Screening and identification of critical biomarkers in erectile dysfunction: evidence from bioinformatic analysis

Introduction

Erectile dysfunction is a medical condition in which a man is unable to maintain an erection for vaginal intercourse (1993). Actually, a description of poor penile erection was firstly mentioned as early as 5,000 years ago in an ancient Egyptian literature (Shamloul & Ghanem, 2013). Erectile dysfunction is known to increase with age. In a cross-sectional study based on community, the prevalence of this condition, in men in the 40–49 age group, was about 5% for severe dysfunction (Feldman et al., 1994). For men in the 70–79 age group, the prevalence rates were 15% and 34%, for complete or severe dysfunction, respectively. By 2025, it is speculated that the global prevalence of ED will reach up to 322 million cases (Bacon et al., 2003). Men aged older than 40 years are more susceptible to erectile dysfunction. Several other medical diseases are linked to erectile dysfunction such as diabetes mellitus (Brown et al., 2005), tobacco use, central neuropathologic conditions, cardiovascular disease (Billups, 2005), lower urinary tract symptoms of benign prostatic hyperplasia (McVary, 2005), and metabolic syndrome (Kupelian et al., 2006).

The vessels of the vascular system (and lymphatics) contain a single-cell-layered lining called the endothelium. Based on its uniqueness and functions, the endothelium is thought to be an organ (Aird, 2004). The parts of the endothelium are diverse regarding biochemical properties, functional characteristics, and cell surface markers (Ho et al., 2003; Kallmann et al., 2002). The understanding of the heterogeneity of this factor in the endothelial cells (ECs) mainly originates from several examinations of various body tissues and organ specimen by microarray analysis (Chi et al., 2003; Ho et al., 2003; Kallmann et al., 2002; Podgrabinska et al., 2002).

Endothelial dysfunction is considered as the genesis of ED, especially that of the systemic vasculature (Gerber et al., 2015). Plenty of studies have been carried out to investigate the function of the endothelium both during physiological erection and also in the ED state (Hurt et al., 2002). The endothelium not only functions as a passive barrier but also as a critical regulator of muscle tone and vascular circulation, which are coordinated by signals originating from mechanical stimuli, humoral and neural sources.

In order to facilitate research into the progression of ED in terms of the functional pathways and differentially expressed genes (DEGs), many technologies such as bioinformatics analysis and microarray technology have been widely used. However, because it is exceedingly difficult to obtain penile tissue specimens from patients with ED, there are few studies of human tissue. Thus, here, the DEGs between non-ED tissues and ED tissues were searched and obtained by screening the Gene Expression Omnibus (GEO) for mRNA microarray datasets. Subsequently, protein-protein interaction (PPI) networks, pathway enrichment analyses of the Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO), were performed to explore the mechanisms that underlie the pathogenesis of ED. In conclusion, we identified 618 DEGs and 25 hub genes as possible biological markers for ED.

Materials & Methods

Results

Identification of DEGs in the ED

DEGs (618 in GSE10804) were identified (Fig. 1A) after standardizing the microarray data. The dataset consisted of 430 downregulated genes and 188 upregulated genes between ED endothelial cells and non-ED endothelial cells.

Selection and analysis Hub gene

We found 25 hub genes with degrees ≥10. Table 4 shows the list of the functions, abbreviations, and names of the genes. The analysis of the biological processes of the 25 genes is shown in Figs. 2C and 2D. For the determination of the categorization of the hub DEGs, the DAVID tool was used. Results of GO analyses demonstrated that changes in BP of hub DEGs remarkably emerged in cell responses to amino acids stimuli, and inflammatory response (Table 3, Figs. 2C and 2D) Changes in molecular function (MF) primarily enhanced in extracellular matrix structural constituent and chemokine activity (Table 3, Figs. 2C and 2D). Changes in cell component (CC) mainly increased in the collagen trimer and proteinaceous extracellular matrix (Table 3, Figs. 2C and 2D). Based on the analysis of the KEGG pathway, we found that the enrichment of hub DEGs was mainly in Protein digestion and absorption (Fig. 3A), ECM-receptor interaction (Fig. 3B) and PI3K-Akt signaling pathway (Table 3, Figs. 2C, 2D and 3C).

Table 3:

GO and KEGG pathway enrichment analysis of hub DEGs in ED samples.

Term	Description	Count in gene set	Gene ratio	P-value
GOTERM_BP_DIRECT
GO:0071230	Cellular response to amino acid stimulus	5	20	2.53E−07
GO:0006954	Inflammatory response	7	28	4.93E−07
GO:0030199	Collagen fibril organization	4	16	1.06E−05
GO:0070208	Protein heterotrimerization	3	12	4.91E−05
GO:0006955	Immune response	5	20	1.44E−04
GOTERM_MF_DIRECT
GO:0005201	Extracellular matrix structural constituent	8	32	4.34E−14
GO:0008009	Chemokine activity	3	12	1.84E−03
GOTERM_CC_DIRECT
GO:0005581	Collagen trimer	7	28	3.48E−11
GO:0005578	Proteinaceous extracellular matrix	8	32	1.54E−09
GO:0005615	Extracellular space	9	36	2.31E−05
GO:0005587	Collagen type IV trimer	3	12	3.31E−05
GO:0005584	Collagen type I trimer	2	8	3.04E−03
KEGG_PATHWAY
ptr04974	Protein digestion and absorption	11	44	2.40E−14
ptr04512	ECM-receptor interaction	10	40	2.56E−12
ptr04151	PI3K-Akt signaling pathway	12	48	1.40E−09
ptr04510	Focal adhesion	10	40	5.78E−09
ptr05146	Amoebiasis	8	32	2.31E−08

DOI: 10.7717/peerj.8653/table-3

Notes:

Abbreviations

GO

Gene Ontology

KEGG

Kyoto Encyclopedia of Genes and Genomes

DEGs

differentially expressed genes

ED

erectile dysfunction

Table 4:

Functional roles of 25 hub genes with degree ≥10.

Gene symbol	Full name	P-Value	logFC
C3	Complement component 3	3.36E−03	1.324
CCR5	C-C motif chemokine receptor 5	6.96E−03	−1.395
COL1A1	Collagen type I alpha 1 chain	1.78E−09	−10.353
COL1A2	Collagen type I alpha 2 chain	1.36E−06	−4.873
COL21A1	Collagen type XXI alpha 1 chain	9.97E−03	−1.804
COL2A1	Collagen type II alpha 1 chain	4.85E−05	−1.92
COL3A1	Collagen type III alpha 1 chain	1.16E−04	−2.991
COL4A1	Collagen type IV alpha 1 chain	6.43E−03	−1.865
COL4A2	Collagen type IV alpha 2 chain	1.85E−03	−1.871
COL4A4	Collagen type IV alpha 4 chain	5.07E−03	−1.064
COL5A3	Collagen type V alpha 3 chain	9.11E−03	−2.508
COL6A1	Collagen type VI alpha 1 chain	4.87E−05	−4.282
COL6A2	Collagen type VI alpha 2 chain	1.65E−05	−5.482
COL6A3	Collagen type VI alpha 3 chain	7.16E−07	−7.906
CXCL11	C-X-C motif chemokine ligand 11	2.30E−03	−1.368
CXCL12	C-X-C motif chemokine ligand 12	4.03E−03	−2.309
CXCL5	C-X-C motif chemokine ligand 5	1.29E−03	−2.202
CXCL6	C-X-C motif chemokine ligand 6	4.23E−03	−1.948
DRD2	Dopamine receptor D2	1.71E−03	−1.038
FPR2	Formyl peptide receptor 2	9.28E−04	1.731
GABBR2	Gamma-aminobutyric acid type B receptor subunit 2	3.17E−06	2.374
GNAI1	G protein subunit alpha i1	1.73E−04	−1.107
GNG7	G protein subunit gamma 7	4.59E−03	−1.475
LEPREL1	P3H2 (LEPREL1) prolyl 3-hydroxylase 2	–	–
LPAR1	Lysophosphatidic acid receptor 1	2.28E−03	−2.632

DOI: 10.7717/peerj.8653/table-4

Discussion

Men aged older than 40 years are more susceptible to erectile dysfunction. Several other medical diseases are linked to erectile dysfunction such as diabetes mellitus (Brown et al., 2005), tobacco use, central neuropathologic conditions (e.g., hemorrhagic or ischemic stroke and Parkinson’s disease), cardiovascular disease (Billups, 2005), lower urinary tract symptoms of benign prostatic hyperplasia (McVary, 2005), metabolic syndrome (Kupelian et al., 2006). More than 70% of coronary artery disease patients present with moderate to severe erectile dysfunction which might occurred before cardiovascular events. Moreover, endothelial dysfunction is also linked with atherosclerosis (Billups, 2005). The likelihood of developing erectile dysfunction tends to increase with prolonged diabetes probably due to a high level of glycated hemoglobin. Some drugs are also suspected to cause erectile dysfunction based on a clinical diagnosis of 25% of the patients diagnosed with erectile dysfunction (Thomas et al., 2003). However, little is understood about the molecular mechanism of ED at present. The NO-cGMP signaling pathway has been shown to be critical in the pathogenesis of ED (Garcia-Cardoso et al., 2010). These results have also been applied in clinical practice, as selective inhibitors of 5-type phosphodiesterase specific to cGMP, such as sildenafil and tadalafil have achieved significant clinical effects in the treatment of ED. RhoA/Rho-kinase contributes to the development and progression of ED. It has been demonstrated that the vasodilation of NO may be due to the reduction of calcium levels by NO-cGMP-PKG on one hand, and the relaxation effect by blocking the RhoA/Rho-kinase calcium-sensitive pathway on the other hand (Jin et al., 2006). Recently, it has also been found that the MAPK signaling pathway was demonstrated to affect the NO-cGMP signaling pathway by regulating related kinases, which indirectly leads to the occurrence of ED. The primary treatments for erectile dysfunction are oral PDE5-Is. Other interventions are psychotherapy, penile devices, testosterone therapy, injection therapies, and psychotherapy, etc (1993). But there is no curable medications for ED that can be completely cured, which may be one of the reasons for the poor patient prognosis. Therefore, potential markers for diagnosis and treatment with high efficiency are urgently demanded. Microarray technology allows us to explore the genetic alterations in the ED and there is evidence surporting it to be a useful method for identifying new biomarkers in other diseases. An article conducted a large-scale genome-wide association study of erectile dysfunction in the multiethnic Kaiser Permanente Northern California Genetic Epidemiology Research in Adult Health and Aging cohort and UK Biobank. They find the single-minded family basic helix-loop-helix transcription factor 1 (SIM1) gene is a mechanism that is specific to sexual function (Jorgenson et al., 2018).

Here, we analyzed the mRNA microarray datasets to identify DEGs in ED cells and non-ED cells. 618 DEGs were obtained in all datasets, including 430 downregulated genes and 188 upregulated genes. The reciprocities among the DEGs were searched out by KEGG enrichment analyses. The DEGs largely had enrichment in positively regulating transcription from RNA polymerase II promoter, cell adhesion, calcium ion binding, receptor binding, Akt signaling pathway, receptor interaction, Protein digestion, and absorption. Lianjun Pan and Jiehua Ma speculated that many lncRNAs are involved in the regulation of muscle contraction and relaxation through ion channel activity. They suggested that a muscle contraction disorder induced by abnormal ion channel activity, collagen deposition and fibrosis might play a critical role in the pathogenesis of ED (Pan et al., 2015). Changes to the endothelial cell–cell junction integrity like in the case of increases in vascular permeability and decreased cell–cell junction proteins were observed in the cavernous tissue of the diabetic mouse. The presence of hollow junctions in diabetes state may explain the high prevalence of erectile dysfunction that occurs before other cardiovascular conditions (Ryu et al., 2013). Large conductance, voltage, and Ca2+ activated K+ channels are widely present in many cells of the body, which play a part in regulating the neuronal excitability and smooth muscle tone. When their function is compromised, they may bring about the development of erectile dysfunction, hypertension and overactive bladder (Carmona et al., 2016). Akt plays an essential role in regulating cell homeostasis through its effects on several downstream effectors (Carmona et al., 2016). Prior studies have shown that phosphorylation of eNOS by Akt causes prolonged NO production and maximal construction and influences penile erection (Zhang et al., 2011). Similarly, it was found that PI3K/Akt-eNOS activation mediated the effects of the A2B adenosine receptor on penile erection (Ryu et al., 2013).

We selected 25 DEGs as hub genes with difference degrees ≥10. Among these hub genes, COL3A1 showed the highest node degrees with 13. And other hub genes also confirmed the top node degrees with 12 (COL1A1, COL1A2, COL21A1, COL2A1, COL4A1, COL4A2, COL4A4, COL5A3, COL6A1, COL6A2, COL6A3, CXCL12, LEPREL1) and 11 (C3, CCR5, CXCL11, CXCL5, CXCL6, DRD2, FPR2, GABBR2, GNAI1, GNG7, LPAR1). A study identification of oxidative stress-induced gene expression profiles in cavernosal endothelial cells, cxcl12 showed a corresponding variation in the CECs and ED rat model compared with the results of the gene microarray analysis (Hu et al., 2015). While ORP in NHPs produced persistent erectile and urinary tract dysfunction. Periurethral injection of CXCL-12 was feasible and improved both urinary incontinence and erectile dysfunction and suggests (Zambon et al., 2018). A study about five immune agents (IgG, IgM, IgA, C4, and C3) was collected on the basis of the Fangchenggang Area Male Health and Examination Survey (FAMHES)and methods of traditional cross-sectional analysis were used. However, there was no clear relationship between ED risk in the baseline analyses and the tested indexes (C3: P = 0.737) (Chen et al., 2014). Another study on sexual dysfunction in male schizophrenia showed that correlation between two polymorphisms in the genes for the DRD2 and scores on questionnaires for erectile and sexual dysfunction were studied. The functional −141Ins/Del promoter region polymorphism of DRD2 was remarkably correlated with sexual dysfunction (Zhang et al., 2011). The remaining genes did not retrieve research literature related to erectile dysfunction.

Conclusions

In conclusion, 618 DEGs and 25 hub genes with the possibility to be considered as diagnostic biomarkers for ED were identified However, additional investigations should be carried out to explore the biological function of these genes in the ED.