Genome-Wide Gene Expression Profiles

As discussed previously, the hypothesis-based approaches aim to either prove or disprove the basic proposition. On the other hand, a hypothesis-neutral genome-wide approach can yield an unbiased and testable new, wide angle networks, novel hypotheses, and provide the opportunity to integrate, elucidate and discover the functional association between differentially expressed genes. As shown in Table 3.5, nine reports of genome-wide gene expression profiles in ovarian and peritoneal endometriosis provide significant leads about their specific genomic signatures. A few significant issues emerged from this analysis, which are discussed in the following sections.

Firstly, there appears a need to exercise caution in defining the quality of the control samples since endometrial physiology based on genomic expressions are pathognomic even in cases of uterine fibroids and polyps and thus do not fall under the umbrella of “normal" endometrium [144].

Secondly, the three subtypes of endometriosis, ovarian, peritoneal, and rectovaginal or adenomyosis show clear distinctions both clinically and in their molecular phenotypes [18, 19, 145-149]; thus critical interpretation of the “molecular signature" of endometriosis should take into consideration case-specific data based on the nature of lesions [138,148-155] despite a certain degree of overlap in lesion sites in peritoneal and ovarian foci [145,156] and to severity stage of the disease [152-154,157].

Reference

Major results and comments

Arimoto et al. 2003 [150]

26 genes upregulated in proliferative phase and 30 genes upregulated in ectopic tissues included TGFpi, HLA antigens, and complement factors. Common downregulated genes included TP53, GADD 34, GADD 45A, and GADD 45B.

Differential genomic expressions in phase specific manner in eutopic versus ectopic tissues for ovarian endometriosis.

Hever et al. 2007 [151]

53 genes upregulated in ectopic compared with eutopic samples. Genes associated with immune responses were affected suggesting endometrioma as an immune disorder.

Eyster et al. 2007 [159]

Differential gene expressions in ovarian endometrioma versus peritoneal implants, as Well as, differential expression of genes in eutopic versus ectopic for several functional categories observed.

No difference in gene expressions for ERp, aromatase, 17pHSD, IL-6, IL-8, and Tenascin between eutopic and ectopic tissues.

Inflammatory genes form the largest functional group displaying differential expressions with involvement of WNT and bone morphogenetic pathways.

Borghese et al. 2008 [152]

Ectopic endometrioma showed dysregulated gene expressions compared with autologous eutopic tissue samples; 5600 differentially regulated genes clustered into 55 functional groups.

Lack of oncogenic potential along with down regulated HOX A and В genes with repressed cell cycle genes.

Zafrakas et al. 2008 [153]

Upregulated retinoic acid responder 1, 2 in endometrioma indicative of high degree of differentiation; cell cycle, metabolism, and homeostasis related genes down regulated. Eutopic but not ectopic tissue showed high telomerase reverse transcriptase enzyme activity.

Ovarian endometriosis showed highly differentiated expressional features with no malignant potential and overt genomic aberration.

0Continued)

Reference

Major results and comments

Aghajanova et al. [157]

High expressions of EGFR and ERBB2 in early secretory phase. Dysregulated PI3K/AKT, JAK/STAT, SPK/JNK and МА? К pathways in severe endometriosis.

Dysregulated progesterone and cAMP regulated genes and genes for thyroid hormone metabolism and action in endometriosis eutopic tissue compared with normal tissue.

Khan et al. 2012 [154]

A cohort of 28 genes with high degree of predictability for ovarian endometriosis in fertile women. Although overt oncogenic potential absent, but genes associated with gynaecological cancer highly expressed in ectopic tissue. No change in aromatase expression in eutopic and ectopic tissues, however, high expression of StAR, and SF1 in ectopic tissue.

Dysfunctional immuno-neuro-endocrine behaviour appears critical in endometriosis. Dysregulation of CLOCK, ESR1, and cMYC, all of them being major transcription factors appear as significant causative factors in pathogenesis of ovarian endometriosis.

Monsivais et al. 2012 [155]

Abnormalities in pathways regulating metabolism and action of prostaglandins and glucocorticoids with critical involvement of TNF in endometriotic stromal cells.

Characteristics pro-inflammatory milieu in ectopic lesion sites.

Tamaresis et al. 2014 [144]

Genes related to ER, AR, immune activation, CCR3 and CXCR4 pathways, stress, and inflammatory responses in endometriosis differentially regulated as compared to normal endometrium, and non-endometriotic endometrium with uterine/pelvic pathology.

Classifier analysis of genomic data allowed detection of stage-specific pelvic endometriosis in hormone-dependent and independent milieu.

3 Molecular Biology of Endometriosis 95

Thirdly, the issue of fertility and subfertility/infertility has its bearing on the molecular phenotyping. Endometriosis has long been associated with infertility, however, the mechanism by which it affects fertility are not fully understood, the estimated fecundity rate in women with untreated endometriosis is between 2-10% [158]. Endometrial transcriptome in infertility indicates the involvement of inflammatory genes and bone and morphogenetic pathways in ovarian endometrioma [159], differential expression of genes in the transcriptional regulation by RNA editing protease dependent G-protein signaling cascades and receptor tyrosine mediated signaling in eutopic tissue of stage IV endometrioma [160]. Aghajanova and colleagues [157] provided evidence of heightened expressions for EGFR mRNA and protein and ERBB2 in early secretory phase with severe endometriosis, which are associated with dysregulation of second-messenger signaling pathways that included PI3K/AKT, JAK/STAT, SPK/JNK, and MAPK, confirming earlier reports [161]. Tamaresis et al. [144] with the use of machine learning and high dimensional analysis tools provide a classifier based analysis of genomic data to detect and to stage pelvic endometriosis in hormone dependent and independent milieu.

Finally, for understanding gene to gene causal interactions it is necessary to undertake differential expression and differential co-expression analysis to search for groups of genes that show similarities among the different conditions [162,163]. In a study reported from our laboratory, gene expressions in eutopic and ectopic tissues in ovarian endometriosis were thus examined [154]. To improve upon the two-fold change at P <0.05 between two groups of tissue samples generally considered significant, Khan et al. [154] then sought to circumvent biases and inadequacies in interpretation and discovery [164,165] by employing a three-fold change at P < 0.01 as the pre-set filter for differential expression (DE) of individual genes between groups followed by pathway network-based enrichment analysis, and for differential coexpression (DC) analysis of K-mean based expressional clusters followed by gene set enrichment analysis for post hoc analysis of autologous data set in fertile pool of women suffering from ovarian endometriosis. The GSEA model for interpreting transcriptomic data derives its power by focusing on gene sets, that is, groups of genes that share common biological function, chromosomal location, or regulation [166]. Based on the GSEA model eutopic endometrium in stage IV ovarian endometriosis in fertile women of

Indian origin was found transcriptionally dysfunctional in mediating immune-neuro-endocrine responses with vulnerability to give rise to ectopic foci through a pathways network of several transcription factors including CLOCK-ESR1-MYC in the pathoetiology of ovarian endometriosis [154]. Figures 3.1 and 3.2 provide knowledge-based construction of the pathways-network of transcription factors putatively associated with pathogenesis of endometriosis.

 
Source
< Prev   CONTENTS   Source   Next >