtered, focusing on the highly significant BP terms associated with angiogenesis.

4.1.5 Protein–protein interaction network analysis

To map potential interactions among proteins encoded by the filtered DEGs and to identify hub nodes, angiogenesis-related DEGs from the enrichment step were queried in STRING (https://string-db.org/; v12, accessed July 2023) (33). Only edges with combined score > 0.15 were retained for network construction. The resulting PPI networks was then visualized using Cytoscape software (version 3.10.2). Topological metrics were then computed with the CentiScaPe app (34) and the degree (number of connections) of each node was calculated; genes with degree >5 were defined as hub genes.

4.1.6 Selection of candidate genes

Biomarkers (mRNAs and miRNAs) were selected through a structured, multi-step integrated bioinformatics pipeline and previous literature validation studies designed to prioritizing relevance to diabetic nephropathy or acute coronary pathogenesis, functional annotations, and prior evidence of differential expression (Supplementary Table S1).

From the hub set, we prioritized FZD5 (Frizzled class receptor 5) and GTF2I (General Transcription Factor II-I) to assemble a targeted co-regulatory network. Support for their relevance derives from prior studies (25–27) and from public resources the Comparative Toxicogenomics Database (http://ctdbase.org/) and Gene Cards (https://www.genecards.org/; accessed October 2024) which annotate these genes as linked to angiogenesis and implicated in acute coronary syndrome and diabetic nephropathy progression. The curated genes were subsequently submitted to STRING to construct the protein–protein interaction (PPI) network.

4.1.7 Prediction of candidate microRNAs

Predicted interactions between miRNAs and the selected candidate genes were generated using miRWalk 3.0 (http://mirwalk.umm.uni-heidelberg.de/). Functional implications of the selected miRNA were then evaluated with DIANA tools miRPath v4 module (http://www.microrna.gr/miRPathv4), which tests enrichment of its targets across defined biological pathways.

4.1.8 Prediction of candidate long noncoding RNAs

LncBase predicted version 3 (DIANA Tools - miRNA-lncRNA interactions (uth.gr) was used to predict interactions between long noncoding RNAs (lncRNAs) and the chosen candidate genes, Additional annotation and verification were obtained from Gene card(GeneCards - Human Genes | Gene Database | Gene Search). We confirm the selected lncRNA from another database (LNCipedia database) (https://ngdc.cncb.ac.cn/databasecommons/database/id/24).

4.2 Participants and study groups

The study enrolled 167 participants distributed into four groups: 56 Patients who fulfilled the American Diabetes Association’s (ADA) T2DM criterion and had no cardiovascular disease, 25 Patients who had a cardiovascular disease only, 26 Patients who fulfilled the American Diabetes Association’s (ADA) T2DM criterion and has cardiovascular disease and 60 Individuals with normal blood glucose levels who have never had diabetes or any kind and Cardiovascular diseases.

The study cases were enrolled from Cardiology and Endocrinology Department Ain Shams University. The study protocol was approved by the Research Ethics Committee, Faculty of Medicine, Ain Shams University (FMASU R 42/2024). Written informed consent was obtained from all participants in accordance with the Declaration of Helsinki after clear explanation of the study aims, procedures, and potential risks. Data confidentiality was maintained throughout to safeguard participant privacy.

Exclusion criteria of the study included patients with other kinds of diabetes mellitus, severe liver dysfunction, acute infections, active neoplasm. pregnancy patients, Breast feeding patients with mental disorder, autoimmune disease, Patients that are uncooperative and refuse to give consent, Patients that are less than 35 years old and Patients that are related to angiogenesis disease such as numerous malignant, inflammatory, infectious and immune disorders.

Venous blood was obtained from all participants. Serum was separated by centrifugation at 4, 000 rpm for 20 min, aliquoted, and stored at −80 °C until analysis. A multifunctional biochemistry analyzer (AU680, Beckman Coulter Inc., Kraemer Blvd., Brea, CA 92821, USA) was used to assess serum lipid profile, liver function tests, CKMB, Troponin, HBA1C, Insulin level, post prandial glucose and fasting glucose. HOMA-IR calculated as (Fasting insulin (μU/L) × fasting glucose (nmol/L)/22.5) (35).

4.3 RNA isolation and cDNA preparation

Total RNA was isolated from serum using the miRNeasy Mini Kit (Qiagen, Hilden, Germany, cat. no. 217084) according to the manufacturer’s instructions. RNA yield and purity were quantified on a Qubit 3.0 Fluorometer (Invitrogen, Life Technologies, Malaysia) with the Qubit™ dsDNA HS Assay Kit and the Qubit™ RNA HS Assay Kit (cat. nos. Q32851 and Q32852, respectively). cDNA was then synthesized from the purified RNA using a Rotor gene Thermal cycler (Thermo Electron Waltham, MA) and the QuantiTect Reverse Transcription Kit for mRNA and lncRNA (Qiagen, Hilden, Germany, cat. no. 205311) and the miRCURY LNA RT Kit (Qiagen, Hilden, Germany, cat. no. 339340) for miRNA in reference to the kit’s protocol.

4.4 Quantitative RT-PCR analysis of target mRNAs/miRNA/lncRNA

Prior studies demonstrating ACTB, GAPDH showed the most stable expression. stable expression in human blood and vascular tissues under metabolic stress (PMID: 38766348, PMID: 37223013). We employed geometric mean normalization (GAPDH + ACTB) to minimize individual gene fluctuations, as recommended for metabolic disease studies (36). Reference gene stability and assay performance are summarized in Supplementary Table S2.

mRNA targets (FZD5, GTF2I) were quantified using gene-specific primer assays in combination with the QuantiTect Multiplex PCR Kit (Qiagen, Hilden, Germany, cat. no. 249900; assay IDs QT00200886 and QT01677305), with GAPDH and ACTB serving as internal references. For miRNA measurements, hsa-miR-1976 was amplified with the miRCURY LNA SYBR Green PCR Kit (Qiagen, cat. no. 339345) and the corresponding assay (Cat. No. 339350; ID: ZP00000388), and expression was normalized to SNORD44. LINC02210 (lncRNA) levels were determined using the RT2 lncRNA qPCR Assay (Qiagen, cat. no. 330701), with GAPDH as the reference control. Thermal cycling conditions were 95 °C for 2 min, followed by 45 cycles of 95 °C for 5 s and 60 °C for 10 s. Relative expression was computed by the Livak method (RQ = 2^−ΔΔCt) (37), and reactions were run on an Applied Biosystems 7500 Fast System (37). All primer assays utilized in this study were sourced from Qiagen, Germany (Supplementary Table S3).

4.5 Statistical analysis

All analyses were conducted in SPSS v29 (IBM, Chicago, USA). Continuous variables are summarized as median [IQR] for non-normally distributed data and mean ± SD for normally distributed data. Normality was examined with the Shapiro–Wilk test. Between-group comparisons used Kruskal–Wallis with Dunn’s post-hoc procedure for nonparametric outcomes, and one-way ANOVA with Tukey’s post-hoc test for parametric outcomes. Demographic characteristics and clinical predictors of T2DM+IHD were evaluated within this framework. Two-sided p < 0.05 was considered statistically significant. Multicollinearity was assessed via correlation matrices. Covariates were selected a priori based on clinical relevance.

4.6 Measures to overcome risks of overfitting

4.6.1 Feature selection rationale

Gene candidates were prioritized through a biology-driven strategy focusing on hypoxia-responsive angiogenesis pathways implicated in T2DM and ischemic heart disease (IHD) pathogenesis. Targets such as FZD5 were selected based on pathway enrichment analyses and prior literature evidence of their roles in endothelial dysfunction (38). Biomarker inclusion criteria required both statistical significance (adjusted p<0.05) and biological relevance (≥2-fold differential expression), ensuring alignment with disease mechanisms while minimizing false discovery.

4.6.2 Experimental design

Technical reproducibility was ensured through triplicate PCR measurements for all samples, achieving coefficient of variation (CV) values <1.5% for cycle threshold (Ct) values (Table 9). Biological replicates were incorporated to account for inter-individual variability inherent in human studies. Statistical analyses employed ANOVA & Kruskal-Wallis tests (for non-normally distributed data) with post-hoc correction to address multiple comparisons. A priori power analysis (α=0.05, β=0.20) confirmed adequate sample size to detect ≥2-fold expression differences, aligning with clinically relevant thresholds in metabolic disease research.

4.6.3 Reproducibility metrics

Stringent quality control included evaluation of intra-assay (within-run) and inter-assay (across-run) variability, with Ct standard deviations maintained at ≤0.33 and ≤0.65, respectively (Table 9). These metrics, combined with primer efficiencies of 90–105% (Supplementary Table S3), met MIQE guidelines for qPCR reliability. The low CV% values (<1.5%) across all targets underscore the technical precision of our experimental workflow.

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

Ethics statement

The studies involving humans were approved by the Ain Shams University, Faculty of Medicine’ research Ethical Committee (FMASU R 42/2024). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

Author contributions

ES: Writing – review & editing, Writing – original draft. MM: Writing – original draft, Writing – review & editing, Funding acquisition, Supervision, Data curation. MS: Writing – review & editing, Writing – original draft. MH: Writing – original draft, Writing – review & editing, Supervision, Investigation, Software.

Funding

The author(s) declare financial support was received for the research and/or publication of this article. This research was funded by Diabetes and Acute coronary syndrome under project number ASERT 2022/2.

Acknowledgments

We acknowledge TASH center (translational and applied science hub) for resources availability.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declare that no Generative AI was used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fendo.2025.1687145/full#supplementary-material

References

1. Liu P, Zhang Z, Cai Y, Li Z, Zhou Q, and Chen Q. Ferroptosis: Mechanisms and role in diabetes mellitus and its complications. Aging research review (2024). doi: 10.1016/j.arr.2024.102201

PubMed Abstract | Crossref Full Text | Google Scholar

2. WHO. (2024). Available online at: https://www.who.int/news-room/fact-sheets/detail/diabetes (Accessed November 2024).

Google Scholar

3. Egypt Diabetes Statistics & Prevalence | IDF Atlas. (2025). Available online at: https://diabetesatlas.org/data-by-location/country/Egypt/ (Accessed April 9, 2025).

Google Scholar

4. American Diabetes Association. Classification and Diagnosis of Diabetes: Standards of Medical Care in Diabetes (2021). Available online at: https://diabetesjournals.org/care/article/44/Supplement_1/S15/30859/2-Classification-and-Diagnosis-of-Diabetes (Accessed March 2025).

Google Scholar

5. Sun H, Saeedi P, Karuranga S, Pinkepank M, Ogurtsova K, Dunca BB, et al. Diabetes Atlas: Global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045. Diabetes Res Clin Pract. (2022) 183:109119. doi: 10.1016/j.diabres.2021.109119

PubMed Abstract | Crossref Full Text | Google Scholar

6. Ali HS, Kamel MM, Agwa SHA, Abdel Hakeem MS, El Meteini MS, and Matboli M. Analysis of mRNA-miRNA lncRNA differential expression in prediabetes/type 2 diabetes mellitus patients as potential players in insulin resistance. Front Endocrinol (Lausanne) (2023) 14:1131171. doi: 10.3389/fendo.2023.1131171

PubMed Abstract | Crossref Full Text | Google Scholar

7. Wei S, Ma F, Feng S, and Ha X. Integrating transcriptomics and proteomics to understand the molecular mechanisms underlying the pathogenesis of type 2 diabetes mellitus. Genomics (2024) 16:110964. doi: 10.1016/j.ygeno.2024110964

PubMed Abstract | Crossref Full Text | Google Scholar

8. Zakir M, Ahuja N, Surksha MA, Sachdev R, Kalariya Y, Nasir M, et al. Cardiovascular complications of diabetes: from microvascular to macrovascular pathways. Cureus (2023) 15:e45835. doi: 10.7759/cureus.45835

PubMed Abstract | Crossref Full Text | Google Scholar

9. Schiborn C and Schulze MB. Precision prognostics for the development of complications in diabetes. Diabetologia (2022) 65:1867–82. doi: 10.1007/s00125-022-05731-4

PubMed Abstract | Crossref Full Text | Google Scholar

10. Michalopoulou E, Thymis J, Lampsas S, Pavlidis G, Katogiannis K, Vlachomitros D, et al. The triad of risk: linking MASLD, cardiovascular disease and type 2 diabetes; from pathophysiology to treatment. Clin Med. (2025) 14:428. doi: 10.3390/jcm14020428

PubMed Abstract | Crossref Full Text | Google Scholar

11. Bielska A, Niemira M, Bauer W, Sidorkiewicz I, Szałkowska A, Skwarska A, et al. Serum miRNA profile in diabetic patients with ischemic heart disease as a promising non-invasive biomarker. Front Endocrinol. (2022) 13:888948. doi: 10.3389/fendo.2022.888948

PubMed Abstract | Crossref Full Text | Google Scholar

12. Sakhuja A, MBBS, MS, FACP, FASN, and FCCP. Novel Biomarkers of Kidney Function: introduction and overview, characteristics of an ideal marker for kidney disease, biomarker of acute kidney injury (2021). Available online at: https://emedicine.medscape.com/article/1925619-overview (Accessed February 2025).

Google Scholar

13. Galicia UN, Benito GA, Larrea VSJA, Siddiqi SH, Kepa B, Ostolaza UH, et al. Pathophysiology of type 2 diabetes mellitus. Int J Mol Sci. (2020) 21:6275. doi: 10.3390/ijms21176275

PubMed Abstract | Crossref Full Text | Google Scholar

14. Jazieh C, Arabi TZ, Belal Z, Sabbah N, Alsaud AW, Alkattan K, et al. Unraveling the epigenetic fabric of type 2 diabetes mellitus:pathogenic mechanisms andtherapeutic implications. Front Endocrinol (Lausanne) (2024) 15:1295967. doi: 10.3389/fendo.2024.1295967

PubMed Abstract | Crossref Full Text | Google Scholar

15. Wang J. The role of the lysine-specific demethylase 1 in the development of hepatocellular carcinoma. (2022).

Google Scholar

16. Yi YS. MicroRNA-mediated epigenetic regulation of inflammasomes in inflammatory responses and immunopathologies. Semin Cell Dev Biol. (2024) 154:227–38. doi: 10.1016/j.semcdb.2022.11.006

PubMed Abstract | Crossref Full Text | Google Scholar

17. Khan IA, Rashid S, Singh N, Rashid S, Singh N, Gunjan D, et al. Panel of serum miRNAs as potential non-invasive biomarkers for pancreatic ductal adenocarcinoma. Sci Rep. (2021) 11:2824. doi: 10.1038/s41598-021-82266-5

PubMed Abstract | Crossref Full Text | Google Scholar

18. Yee Kong. AS, Lai K-S, Lim S-HE, Sivalingam S, Loh JY, and Maran. SMiRNA in ischemic heart disease and its potential as biomarkers: A comprehensive review. Int J Mol Sci. (2022) 23:9001. doi: 10.3390/ijms23169001

PubMed Abstract | Crossref Full Text | Google Scholar

19. Sheykhhasan M, Dermani F, Gheibi N, Datta I, Sadeghi H, and Khoei SG. The emerging role of lncRNA AWPPH in multiple cancers: A review study. Current Molecular Medicine (2023) 25:237–68. doi: 10.2174/1566524023666230816163031

PubMed Abstract | Crossref Full Text | Google Scholar

20. Geng M, Liu W, Li J, Yang G, Tian Y, Jiangand X, et al. LncRNA as a regulator in the development of diabetic complications. Front. Endocrinol (2024) 15:2024. doi: 10.3389/fendo.2024.1324393

PubMed Abstract | Crossref Full Text | Google Scholar

21. Nguyen TT, Wei S, Thu NT, Jo Y, Zhang Y, Park MJ, et al. Mitochondria-associated programmed cell death as a therapeutic target for age-related disease. Exp Mol Med. (2023) 55:1595–619. doi: 10.1038/s12276-023-01046-5

PubMed Abstract | Crossref Full Text | Google Scholar

22. Qiu Z, Hoon Lee D, Lu Q, Li R, Zhu K, Li L, et al. Associations of Regional Body Fat with Risk of Cardiovascular Disease and Mortality Among Individuals with type 2 diabetes. J Clin Endocrinol Metab (2025) 110:e372-e81. doi: 10.1210/clinem/dgae192

PubMed Abstract | Crossref Full Text | Google Scholar

23. Li Y, Liu Y, Liu S, Gao M, Wang W, Chen K, et al. Diabetic vascular diseases: molecular mechanisms and therapeutic strategies. Signal Transduct Target Ther. (2023) 8:152. doi: 10.1038/s41392-023-01400-z

PubMed Abstract | Crossref Full Text | Google Scholar

24. Chen X, Shi C, He M, Xiong S, and Xia X. Endoplasmic reticulum stress: molecular mechanism and therapeutic targets. Signal Transduct Target Ther. (2023) 8:352. doi: 10.1038/s41392-023-01570-w

PubMed Abstract | Crossref Full Text | Google Scholar

25. He Y, Chen X, Liu M, Zuo L, Zhai Z, Zhou L, et al. The potential DNA methylation markers of cardiovascular disease in patients with type 2 diabetes. BMC Med Genomics. (2023) 16:242. doi: 10.1186/s12920-023-01689-3

PubMed Abstract | Crossref Full Text | Google Scholar

26. Agha G, Michael M, Mendelson CK, Caviness W, and Joehanes R. Blood leukocyte DNA methylation predicts risk of future myocardial infarction and coronary heart disease. Circulation. (2019) 140:645–57. doi: 10.1161/CIRCULATIONAHA.118.039357

PubMed Abstract | Crossref Full Text | Google Scholar

27. Zheng H, Li W, Huang G, Zhu H, Wen W, Liu X, et al. Secreted frizzled-related protein 2 ameliorates diabetic cardiomyopathy by activating mitophagy. Biochim Biophys Acta Mol Basis Dis. (2024) 1870:166989. doi: 10.1016/j.bbadis.2023.166989

PubMed Abstract | Crossref Full Text | Google Scholar

28. Edgar R, Domrachev M, and Lash AE. Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Res. (2002) 30:207–10. doi: 10.1093/nar/30.1.207

PubMed Abstract | Crossref Full Text | Google Scholar

29. Woroniecka KI, Park ASD, Mohtat D, Thomas DB, Pullman JM, and Susztak K. Transcriptome analysis of human diabetic kidney disease. Diabetes (2011) 60:2354–69. doi: 10.2337/db10-1181

PubMed Abstract | Crossref Full Text | Google Scholar

30. Corti R, et al. Differential gene expression in thrombus-derived white blood cells of patients with acute coronary syndrome, arrayexpress-repository(2011). Available online at: https://www.ebi.ac.uk/arrayexpress/experiments/E-GEOD-19339 (Accessed March 2024).

Google Scholar

31. R Development Core Team. “R: A Language and Environment for Statistical Computing. Vienna, Austria: The R Foundation for Statistical Computing (2011). Available online at: http://www.R-project.org/ (Accessed March 2024).

Google Scholar

32. Pathan M, Keerthikumar S, Ang CS, Gangoda L, Quek CY, Williamson NA, et al. FunRich: An open access standalone functional enrichment and interaction network analysis tool. Proteomics. (2015) 15:2597–601. doi: 10.1002/pmic.201400515

PubMed Abstract | Crossref Full Text | Google Scholar

33. Szklarczyk D, Gable AL, Lyon D, Junge A, Wyder S, Huerta-Cepas J, et al. STRING v12: Protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res. (2019) 47:D607–13. doi: 10.1093/nar/gky1131

PubMed Abstract | Crossref Full Text | Google Scholar

34. Scardoni G, Petterlini M, and Laudanna C. Analyzing biological network parameters with CentiScaPe. Bioinformatics. (2009) 25:2857–9. doi: 10.1093/bioinformatics/btp517

PubMed Abstract | Crossref Full Text | Google Scholar

35. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, and Turner RC. Homeostasis model assessment: insulin resistance and β-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. (1985) 28:412–9. doi: 10.1007/BF00280883

PubMed Abstract | Crossref Full Text | Google Scholar

36. Vandesompele J, Preter K, Pattyn F, Poppe B, Van Roy N, Paepe A, et al. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Geome Biol. (2002) 3:research0034. doi: 10.1186/gb-2002-3-7-research0034

PubMed Abstract | Crossref Full Text | Google Scholar

37. Livak KJ and Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2– ΔΔCT method. Methods. (2001) 25:402–8. doi: 10.1006/meth.2001.1262

PubMed Abstract | Crossref Full Text | Google Scholar

38. Ortiz-Martınez M, Gonzalez-Gonzalez M, Martagon AJ, Hlavinka V, Willson RC, and Rito-Palomares M. Recent developments in biomarkers for diagnosis and screening of type 2 diabetes mellitus. Curr Diabetes Rep. (2022) 3:1–21. doi: 10.1007/s11892-022-01453-4

PubMed Abstract | Crossref Full Text | Google Scholar