Integrated Analysis of Immunocyte Inltration and Differential Gene Expression in Hypertrophic Obstructive Cardiomyopathy

Objective: Hypertrophic obstructive cardiomyopathy (HOCM) is one of the main reasons for sudden cardiac death (SCD) of young people. Researches have revealed that immune-related genes are closely relevant with HOCM. Therefore, it is important to explore the key immune regulatory mechanisms and biomarkers of HOCM. Methods: We used many bioinformatics methods, including linear models for microarray analysis (LIMMA), protein-protein interaction (PPI) network, Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes pathway (KEGG), and CIBERSORT to assess the key pathway and hub genes involved in HOCM. Furthermore, expression levels of hub genes were validated in human tissue. Results: Our results showed that the degree of inltration of ve immune cells were linked to HOCM, including monocytes, macrophages M2, NK cell resting, B cells native, and T cells regulatory (Tregs). A total of 7 hub genes (CCL2, CXCL8, FOS, MAP2K1, NFKBIA, STAT3, and TNFRSF1A) were identied and validated by qt-PCR. The core genes including CCL2, MAP2K1, NFKBIA, STAT3, and TNFRSF1A are closely related to monocytes inltration in HOCM. Conclusion: Taken together, our research will provide useful information to explore the immune mechanisms underlying HOCM and the potential targets for therapy. The candidate genes CCL2, MAP2K1, NFKBIA, STAT3, and TNFRSF1A were involved in the regulation of monocytes tissue inltration, which is closely related to the HOCM.


Introduction
Hypertrophic cardiomyopathy (HCM) has been one of the main reasons for sudden cardiac death (SCD) of young people [1,2]. The primary cause of SCD in HCM is the left ventricular out ow tract obstruction (LVOTO) [3,4]. About 25-70% of HCM patients demonstrate LVOTO, which is called hypertrophic obstructive cardiomyopathy (HOCM) [5][6][7]. In HCM, most patients remain asymptomatic or mildly symptomatic in the whole life, whereas patients have apparent symptoms in HOCM, consisted of dyspnea, exercise intolerance, chest pain, palpitations, and syncope. Moreover, SCD might be the initial symptom of unfortunate young individuals with HOCM. Therefore, it is essential to study the mechanism of HOCM.
In the heart, there are many cell types, composed by cardiomyocytes and non-cardiomyocytes including immune cells, endothelial cells and broblasts [8]. Multiple reports suggest that rolling and in ltration of circulating and resident immune cells, particular of monocytes, macrophages, and natural killer T lymphocytes, is a critical initiating event in the pathogenesis of HOCM [9,10]. Understanding the speci c functional phenotype of immune cells is as necessary as the recognition of speci c immune cell populations. Resident and circulating immune cells can produce numerous cytokines associated cardiac hypertrophy and heart failure [10]. HOCM is a complex process involving a variety of cell types that interact with each other in the heart and circulation. It is very important to fully understand the recognition and the role of immune-related cell types in pathophysiological response.
In this study, we used many bioinformatics methods to assess the immune-related pathways, and immune cell subtypes in HOCM ( Figure. 1). Furthermore, we used human samples to validate our results. We explored the mechanisms of the in ltration of immune cells and provided useful information to explore the potential targets for therapy in HOCM. Statistical analysis were performed with the R (version 3.6.0).

Identi cation of differentially expressed genes
Raw data was processed using "affay" package of R language. Next, the Benjamini-Hochberg method was used to adjust p-values and thus the false discovery rate (FDR) and fold-change (FC) were calculated [12,13]. Finally, genes expression values of the |log 2 FC| >1.5 and adjusted p < 0.05 for ltering differential expressed genes (DEGs) were set.

Gene Ontology (GO) and pathway analysis
GO enrichment of DEGs was performed using the clusterPro ler algorithm. The results of GO and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment for DEGs were obtained from Metascape (http://metascape.org/gp/index.html) database. GO includes biological process (BP), cellular component (CC) and molecular function (MF). GO terms for which p < 0.05 were considered to be signi cant. And, the ltered the genes enriched in KEGG pathway associated with immunity and in ammation were subsequently used to construct the protein-protein interaction (PPI) biological networks based on the STRING online database (V 11.0) (https://string-db.org/) with the nodes association con dence score > 0.4. In addition, the Cytoscape software (V3.7.1) was used to visualize and evaluate interactions and identifying the hub genes in functional networks [14].

Immune-related pathways and cells subtypes analysis
To characterize the immune cell subsets in HOCM, we've applied the CIBERSORT (https://cibersort.stanford.edu/) estimate software to quantify the immune cell fractions for the gene expression matrix derived from HOCM samples [15]. The advantage of CIBERSORT algorithm in the aspects of characterizing the in ltration of various immune cell subtypes using RNA mixtures from any tissues with a systems-level insight [16]. In addition, we performed a Pearson correlation clustering analysis of hub genes and differential cell subtype in ltration values.
Meanwhile, we applied DAVID (https://david.ncifcrf.gov/home.jsp) database and DisGeNET Curated online database (http://disgenet.org/) to annotate the hub genes and discover the human diseases associated with hub genes. Subsequently, to identify the transcription factor (TF) of the hub genes, the plug-in iRegulon for Cytoscape software was applied, with the default parameters. Herein, we chose the top 3 regulators with the highest NES value to construct the regulatory network involved in HOCM.

Validation of hub genes
Additionally, total RNA from human heart IVS tissue samples was isolated and puri ed using TRIzol method, and reverse-transcribed using the PrimeScript RT Master Mix (TAKARA) according to the manufacturer's instructions. Real-time PCR was performed using SYBR Premix Ex Taq II (TAKARA) and a 7500 Real-Time PCR System (Life Technologies) according to the manufacturer's instructions. The expression level of each sample was internally normalized against that of the glyceraldehyde 3-phosphatedehydrogenase (GAPDH). The relative quantitative value was calculated using 2 − ΔΔCt method. Each experiment was performed in triplicate. The primers used in Real-time PCR were listed in Table A

Identi cation of DEGs in HOCM
Comparing the healthy and HOCM IVS samples 405 genes with different expression (|Log 2 FC|>1.5, adjusted p-value < 0.05) were identi ed, including 254 up regulated and 151 down regulated ( Fig. 2A and Table A

Functional analysis of DEGs and identi cation of hub genes
In enrichment analysis for DEGs, we found that GO terms of top 5 BP including muscle system process, heart contraction, extracellular structure organization, heart process and positive regulation of in ammatory response were closely related to DEGs associated with HOCM ( Fig. 2D and Table A.3), whereas GO terms of top 5 CC consisting of collagen-containing extracellular matrix, extracellular matrix, secretory granule lumen, cytoplasmic vesicle lumen and vesicle lumen were closely related to DEGs associated with HOCM ( Fig. 2E and Table A.3). Meanwhile, GO terms of top 5 MF comprising integrin binding, extracellular matrix structural constituent, collagen binding, structural constituent of cytoskeleton and metalloendopeptidase inhibitor activity were closely related to DEGs associated with HOCM ( Fig. 2F and Table A.3).
Regarding the KEGG pathway enrichment, the DEGs were signi cantly enriched in pathways including phagosome, pathogenic Escherichia coli infection, complement and coagulation cascades, apoptosis, AGE-RAGE signaling pathway in diabetic complications and HIF-1 signaling pathway ( Fig. 3A and Table A.4).
Interestingly, we also found that DEGs enriched 8 immune-related pathways, including NF-kappa B signaling pathway, TNF signaling pathway, IL-17 signaling pathway and Toll-like receptor signaling pathway and so on. These results are illustrated in Fig. 3A. These HOCM-related pathways were related to immunityin ammation response. Hierarchical clustering analysis of the genes of immune-related pathways in HOCM is shown in Fig. 3B. After submitting the genes enriched in immune-related pathways to the STRING database (https://string-db.org/), 7 PPI nodes were obtained, with a con dence threshold greater than 0.4. After analyzed by Cytoscape software as an undirected method, the 7 nodes of the PPI network was considered to be central agents, which were CCL2, CXCL8, FOS, MAP2K1, NFKBIA, STAT3, and TNFRSF1A (Fig. 3C). Compared with the healthy group, Through expression levels of these 7 hub genes were lower expressed in HOCM group (all p < 0.05, Fig. 3D). In addition, CCL2 is the most differentially expressed gene.

Immunocyte in ltration and hub immunocyte detection
To characterize the immunocyte status of HOCM, we've applied the CIBERSORT algorithm to quantify the immune cell fractions for the gene expression matrix derived from cardiac samples. The overall immune in ltration landscape of HOCM tissues is shown in Fig. 4A, B and Table A

Identi cation of the relationship between the hub gene and key immune cells subtypes in HOCM
Through correlation analysis, we found that the 5 hub genes (CCL2, MAP2K1, NFKBIA, STAT3, TNFRSF1A) in immune-related pathways had the strong correlations with the degree of in ltration of monocytes (Fig. 4C). This suggests that, to some extent, monocytes may play an important role in HOCM.

Investigating the Functional Role and TF of Hub Genes
To further understand how the hub genes were correlated with HOCM, we applied DAVID and Metascape online database to explore the biological function and associated pathways. The results of GO term enrichment in hub genes indicated that the response to lipid, response to lipopolysaccharide, and response to cytokine were mainly enriched ( Table 1). The results of pathways were TNF signaling pathway, interleukin-10 (IL-10) signaling, chemokine signaling pathway, signaling by interleukins, Cadmium induces DNA synthesis and proliferation in macrophages, IL-17 signaling pathway, Toll-like receptor signaling pathway and cytokine Signaling in Immune system ( Table 2). We used DisGeNET Curated online database (http://disgenet.org/) to discover the human diseases associated with hub genes, which contained cardiomyopathy, familial idiopathy, and so forth (Table 3).   Finally, we predicted TFs and found that BCL3 transcription coactivator (BCL3), activating transcription factor 1 (ATF1) and GATA binding protein 5 (GATA5), as the master regulators of the hub genes are involved in HOCM (Fig. 4D).

Human tissue for gene expression validation
To test gene expression, the expression of the hub genes was accessed by Q-PCR in human cardiac tissues Therefore, the PCR results partly supported the reliability of our analysis.

Discussion
HOCM is a kind of HCM, characterized by LVOTO caused by the cardiac hypertrophy of IVS. Emerging evidence has revealed the central role of immune-related pathways in cardiac hypertrophy. Our research found that the in ltration of monocyte, macrophages M2, naïve B cells, NK cells resting, and Tregs in cardiac tissues were closely associated with HOCM. Through further analysis, our study also suggested that CCL2, MAP2K1, NFKBIA, STAT3, TNFRSF1A may be the core regulatory genes of immune-related pathways and were closely correlated with the degree of in ltration of monocytes, indicating that these genes may be critical regulatory markers in HOCM.
A greater understanding of the immune system itself has accelerated the progress in de ning the cell populations of the immune system that plays a role in HOCM. Monocytes are key innate immune system mediators of in ammatory responses. In our analysis, the monocytes in HOCM samples had the most signi cant difference compared with healthy samples. Meanwhile, we found that the monocyte had the moderate correlation with 5 hub genes. In addition, CCL2 and TNFRSF1A were validated by PCR. Monocytes can differentiate into macrophages and have a proved signi cant relationship with HF and myocardial infarction in both animal and human research [17,18]. Additionally, monocytes in ltration can cause cardiac hypertrophy and remodeling [19]. Considering the vital participation of monocytes, immune regulation is a promising treatment direction for HOCM and more work is needed to reveal the detailed mechanism. Our results revealed that monocytes may play an important role in HOCM and may serve as a predictive tool in HOCM progression.
Activation of natural killer T cells can attenuate myocardial infarction-induced cardiac remodeling [20], and inhibition of T cell immune activity can ameliorate maladaptive cardiac remodeling in mouse model [21]. Meanwhile, macrophages have a much more substantial role in regulating cardiac hypertrophy and remodeling during heart injury [10]. According to the function of macrophages, it is classi ed into two types, type M1 (classically activated) and type M2 (alternatively activated). Type M1 macrophage can secrete TNF, IL-1, IL-12, and other chemokines to play a proin ammatory function, and the type M2 macrophage mainly secretes anti-in ammatory factors, like epidermal cell growth factor (EGF) and transforming growth factor β (TGF-β) in the late stage of in ammation [22,23]. M1 macrophages mainly facilitate tissue destruction; M2 macrophages promote tissue remodeling and repair, and previous studies showed an increase in M2 macrophage in ltration in myocardium promotes brosis [23][24][25], so altering macrophage phenotype in the heart may be a potential direction to modify cardiac brosis. In consequence, our results suggested the higher relative in ltration value of M2 and NK cells may in uence the pathogenesis of HOCM.
Resting naïve B cell is de ned as B cell before activation and further differentiation into speci c subtypes. At present, we have not found any report about the direct correlation of resting naïve B cells and cardiac hypertrophy or brosis. Zouggari Y et al. found that B-cell depletion can decline left ventricular brosis and cardiac function in the model of myocardial infarction [26]. Furthermore, adoptive transfer of activated Treg cells to cardiac hypertrophic mouse model alleviated CD4+, CD8 + T cells, and macrophages in ltration into the heart and alleviated cardiac hypertrophy and brosis [27]. In our research, the higher relative in ltration values of Tregs and naïve B cells may be associated with the pathogenesis of HOCM, consistent with the previous studies.
There are a few studies investigating cardiac immune-related genes in HOCM patients. STAT3 has a key role in in ammation that underlies cardiovascular disease and impacts on cardiac structure and function and is important for maintaining endothelial cell function and capillary integrity with aging and hypertension [32]. Besides, STAT3 also involves the cardiac hypertrophy and brosis in TAC mouse models [33,34]. Duerr GD et al. suggested that in the cardiac hypertrophy group CCL2 and CXCL8 had increased expression and anti-in ammatory IL-10 had a suppression which indicated the persistent in ammatory reaction [35]. About TNFRSF1A, it is a member of TNF receptor protein and has have effects on remodeling, hypertrophy, in ammation, and apoptosis in HF [36].
Finally, the TFs analysis results show that GATA5, BCL3, and ATF1 were signi cantly predicted in hub genes' regulatory network. Herein, GATA5, a zinc-nger transcription factor essential for cardiovascular development and structural remodeling [37], was the sole and potent transactivator for the β-myosin heavy chain promoter, can bind to nuclear factors induced by leukemia inhibitory factor stimulation during myocardial cell hypertrophy [38]. ATF1, a member of activating transcription factor (ATF) subfamily and basic-region leucine zipper family, is a key driver of human plaque monocytes to acquire the atheroprotective macrophage state [39]. That is to say, ATF1 can promote monocytes to differentiate into a macrophage, in accordance with our analysis.

Conclusion
Taken together, our research found that the in ltration of monocyte, macrophages M2, naïve B cells, NK cells resting, and Tregs in cardiac tissues were closely associated with HOCM progression. In addition, CCL2, MAP2K1, NFKBIA, STAT3, and TNFRSF1A were involved in the regulation of monocytes tissue in ltration, which was closely associated with the process of HOCM. Most of hub regulators were validated in previous researches. However, further experimental evidence concerning the mechanism is still needed.

Declarations
Ethical Approval and Consent to participate Ethics approval and consent to participate. Clinical tissue specimens were obtained from the Guangdong Provincial People's Hospital. This study was approved by the ethics committee of the Guangdong Provincial People's Hospital (Grant No. GDREC2016255H).

Consent for publication
Not applicable.

Availability of supporting data
The dataset supporting the conclusions of this article is available in the GEO repository (http:// www.ncbi.nlm.nih.gov/geo/, GEO accession number: GSE39461).