Elucidating the Histone Deacetylase Gene Expression Signatures in Peripheral Blood Mononuclear Cells That Correlate Essential Cardiac Function and Aid in Classifying Coronary Artery Disease through a Logistic Regression Model

A proinflammatory role of HDACs has been implicated in the pathogenesis of atherosclerosis as an emerging novel epigenetic diagnostic biomarker. However, its association with the clinical and cardiovascular function in coronary artery disease is largely unknown. The study aimed to profile the gene expression of HDAC1–11 in human peripheral blood mononuclear cells and to evaluate their influence on hematological, biochemical, and two-dimensional echocardiographic indices in CAD. The HDAC gene expression profiles were assessed in 62 angioproven CAD patients and compared with 62 healthy controls. Among the HDACs, upregulated HDACs 1,2, 4, 6, 8, 9, and 11 were upregulated, and HDAC3 was downregulated, which was significantly (p ≤ 0.05) linked with the hematological (basophils, lymphocytes, monocytes, and neutrophils), biochemical (LDL, HDL, and TGL), and echocardiographic parameters (cardiac function: biplane LVEF, GLS, MV E/A, IVRT, and PV S/D) in CAD. Furthermore, our constructed diagnostic model with the crucial HDACs establishes the most crucial HDACs in the classification of CAD from control with an excellent accuracy of 88.6%. Conclusively, our study has provided a novel perspective on the HDAC gene expression underlying cardiac function that is useful in developing molecular methods for CAD diagnosis.


Introduction
Coronary artery disease (CAD) continues to be a leading cause of mortality globally [1].CAD is caused by atherosclerosis.Plaque formation and progression result from the intricate interaction of cellular molecules, exacerbated by traditional risk factors such as diabetes, smoking, hypertension, and obesity.These traditional risk factors significantly impact the endothelium and increase the disease burden, resulting in significant complications [2].Clinical symptoms, an electrocardiogram (ECG) [3,4], stress testing, cardiac computed tomography (CT), and molecular markers [5] are used to diagnose coronary The institutional human ethics committee of the Chettinad Academy of Research and Education (IHEC/10-17/Proposal No. 372) was obtained for this study.All participantssigned informed consent before the initiation of the investigation process.Study participants were recruited at Chettinad Super Specialty Hospital between August 2019 and December 2019 and underwent a coronary angiogram with angina symptoms.Each study participant's demographic and clinical histories were recorded based on a questionnaire approved by the institution's institutional review board.By the inclusion and exclusion criteria, 124 participants were recruited for the study.The inclusion criteria for the study participants include the following:(1) Participants belonging to South Indian ancestry of both genders.(2) Physiological status was diagnosed or classified by a qualified senior consultant cardiologist based on the coronary angiogram (for CAD, a lesion greater than 30% in the primary coronary artery or main branches, and participants without CAD, lesions below 30%).The exclusion criteria include the following: 1. previous history of myocardial infarction (MI); 2. heart failure; 3. arterial revascularization; 4. rheumatic disease; 5. cardiomyopathies; 6. pericardial diseases; 7. a severe systemic inflammatory disorder; and 8. acute or chronic respiratory diseases.

Gene Expression Profiling of HDAC
Peripheral blood (5 mL) was collected from the study participants (62 CAD and 62 control) in the K2EDTA BD Vacutainer (BD Vacutainer ® ) at Cath ICU, Department of Cardiology, Chettinad Super Specialty Hospital.Following the manufacturer's protocol, histopaque (Cat#1077, Sigma-Aldrich, St. Louis, MO, USA) was used to isolate the PBMCs from the collected whole blood.Total RNA was extracted from PBMCs (1.30 × 10 6 cells/mL) using Trizol reagent (cat#15596026, Invitrogen, Waltham, MA, USA).The Nanodrop 2000 Spectrophotometer (Thermo Scientific, Waltham, MA, USA) was used to quantify the collected RNA.The reverse transcription was carried out for cDNA synthesis using a SuperScript III First-Strand (Life Technologies, Carlsbad, CA, USA).The gene expression was performed using ABI-7000 (Applied Biosystems, Waltham, MA, USA) with SYBR green master mix (Takara, Kusatsu, Japan) and gene-specific primers for HDACs (Table 1).The HDAC gene expressions were normalized with the housekeeping gene (GAPDH), and the expression was calculated by the ∆∆Ct method.

Echocardiography Imaging
Transthoracic 2D echocardiography was performed using Esaote (MyLabTM25Gold) according to the guidelines of the American Society of Echocardiography [10,11].Left ventricular structural indices such as left ventricular interventricular septal, posterior wall thickness, left ventricular end-systolic dimension, and left ventricular end-diastolic dimension were measured using M-mode echocardiography at the chordae tendineae level to determine left ventricular ejection fraction (biplane LVEF).Further, the mass index of the left ventricle (LVMI) was calculated using the Devereux 1987 method.The biplane Simpson's technique and global longitudinal strain (GLS) were utilized to determine the left ventricle's systolic functions.Furthermore, pulsed wave transmitral flow and tissue Doppler velocity were implemented to determine the left ventricular diastolic functions based on early and late transmitral flow velocity, mitral inflow E/A ratio (MV E/A), mitral peak lateral (L E/e ) and septal myocardial early diastolic velocity (S E/e ), isovolumetric relaxation time (IVRT), pulmonary vein AR duration (PV AR), and pulmonary venous systolic velocity/diastolic velocity ratio (PV S/D).

Diagnostic Model Using Binary Logistic Regression
To assess the diagnostic value of the HDACs related to anthropometric, biochemical, and imaging parameters, the diagnostic model was constructed by fitting the expression value of the significant HDACs into a binary logistic regression model (glm2 package version 1.2.1, R package).Youden's J index was used to find the optimal threshold.All 124 participants were classified into training (70%) and test set (30%).The training cohort was utilized for constructing the model, while the testing cohort was employed to assess the model's performance.Then, the receiver operating characteristic (ROC) curves were utilized to evaluate the model's efficacy.

Statistical Examination
The characteristics of participants, such as anthropometric, biochemical, molecular, and imaging parameters, were represented as the mean ± standard deviation.The difference between the CAD and control was determined using a student t-test for each characteristic except gender data.The categorical data (gender) were assessed throughthe Chi-square test.The Pearson correlation was implemented to assess the association between the characteristics.All statistical analyses were performed using SPSS (version 21), and significance was determined if the p-value was less than 0.05.

Gene Expression Profiling of HDACs in PBMCs
The gene expression of HDACs was evaluated in PBMCs and compared between CAD and control groups.As shown in Table 3, CAD exhibited a significant increase in HDAC1, HDAC2, HDAC4, HDAC6, HDAC8, HDAC9, and HDAC11 expression, while HDAC3 expression decreased compared to the control.In contrast, HDAC5, HDAC7, and HDAC10 failed to reach the minimal significance level between groups.

Assessment of 2D Echocardiographic Imaging
Using echocardiography, the structure and functional behavior of the heart were evaluated in 124 participants.Table 4 displays the outcome of echocardiographic indices.In CAD, the echocardiographic indices of LVMI, PV S/D, and IVRT were significantly increased.In contrast, the ratios of MV E/A, S E/e , L E/e, PV AR, biplane LVEF, and GLS weredecreased compared to the control.Consequently, the statistical analysis of echocardiographic indices confirms the left ventricular dysfunction in CAD relative to healthy individuals.

Diagnostic Model with Crucial HADCs for CAD Classification
A logistic regression analysis was performed after the screening of HDACs based on the correlation significance with hematological, biochemical, and echocardiographic indices.The diagnostic model was created using the R-package version 3.2.0,incorporating the variables HDAC1, HDAC2, HDAC3, HDAC4, HDAC6, HDAC8, HDAC9, and HDAC11.The model utilizes a training set including 70% of the randomly selected 124 people and a test set consisting of the remaining 30%.The diagnostic regression equation was constructed as logit (P) = −16.937+ 0.763×HDAC11 + 1.595×HDAC9 + 0.298×HDAC8 + 1.652 × HDAC4 − 0.676 × HDAC3.The diagnostic capabilities of this equation were evaluated using ROC analysis, as shown in the figures.The training set yielded an accuracy of 88.6% with a 0.94 area under the curve (AUC), a specificity of 86.4%, and a sensitivity of 90.9% (Figure 3A).Similarly, the test set demonstrated an accuracy of 83.3% with a 0.95 AUC, a specificity of 72.2%, and a sensitivity of 94.4% (Figure 3B).These results provide evidence that these HDACs possess strong predictive capabilities of CAD.

Diagnostic Model with Crucial HADCs for CAD Classification
A logistic regression analysis was performed after the screening of HDACs based on the correlation significance with hematological, biochemical, and echocardiographic indices.The diagnostic model was created using the R-package version 3.2.0,incorporating the variables HDAC1, HDAC2, HDAC3, HDAC4, HDAC6, HDAC8, HDAC9, and HDAC11.The model utilizes a training set including 70% of the randomly selected 124 people and a test set consisting of the remaining 30%.The diagnostic regression equation was constructed as logit (P) = −16.937+ 0.763 × HDAC11 + 1.595 × HDAC9 + 0.298 × HDAC8 + 1.652 × HDAC4 − 0.676 × HDAC3.The diagnostic capabilities of this equation were evaluated using ROC analysis, as shown in the figures.The training set yielded an accuracy of 88.6% with a 0.94 area under the curve (AUC), a specificity of 86.4%, and a sensitivity of 90.9% (Figure 3A).Similarly, the test set demonstrated an accuracy of 83.3% with a 0.95 AUC, a specificity of 72.2%, and a sensitivity of 94.4% (Figure 3B).These results provide evidence that these HDACs possess strong predictive capabilities of CAD.

Discussion
The emphasis of our research has been to observe the types of epigenetic alterations in patients presenting with angina pectoris that are further grouped as angioproven CAD and healthy controls.The scientific application of this perception can immensely support

Discussion
The emphasis of our research has been to observe the types of epigenetic alterations in patients presenting with angina pectoris that are further grouped as angioproven CAD and healthy controls.The scientific application of this perception can immensely support the development of effective diagnostic biomarkers for CAD.Earlier clinical studies on animals and human atherosclerotic tissues have demonstrated the epigenetic modification that plays a predominant role in the regulatory network of inflammation, oxidative stress, and vascular smooth muscle cell proliferation, contributing to vascular diseases such as atherosclerosis and restenosis [12].Evidence of HDAC modification in atherosclerosis has been well-established as a valid target for novel therapeutic approaches [13].Yet, the increased understanding of HDAC regulatory activity in atherosclerosis as a diagnostic biomarker associated with echo imaging, biochemical, and hematological analysis is unseen.Hence, to our knowledge, the present study is the first report demonstrating HDAC profiling and altered gene expression in the PBMCs of patients with CAD and healthy controls.Our study demonstrated the significant variation in HDACs 1-11 activity between the angioproven CAD and control individuals.Regarding HDAC profiling, our t-test analysis identified certain HDACs as the exclusive significant factor.Further, using the Pearson correlation, the significant HDACs and biochemical, hematological, and echocardiographic parameters were analyzed and proven to be possible diagnostic markers for CAD.
Histone modification profiling in human PBMCs was performed between angioproven CAD and healthy participants to discover a novel selective diagnostic marker for CAD.Interestingly, by profiling HDACs 1-11, we observed changes in the HDAC gene expression and their statistical significance in the PBMCs between CAD and healthy participants.Several animal, cell, and human model studies have highlighted that class I HDAC is predominant in mediating proinflammatory molecules, inflammation, and endothelial dysfunction and regulating VSMC proliferation, thrombus formation, and atherosclerosis [14].In our study, Class I HDACs (1-3, and 8) were solely significant and dysregulated in the study group.Studies by Manea et al. revealed that HDAC1 and 2 were upregulated in human atherosclerotic aorta/carotid arteries, and ApoE / mice thereby witnessed the dominant role of HDAC1 and 2 in the development of atherosclerotic plaque [15].Yao et al. have reported that the levels of class I HDAC3 were downregulated, which legalizes the cyclic strain and promotes the migration and proliferation of VSMCs in the development of atherosclerosis [16].Another study by Kee et al. demonstrated that the inhibition of HDAC8 in a mouse model lowers blood pressure, reduces the aortic wall thickness, and increases vascular relaxation, resulting in the inhibition of inflammation that leads to atherosclerosis [17].In our present study, we observed statistically significant upregulation of HDAC1, 2, and 8 and downregulation of HDAC3 in CAD participants.
Furthermore, Class II HDACs (4, 5, 6, 7, 9, and 10) were analyzed, resulting in upregulated expressions of histone acetylation, of which HDACs 4, 6, and 9 were statistically significant among the study participants.Numerous studies [18][19][20][21] have proved that HDAC4 promotes VSMC proliferation, migration, and atherosclerotic plaque formation.In addition, HDAC4 was involved in VSMC proliferation, which plays a crucial role in vascular calcification [22] and the inflammatory response [23].Likewise, increasing evidence has supported our study results that HDAC6 plays a decisive role in endothelial dysfunction [24], oxidative stress [25], and inflammation [26] and has a protective role in promoting vascular homeostasis [27].Among all the classical HDACs, HDAC9 is the most well-studied individual subtype for its association with atherosclerosis in animal and human models.It has been reported that an increase in HDAC9 gene expression in human internal carotid [28], plasma/coronary artery disease [29,30], the blood of patients with large atherosclerotic lesions [31], and plaque vulnerability [32] results in plaque formation and the development of atherosclerosis that further causes severe cardiac events.However, our study was in line with other studies in which the levels of HDAC were upregulated and highly significant among our CAD participants.Although the HDACs (5, 7, and 10) showed dysregulation in CAD, they were not statistically significant.Of note, HDAC11 is the only member of Class IV that has been least explored and analyzed in the development of atherosclerosis in humans and animals.Our study showed that Class IV (HDAC11) levels were significantly upregulated among CAD participants.In contrast to our investigation, Zhang and Ge et al. and Yanginler and Logie et al. [33,34] demonstrated that HDAC11 potentially treats atherosclerosis.Hence, these findings determine the dysfunction of HDAC11 in the pathogenesis of atherosclerosis and CAD, which must be further studied in detail.
In addition to HDAC gene expression profiling, we have also analyzed the association of significantly altered HDACs (1-4, 6, 8, 9, and 11) with the hematological and biochemical parameters that play a vital role in the cascade of events that leads to atherosclerosis.Despite all these studies, the association between HDAC levels and hematological/biochemical parameters related to CAD has not been explored.In our present investigation, our data showed that dysregulated levels of HDACs such as HDAC2 (basophils, lymphocytes, monocytes, neutrophils, HDL, and LDL), HDAC3 (HB, PCV, neutrophils, HDL, and LDL), HDAC9 (TLC, basophils, lymphocytes, monocytes, neutrophils, HDL, and LDL), HDAC1 (basophils and neutrophils), HDAC6 (SBP, DBP, basophils, monocytes, and T. cholesterol), HDAC8 (HDL and LDL), and less known HDAC11 was correlated with the hematological/biochemical parameters (TLC and monocytes) of CAD patients.Studies such as Dorneles et al. and Chi et al. provided good evidence for our data by demonstrating an imbalance of HDAC2 expression levels in obese patients [35] and the role of HDAC6 in cardiac dysfunction regulated by angiotensin II [36].Hence, the present experimental association of dysregulated HDACs with the hematological and lipid profiles of CAD indicates the crucial pathological role of HDACs in atherosclerosis [37], which can potentially serve as a diagnostic biomarker for CAD with clinical benefits.
To address whether dysregulated HDAC gene expression distresses the left ventricular mass diastolic and systolic function, we validated the statistically significant profiled HDACs with 2D transthoracic echocardiographic imaging.Left ventricular mass index is recognized as one marker of cardiovascular risk in patients without CAD.Abdi-Ali et al. [38] and Kee et al. [39] ruled out the significant association of LVMI with cardiac hypertrophy and future cascades of cardiac events.Our data found a similar association between dysregulated HDACs 1, 2, and 6 and LVMI in the CAD group, demonstrating the regulatory mechanism of HDAC activity in LV mass hypertrophy.Further, the assessment of LV function showed a significant association of dysregulated HDACs 1-4, 6, 8, 9, and 11 with systolic and diastolic function.Several studies in CAD patients have elucidated that subclinical LV systolic and LV diastolic impairments are independent markers for predicting CAD.Our data results were in line with a few studies performed on both animal and human models by Kimbrough et al., Chen et al., and Jeong et al. [40][41][42] that illustrate that dysregulated HDACs promote endothelial dysfunction and contribute to the pathological process that leads to CAD and mediated cardiac events.Altogether with prior research, alterations in HDAC mRNA expression are detected in numerous CAD-associated cells, including endothelial cells, smooth muscle cells, and cardiomyocytes.Comparable patterns of gene expression were identified in peripheral blood mononuclear cells (PBMCs) that correlate with cardiac function in CAD.Considering the CAD association of these HDACs, we constructed a diagnostic model with significant HDACs that showed accuracy greater than 83% in both training and testing sets, which showed that our model had a strong clinical application value.To our knowledge, no existing study has developed a diagnostic model utilizing HDAC gene expression.Therefore, this diagnostic model holds significant novelty and potential for application in laboratory settings.Overall, this study has demonstrated that the process of histone deacetylation has the potential to function as a diagnostic biomarker for coronary artery disease.

Conclusions
In conclusion, our study evaluated the gene expression levels of various classes of HDACs in the human peripheral mononuclear cells in patients with CAD and compared them with the healthy control group.It was observed that significant alterations in HDACs 1-4, 6, 8, 9, and 11 were associated with the critical hematological, biochemical, and cardiac indicators of CAD.Further utilizing the HDACs, the diagnostic model was generated, which showed significant accuracy in classifying CAD from healthy normal.

Figure 1 .
Figure 1.Correlation analysis of significantly altered hematological and biochemical features with HDACs (1-4, 6, 8, 9, and 11) in CAD participants.The yellow color represents an insignificant association between the analyzed features, and the red color indicates a significant association.Of those, the mathematical sign + and − represent positive and negative correlation, respectively.

Figure 1 .
Figure 1.Correlation analysis of significantly altered hematological and biochemical features with HDACs (1-4, 6, 8, 9, and 11) in CAD participants.The yellow color represents an insignificant association between the analyzed features, and the red color indicates a significant association.Of those, the mathematical sign + and − represent positive and negative correlation, respectively.

Figure 2 .
Figure 2. Correlation analysis of significantly altered echocardiographic indices with HDACs (1-4, 6, 8, 9, and 11) in CAD participants.The yellow color represents an insignificant association between the analyzed features, and the red color indicates a significant association.Of those, the mathematical sign + and − represent positive and negative correlation, respectively.

Figure 2 .
Figure 2. Correlation analysis of significantly altered echocardiographic indices with HDACs (1-4, 6, 8, 9, and 11) in CAD participants.The yellow color represents an insignificant association between the analyzed features, and the red color indicates a significant association.Of those, the mathematical sign + and − represent positive and negative correlation, respectively.

Biomedicines 2023 , 12 Figure 3 .
Figure 3. Area under the ROC curve graphs for the CAD diagnostic model's training (A) and test (B) sets based on HDAC gene expression levels.The color code (blue and red) represent the area under the curve (AUC).

Figure 3 .
Figure 3. Area under the ROC curve graphs for the CAD diagnostic model's training (A) and test (B) sets based on HDAC gene expression levels.The color code (blue and red) represent the area under the curve (AUC).

Table 1 .
List of primers used in our study.

Table 2 .
Characteristics of the study population in anthropometric, hematological, and biochemical features.

Table 3 .
HDAC gene expression of study participants.
p-value < 0.05 considered as significant.

Table 4 .
Characteristics of the study population in echocardiographic parameters.
p-value < 0.05 considered as significant.