Integrated Eutopic Endometrium and Non‐Depleted Serum Quantitative Proteomic Analysis Identifies Candidate Serological Markers of Endometriosis

proteomics, iTRAQ, LC-MS 29 2

30

Endometriosis affects about 4 % of women in the reproductive age and is 31 associated with subfertil ity. The aim of the present study was to examine the quantitative 32 proteomic profile of eutopic endometrium and serum from women with endometriosis 33 compared to controls in order to identify candidate disease-specific serological markers. 34

Eutopic endometrium and serum from patients with endometriosis (n=8 for tissue 35 and n=4 for serum ) was respectively compared t o endometrium and serum from females 36 without endometriosis (n=8 for tissue and n=4 for serum ) using a shotgun quantitative 37 proteomics method. All study participants were at the proliferative phase of their menstrual 38 cycle. 39

At the tissue and serum level, 1,214 and 404 proteins were differentially expressed 40 (DEPs) in eutopic endometrium and serum respectively of women with endometriosis vs. 41 control. Gene ontology analysis showed that terms related to immune response | 42 inflammation, cell adhesion | migration and blood coagulation were significantly enriched in 43 the DEPs of eutopic endometrium as well as serum. Twenty-one DEPs had the same trend of 44 differential expression in both matrices and can be further examined as potential disease- and 45 tissue-specific serological markers of endometriosis. 46

The present in-depth proteomic profiling of eutopic endometrium and serum from 47 women with endometriosis identified promising serological markers that can be further 48 validated in larger cohorts for the minimally invasive diagnosis of endometriosis. 49 50 51 52 53 54 55 56 57 3

58 Endometriosis is a gynaecological condition in which endometrial glands and stroma is 59 implanted outside the uterine cavity, usually on the ovaries, Fallopian tubes and surrounding 60 tissue within the peritoneal cavity (1). Endometriosis affects approximately 3 to 4% of women 61 in the reproductive age (2) and the most common symptoms include pelvic pain, especially 62 during menstruation, and subfertility (3). 63 The exact pathophysiology of endometriosis, related to infertility i s still unknown. 64 Endometriosis can be detrimental to fertility directly by distorting tubo-ovarian anatomy (4), or 65 indirectly by invoking inflammatory (5) and oxidative damage (6,7) on the oocytes resulting in 66 poorer quality oocytes. A non- invasive method of diagnosis is not available and currently, 67 endometriosis can be only definitively diagnosed through laparoscopic surgery. Transvaginal 68 sonography (TVS), as described in the consensus statement of the International Deep 69 Endometriosis Analysis (IDEA) group, can also be used as a first -line imaging technique in 70 order to examine women with suspected endometriosis (8). 71 Endometriosis persists in both the proli ferative and secretory phases of the menstrual 72 cycle. Rai et al. (9) reported an altered endometrial proteomic profile between proliferative and 73 secretory/luteal phases of the menstrual cycle. Previous biomarker discovery studies have 74 focused primarily on the secretory phase of the menstrual cycle, at the time where there is a 75 significant level of protein turnover, modification and regeneration. The secret ory or luteal 76 phase can vary in individuals and in the context of fertility and implantation, the ‘luteal phase 77 defect’ (10), coupled with the recent evidence around the non- specific timeframe of the 78 ‘implantation window’ within the secretory phase of the menstrual cycle (11), means that the 79 proteins related to the secretory phase are much more heterogeneous and may inadvertently 80 conceal the discovery of non- menstrual cycle related endometriosis specific m arkers. For 81 these reasons, in the present study, we focused on the proliferative phase for both controls 82 and patients with endometriosis. 83 Non-targeted global proteomics, supported by recent technological advances in mass 84 spectrometry, is gradually becoming an indispensable analytical tool in clinical research since 85 4 the unbiased protein expression profiling of tissue or serum/plasma can provide novel 86 endophenotypic insight for a given pathophysiological state with unsurpassed analytical 87 confidence. Such a strategy also provides great promise in the detection of novel diagnostic, 88 prognostic and therapeutic targets that can eventually influence clinical practice (12-15). 89 There is a limited number of studies that have examined the global proteomic portrait of 90 eutopic endometrium in women with endometriosis (16-18), and the serum/plasma proteomic 91 profile of endometriosis patients ( 19-21) in order to identify tissue or blood level biomarkers 92 for the diagnosis of endometriosis . However, the integrated quantitative global proteomic 93 analysis of eutopic endometrium and non -depleted serum samples from women with 94 endometriosis for the identification of candidate tissue- and disease-specific biomarkers using 95 isobaric tags and state -of-the-art ultra-high precision LC -MS based methods has not been 96 reported to date. 97 The aim of the present study was to apply an in- depth quantitative proteomics 98 methodology in combination with comprehensive bioinformatics analysis t o eutopic 99 endometrium and serum from women with endometriosis during the proliferative phase of the 100 menstrual cycle compared to healthy controls in order to identify potential serological markers 101 for the minimally invasive diagnosis of endometriosis . An ov erview of the study workflow is 102 presented in Figure 1. 103 104

105 Data recording, sample collection and tissue storage in this study were performed 106 according to the World Endometriosis Research Foundation (WERF) Endometriosis Phenome 107 and Bioban king Harmonisation Project (EPHect) (22-24). This study has institutional and 108 regional review board approval by the University Hospital Southampton (RHMO&G160) and 109 Hampshire B ethical committees (MREC08/ HO502/162). 110 111 112 113 5 Inclusion and exclusion criteria 114 Women in the endometriosis group had a laparoscopic diagnosis of endometriosis 115 (laparoscopy or laparotomy) with the disease stage documented according to the ASRM 116 classification [Stage I: minimal; Stage II: mild; Stage III: moderate; Stage IV: severe] (25, 26). 117 Women undertaking endometrial biopsy had transvaginal ultrasonography or hysteroscopic 118 inspection of their uterine cavity and this did not reveal any endometrial pathology. Patients 119 with pelvic inflammatory disease were excluded from the study. The control group consisted 120 of women with no endometriosis as diagnosed by a negative laparoscopy. Since the 121 endometrial proteomic profile may vary in the different phases of the menstrual cycle, all 122 participants (patients with endometriosis and healthy control s) were consistently at the non-123 menstruating proliferative phase of the menstrual cycle. Subfertility was defined as trying to 124 conceive for more than 1 year without a successful outcome, while having regular sexual 125 intercourse and not using any contraceptive methods. 126 Women were excluded from the study if they were age 45 years old and above, at the 127 secretory phase of the menstrual cycle (15-28 day of menstrual cycle), on hormonal treatment 128 within three months prior to the procedure, had a BMI of more than 30, or a current smoker. 129 A systematic review and meta-analysis showed no association between smoking status and 130 the development of endometriosis (2 7). However, smokers were excluded from our study 131 because smoking has been shown to alter the blood plasma/serum proteomic profile (28). 132 Due to the small number of subjects included in the present study we would be unable to 133 correct for this potential confounder. 134 We selected women with regular cycles in order to more accurately define the 135 proliferative phase. We included women with a history of regular menstrual cycles, and 136 confirmed their stage of menstrual phase by their retrospective last menstrual date and the 137 prospective date of menstruation. This may mean we excluded women with endometriosis 138 and irregular cycles, but as menstrual cycle regularity has not been found to be significantly 139 associated with endometriosis (29 ), we do not expect this inclusion criterion to significantly 140 confound our results. 141 6 Patient recruitment 142 This study was performed at the Princess Anne Hospital, Southampton where suitable 143 candidates were given an information sheet outlining the study and signed a consent form. 144 Patients were grouped into those with endometriosis and those without (control) in accordance 145 with the findings during laparoscopy. The findings of the laparoscopy were documented in the 146 proforma. Whenever possible photographic evidence was obtained. 147 148 Endometrial tissue collection 149 Endometrial tissue was collected using endometrium sampler (Endocell®, Wallach, 150 USA). Sample collection was performed before any uterine manipulation or procedure. 151 Endometrial tissues that were suctioned in the tube were collected into individual falcon tubes 152 containing normal saline. The procedure was repeated at least twice or until an adequate 153 tissue sample was obtained. 154 155 Processing of endometrium sample and storage 156 The collected tissue samples were processed up to 4 hours from the collection. Tissues 157 were transferred into a petri dish and were gently teased apart with a tissue forceps and then 158 washed repeatedly with Phosphate Buffered Saline (PBS) to remove any blood. Healthy 159 tissues that were free from blood were cut into smaller pieces (approximately 15mm in length) 160 using a pair of tissue scissors. The processed tissues were then transferred into at least 3 161 separate Cryovials (Greiner, UK). These vials were snap frozen in -80 °C freezer. 162 Endometrium was transported on solid carbon dioxide (dry ice) inside a polystyrene box. 163 164 Quantitative proteomics sample processing 165 Two independent multiplex experiments were performed to include specimens from 16 166 subjects (n=8 controls; n=8 females with endometriosis). Specimens were dissolved in 0.5 M 167 triethylammonium bicarbonate, 0.05% sodium dodecyl sulphate and subjected to pulsed 168 probe sonication (Misonix, Farmingdale, NY, USA). Lysates were centrifuged (16,000 g, 10 169 7 min, 4oC) and supernatants were measured for protein content using infrared spectroscopy 170 (Merck Millipore, Darmstadt, Germany). Lysates were then reduced, alkylated and subjected 171 to trypsin proteolysis. Peptides were labelled using the eight -plex isobaric Tag for Relative 172 and Absolute Quantitation (iTRAQ) reagent kit (Label assignment, Experiment A: 113=control 173 1, 114=control 2, 115= control 3, 116= control 4, 117= endometriosis patient 1, 118= 174 endometriosis patient 2, 119= endometriosis patient 3, 121= endometriosis patient 4 ; 175 Experiment B: 113=control 5, 114=control 6, 115= control 7, 116= control 8, 117= 176 endometriosis patient 5, 118= endometriosis patient 6 , 119= endometriosis patient 7, 121= 177 endometriosis patient 8 ) and analysed using multi -dimensional liquid chromatography and 178 tandem mass spectrometry as reported previously by the authors (30-34). 179 180 Serum procurement and proteomic analysis 181 The procurement and handling of sera was in accordance with the recommendations of 182 the Standard Operating Procedure Integration Working Group (SOPIWG) as adopted by the 183 author’s method ( 35). One eight-plex s erum proteomics experiment was performed (n=4 184 controls; n=4 patients with endometriosis). Serum specimens were freshly thawed and 185 vortexed for 2 minutes. For each participant, 100uL of unprocessed serum were mixed with 186 400uL 6M Guanidine Hydrochloride and subjected to global quantitative serum proteomic 187 analysis using our reported depletion-free methodology (12-14). In summary, high -188 performance Size Exclusion Chromatography using three serially connected Waters KW-804 189 columns at 0.75 ml/min flow rate and 30°C was used to separate the proteins based on their 190 molecular weight differences. The separ ated low -molecular weight protein segments 191 (molecular weight cutoff 3 kDa) were dialysis purified and lyophilized to dryness. One-hundred 192 μg of protein from each sample was subjected to trypsin proteolysis and the peptides were 193 chemically labelled using the eight-plex iTRAQ reagent kit (Label assignment, 113=control 9, 194 114=control 10, 115= control 11, 116= control 12, 117= endometriosis patient 9, 118= 195 endometriosis patient 10 , 119= endometriosis patient 11 , 121= endometriosis patient 12 ), 196 pooled, and offline fractionated with high pH C4 reverse phase chromatography. Each fraction 197 8 was analysed using ultra- high performance low pH C 18 nano-liquid chromatography 198 hyphenated with high- resolution tandem mass spectrometry using the FT -Orbitrap Elite 199 platform. 200 201 Database searching 202 Unprocessed raw files were submitted to Proteome Discoverer 1.4 for target decoy 203 search against the UniProtKB homo sapiens database comprised of 20,159 entries (release 204 date January 2015), allowing for up to two missed cleavages, a prec ursor mass tolerance of 205 10ppm, a minimum peptide length of six and a maximum of two variable (one equal) 206 modifications of; iTRAQ 8-plex (Y), oxidation (M), deamidation (N, Q), or phosphorylation (S, 207 T, Y). Methylthio (C) and iTRAQ (K, Y and N -terminus) were set as fixed modifications. FDR 208 at the peptide level was set at <0.05. Percent co-isolation excluding peptides from quantitation 209 was set at 50. Reporter ion ratios from unique peptides only were taken into consideration for 210 the quantitation of the respective protein. The iTRAQ ratios of proteins were median -211 normalized and log2transformed. 212 A one- sample Student’s T- Test was performed to identify differentially expressed 213 proteins in tissue and serum samples from endometriosis patients vs. controls. Significance 214 was set at p ≤ 0.05. Only proteins with a one- sample Student’t T-Test p-value<0.05, a mean 215 iTRAQ log2ratio higher than ±0.3 and identified with at least two unique peptides in adherence 216 to the Paris Publication Guidelines for the analysis and documentation of peptide and protein 217 identifications (http://www.mcponline.org/site/misc/ParisReport_Final.xhtml), were 218 considered differentially expressed and subjected to bioinformatics analysis. All mass 219 spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via 220 the PRIDE partner repository with the dataset identifier PXD009090 (eutopic endometrium 221 proteomic analysis) and PXD011091 (serum proteomic analysis). 222 223 224 225 9 Bioinformatics analysis 226 DAVID (https://david.ncifcrf.gov/), STRING (https://string-db.org/), BiNGO in Cytoscape 227 and MetaCore (Clarivate Analytics, Philadelphia, PA, USA) were applied to differentially 228 expressed proteins in order to identify over-represented gene ontology terms, pathway maps 229 and direct protein interaction networks in endometriosis vs. control. P -values ≤ 0.05 were 230 considered significant. 231 232

233 Twenty-four patients were recruited between September 2013 and September 2015. Of 234 these, eutopic endometrium from 16 subjects was used for the tissue proteomic analysis (n=8 235 patients with endometriosis; n=8 controls) and serum from eight subjects for the serum 236 proteomics analysis (n=4 patients with endometriosis; n=4 controls). The clinical 237 characteristics of the participants are presented in Table 1. All patients were in the proliferative 238 phase and had a regular menstrual cycle. There was no significant difference in age, body 239 mass index, and baseline FSH between the two groups. 240 241 Tissue and serum proteomic analysis 242 Tissue proteomic analysis resulted in the profiling of 10,929 proteins whereas serum 243 proteomic analysis quantitatively identified 2,010 proteins (peptide FDR p<0.05). Of these, 244 1,214 ( Supplementary Table 1) and 404 ( Supplementary Table 2) were identified as 245 differentially expressed at the tissue and serum level respectively and were further subjected 246 to bioinformatics analysis. Forty-four DEPs were common between the two matrices, 21 of 247 which with the same trend of differential expression (i.e. up-regulated or down- regulated in 248 endometriosis vs. control at both tissue and serum level). 249 DAVID gene ontology analysis of the tissue and serum DEPs showed a significant 250 enrichment for gene ontology terms related to Immune response | Inflammation, Cell adhesion 251 | Migration , Blood coagulation and other terms (e.g. receptor -mediated endocytosis, high -252 density lipoprotein particle remodelling and G2/M transition of mitotic cycle) in both matrices 253 10 (Figure 2A). Forty-four DEPs were observed at both tissue and serum level and these are 254 presented in heatmap format in Figure 2B. The 21 proteins with the same trend of modulation 255 at both tissue and serum level are highlighted in grey. Ingenuity Pathway Analysis showed 256 that carbohydrate | lipid metabolism and organ development protein networks were enriched 257 in the 21 DEPs analysed in tissue and serum of patients with endometriosis vs. control (Figure 258 3). 259 260

261 The present study reports the integrated quantitative proteomic profiling of eutopic 262 endometrial tissue and non-depleted serum from women diagnosed with endometriosis 263 compared to healthy controls. Bioinformatics analysis of differentially expressed proteins 264 (DEPs) showed a significant enrichment for processes related to immune 265 response/inflammation, cell adhesion/migration, blood coagulation in both matrices, in 266 keeping with the known inflammatory and adhesive nature of endometriosis. 267 Abnormalities in immune responses have been suggested to play an important role in 268 the perpetuation of endometriosis (36, 3 7). Endometrial cells in the peritoneal cavity can 269 escape clearance from immune cells through a mechanism coined as “immunoescaping” (38). 270 Dysregulation of immune response can thus allow the proliferation, implantation and 271 angiogenesis of ectopic endometrial tissue (39). Furthermore, previous studies of peritoneal 272 fluid from patients with endometriosis have shown disease- related abnormalities in the 273 immune response (40, 41). 274 Studies have shown that eutopic endometrial stromal cells from females with 275 endometriosis exhibit an altered cell -adhesion molecular profile compared to stromal cells 276 from healthy controls (42). Extracellular matrix has been shown to control cell proliferation, 277 differentiation and apoptosis (43). 278 Two proteins were found to be up-regulated in both the eutopic endometrium and serum 279 proteomic analysis of patients with endometriosis vs. control, Na(+)/H(+) exchange regulatory 280 cofactor NHERF-1 (NHERF-1) (gene name SLC9A3R1) and thymosin beta-4 (Tb 4) (gene 281 11 name TMSB4X). Increased expression of a particular protein is more easily and reliably 282 detected compared to lower expression levels, thus these two proteins may represent the 283 most promising serological markers of endometriosis for further larger scale investigative 284 studies. 285 NHERF-1 is a scaffold protein expressed primarily in the plasma membrane of polarized 286 epithelial cells and mediates signals connecting the membrane to the cytoskeleton. The role 287 of NHERF-1 in uterine physiology remains unknown, with few studies reporting its involvement 288 with pathological conditions such as endometrial cancer and polycystic ovaries syndrome 289 (PCOS). NHERF-1 contributes to the organization of microvilli in polarized epitheliums, but 290 also regulates the acti vity of growth factor receptors, ion channels and the endocytic 291 machinery (44-46). A study showed that NHERF-1 expression is transcriptionally regulated by 292 oestrogens in human endometrium, and that it is expressed at higher levels during the 293 proliferative phase of the menstrual cycle (47). The role of NHERF-1 in endometriosis warrants 294 further investigation. 295 Tb4, a member of the beta- thymosins family, is an N -terminally acetylated peptide 296 composed of 43 amino acid residues (48). Tb4 interacts with monomeric actin (48) and 297 modulates actin polymerization (49). As a secreted factor, Tb4 has been found to modulate 298 the immune response and participate in hormonal activities (48, 50). Tb4 is also involved in 299 inflammatory response, angiogenesis, blood coagulation, wound healing and apoptosis (51-300 54). Using a mouse model, Kawahara et al. (55) showed that Tb4 over-expression could 301 participate in musculature disintegration and the development of adenomyosis . The role of 302 Tb4 in endometriosis should be assessed in future studies. 303 The main limitation of the study is its small size. Power calculation of sample size (n=16 304 for tissue analysis and n=8 for serum analysis) was based on ensuring a statistical power of 305 over 0.7, taking into consideration a 30% measurement error and a log2ratio fold change > 0.3 306 between biological replicates, as reported in a similar simulation study ( 56). Validating the 307 proteins at the tissue and serum level in a larger cohort using mass spectrometry or an 308 alternative analytical method to mass spectrometry (e.g. ELISA or western blot) to confirm 309 12 their clinical utility was beyond the scope of the present study and constitutes a future 310 perspective. 311 In conclusion, the integrated eutopic endometrium and serum global proteomic profiling 312 identified candidate serological targets that can be further validated for their clinical utility in 313 the non-invasive diagnosis of endometriosis. 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 13

338 We are indebted to Mr. Roger Allsopp, Mr. Derek Coates and Hope for Guernsey for 339 establishing the clinical mass spectrometry infrastructure at the University of Southampton. 340 The authors are grateful to the support of King Saud University, Deanship of Scien tific 341 Research Chair, Prince Mutaib Bin Abdullah Chair for Biomarkers of Osteoporosis, College of 342 Science, as well as the Visiting Professor Program of King Saud University, Riyadh, Saudi 343 Arabia. 344 345 Disclosure of interests 346 The authors declare no conflict of interest 347 348 Contribution to Authorship 349 AM performed experiments, analysed/ interpreted data and wrote manuscript; MH collected 350 samples, performed experiments, analysed/ interpreted data; MF, DJGB and JT performed 351 experiments and analysed data; SDG and YC designed study, supervised the execution of 352 experiments, interpreted the experimental results and wrote manuscript. 353 354 Details of ethics approval 355 This study has institutional and regional review board approval by the University Hospital 356 Southampton (RHMO&G160) and Hampshire B ethical committees (MREC08/ HO502/162) 357 (Approval date: 24 October 2008). 358 359 Funding 360 JT was supported by the China Scholarship Council and the China Postdoctoral Science 361 Foundation (2013T60260). 362 363 364 365 14

366 367 1. Dunselman GA, Vermeulen N, Becker C et al. (2014) ESHRE guideline: management of 368 women with endometriosis. Hum Reprod 29, 400-412. 369 2. Buck Louis GM, Hediger ML, Peterson CM et al. (2011) Incidence of endometriosis by study 370 population and diagnostic method: the ENDO study. Fertil Steril 96, 360-365. 371 3. Zondervan KT, Becker CM, Koga K et al. (2018) Endometriosis. Nat Rev Dis Primers 4:9. 372 4. Young VJ, Brown JK, Saunders PT et al. (2013) The role of the peritoneum in the 373 pathogenesis of endometriosis. Hum Reprod Update 19, 558-569. 374 5. Gazvani R, Templeton A (2002) Peritoneal environment, cytokines and angiogenesis in the 375 pathophysiology of endometriosis. Reproduction 123, 217-226. 376 6. Hamdan M, Jones KT, Cheong Y et al. (2016) The sensitivity of the DNA damage 377 checkpoint prevents oocyte maturation in endometriosis. Sci Rep 6, 36994. 378 7. Matsuzaki S, Schubert B (2010) Oxidative stress status in normal ovarian cortex 379 surrounding ovarian endometriosis. Fertil Steril 93, 2431-2432. 380 8. Guerriero S, Condous G, van den Bosch T et al. (2016) Systematic approach to sonographic 381 evaluation of the pelvis in women with suspected endometriosis, including terms, definitions 382 and measurements: a consensus opinion from the International Deep Endometriosis Analysis 383 (IDEA) group. Ultrasound Obstet Gynecol 48, 318-332. 384 9. Rai P, Kota V, Sundaram CS et al. (2010) Proteome of human endometrium: Identification 385 of differentially expressed proteins in proliferative and secretory phase endometrium. 386 Proteomics Clin Appl 4, 48-59. 387 10. Schliep KC, Mumford SL, Hammoud AO et al. (2014) Luteal phase deficiency in regularly 388 menstruating women: prevalence and overlap in identification based on clinical and 389 biochemical diagnostic criteria. J Clin Endocrinol Metab 99, E1007-1014. 390 11. Tan J, Kan A, Hitkari J et al. (2018) The role of the endometrial receptivity array (ERA) in 391 patients who have failed euploid embryo transfers. J Assist Reprod Genet 35, 683-692. 392 12. Al -Daghri NM, Manousopoulou A, Alokail MS et al. (2018) Sex-specific correlation of 393 15 IGFBP-2 and - 3 with vitamin D status in adults with obesity: A cross -sectional serum 394 proteomics study. Nutr Diabetes [Epub ahead of print] 395 13. Manousopoulou A , Scorletti E, Smith DE et al. (2018) Marine omega- 3 fatty acid 396 supplementation in non- alcoholic fatty liver disease: Plasma proteomics in the randomized 397 WELCOME* trial. Clin Nutr doi: 10.1016/j.clnu.2018.07.037. [Epub ahead of print] 398 14. Zeidan B, Manousopoulou A, Garay-Baquero DJ et al. (2018) Increased circulating resistin 399 levels in early-onset breast cancer patients of normal body mass index correlate with lymph 400 node negative involvement and longer disease free survival: a multi -center POSH cohort 401 serum proteomics study. Breast Cancer Res 20, 19. 402 15. Manousopoulou A, Hayden A, Mellone M et al. (2018) Quantitative proteomic profiling of 403 primary cancer -associated fibroblasts in oesophageal adenocarcinoma. Br J Cancer 118, 404 1200-1207. 405 16. Siva AB, Srivastava P, Shivaji S (2014) Understanding the pathogenesis of endometriosis 406 through proteomics: recent advances and future prospects. Proteomics Clin Appl 8, 86-98. 407 17. Ten Have S, Fraser I, Markham R et al. (2007) Proteomic analysis of protein expression 408 in the eutopic endometrium of women with endometriosis. Proteomics Clin Appl 1, 1243-51. 409 18. Marianowski P, Szymusik I, Malejczyk J et al. (2013) Proteomic analysis of eutopic and 410 ectopic endometriotic tissues based on isobaric peptide tags for relative and absolute 411 quantification (iTRAQ) method. Neuro Endocrinol Lett 34, 717-21. 412 19. Hwang JH, Lee KS, Joo JK et al. (2014) Identification of biomarkers for endometriosis in 413 plasma from patients with endometriosis using a proteomics approach. Mol Med Rep 10, 725-414 30. 415 20. Long X, Jiang P, Zhou L et al. (2013) Evaluation of novel serum biomarkers and the 416 proteomic differences of endometriosis and adenomyosis using MALDI -TOF-MS. Arch 417 Gynecol Obstet 288, 201-5. 418 21. Seeber B , Sammel MD, Fan X et al. (2010) Proteomic analysis of serum yields six 419 candidate proteins that are differentially regulated in a subset of women with endometriosis. 420 Fertil Steril 93, 2137-44. 421 16 22. Fassbender A, Rahmioglu N, Vitonis AF et al. (2014) World Endometriosis Research 422 Foundation Endometriosis Phenome and Biobanking Harmonisation Project: IV. Tissue 423 collection, processing, and storage in endometriosis research. Fertil Steril 102, 1244-1253. 424 23. Rahmioglu N, Fassbender A, Vitonis AF et al. (2014) World Endometriosis Research 425 Foundation Endometriosis Phenome and Biobanking Harmonization Project: III. Fluid 426 biospecimen collection, processing, and storage in endometriosis research. Fertil Steril 102, 427 1233-1243. 428 24. Vitonis AF, Vincent K, Rahmioglu N et al. (2014) World Endometriosis Research 429 Foundation Endometriosis Phenome and Biobanking Harmonization Project: II. Clinical and 430 covariate phenotype data collection in endometriosis research. Fertil Steril 102, 1223-1232. 431 25. American Society for Reproductive Medicine (1997) Revised American Society for 432 Reproductive Medicine classification of endometriosis: 1996. Fertil Steril 67, 817-821 433 26. Johnson NP, Hummelshoj L, Adamson GD et al. (2017) 434 World Endometriosis Society consensus on the classification of endometriosis. Hum Reprod 435 32, 315-324 436 27. Bravi F, Parazzini F, Cipriani S et al. (2014) Tobacco smoking and risk of endometriosis: 437 a systematic review and meta-analysis. BMJ Open 4, e006325. 438 28. Bortner JD Jr, Richie JP Jr, Das A et al. (2011) Proteomic profiling of human plasma by 439 iTRAQ reveals down-regulation of ITI-HC3 and VDBP by cigarette smoking. J Proteome Res 440 10, 1151-1159. 441 29. Parazzini F, Cipriani S, Bianchi S et al. (2008) Risk factors for deep endometriosis: a 442 comparison with pelvic and ovarian endometriosis. Fertil Steril 90, 174-9. 443 30. Billiard F, Karaliota S, Wang B et al. (2018) Delta-like Ligand-4-Notch Signaling Inhibition 444 Regulates Pancreatic Islet Function and Insulin Secretion. Cell Rep 22, 895-904. 445 31. Nguyen TL, Duchon A, Manousopoulou A et al. (2018) Correction of cognitive deficits in 446 mouse models of Down syndrome by a pharmacological inhibitor of DYRK1A. Dis Model Mech 447 11, pii: dmm035634. 448 17 32. Manousopoulou A, Koutmani Y, Karaliota S et al. (2016) Hypothalamus proteomics from 449 mouse models with obesity and anorexia reveals therapeutic targets of appetite regulation. 450 Nutr Diabetes 6, e204. 451 33. Manousopoulou A, Gatherer M, Smith C et al. (2017) Systems proteomic analysis reveals 452 that clusterin and tissue inhibitor of metalloproteinases 3 increase in leptomeningeal arteries 453 affected by cerebral amyloid angiopathy. Neuropathol Appl Neurobiol 43, 492-504. 454 34. Manousopoulou A, Woo J, Woelk CH et al. (2015) Are you also what your mother eats? 455 Distinct proteomic portrait as a result of maternal high -fat diet in the cerebral cortex of the 456 adult mouse. Int J Obes (Lond) 39, 1325-1328. 457 35. Tuck MK, Chan DW, Chia D et al. (2009) Standard operating procedures for serum and 458 plasma collection: early detection research network consensus statement standard operating 459 procedure integration working group. J Proteome Res 8, 113-7. 460 36. Christodoulakos G, Augoulea A, Lambrinoudaki I et al. (2007) Pathogenesis of 461 endometriosis: the role of defective 'immunosurveillance'. Eur J Contracept Reprod Health 462 Care 12, 194-202. 463 37. Riccio LGC, Baracat EC, Chapron C et al. (2017) The role of the B lymphocytes in 464 endometriosis: A systematic review. J Reprod Immunol 123, 29-34. 465 38. Vetvicka V, Lagana AS, Salmeri FM et al. (2016) Regulation of apoptotic pathways during 466 endometriosis: from the molecular basis to the future perspectives. Arch Gynecol Obstet 294, 467 897-904. 468 39. Matarese G, De Placido G, Nikas Y et al. (2003) Pathogenesis of endometriosis: natural 469 immunity dysfunction or autoimmune disease? Trends Mol Med 9, 223-228. 470 40. Wang XM, Ma ZY, Song N (2018) Inflammatory cytokines IL-6, IL-10, IL-13, TNF-α and 471 peritoneal fluid flora were associated with infertility in patients with endometriosis. Eur Rev 472 Med Pharmacol Sci 22, 2513-2518. 473 41. Sikora J, Ferrero S, Mielczarek -Palacz A et al. (2018) The Delicate Balance between 474 the Good and the Bad IL-1 Proinflammatory Effects in Endometriosis. Curr Med Chem 25, 475 2105-2121. 476 18 42. Klemmt PA, Carver JG, Koninckx P et al. (2007) Endometrial cells from women with 477 endometriosis have increased adhesion and proliferative capacity in response to extracellular 478 matrix components: towards a mechanistic model for endometriosis progression. Hum Reprod 479 22, 3139-3147. 480 43. Frisch SM, Francis H (1994) Disruption of epithelial cell -matrix interactions induces 481 apoptosis. J Cell Biol 124, 619-626. 482 44. Bretscher A, Chambers D, Nguyen R et al. (2000) ERM-483 Merlin and EBP50 protein families in plasma membrane organization and function. Annu Rev 484 Cell Dev Biol 16, 113-43. 485 45. Morales FC, Takahashi Y, Kreimann EL et al. (2004) Ezrin-radixin-moesin (ERM)-486 binding phosphoprotein 50 organizes ERM proteins at the apical membrane of 487 polarized epithelia. Proc Natl Acad Sci U S A 101, 17705-10. 488 46. Shenolikar S, Voltz JW, Cunningham R et al. (2004) Regulation of ion transport by 489 the NHERF family of PDZ proteins. Physiology (Bethesda) 19, 362-9. 490 47. Reczek D, Berryman M, Bretscher A (1997) Identification of EBP50: A PDZ-491 containing phosphoprotein that associates with members of the ezrin-radixin-moesin family. 492 J Cell Biol 139, 169-79. 493 48. Mannherz HG, Hannappel E (2009) The beta-thymosins: intracellular and extracellular 494 activities of a versatile actin binding protein family. Cell Motil Cytoskeleton 66, 839-51. 495 49. Dominguez R (2007) The beta-thymosin/WH2 fold: multifunctionality and structure. Ann N 496 Y Acad Sci 1112, 86-94. 497 50. Paulussen M, Landuyt B, Schoofs L et al. (2009) 498 Thymosin beta 4 mRNA and peptide expression in phagocytic cells of different mouse 499 tissues. Peptides 30, 1822-32. 500 51. Young JD, Lawrence AJ, MacLean AG et al. (1995) Thymosin beta 4 sulfoxide is an anti-501 inflammatory agent generated by monocytes in the presence of glucocorticoids. Nat Med 5, 502 1424-7. 503 19 52. Smart N, Rossdeutsch A, Riley PR (2007) Thymosin beta4 and angiogenesis: modes of 504 action and therapeutic potential. Angiogenesis 10, 229-41. 505 53. Sosne G, Szliter EA, Barrett R et al. (2002) Thymosin beta 4 promotes corneal wound 506 healing and decreases inflammation in vivo following alkali injury. Exp Eye Res 74, 293-9. 507 54. Safer D, Golla R, Nachmias VT (1990) Isolation of a 5-kilodalton actin-508 sequestering peptide from human blood platelets. Proc Natl Acad Sci U S A 87, 2536-40. 509 55. Kawahara R, Matsuda M, Imaoka T et al. (2003) Up-regulation of thymosin beta 4 gene 510 expression in experimentally-induced uterine adenomyosis in mice. In Vivo 17, 561-5. 511 56. Levin Y (2011) The role of statistical power analysis in quantitative proteomics. Proteomics 512 11, 2565–7. 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 20 Table and Figure Legends 531 532 Table 1. Clinical characteristics of study participants 533 534 Figure 1. Study design 535 Figure 2. A. DAVID gene ontology analysis of the DEPs showed a significant enrichment for 536 gene ontology terms related to Immune response | Inflammation, Cell adhesion | Migration, 537 Blood coagulation in both tissue and serum from patients with endometriosis vs. control. B. 538 Heatmap of fourty-four DEPs observed at both tissue and serum level. Proteins with the same 539 trend of modulation at both matrices are highlighted in grey. 540 Figure 3. Ingenuity Pathway Analysis showed that carbohydrate | lipid metabolism and organ 541 development protein networks were enriched in the 21 DEPs analysed in tissue and serum of 542 patients with endometriosis vs. control 543 544