STEM I

Our first long-term project in STEM is the independent research project! We are allowed to chose any STEM topic of interest to pursue research and experimentation to investigate our research question. The final products of our STEM projects are then presented at a STEM Fair in February.



A Meta-Analysis on Endometriosis Biomarkers using GWAS

Endometriosis, a chronic disease defined by the presence of endometriotic lesions outside of the uterus. This disease is found in 5-10% of women at reproductive age yet is majorly under researched. My research aims to identify biomarkers of Endometriosis in order to expand knowledge on the disease and access to treatment and rapid diagnosis.

Abstract

Endometriosis is a chronic illness defined by the presence of endometrial tissue in the form of lesions outside the uterus. Endometriosis is a multifactorial disease that affects 5-10% of women of reproductive age and is the leading cause of infertility (Chauhan et al., 2022). Yet, despite how widespread the disease is, it is majorly under researched. Furthermore, there is much to be desired in treatment, prevention, and diagnosis methods. The aim of this research project is to address the most significant needs of endometriosis patients: knowledge and accessibility. This study investigated a possible genetic component that makes patients susceptible to Endometriosis. To do this, computational biology was conducted using Genome Wide Association Studies (GWAS) and the Gene Expression Omnibus (GEO) to identify biomarkers of Endometriosis. Then, prevention strategies were be developed based on these biomarkers. The results of this study include a set of potential biomarkers for endometriosis, as well as a discussion on how their relationship with the Endometriosis disease. At the end of this study, multiple non-invasive biomarkers are developed for the purpose of detecting endometriosis via tissue or urine sample.

Graphical Abstract

Graphical Abstract

Research Proposal

Research Question

What are potential biomarkers for Endometriosis?

Hypothesis

Biomarkers will target to contributing factors of Endometriosis such as immune system—including interleukins and macrophages, inflammation, and hormones.

Background

background Infographic

Endometriosis is a women’s health disease that is present in 5-10% of women at reproductive age, and inflammation because of the disease can lead to infertility (Chauhan et al., 2022). Superficial Peritoneal Endometriosis (SPE), the most common form of Endometriosis where lesions are found in the peritoneal cavity (Imperiale et al., 2023). The main theory regarding the cause of Endometriosis is the Retrograde Menstruation Theory, developed by Dr. John Sampson. It is a phenomenon where endometrial tissue becomes menstrual blood and travels up the fallopian tubes and into the peritoneal cavity. This causes the blood to invaginate into its environment in the form of endometriosis lesions (Sampson, 1927). Retrograde Menstruation is not the sole cause of Endometriosis as not all women who experience this phenomenon develop Endometriosis (Lamceva et al., 2023). Other contributing factors for Endometriosis are genetics, hormonal imbalance, inflammation, and immune dysfunction (Appendix A). The illness has four stages, with tissue scarring worsening as the stages progress (Chauhan et al., 2022). SPE can only be diagnosed via a laparoscopy where surgeons remove any endometrial lesions present in the peritoneal cavity (Chauhan et al., 2022). It is important to note that endometrial lesions could grow back even after surgical removal, with a recurrence rate of 30-50% (Imperiale et al., 2023). The current clinically approved method of treatment is by use of GnRH agonists, which inhibit lesion growth, by stopping estrogen production and repressing ovulation (Nishimoto-Kakiuchi et al., 2023). However hormonal treatments like GnRH agonists can have side effects such as menopausal symptoms, metrorrhagia, weight gain, depression, mood swings, migraines, and bone density loss (Shan et al., 2017). There are several reasons why hormonal treatment is not feasible for patients. To name a few examples, it is not possible for someone to get pregnant while on hormonal medication and some patients may be at greater risk for side effects caused by GnRH agonists. Although there is a basic understanding of how to diagnose and treat Endometriosis, patients still struggle with the limited variety of treatment methods, negative symptoms of GnRH agonists, and access to a diagnosis. Research is needed to improve diagnosis methods and give patients access to more affordable and less invasive diagnosis methods. Furthermore, with a better understanding of the nature of Endometriosis, researchers could potentially develop and clinically approve more natural treatment techniques to inhibit lesion growth. While around 51% of Endometriosis cases are inherited either maternally or paternally, there is no one gene that indicates the likelihood of having Endometriosis as it is a multifactorial disease (Nyholt et al., 2012). This study focuses on investigating possible biomarkers that could be used to detect SPE. SPE was chosen as the focus for this study as it is the most common form of the disease


Procedure

Procedure Infographic volcano plot

The preliminary data for this study used data from the Gene Expression Omnibus (GSE23713). The Genome Wide Association Study EFO_0001065 was used with compiled associations from multiple studies. Microsoft Excel was used to analyze data and create volcano plots. The compiled data from GWAS studies was imported to Excel (A) and sorted by significance based on the p-value cut off (5E-8) as used by Nyholt et al. in 2012 (B). RNA sequences that were associated with Ovarian endometriosis or Deep Infiltrated endometriosis were deleted as they are not the focus of this study. The log Fold Change value of each RNA sequence was determined by the logarithm of each corresponding beta value. The top 20 upregulated RNA sequences and top 20 downregulated RNA sequences were determined by the RNA sequences with the top 20 logFC values and top 20 logFC values (C). The data was then plotted as a volcano plot to depict statistical significance in RNA sequences in SPE patients (D). Then, highlighted RNA sequences were converted to their mapped genes. Finally, those mapped genes were looked up on the GWAS database to find traits associated with the expression of those genes (E).


Upregulated genes Downregulated genes


From these highlighted genes and their associated functions, we can see what quantitative measures are associated with each gene, highlighted in red. White blood cells within the immune system are not viable to be considered as biomarkers as they constantly change (The Editors of Encyclopaedia Britannica). Similarly, blood cell counts are not proven to be viable biomarkers for endometriosis so they can also be canceled out. We’re left with protein counts like CCL5, PHF tau, and RANTES as well as measures attainable by body scans, like IDP dMRI TBSS ICVF Sagittal stratum R, visceral adipose tissue measurement, QRS duration. Associated phenotypes measured via blood tests include triglyceride levels, high density lipoprotein cholesterol measurement, atrial natriuretic factor measurement, creatinine measurement, sex hormone-binding globulin measurement (SHBG), von Willebrand factor (VWF) measurement, 2-hydroxyadipate levels, factor VII measurement, Atrial natriuretic factor (ANF), IGFBP-3 measurement, aspartate aminotransferase measurement, serum alanine aminotransferase measurement, quinolinic acid measurement, ferritin measurement, alkaline phosphatase measurement (ALP), and Gamma glutamyl transferase (GGT) levels. Measures attainable via vaginal samples include vaginal microbiota measurement. Markers measured with urine samples are ST2 protein, non-albumin protein levels, and 3-hydroxypropylmercapturic acid (3-HPMA) measurement. Platelet count during third trimester of pregnancy is also a function of the downregulated gene CLIC4, this could mean that pregnant women could also detect their Endometriosis.

Biomarkers

Discussion

All these quantitative measures have the potential to be biomarkers for Endometriosis with further research. Through our data mining, we were able to come up with a list of potential biomarkers that can be detected via blood, urine, body scans, or vaginal samples as visualized in Figure 3. Due to these findings coming from a meta-analysis of many GWAS studies, conditions are not necessarily the same. For example, genomic data could have been collected from patients at different stages of their menstrual cycle, which will affect hormone coding genes. Furthermore, genomic data could have been collected from patients at different stages of endometriosis and with different types of endometrioses. This error could have happened because GWAS only compiles data as specific as “Endometriosis” and not “Stage I Superficial Peritoneal Endometriosis”. These findings also have incomplete validation. This study reinforces current knowledge on noninvasive biomarkers for Endometriosis as well as introduces new potential biomarkers. Anastasiu et al. in 2020 summarizes current biomarkers in their study, including IGFBP-3, VEGF, and HLA, which were all found in the current study. In the future, upregulated and downregulated genes should be investigated to depict their relationship with the endometriosis disease. Further validation should be done on the meta-analysis to determine the validity of this study’s work. If the discovered genes are indeed valid biomarkers for endometriosis, they could be used in a logical regression to predict the likelihood of Endometriosis. Single cell RNA sequence databases will also be investigated to develop more biomarkers for endometriosis.

This study developed a list of potential noninvasive biomarkers for endometriosis through a meta-analysis of GWAS relating to Endometriosis. GWAS data was data-mined for the top 20 upregulated and downregulated genes, and a volcano plot was created to visualize data. This process resulted in a collection of potential biomarkers to be further investigated for the end goal of speeding up the diagnosis process. If these findings are proved valid, these biomarkers could lead to the development of more treatment methods for endometriosis (Appendix B).

References

Project Poster

STEM fair poster