Metabolite changes in patients with endometriosis: new potential diagnostic and therapeutic targets

Introduction Although invasive, endometriosis is a benign multifactorial gynecological disease and is one of the most frequently encountered gynecological disorders, characterized by the presence of endometrial glands and stroma outside the uterine cavity. Despite being a benign condition, endometriosis exhibits several biological behaviors similar to cancers, such as neovascularization, cell invasion, and overgrowth. Endometriosis affects approximately 6%–10% of women of reproductive age with 30%–50% of patients suffering from pelvic pain and infertility. Endometriosis is a complex estrogen-dependent disease[1]. Due to the lack of effective diagnostic techniques, surgery remains the gold standard of diagnosis. Although multiple theories exist regarding the etiology of endometriosis and its high prevalence, the details of its underlying mechanisms remain under debate. Endometriosis is widely accepted to adversely affect fecundity and pregnancy potential, and patients usually need the support of assisted reproductive techniques to achieve pregnancy. Nevertheless, patients with endometriosis have poorer outcomes when undergoing intracytoplasmic sperm injection (ICSI)/in vitro fertilization (IVF) than women with tubal factor only[2]. Endometriosis is recognized as not just a local disorder but also a chronic systemic disease[3]. Recent studies have shown a definite trend toward more complex metabolomic analysis of endometriosis. Metabolites, as the substrates and products of metabolism, are involved in essential cellular functions such as energy production and storage, cell apoptosis, and signal transduction[4]. The development of metabolomic analysis is helping to increase our understanding of metabolism in various states and diseases[5]. Metabolomics can be used to obtain valuable discoveries in biofluids, cells, and tissues because of advances in metabolite measurement technologies such as nuclear magnetic resonance (NMR) spectroscopy and mass spectrometry (MS)[6]. Metabolomics are informative and consistent technologies that can distinguish the presence of a disease and monitor disease progression[7]. Although the metabolomic approach has been applied to the study of endometriosis, this application remains in the preliminary stage, despite the fact that studies have found metabolite profile changes in serum[8,9], endometrial fluid[10], follicular fluid[2,11], and peritoneal fluid[12,13] of patients with endometriosis. In this review, we provide a comprehensive overview of the metabolomic analysis of biological fluid and tissues, including peripheral blood, eutopic endometrium, ectopic lesions, follicular and peritoneal fluid, from patients with endometriosis. The metabolomic profile alterations of these patients are related to the pathological changes of the disease. These studies comprise single biochemical marker studies and recent metabolomics studies. We illustrate that metabolomics have the high sensitivity and specificity required to differentiate between the disease conditions of endometriosis. Furthermore, certain pivotal metabolites are expected to be identified as novel and effective targets for treating endometriosis. Metabolic markers in the peripheral blood of patients with endometriosis To make the diagnosis of endometriosis less invasive and more accessible, such as by avoiding laparoscopy, the precise identification of serum biomarkers for endometriosis is required[14,15]. Peripheral blood metabolites can be used as relatively accessible clinical samples that reflect the entire metabolic levels of the body. Therefore, these are often used as the first step in screening disease biomarkers[16]. The technique of peripheral blood metabolomics analysis is mature and may provide a new direction for developing a noninvasive adjuvant diagnostic tool. Oxidative stress markers Several untargeted NMR-based metabolomics studies have analyzed blood samples of patients with or without endometriosis[8,13,17] and revealed the altered serum metabolomics profiles of these patients compared to those of the healthy controls. The serum samples of patients with endometriosis exhibited elevated levels of lactate, 3-hydroxybutyrate, 2-hydroxybutyrate, leucine, valine, alanine, threonine, lysine, succinic acid, glycerophosphatidylcholine, and creatine, as well as reduced levels of lipids, isoleucine, glucose, and arginine compared with those healthy controls, which reflected the impaired pyruvate metabolism and increased oxidative stress in endometriosis[8,18]. The increased creatine concentration, which helps protect tissues from the harmful effects of free radicals, is of considerable significance. Even in the presence of oxygen, malignant cells produce high levels of lactate, which is recognized as the “Warburg effect.” Patients with endometriosis may also experience this biological process because this behaves in a similar manner as cancer[19,20]. Therefore, the lactate accumulation and glucose depletion infer that enhanced anaerobic glycolysis occurs in endometriosis and that elevated 2-hydroxybutyrate levels may be a valuable oxidative stress marker for detecting the disease[8]. The increased anaerobic glycolysis in endometriosis may also relate to enhanced oxidative stress. The increased creatine and estradiol concentrations in patients with endometriosis compared with those in controls fulfill the high energy demands of ectopic cells[21]. Nitric oxide synthase (NOS) converts L-arginine to L-citrulline and nitric oxide (NO), and the decreased levels of L-arginine indicate the impairment of the arginine–NO pathway. Oxidative stress can destroy local tissue and promote disease aggressiveness, which is interrelated with the development of endometriosis. Jana et al.[17] revealed that the citrate, succinate, serum lactate, and pyruvate levels were increased in patients with endometriosis, indicating that the glycolysis is vigorous in endometriosis patient. The accumulation of lactate and pyruvate reflects the enhanced glycolysis and mitochondrial dysfunction in these patients[22]. These studies also observed the increased production of reactive oxygen species (ROS), advanced oxidation protein products, lipid peroxidation, as well as decreased total antioxidant capacity and superoxide dismutase (SOD), catalase, and glutathione levels in the serum of patients with endometriosis compared with those of healthy controls[17,23]. In normal cells, multiple antioxidants scavenge ROS, whereas the insufficient SOD and catalase levels in patients with endometriosis indicate a decreased scavenging efficiency, leading to excessive ROS in the system. The excessive ROS in endometriosis may be due to a dysfunction of the mitochondrial respiratory system, indicating oxidative stress condition in patients with endometriosis[17]. Lipids and amino acids Vouk et al.[24] used MS to analyze serum samples from 40 patients with endometriosis and 52 healthy controls. The study reported that eight differential lipid metabolites, regarded as relevant biomarkers, were elevated in patients with endometriosis. The elevated levels of sphingomyelins and phosphatidylcholines in endometriosis are relevant to the suppression of apoptosis and affect the signaling pathways of lipids; consequently, they have been suggested as possible biomarkers for endometriosis. These results also showed that a model involving hydroxy-sphingomyelin and the ratio of phosphatidylcholine to ether phospholipid had considerable sensitivity and specificity for diagnosing endometriosis[24]. Ghazi et al.[25] applied a 1H NMR-based metabolomics approach to explore metabolic changes in the serum of patients with endometriosis. They found that compared with levels in the controls, those of several metabolites such as 2-methoxyestradiol, 2-methoxyestrone, dehydroepiandrosterone, androstendione, aldosterone, and deoxy corticosterone were significantly elevated in infertile patients with endometriosis, indicating the increased expression of estradiol, which promotes the growth of ectopic lesions[25]. Significantly higher concentrations of lipoproteins were also found in patients with endometriosis. Oxidative stress increases the total cholesterol (TC) levels of patients with advanced-stage endometriosis. The increased levels of low-density lipoprotein (LDL), non–high-density lipoprotein (non-HDL), triglycerides, and TC and a lower HDL:TC ratio and serum vitamin E level were demonstrated compared with those in healthy controls[26]. Vitamin E is a potent antioxidant, and infertile patients with endometriosis may have a higher consumption of Vitamin E; therefore, lower serum vitamin E levels reflect the increased oxidative reactions in these patients[27]. The metabolism of sphingolipids is altered in serum, endometrial tissue, and peritoneal fluid of patients with endometriosis. Numerous sphingolipids, including glucosylceramides and ceramides, accumulate in the peripheral blood and peritoneal fluid of these patients. Glucosylceramide is a mitogenic factor that may be an important substance for the ectopic lesion to survive[28]. Moreover, altered lipid expression also contributes to the oxidative stress condition. Patients with endometriosis also have increased concentrations of ketone bodies[29], glycerophosphocholine, choline-containing metabolites, lipoproteins, and fucose and reduced creatinine concentrations in plasma than found in healthy controls[9]. The increased fucose levels in the plasma metabolomic profile of patients with endometriosis may be associated with the initiation and progression of local and distant ectopic focus[9]. Glutamine levels were found to be increased in the serum of patients with endometriosis[30,31]; the level of glutamine is also increased in specific brain regions (insula) of patients with endometriosis with chronic pelvic pain, suggesting a role in the onset or deterioration of pain[31]. In addition, for diagnosis of stage I endometriosis, serum alanine has 90% sensitivity and 58% specificity while in Stage II endometriosis, alanine, leucine, lysine, proline, and phenylalanine levels are changed in the serum and may be suitable as diagnostic markers of this disease[18]. However, in another study, no distinguishable differences in the serum metabolome between patients with or without endometriosis were observed, which increases the challenges associated with using metabolic methods to identify this heterogeneous disease[32], and further research is still needed. A schematic describing this section is shown in Fig. 1. Metabolic markers in the eutopic endometrium and ectopic lesions of patients with endometriosis In endometriosis, the proliferation of uterine endometrial cells outside the uterine cavity requires additional biological energy and biosynthesis. The eutopic endometrium and ectopic lesions of patients with endometriosis thus have different metabolic profiles that are associated with disease development and progression, and the concentrations of various metabolites are significantly altered in all disease stages[18]. Large amounts of metabolites are produced in endometriotic tissues by abnormally active processes, including inflammation, excessive estrogen and prostaglandin production, and/or progesterone resistance. Inflammation of ectopic foci is closely related to the overexpression of prostaglandins, metalloproteinases, cytokines, and chemokines. Autoregulatory positive-feedback loops further promote their accumulation in established lesions[33]. Lipids Lipids play an essential role in almost all pathological processes, including inflammation, oxidative stress, proliferation, and angiogenesis, which are all involved in the pathogenesis of endometriosis[13]. Lipids were found to be involved in the implantation, hyperproliferation, and decreased apoptosis of endometriotic tissue and also contributed to the pain syndromes of patients with endometriosis[34,35]. Adamyan et al.[36] applied direct MS to analyze ectopic lesions of patients with endometriosis in a prospective observational cohort study on 50 patients with ovarian cysts and peritoneal endometriosis. Conspicuous differences were identified in lipid levels between endometriotic lesions and the eutopic endometrium. Ectopic tissues have high levels of sphingolipids, phospholipids, and fatty acids (di- and triglycerides), indicating that MS may be an efficient tool for identifying endometriotic tissue. These differentially expressed lipids may be specific for identifying endometriotic lesions but cannot differentiate between peritoneal and ovarian endometriosis. However, they did not detect the endometrium of healthy patients. Thus, lipids may not be of use in distinguishing the presence of endometriosis[36]. An abnormal sphingolipid metabolic pathway may promote endometrial cell proliferation, causing implantation and growth of ectopic lesions. Sphingosine-1-phosphate, a bioactive sphingolipid metabolite, can promote cell proliferation and migration and resistance to apoptosis and stimulate angiogenesis. Local immune cells are stimulated by sphingosine-1-phosphate to express tumor necrosis factor-alpha (TNF-α) and interleukin-8 (IL-8), which consequently promote cyclooxygenase-2 (COX-2) and prostaglandin production. Oxidative stress can also induce COX-2 production to promote the synthesis of prostaglandins, which are associated with pain and infertility in endometriosis[36,37]. Endometriotic lesion size and number of implants are positively correlated with COX-2 levels, and lower COX-2 levels can reduce lesion sizes[38]. The increased choline glycerophospholipid (ChoGpl) level is necessary for phospholipase A2 (PLA2G2A) synthesis, and ChoGpl is often overexpressed in endometriotic lesions; lysophosphatidic acid is involved in cell proliferation in endometriosis, and phospholipase A2 can promote its production[39]. In malignant tissue, ChoGpl is closely related to high cell proliferation, and therefore be considered as a potential biomarker of endometriosis[35,36]. Fatty acids are sources of signaling molecules and are esterified to phospholipids and involved in energy supply, which is essential for the rapid proliferation of ectopic endometrial cells[40]. Sphingolipids and ChoGpl were found to be more abundant in patients with endometriosis[35]. The biosynthesis and metabolism of bioactive sphingolipids regulate a series of cellular processes such as cell proliferation, migration, and apoptosis, and imbalances in their metabolism, which is involved in the pathology of endometriosis[28]. The signal pathway of endogenous lipids associated with sphingolipids is altered in the endometrial tissues of patients with endometriosis[28]. Saturated diacylglycerols and saturated triacylglycerols levels were shown to be reduced in the endometrial fluid of patients with endometriosis[10]. The endometrial stromal cells of eutopic endometrium and ectopic foci from patients with endometriosis have a damaged ceramide-signal pathway that is closely associated with apoptosis, thus facilitating the growth of endometriotic cells[41]. Increased levels of acylcarnitine (AC) are present in the endometrial fluid of patients with endometriosis. Previous lipidomic studies showed that increased AC levels are related to cell β-oxidation dysfunction which is associated with inflammation[24]. Vouk et al.[24] found that the continuous process of denervation followed by reinnervation in ectopic foci was accompanied by elevated sphingomyelin levels, indicating that the elevated sphingomyelin levels may be a major source of pain in patients with endometriosis. These results indicate that the increased production of sphingolipids and phosphatidylcholines and the imbalances in their metabolic pathways may be directly involved in the implantation of endometrial stromal cells and play a role in the heightened proliferative and decreased apoptosis of endometriotic cells, causing lesion growth and associated pain. Investigations of the role of oxidative metabolism in the pathogenesis of endometriosis showed that endometriotic lesions had higher levels of oxidative damage products, such as mitochondrial DNA rearrangement, 8-OH-deoxyguanosine and lipoperoxide, than the normal endometrium[42]. Patients with deep infiltrating endometriosis (DIE) had a lower LDL-cholesterol expression compared with healthy controls, representing the increasing uptake level of cholesterol by endometriotic tissues. Low-density lipoprotein receptors (LDL-R and LRP-1) were more highly expressed in endometriotic lesions than in the endometrium of the same patient or healthy patients while the LDL-R gene had a higher expression during the secretory phase of the menstrual cycle in the eutopic endometrium of patients with endometriosis[19]. Individuals with high body mass index (BMI) also have increased LRP-1 mRNA expression in endometriotic lesions. Therefore, decreasing the BMI by reducing weight is critical[19]. Amino acids and lactate A lipidomic profiling study of endometrial fluid, which may reflect the metabolism of eutopic endometrium, was performed on patients with ovarian endometriosis[10]. Ovarian endometriosis was associated with 123 significantly differentially expressed metabolites in the endometrial fluid compared with controls (at the time of embryonic implantation). Using these 123 differentially expressed metabolites, a predictive model was generated with the significant diagnostic accuracy of approximately 86% for ovarian endometriosis, suggesting that endometrial fluid analysis represents a minimally invasive approach for endometriosis diagnosis[10]. Cysteine levels are significantly increased in patients with endometriosis, which is important because this amino acid is a precursor to protein and antioxidant glutathione biosynthesis and can also maintain physiological redox conditions. The increased cysteine levels in endometrial fluid suggest alterations to redox balance. In addition, the levels of sphingolipids and ceramides were lower in the endometrial fluid of patients with endometriosis than those in healthy controls[10]. The levels of various purines and amino acids, such as hypoxanthine, L-arginine, L-tyrosine, leucine, lysine, inosine, omega-3 arachidonic acid, guanosine, xanthosine, lysophosphatidylethanolamine, and asparagine, were also upregulated in the eutopic endometrium of patients with early endometriosis[43]. Patients with early endometriosis had lower phenylalanine and uric acid levels in endometrial tissue than healthy patients[18,43]. Endometriotic cells have high energy requirements, a property similar to that found in cancer cells[44], and the enhanced glycolysis levels could be caused by the Warburg effect seen in tumorigenesis, which promotes rapid cell proliferation. Increasing lactate production in peritoneal fluid may “feed” ectopic endometrial cells, facilitating their survival and implantation and promoting their invasion of the peritoneum to form endometriosis lesions[45]. The increasing amounts of lactate can also activate TGF-β, which induces myofibroblast differentiation, leading to the epithelial to mesenchymal transition in the adjacent peritoneum or lesions, further facilitating the invasion of ectopic endometrial cells[46]. In patients with severe endometriosis, the endometrial tissue exhibits decreased formate levels. One possible explanation for this is that in these patients, formate acts as an essential metabolite for purine biosynthesis because of the increased angiogenesis and tissue growth; therefore, formate is likely incorporated into the synthesis of purine nucleotides, leading to its decreased levels[18]. Enzymes The levels of PLA2G2A seem to favor an elevated platelet-activating factor synthesis in ectopic endometrium. Ovarian ectopic lesions exhibit PLA2G2A mRNA levels that are more than 2000-fold higher than those in the normal endometrium[47]. Previously, aromatase was shown to be highly expressed in endometriosis foci, although this is low or even absent in the endometrium of healthy women[1]. The expression levels of phosphatases sphingosine-1-phosphate phosphatase 2 gene (SGPP2), sphingosine-1-phosphate lyase 1 gene (SGPL1), and sphingosine-1-phosphate receptor 3 gene (S1PR3) were decreased in the eutopic and ectopic endometrium of patients with endometriosis, whereas that of SGPP1, SPHKAP, S1PR2, and S1PR1 was increased compared with that in controls, indicating that this pathway may play an important role in the growing and survival of ectopic lesions[48]. Glycolysis-associated genes such as hypoxia-inducible factor-1A (HIF1A), pyruvate dehydrogenase kinase 1 (PDK1), and lactate dehydrogenase A (LDHA) are more highly expressed in endometriosis lesions than in the eutopic endometrium, and a higher HIF1A and glucose transporter 1 (SLC2A1) expression is also found in the adjacent peritoneum than that in control peritoneum. TGF-β1 exposure increased the production of lactate in mesothelial cells, which increased the mRNA and protein levels of HIF1A, as well as those of glycolysis-associated genes LDHA, PDK1, and SLC2A1. The metabolic phenotype changes in ectopic lesions and peritoneal mesothelium of patients with endometriosis may favor disease development. The peritoneum adjacent to lesions can use glycolysis as one way to produce energy, and endometriotic cells assimilate lactate expressed by peritoneal cells to produce an abnormal microenvironment that promotes the establishment and progression of ectopic foci[45]. Furthermore, pyruvate kinase M1/2 is a typical glycolysis enzyme whose expression is increased in biopsies of the ectopic endometrium, indicating the decreased cellular respiration[49]. Further studies are warranted to confirm or assess the suitability of these metabolites as biomarkers for endometriosis. The schematic for this section is shown in Fig. 2. Metabolic markers in the follicular fluid of patients with endometriosis The follicular fluid provides a complex microenvironment surrounding the oocyte and strongly regulates oocyte quality, fertilization, and embryo development[50]. Approximately 30%–50% of patients with endometriosis are infertile and have hampered follicular development[51]. Mitochondrial dysfunction in endometriosis lowers oocyte quality and potentially affects the fertility of patients with endometriosis[52]. When undergoing ICSI/IVF treatment, patients with endometriosis with poor oocyte quality tend to have adverse pregnancy outcomes. Women with endometriosis undergoing IVF treatment have decreased retrieved oocyte numbers during ovarian stimulation and often need higher gonadotropin doses than women without endometriosis[53]. NMR-based metabolic profiling could be used as an effective approach to identify the abnormal metabolism of carbohydrates, lipids, and amino acids in the follicular fluid of patients with endometriosis. In the follicular fluid, several biochemical characteristics may be beneficial in determining oocyte quality and subsequently obtaining fertilization and potential embryo development[12]. Amino acids and lactate The follicular fluid theoretically reflects the entire follicle metabolism. During recent years, prominent advances have been made in finding metabolomic predictors of oocyte quality in follicular fluid and have shown promising results. The metabolites identified as oocyte competence indicators include phosphocholine, leucine/isoleucine, glutamine, and HDL[11,54]. Recent studies emphasized the use of follicular fluid metabolomics in the area of endometriosis infertility, reporting an unfavorable metabolic profile as being indicative of oocyte competence and also for the diagnosis of endometriosis. The follicular fluid of women with endometriosis has significantly increased valine concentrations compared with that of controls, and the concentration of valine in the follicular fluid was markedly higher in women with repeated IVF failure than in those who achieved pregnancy after the first IVF treatment[54]. The follicular fluid composition differs between patients with endometriosis and controls, and a specific metabolic signature is found in different endometriosis phenotypes, including DIE and ovarian endometrioma (OMA)[11]. Tyrosine conversion into acetoacetate contributes to ketone body production, which may explain the lower tyrosine concentration in the follicular fluid from patients with DIE[11]. Higher concentrations of threonine in the endometriotic follicular fluid indicate that this essential amino acid will probably not be able to enter the tricarboxylic acid (TCA) cycle. The decreased concentration of citrate, an intermediate in the TCA cycle, in the follicular fluid of patients with endometriosis further supports this hypothesis, suggesting that an enhanced cytosolic degradation is present[11,55]. Differences in the concentrations of amino acids such as threonine and glutamine could also discriminate the presence or absence of OMA in women with DIE[11]. In the follicular fluid of patients with DIE, the concentrations of amino acids such as tyrosine and alanine were lower than those in the control[11,56]. Anaerobic glycolysis shifts toward lactate production as alanine is converted to pyruvate; therefore, decreased levels of alanine coincide with increased levels of pyruvate and lactate. Alanine has been shown to improve meiotic maturation, and its concentration may be important for the development of oocytes[56]. The anaerobic glycolysis pathway is important in the pathological process of endometriosis and may be associated with the metabolite composition of follicular fluid, which is consistent with findings from other biological fluids of endometriosis[55,57]. In the follicular fluid of patients with DIE, the concentrations of glucose and creatine were lower than those in control patients, whereas the concentrations of lactate, pyruvate, free fatty acids, and lipids were higher than those in the control group[11,56]. The decreased creatine concentration supports the shift in lipid metabolism toward β-oxidation[11]. Glucose and lactate have been identified as oocyte competence metabolite indicators[11]. The follicular fluid from patients with ovarian endometriosis had lower glucose concentrations and higher lactate and pyruvate concentrations than that from the controls[58], suggesting that in patients with ovarian endometriosis, the metabolism relies on anaerobic glycolysis, reflecting a reduced functional TCA cycle. The increased lactate level in the follicular fluid of patients with endometriosis is attributed to increased anaerobic glycolysis in granulosa cells[55,56]. Succinate is an intermediate in the TCA cycle, and if the concentrations are decreased, threonine will not enter the TCA cycle; these results further supported the role of anaerobic glycolysis in endometriosis[21]. Ketone bodies and lipids In ovarian endometriosis, the highly proliferative ectopic cells require considerable energy, which promotes ketone body production in the follicular fluid[29], as demonstrated by the increased concentration of ketone bodies in the follicular fluid of patients with endometriosis compared with that in the controls[11]. Ketone bodies are byproducts of the glutathione oxidation pathway and energy-rich compounds. They can also promote inflammation as well as oxidative stress[59] and are associated with glycolysis, providing a source of energy for oocyte growth[60]; therefore, the increase in ketone body concentration in endometriosis is consistent with the characteristics of this disease. The decreased glucose concentration represents the activation of anaerobic glycolysis. Metabolic analysis has suggested that increased glycerol and ketone body concentrations in patients with OMA indicated that lipolysis activation is followed by β-oxidation in the follicular fluid. These results indicate that mitochondrial dysregulation occurs in endometriosis[11]. Three metabolites are closely related to endometriosis-associated infertility: phytosphingosine, lysophosphatidylcholine (lysoPC) (18:2), and lysoPC (18:0). LysoPC (18:2) and lysoPC (18:0) levels have been shown to be elevated in the follicular fluid of patients with endometriosis, whereas phytosphingosine levels were reduced. Phytosphingosine is involved in sphingolipid metabolism, suggesting that sphingolipid metabolism was abnormal in patients with endometriosis[61]. In the follicular fluid of patients with endometriosis, the higher level of lysoPC may be associated with the low conception rate. The decreased phytosphingosine level can also increase the risk of type 2 diabetes mellitus in patients with endometriosis[62]. These differential metabolites may offer a new approach for the differential metabolite profile to noninvasively assess the oocyte developmental potential of patients with endometriosis[61]. Oxidative stress markers A moderate ROS level is required in physiological conditions, but an imbalance between pro-oxidant and antioxidant activities may harm folliculogenesis and embryo development[23]. Patients with endometriosis have higher follicular fluid oxidative stress levels[2]. Several findings have indicated that levels of oxidative stress markers are increased while those of antioxidant markers are decreased in the follicular fluid of patients with endometriosis, indicating that oxidative stress affects the follicular fluid composition and follicular development in these patients[23,27,63]. Patients with endometriosis have lower vitamin C levels than infertile patients without this disease. Vitamin C plays an important role against oxidative stress, and its lack could cause ovarian atrophy and extensive follicular atresia. In patients with endometriosis, the lower vitamin C levels in the follicular fluid may be due to the excessive antioxidant consumption when attempting to neutralize excess ROS, and thus the low levels of antioxidant capacity may be the cause of the excess ROS in the follicular fluid of patients with endometriosis[23]. Patients with endometriosis have significantly decreased vitamin E levels before ovulation probably because antioxidants are consumed in oxidation reactions. After ovulation induction, compared with that in controls, patients with endometriosis not only have increased lipid peroxidation but also maintained lower vitamin E levels in the follicular fluid, which compromises oocyte quality in these patients[27]. In human follicles, the normal expression of SOD plays a role in ovulation and oocyte development, which can be positively correlated with the number of oocytes retrieved, mature oocytes, and fertilized oocytes[23]. The SOD concentration in a follicular fluid may also be a good prognostic marker in IVF. Infertile women with endometriosis have reduced antioxidant capacity in their follicles[23]. Zinc, copper, and selenium are important for follicle development, and a significant deficiency of copper and zinc in patients with endometriosis could cause a decrease in SOD activity. Women with endometriosis who became pregnant following IVF had higher intrafollicular zinc levels[63], whereas nonpregnant women with endometriosis have significantly decreased zinc levels, suggesting that zinc deficiency in the follicular fluid of patients with endometriosis is a potential risk factor associated with pregnancy failure[63]. Selenium and glutathione peroxidase (GPx) have a positive correlation, and their activity was significantly lower in infertile patients with endometriosis than in the controls. The decreased levels of intrafollicular selenium/GPx in patients with endometriosis enhance oxidative stress in oocytes, affecting their development[63]. Endometriosis has been reported to be closely related to ovarian oxidative imbalance. A recent study found that the levels of malondialdehyde (MDA) and resolvins (RvD1) are increased in the follicular fluid and thus could act as predictors of oocyte quality to influence reproductive ability[64]. Several studies demonstrated that the follicular fluid of patients with endometriosis exhibited decreased concentrations of ROS, NO, and lipid peroxidation but an increased level of total antioxidant capacity[23,27,63]. NO is a negative parameter for IVF outcomes of patients with endometriosis. The high intrafollicular ROS/NO levels in patients with endometriosis are associated with poor oocyte and embryo quality[63]. The follicular fluid of infertile patients with endometriosis has higher MDA levels than the controls, indicating a disequilibrium between ROS production and removal, which partly contributes to poor oocyte quality[27]. Too low or too high glucose levels in the follicular fluid were associated with increased ROS levels and are detrimental to oocyte and follicle maturations[65]. Aberrant sphingolipid metabolism in the peritoneal fluid may, via ROS, compromise oocyte development and infertility in women with endometriosis[28]. A certain level of ROS is necessary under physiological conditions; however, the imbalance between pro-oxidants and antioxidants in the follicles of patients with endometriosis also has a detrimental effect on oocyte maturation and embryo development[23]. Insufficient or increased ROS levels in endometriosis may have an adverse effect on oocyte quality and embryo development in vivo, causing adverse pregnancy outcomes[66,67]. The composition of the follicular fluid is correlated with oocyte development and the rate of implantation after fertilization, so these results may help improve understanding of the high rates of infertility in patients with endometriosis. Therefore, based on metabolomic profiling of follicular fluid in patients with endometriosis undergoing IVF, high-quality oocytes could be selected. The schematic of this section is shown in Fig. 3. Metabolic markers in the peritoneal fluid of patients with endometriosis In the endometriosis microenvironment, the peritoneal fluid communicates with the endometriotic lesions, indicating that the peritoneal fluid could be a critical source for discovering biomarkers of peritoneal endometriosis. Biomarkers in the peritoneal fluid and peripheral blood are not often directly correlated[68]. Current efforts have focused on peritoneal fluid components, including metabolites that contribute to the pathogenesis of endometriosis. Lipids The peritoneal fluid of patients with endometriosis is rich in proteins and lipids[68]. Compared with those in the controls, patients with severe endometriosis have significantly higher levels of glucosylceramide in the peritoneal fluid[28]. Eleven sphingolipids with significantly elevated levels were identified in the peritoneal fluid of infertile patients with severe endometriosis compared with those in infertile patients without endometriosis[12]. In the peritoneal fluid of patients with endometriosis, the levels of ceramides such as sphingolipids are elevated and can promote endometrial stromal cell migration[69]. Endometriosis-associated infertility is associated with the accumulation of very long-chain ceramides in the peritoneal fluid, which may compromise murine MII oocyte maturation by altering mitochondrial superoxide production[12]. In the peritoneal fluid of infertile patients with endometriosis, the concentrations of lipid peroxidation end products, including MDA, 8-isoprostane, and 25-hydroxycholesterol, were increased[70]. PLA2G2A overexpression has been shown in the peritoneal fluid and ectopic endometrium of patients with endometriosis[47,71]. The peritoneal fluid of patients with endometriosis has high levels of sphingomyelins and phosphatidylcholine; thus, the inhibition of PLA2G2A signaling to reduce sphingomyelins and phosphatidylcholine metabolism could be a potential treatment[24]. The lipoproteins in the peritoneal fluid have also been studied, and the peritoneal fluid of endometriosis subjects contains substantial quantities of LDL in a slightly oxidized form that was rarely seen in the controls[26,34]. Compared with controls, patients with endometriosis have 125 significantly altered metabolite ratios. The two most prominent diagnostic ratios are the ratio of carnitine to phosphatidylcholine (C0/PC ae C36:0) and the ratio between two types of phosphatidylcholines (PC aa C30:0/PC ae C32:2). Hence, these metabolite ratios may have the potential to be biomarkers for semi-invasive endometriosis diagnostics[13]. Amino acids and lactate Phenylalanyl-isoleucine is a dipeptide comprising phenylalanine and isoleucine that is involved in intracellular signal transduction. The levels of phosphatidylcholine and phenylalanyl-isoleucine were distinctly increased in the peritoneal fluid of patients with endometriosis compared with those in controls, indicating a possible altered metabolic state that promotes the growth of ectopic lesions[72]. In a recent study using untargeted metabolomics, phenylalanyl-isoleucine was identified as a novel potential mini-invasive diagnostic biomarker of peritoneal endometriosis. The level of phenylalanyl-isoleucine was reflected in the circulation and was increased in serum with an 81.4% diagnostic value, identifying a connection between the circulation and the dysregulated peritoneal cavity in women with endometriosis[68]. The concentrations of lactate were found to be increased in the peritoneal fluid of patients with endometriosis[45]. TGF-β is considered to play an important role in the pathology of peritoneal endometriosis since this can induce the Warburg effect in tumors and can increase lactate concentrations to participate in the development of endometriosis through promoting cell migration, invasion, angiogenesis, and immune escape. Compared with those in nonendometriosis controls, the levels of secreted TGF-β1 were increased in the peritoneal fluid of patients with endometriosis[45]. The concentrations of lactate in peritoneal fluid positively correlated with those of TGF-β1 across the menstrual cycle, forming a positive feedback loop[45]. The lactate levels are positively correlated to metastases and may act as a useful biomarker for diagnosing endometriosis, and the in vivo assessment of lactate levels could therefore be used to detect responsiveness to treatment. Furthermore, Horne et al.[73] used a PDK inhibitor dichloroacetate in a murine model to reduce the secretion of lactate in peritoneal fluid. This metabolic drug may become a novel therapeutic drug among women of childbearing age[73]. Oxidative stress markers Oxidative stress markers are present in abundance in the peritoneal fluid of patients with endometriosis[74]. However, the underlying mechanism of oxidative stress involved in endometriosis remains obscure. Studies have underlined that oxidized lipids mediate proinflammatory or anti-inflammatory responses by modulating innate immune cells such as peritoneal macrophages[75,76]. The peritoneal fluid of patients with endometriosis has an increased number of macrophages and activated macrophage products such as TNF and interleukins, and the activated macrophages and LDL together may cause oxidative stress in the peritoneal cavity of patients with endometriosis[33]. Metabolic pathways and inflammatory processes may engage in complex cross-talk in the pelvic cavity, which is possibly responsible for physiological changes associated with oxidative stress and endometriosis. Further investigations of the mechanisms between inflammation and metabolism, and their role in shaping the peritoneal environment are essential to a deeper understanding of this disease. The immune cells in the peritoneal fluid activated by the presence of ectopic foci need vast amounts of energy and consume high levels of glucose, glutamine, ketone bodies, and fatty acids, which are markedly increased in the peritoneal fluid of patients with endometriosis. Importantly, the oxidation level of polyunsaturated fatty acid is also increased in peritoneal fluid from patients with endometriosis[30]. The SOD levels, mean activity of SOD, total antioxidant capacity, and enzyme activity are decreased in the peritoneal fluid of patients with endometriosis compared with those in controls, especially in infertile patients with endometriosis[77]. The peritoneal fluid from patients with endometriosis also has lower levels of lipid antioxidants, for example, vitamin E[34]. The concentrations of ROS and lipid peroxide are increased in patients with endometriosis, indicating that ROS may be involved in the progression of the disease[75,77]. Iron overload has been demonstrated in patients with endometriosis. Iron and copper concentrations are increased in the peritoneal fluid of patients with endometriosis compared with those in women without endometriosis, and iron-induced oxidative stress may compromise the fertility of these women[78,79]. The schematic of this section is shown in Fig. 4.

Despite extensive studies in the pathophysiology of endometriosis, the etiology and pathogenesis of endometriosis remain elusive. Current research is moving forward to seek global metabolic patterns of endometriosis. Different metabolite levels have emerged in women with endometriosis compared with those without this disease, affirming the potential of metabolites as early diagnostic and prognostic biomarkers of endometriosis. Metabolomics has become an informative technology for analyzing several diseases, although its applications in human endometriosis have some limitations. For example, comparing results between different platforms (liquid chromatography and MS) remains difficult given the diverse nature of metabolic molecules, and discriminating metabolites must be first identified and targeted, and then quantitative analysis used to confirm the results. Endometriosis sub-phenotype pathophysiologies may present different metabolism and metabolite levels, and biomarkers need to be identified for each endometriosis sub-phenotype. Although the difficulties of assessing novel biomarkers of endometriosis using metabolomics remain challenging, recent research also provides sufficient feasibility and foresight for this development[30,68]. A metabolism-altering drug has been proven to be effective for endometriosis, where vitamin E supplementation not only reduces inflammatory and oxidative stress but also relieves pain symptoms in patients with endometriosis[80]. The pain induced by oxidatively modified lipoproteins could be reduced by using antioxidants (vitamin E and N-acetylcysteine) along with nonsteroidal antiinflammatory drugs or other analgesics[34]. Due to their antiproliferative, anti-inflammatory, antioxidant, and antiangiogenic effects, statins also have the potential to treat endometriosis[81]. Overall, continuous exploration using metabolomics will strengthen our understanding and enable the discovery of novel approaches for future diagnostics and improvement of therapies for endometriosis. Acknowledgments None. Author contributions X.Q.W. and D.J.L. did conceptualization. L.Y.C., X.X.H., D.J.L., and X.Q.W. drafted the manuscript. X.Q.W. and D.J.L. revised the manuscript and approved the final version. Funding(s) This work was supported by the National Natural Science Foundation of China (81971456) to X.Q.W., National Natural Science Foundation of China (32070915) to D.J.L., and National Key Basic Research Project (2015CB943300) to D.J.L. Conflicts of interest All authors declare no conflicts of interest. All the figures were generated using BioRender software (https://www.biorender.com). Da-Jin Li is Editorial Board member of Reproductive and Developmental Medicine. The article was subject to the journal's standard procedures, with peer review handled independently of these Editorial Board member and their research groups. Data availability statement Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.

Endometriosis; Metabolites changes; Eutopic endometrium; Ectopic lesions