Effect of Endometrial Biopsy on Uterine Health in Tropically Adapted Beef Cattle

Effect of endometrial biopsy on uterine health in tropically adapted beef cattle

Helbert Orlando Ramírez-Garzon

DVM, MSc

A thesis submitted for the degree of Doctor of Philosophy at

The University of Queensland in 2017

School of Veterinary Science

Abstract

In cattle, early embryonic loss during the first two to three weeks after fertilisation is recognised as the major cause of reproductive failure, with consequent significant impacts on business performance of beef and dairy farms. Inadequate endometrial receptivity is considered an important cause of early embryonic mortality in cattle. A pilot experiment was conducted to compare the outcome of performing different endometrial sampling procedures in tropically adapted (Bos indicus x Bos taurus) heifers, and to select the most suitable method to collect samples for investigation of endometrial gene expression. Endometrial biopsy (EB) using a biopsy device with a diameter similar to an artificial insemination gun (10366L Karl Storz®, Germany) was found to be the most reliable method of collecting endometrial samples in these heifers, with the samples recovered yielding a mean concentration of RNA of 72.9 ±39.3 ng/µl and mean RNA integrity number (RIN) of 7.3 ±1.1.

To assess the likely safety of conducting EB in cattle a systematic review of published information assessing the impact of EB on subsequent pregnancy rates was conducted. Using the Preferred Reporting Items for Systematic reviews and Meta-Analysis (PRISMA) protocol, online searches identified 1164 possible studies. After applying the exclusion criteria, six studies allowed the comparison of pregnancy rates between sampled (n= 1,208) and non-sampled cows (n=2,393). The results showed that EB can be performed before breeding without adversely affecting the likelihood of cows becoming pregnant.

A major objective of the research was to determine whether EB and evaluation of endometrial gene expression could be used to identify tropically adapted embryo transfer (ET) recipients more likely to become pregnant. Three potential EB scenarios were investigated, 1) EB on Day 7 of the oestrous cycle prior to the cycle when ET would be performed, 2) EB on Day 4 of the oestrous cycle in which ET would be performed i.e. EB 3 days before ET, and 3) EB on the same day when ET would be performed.

The first study assessed the morphological and inflammatory changes induced by endometrial biopsy. Pre-synchronised heifers underwent EB from the horn ipsilateral to the corpus luteum (CL) either at Day 7 of the previous cycle (Group 1 n=10) followed by administration of a luteolytic dose of prostaglandin F2α or at Day 4 (Group 2, n=7) of the present cycle. All heifers were slaughtered on Day 7 of the present cycle. The site of EB was visually evident in 40% (4/10) of Group 1 and 42% (3/7) of Group 2 heifers. Histological assessment revealed no evidence of changes indicative of endometritis in either the biopsied or non-biopsied horns in both groups. The histological

changes observed included mild mononuclear infiltration of the epithelium and mild vascular changes associated with haemorrhage and congestion.

Initial investigation of the endometrial expression of the pro-inflammatory cytokine IL1β and wound healing cytokine TGFβ genes found no significant difference between biopsied and non- biopsied horns for both Group 1 (P=0.8) and Group 2 (P=0.125) heifers. However, expression of IL1β in the biopsied horns of Group 2 were greater than that detected in Group 1 horns (P=0.031, P<0.05).

Subsequently, a comprehensive assessment of the expression of pro-inflammatory and wound healing genes was conducted using a customised array. Whilst differences in gene expression between the biopsied and non-biopsied horns 3 or 10 days after were not different (P >0.05), the cytokine profile 3 days after biopsy suggested up-regulation of pro-inflammatory cytokines, whereas the inflammatory response 10 days after biopsy suggested up-regulation of wound healing genes. Further, there were no significant differences in gene expression between control horns and the biopsied and non-biopsied horns. This study indicates that the trauma caused by EB has only a short-term minor effect on uterine health.

A study was conducted to investigate whether there were any ultrasonographic changes in the uterus following EB either at Day 4 (Group 1 n=5) or at Day 7 (Group 2 n=8) after oestrus. Examination of images of the uterus before and after EB, demonstrated that regardless of the day of EB approximately 55 % of heifers had ultrasonographic evidence of some fluid accumulation in the lumen of the uterus 48-72 h after biopsy.

In the final study we assessed whether EB affected CL lifespan. Tropically adapted beef heifers were biopsied either at Day 4 or at Day 7 post-oestrus, and blood samples were collected during the oestrous cycle to measure the concentrations of progesterone (P4) and the prostaglandin F2α metabolite 13,14-dihydro-15-keto-PGF2α (PGFM). The results showed that oestrous cycle length was not affected by EB conducted at Day 4 or at Day 7, and the procedure did not induce the release of luteolytic pulses of PGF2α. Further, there was no evidence of a significant decline in P4 associated with EB and P4 concentrations remained above 2ng/ml until Day 18. In summary, we conclude that conducting EB in tropically adapted beef heifers using the Storz® device is a safe procedure that does not adversely affect uterine health or ovarian function, and yields sufficient tissue to enable assessment of endometrial receptivity.

iii

Declaration by author

This thesis is composed of my original work, and contains no material previously published or written by another person except where due reference has been made in the text. I have clearly stated the contribution by others to jointly-authored works that I have included in my thesis.

I have clearly stated the contribution of others to my thesis as a whole, including statistical assistance, survey design, data analysis, significant technical procedures, professional editorial advice, and any other original research work used or reported in my thesis. The content of my thesis is the result of work I have carried out since the commencement of my research higher degree candidature and does not include a substantial part of work that has been submitted to qualify for the award of any other degree or diploma in any university or other tertiary institution. I have clearly stated which parts of my thesis, if any, have been submitted to qualify for another award.

I acknowledge that an electronic copy of my thesis must be lodged with the University Library and, subject to the policy and procedures of The University of Queensland, the thesis be made available for research and study in accordance with the Copyright Act 1968 unless a period of embargo has been approved by the Dean of the Graduate School.

I acknowledge that copyright of all material contained in my thesis resides with the copyright holder(s) of that material. Where appropriate I have obtained copyright permission from the copyright holder to reproduce material in this thesis.

Publications during candidature

- Peer reviewed papers: - Ramirez-Garzon, O., Satake, N., Lyons, R.E., Hill, J., Holland, M.K.& McGowan, M. Endometrial biopsy in Bos indicus beef heifers. Reprod Dom Anim.2017;00:1–3. https://doi.org/10.1111/rda.12944 (Published). - Conference abstracts: - Ramirez-Garzon O., Satake N., Lyons R. E., Palmieri C., Hill J., Gallego-Lopez C., Holland M. K., McGowan M. (2016) 63. Effect of endometrial biopsy on uterine health of tropically adapted beef cattle. Reproduction, Fertility and Development 29, 139-139 Publications included in this thesis

- Ramirez-Garzon, O., Satake, N., Lyons, R.E., Hill, J., Holland, M.K.& McGowan, M. Endometrial biopsy in Bos indicus beef heifers. Reprod Dom Anim.2017;00:1–3. https://doi.org/10.1111/rda.12944 incorporated as Appendix 1.

Contributor Statement of contribution

Ramirez-Garzon Designed and conducting experiments (60%)

Wrote and edited the paper (50%)

Satake, Nana Drafting and critical revision (50%)

Wrote and edited paper (30%)

Lyons, RE Drafting and critical revision (20%)

Hill, J Design and conducting experiments (10%)

Drafting and critical revision (10%)

Holland, M.K. Drafting and critical revision (20%)

McGowan, M Designed and conducting experiments (30%)

Wrote and edited paper (20%)

Contributions by others to the thesis

No contribution by others

Statement of parts of the thesis submitted to qualify for the award of another degree

None

iii

Acknowledgements

I would first and foremost like to thank Professor Michael Holland and Michael McGowan, principal supervisors who gave me the resources, unconditional support and guidance throughout my PhD studies. I would also like to thank my co-supervisors, Professor Jonathan Hill and Dr Russell Lyons for their valuable feedback and suggestions given to improve my work and critically reviewing of this manuscript. Special mention to Dr Nana Satake for her time and endless support during the pursuit of this degree. Her supervision was invaluable and words cannot fully express my gratitude for helping me. The financial support for these studies came from the Department of Science, Technology and Innovation, COLCIENCIAS, Colombia. I would also like to acknowledge the Graduate School for the fee scholarship during the last year. Also, I would like to thank Dr Gry Boe-Hansen for the analysis of ultrasound images, Dr Nancy Phillips for the collection of endometrial samples, Dr Chiara Palmieri and Dr Carolina Gallego for the assessment of the histology slides, Dr John Al-Alawneh and Elisa Girola for the support and advice for the statistical analysis, Dr Tamara Keeley for carrying out the PGFM experiment and to Ricardo Soares for his assistance in the systematic review. I wish to express my thanks to Tom Conolly, the herd manager at the Pinjarra Hills research station, for the assistance in practical arrangements during the experiments. All my gratitude to all the fellow graduate and undergraduate students for their generous support throughout this project.

I was surrounded by good friends and colleagues in the Vet school: Andres, Matias, Carlos, Willy and Des. The multicultural Colombian-Latin-Worldwide community of soccer, running, parties, BBQ’s & housemates. Hanging out with all of you during this long journey was priceless. Special thanks to Gonzalo, Vivian, Lina, Gloria, Angelita, Diego, Sofi, Pipe, La Tati po, etc etc. Thank you ALL for being so supportive especially during the hardest time of thesis-writing.

Special mention to My Little Havana family: ‘After a tough moment in life, I found a therapist. Her name is Salsa’. Well done team!”

Lastly, I would like to thank my friends back in Colombia and around the world (Diana, Mono, MP, YK, Marin, Cata, Pili) for being there 24/7 to listen when I needed an ear and to my wonderful family-mom, dad, siblings and nephews- for their love and unconditional support in all these years through my PhD.

Keywords

Beef cattle, uterine health, endometrial biopsy, histopathology, endometritis, gene expression, progesterone, PGFM

Australian and New Zealand Standard Research Classifications (ANZSRC)

ANZSRC code: 070206 Animal Reproduction 70%

ANZSRC code: 070703 Veterinary Diagnosis and Diagnostics 20%

ANZSRC code: 070705 Veterinary Immunology 10%

Fields of Research (FoR) Classification

0702 Animal Production 70%

0707 Veterinary Sciences 30%

Table of Contents

Abstract ...... ii

Declaration by author...... iv

Publications during candidature ...... ii

Publications included in this thesis...... ii

Contributions by others to the thesis ...... iii

Statement of parts of the thesis submitted to qualify for the award of another degree ...... iii

Acknowledgements ...... iv

Keywords ...... v

Australian and New Zealand Standard Research Classifications (ANZSRC) ...... v

Fields of Research (FoR) Classification ...... v

Table of Contents ...... vi

List of Figures ...... ix

List of Tables ...... xiii

List of Abbreviations used in the thesis ...... xv

Introduction ...... xvii

CHAPTER 1. REVIEW OF LITERATURE ...... 1

1.1 The bovine uterus ...... 2

1.2 Bovine oestrous cycle...... 3

1.2.1 Role of progesterone on embryo development and survival ...... 7

1.2.2 Luteolysis ...... 10

1.3 Embryo mortality and endometrial receptivity ...... 11

1.4 Uterine assessment for endometrial receptivity...... 15

1.4.1 Non-invasive uterine assessment of endometrial receptivity ...... 15

1.4.2 Invasive uterine assessment ...... 16

1.5 Effects of endometrial sampling on uterine health ...... 28

1.5.1 Systematic review of the effect of different methods of uterine sampling on subsequent pregnancy rates ...... 31

1.6 Aims and objectives of the thesis ...... 44

CHAPTER 2. INVESTIGATION OF THE PRACTICAL EFFICACY OF DIFFERENT UTERINE SAMPLING METHODS ...... 45

2.1 Evaluation of different methods of uterine sampling from tropically adapted beef heifers 47

2.1.1 Introduction ...... 48

2.1.2 Material and methods ...... 50

2.1.3 Results ...... 55

2.1.4 Discussion ...... 58

CHAPTER 3. INVESTIGATION OF THE IMPACT OF ENDOMETRIAL BIOPSY ON ENDOMETRIAL HEALTH ...... 62

3.1 Histological response to endometrial biopsy ...... 65

3.1.1 Introduction ...... 66

3.1.2 Materials and Methods ...... 67

3.1.3 Results ...... 70

3.1.4 Discussion ...... 76

3.2 Inflammatory gene response to endometrial biopsy...... 79

3.2.1 Introduction ...... 81

3.2.2 Materials and Methods ...... 84

3.2.3 Results ...... 91

3.2.4 Discussion ...... 99

3.3 Ultrasonographic changes in uterine horns after endometrial biopsy ...... 103

3.3.1 Introduction ...... 104

3.3.2 Material and Methods ...... 105

3.3.3 Results ...... 108

3.3.4 Discussion ...... 113

vii

CHAPTER 4. INVESTIGATION OF THE IMPACT OF ENDOMETRIAL BIOPSY ON CORPUS LUTEUM FUNCTION ...... 115

4.1 Introduction ...... 117

4.2 Material and Methods...... 121

4.2.1 Animals and management...... 121

4.2.2 Synchronisation of oestrus: ...... 121

4.2.3 Experimental design ...... 121

4.2.4 Blood sampling...... 122

4.2.5 Progesterone assessment ...... 122

4.2.6 PGFM assessment ...... 122

4.2.7 Statistical analysis ...... 123

4.3 Results ...... 125

4.3.1 Experiment 1: Transrectal manipulation of uterine tract and placement of biopsy device within the uterine horns...... 125

4.3.2 Experiment 2: Endometrial biopsy at Day +4 (B4) or Day +7 (B7) post-oestrus...... 129

4.4 Discussion ...... 132

CHAPTER 5. GENERAL DISCUSSION ...... 137

5.1 Future studies ...... 144

References...... 146

Appendix A. Technical note ...... 162

Appendix B. Abstract conference ...... 162

Appendix C. Progesterone assessment ...... 163

Appendix D. PGFM assessment ...... 167

viii

List of Figures

Figure 1.1 Uterine and ovarian changes during oestrous cycle in the cow. Adapted from (Eurell and Frappier, 2013).P4: Progesterone; E2: Oestrogens...... 4 Figure 1.2 Early embryonic growth and conceptus elongation (Adapted from Spencer (2013). The period of conceptus elongation in bovine embryos is highlighted...... 8 Figure 1.3 Cytobrush device covered with a plastic sheath designed for cattle (Cytology Brush; Minitube GmbH, Germany)...... 18 Figure 1.4 a) Ghent device® for aspiration of luminal fluid. Outer-inner tube 50 cm length with a flexible post-cervical insemination catheter x70 cm length b) Uterine lavage (UL) using Ghent device®. The cannula was inserted via transvaginal into the uterine horn and the fluid was collected by aspiration using a 50 ml syringe...... 21 Figure 1.5 Biopsy devices used for uterine sampling in cattle a)Ten Thije’s model b) Folmer Nielsen’s model (Oskerven, 1956) c) Storz® device d) Hauptner Yeoman Uterine Biopsy Punch. 23 Figure 1.6 Biopsy Storz® device (10366L Karl Storz, Germany) with round cup single action jaws, 2.5 mm, length 55 cm; b) Bovine uterine tissue collected using different devices. Left-hand side, Storz® device, right-hand side, Yeoman Hauptner® ...... 24 Figure 1.7 Histopathological criteria for endometritis diagnosis (Meira et al., 2012, permission requested)...... 27 Figure 1.8 Selection process of papers for systematic review according to PRISMA guidelines. .... 35 Figure 2.1 Oestrus synchronisation protocol used in crossbreed heifers/cows: Day -9 Insertion of intravaginal P4 releasing device (IPRID, Cue-mate 0.786 g) and Oestradiol Benzoate (OB), 1mg, IM. Day -2: Prostaglandin F2α analogue (PGF2α, Cloprostenol 250 µg IM) and Day -1: Oestradiol Benzoate (OB, 1mg IM). Day 0: Oestrus ...... 51 Figure 3.1 Experimental design for biopsy procedures performed during the prior cycle (B7-O-T7) or during ongoing cycle (B4-T7). B7: Biopsy at Day 7 post-oestrus. O: Oestrus. T7: Reproductive tracts at Day 7 post-oestrus. B4: Biopsy at Day 4 post-oestrus...... 68 Figure 3.2 Gross abnormalities in biopsied uteri collected from (a) B7-O-T7 and (b-d) B4-T7 heifers. (a,b) The biopsied spot is grossly evident as a focal endometrial red discoloration suggestive of an underlying inflammatory process and/or circulatory disorder (arrows);(c) Serosa congestion (d): In one case, a severe locally extensive dark red haemorrhage was observed (arrow)...... 72 Figure 3.3 Histopathological assessment in B7-O-T7 heifers: (a) Normal tissue, 4X (b) Loss of epithelium and slight oedema and haemorrhage in submucosa layer, 4X. B4-T7 heifers: (c.d) Loss of epithelium, focal haemorrhage in submucosa layer, presence of macrophages containing ix

hemosiderin 4X&10X (e) Slight glandular dilation and moderate periglandular fibrosis 10X (f) Mononuclear infiltration in glandular epithelium 40X (g-h) Vascular congestion 10X. H&E ...... 74 Figure 3.4 Fold change in endometrial gene expression of IL1β in biopsied and non-biopsied horns at 3 days (B4-T7) and 10 days (B7-O-T7) after biopsy compared to non-biopsied heifers. BH: Biopsied horn, NBH; Non-biopsied horn. B4 (Biopsy at Day 4 post-oestrus), B7: (Biopsy at Day 7 post-oestrus); O: (Oestrus); T7: (Endometrial post-mortem tissue at Day 7 post-oestrus)...... 91 Figure 3.5 Fold change in endometrial gene expression of TGFß in biopsied and non-biopsied horns in heifers biopsied 3 days (B4-T7) or 10 days (B7-O-T7) previously. BH: Biopsied horn, NBH; Non-biopsied horn. B4 (Biopsy at Day 4 post-oestrus), B7: (Biopsy at Day 7 post-oestrus); O: (Oestrus); T7: (Endometrial post mortem tissue at Day 7 post-oestrus)...... 92 Figure 3.6 Comparison (mean ±SE) of endometrial gene expression at Day 4 (b4) and at Day 7 (b7). Gene expression were normalised against internal control GAPDH expressed as ΔCt., CSF2: Colony stimulating factor 2; IL1R1: Interleukin 1 receptor, Type 1, IL6: Interleukin 6,: PTGS2: Prostaglandin endoperoxidase synthase 2; SPP1: Osteopontin; TGFB1: Transforming growth factor beta 1; TNF: Tumor necrosis factor; VEGFA: Vascular endothelial growth factor A Comparison (mean ±SE) of endometrial gene expression of beef heifers at Day 4 (b4) and at Day 7 (b7)...... 93 Figure 3.7 Gene expression of endometrial pro-inflammatory and wound healing cytokines (mean±SE) from biopsies collected at Day 7 post-oestrus and biopsied (T7b) and non-biopsied horn (T7nb) at Day 7 post-oestrus. Gene expression was normalised against internal control GAPDH expressed as ΔCt ...... 94 Figure 3.8 Fold change expression of inflammatory genes between biopsied (T7b) and non-biopsied horn (T7nb) at Day 7 post-oestrus, 10 days after injury. Relative gene expression for every transcript was calculated using the 2- ΔΔ Ct using GAPDH as housekeeping gene. Each bar represents the mean +/− SEM of six animals...... 95 Figure 3.9 Comparison of inflammatory gene expression at Day 4 (mean±SE) post-oestrus and biopsied (B7b) and non-biopsied horn at Day 7 post-oestrus. Gene expression was normalised against internal control GAPDH expressed as ΔCt...... 96 Figure 3.10 Fold change expression of inflammatory genes between biopsied (B7b) and non- biopsied horn (B7nb) at Day 7 post-oestrus, 3 days after injury. Relative gene expression for every transcript was calculated using the 2- ΔΔ Ct using GAPDH as housekeeping gene. Each bar represents the mean +/− SEM of five animals...... 97 Figure 3.11 Comparison of gene expression (mean±SE) in biopsied and non-biopsied horns in B4- B7 and B7-O-T7. B4b (Biopsied horn at Day 4 post-oestrus) B4nb (Non-Biopsied horn at Day 4 post-oestrus) T7b (Post-mortem uterine tissue biopsied at Day 7 post-oestrus) T7nb (Post-mortem uterine tissue non-biopsied at Day 7 post-oestrus)...... 98 x

Figure 3.12 Scanning of uterine tract for presence of uterine fluid by transrectal ultrasound (US) in heifers biopsied at Day +4 (B4, n=6) and Day +7 (2xB7, n=8) and non-biopsied controls (n=3). For all groups, the first uterine scanning was performed at Day -7 of synchronised protocol which corresponds to the day of CIDR removal and the injection of PGF2α to induce oestrus that occur 48-72 h after injection (Day of oestrus=Day 0). B4 heifers were scanned at Day -7 and Day +4 (before biopsy) and Day +7 (after biopsy). 2xB7 heifers were scanned at Day -7 and Day +7 (before biopsy) and Day +9, +11, +11, +13, +16 (after biopsy). Control heifers were scanned at Day -7, Day +7, Day +9, D+13 and D+16...... 106 Figure 3.13 Ultrasonographic images of uterine horns collected during the study. Intraluminal non- echogenic (black) fluid collections represent categories (0-4) after EB...... 107 Figure 3.14 Comparison of scores (±SD) given by Observer 1 for the assessment of accumulation of uterine fluid in heifers biopsied at Day 4, Day 7 and non-biopsied (control).* P< 0.05 ...... 112 Figure 3.15 Comparison of scores (±SD) given by Observer 1 for the assessment of accumulation of uterine fluid in heifers biopsied at Day 4, Day 7 and non-biopsied (control).* P< 0.05...... 112 Figure 4.1 Individual variations in plasma PGFM (pg/ml) concentration in biopsied heifers# 9 and # 11 from ...... 124 Figure 4.2 a) P4 (ng/ml) measured from Day 3 to 20 of oestrous cycle (n=9). Five heifers (in red) were subjected to a UM at Day+4 of oestrous cycle, while remaining four heifers (in blue) were subjected to a UM at Day 7 of .b) Observed and fitted values for the model described in section 4.1.2.6 showing differences in P4 between treatments (UM4 and UM7). P4 for each treatment is plotted as a function of DOE (Day of oestrous). Each treatment is represented by a different colour. Observed values are plotted as dots. Fitted values are plotted as solid lines and represent the effect of DOE on P4 for each treatment. The shaded areas represented the 95 % point wise confidence intervals...... 126 Figure 4.3 PGFM (mean ± SE) measured from -24 h before UM to + 24 h after manipulation (n=9) a) Boxplot of proportions of PGFM before and after manipulation (UM4 and UM7 combined) On each box, the dot is the median, the base of the box the 25th and the top of the box the 75th percentiles, the whiskers extend to the most extreme data points not considered outliers, and any outliers are plotted individually.b) Comparison of PGFM concentration in UM4 and UM7 before and after UM. Five heifers (in red) were subjected to a UM at Day+4 of oestrous cycle (UM4), while remaining four heifers (in blue) were subjected to a UM at Day +7 of oestrous cycle (UM7)...... 127 Figure 4.4 a) P4 (ng/ml) measured from Day 3 to 20 of oestrous cycle (n=11). Six heifers (in red) were subjected to EB at Day+4 of oestrous cycle (B4), while remaining five heifers (in blue) were subjected to EB at Day 7 of oestrous cycle (B7). b) Observed and fitted values for the model xi

described in section 4.1.2.6 showing differences in P4 between treatments (B4 and B7). P4 for each treatment is plotted as a function of DOE (Day of oestrus). Each treatment is represented by a different colour. Observed values are plotted as dots. Fitted values are plotted as solid lines and represent the effect of DOE on P4 for each treatment. The shaded areas represented the 95 % point wise confidence intervals...... 130 Figure 4.5 PGFM (pg/ml) measured from -24 h before EB to + 84 h after EB (n=11) a) Boxplot of proportions of PGFM before and after biopsy (B4 and B7 combined). On each box, the dot is the median, the edges of the box are the 25th and 75th percentiles, the whiskers extend to the most extreme data points not considered outliers, and outliers are plotted individually.b) Comparison of PGFM concentration in B4 and B7 before and after EB. Six heifers (in red) were subjected to EB at Day+4 of oestrous cycle (B4), while remaining five heifers (in blue) were subjected to a UM at Day +7 of oestrous cycle (B7)...... 131 Figure C1 Representative calibration curve for Progesterone assessment by LCMS……….165 Figure C2 Instrument stability across the 3QC concentrations all less than 15%. QC is assess after a set of every 20 samples (so batch of 100 there are 6 points) b) at the retention time stability ….166 Figure D1 Parallelism of serial plasma dilutions (% binding) of PGFM against standard curve…168

xii

List of Tables

Table 1.1 Characteristics of studies excluded for systematic review ...... 36 Table 1.2 Pregnancy rates after endometrial sampling in cattle...... 40 Table 2.1 A subjective scoring system to assess ease of traversing through the cervical canal into the uterine horn...... 52 Table 2.2 Qualitative characteristics of endometrial sampling methods, device characteristics and samples recovered in heifers...... 56 Table 2.3 RNA yield and quality of endometrial tissue recovered by Storz® device from post- mortem uteri...... 57 Table 3.1 Mean (±SD) scores given by pathologists in biopsied (BH) and non-biopsied horns (NBH) in B4-T7 and B7-O-T7 and control. Mean comparisons are given between rows in BH and NBH ...... 71 Table 3.2 Morphological changes described by pathologist in B7-O-T7 and B4-T7 heifers ...... 75 Table 3.3 Sequences of primers for bovine cytokines and GAPDH in quantitative real-time polymerase chain reaction (qRT-PCR)...... 89 Table 3.4 Customised RT2 bovine inflammatory array (Qiagen Cat#: CAPB-13580,) ...... 90 Table 3.5 Individual mean score given by two observers for assessment of accumulation of uterine fluid in heifers biopsied at Day 4 (Group 1, B4 n=5). Heifers were scanned prior to biopsy the day of Cue-Mate removal (Day -7) and then the day of biopsy (Day+4, before the procedure). After biopsy, the uterus was scanned at Day +7. Score system used from 0-4: 0 Uterine fluid absent, 1: 1- 2 mm uterine fluid; 2: 2-3 mm uterine fluid; 3: 3-4 mm uterine fluid; 4: > 5 mm uterine fluid. (-) Negative for scores 0 and 1; (+) Positive for scores ≥2 ...... 108 Table 3.6 Individual mean score given by two observers for assessment of accumulation of uterine fluid in heifers biopsied at Day 7 (Group 2, 2xB7 n=8) in both horns. Heifers were scanned prior to biopsy the day of Cue-Mate removal (Day -7) and then the day of biopsy (Day+7, before the procedure). After biopsy, the uterus was scanned at Day +9, Day +13, Day +16. Score system used from 0-4: 0 Uterine fluid absent, 1: 1-2 mm uterine fluid; 2: 2-3 mm uterine fluid; 3: 3-4 mm uterine fluid; 4: > 5 mm uterine fluid. (-) Negative for scores 0 and 1; (+) Positive for scores≥ 2 ...... 109 Table 3.7:Individual mean score given by two observers for assessment of accumulation of uterine fluid in non-biopsied heifers (Control n=3).Heifers were scanned prior to biopsy the day of Cue- Mate removal (Day -7) and then the uterus was scanned at Day +7, Day +9, Day +13, Day +16. Score system used from 0-4: 0 Uterine fluid absent, 1: 1-2 mm uterine fluid; 2: 2-3 mm uterine

xiii

fluid; 3: 3-4 mm uterine fluid; 4: > 5 mm uterine fluid. (-) Negative for scores 0 and 1; (+) Positive for scores ≥2 ...... 110 Table 3.8 Score mean (±SD) assessment of accumulation of uterine fluid in heifers biopsied at Day 4 (Group 1 B4, n=5); at Day 7 in both horns (Group 2 2XB7, n=8); and non-biopsied heifers (Control, n=3) given by two observers (OBS1, OBS2). (B4). Heifers were scanned prior to biopsy the day of Cue-Mate removal (Day -7) and then the day of biopsy (Day+4, before the procedure). After biopsy, the uterus was scanned at Day +7. 2XB7 heifers were scanned prior to biopsy the day of Cue-Mate removal (Day -7) and then the day of biopsy (Day+7, before the procedure). After biopsy, the uterus was scanned at Day +9, Day +13, Day +16. Control heifers were scanned prior to biopsy the day of Cue-Mate removal (Day -7) and then the uterus was scanned at Day +7, Day +9, Day +13, Day +16. Score system used from 0-4: 0 Uterine fluid absent, 1: 1-2 mm uterine fluid; 2: 2-3 mm uterine fluid; 3: 3-4 mm uterine fluid; 4: > 5 mm uterine fluid. P-value <0.05 indicate significant difference in scores before and after biopsy (Group 1: D+4 vs. D+7; Group 2: D7 vs. D9)...... 111 Table 4.1 Parameter estimates for the GAMMs specified in Equation (1)...... 125 Table 4.2 Parameter estimates for the GAMMs specified in Equation 2...... 129

xiv

List of Abbreviations used in the thesis

AA arachidonic acid AI artificial insemination AT annealing temperature B-mode brightness mode bp base pair BW bodyweight CB cytobrush CH corpus haemorrhagicum CL corpus luteum cm centimetre COX 1-2 cyclooxygenase 1-2 Cq number of amplification cycles CV coefficient of variation d day DF dominant follicle DNA desoxyribonucleic acid DNase desoxyribonuclease EB endometrial biopsy et al. et alii ERA endometrial receptivity arrays E2 oestradiol ESR1 oestrogen receptor FGF fibroblast growth factor Fig. figure FSH follicle stimulating hormone g gravitational acceleration GE glandular epithelial (um) GnRH gonadotropin-releasing hormone h hour IVP In vitro produced i.e. Id est IFNτ interferon-tau IL interleukin IM. intramuscular kg kilogram LPS lipopolysaccharides LE luminal epithelium LH luteinizing hormone mg milligram MRP maternal recognition of pregnancy MHz megahertz min minute ml milliliter mm millimeter n number ng nanogram

NSAID non-steroidal anti-inflammatory drugs NO nitric oxide OXT1 oxytocin receptor P4 progesterone PAGs pregnancy-associated glycoproteins PCR polymerase chain reaction pg picogram PGR progesterone receptor PGFM 15-keto-13,14-dihydro-PGF2α PGF2α prostaglandin F2alpha QC quality control r correlation coefficient r2 coefficient of determination RNA ribonucleic acid RNase ribonuclease RIF recurrent implantation failure RT-PCR reverse transcription polymerase chain reaction SD standard deviation SEM standard error of the mean TLR toll-like receptor TNF-α tumor necrosis factor alpha UL uterine lavage VEGF vascular endothelial growth factor WOI window of implantation ZP zona pellucida ΔCq cycle number °C degree Celsius % percent μg microgram

xvi

Introduction

In cattle, early embryonic losses are recognised as one of the main causes of reproductive failure with significant economic impact in the industry. Early embryo loss has been defined as the loss of the embryo between fertilization and day 24 (Committee on Bovine Reproductive Nomenclature, 1972). The rate of early embryonic losses range between 20-45% (Humblot, 2001) where almost 70-80% of these losses occur between 8-16 days after fertilisation (Dunne et al., 2000, Diskin et al., 2011). External factors or maternal conditions that affect embryo development during the first two to three weeks of pregnancy might induce early embryonic mortality. External environmental conditions such heat stress affect embryo survival (Sartori et al., 2002, Hansen, 2009, Diskin et al., 2016).The maternal conditions that induce early embryonic mortality includes: a) genetic mutations (Diskin et al., 2016), b) reduced oocyte competence and changes in follicular dynamics (Lonergan et al., 2016); c) inadequate circulating levels of P4/E2 before and after insemination (Wiltbank et al., 2016) d) high levels of milk production and nutritional imbalances such negative energy balance during early postpartum (Lucy, 2001, Sartori et al., 2002, Wiltbank et al., 2016) e) infectious diseases (Hansen, 2011) and f) inadequate uterine environment and endometrial receptivity (Berg et al., 2010; Spencer, 2013).

Endometrial receptivity refers to the ability of the maternal endometrium to accept a normal embryo for attachment. This is regulated by secretion of ovarian steroidal hormones, oestrogens and progesterone. After the embryo enters the uterus on days 4 to 6 after fertilisation (Betteridge and Fléchon, 1988) there begins a coordinated and synchronised molecular cross-talk between the conceptus and the receptive endometrium. This molecular interaction is necessary to prepare the endometrium to support embryo development, elongation, maternal recognition of pregnancy and establish a successful attachment and ongoing pregnancy (Bazer et al., 2012).

In human reproduction, endometrial receptivity arrays (ERA) have been developed to assess the endometrial transcriptome during a limited period (window of implantation, WOI) when the uterus is receptive to implantation of the blastocyst (Miravet-Valenciano et al., 2015). These arrays have been used for clinical application in women with recurrent implantation failure (RIF) (Mahajan, 2015) where the endometrial expression or suppression of certain genes during WOI predicts the likelihood of pregnancy. However, the employment of uterine receptivity transcriptome arrays in cattle is still a challenge. Differences in studies between breeds, age, physiological uterine conditions and methods of uterine sampling, have limited the application of ERA in bovine

xvii

Introduction reproduction because so far there is no agreement on the identification of essential endometrial markers. ERA test could be implemented in the cattle industry in assisted reproductive technologies such as artificial insemination (AI) and embryo transfer (ET). The identification of molecular markers suggestive of adequate endometrial receptivity before AI or ET would be valuable for improving the selection of animals with higher chances of pregnancy success, decreasing pregnancy losses associated with inadequate uterine environment.

Uterine health is currently assessed by the traditional non-invasive methods of transrectal palpation, ultrasound and the score of vaginal mucus (Sheldon et al., 2006). However, these methods are a subjective assessment of endometrial receptivity. Hence, the use of more objective assessment requiring invasive methods for sampling the endometrium is needed. These sampling methods include uterine lavage (UL), cytobrush (CB) or endometrial biopsy (EB). These procedures employ a transvaginal device guided through the cervix under transrectal manipulation into the uterine body or horns to collect the sample required (cells, fluid or tissue).

An adequate model to predict endometrial receptivity would require 1) animals whose uterine environment has not been modified by breeding, gestation, calving or affected by production or metabolic diseases, 2) a method that provides consistently adequate material to perform transcriptome analysis 3) a procedure that ensures that the uterine environment and the subsequent reproductive performance is not affected after sampling and finally, 4) a time point during the oestrous cycle that allows sample collection and analysis before the day of breeding or ET. Assessing the uterus at Day 7, around the time of embryo arrival, would be indicative of the uterine environment to which the embryo is first exposed during development.

In this regard, different studies have been performed using cycling heifers to identify candidate markers for uterine receptivity based on pregnancy outcome (Geary et al., 2016). Differences in endometrial transcriptome profile between pregnant and non-pregnant heifers have been identified by assessing endometrial samples collected using CB from the previous cycle (Salilew-Wondim et al., 2010) or in the same cycle at 6 d after AI by EB from the horn contralateral to the corpus luteum (CL) (Pugliesi et al., 2014, Binelli et al., 2015). However, these studies failed to evaluate the effect that the mechanical injury induced by sampling has on the uterine environment and the effect of sampling the uterus during dioestrus on CL lifespan. EB is an invasive method of sample collection that causes a mechanical injury with disruption of the epithelial barrier and uterine environment (Gnainsky et al., 2010, Snider et al., 2011). After mechanical injury or trauma, the immune system starts wound healing, to replace damaged tissue (Salamonsen, 2003). In women, it has been demonstrated that EB performed one cycle before ET, induces an inflammatory reaction with the

xviii

Introduction subsequent influx of inflammatory cells and production of cytokines and growth factors that actually improves the subsequent pregnancy rates (Barash et al., 2003, Almog et al., 2010, Gnainsky et al., 2010, Granot et al., 2012).

Preliminary studies in cattle (Zaayer and Van der Horst, 1986) and mares (Baker et al., 1981) reported that endometrial biopsies during metoestrus or dioestrus affect the lifespan of the CL (CL) decreasing progesterone concentration, shortening the oestrus length and affecting the endometrial environment. Causing endometrial injuries when the CL is fully matured, as in dioestrus, may trigger an endogenous release of PGF2α resulting in lysis of the CL (Snider et al., 2011).

Therefore, this thesis aimed to establish the effects subsequent to endometrial biopsy on uterine health and CL lifespan. To achieve the overall objective, the specific sub-objectives of this research thesis are:

1. Compare the feasibility of performing endometrial sampling procedures in heifers under field conditions and assess whether the sample recovered provides sufficient material to perform further transcriptome assessment. 2. Assess the uterine morphological changes at Day 7 of oestrus, after performing an endometrial biopsy at Day 7 post-oestrus of the preceding cycle or at Day 4 of the ongoing cycle. 3. Evaluate the uterine inflammatory response induced by mechanical injury from an endometrial biopsy. 4. Assess the effect of endometrial biopsy on CL lifespan.

xix

CHAPTER 1. REVIEW OF LITERATURE

Chapter 1 Review of literature

Introduction

The purpose of this chapter is to summarise the theory behind endometrial sampling procedures in cattle, focusing on endometrial biopsy and the subsequent effects on uterine health. Initially, the morphological description of the uterus and the physiology of oestrous cycle and embryo development conditions are described. Thereafter, the review is focused on the role that endometrial receptivity and progesterone secreted from CL have on embryo development and survival. Considering that inadequate endometrial receptivity is one of the causes of early embryo mortality, the methods of endometrial assessment are described so as to highlight the potential of endometrial biopsy to predict endometrial receptivity.

1.1 The bovine uterus

In mammals the female reproductive tract consists of: a pair of ovaries, oviducts, a uterus, vagina, vestibule and external genitalia. The uterus in cattle is an organ consisting of the cervix, the uterine body and two uterine horns, bound together by the intercornual ligament. The cervix is a thick- walled, smooth muscle sphincter that remains tightly closed except during oestrus or parturition. The uterine body is 2-4 cm long and extends from the internal cervical os to the bifurcation of the uterine horns. The uterine horns are 20 to 45 cm long and are the site for embryo implantation and foetal development (McEntee, 1990).The uterus provides an adequate environment for conceptus survival and development during pregnancy, and at the end of gestation the delivery of the foetus (Spencer et al., 2012).

The uterine wall is composed of three layers (1) the mucosa or endometrium (2) the muscularis or myometrium and (3) the serosa or perimetrium (Eurell and Frappier, 2013).The endometrium contains a pseudoestratified columnar luminal epithelium (LE) and single or branched secretory uterine glands. The submucosa layer or lamina propria consist of loose connective tissue, blood vessels and uterine glands that open into the lumen to secrete or transport substances required for embryo development and successful attachment (Mullen et al., 2012). The endometrium contains intercaruncular (glandular) and caruncular (non-glandular) areas. During gestation, the endometrial caruncles fuse with chorionic cotyledons to form the placentomes, which are the maternal site for nutrients and exchange of gases between the maternal and foetal circulations (Samuelson, 2007, Bacha and Bacha, 2011). Beneath the endometrium are two layers of smooth muscle or myometrium that control uterine motility and a serous membrane or perimetrium (Samuelson, 2007).

Chapter 1 Review of literature

1.2 Bovine oestrous cycle

The oestrous cycle is a sequence of events that occurs between two successive periods of sexual receptivity (heat or oestrous). Bovine females are seasonally polyoestrous with a 21±3 days oestrous cycle length (Savio et al., 1990, Peter et al., 2009, Forde et al., 2011b). The oestrous cycle length is similar in B. indicus and B. taurus breeds (Bó et al., 2003) and generally is longer in adult cows than young heifers (23 vs. 20.8 days; respectively P< 0.05) (Sartori et al., 2004).The length of the cycle depends on the number of waves of follicular growth (2 to 4). Most commonly, heifers and cows have two to three waves of follicular growth during the oestrous cycle. In heifers or cows with three follicular waves the interoestrous interval is longer than females with two waves (Ginther et al., 1989, Sartori et al., 2004).The number of follicular waves varies within the same animal, between cycles affecting the interoestrous interval (Figueiredo et al., 1997).

The oestrous cycle consists of follicular and luteal phase. During the follicular phase (4 to 6 days duration), the growing dominant follicle (DF) secretes oestradiol (E2) whereas, in the luteal phase (14 to 18 days duration), the corpus luteum (CL) secretes progesterone (P4) (Forde et al., 2011b, Sartori and Barros, 2011). These changes in hormonal secretion modify the sexual behaviour and induce morphological changes in the uterus, hence dividing the oestrous cycle into four stages: oestrus, metoestrus, dioestrus and pro-oestrus. The follicular phase includes the pro-oestrus and oestrus stages, whereas the luteal phase includes metoestrus and dioestrus stages) (Forde et al., 2011b). During oestrus or heat (Day 0), elevated concentrations of systemic E2 secreted by the DF, induce the behavioural changes associated with sexual receptivity, endometrial oedema and myometrial contractility. E2 stimulate the secretion of oestral mucus from LE and the proliferation, growth and branching of endometrial glands (Mullins and Saacke, 1989). This secretory activity is accompanied by a massive migration of white blood cells (especially neutrophils) to clear the uterus of dead sperm cells, bacteria and cellular debris deposited during breeding (Ohtani et al., 1993) (Figure 1.1 b). During metoestrus phase (Day 1 to 4), the CL is formed secreting the P4 necessary to establish and maintain the pregnancy and the recruitment of first follicular wave starts (Eurell and Frappier, 2013) (Figure 1.1c). In metoestrus, the gradual increase in P4 stimulates endometrial glandular proliferation which is accompanied with significant branching and coiling. The synthesised secretions are accumulated in the glandular epithelial (GE) to be transported into the lumen during the maximum secretory phase in dioestrus (Wang et al., 2007).

Chapter 1 Review of literature

Figure 1.1 Uterine and ovarian changes during oestrous cycle in the cow. Adapted from (Eurell and Frappier, 2013).P4: Progesterone; E2: Oestrogens.

Chapter 1 Review of literature

The onset of the luteal phase or dioestrus (Day 5 to 16-17) is marked by the presence of a fully mature and functional CL that secretes P4. Peripheral plasma P4 concentration above 1 ng/ml is suggestive of a fully functional CL and the cow is classified as cyclic. Peripheral levels below 1 ng/ml indicate a non-functional or a regressing CL suggestive of either anoestrous or degenerated CL after luteolysis (Chebel et al., 2006). Peripheral P4 concentrations start to increase approximately by Day 4 post-oestrus, reaching the maximal concentration by Day 8–10 (Robinson et al., 2008). The function and development of the CL is partly influenced by blood flow which is in turn mediated by the presence of growth factors (vascular endothelial growth factor, VEGF and fibroblast growth factor, FGF) since they both promote vascular growth and angiogenesis creating a network of capillaries in the luteal cells (Yamashita et al., 2008). Acosta et al. (2003) in cows and Ginther et al. (2007) in mares had found a gradual increase in the CL blood flow, CL volume and systemic P4 concentration during the first week after ovulation which was associated with the development of the CL and its potential to produce and secrete P4.

At endometrial level, luteal P4 stimulates endometrial glands in LE and GE to secrete into the lumen a complex mixture of nutrients and signalling molecules, called histotroph or uterine milk (Brooks et al., 2014). The histotroph is composed of a mixture of proteins, enzymes, ions, growth factors, cytokines, lymphokines and hormones that provide the nutrients that are essential for the embryo’s development and growth within the uterine lumen during early pregnancy (Figure 1.1 d) (Igwebuike, 2006).

During early and mid-luteal phase, luteal P4 act via progesterone receptor (PGR) to suppress the expression of oxytocin receptors (OXT1) and E2 receptors (ESR1) in the endometrium. P4 stimulates the accumulation of phospholipids in luminal (LE) and glandular epithelium (GE) (Arosh et al., 2002). In late dioestrus, if pregnancy does not occur endometrial pulses of prostaglandin F2- alpha (PGF2α) are released inducing structural and functional regression of both the CL (luteolysis) and endometrial glands, decreasing both the systemic levels of P4 and the glandular secretory activity, respectively (Ohtani et al., 1993, Shaham-Albalancy et al., 1997). During luteolysis, P4 induces down-regulation of the receptor in LE and GE, allowing the expression of ESR1 and OXT1 in the uterus. The accumulated membrane phospholipids in LE and GE, are converted into arachidonic acid (AA) by the activation of phospholipase A2 which in turn is converted into PGF2α by the enzymes cyclooxygenase (COX-1 and COX-2) (Okuda et al., 2002, Arosh et al., 2002).

The luteal phase is terminated by the secretion of pulses of PGF2α from the endometrium. These pulses are short in duration (1 to 5 h) occurring nearly every 12 h and lasting for 2-3 days (Arosh et al., 2002, Mann and Lamming, 2006). Since PGF2α has a very short half-life due to its metabolism

Chapter 1 Review of literature in lungs, where almost 65% is metabolised in one pass (Thatcher et al., 1986), the main indicator of PGF2α release into circulation is the assessment of the main metabolite, 13,14-dyhydro-15-keto- PGF2α (PGFM) (Ginther et al., 2010) which is more stable than PGF2α. Thus, a peripheral concentration of PGFM is a reliable indicator of uterine production of PGF2α. Once is secreted, uterine PGF2α is transported to the ipsilateral ovary that contains the CL and secreted into the uterine vein, which joins the ovarian vein to form the utero-ovarian vein. The utero-ovarian vein lies in close proximity to the ovarian artery. Through a vascular counter current exchange mechanism, PGF2α is transported into the ovarian artery (Lee et al., 2010, Forde et al., 2011b). Within the CL, PGF2α suppress the expression of angiogenic factors such as vascular epithelial growth factor (VEGF) and increase the expression of vasoactive factors such as endothelin-1 (EDN1), angiotensin II (ANG II) and luteal PGF2α (Miyamoto et al., 2005), that produce a vasoconstriction decreasing luteal blood flow, and results in initiation of luteolysis (Miyamoto et al., 2009, Berisha et al., 2010, Scully et al., 2015).

The pro-oestrus (Day 17-21) starts with the regression of the CL and the subsequent decline of P4 concentration that allows the development of the ovulatory follicle for the subsequent oestrous cycle and results in a return to normal cyclicity (Ginther et al., 2009). In pro-oestrus, the circulating E2 stimulates a proliferative response at the endometrium, increasing both branching and size of the uterine glands and the proliferation of ciliated epithelial cells and glandular epithelium (Figure 1.1 a) (Ohtani et al., 1993, Wang et al., 2007).

In conclusion, the concentration and regulation of ovarian and uterine hormones, changes depending on the phase of oestrous cycle via hypothalamic-pituitary-ovarian (HPO) neuroendocrine signalling. The oestrous cycle is regulated by both the CL function and the P4 secretion and E2 from the DF that together regulate the uterine environment. At the morphological level, the cyclic changes in ovarian hormonal secretion modify the epithelia height and uterine gland features during the oestrous cycle. Oestradiol stimulates the proliferation, growth and branching of the glands whilst P4 stimulate the secretory activity of endometrial glands.

Chapter 1 Review of literature

1.2.1 Role of progesterone on embryo development and survival

After the embryo enters the uterus, 4 to 6 days after fertilisation (Betteridge and Fléchon, 1988) a coordinated dialogue between the embryo and the endometrium starts. The endometrial-conceptus interaction is necessary to prepare the endometrium to support embryo development, elongation, maternal recognition of pregnancy and establish a successful attachment and ongoing pregnancy (Bazer et al., 2012).

Circulating P4 concentrations represent a balance between the production of P4, primarily by the corpus luteum (CL), and the metabolism by the liver (Wiltbank et al., 2014). There is no direct effect of P4 on embryo development (Clemente et al., 2009). P4 acts indirectly on conceptus development through changes in the endometrial transcriptome, which are necessary for establishment of endometrial receptivity (Forde et al., 2009, Lonergan et al., 2016). At the uterine level, the post-ovulatory rise in P4 creates the uterine microenvironment to support early embryonic development through activation of P4 receptors (PR).These receptors are expressed in LE and GE during early luteal phase (Spencer et al., 2004b).Increasing concentrations of P4 after ovulation enhance conceptus elongation (Beltman et al., 2009a, Forde et al., 2011c), while lower P4 concentrations in metoestrus retarded embryonic development in cattle (Larson et al., 1997, Mann and Lamming, 2001, Stronge et al., 2005).

Larson et al. (1997) showed that in cows that became pregnant, the onset of luteal phase after breeding occurred earlier than non-pregnant cows (3.27 vs. 4 days, respectively p<0.05) suggesting that as little as one day delay in the post-ovulatory rise of P4 has adverse consequences for embryo development and survival. Using logistic regression analysis from post-insemination P4 milk concentration and embryo survival in 871 inseminations in dairy cows, Stronge et al. (2005) found that low P4 levels during Days 5-7 (after insemination) was associated with low fertility. This linear relationship between milk P4 and embryo survival was also reported by McNeill et al. (2006) who found a positive relationship between P4 concentrations in milk on Days 4 to 6 after AI on embryo survival rate indicating that sub-optimal levels of P4 during metoestrus and early dioestrus impaired embryo development. Similarly, Mann and Lamming (2001) found that inseminated cows with confirmed embryos at Day 16 post –AI, showed earlier increased in P4 concentrations (>1ng/ml) than inseminated, non-pregnant cows (Day 4.9 ±0.2 vs 6.2 ±0.4 , respectively) (P<0.001).

Chapter 1 Review of literature

The effects of low concentrations of circulating P4 after ovulation have been associated negatively with conceptus growth, elongation and embryo survival (Beltman et al., 2009b, Lonergan et al., 2016). After hatching from the zona pellucida (ZP), the bovine conceptus (embryo and associated extraembryonic membranes) undergoes a phase of trophectoderm expansion and development, known as elongation. Elongation is a uterine regulated process which is mediated by P4 secreted from the CL (Clemente et al., 2009). In the bovine, the elongation process starts at Day 13 and lasts until Day 16. During this stage, the conceptus elongates from a spherical shape to a tubular and filamentous form, increasing almost 300-fold in size from 2 mm at Day 13 to 60mm at Day 16, and finally reaching 200 mm size by Day 19, occupying the entire length of the uterine horn ipsilateral to the CL (Spencer, 2013) (Figure 1.2).

The elongation is regulated by the histotroph secretion from GE and LE. In vitro produced (IVP) embryos are able to successfully develop until the blastocyst stage (Day 7), but elongation does not occur in vitro (Igwebuike, 2006, Clemente et al., 2009). The importance of the uterine glands contribution to the histotroph composition and conceptus elongation has been assessed using in vivo sheep uterine glandular knock-out (UGKO) models (Gray et al., 2001). Ewe lambs were treated with P4 implants from birth to 8 weeks. The growth and development of the endometrial glands was inhibited. After natural breeding, adult UGKO ewes were unable to maintain a pregnancy due to the absence of endometrial glands and their secretions which resulted in an inadequate conceptus elongation and hence the failure of pregnancy.

Figure 1.2 Early embryonic growth and conceptus elongation (Adapted from Spencer (2013). The period of conceptus elongation in bovine embryos is highlighted.

Chapter 1 Review of literature

Elongation of the bovine conceptus is a prerequisite for maternal recognition of pregnancy (MRP) and attachment (Bazer et al., 2008). The maternal recognition of pregnancy (MRP) is defined as the physiological process whereby the conceptus signals its presence to the maternal system and prolongs the lifespan of the CL (Spencer et al., 2004a). The trophoblast of the bovine conceptus secretes interferon-tau (IFNτ) between Day 10 and 25 of gestation and is the signal for MRP in ruminants (Bazer et al., 1991). IFNτ is a type-I IFN family member that is expressed in mononuclear trophectoderm cells of the ruminant conceptus and acts locally on LE and GE and stroma to prevent PGF2α synthesis and the cascade of the luteolytic mechanism (Roberts et al., 2008, Bazer et al., 2008).

Studies on exogenous P4 supplementation during the first-week post-insemination have been proposed as an approach to increase embryo growth and survival. Beltman et al. (2009a) found that supplementing beef heifers with exogenous P4 during metoestrus (P <0.04) or early dioestrus (P< 0.06) had a positive relationship with embryo survival rates. Similar results were described by Wallace et al. (2011) who found significantly higher pregnancy rates in P4-supplemented beef heifers treated with human chorionic gonadotropin (hCG 1000 IU) compared to non-treated animals (66.3 vs. 55.5%, respectively, P< 0.001). In the experiment of Beltman et al. (2009a) the embryo length assessed at Day 25 was similar in treated and non-treated heifers (29.5 vs. 30.2 cm, respectively). However, when the length was assessed earlier, shortly after MRP (Day 16), the embryos recovered from P4-treated cows were fourfold longer (P<0.01) than embryos from non- treated cows. In addition, the uterine fluid recovered from P4-treated cows, showed a six-fold higher concentration of IFNτ than the uterine fluid from non-treated cows (Wallace et al., 2011).

In conclusion, the effect of circulating concentrations of P4 results in changes in histotroph secretion and thus endometrial environment that affect conceptus elongation during the second week after fertilisation. The positive effect on pregnancy rates after exogenous P4 supplementation has been attributed to the early rise in plasma P4 after hormone administration (within 24 h after treatment), the formation of an accessory CL and the higher conceptus secretion of IFNτ in supplemented animals. Thus, larger embryos would secrete more IFNτ and would be more effective in suppressing the luteolytic cascade, crucial for pregnancy success (O’Hara et al., 2014).

Chapter 1 Review of literature

1.2.2 Luteolysis

In non-pregnant ruminants, luteolysis occurs during the late luteal phase (Days 16 to 18). The systemic level of P4 decreases, allowing the development of the ovulatory follicle from the subsequent oestrous cycle and the normal ciclicity returns. During luteolysis, the endometrial cells synthesize pulses of PGF2α that induces structural and functional regression of the CL decreasing the P4 levels, (Ginther et al., 2009).

During dioestrus, luteal P4 stimulates the accumulation of phospholipids in luminal (LE) and glandular epithelium (GE). If pregnancy does not occur, E2 secreted from the dominant follicle induces the activation of phospholipase A2 (PLA). PLA changes the membrane phospholipids into arachidonic acid, which in turn is converted into Prostaglandin H2 (PGH2) by the enzymes cyclooxygenase (COX-1 and COX-2). Finally, the enzyme PGF synthase (PGFS) catalyse the conversion of PGH2 to PGF2α (Okuda et al., 2002, Arosh et al., 2002).

Luteolysis is a uterine dependent process that responds to the effects of P4, E2 and oxytocin (OXT) acting through their endometrial receptors. During early and mid-luteal phase, oxytocin receptor (OXT1) and estrogen receptor (ESR1) are suppressed by the presence of luteal P4. If pregnancy does not occur, progesterone receptors (PGR) are down-regulated in the uterus and ESR1 and OXT1 are expressed in the uterus. This change in endometrial receptors is mediated by the increase of E2 secreted from the dominant follicle and the release of hypophyseal and luteal OXT, generating a pulsatile release of luteolytic PGF2α (Robinson et al., 2001, Mann and Lamming, 2006).

These PGF2α pulses are short in duration (1 to 5 h) occurring nearly every 12 h and lasting for 2-3 days (Arosh et al., 2002, Mann and Lamming, 2006). Since PGF2 has a very short half-life due to its metabolism in lungs and endometrium (Thatcher et al., 1986), the main indicator of PGF2 release into circulation is the assessment of their main metabolite, 13,14-dyhydro-15-keto-PGF2α (PGFM) (Ginther et al., 2010). Once is secreted, uterine PGF2α is transported to the ipsilateral ovary that contains the CL. PGF2α is secreted into the uterine vein, which join the ovarian vein to form the utero-ovarian vein. The utero-ovarian vein lies in close proximity with the ovarian artery. Through vascular counter current exchange mechanism, PGF2α is transported into the ovarian artery (Lee et al., 2010, Forde et al., 2011b). Within the CL, PGF2 suppress the expression of angiogenic factors such as vascular epithelial growth factor (VEGF) and increase the expression of vasoactive factors such as endothelin-1 (EDN1), angiotenisin II (Ang II) and luteal PGF2

Chapter 1 Review of literature

(Miyamoto et al., 2005), that produces a vasoconstriction thus results in a decrease in blood flow and initiation of luteolysis (Miyamoto et al., 2009, Berisha et al., 2010).

The effects of PGF2α are also exerted at endometrial glandular level. In case of luteolysis, the endometrial glands regress along with the CL, involute gradually and decrease the secretory activity, starting again the proliferative glandular stage (Ohtani et al., 1993, Shaham-Albalancy et al., 1997).

1.3 Embryo mortality and endometrial receptivity

Embryonic losses which can occur throughout pregnancy are recognised as the major cause of reproductive failure in the cattle industry. The decline in fertility and the subsequent reduction of calving rates have a negative impact on the reproductive performance affecting production and profitability (Geary, 2005). The fertility rates in dairy and beef cattle are similar and range between 85-90% (Wiltbank et al., 2011). However, the calving rates in high-producing dairy cows are lower (35-40%), compared to the full-term pregnancy reported in dairy heifers and beef cattle (55-60%) (Diskin and Sreenan, 1980, Santos et al., 2004, Lonergan et al., 2016). The Committee of Bovine Reproductive Nomenclature in 1972 defined two types of embryo mortality (EM): early embryo mortality (EEM) known as the losses that occur between fertilisation and day 24 of gestation, and late embryo mortality (LEM) as losses that occur beyond day 25 to day 45 of gestation. Any subsequent loss of pregnancy beyond day 45 is classified as foetal loss. Since then, that terminology has been used worldwide to define embryo losses. Recently, some authors (Walsh et al., 2011, Wiltbank et al., 2016) have proposed some modifications to define embryo mortality based on the crucial periods that the embryo faces within the reproductive tract to establish a successful full-term pregnancy. Thus, Walsh et al. (2011) suggested including an initial stage termed very early embryo mortality to define the losses that occur after breeding and within the first week of gestation. This stage comprises the oviductal transport, genome activation, and early blastocyst development. In addition, Wiltbank et al. (2016) suggested extending the concept of an early embryonic period from day 8 to day 27 to be able to diagnose pregnancy by either ultrasound or pregnancy-associated glycoproteins (PAGs) and hence determine whether the early loss occurred during this period. Finally, they suggested a third period of embryo mortality corresponds to the second month of pregnancy -days 28 to 60- where critical events include the attachment of maternal and allantoic membranes and placentome development occur. This scheme better aligns loss with major physiological events and hence suggests the cause of the loss.

Chapter 1 Review of literature

Between 70-80% of the total embryo loss occurs during the second week of pregnancy between 8- 16 days after breeding, before MRP (Dunne et al., 2000, Diskin et al., 2011). Alterations of the uterine environment at the time of embryo arrival may reduce the endometrial receptivity, hence the quality of the molecular cross-talk dialogue, thus affecting embryo development (Bazer et al., 2012). The uterus is sterile or clear of pathogens (Sheldon et al., 2008). Anatomically, the cervix, vestibule, vagina and vulva act as a physical barrier to protect the uterus against ascending bacteria, isolating the uterus from the external environment (Sheldon and Dobson, 2004). However, after mating or parturition, the presence of pathogens increase and those animals unable to eliminate bacteria may develop clinical or subclinical endometritis (Herath et al., 2006).

In cattle, the most common condition that affects endometrial receptivity is endometritis (clinical or subclinical). Clinical endometritis is defined as an inflammation of the endometrium with vulvar discharge detected after 21 days postpartum, without systemic clinical signs. Subclinical endometritis is defined as an inflammation of the endometrium with no vulvar discharge (Sheldon et al., 2006).

Endometrial receptivity is compromised by the presence of pathogenic bacteria and bacterial products such as gram-negative bacterial endotoxin (lipopolysaccharide, LPS) or pro-inflammatory mediators (PGF2α or Nitric Oxide, NO), compromising early embryo development (Soto et al., 2003, Schwarz et al., 2008, Gilbert, 2011). After contact with endotoxin LPS, endometrial immune receptors (toll-like receptors, TLR), recognize the pathogen and trigger the expression of pro- inflammatory cytokines (TNFα, IL1β, IL6) and chemokines (IL8) which in turn attract and activate neutrophils and macrophages to the site of inflammation (Ahmadi et al., 2005, Galvao et al., 2011, Gilbert, 2011). Hill and Gilbert (2008) demonstrated that embryos exposed to pro-inflammatory cytokines in the culture media had a reduced number of trophectoderm cells thereby impairing their ability to produce IFNτ required for successful MRP.

In addition, the presence of NO in the culture medium although not affecting blastocyst development, increases the apoptosis incidence in trophectoderm cells, hence limiting the embryo secretion activity of IFNτ (Schwarz et al., 2008). Therefore, in cows with either subclinical or clinical endometritis during the first two weeks of pregnancy, embryo attachment is compromised. This is because endometrial receptivity is disturbed and the quantity of IFNτ secretion by the embryo is impaired, affecting the maternal recognition of pregnancy.

Therefore, uterine pathologies such as puerperal metritis, clinical endometritis and subclinical endometritis induce endometrial damage that compromises embryo attachment success, hence reducing the probability of successful conception, increasing calving to conception interval and 12

Chapter 1 Review of literature number of cows culled due to conception failure (Sheldon et al., 2006, Lopez-Helguera et al., 2012).

As discussed above, initial studies of embryonic losses in cattle sought to find the correlation between attachment failure and uterine health caused by clinical or subclinical endometritis based on the percentage of PMN cells detected in cytology swabs (Gabler et al., 2010). However, more attention has been given to endometrial gene expression at the time of embryo arrival as a more accurate indicator of uterine status. Thus, the candidate genes approach seems to be more reliable for defining new quality criteria for an optimal endometrium.

A large number of studies in vitro have described the culture conditions and embryo requirements from fertilisation up to the blastocyst stage (Besenfelder et al., 2010, Block et al., 2011, Ghanem et al., 2011). Likewise, endometrial transcriptome studies (Bauersachs et al., 2005, Mitko et al., 2008, Forde et al., 2011a) identified genes differentially expressed during the oestrous cycle. These genes are especially related to cell motility, cytoskeleton, extracellular matrix (ECM), ECM remodelling, and cell growth are differentially expressed. Genes expressed before MRP are designed to contribute to conceptus elongation and prepare the endometrium for attachment.

Recently, transcriptome studies had provided a description of specific endometrial genes and pathways suggested for endometrial receptivity based on pregnancy outcome. The identification of these molecules could be used as a prognostic factor to screen and select recipients with endometrial characteristics that favour receptivity. Salilew-Wondim et al. (2010) and recently Binelli et al. (2015) described the endometrial gene expression around Day 7 post-oestrus in pregnant and non-pregnant cows. Although different methods for sample collection and platform analysis were used in both studies, they concluded that differences in gene expression exist in the number, types of genes and pathways expressed between pregnant cows with receptive endometrium and non-pregnant cows with non-receptive endometrium at Day 7 post-oestrus.

Binelli et al., (2015) found a higher number of down-regulated genes (180) than up-regulated genes (36) in the uterus of pregnant cows at Day 6 after mating. Based on the high number of down- regulated genes, he suggests that the embryo requires a “low” gene active uterine environment during the initial stage of development (Baumann et al., 2007). This theory of ‘quiet’ uterine environment seems to be possible considering the low growth in terms of cellular mass during the second week of embryo development which is occurring at this time, before the elongation starts (Spencer, 2013) (See Figure 1.2). To achieve this stage, the uterus reduces the concentration of nutrients in its environment, encouraging the embryo to use intrinsic resources (Leese, 2003). Similarly, Moran et al., (2017) found significantly different expression of 400 genes in the 13

Chapter 1 Review of literature transcriptome of endometrial biopsies collected on Day 7 of the oestrous cycle from cows of high and low fertility suggesting an association between gene expression in the uterine environment and genetic merit for fertility in dairy cows.

An opposite gene pattern was described by Salilew-Wondim et al. (2010) who found that almost 600 genes were up-regulated to some extent in the receptive endometrium (those resulting in successful pregnancy) and 500 genes were down-regulated in the non-receptive endometrium. Several molecules such as adhesion molecules, cytokines, growth factors, solute carriers and those related with cell motility, extracellular matrix (ECM) remodelling and cell growth contribute to a receptive uterus and prepare the uterus for attachment. These studies with very different outcomes suggest that the identification of a single marker or even a small group of markers indicative of endometrial receptivity is still a challenge in cattle.

Similarly, in women it has been demonstrated that the endometrial gene expression pattern at the time of embryo implantation is different in fertile women from women with unexplained infertility (Altmäe et al., 2010). Endometrial receptivity arrays (ERA) have been developed in women to predict the likelihood of pregnancy and are currently used for clinical application in women with recurrent implantation failure (RIF) especially in IVF treatments (Díaz-Gimeno et al., 2013). ERA’s consider the individual variation in cycle’s length that modifies the most receptive time for implantation, assessing the endometrial transcriptome when the uterus is most receptive to implantation during the window of implantation, (WOI) between 5 to 7 days after ovulation (Miravet-Valenciano et al., 2015). Endometrial tissue is collected before ET from the anterior wall of the uterine cavity using a Pipelle catheter and the gene expression is analysed in predictor software. After principal component analysis (PCA) is applied to data, samples with similar trends in their gene expression profiles tend to cluster close together in the plot showing whether the uterus is at the most receptive time for implantation (Altmäe et al., 2010, Díaz-Gimeno et al., 2013). Similar arrays are not yet available in cattle but the women clearly show their potential. The main limitation for implementation and development of these microarrays in cattle is the length of time that the embryo stays within the uterus before implantation (from Day 4-5 until Day 18) and the morphological and dynamic changes that occur in both the embryo and uterine environment during conceptus development.

Therefore, the assessment of uterine environment and early detection of uterine dysfunction or disease in individual cows, before breeding or ET, would significantly improve herd reproductive performance which is vital for achieving an optimal per capita return for breeding efforts.

Chapter 1 Review of literature

1.4 Uterine assessment for endometrial receptivity

The traditional examination of physiological and pathological conditions of the bovine uteri relies on morphological features detected by rectal palpation, evaluation of vaginal discharges using a vaginoscope or Metricheck®,and the visualization of reproductive organs by transrectal ultrasound (Berlanga et al., 2011). These quick, generally harmless and low-cost methods for on-farm diagnosis are common in clinical practice, although their accuracy in terms of prediction of successful attachment is subjective and limited (Lopez-Helguera et al., 2012, Madoz et al., 2014).

1.4.1 Non-invasive uterine assessment of endometrial receptivity

Rectal palpation relies on the ability of the examiner to manually compare the symmetry and tone of the uterine horns and detect uterine content within horns (Leutert et al., 2013). This method is the most time effective in terms of offering an immediate on-farm diagnosis, but its accuracy for diagnosis of uterine disorders is limited (Barlund et al., 2008).Vaginal examination is a rapid, common method used to diagnose vaginal discharges from the cervical os or the floor of the vagina. Discharges can be collected and visually examined either by a gloved hand, the Metricheck® device or vaginoscopy (McDougall et al., 2007, Leutert et al., 2012). Based on the discharge characteristics (colour), the discharge obtained is given a score from 0 to 3 based on the amount of pus that the discharge contains. Thus, zero corresponds to translucent mucus, and 3 corresponds to a purulent exudate that generally reflects a high number of pathogenic bacteria and is normally associated with uterine infections (Williams et al., 2005). This scoring system has been correlated with the presence of pathogenic bacteria and the prognostic value for pregnancy success (Sheldon et al., 2006). Postpartum cows with clear discharge (score 0) had higher non-return rate at 90 days (56%) than cows with abnormal vaginal secretion (score >1) (48% P< 0,05) (Lambertz et al., 2014). However, vaginal discharges are not always indicative of uterine infection because many cases of vaginitis or cervicitis can occur in parallel without uterine disorders and vice versa. Comparing the cytological results obtained by CB collection from the uterus and the cervix, Hartmann et al. (2016) found that 66% of cows diagnosed with cervicitis did not present endometritis. Deguillaume et al. (2012) found in post-partum cows, 32 % of cows presented both cervicitis and endometritis, 11% presented only cervicitis and 13% suffered only from endometritis. Both studies confirmed that cervicitis and endometritis are independent diseases, that affect the reproductive performance (Dubuc et al., 2010, Šavc et al., 2016) and suggest that the clinical evaluation of the reproductive tract should not be based only in the assessment of vaginal discharge as indicative of endometritis, but it should assess endometrium, cervix and vagina (Dubuc et al., 2010).

Chapter 1 Review of literature

Transrectal ultrasound has been used for more than 30 years in farm animals as a non-invasive tool for visual access to the reproductive tract to evaluate physiological and pathological conditions (Pierson and Ginther, 1988). However, uterine ultrasonography is not a reliable method for diagnosis of uterine pathologies by itself (Polat et al., 2015). Ultrasound provides information about uterine content, endometrial thickness and cervical size and recently, with the use of colour Doppler ultrasound (Polat et al., 2015) and contrast-enhanced analysis (Schmauder et al., 2008), it is possible to assess the increase in uterine blood flow, the tissue echotexture and persistence of oedema in cases of acute endometritis after injury (Debertolis et al., 2016).

Several studies have assessed the accumulation of uterine fluid by ultrasound during the oestrous cycle (Fissore et al., 1986, Kastelic et al., 1991) early pregnancy (Scully et al., 2015) and postpartum (Mateus et al., 2002, Kasimanickam et al., 2004, Barlund et al., 2008). During the oestrous cycle, there is limited uterine fluid in the luteal phase, but the fluid increases many-fold during the periovulatory period and when the cow enters oestrus. In non-bred heifers, Kastelic et al. (1991) scanned the uterus throughout the oestrous cycle and found that the uterine fluid content was about 0.2 ml from Days 10 to Days 16, whereas at Day 18 it increased to up to 3.5 ml. In postpartum cattle and mares that have recently been bred, the presence of fluid in the uterine lumen has been considered a positive sign for the diagnoses of subclinical endometritis (Troedsson, 1999, Barlund et al., 2008). In a recent study, Jaureguiberry et al. (2017) found that in 20% of repeat breeder cows (healthy cows with > 3 AI) the presence of fluid (≥2 mm) in uterine lumen, was associated with lower conception rates (odds ratio=0.46). The uterine fluid accumulated in the lumen may interfere with embryo development or conceptus attachment (Chien et al., 2002, Lu et al., 2013).

1.4.2 Invasive uterine assessment

The non-invasive methods of uterine assessment may be followed with more invasive investigative sample collection methods in suspect cases after initial screening. Cattle practitioners seldom collect uterine samples in the field because the collection is laborious and expensive, as sample processing and interpretation requires a specialised laboratory. There is also the concern that the collection of uterine samples, and subsequent tissue trauma, may jeopardise the sperm or embryo survival in the uterine environment due to inflammatory responses which could negatively affect conception rates (Zaayer and Van der Horst, 1986, Sheldon et al., 2006). In other species, such as mares and women, uterine sampling is a routine procedure in cases of infertility and the mechanical injury caused by uterine sampling does not affect (Watson and Sertich, 1992) nor impair (Potdar et al., 2012) pregnancy rates. However, the impact of these procedures on the reproductive 16

Chapter 1 Review of literature performance of cattle is unknown. Uterine sampling using invasive procedures is still limited and a number of different approaches have been described to recover endometrial samples for assessment of uterine health.

Methods include the use of a cytobrush (CB), Cytotape (CT), uterine lavage (UL) and endometrial biopsy (EB). These invasive methods allow the collection of epithelial and inflammatory cells (CB, CT, UL), luminal secretions (UL) and uterine tissue (EB). The samples obtained can be used to accurately diagnose the uterine status by using a single or combined assessment with cytological smears (Kasimanickam et al., 2004, Pascottini et al., 2015), bacteriological culture (Bonnett et al., 1991a), protein analysis (Beltman et al., 2014), histopathology (Chapwanya et al., 2010), and gene expression (Munoz et al., 2012) These tests are better indicators of uterine health than the subjective non-invasive methods. Here I describe three common approaches with different degrees of invasiveness which I used experimentally.

1.4.2.1 Collection of endometrial cells

Cytobrush (CB) is a minimally invasive method used to collect endometrial epithelial cells, bacteria and infiltrating PMNs from superficial layers (See Figure 1.3). Two versions of the CB device can be employed in cattle practice. Breeding companies offer a disposable semi-rigid device with a small sterile brush attached (2 cm length, 0.5 cm diameter) which is ideal to sample adult cows (Cytology Brush; Minitube GmbH, Germany). In the other approach the device is modified for use in large animals. A pap endocervical brush (employed in human gynaecology) is attached into a rigid, stainless steel gun device used for AI.: In both cases, once the brush is assembled and the device covered with a sanitary sheath, the CB is passed through the vagina to the external cervical os, and advanced through the cervix into the uterine horns. At this point, the brush is exposed and rotated clockwise along the endometrial layer to collect endometrial cells. The CB is retracted into the stainless steel device, prior to removal from the uterus (Kaufmann et al., 2009, Overbeck et al., 2011).Slides for cytology are prepared by rolling the brush onto a microscope slide and air dried.

Chapter 1 Review of literature

Figure 1.3 Cytobrush device covered with a plastic sheath designed for cattle (Cytology Brush; Minitube GmbH, Germany).

The advantages of using CB include the high success rate of obtaining samples and the quality of the material recovered. Plontzke et al. (2010) and Sens and Heuwieser (2013) reported than more than 95% of the attempts for recovering a sample using CB were successful. Kasimanickam et al. (2005c) found that assessing cytological smears obtained by CB was more reliable than cytological examination of UL because endometrial cells were damaged during the centrifugation process and also due to the increased time needed to recover the fluid in UL. Although CB is a less harmful procedure with low likelihood of causing a deleterious effect on endometrial health or function, it is still uncertain whether CB increases the number of PMN in the endometrium after sampling (Deguillaume et al., 2012). In addition, CB only collects cells from superficial layers from the luminal epithelium and thus, specific changes in the submucosa glandular region may not be detected when analysing samples obtained by this method (Ulbrich et al., 2013).

Chapter 1 Review of literature

Cytology has been developed and implemented for the diagnosis of uterine health. A scraping of the endometrial surface is taken and then microscopically the ratio of polymorphonuclear leukocytes (PMNs) to epithelial cells is estimated and expressed as a percentage of PMN which can then be used to categorise cows as healthy or with subclinical or clinical endometritis (Barański et al., 2012, Madoz et al., 2014). Controversy exists about the interpretation of cytology samples. Although cytology is a simple method, easy to use, that provides valuable information about the intrauterine environment, it is not considered the standard gold method to diagnose endometrial status. This is basically due to the lack of agreement about the threshold values of PMN cells and the possible false negatives that can be diagnosed. In subclinical endometritis, the threshold values mentioned in different studies ranged from 5 to 18% (Kasimanickam et al., 2004, Barlund et al., 2008, Barański et al., 2012). These differences in the percentages of PMN between the studies are based on the inflammatory changes in the uterus after parturition showing higher percentage of PMN during first weeks postpartum, decreasing by the time of uterine involution (Barański et al., 2012). In addition, the sensitivity and specificity of cytological samples may result in false negatives (Ahmadi et al., 2006, Galvao et al., 2011, Ghasemi et al., 2012). During oestrus and metoestrus, the increase in uterine flow caused by follicular E2 secretions triggers a physiological influx of PMN cells. Ahmadi et al., (2006) found 17.6% infiltration of PMN (P<0.05) during metoestrus in samples taken from the cervix of healthy heifers without signs of uterine infection which were above the threshold values used to diagnose subclinical endometritis. Opposite results were described by Madoz et al. (2014) who found less than 5% of PMN during the oestrous cycle collecting the sample from the uterine body. Overall, these studies indicate that for the cytology interpretation the stage of the oestrous cycle or postpartum stage and the site of sampling should be considered during in the assessment of uterine function.

1.4.2.2 Collection of uterine fluid

Collecting uterine fluid offers a non-disruptive and minimally invasive approach to study uterine health. Two methods for uterine fluid collection have been described: uterine lavage (UL) and aspiration of luminal uterine fluid (ALF) (See Figure 1.4).

UL is the most common method employed to recover uterine fluid by flushing the uterus with saline solution generally using a similar non-surgical approach to that used to recover embryos. Briefly, for UL a sterile catheter is introduced into the uterine horn and 20-50 mL of sterile 0.9% sodium chloride solution are infused through the catheter. The uterus is gently massaged and the ﬂuid is recovered by aspiration into a 50 ml syringe (Barlund et al., 2008, Machado et al., 2012, Cheong et al., 2012). UL provides a more representative sample of luminal contents than a swab or CB

Chapter 1 Review of literature because it harvest cells from a larger endometrial area. A disadvantage of the technique is that often not all the volume is recovered after flushing. In mares, Cocchia et al. (2012) recovered only 50- 60% of the initial volume infused (20 ml) whereas in postpartum cows, Barlund et al. (2008) recovered 20% after flushing with 10 ml SSF. Kasimanickam et al. (2005a) failed to recover the lavage fluid in 17% of the cows flushed. Generally, a bigger volume of between 20-50 ml gives the best recovery. In addition, centrifugation may impair the quality and morphology of the recovered cells and the uterine mucus attached to the cells interferes with the cellular fixation to the slides, hence cytological smears can be excluded because of the poor quality of the samples affecting the diagnosis or the uterine disease (Barlund et al., 2008).

It is also possible to collect undiluted uterine luminal fluid by aspiration of the content of the uterine horn using a flexible catheter similar to the one used for deep intrauterine insemination. Velazquez et al. (2010) claims that one of the disadvantages with fluid aspiration is the marked variation in volume recovered within animals (5-400 µl/animal) since the quantity of uterine fluid varies with stage. Uterine secretions are higher during the secretory luteal phase than the proliferative follicular phase (Gray et al., 2002). Therefore, with such small amounts of fluid recovered, the sample needs to be diluted and this should be considered depending on the analysis desired.

The collection of uterine fluid (UF) also allows the recovery of endometrial cells for cytological assessment in diagnosing endometritis. However, the yield of endometrial cells recovered from uterine fluid is less and this affects subsequent cytological analysis and this is worse than in samples recovered from CB (Kasimanickam et al., 2005b, Barlund et al., 2008, Cocchia et al., 2012). As mentioned before, the composition of UF is a complex mixture of molecules and substances derived from genes expressed in the endometrium or transported from maternal blood (Beltman et al., 2014). The proteome profile of the histotroph has been widely studied in cattle. The proteome varies depending on the method used for recovery. By UL, a more extensive uterine area is flushed, hence UF is more diluted, whereas by aspiration the neat fluid is specifically collected at the site of sampling without dilution (Hannan et al., 2012). The proteome expression of the histotroph varies during the oestrous cycle, early pregnancy and pathological conditions (Berendt et al., 2005, Choe et al., 2010, Faulkner et al., 2013). Faulkner et al. (2013) compared uterine flushings from early and late dioestrus. They found that 20 proteins were highly expressed in late dioestrus (Day 15) compared to the protein profile expressed in early dioestrus (Day 7) confirming the dynamic changes of histotroph secretion which is regulated by P4. The proteome of the histotroph also changes with the presence of the embryo in the uterus. Berendt et al. (2005) and Ledgard et al. (2009) found a higher abundance of proteins needed for trophoblast invasion into the endometrium in pregnant animals during the pre-attachment period (16 to 18 days). 20

Chapter 1 Review of literature

Figure 1.4 a) Ghent device® for aspiration of luminal fluid. Outer-inner tube 50 cm length with a flexible post-cervical insemination catheter x70 cm length b) Uterine lavage (UL) using Ghent device®. The cannula was inserted via transvaginal into the uterine horn and the fluid was collected by aspiration using a 50 ml syringe.

Chapter 1 Review of literature

1.4.2.3 Collection of endometrial tissue

Endometrial biopsy involves the removal of a portion of the lining of the uterus for microscopic, bacteriologic and molecular assessment. This technique has been used for more than thirty years in mares (Ricketts, 1975) and cattle reproduction (Studer and Morrow, 1978), as a diagnostic procedure for uterine health. In mares, EB is routinely performed as part of the breeding soundness examination (BSE). The technique of collecting endometrial biopsies has employed a variety of instruments. Skjerven (1956) summarised the biopsy devices that were available in the 50’s. These devices consisted of an outer tube with a window in the tip and an inner sharpened tube connected to an external syringe. The sample was obtained by suctioning the endometrial tissue into the window but only portions of the endometrium were obtained. Further on, Skjerven (1956) designed a 52 cm single tube device with a rounded cutting tip’s jaw that was manipulated with an external hand-grip that allowed clipping off the biopsy tissue, measuring 4x3x2 mm. (Figure 1.5).

Most of the devices used to collect uterine tissue in cattle have been adapted from those used in mares where the transvaginal per cervical approach allows the use of rigid large outer diameter forceps (above 0.8 cm). In the 70’s Ricketts (1975) described the use of a 40-cm long Yeomanns’ basket punch device or 60-cm long Jackson forceps in the diagnosis of endometrial diseases in mares. Most of the studies in cattle have used these devices long (40-60 cm) rigid or semi-flexible teeth or punch jaw-forceps designed for mares (Bonnett et al., 1991b, Chapwanya et al., 2010, Meira et al., 2012). The size and amount of tissue recovered varied with the cutting tip area. Basket- jawed devices (i.e. Yeoman Hauptner device) collect uterine fragments of approximately 2x1 cm size (Ricketts, 1975) with approximately 300-500 mg weight of tissue (Studer and Morrow, 1978) (Figure 1.6).

Chapter 1 Review of literature

Figure 1.5 Biopsy devices used for uterine sampling in cattle a)Ten Thije’s model b) Folmer Nielsen’s model (Oskerven, 1956) c) Storz® device d) Hauptner Yeoman Uterine Biopsy Punch.

Chapter 1 Review of literature

The biopsy procedure is performed mainly by the transcervical approach. Before the procedure, the vulva and surrounding areas are cleaned and epidural anaesthesia is provided to decreased rectal contractions and minimise pain (Williams et al., 1987, Mann and Lamming, 1994). To avoid uterine contamination, the device is covered with a sanitary sheath while advancing through the vagina into the cervical orifice, guided by manipulation per rectum. The protective sheath is ruptured at the external cervical os and the device guided into the uterine horn/body. The forceps jaws with the cutting edges are opened and a piece of the endometrium is pressed into the jaws. The jaws are closed, and the device rotated 90 degrees to trap the tissue (Messier et al., 1984, Bourke et al., 1997).

Figure 1.6 Biopsy Storz® device (10366L Karl Storz, Germany) with round cup single action jaws, 2.5 mm, length 55 cm; b) Bovine uterine tissue collected using different devices. Left-hand side, Storz® device, right-hand side, Yeoman Hauptner®

Chapter 1 Review of literature

The histopathological assessment of the biopsied tissue is considered the gold standard method for uterine health diagnosis. The collection of uterine tissue by EB provides information about the histological status of all the endometrial layers. The proper visualization of the mucosa and submucosa layers allows the identification of inflammatory cells and the assessment of uterine glands and blood vessels (Galvao et al., 2011, Madoz et al., 2014). In mares, EB is also used as a prognostic indicator of the ability of the mare to carry a foal to term (Snider et al., 2011). Kenney (1978) developed a grading system to evaluate uterine tissue to facilitate the description of endometrial inflammatory patterns and the severity of inflammation. Although this system was first suggested in the 70s it is still in use by equine practitioners.

In cattle, the assessment of uterine tissue is considered the most accurate method to diagnose uterine health (Meira et al., 2012), however, studies in postpartum cattle failed to positively correlate biopsy scores and future fertility (Studer and Morrow, 1978, Madoz et al., 2014). Nevertheless, EB is seen as the most informative approach for assessment of cattle uterine health and function. Contrary to the standardised and validated method of histopathological assessment in mares, different systems have been described to evaluate uterine samples in cattle. Chapwanyama et al., (2009) suggested a simplified numeric scale method (from 0 to 3), based on infiltration of inflammatory cells (PMNs, lymphocytes and monocytes) with 0 score being considered as healthy uterus, 1 score low-grade lymphocytic infiltration, 2 score moderate lymphocytic-rich infiltration and 3 score indicates severe inflammation with high cellular infiltrates (neutrophils, monocytes) associated with vascular changes. Other studies (Singh et al., 1983, Azawi et al., 2008) had shown that specific morphological changes such as lack of columnar epithelium in the mucosa, the presence of haemorrhage and hemosiderin in lamina propria, periglandular fibrosis, and oedema may be suggestive of endometritis. A more detailed evaluation system has been mentioned by Meira et al., (2012). They proposed a summation score system (from 1 to 26) involving the assessment of epithelium, the lamina propria, endometrial glands and endometrial vessels. They established numerical threshold values according to the area being assessed. Thus, epithelial lesions were scored from 1 to 11, lamina propria from 0 to 8, endometrial glands from 0 to 4 and vessels from 0 to 3. Specimens with a total score above 15 were considered to exhibit endometritis. Using this scoring system, they found 78% sensitivity and 94% specificity for the diagnosis of clinical endometritis which demonstrates that it is an appropriate method to assess endometrial health (See Figure 1.7). Even though the biopsy provides enough material for histopathological and molecular assessment, one question arises from the sample obtained: Is the sample representative of the health of the entire endometrium? A few papers have dealt with these two questions of representativeness and repeatability of the technique. Ricketts (1975) suggested collecting one mid-horn biopsy as

Chapter 1 Review of literature being a diagnostic of the whole endometrium. Blanchard et al. (1987) found low inter-sample variability comparing biopsies taken from both horns in healthy mares, indicating that one sample would be enough to assess endometrial health. In contrast, the representativeness of biopsy apparently varies in pathological conditions. Fiala et al. (2010) evaluated the agreement between two-blinded observers in the diagnosis of endometriosis in biopsies taken from three different uterine positions in mares. The results showed that in just 30% of the cases, the three biopsies had the same grade of endometriosis between observers, indicating that one biopsy alone is not enough to diagnose degenerative changes in the mare’s uterus. No studies comparing biopsies taken from different places within and between horns have been done in cattle. Madoz et al. (2014) assessed the interobserver agreement evaluating biopsies in a scale 1 to 4 for diagnosis of endometritis. They found high agreement between two blinded observers with a weighted kappa coefficient of 0.854 (P< 0,001). Hence, how representative a single biopsied sample from one point in the endometrium is of the whole endometrium remains to be investigated in cattle and requires further research.

It is estimated that endometrial tissue is recovered by biopsy in cattle in 60-85% of the total attempts (Singh et al., 1983, Bonnett et al., 1991b, Mann and Lamming, 1994, Meira et al., 2012). Several factors, like biopsy instrument, the ability of operator, time of sampling (i.e. days postpartum, stage of the oestrous cycle) and individual animal variation could affect the successful rate of always obtaining a representative, high-quality uterine sample for histological analysis (Singh et al., 1983; Bonnett et al., 1991). However, the recovery of the endometrial sample does not always guarantee that all the biopsies are readable or considered of good quality. For the diagnosis of endometritis in cattle, the pathologist assesses the epithelium continuity and infiltration of inflammatory cells, vascular congestion, the integrity of uterine glands and cellular infiltration in submucosa layers (Chapwanya et al., 2009, Meira et al., 2012). Thus, if one of these components is absent, the sample is considered unreadable and difficult to interpreted (Bonnett et al., 1991a). Madoz et al. (2014) found that 16% of the biopsies collected by them were unreadable by a pathologist when taken from postpartum cows for this specific reason.

Chapter 1 Review of literature

Figure 1.7 Histopathological criteria for endometritis diagnosis (Meira et al., 2012, permission requested).

Chapter 1 Review of literature

1.5 Effects of endometrial sampling on uterine health

Despite the fact that this technique is a reliable method to assess endometrial health, the use of uterine sampling procedures is still controversial since it is considered costly, the assays time consuming and requires referring to a laboratory for specialised analysis (Westermann et al., 2010). In addition, some potential adverse effects of sampling on uterine health have been mentioned in the literature.

After performing the uterine sampling procedure, morphological changes may be detected as a result. The degree of endometrial injury and trauma varies according to the sampling method from none to moderate. Using uterine lavage (UL) it is possible to collect endometrial cells from the lumen with almost no impact on the superficial layers. Cytobrush (CB) removes endometrial and inflammatory cells from superficial layers by rolling the brush, against the endometrium causing mild damage to the endometrium. However, the impact on endometrial health and function of the mild trauma caused by UL and CB is uncertain. There are no studies describing either the pressure with which the CB was rolled against the endometrium or the massage pressure that was applied to the uteri to recover the fluid. Therefore, there is no certainty whether after performing these procedures any inflammatory response is induced. In contrast, EB disrupts the superficial layers and this may permit bacteria to invade the endometrium thus causing an inflammatory response after the procedure (LeBlanc et al., 2002). Mann and Lamming (1994) reported the persistence of small scars in the endometrium of cows three weeks after biopsy using the Hauptner device. The clinical exam of 13 biopsied cows did not reveal any sign of uterine or systemic illness within 48 h after the biopsy procedure was performed suggesting infection is a rare occurrence (Chapwanya et al., 2010).

In a recent study, Pugliesi et al., (2014) scanned the uterus after EB and found an increase in endometrial vascular perfusion at 6 h that returned to normal 24 h later. This acute increase in uterine blood flow may reflect a transitory inflammatory reaction in uterine horns after the biopsy procedure. After the mechanical injury or trauma occurs, sterile endometritis can be induced (Granot et al., 2012) and the immune system starts the process of wound healing, where the destroyed tissue is replaced by functional living tissue. After performing biopsy some bleeding may occur (Sheldon et al., 2006) but it’s not clear how long it lasts. Normally, after injury or trauma a short bleeding phase (2-3 h) will occur. The length of bleeding will vary based on the tissue and size of trauma. Highly vascularized tissues will bleed longer than poorly vascularized tissues (Salamonsen, 2003). Thereafter, the immune system induces a rapid onset of a sterile inflammatory reaction (1-3 days) which is characterised by a simultaneous activation of vascular and cellular cascades. 28

Chapter 1 Review of literature

Contrary to microbial infection, where the released bacterial products trigger an inflammatory response, the dying cells release intra and extracellular matrix molecules (damaged-associated molecular patterns, DAMPs) after mechanical injury (Rock et al., 2010, Kono et al., 2014). The immune cells identify these non-pathogenic DAMPs through specialised pattern recognition receptors (PRRs), the toll-like receptors (TLR 1-10) (Kannaki et al., 2011, Shen et al., 2013). Inflammation causes changes in the vascular blood flow and induces the recruitment and migration of PMN cells accompanied by the release of pro-inflammatory cytokines and chemokines (Li et al., 2007). During initial stages of sterile inflammation, neutrophils release pro-inflammatory cytokines such as interleukin IL1, IL6, tumour necrosis factor alpha (TNFα) and osteopontin (SPP1) (Dhaliwal et al., 2001, Herath et al., 2006, Singh et al., 2008, Chen and Nunez, 2010). The pro- inflammatory cytokine IL1α is released in response to the presence of necrotic endometrial cells signalling through the receptor IL1R to induce the release of other chemokine ligands (CXCL 1,-2 and-3). These chemokines are produced by the endometrial epithelium to stimulate neutrophil recruitment (Gabler et al., 2009, Herath et al., 2009, Heppelmann et al., 2015). High endometrial concentrations of IL6 have been detected in sterile inflammation (Chen and Nunez, 2010). Once secreted, IL6 promotes neutrophil maturation and the differentiation of monocytes into mature macrophages and natural killer cells (Singh et al., 2008). It is suggested that IL6 regulates the transition from a neutrophil dominated phase in the initial stages of inflammation to a macrophage- domination during the resolution phase (Tizard, 2012). TNFα stimulates the production of cytokines and chemokines such as IL8 or CXCL8 which is a chemokine that attracts neutrophils to the site of injury or infection in the uterus. The secretion of TNFα induces inflammation by stimulating endometrial cells to produce another cytokine, osteopontin (SPP1), which in turn recruits and activate macrophages (Granot et al., 2012).

The duration and magnitude of the immune response regulate the healing mechanism to avoid persistent tissue inflammation. During the tissue formation phase (lasts days-weeks), growth factors and cytokines secreted by immune cells, restore tissue structure and functionality of the damaged area with the re-establishment of blood flow in the region (Diegelmann and Evans, 2004) For effective inflammatory resolution, mononuclear macrophages develop anti-inflammatory properties changing from a phagocytic mechanism during acute inflammatory phase to a repair phase (Diegelmann and Evans, 2004). Thus, macrophages stimulate the release of multifunctional cytokines that phagocytose necrotic or apoptotic neutrophils (transforming growth factor beta TGFβ) (Sugawara et al., 2010, Godkin and Dore, 1998) avoiding the recruitment of more neutrophils to the inflammatory site. The cytokines released includes both immunosuppressive and anti-inflammatory cytokines (interleukin-10 IL10), vasoactive molecules (prostaglandin–

Chapter 1 Review of literature endoperoxide synthase 2, PTGS2), factors that enhance tissue resolution (granulocyte-macrophage colony stimulating factor, GM-CSF2) and angiogenesis (vascular endothelial growth factor, VEGF) (Tizard, 2012). TGFβ regulates tissue remodelling by stimulating cell proliferation and angiogenesis (Godkin and Dore, 1998). GM-CSF2 is a cytokine that plays important roles in the defence against uterine infection stimulating the phagocytic activity of neutrophils but inhibiting their migration. The injured endometrial epithelial and stromal cells secrete vasoactive molecules or pro-inflammatory prostaglandins. PTGS2 or COX2 is a key enzyme in the production of prostaglandins (prostaglandin F2α, PGF2α and prostaglandin E2, PGE2) (Arosh et al., 2002). The repair mechanism also includes a remodelling of the vasculature and restoration of oxygen to the hypoxic tissue. The phagocytosis of apoptotic neutrophils by macrophages stimulates the secretion of VEGF, which enhances tissue angiogenesis, revascularisation and wound repair (Tasaki et al., 2010).

In mares, it is reported that EB taken at Day 4 post-oestrus induced a shortening of the interoestrous interval due to the endogenous release of endometrial prostaglandin that results in premature luteolysis (Baker et al., 1981, Snider et al., 2011). PGF2α and its metabolite PGFM are not only secreted during physiological luteolysis (Ginther et al., 2010) but they are also secreted after transrectal uterine manipulation (Velez et al., 1991), after performing AI/ET procedures (Scenna et al., 2005, Koblischke et al., 2008) or in cases of endometrial infections in cows (Del Vecchio et al., 1992, Archbald et al., 1998, Mateus et al., 2003) and mares (Watson et al., 1987). To evaluate whether the CL lifespan is affected by the biopsy procedure in cattle, Mann and Lamming (1994) did a repeated collection of endometrial tissues in cyclic cows (2 days interval) in the latter stages of the luteal phase (Days 13 to 17 of the ). They did not find a significant difference in the levels of PGFM released after the procedure, suggesting low or no effect of the procedure on luteolysis. Similarly, Boos et al. (1996) collected endometrial biopsies at different stages of the in five Holstein cows at Days 1, 8, 15 and 19 from the horn ipsilateral to the CL retrieving tissue specimens of 0.4 to 0.6 cm in diameter. The cycle length was 21±1 day and no signs of inflammation were detected in the biopsies.

Chapter 1 Systematic review 1.5.1 Systematic review of the effect of different methods of uterine sampling on subsequent pregnancy rates

One of the main general concerns about manipulation and sampling the uterus in maiden heifers before breeding is any subsequent effect on reproductive performance (Zaayer and Van der Horst, 1986). The mechanical injury and trauma caused using the transcervical approach by metallic devices and the sample collection per se may affect the uterine environment, compromising the survival of sperms cells and the embryo. Some controversy exists about the effect of collecting endometrial samples on subsequent reproductive performance. In subfertile women, local endometrial injuries in the cycle before ET improved the pregnancy rates (Barash et al., 2003, Gnainsky et al., 2010). In cattle, some studies suggest that these procedures impair uterine function (Zaayer and Van der Horst, 1986, Etherington et al., 1988) whilst others report satisfactory pregnancy rates but without making a comparison with control non-sampled group (Rhoads et al., 2008, Chapwanya et al., 2010). Thus, in this section sought to systematically review and summarise existing evidence related to the impact of invasive techniques available for endometrial sampling on pregnancy rates in cattle.

Chapter 1 Systematic review

Abstract

Uterine infections are a common cause of failure of confirmed pregnancy after mating in cattle, particularly in dairy cattle. Accurate diagnosis is important to reduce the economic losses caused by endometritis. Invasive sampling methods for uterine assessment are required to collect cells, fluid or tissue for further analysis. However, discrepancies exist about the effect that mechanical injuries will cause on the uterine environment and the subsequent impact on likelihood of confirmed pregnancy.

This systematic review compared the pregnancy rates in cattle which have undergone uterine lavage (UL), cytobrush (CB) or endometrial biopsy (EB) procedures. Using the Preferred Reporting Items for Systematic reviews and Meta-Analysis (PRISMA) protocol, relevant databases including Pubmed, Web of Science, CAB Abstracts VetMed Resource –Ruminants, Science Direct and Scopus databases were searched. The outcome measured was pregnancy rate after collection of uterine sample(s). A total of six studies, including a total of 3,601 cows, fulfilled the inclusion criteria for the systematic review and allowed the comparison of pregnancy rates between sampled (n= 1,208) and non-sampled cows (n=2,393). The data showed large heterogeneity in population and procedures, hence comparisons between studies were not possible. The results of the systematic review showed that there is not enough data to support the evidence that uterine sampling procedures around the time of breeding are safe for fertilisation and embryo development.

Chapter 1 Systematic review 1.5.1.1 Introduction

High reproductive performance in production animals such as domestic cattle is vital for achieving optimal per capita return. Uterine pathologies such as puerperal metritis, clinical endometritis and subclinical endometritis induce endometrial damage and alter the intrauterine environment predisposing to fertilisation failure and early embryonic losses, thus increasing both calving to conception interval and culling rates (Sheldon et al., 2006, Lopez-Helguera et al., 2012). Therefore, early detection of these uterine infections in individual cows, before breeding or ET is critical.

The diagnosis of uterine diseases often relies on the presence of clinical symptoms such as vulvar or cervical discharge or accumulation of uterine fluid (Gilbert et al., 2005). These clinical signs are usually detected by transrectal palpation, ultrasound and/or vaginoscopy (Sheldon et al., 2006, de Boer et al., 2014). These quick, low-cost diagnostic methods are commonly employed in clinical practice, providing a subjective indication of uterine health. Hence, more uterine invasive investigative sample collection methods such as cytobrush (CB), uterine lavage (UL) and endometrial biopsy (EB) are required. These allow the collection of epithelial and inflammatory cells (UL, CB), luminal secretions (UL) and uterine tissue (EB).

The degree of endometrial injury and trauma varies with the method of sampling from negligible using UL or CB to potentially moderate damage when performing EB. By UL and CB, it is possible to collect endometrial cells from the lumen of each uterine horn, whereas by EB a full thickness sample of the endometrium and in some cases a portion of the underlying myometrium is recovered (depth of tissue varies from 0.4 to 1cm ) (Rhoads et al., 2008, Chapwanya et al., 2010). The impact of these procedures on fertilisation and early embryo development is unclear (Zaayer and Van der Horst, 1986, Sheldon et al., 2006). In mares and women, uterine sampling is a routine procedure which does not apparently adversely affect the likelihood of the sampled female becoming pregnant (Watson and Sertich, 1992). However, the impact of these procedures on the subsequent reproductive performance of cattle is unknown. The objective of this study was to systematically review and summarise existing evidence related to the impact of these uterine sampling procedures on the likelihood of sampled females becoming pregnant.

Chapter 1 Systematic review 1.5.1.2 Materials and Methods

1.5.1.2.1 Search strategy and selection criteria

A systematic review was conducted of studies investigating the relationship between pregnancy rates after collection of uterine samples from cattle by different methods (endometrial biopsy, cytobrush and uterine lavage), using the Preferred Reporting Items for Systematic reviews and Meta-Analysis (PRISMA) protocol. Online searches of literature databases were carried out in Pubmed, Web of Science, CAB Abstracts VetMed Resource –Ruminants, Science Direct and Scopus with no date limitations. A combination of the following search words were used to generate a subset of citations: cattle and (cow* or heifer* or bov* or buffal* or bos) and (uteri* or endometr*) and (cytology* or cytobrush* or cotton swab* or uterine lavage* or uterine aspiration or biops*) and (fertili* or pregnancy rates* or reproductive performance*). Primary papers written in English, Portuguese and Spanish were accepted from peer reviewed journals with no date restrictions. The latest date for search was February 10, 2016.

Eligibility criteria included: 1) all types of studies including observational, experimental and descriptive; 2) all study settings and countries; 3) studies where an endometrial sample was collected from cattle regardless of uterine health before sampling (with or without endometrial diseases) and 4) studies reporting pregnancy rate after the procedure was performed. The full text was then examined and retained if: 1) uterine samples were collected either by endometrial biopsy, cytobrush or uterine lavage, 2) the pregnancy rates reported were associated with the procedure itself and, 3) the pregnancy rates of sampled animals were compared to non-treated animals. Duplicate citations were excluded. Citations were also excluded if the information was published in reviews, book chapters or conference proceedings, although their reference lists were examined for additional studies not identified by the primary search strategy. Studies from non-cattle species (goat, sheep, horse and human) were also excluded. Pregnancy rate (%; PR) was the primary outcome measure.

1.5.1.2.2 Data extraction

All articles selected from the electronic searches and data extraction were assessed by two authors (Orlando Ramirez-Garzon, ORG and Ricardo Soares Magalhaes, RSM). The final decision on the studies to include in the analysis was ORG’s. The following information from each study was tabulated: first author, year of publication, country, season, study population, number of subjects, physiological status, method of sampling, type of sampling device used, number of samplings and pregnancy rate after sampling.

Chapter 1 Systematic review 1.5.1.3 Results

The studies were selected and reported according to the PRISMA 2009 guidelines (Figure 1.8). A total of 1164 studies were identified. After reading the title and applying the exclusion criteria, 1002 studies were excluded. A total of 162 citations were screened and 116 were excluded after reading the abstracts.

Figure 1.8 Selection process of papers for systematic review according to PRISMA guidelines.

Forty-six publications were retrieved for a full-text appraisal, with 40 subsequently excluded because they did not meet the predefined inclusion criteria (Table 1.1). From the excluded manuscripts, seven did not have a control group, two did not sample the uterus, five did not report the pregnancy rates after sampling, two published the same data in two different papers, 24 used the sampling methods to establish threshold values for endometritis, and the resultant pregnancy rates obtained were only for cattle diagnosed with endometritis.

Chapter 1 Systematic review Table 1.1 Characteristics of studies excluded for systematic review

Type of Study Exclusion criteria intervention

Bacha and Regassa (2010) UL Pregnancy rates from cows diagnosed with endometritis diagnosed by cytology from uterine lavage

(Barański et al., 2012) CB Established cytology thresholds of subclinical endometritis by CB (Baranski et al., 2013) CB Determine impact of cytological endometritis on fertility rates (Barlund et al., 2008) CB, UL Compared thresholds of different techniques for diagnosis of endometritis

(Barrio et al., 2015) CB Diagnosed subclinical endometritis by Cytobrush and determined the impact on reproductive performance

(Brodzki et al., 2014) CB Pregnancy rates were related to presence/absence of subclinical endometritis

(Carneiro et al., 2014) CB Evaluate reproductive performance in cows diagnosed with subclinical endometritis

(Carneiro et al., 2013) UL Evaluate the cytological endometritis diagnosed by uterine lavage in reproductive performance

(Carvalho et al., 2013) CB Assess impact of neutrophils before superovulation on embryo recovery rates

(Chapwanya et al., 2010a, Chapwanya et al., 2009b, EB No control group. Same data in three studies Chapwanya et al., 2012)

(Cheong et al., 2011, Cheong et a) Determine risk factors for subclinical endometritis and its effect on reproductive performance b) Determine the use UL al., 2012) of leucocyte esterase strip as indicator of endometritis

(Couto et al., 2013) CB, UL Used uterine fluid to measure leucocyte esterase activity to diagnose endometritis (de Boer et al., 2015) CB Assessed inflammatory cells and reproductive performance

(Deguillaume et al., 2012) CB Correlate samples taken from the cervix and endometrium with endometritis. Provided threshold values for cervicitis

(Denis-Robichaud and Dubuc, CB Used leucocyte esterase test for endometritis diagnosis and its effect on reproductive performance 2015)

Chapter 1 Systematic review

(Dubuc et al., 2010) CB Assessed cows with cytological endometritis and evaluate the reproductive performance

(Hartmann et al., 2016) EB+ CB Correlation between cytological cervicitis and endometritis and reproductive performance

(Kasimanickam et al., 2004, a) CB b) None a) Validated cytology for diagnosis of endometritis b) They did not use any uterine sampling procedure Kasimanickam et al., 2006)

(Katagiri and Takahashi, 2006) EB No control group

(Lopez-Helguera et al., 2012) CB No control group

(Machado et al., 2012) UL Used optic density of uterine fluid to diagnose clinical endometritis. Fertility rate is not mentioned

CB Used endometrial cells to establish cut-off values of endometritis in grazing cows

(McDougall et al., 2007, a) None b) CB a) Did not analyse any sampling procedure in the uterus b) Correlated %PMN and fertility rates McDougall et al., 2011) (Plontzke et al., 2010) CB Assess % PMN for subclinical endometritis and pregnancy rates (Prunner et al., 2014) CB Fertility rate determined by postpartum uterine health (Ricci et al., 2015) UL Used %PMN to determine subclinical endometritis (Salasel et al., 2010) UL Used % PMN to determine subclinical endometritis in repeat breeder cows (Santos et al., 2009) UL Determine prevalence of subclinical endometritis in B. indicus cows (Senosy et al., 2012, Senosy and CB Fertility rates based on uterine health diagnosed during postpartum Osawa, 2013) (Sens and Heuwieser, 2013) CB No control group

(Studer and Morrow, 1978) EB Pregnancy rate is not mentioned after procedure (Werner et al., 2012)] CB Pregnancy rate is not mentioned after procedure (Westermann et al., 2010) CB Pregnancy rate is not mentioned after procedure (Zaayer and Van der Horst, EB Pregnancy rate is not mentioned after procedure 1986)

Chapter 1 Systematic review Six studies including 3,601 animals (sampled=1,208; not sampled=2,393) met the inclusion criteria and the PR were compared between sampled animals and non-sampled animals. The characteristics of the selected studies are summarised in Table 1.2. Three of these were prospective cohort studies (Kaufmann et al., 2009, Cheong et al., 2012, Pugliesi et al., 2014), two studies were randomized controlled studies (RCS) (Etherington et al., 1988, Thome et al., 2016) and one a case-control study (Goshen et al., 2012). In the prospective cohort studies, the experiments compared PR of cytobrush (Kaufmann et al., 2009), uterine flushing (Cheong et al., 2012) or uterine flushing and EB (Pugliesi et al., 2014) with PR in non-sampled animals. In the RCS, one study compared EB (Etherington et al., 1988) and the other compared uterine flushing (Thome et al., 2016) with PR with no uterine sampling. In the case-control study (Goshen et al., 2012), the effect of EB in postpartum cows was compared with non-biopsied cows.

1.5.1.3.1 Uterine lavage studies

Cheong et al. (2012) performed a prospective cohort study comparing the effect of uterine lavage to collect endometrial cells (n=705) with no uterine sampling (n=1992) on the reproductive performance of healthy Holstein cows. The selection criteria included primiparous and multiparous cows within 40-60 d postpartum, not inseminated and without vaginal discharge or systemic illness. The reproductive performance was assessed during a 210 day period after uterine sampling. The mean interval from sampling to first service was 19.4 days. In primiparous cows, the PR to first service was lower in sampled cows compared to cows which were not sampled (31.2% vs. 36.5% OR for pregnancy =1.03; 95% C.I. 0.80-1.33; P= 0.82), whereas in multiparous cows, PR was similar in both groups (29.1% and 28.1% sampled and non-sampled cows, respectively).

In a randomised controlled study, Thome et al. (2016) evaluated the effect of collecting endometrial cells by uterine lavage in postpartum Nellore cows (50-70 days postpartum). In 35 cows, the uterine lavage was performed 4 h after timed-artificial insemination whilst 93 were not sampled. No significant differences in PR were found between sampled and non-sampled groups (54.2% vs. 56.7%, respectively P> 0.05).

In a prospective cohort study, Pugliesi et al. (2014) evaluated the effect of uterine flushing 6 days after timed artificial insemination on pregnancy rates at Day 30 and 60. The uterine fluid was collected after infusion of 20 ml PBS, from the horn contralateral to the CL from multiparous B. indicus cows. After the procedure, all cows received a non-steroidal anti-inflammatory treatment (Flunixin Meglumine, 1.1 mg/kg BW, IM) and an antibiotic (Penicillin-Streptomycin 6 000 000 IU) drugs. The PR were similar (P> 0.1) between flushed and non-sampled cows at Day 30 (28.6% (n=10/35), and 40.5% (n=15/37), respectively), but the pregnancy rates decreased significantly in 38

Chapter 1 Systematic review flushed cows (P<0.004) compared to control at Day 60 (17.1% (n=6/35) and 40.5% (n=15/37), respectively).

1.5.1.3.2 Cytobrush studies

In a prospective cohort study, Kaufmann et al. (2009) evaluated the effect of CB sampling on the uterus 4 h after artificial insemination on pregnancy rate to first service in cows calved at 65 days. PR was similar for sampled and non-sampled cows (43.3% vs. 41.7%, P> 0.05) and PR was higher in primiparous than multiparous cows (54.3 vs. 38.5% P< 0.05).

1.5.1.3.3 Endometrial biopsy studies

In a case-control study, Goshen et al. (2012) randomly selected fifty-four Holstein cows calved approximately 67 days previously to undergo EB; 157 control cows were paired matched to the sampled cows. The effect of the biopsy on PR to first artificial insemination was calculated using binary logistic regression. The interval from biopsy to first AI was 40.5 days (range 5-111 days). The PR and days from calving to conception in biopsied cows (44.4%; 147.3 days) did not differ significantly from control cows (38.9%, 150.8 days).

Etherington et al. (1988), conducted a randomised controlled trial on 130 postpartum dairy cows and evaluated the effect of postpartum uterine biopsy on PR to first AI and calving to conception interval. All cows were biopsied between on Day 26 and/or 40 postpartum. Uterine biopsy increased the interval from calving to first service (89 days biopsied cows versus 81.5 days for control cows; P= 0.07). However, the PR to first AI for biopsied cows (n=92; 37%) was not significantly different from non-biopsied cows (n=69; 39%).

In a prospective cohort study, Pugliesi et al. (2014) evaluated the effect of collecting EB 6 days after timed artificial insemination, from the horn contralateral to the CL on pregnancy rates at Day 30 and 60. Uterine tissue was collected using the Hauptner and Eppendorfer device. After the procedure, all B. Indicus received a non-steroidal anti-inflammatory treatment (Flunixin Meglumine, 1.1 mg/kg BW, IM) and an antibiotic (Penicillin-Streptomycin 6 000 000 IU) drugs. The PR were similar (P> 0.1) between biopsied and control in both Day 30 (31.6 and 40.5%, respectively), and Day 60 (26.3 and 40.5%, respectively).

Chapter 1 Systematic review Table 1.2 Pregnancy rates after endometrial sampling in cattle.

NM: Not-mentioned: RCS: Randomised controlled study; CS: Cohort study: CC: Case control study. B: Beef; D: Dairy; Nel: Nellore;; Guz: Guzerat;Pri-Mul: Primiparous-Multiparous; I: II: III: IV:UL: Uterine lavage; EB: Endometrial biopsy; CB: Cytobrush *Time of breeding after intervention; ** 30 days of pregnancy assessment after breeding; *** 60 days of pregnancy after

Chapter 1 Systematic review

1.5.1.4 Discussion

The aim of this review was to determine the likely impact of uterine sampling procedures on the subsequent PR of treated cattle. Our results indicate that PR were similar between non-sampled and sampled animals using cytobrush (CB), uterine lavage (UL) and endometrial biopsy (EB) if the procedure was performed before breeding or a few hours after insemination. The heterogeneity of the selected studies in terms of type of cattle used (dairy and beef), physiological stage during sampling (early postpartum (Etherington et al., 1988) or late postpartum > 45 days (Goshen et al., 2012) and the interval from sampling to breeding (i.e. 4 h (Kaufmann et al., 2009, Thome et al., 2016) or. 20-40 days (Cheong et al., 2012, Goshen et al., 2012) did not allow comparison between different methods to determine whether one method affects the PR more than any other. We also found little data available in the literature that had estimated the effect of uterine sampling procedures on pregnancy rates. Most studies had been performed collecting uterine samples in postpartum period during uterine involution and the long interval from sampling to AI (>30 days) (Goshen et al., 2012) difficult the interpretation on PR.

These results suggest that the potential injury caused by sampling procedures before breeding or few hours after AI has a minimal effect on the uterine environment and by the time of embryo arrival, the uterus is healed. Color Doppler transrectal ultrasonography has been used in cattle to assess uterine blood flow after endometrial injury as an indicator of inflammatory response after artificial insemination (Oliveira et al., 2014) or endometrial sampling (Pugliesi et al., 2014, Thome et al., 2016, Debertolis et al., 2016). An increase in uterine tissue perfusion was observed within 4 h of the procedure but the uterine hemodynamic returned to baseline by 24 h, indicating that these procedures may induce a short acute inflammatory response. Although previous reports indicate that performing UL, induces endometrial irritation caused by the fluid (Brook, 1993) or by the device (Kasimanickam et al., 2005b), studies in mares (Ball et al., 1988, Linton and Sertich, 2016) and women (Hannan et al., 2012) have demonstrated that UL did not induce significant morphological changes to the endometrial tissue (Koblischke et al., 2008). Cheong et al. (2012) and Thome et al. (2016) performed UL and CB 4 h. after insemination and found that the procedures did not affect the PR in treated mares (Brinsko et al., 1991). Similarly, Salilew-Wondim et al. (2010) collected CB samples from the uterine body in Simmental heifers during the pre-transfer cycle at Day 7 and 14 of the oestrous cycle before ET on Day 7 of the subsequent cycle (nearly 14 days apart from the last procedure). The PR obtained after ET was 41% which is similar to that achieved after transfer of in vivo-derived embryos (Pontes et al., 2009).Therefore, it seems likely that UL and CB did not adversely affect fertilisation and early embryo development when it’s performed before AI. 41

Chapter 1 Systematic review

However, it is still uncertain whether sampling the uterus close to the time of embryo arrival in the uterine horn (Day 5-7) adversely affects embryo survival. In this review only one study (Pugliesi et al., 2014) assessed PR after performing UL or EB from the horn contralateral to the CL six days after insemination. Although PR were not significantly different at Day 30 between sampled and non-sampled cows, the PR at Day 60 were significantly lower in cows that had UL performed 6 days after insemination. Whether this increased late embryonic-early foetal mortality can be attributed to either the early removal of unknown histotrophic factors subsequently needed for conceptus later stage development or even survival or the uterine inflammatory reaction itself induced by UL is uncertain. The histotroph contains proteins, amino acids and lipids essential for early embryo development (Ribeiro et al., 2016), and any induced change in the histotroph as a result of uterine sampling may subsequently adversely affect early placental development (Dorniak et al., 2013). However, aspiration of histotroph has been performed in women on the same day of ET without affecting PR (van der Gaast et al., 2003, Boomsma et al., 2009a, Salamonsen et al., 2013). Differences in attachment process between women and cattle may explain these discrepancies.

It is possible to speculate that after using more invasive procedures such as EB, the trauma and subsequent inflammatory and wound healing reaction, will adversely affect embryo development and survival. However, the findings from the studies present in this review showed that PR after EB were not affected by the type of cattle (beef or dairy), uterine and ovarian stage (uterine involution/anoestrus), parity (heifers or multiparous cows), number of interventions (single or multiple procedures) and the type of device used. Etherington et al. (1988) found that cows that underwent uterine biopsies from both horns at Day 26 and/or Day 40, had a lengthened calving to conception interval compared to non-biopsied cows (135 vs. 115 days, respectively P=0,03), and prolonged interval from calving to first service compared to non-biopsied cows (89 vs. 81 days, respectively P=0,07). However, they did not find significant variation in the pregnancy rates between biopsied and non-biopsied cows (37 vs. 39% respectively). In a larger study in high-yield Holstein cows (n=54), Goshen et al. (2012), evaluated the effect of biopsying high-yield milking cows after uterine involution (between 44-104 days postpartum) on pregnancy rates. They found that biopsied cows tended to have higher pregnancy rate when breeding the cows around 40 days after biopsy than non-biopsied cows (44.4 vs. 38.9, P=0.146, respectively). These results suggest that performing biopsies before the completion of uterine involution, may slow down the complete recovery of the endometrium postpartum but if the biopsy is performed after involution apparently it does not affect the conception rates.

Chapter 1 Systematic review

In other studies that were not considered in the systematic review for not fulfilling the selection criteria, pregnancy rates have been assessed after EB. Chapwanya et al. (2010) evaluated the effect of three consecutive endometrial biopsies in the same uterine horn at Day 15, 30 and 60 on pregnancy rates in postpartum cows (n=13). They reported pregnancy rates of 77% performing the first AI 30 days after the last biopsy. Similarly, Rhoads et al. (2008) collected three endometrial biopsies/cow (n=33) from both horns at different time (3 days before oestrus, during oestrus and 4 days after insemination) and reported pregnancy rates in biopsied cows of 52% which is considered within the normal range of pregnancy rates in dairy postpartum cows bred by AI. Studies in infertile women with repeated implantation failure (RIF) have demonstrated improvement in PR after a mechanical injury caused by EB (Potdar et al., 2012). It is reported that causing mechanical trauma to the endometrium in the cycle before ET (Barash et al., 2003, Karimzadeh et al., 2009) or during the ongoing cycle (Boomsma et al., 2009a, Gnainsky et al., 2010, Huang et al., 2011), actually improves endometrial receptivity and pregnancy success. A systematic review, (Potdar et al., 2012) reported the pregnancy outcomes of seven controlled studies (n=2062) in women with RIF. They found that subfertile women, who underwent endometrial trauma (hysteroscopy or EB) in the cycle before ET, were 70% more likely to get pregnant compared with control, non-treated patients suggesting a positive effect of injury on endometrial receptivity. It was suggested that the improved PR after mechanical trauma was due to a pseudo-decidual reaction and the development of a positive injury-induced inflammatory reaction that improves implantation (Karimzadeh et al., 2009, Gnainsky et al., 2010, Almog et al., 2010). To date, the molecular and cellular repair mechanism reported in women after endometrial injury, have not been studied in cattle. In conclusion, there is not enough evidence in the literature to conclude whether uterine sampling procedures around the time of breeding are safe for fertilisation and embryo development. Differences in the type of procedure and the time of sampling difficult make further comparisons between studies. Further studies with much larger numbers of cattle are needed to verify the effect of endometrial sampling on pregnancy rates in cattle.

Chapter 1 Aims and objectives

1.6 Aims and objectives of the thesis

It is evident from the literature that endometrial biopsy despite being an invasive technique provides useful information about uterine health but it has been infrequently used in cattle for the study of endometrial receptivity. The identification of essential endometrial markers to predict the likelihood of pregnancy in cattle is still a challenge. The information on endometrial receptivity has been derived mostly from studies performed in postpartum dairy cattle where the uterine environment has been modified, using moderately invasive methods that assess endometrial cells and uterine fluid, which are not the most sensitive type of sample indicative of the endometrial status.

Thus, an adequate model to identify biological markers predictive of receptivity would require 1) collecting samples from animals whose uterine environment has not been modified, 2) using a procedure that provides consistent material for transcriptome analysis 3) without affecting further reproductive performance 4) at a time point in the that is synchronised with the embryo stage.

The hypothesis being tested by this research is that performing endometrial biopsy in beef heifers before the time of embryo arrival in the uterus would not affect the subsequent endometrial function and thus development and attachment of the embryo. The specific aims of this thesis are:

1. To determine whether performing uterine biopsy in beef heifers using the Storz device is a reliable and consistent method to recover samples for transcriptome analysis. 2. To characterize the uterine morphological changes after endometrial biopsy 3. To evaluate the duration and severity of inflammation in the endometrium after biopsy 4. To determine whether endometrial biopsy performed during metoestrus or dioestrus effects CL lifespan.

Chapter 2 Uterine sampling methods

CHAPTER 2. INVESTIGATION OF THE PRACTICAL EFFICACY OF DIFFERENT UTERINE SAMPLING METHODS

Chapter 2 Uterine sampling methods Introduction

Collecting endometrial samples in heifers is not a common procedure in cattle practice. In the development of a model of endometrial receptivity, recovering samples from maiden heifers where the uterine conditions have not been modified by breeding, pregnancy or parturition would be a benefit. The main challenges for obtaining samples in the study of endometrial receptivity in heifers are 1) to access the uterus through the narrow cervical lumen and 2) to recover a good quality sample to perform further molecular procedures. Different approaches to collect endometrial samples using different type of device have been described. The first study in this chapter aimed to select the most consistent and reliable method for uterine sampling in heifers and determine whether the samples collected provide enough high-quality material to perform transcriptome analysis.

Chapter 2 Uterine sampling methods 2.1 Evaluation of different methods of uterine sampling from tropically adapted beef heifers

Abstract

This study aimed to compare the feasibility of performing invasive endometrial sampling procedures in heifers under field conditions and to select the most suitable method to collect endometrial samples for transcriptome assessment. In Experiment 1, an optimised ovulation synchronisation protocol was applied (n=19) to B. indicus heifers and uterine samples were collected at Day 7 of the stimulated oestrous cycle using four techniques: uterine lavage (UL n=5), aspiration of luminal fluid (ALF n=5), cytobrush (CB n=5) or endometrial biopsy (EB, n=4). Sample specimens were successfully recovered by UL (3/5), CB (5/5) and EB (3/4) whereas ALF was unsuccessful (0/5). Although logistically, CB was the most reliable technique, the excessive handling and manipulation required to set the device and collect the sample increased the probability of DNA contamination. In Experiment 2, uteri (n=8) from the local abattoir were collected and endometrial tissue was sampled using Storz® device from both horns to optimise RNA extraction and quality as well as sample storage conditions (liquid nitrogen or RNA later). The weight of the biopsied tissue obtained using Storz® device was 32.3 ± 13.6 mg, RNA yield was 72.9 ±39.3 ng/µl with RNA integrity number of 7.3 ±1.1. No significant differences in RNA yield (p-value=0.09) and quality (p-value=0.3) were found between storage in RNA later or liquid nitrogen. In conclusion, EB using the Storz® device was the most efficient and reliable method of uterine sample recovery in heifers for transcriptome analysis.

Chapter 2 Uterine sampling methods 2.1.1 Introduction

Traditionally, cattle practitioners indirectly assess uterine health based upon subjective non-invasive methods of rectal palpation complemented by vaginoscopy and/or transrectal ultrasound (Bonnett and Martin, 1995, Madoz et al., 2013). These procedures are straightforward and the diagnosis of uterine health is obtained rapidly on site without the need for a specialised laboratory. However, these procedures are unable to provide a detailed analysis of endometrial phenotype and its capacity for receptivity of pre-attachment embryos. Thus, moderately uterine invasive procedures such as uterine lavage (UL), aspiration of luminal fluid (ALF), cytobrush (CB) or endometrial biopsy (EB) are designed to collect samples such as endometrial uterine fluid, endometrial cells, or tissue biopsies to improve the assessment of endometrial receptivity. These procedures access the endometrium via a transcervical approach under transrectal guidance to enable the placement of the device into either the uterine body or horns to collect the sample required.

CB collects epithelial and inflammatory cells from superficial layers using a sweeping motion of the brush attached to the device as it is rotated across the uterine surface. After the sample is collected, the CB is retracted into the device, prior to removal from the uterus (Kasimanickam et al., 2005b). This approach is thus gentle and minimally disruptive of uterine function but returns lowest quantities of material for analysis.

Uterine fluid can be collected using a lavage (UL) approach by infusion of saline solutions into the lumen of the uterine horns or directly by aspiration of uterine luminal fluid (ALF). In the UL method, the uterine fluid and superficial endometrial cells are recovered by flushing the uterus with low volume saline solutions (20-50 ml) which are infused through a sterile catheter. Thereafter, the uterus is massaged and the ﬂuid is recovered by aspiration into a 50 ml syringe (Cheong et al., 2012). In the ALF method, the content of luminal uterine horns is aspirated directly using a rounded ridged plastic tip flexible catheter with side openings which is attached to a 50-ml syringe (Velazquez et al., 2010).The features of the flexible catheter decrease the risk of damaging the uterine wall allowing the access to the curled uterus (Verberckmoes et al., 2004). Both UL and ALF are reasonably gentle but in the case of UL substantially less than the volume of saline infused is recovered and thus the process is not particularly efficient. In the case of ALF very small volumes of fluid are recovered and there is a significant chance of minimal uterine trauma and thus contamination by blood.

EB involves the removal of a portion of the lining of the uterus using rigid or semi-flexible teeth or punch jaw-forceps devices with cutting edges. The forceps jaws are opened and a piece of

Chapter 2 Uterine sampling methods endometrium is pressed into the jaws and the tissue is trapped (Chapwanya et al., 2010).This is a more aggressive approach which does result in minor uterine trauma but this is offset to some extent by the fact that the sample contains cells representative of the entire endometrium. The process does not seem to impair fertility in animals in which are artificially inseminated within a few days of biopsy (Goshen et al., 2012). In summary, the ideal method used for sample acquisition for transcriptomic analysis should be fast, repeatable, and yield a high-quality sample with minimal disturbance to endometrial structure and function.

While uterine health examination via transrectal palpation or ultrasonographic visualisation is routinely performed to evaluate heifers before AI or ET (Morotti et al., 2014), endometrial sampling procedures are rarely performed. In heifers, endometrial sampling for microscopic or transcriptomic assessment of uterine function is uncommon because these techniques presents two main issues: (1) The cervix is difficult to transverse using the traditional large devices (Bonnett et al., 1991b) because the cervical luminal diameter in heifers is between 0.2-04 cm (Drennan and Macpherson, 1966), therefore, the use of slim and semi rigid devices is the most appropriate approach to traverse the cervix to collect endometrial samples in cycling heifers and (2) Obtaining a representative sample of sufficient quantity and quality is a challenging step for experiments requiring RNA isolation, since both quantity (yield) and quality (purity or integrity) of the RNA effects the accuracy of the subsequent molecular analysis.

Differences exist between the RNA yields isolated from samples that collect endometrial cells or tissue. These differences may be related to the method of extraction and the disruption of the starting material (tissue or cells). CB and UL harvest cells from the endometrial lining and lower yield is expected whereas EB recovers the entire endometrium tissue which provides higher RNA yield (Fleige and Pfaffl, 2006).The assessment of RNA integrity (RNA integrity number, RIN) is critical to validate the gene expression results. RNA integrity is classified by a numbering system from 1 to 10, 1 being considered as RNA which is significantly degraded possibly reflecting damage to the RNA which could be caused by inadequate sample manipulation, DNA contamination or by post-mortem degradation changes. A score of 10 is considered the most intact, high-quality RNA that is free of contamination (Fleige and Pfaffl, 2006). Thus, this study aimed to select the most efficient and feasible method of uterine sampling in B.Indicus heifers and assess whether the sample recovered by that method, provides good-quality material to perform transcriptome analysis for endometrial receptivity.

Chapter 2 Uterine sampling methods 2.1.2 Material and methods

2.1.2.1 Animals and management

All the animal procedures were approved by the University of Queensland Animal Welfare Committee under permit AE02430. The experiment was carried out at the academic beef unit of the University of Queensland at Pinjarra Hills (Brisbane, QLD). Animals were enrolled from Nov 2013 to Jan 2014. Animals grazed improved tropical pastures rhodes grass (Chloris gayana), creeping bluegrass (Bothriochloa insculpta), green panic (Panicum maximum), Kikuyu grass (Pennisetum clandestinum), and white clover (Trifolium repens) with water ad libitum

Examination and selection of heifers were conducted by transrectal palpation and RT- Ultrasonography (Sonosite M-Turbo ultrasound system, Bothell, WA, USA) using a 5 to 7.5 MHz linear transducer. The cervix, uterine horns and ovaries were assessed by rectal palpation. Animals with anatomical abnormalities in the cervix (tortuous cervical canal) or asymmetry of the uterine horns were discarded. The presence of ovarian dominant structures, such as CL and DF, were assessed by ultrasound and results recorded.

2.1.2.2 Synchronisation of oestrus

Non-pregnant animals were selected to synchronise oestrus according to a protocol used by Butler et al. (2011). Briefly, at Day -9 oestradiol benzoate (1mg, Cidirol®, Genetics, Australia) was administered intramuscularly (IM) and an intravaginal P4 device (0.786g Cue-Mate®, Bioniche Animal Health lab, Australia) was inserted. At Day -2 the Cue-Mate was removed and Cloprostenol (Estromil®, 250 µg Ilium, Australia) was administered IM. At Day -1, oestradiol benzoate (1mg) was administered IM. (Figure 2.1). Day 0 (oestrus) was determined by twice daily observation for signs of heat (mounting/standing behaviour) or through the use of tail paints or Kamar® heatmount detectors. Only animals that showed sign of oestrus (Day 0) after synchronisation were used for analysis.

Chapter 2 Uterine sampling methods

Figure 2.1 Oestrus synchronisation protocol used in crossbreed heifers/cows: Day -9 Insertion of intravaginal P4 releasing device (IPRID, Cue-mate 0.786 g) and Oestradiol Benzoate (OB), 1mg, IM. Day -2: Prostaglandin F2α analogue (PGF2α, Cloprostenol 250 µg IM) and Day -1: Oestradiol Benzoate (OB, 1mg IM). Day 0: Oestrus

2.1.2.3 Experiment 1: Comparison of methods for sampling uterine material in beef heifers

Bos Indicus heifers (n=19; 24-36 months, average liveweight (LW) 491 kg, range 433-543 kg), were enrolled into this study. Heifers were synchronised using the protocol described in Material and Methods (Synchronisation of oestrus). At day 7 after synchronised oestrus (Day 0: Day of oestrus), a uterine sample was taken from the uterine horn ipsilateral to the CL at a depth approximately 3-4 cm beyond the uterine junction, using one of the following methods: CB (n=5), UL (n=5), ALF (n=5) and EB (n=4).

Prior to sample collection, the vulva and surrounding areas were cleansed with water and disinfected with 70% (v/v) ethanol solution before low dose epidural anaesthesia (Lignocaine 0.2 mg/kg Ilium®, Troy Laboratories, NSW, Australia) was administered in the first coccygeal interspace using a 18’’ gauge needle. 51

Chapter 2 Uterine sampling methods The procedures were performed by two experienced technicians. The sampling devices used for CB, UL/ALF or EB were covered with a sanitary sheath (Minitube, VIC, Australia) to minimise transmission of infectious agent while advancing through the vagina into the cervical orifice, guided by per rectum manipulation. To evaluate the feasibility to get through the cervix with the sampling devices, a modified subjective score, originally utilised during ET procedures (Wright, 1981), was used to estimate ease of passing the device through the cervix into the uterine horn. The scoring system is summarised in Table 2.1.

Table 2.1 A subjective scoring system to assess ease of traversing through the cervical canal into the uterine horn.

Category Score Description

Easy + No difficulty noted to transverse the cervix

Moderate ++ Moderate difficulty in advancement. Some resistance.

Considerable difficulty and resistance to advance through the Difficult +++ cervix into the uterine horn.

Not possible ++++ Impossible to get through the cervix

2.1.2.4 Uterine sampling methods

2.1.2.4.1 Cytobrush

The protocol was adapted from Kasimanickam et al. (2004).The cytobrush (CB) (Cytology Brush; Minitube GmbH, Germany) is a disposable semi-rigid device that was inserted into the uterine horn covered with a sanitary sheath while advancing through the vagina into the cervical orifice, guided by manipulation per rectum. The sample was collected 4-5 cm after the uterine bifurcation of the horn ipsilateral to the CL. The brush was rotated clockwise along the uterine wall to collect the endometrial luminal cells and retracted into the plastic cover (See Figure 1.3).

2.1.2.4.2 Aspiration of luminal fluid (ALF)

ALF was collected using the Ghent device® (Verberckmoes et al., 2004) as described by Velazquez et al. (2010). Briefly, the rigid, hollow steel cannula was inserted through the cervix into the entrance of the uterine horn. Then, in the middle part of the device, a flexible catheter used for deep intrauterine insemination in cattle (Minitube, VIC, Australia) was introduced into the uterine horn 52

Chapter 2 Uterine sampling methods through the luminal space of the rigid cannula and luminal fluid was aspirated by negative pressure connecting a 50 ml syringe to the flexible catheter (See Figure 1.4a).

2.1.2.4.3 Uterine lavage (UL)

The protocol was adapted from Kasimanickam et al. (2004). For UL the same Ghent device® used for ALF was utilised to access the uterine horns. Thereafter, 20 mL of sterile 0.9% sodium chloride solution (Baxter I/V fluids Normal Saline 0.9%, Baxter healthcare Pty, NSW, Australia) was infused through the catheter. After gentle uterine massage, the fluid was retrieved by negative aspiration into a 50 ml syringe (See Figure 1.4b).

2.1.2.4.4 Endometrial biopsy

On the day of biopsy, heifers were examined by ultrasound for the presence of a CL. The protocol was adapted from Chapwanya et al. (2010). Briefly, the vulva and surrounding areas were cleansed with iodine solution and water and epidural anaesthesia (Lignocaine 0.2 mg/kg Ilium®, Troy Laboratories, Australia) was administered. To avoid uterine contamination, the biopsy device was covered with a sanitary sheath while advancing through the vagina into the cervical orifice, guided by manipulation per rectum. Biopsy was performed by transcervical insertion of circular cup forceps (10366L Karl Storz®, Germany) (See Figure 1.6). The protective sheath was ruptured at the external cervical orifice and the device guided into the uterine horn ipsilateral to the CL. The forceps jaws with cutting edges were opened and a piece of endometrium was pressed into the jaws. The jaws were closed, and the device rotated 90 degrees to trap the sample.

2.1.2.5 Experiment 2: Assessment of RNA yield from uterine biopsies

To assess whether the endometrial sample obtained by biopsy using the Storz® device provided enough material to perform transcriptome assessment, eight non-pregnant bovine uteri were collected within 4 h of slaughter and then transported on ice to the laboratory. At the laboratory, the Storz® device was placed in the middle region of both uterine horns, the jaws of the device were opened and two samples per horn were collected. One sample was stored at -20º C in RNA later (Qiagen, VIC, Australia) and the other in liquid nitrogen until RNA extraction.

2.1.2.5.1 RNA extraction

Unless explicitly stated, all reagents used in gene expression were purchased from Qiagen (Qiagen Pty Ltd Chadstone Centre, VIC- Australia).Total RNA was extracted using RNeasy mini kit following manufacturer’s instructions. Uterine tissue was removed from RNAlater and approximately 30 mg of uterine tissue was homogenized in Buffer RLT (contains undetermined 53

Chapter 2 Uterine sampling methods concentration of guanidine isothiocycanate). The cleared tissue lysate was mixed with 700 µl 70% ethanol loaded into an RNeasy mini column and centrifuged at 8000 g for 30 sec. Genomic DNA contaminants were removed by performing a DNase treatment (RNase-Free DNase Set). After washing the column with wash buffer, RNA was eluted with 30 µl of nuclease-free water.

2.1.2.5.2 Quantification of RNA yield and quality

Total RNA was quantified by NanoDrop ND®-1000 (Thermo Scientific, MA, USA). The integrity of RNA was assessed with a Bioanalyzer (Agilent Technologies, CA, USA) following manufacturer’s instructions. RIN values above 5.5 (scale 1 to 9) were used as cut off value for qRT- PCR.

2.1.2.5.3 Statistical analysis

Statistical analyses were carried out in R 3.2.1 (R Core Team, 2015). The mean RNA yield for the two storage conditions RNA later and liquid Nitrogen was compared using the student t-test.

Chapter 2 Uterine sampling methods 2.1.3 Results

a. Experiment 1: Comparison of methods for sampling uterine material in beef heifers

In six heifers (6/19; 31.5%) transcervical catheterisation was unable to be performed. In five of those six unsuccessful attempts, the Ghent device® was unable to be passed through the cervix and one attempt failed to get through the cervix using the Storz® device, hence it was not possible to recover any endometrial sample in these six cases.

In the thirteen heifers where uterine access was achieved using at least one method, success rates for the four sampling techniques varied. A 100% success rate was recorded for the CB (n=5/5), compared to 75% EB (n=3/4), 40% successful ALF (n=2/5), and 60% success with UL (n=3/5). Comparison of the success and description of technical difficulties encountered performing the different procedures is summarised in Table 2.2. Overall, with all the techniques, except ALF, it was possible to obtain an endometrial sample. Technical difficulties in collecting the sample were encountered with both UL and CB. When performing UL the main difficulty was that when 20 ml of SSF were infused not all the fluid was recovered. The volume recovered ranged from (5-15 ml). Although CB samples were obtained successfully in all the animals, the original device required modification to facilitate the access into the uterus. The CB device diameter was greater than the cervix orifice found in heifers and this device was not rigid enough to allow adequate transrectal manipulation. Hence, the device was modified by cutting the brush out from the original device and attaching it to the plunger of an insemination gun to provide the required rigidity and to facilitate transrectal manipulation and access to the uterus.

Chapter 2 Uterine sampling methods

Table 2.2 Qualitative characteristics of endometrial sampling methods, device characteristics and samples recovered in heifers.

Successful Uterine Method attempts Sample recovery Comments access (n) Aspiration luminal fluid 2/5(40%) ++ Nil Device is rigid. Lack of luminal fluid recovery (ALF) Device is rigid. Fluid recovered with blood 25-75% recovery of Uterine lavage (UL) 3/5 (60%) ++ (n=2/3). Insufficient volume for further uterine fluid fluid analysis.

Device is semi-rigid. Wider than cervix orifice. Brush causes minor Cytobrush (CB) 5/5 (100%) +++ Excessive handling. The brush needs to be blood contamination assembled into AI gun.

Small tissue Endometrial biopsy (EB) 3/4 (75%) + Rigid. Requires practice to handle small samples (0.3x03cm) + No difficulty. ++ Moderate difficulty; +++ indicated considerable difficulty; ++++ impossibility to access the uterine horns.

Chapter 2 Uterine sampling methods

b. Experiment 2: RNA optimisation from biopsied samples

The weight of the endometrial tissue recovered from post-mortem uterine samples with Storz® device, the mean RNA yield obtained and the RNA integrity number (RIN) are given in Table 2.3. Interestingly, the RNA storage conditions affected neither the RNA yield (P= 0.09) nor the RIN (P=0.3).

Table 2.3 RNA yield and quality of endometrial tissue recovered by Storz® device from post- mortem uteri

RNA yield (ng/µl) Uteri biopsy Overall RIN (±SD) (mg) Total RNA RNA later/RIN Liquid Nitrogen (±SD) (±SD) / RIN (±SD)

81.3 (±48.9) 64.5 (±26.5) 32.3 ± 13.6 72.9 (±39.3) 7.3 (±1.1) 7.5 (±0.4) 7.1 (±1.5) RIN: RNA integrity number

Chapter 2 Uterine sampling methods 2.1.4 Discussion

In cattle, uterine health is most commonly assessed by indirect procedures such as rectal palpation or ultrasonography. To successfully collect endometrial samples using more invasive procedures requires a method of sampling, an appropriate device and special technical skills. In heifers, the main technical challenge in retrieving uterine samples is to traverse the cervix to gain access to the endometrium. In this study, most of the unsuccessful attempts to traverse the cervix (n=6/19), occurred using the Ghent® device (n=5/6) whilst getting through the cervix using the Storz® device failed in one attempt (n=1/6). These failed attempts could partially be explained by the greater external diameter of the Ghent® device (0.6 cm) (Verberckmoes et al., 2004) compared to the diameter of the internal cervical lumen in heifers (0.2-0.4 cm) that makes it physically unlikely to be capable to traverse the cervix. Individual variability in the bodyweight and reproductive tract development could explain how in some heifers accessing the cervix was successfully achieved. Although the animals selected for this experiment were non-bred crossbreed B. Indicus x B Taurus heifers that averaged 491 kg of liveweight, this weight corresponds to nearly the 75-90% of a mature crossbreed cow bodyweight (550-650 kg) (Abeygunawardena and Dematawewa, 2004). Hence, the diameter of the cervical lumen of heavier adult cows is higher than the cervical lumen of lighter heifers which explains the fact that the cervix in only 50% (n=5/10) of the heifers was effectively accessed using the Ghent® device. In contrast, Velazquez et al., (2010) were able to get through the cervix of all six non-lactating dairy cows using the Ghent® device, indicating that age differences affects the cervical passage. In addition, the relative operator inexperience and the limited technical skills and expertise with some of these methods, played an important role not only in the success of transcervical catheterisation but also in the quality of the sample retrieved. The prolonged manipulation of the reproductive tract during cervical passage of metallic devices potentially caused some degree of tissue disruption to the epithelial lining, increasing the risk of endometrial damage and contamination of the sample with blood cells. In this experiment, 2/3 samples obtained by UL and 3/5 by CB were contaminated with blood cells, indicating that some tissue disruption was caused during the procedure. Similarly, Kasimanickam et al. (2005a) and Cocchia et al. (2012) found greater contamination with red blood cells in samples collected by uterine lavage than by CB. This issue should be addressed based on the analysis to be undertaken. For instance, blood contamination recovered from ALF samples in women was a confounder factor when analysing pro-inflammatory cytokines since red cells interfered with the measurements of pro-inflammatory cytokines in uterine fluid aspirated from women (Boomsma et al., 2009b).

Chapter 2 Uterine sampling methods Overall in this study, the correct placement of a device in the uterine horns for sampling was successfully achieved in 68% (13/19) of the heifers. However, being able to get through the cervix does not guarantee that the sample will be collected successfully. In our study, we were unable to collect any uterine fluid by aspiration (ALF 0/2). Performing ALF in adult cows (n=6) at different stages of the oestrous cycle, Velazquez et al. (2010) recovered 5-400 µl of neat uterine fluid by vacuum suction. In this study, we tried to collect uterine fluid using a negative pressure aspiration with a 50 ml syringe from heifers at Day 7 of oestrous cycle without any success. This lack of fluid recovery could be explained partly by the stage of oestrous cycle as sampling was attempted during luteal phase when uterine luminal fluid is lowered compared to follicular phase (Kastelic et al., 1991). Likewise, in this study UL was performed successfully using Ghent® device, but the fluid recovered after uterine massage, varied between heifers from 5-15 ml after flushing the uterus with 20 ml of SSF. Hill and Gilbert (2008) performed UL by infusing 30 ml through an ET catheter, and then 10 to 20 sec later recovered approximately 20 ml of infused fluid. Kasimanickam et al. (2005a) infused a greater volume (60 ml of SSF) into the uterus body of adult cows and found that in 17% (12/70) of the cows, they failed to recover any fluid. Thus, this inconsistency in fluid recovery using either UL or ALF limits the diagnosis of uterine health through protein analysis of uterine secretions.

Techniques for the collection of surface cells for identification of inflammatory and endometrial cells by CB have been reported for large animals (Kasimanickam et al., 2004). However, most of the cytobrushes used to harvest bovine endometrial cells are designed for human use and required modifications before being used in cattle (Kasimanickam et al., 2005b, Barlund et al., 2008). In this study, we used a device specifically designed for cattle (Cytobrush, Minitube, Australia). However, we found that the catheter body was wider than the cervix, semi-rigid and did not allow the proper manipulation to get the device into the cervix of B.indicus x B.Taurus heifers. Although, we successfully recovered endometrial samples in all cases after we modified the device by attaching the brush to an insemination gun, we found that the excessive handling and manipulation of the instrument increased the risk of contamination (blood, faeces, dust) during the procedure, jeopardizing the quality and interpretation of the sample. Therefore, this method of sampling recovery was discarded for further studies.

In this study, EB was successfully achieved in 75% (n=3/4) of the heifers sampled using a human bronchoscope Storz® device. Although the low numbers of animals used in this study limits comparisons with studies of larger experimental size, our success rate was lower than the 86% obtained by Bonnett et al. (1991) (n=97/113; 86%) in postpartum Holstein cows using a Yomann Hauptner device. This discrepancy could be due to the larger cervical canal diameter of postpartum 59

Chapter 2 Uterine sampling methods cows, (between 3-6 cm) (Wehrend et al., 2003), that makes it considerably easier to pass a device through the cervix of the cow compared to passing a device through the smaller cervical lumen in heifers (Drennan and Macpherson, 1966). EB devices have a variety of different diameters and jaw sizes. Wider-diameter biopsy devices such as a Yomann Hauptner device (approximately 1 cm in diameter (Bonnett et al., 1991b, Mann and Lamming, 1994) have been applied in bovine reproduction, mostly for the assessment of uterine involution in postpartum cows and diagnosis of endometritis (LeBlanc et al., 2002). However, a rigid, thin and small-fragment biopsy device is ideal for animals of different ages and physiological status, in order to access the uterus without injury, minimising the tissue damage and increasing the likelihood of rapid healing after injury. In this study, the efficacy of a slimmer, rigid, small-diameter (0.35 cm) stainless steel Storz® device was investigated. The small diameter Storz® device offered an increased probability of traversing the cervix in heifers in order to recover a sample for microscopic and transcriptome assessment of uterine function. One of the advantages of using circular cup forceps (Storz®) as we used in this study, is the retrieval of a small endometrial sample and reduction in endometrial damage. In this experiment, the endometrial tissue recovered was 0.2 x 0.3 cm, weighing 30-40 mg, which was smaller than those obtained in previous studies (0.4 x 0.8 cm size and 100-500 mg) using devices with larger jaw size (i.e.Yomann Hauptner, Eppendorfer) (Mann and Lamming, 1994, Pugliesi et al., 2014). Therefore, the high consistency of the method obtained in the field combined with the physical features of the device (diameter, strength, jaws size) allowed us to select EB as the superior method to collect endometrial samples in beef heifers using the Storz® device.

Once the EB method was selected, we assessed whether the device provided a sample of sufficient quality and quantity to perform transcriptomic analysis. In this study, test biopsy samples obtained from post-mortem uteri with the Stortz® device provided an appropriate amount of total RNA (0.4- 3µg) of high quality (RIN > 7), for transcriptome assessment. Using endometrial sampling with the CB, only between 1-12 µg of total RNA can be obtained (Gabler et al., 2009, Ghasemi et al., 2012). However, CB samples of the endometrial lining harvest only luminal and infiltrating inflammatory cells, whereas biopsy comprises an entire endometrial sample. RNA yield differences between studies may be related to the extraction method and the disruption of the initial material (tissue or cells). The importance of obtaining high-quality RNA samples will be illustrated in a later chapter where the expression of key genes is reported.

In conclusion, it was demonstrated that the collection of endometrial samples in beef heifers was possible. The selection of the sampling method, the features of the device (diameter, rigidity) and potential risks of sample contamination should be considered before sampling. The sampling

Chapter 2 Uterine sampling methods procedures should be performed by an experienced operator which guarantees that the procedure will be performed rapidly, minimising excessive manipulation of the reproductive tract and the subsequent uterine trauma. In addition, contamination of the device and specimens should be prevented, thus excessive handling of the devices, inadequate hygiene before and after the procedure, inadequate storage conditions for the samples retrieved may jeopardise the yield and quality of RNA for further molecular assessments. The selection of thinner and more rigid devices such as Storz® device would be the most appropriate approach to traverse the cervix to collect endometrial samples in cycling heifers to access the uterus in non-adult cows.

Chapter 3 Impact of endometrial biopsy on endometrial health

CHAPTER 3. INVESTIGATION OF THE IMPACT OF ENDOMETRIAL BIOPSY ON ENDOMETRIAL

HEALTH

Chapter 3 Impact of endometrial biopsy on endometrial health

Introduction

In the pilot study, we found that endometrial biopsy (EB) was the most consistent method to collect samples for transcriptome analysis in beef heifers. This procedure allows the morphological diagnosis of endometrial lesions by histopathology (Chapwanya et al., 2009, Meira et al., 2012), the expression of inflammatory genes during the oestrous cycle or postpartum (Fischer et al., 2010, Galvao et al., 2011) and the detection of biomarkers and pathways responsible for endometrial receptivity and pregnancy success by means of molecular techniques (Salilew-Wondim et al., 2010, Salilew-Wondim et al., 2012, Binelli et al., 2015). To ensure successful pregnancy outcomes from AI and ET technologies, assuring high heifer fertility is optimal is crucial for a successful breeding season. Thus, the assessment of uterine function and endometrial receptivity before breeding or ET by EB would improve the reproductive outcomes. However, EB is rarely performed in heifers for microscopic or transcriptomic assessment due to practical reasons (i.e. laborious technique, requires specialised laboratory for processing) and also because it is considered that heifers that have not been exposed to a bull have a sterile uterus, free of pathogens (Fischer- Tenhagen et al., 2012) and performing EB runs the risk of introducing contamination.

To predict the endometrial receptivity, the collection of biopsies should be performed before embryo arrival/transfer (between Days 5 to 7). Hence, we selected two models for investigating the optimum time collecting biopsies: i) biopsying at Day 7 of the cycle before ET followed by induction of oestrus (Day 0) with PGF2α administration (48-72 h after injection) and ii) biopsying at Day 4 of the ongoing cycle in the horn ipsilateral to the CL. Subsequently, in both models, the effect of the biopsy was assessed at Day 7 post-oestrus as indicative of the uterine environment that the embryo will be exposed after arrival. This means that we assessed a long-term response (10 days after injury) biopsying one cycle earlier and a short-term response (3 days after biopsy) biopsying at Day 4 post-oestrus. Screening the inflammatory and healing response at day 7, around the time of putative embryo arrival, would be indicative of the uterine environment that the embryo will be exposed to during its development. Day 4 post-oestrus has been selected as the day of biopsy, in order to avoid the potential trigger of the endogenous release of PGF2α caused by endometrial injuries, thus resulting in lysis of the CL (Snider et al., 2011). At Day 4 of oestrous cycle, the CL is still under development and is not fully matured and would be

Chapter 3 Impact of endometrial biopsy on endometrial health ineffective in mounting a luteolytic response (Howard and Britt, 1990, Murugavel et al., 2003).

In this chapter, we aimed to assess the morphological, molecular and vascular responses on endometrial function in cycling maiden heifers after EB. Therefore, in study 1 we assessed the long and acute morphological changes after biopsy. In study 2, we assessed expression of pro-inflammatory and wound healing genes using customised PCR arrays after biopsy. In study 3, after characterising the morphological and molecular response at Day 7 post-oestrus, we assessed the accumulation of uterine fluid after biopsying during the luteal phase of the oestrous cycle by non-invasive transrectal ultrasound.

Chapter 3 Impact of endometrial biopsy on endometrial health

3.1 Histological response to endometrial biopsy

Abstract

Two experiments were conducted to evaluate the morphological changes induced by endometrial biopsy in cycling tropically adapted beef cattle. In both experiments, a single biopsy was performed on the horn ipsilateral to the CL using a circular cup biopsy forceps. The uterus was collected after slaughter and sections were taken adjacent to the biopsy site.

In Experiment 1: presynchronised heifers (n=10; 24-36 months old, Brahman and Droughmaster) were biopsied on Day 7 post-oestrus (B7), then re-synchronised with cloprostenol (Estromil®, 250 μg IM) to induce oestrus (O), 2 to 3 days after injection. At Day 7 post-oestrus, the reproductive tracts (T7) were collected 10 days after biopsy (B7-O-T7). In Experiment 2: presynchronised heifers (n=7) were biopsied on Day 4 post-oestrus (B4) and reproductive tracts were collected three days later (B4-T7). A quantitative scoring system was used to define histological appearance of the endometrium (from 1 to 25; above 16 indicates endometritis). Heifers (n=5) were synchronised and uterine manipulation was performed as a negative control.

Macroscopic examination showed the uterus collected from B7-O-T7 heifers exhibited a focal, well-demarcated, dark red discoloration of the endometrium (4/10) with no evidence of the biopsy site found in the remaining heifers (6/10). The biopsied uteri from B4-T7 heifers revealed no macroscopic evidence of the biopsied site (4/7), serosa congestion (n=2/7) and one case where a severe endometrial haemorrhage was evident (1/7). Histologically, the blind assessment of biopsied and non-biopsied uterine slides did not reach the threshold score of 16 meaning there was no evidence of endometritis. No significant difference was found between biopsied horns (BH) and non-biopsied horns (NBH) either in B4-T7 or in B7-O-T7, although the overall ranges mean score in B4-T7 heifers was higher compared to B7-O-T7 in both biopsied (4.3 to 5.6 and 2.1 to 3.1, respectively) and in non-biopsied horns(2.1 to 3.2 and 2.3 to 2.5, respectively, P > 0.05).

Although in most cases there was minimal evidence of endometrial damage within 3 to 10 days after endometrial biopsy, the histological assessment of both the biopsied and non- biopsied horns showed no evidence of damage suggestive of endometritis as a result of using the Storz® device.

Chapter 3 Impact of endometrial biopsy on endometrial health

3.1.1 Introduction

After fertilisation, the bovine zygote is transported through the oviduct for 4 to 6 Days before entering the uterus at the morula or blastocyst stage (Hackett et al., 1993). Thus, Day 7 post- oestrus is a critical time-point in the uterine environment encountered by the embryo during early stages of development.Within the first three weeks after fertilisation, the bovine embryo is particularly vulnerable to damage and the uterus must provide a suitable environment in order to facilitate the cross-talk between the conceptus and the endometrium for normal embryo development, growth, and attachment. This preparatory process is often collectively referred to as endometrial receptivity (Bazer et al., 2012, Ulbrich et al., 2013). Inappropriate uterine function with suboptimal endometrial receptivity is considered one of the major causes of embryo mortality (Inskeep and Dailey, 2005).

The bovine endometrium consists of a mucosal surface with a pseudostratified columnar epithelium and submucosal layer (lamina propria) of loose connective tissue that contains uterine glands and blood vessels (Ohtani et al., 1993). Endometritis is a superficial inflammation of the endometrium, extending no deeper than the lamina propria (Bondurant, 1999). Although the usefulness of bovine uterine histology is limited by being laborious and/or expensive to apply in routine reproductive assessment, greater precision in the diagnosis of inflammatory process using biopsy material before AI or ET has emerged as an option to increase reproductive efficiency in the cattle industry. Histologically, endometritis is characterised by increased infiltration of inflammatory cells associated with glandular and vascular changes (Chapwanya et al., 2010). Although the diagnosis of endometritis by histopathology is not difficult, defined criteria and uniformity in the mode of assessment were lacking in more subjective assessment methods (Singh et al., 1983, Bonnett et al., 1991b, McDougall et al., 2007, Chapwanya et al., 2009). Recently, Meira et al. (2012) developed a more detailed semi-quantitative score system, which demonstrated detection of endometritis with 92% specificity (see Figure 1.7).

Questions remain as to the impact of endometrial biopsy on later fertility and this study aims to assess endometrial morphological changes in heifers after biopsy using the small jaw Storz® device. A biopsy was taken at Day 7 post-oestrus of the preceding cycle or at Day 4 of the ongoing cycle then assessment of microscopic changes on Day 7 after oestrus (either 10 or 3 days after biopsy). Day 7 after oestrus is around the time of embryo arrival, which is indicative of the uterine environment to which embryos are first exposed.

Chapter 3 Impact of endometrial biopsy on endometrial health

3.1.2 Materials and Methods

3.1.2.1 Animals and management

All the animal procedures were approved by the Committee Animal Welfare Unit, UQ Research & Innovation, under permit AE02430. The experiments were carried out at the academic beef unit of the University of Queensland (Brisbane, QLD) at Pinjarra Hills.

The procedures for animal management, reproductive assessment for heifer’s selection, synchronisation protocols and collection of endometrial biopsy are described previously in chapter 2 (Material and methods). Twenty-seven tropically-adapted beef heifers (Brahman crossbred heifers with ¾ to 7/8 Brahman content; n=16; and Droughmaster 50% B. indicus x 50% B. taurus, n = 11), aged between 24 to 36 months old, with an average live weight of 521 kg (range 384-590 Kg), were enrolled.

3.1.2.2 Synchronisation of oestrus:

Heifers (n=22) were selected to synchronise oestrus according to an optimised protocol (Butler et al., 2011) described in Chapter 2 (2.1.2.2 Synchronisation of oestrus). Briefly, at Day -9 oestradiol benzoate (1mg, Cidirol®, Genetics Aust, Australia) was administered by intramuscular (IM) injection and an intravaginal progesterone device (0.786g Cue-Mate®, Bioniche Animal Health lab, Australia) was inserted. At Day -2 the Cue-Mate was removed and Cloprostenol (Estromil®, 250 µg Ilium, Australia) was also administered IM. At Day -1, oestradiol benzoate (1mg) was administered IM. Day 0 (oestrus) was determined by twice daily observation for signs of heat. Only animals that showed signs of oestrus after synchronisation were enrolled.

3.1.2.3 Experimental design

a) Experiment 1: Assessment of morphological changes in slaughtered uterine tissue after endometrial biopsy collected one cycle before (B7-O-T7)

Synchronised heifers (Group 1; n=10) were selected for biopsy at Day 7 (B7) of cycle 1 (Day 0= oestrus). After biopsy, heifers were injected with Cloprostenol (Estromil®, 250 μg IM Troy Laboratories, NSW, Australia) to induce oestrus (O) which occurred 48-72 h after Cloprostenol administration. At Day 7 post-oestrus (T7) of cycle 2, heifers were slaughtered and the reproductive tracts recovered. This equates to 10 days after biopsy (B7-O-T7). In the

Chapter 3 Impact of endometrial biopsy on endometrial health control group (n=2), heifers underwent synchronisation protocols and uterine transrectal manipulation, but were not biopsied (Figure 3.1).

b) Experiment 2: Assessment of morphological changes in slaughtered uterine tissue after endometrial biopsy collected during ongoing cycle (B4-T7)

Synchronised heifers (Group 1; n=7) were biopsied at Day 4 post-oestrus (B4). At Day 7 post-oestrus, heifers were slaughtered and reproductive tracts (T7) were collected, 3 days after biopsy (B4-T7). Control heifers (n=3) were synchronised but not biopsied (Figure 3.1).

Figure 3.1 Experimental design for biopsy procedures performed during the prior cycle (B7- O-T7) or during ongoing cycle (B4-T7). B7: Biopsy at Day 7 post-oestrus. O: Oestrus. T7: Reproductive tracts at Day 7 post-oestrus. B4: Biopsy at Day 4 post-oestrus.

3.1.2.4 Macroscopic evaluation

The reproductive tracts collected at Day 7 post-oestrus (T7) were transported on ice and submitted for a macroscopic evaluation within 4 h after slaughter. The uterus was incised and examined grossly and an attempt to identify the biopsy spot was made for each uterus. Three samples of uterine tissue (1-2 cm) were collected from intercaruncular areas: two from the biopsied horn and the remainder from the opposite, non-biopsied horn. When no biopsy spot

Chapter 3 Impact of endometrial biopsy on endometrial health was identified, the sample was taken from the middle of the recorded biopsied horn. Each sample was further divided into two pieces and either immersed in Bouin’s solution (Sigma- Aldrich, St Louis, MO, USA) or RNA later (Qiagen, VIC, Australia) for histology (Study 1) and qRT- PCR (these material will be used further in Study 2), respectively.

3.1.2.5 Histopathology

Uterine samples were initially fixed in 100% Bouin’s solution and 24 h after samples were transferred to 70% (v/v) ethanol, then embedded in paraffin and 5 µm thick sections were stained with haematoxylin and eosin (HE). The histopathological assessment was carried out independently by two specialist veterinary pathologists, blind to sample IDs and the design of the biopsy experiments. Slides were scored according to the scheme of Meira et al. (2012) developed for use in cows. Briefly, endometrial sections were scored for a total of 26 points dependent on the summarised characteristics of the surface epithelium (from 1 to 11), the lamina propria (from 0 to 8), endometrial glands (from 0 to 4) and vascular inflammatory status (from 0 to 3) (Figure 1.7). The epithelium and lamina propria were assessed based on the integrity, type and intensity of inflammatory cells and presence of lymphocytic aggregates. Changes in endometrial glands (atrophy, dilation, and fibrosis) and vessels (degeneration, haemorrhage and hemosiderin) were also considered in the evaluation. Scores above the threshold value of 15/26 were diagnosed as exhibiting endometritis.

3.1.2.6 Statistical analysis

Histopathology data scores were tested for normality using the Shapiro test, for homogeneity of variance by the Breush-Pagan test and independency by Durbin-Watson test. Normally distributed data were analysed using Wilcoxon Sign Rank Test. A reliability analysis using the Kappa statistic (value between-1 and 1; k) was performed to determine consistency among pathologists in the diagnosis of endometritis (score >16). The confidence interval was 95% and P- value <0.05 for significance. Statistical analyses were carried out in R 3.2.1 (R Core Team, 2015).

Chapter 3 Impact of endometrial biopsy on endometrial health

3.1.3 Results

3.1.3.1 Macroscopic evaluation

Gross examination of 4/10 uteri collected 10 days after biopsy (B7-O-T7 heifers) exhibited a focal, well-demarcated, dark red discoloration of the endometrium suggestive of chronic haemorrhage and a recent biopsy (Figure 3.2a). In the uterine horns of the remaining 6 heifers, the biopsy area could not be identified. The macroscopic evaluation of uteri biopsied 3 days previously (B4-T7 heifers) revealed either no macroscopic evidence of the biopsied spot (4/7), minor evidence of a biopsy spot in endometrium and serosa congestion (2/7) (Figure 3.2 b,c) or severe endometrial haemorrhage (1/7). Uteri from control did not show any lesion (Figure 3.2d).

3.1.3.2 Histopathology

Overall, both pathologists agreed on the assessment of the biopsied and non-biopsied uterine slides with neither of them reaching the threshold score level of 16 for the diagnosis of endometritis. Thus, analysing the scores given (below 16) according to the method of Meira et al. (2012), the observers showed fair agreement in the score given to the slides (κ=0.35) in the histopathological assessment in both experiments (Table 3.1). This reduced consistency between observers could have been affected by the number of variables (>10) available to score every slide. In B7-O-T7 heifers, there was no significant difference (P> 0.05) between the scores given by the pathologists either in the biopsied or non-biopsied horns (score range from 2.1 to 3.1 in BH and 2.3 to 2.5 in NBH). The controls did not show significant difference compared to NBH from B7-O-T7 and B4-T7.

The main lesions found in B7-O-T7 were mild epithelial damage accompanied by infiltration of mononuclear cells into both epithelium and lamina propria, mild fibrosis in endometrial glands and vascular changes associated with the presence of haemorrhage and hemosiderin in Table 3.2. These lesions were focal and did not extend into adjacent endometrium (Figure 3.3b).

Chapter 3 Impact of endometrial biopsy on endometrial health

Table 3.1 Mean (±SD) scores given by pathologists in biopsied (BH) and non-biopsied horns (NBH) in B4-T7 and B7-O-T7 and control. Mean comparisons are given between rows in BH and NBH

p >0.05 no significant

Although the mean scores given in B4-T7 heifers was higher compared to B7-O-T7 there was no significant difference (P > 0.05) between the scores given by the two pathologists in the biopsied horns nor in the non-biopsied horn. The scores in B4-T7 range from 4.36 to 5.57 in BH and 2.14 to 3.14 in NBH whilst in B7-O-T7 the scores range from 2.1 to 3.1 in BH and 2.35 to 2.55 in NBH. The morphological changes described by assessors are summarised in Table 3.2. These changes include transmigration of mononuclear cells into the epithelial layer, mild epithelium damage (Figure 3.3b) vascular changes associated with haemorrhage and hemosiderin (Figure 3.3 c,d) (Figure 3.3 c,d) dilation and mild fibrosis of endometrial glands (Figure 3.3e), infiltration of mononuclear cells in uterine glands (Figure 3.3f) and vascular congestion (Figure 3.3g,h).

Chapter 3 Impact of endometrial biopsy on endometrial health

Figure 3.2 Gross abnormalities in biopsied uteri collected from (a) B7-O-T7 and (b-d) B4-T7 heifers. (a,b) The biopsied spot is grossly evident as a focal endometrial red discoloration suggestive of an underlying inflammatory process and/or circulatory disorder (arrows);(c) Serosa congestion (d): In one case, a severe locally extensive dark red haemorrhage was observed (arrow).

Chapter 3 Impact of endometrial biopsy on endometrial health

Figure 3.3 Histopathological assessment in B7-O-T7 heifers: (a) Normal tissue, 4X (b) Loss of epithelium and slight oedema and haemorrhage in submucosa layer, 4X. B4-T7 heifers: (c.d) Loss of epithelium, focal haemorrhage in submucosa layer, presence of macrophages containing hemosiderin 4X&10X (e) Slight glandular dilation and moderate periglandular fibrosis 10X (f) Mononuclear infiltration in glandular epithelium 40X (g-h) Vascular congestion 10X. H&E

Chapter 3 Impact of endometrial biopsy on endometrial health

Table 3.2 Morphological changes described by pathologist in B7-O-T7 and B4-T7 heifers

B4-T7 (n=7) B7-O-T7 (n=10) Score Observer 1 Observer 2 Observer 1 Observer 2 Epithelium 42% (1-11) n % n % n % n % Height Cuboidal 2 1 14% 0 0 0 0% 1 10% Height Flattened 3 2 29% 0 0% 0 0% 0 0% Epithelial damage Mild 1 5 71% 1 14% 3 30% 0 0% Epithelial damage Ulcer 3 0 0% 1 14% 0 0% 0 0% Inflammatory cell type Mononuclear 1 4 57% 0 0% 1 10% 6 60% Infiltrate intensity Mild 1 0 0% 1 14% 2 20% 6 60% Lamina Propria 31% (0-8) Inflammatory cell type Mononuclear 1 1 14% 5 71% 1 10% 8 80% Infiltrate intensity Mild 1 2 29% 2 29% 0 0% 1 10% Lymphocytic 1 0 0% 0 0% 2 20% 3 30% aggregation Mild Endometrial gland 15% (0-4) Atrophy or dilation Present 1 5 71% 2 29% 4 40% 2 20% Fibrosis Mild 1 3 43% 4 57% 5 50% 3 30% Fibrosis Moderate 2 3 43% 0 0% 1 10% 0 0% Fibrosis Severe 3 1 14% 0 0% 0 0% 0 0% Vascular 12% (0-3) Vessel degeneration Present 1 0 0% 1 14% 0 0% 0 0% Haemorrhage Present 1 3 43% 0 0% 2 20% 0 0% Hemosiderin 1 2 29% 0 0% 0 0% 4 40% macrophages Present

Chapter 3 Impact of endometrial biopsy on endometrial health

3.1.4 Discussion

This study assessed the histological changes occurring in the uterus at Day 7 of the oestrous cycle after an endometrial biopsy performed during the prior (B7-O-T7) or the same cycle (B4-T7).

Endometrial biopsy is a technique in which a sample of endometrial tissue is removed for microscopic and/or molecular assessment of uterine function. Endometrial biopsy can be performed using devices with different jaw size and diameter. One of the advantages of the circular cup forceps (Storz®) used in this study was both the small diameter of the device and the small sample size retrieved which minimised disruption to the endometrium. The size of endometrial tissue recovered was 0.2 x 0.3 cm which was smaller compared to samples obtained using larger jaw devices (Yomann Hauptner, Eppendorfer: 0.4 x 0.8 cm size) (Mann and Lamming, 1994, Pugliesi et al., 2014). The small amount of tissue recovered with the Storz® device might explain the minimal number of gross abnormalities detected in the uterine horns after biopsy or alternatively, the small sample size causes minimal disruption to the uterus.

The assessment of the reproductive tracts post-mortem revealed that the biopsy spot was evident in 42% (3/7) of B4-T7 heifers and 40% (4/10) in B7-O-T7 heifers. Mann and Lamming (1994) reported the persistence of small scars in the endometrium of cows three weeks after biopsy using the Hauptner device. In other species, uterine wound healing has been assessed after surgical injury. In our study, only one heifer showed a focal and severe endometrial haemorrhage at Day 7 post-oestrus, with the biopsy taken 3 days earlier. This lesion may have been caused by the procedure causing injury to large endometrial vessels or alternatively, by the physiological rupture of blood vessels occurring after ovulation during metoestrus, which is accompanied by haemorrhagic lesions beneath the epithelial surface (Eurell and Frappier, 2013).

In addition, the small diameter of the Storz® device (0.3 cm) facilitated access to the uterus of beef heifers. Thus, a rigid, thin and small-fragment biopsy device is preferred for biopsying animals of different ages and physiological status. The Storz® device allowed access to the uterus of maiden heifers, minimising tissue damage and increasing the likelihood of rapid healing after injury.

Chapter 3 Impact of endometrial biopsy on endometrial health

In our experiment, the histopathological examination of the biopsied horn site, its adjacent segment and the opposite, non-biopsied horn, did not reveal cellular infiltration, glandular or vascular changes as would be present in endometritis. None of the samples obtained a score above 16, which is the threshold value for the histopathological diagnosis of endometritis according to Meira’s method (Meira et al., 2012). However, previous studies (Singh et al., 1983, McDougall, 2005, Azawi et al., 2008) had shown that specific morphological changes such as lack of columnar epithelium in the mucosa, the presence of haemorrhage and hemosiderin in lamina propria, periglandular fibrosis, and oedema may be suggestive of endometritis. These histological alterations were found sporadically in both B4-T7 and B7-O- T7 biopsied uteri, but the overall score using this method was not diagnostic of clinical endometritis.

Considering that the endometrial tissue recovered during biopsy represents less than 1% of the uterine tissue (Snider et al., 2011), we aimed to elucidate whether one sample is representative of overall uterine function. The results in our study did not reveal significant differences in the histological assessment in samples taken at or near the biopsy site nor in either the biopsied horn or non-biopsied horn in both experiments. In mares, Blanchard et al. (1987) found low inter-sample variability comparing biopsies taken from both horns in normal cycling mares, indicating that one sample would be enough to assess endometrial health. Similarly, in physiological conditions such as uterine involution Bonnett et al. (1991b) did not find histological differences between gravid and non-gravid horn in normal postpartum cows. In contrast, the representativeness of biopsy samples apparently varies in pathological conditions. Fiala et al. (2010) evaluated the agreement between two-blind observers in biopsies taken from three different uterine sites for the diagnosis of endometriosis in mares. They found that in only 30% of the cases were the three biopsies diagnosed with the same grade of endometriosis by the observers, indicating that the distribution of endometriosis was variable and one biopsy site did not provide enough precision for an accurate diagnosis of degenerative fibrotic changes in a mare’s uterus. Since our results showed that histologically none of the uteri of biopsied heifers developed infectious endometritis, we can presume that assessing a single biopsy by histology is representative of the entire uterus and indicative of uterine function in healthy animals. However, we cannot conclude with our results whether a single biopsy would be representative of the entire uterus in animals having pathological conditions.

Chapter 3 Impact of endometrial biopsy on endometrial health

In conclusion, the histological assessment did not show evidence in either biopsied or non- biopsied horns of PMN cells either 3 or 10 days after injury, indicating the procedure did not cause infectious endometritis. The inflammatory changes reported in glands and stroma which are suggestive of endometrial damage, were localised to the site of injury, not disseminated to the adjacent site along the biopsied horn or the opposite horn. It was also evident that there is no morphological difference in the assessment of biopsies collected from the biopsied or non-biopsied horn indicating that one single sample is representative of uterine health. Despite morphologically the uterine tissue at Day 7 post-oestrus returns to normal function by the time of embryo arrival after biopsy, further studies are necessary to characterise the acute and long-term molecular response after mechanical injury.

Chapter 3 Impact of endometrial biopsy on endometrial health

3.2 Inflammatory gene response to endometrial biopsy

Abstract

The effect of the mechanical injury caused by EB on endometrial inflammatory gene expression around the time of embryo arrival has not been assessed in bovine heifers. This study aimed to assess the effects of a single EB performed before the time of embryo arrival at Day 7 post-oestrus of the preceding cycle and compares this to an EB at Day 4 of the ongoing cycle by the evaluation of changes in inflammatory gene expression in the uterus at Day 7 of in beef heifers. In Experiment 1, pre-synchronised heifers (Brahman and Droughmaster) were biopsied at Day 7 of the preceding cycle and injected with Cloprostenol (Estromil®, 250 μg IM) to induce oestrus (O) which occurred 48-72 h later. At Day 7 post- oestrus (T7) of the ongoing cycle, 22 heifers were slaughtered. This included 10 heifers that underwent single EB at Day 7 post-oestrus of the preceding cycle (B7-O-T7), 7 heifers biopsied at Day 4 post-oestrus of the ongoing cycle (B4-T7) and the 5 non-biopsied control heifers. The reproductive tracts were recovered for gene expression of pro-inflammatory IL1β and wound healing TGFβ genes. In Experiment 2, a subset of RNA samples obtained from biopsied-slaughtered heifers from Experiment 1 was selected to assess inflammatory gene expression using RT2 customised gene array (Qiagen, MD, USA). To assess the changes in gene expression in uterine horns 3 days after EB, two groups of synchronised Droughtmaster heifers (n=12, 26 months, 387 ± 22 Kg LW) were randomly selected for EB. In group 1 (n=6; B4-B7), heifers were initially biopsied at Day 4 post-oestrus (B4) from the horn ipsilateral to the CL post-oestrus and 3 days later, a second biopsy was retrieved from both the previously biopsied horn (B7b) and the non-biopsied horn (B7nb). Heifers in the control group (n=6; B7) were each horn biopsied at Day 7 post-oestrus. Biopsies were stored in RNA later (Qiagen, VIC, Australia) to assess in a RT2 Profiler PCR arrays pro-inflammatory cytokines (IL1R, TNF, IL6, SPP1), markers for tissue repair (VEGF, TGFβ, CSF2) and vasoactive enzymes (PTGS2).

In Experiment 1, neither IL1β expression nor TGFβ were different between biopsied and non-biopsied horns at either 3 (B4-T7) or 10 days (B7-O-T7) after biopsy (P>0.05). However, IL1β expression was significantly higher in the biopsied horn at 3 compared to 10 days after biopsy (P=0.031). In Experiment 2, the change of initial gene expression within the same animal over time of pro-inflammatory and wound healing genes did not differ

Chapter 3 Impact of endometrial biopsy on endometrial health statistically (P>0.05) in the biopsied and non-biopsied horns assessed at 3 and 10 days after EB. Overall, the pattern of expression 3 days after EB showed up-regulation of pro- inflammatory cytokines (IL1α, IL1R, IL6, TNFα and PTGS2) and down-regulation of wound healing genes (TGFβ and VEGF). The pattern of gene expression 10 days after EB, VEGF was up-regulated in both biopsied and non-biopsied horns whilst CSF2, TNF, SPP1, TGFB1 IL6 and PTGS2 were down-regulated in both horns.

These results suggest that EB did not induce significantly changes in the expression of targeted inflammatory genes by the time of embryo arrival performing a single biopsy either during ongoing metoestrus or during the cycle prior to ET and by the time of transfer the endometrium will be healed, thus not affecting embryo development and survival.

Chapter 3 Impact of endometrial biopsy on endometrial health

3.2.1 Introduction

The immune system induces a rapid onset inflammatory reaction to restore functionality of damaged tissue (Salamonsen, 2003, Weiss et al., 2009). This inflammatory reaction, which may last 1 to 3 days, is characterised by acute changes in vascular blood flow accompanied by recruitment and migration of inflammatory cells with the subsequent release of cytokines, chemokines and growth factors needed for wound healing (Li et al., 2007). Endometrial biopsy (EB) causes localised mechanical injury and tissue damage that triggers a sterile inflammatory response (Granot et al., 2012). During the tissue repair phase, which may last days to weeks, growth factors and cytokines are secreted by immune cells, they help restore tissue structure and functionality of the damaged area with the re-establishment of blood flow into the region (Diegelmann and Evans, 2004).

Mechanical injury stimulates the release of intra and extracellular matrix molecules, known as damaged-associated molecular patterns (DAMPs), from dying cells. Immune cells identify these non-pathogenic DAMPs in the extracellular matrix through a series of pattern recognition receptors (PRRs) especially toll-like receptors (TLR 1-10), that trigger inflammatory signalling pathways (Kannaki et al., 2011, Thompson et al., 2011, Shen et al., 2013). In the bovine uterus, the damage to the endometrial epithelium activates the first part of the immune response against injured cells. Endometrial cells express TLR-2, -4 and -6 which bind to DAMPs activating injury-specific signalling pathways (Chen and Nunez, 2010, Kannaki et al., 2011, Shen et al., 2013). The cascade of inflammatory events stimulates the production of cytokines, chemokines and vasoactive mediators by the recruitment of local and systemic polymorphonuclear (PMN) neutrophils to the injured tissue to remove cellular debris (Weiss et al., 2009, Kannaki et al., 2011). During initial stages of sterile inflammation, neutrophils release pro-inflammatory cytokines such as interleukin IL1, IL6, tumour necrosis factor alpha (TNFα) and osteopontin (SPP1) (Dhaliwal et al., 2001, Herath et al., 2006, Chen and Nunez, 2010, Healy et al., 2014) and chemokine ligands (CXCL 1,-2 and-3) to stimulate further neutrophil recruitment and migration to the site of injury in the uterus. There they promote neutrophil maturation and differentiation of monocytes into mature macrophages and natural killer cells (Gabler et al., 2009, Herath et al., 2009, Heppelmann et al., 2015).

For effective inflammatory resolution, mononuclear macrophages develop anti-inflammatory properties changing from a phagocytic mechanism during acute inflammatory phase to a repair phase (Diegelmann and Evans, 2004). Thus, macrophages stimulate the release of

Chapter 3 Impact of endometrial biopsy on endometrial health multifunctional cytokines (i.e transforming growth factor beta TGFβ) that phagocytose necrotic or apoptotic neutrophils hence, avoiding the recruitment of more neutrophils to the inflammatory site (Godkin and Dore, 1998, Sugawara et al., 2010). Additional cytokines and growth factors released during the resolution phase include immunosuppressive and anti- inflammatory interleukin-10 (IL10) that inhibits the production of pro-inflammatory cytokines (Schneider et al., 2004); factors that stimulates the phagocytic activity of neutrophils and enhance tissue resolution (granulocyte-macrophage colony stimulating factor, GM-CSF2 (Becher et al., 2016); key vasoactive enzyme molecules which mediate the synthesis of prostaglandins (prostaglandin–endoperoxide synthase 2, PTGS2) (Arosh et al., 2002).) and growth factors that enhances angiogenesis, revascularisation and wound repair (vascular endothelial growth factor, VEGF) (Tasaki et al., 2010) (Tizard, 2012).

During early pregnancy, growth factors and cytokines are expressed at the maternal- embryo interface by both endometrial and embryonic cells. A successful attachment requires the expression and balance of cytokines and growth factors involved in the dialogue between the conceptus and the uterus from fertilisation until attachment. Beyond that, they are essential during the foetal stage (Martal et al., 1997). Whilst the endometrial over-expression or imbalances of some pro-inflammatory cytokines such as IL1α/β, TNFα, IL6 or inflammatory mediators such as PTGS2 have shown deleterious effect on embryo development (Hill and Gilbert, 2008, Koblischke et al., 2008, Gilbert, 2011, Prins et al., 2012), the expression of other cytokines (SPP1) and growth factors (VEGF, CSF2, TGFβ) have a positive effect on embryo development and embryo survival (Martal et al., 1997, Robertson, 2007, Gnainsky et al., 2010). Cytokines and growth factors interact with cell adhesion molecules (i.e. integrins) expressed in trophoblast, participate in tissue remodelling and angiogenesis events during attachment or activate intracellular signalling transduction pathways (i.e. Janus kinase/signal transducers and activators of transcription, JAK/STAT) improving attachment (Guzeloglu- Kayisli et al., 2009).

The effect of the mechanical injury caused by EB on endometrial inflammatory gene expression around the time of embryo arrival has not been assessed in bovine heifers. Therefore this study assesses the effects of a single EB performed before the time of embryo arrival at Day 7 post-oestrus of the preceding cycle and compares this to an EB at Day 4 of the ongoing cycle by the evaluation of changes in inflammatory gene expression in the uterus at Day 7 of in beef heifers. Therefore, this study will characterise the long (10 d) and short- term (3 d) changes in endometrial gene expression by assessing pro-inflammatory cytokines

Chapter 3 Impact of endometrial biopsy on endometrial health

(IL1R, TNF, IL6, SPP1), markers for tissue repair (VEGF, TGFβ, CSF2) and vasoactive enzymes (PTGS2) after mechanical injury induced by biopsy using the Storz® device.

Chapter 3 Impact of endometrial biopsy on endometrial health

3.2.2 Materials and Methods

3.2.2.1 Animals and management

The procedures for animal management, reproductive assessment for heifer’s selection, synchronisation protocols and biopsy procedure are described previously in Chapter 2 (Material and methods).

3.2.2.2 Experimental design

a) Experiment 1: Endometrial gene expression of pro-inflammatory (IL1β) and wound healing (TGFβ) cytokines after EB in biopsied and non-biopsied (control) heifers by qRT-PCR

The description of the heifers (age, weight, and breed), oestrus synchronisation protocol, EB procedure and collection of uterine tracts has been described in Study 1-(Animals and management). This experiment aimed to evaluate whether EB induces local changes in the inflammatory gene expression (IL1β and TGFβ) after long (10 d) or short term (3d) assessment in both biopsied and non-biopsied horns.

Heifers biopsied in the preceding cycle were injected with Cloprostenol (Estromil®, 250 μg IM) on the day of biopsy to induce oestrus (O) which occurred 48-72 h later. At Day 7 post- oestrus (T7) of the ongoing cycle, 22 heifers were slaughtered and the reproductive tracts recovered. This included 10 heifers that underwent single EB at Day 7 post-oestrus of the preceding cycle (B7-O-T7), 7 heifers biopsied at Day 4 post-oestrus of the ongoing cycle (B4-T7 and the 5 non-biopsied control heifers. In B7-O-T7 heifers involved the induction of oestrus after EB, with the subsequent rising levels of E2. In B4-T7 heifers, the EB is performed when the reproductive tract is under influence of rising levels of P4 secreted from the CL.

The reproductive tracts were collected at day 7 post-oestrus (T7), and submitted for assessment within 4 h after slaughter. Three uterine specimens of intercaruncular tissue (1-2 cm) were collected per animal. Two specimens were collected from the biopsied horn (the

Chapter 3 Impact of endometrial biopsy on endometrial health biopsy spot and adjacent segment) and the remainder was collected from the opposite, non- biopsied horn. Specimens were divided into two pieces and were stored in either RNA later (Qiagen, VIC, Australia) for qRT- PCR or immersed in Bouin’s solution (Sigma-Aldrich, St Louis, MO, USA) for histology (described in study 1 Materials and Methods).

b) Experiment 2 Analysis of inflammatory gene expression after EB using a customized PCR-array

A subset of RNA samples (n=7) obtained from biopsied-slaughtered heifers from Experiment 1 (B7-O-T7) was selected to assess inflammatory gene expression using RT2 customised gene array (Qiagen, MD, USA). This experiment aimed to evaluate whether EB induces changes in the inflammatory gene expression at days 10 post-biopsy within the same individual.

Samples from B4-T7 heifers in Study 1 were discarded due to low RNA quality RIN <7, likely caused by inadequate storage conditions. To replace those samples, two groups of synchronised Droughtmaster heifers (n=12, 26 months, 387 ± 22 Kg LW) were randomly selected for EB. In group 1 (n=6; B4-B7), heifers were initially biopsied at Day 4 post- oestrus (B4) from the horn ipsilateral to the CL. post-oestrus. Three days later, a second biopsy was retrieved from both the previously biopsied horn (B7b) and the previously non biopsied horn (B7nb). In this experiment, heifers in control group (n=6, B7) were biopsied on each horn at Day 7 post oestrus. In this experiment we used as control group (n=6; B7), heifers where each horn was biopsied at Day 7 post-oestrus. This experiment aimed to assess the acute changes in inflammatory gene expression in uterine horns 3 days after EB.

3.2.2.3 Gene expression assessment from endometrial tissue

3.2.2.3.1 RNA extraction and quantification

Total RNA was extracted from samples for analysis of IL1β and TGFβ using the RNeasy mini kit (Qiagen, VIC, Australia) following manufacturer’s instructions. The protocol is described in Chapter 2 (2.1.2.5.1). The yield and purity of total RNA from biopsies and post- mortem uterine strips were quantified using the Bioanalyzer (Agilent Technologies, CA, USA), and further assessed based upon the RNA integrity number generated by the Bioanalyzer. After total RNA was extracted and assessed, it was stored at -80 °C until required (2.1.2.5.2).

Chapter 3 Impact of endometrial biopsy on endometrial health

3.2.2.3.2 qRT-PCR

Gene transcripts were measured by quantitative real-time polymerase chain reaction (qRT- PCR). Sets of primers for cytokines and an endogenous control, glyceraldehyde-3-phosphate dehydrogenase (GAPDH), were designed using Primer3Plus (Untergasser et al., 2007) from published sequences

Table 3.3). PCR experiments were conducted in triplicate for each transcript. For quantitative analysis of RNA, the QuantiTect SYBR Green RT-PCR kit and the real-time thermocycler Rotor-Gene 6000 (Corbett Life Science, USA) were used. Reactions were performed in 25 µl reaction tubes with 12.5 µl 2xQuantiTectSYBR Green RT-PCR master-mix, 0.25 µl QuantiTect RT Mix, 2.5 µl forward primer (12.5 µM), 2.5 µl reverse primer (12.5 µM), and 100 ng total RNA. To test for the absence of genomic DNA, reactions containing all components excluding QuantiTect RT Mix, were carried out at the same time as negative controls. The real-time RT-PCR protocol included the following default cycling conditions: 50°C for 30 min (reverse transcription), 95°C for 15 min (polymerase activation), 40 cycles of 95°C for 15 s (melt) and 6 °C for 30 s (anneal) and 72°C for 30 s (extension).The melting points of the amplified PCR products obtained confirmed specific amplification. Negative controls included a no template control (NTC) omits any DNA or RNA template from a reaction, no reverse transcriptase control (NRT). These controls did not show a signal, verifying that amplicons were not derived from contamination or genomic DNA.

3.2.2.3.3 Real-time PCR customised array (Experiment 2):

A customised RT2 Profiler PCR arrays (Cat#: CLAB20712R, Lot HZ14, Qiagen, CA, USA) (

Table 3.4) which could simultaneously assess eight genes related to inflammation was used. Each array disc contained 1 housekeeping gene (GAPDH) and 3 controls (1 positive-PCR, 1 reverse transcription, 1 genomic DNA). For reverse transcription, 0.8 µg total RNA was used in a final volume of 20 µl with an RT2 First Strand Kit (Cat 330401, Qiagen, Vic, Australia). The RT2 Profiler PCR Array Test was performed following manufacturers recommendations. Each cDNA sample (20 µl) was diluted with 91µl of RNase-free water. Then, 12.75 µl of CDNA synthesis reaction was mixed with 143.75 µl of, (Cat 330600, Qiagen, Vic, Australia) and 131 µl of RNase-free water to obtain a total volume of 287.5 µl. A total of 20 µl per well was added into the 96- well array disc. Real-time PCR were carried out using the real-time

Chapter 3 Impact of endometrial biopsy on endometrial health thermocycler Rotor-Gene 6000 (Corbett Life Science, USA) under the following cycling conditions: 95 oC for 10 mins followed by 40 cycles of 95 oC for 15 s and 60 oC for 30s. Ct values were uploaded onto the website http://www.qiagen.com/au/shop/genes-and- pathways/data-analysis-center-overview-page/ for online analysis

3.2.2.4 Statistical analysis

Gene expression was calculated as described by Galvao et al. (2011). Briefly, all Ct values for genes were normalised to the housekeeping gene GAPDH. The fold expression of each gene was estimated using the 2- ΔΔ Ct method. The average ΔCt for samples collected from non-biopsied heifers (control) was used as the calibrator. The ΔΔCT for each sample was calculated by subtracting the average ΔCT of the gene of interest (i.e TGFß biopsied) from the calibrator (TGFß calibrator) and then n-fold (2–ΔΔCt) could be calculated. The n-fold represents the endometrial gene expression in biopsied heifers relative to samples from non- biopsied heifers, normalized to the endogenous control GAPDH.

For all the analyses, statistical analyses were carried out in R 3.2.1 (R Core Team, 2015). A confidence interval of 95% was applied, and P-values were used to assess statistical significance using a significance level of α = 0.05.

a) Experiment 1:

For gene assessment, the level of RNA was normalised against GAPDH using the Shapiro test. Normalised means were analysed by the Wilcoxon signed rank test. Comparisons between biopsied and non-biopsied horns were conducted for every transcript B4-T7 and B7- O-T7 after biopsy. Then, comparisons between B4-T7 biopsied and non-biopsied horns were contrasted with B7-O-T7 biopsied and non-biopsied horns for every transcript.

b) Experiment 2

The average ΔCt for uterine samples collected at Day 4 (B4) or Day 7 (B7; control) were used as calibrator. To assess for differences in mean gene expression between B4 and B7 samples, a student t-test was performed. The mean gene expression was tested for normality (Shapiro-Wilk normality test), homogeneity of the variance (Bartlett's test) and sphericity (Mauchly's test). Outliers were removed from the dataset.

To evaluate change in inflammatory gene expression at Day 7 (B7) compared to gene expression of uterine tissue collected 10 days later (B7-O-T7), normally distributed data were 87

Chapter 3 Impact of endometrial biopsy on endometrial health analysed by ANOVA for repeated measurements. Comparisons between the original biopsy (B7) and biopsied (T7b) and non-biopsied horns (T7nb) were conducted for every transcript after biopsy. To evaluate whether inflammatory gene expression at Day 4 (B4) was significantly different to gene expression of biopsies collected 3 days after (B7), normally distributed data were analysed by ANOVA for repeated measurements. Comparisons between the original biopsy (B4) and biopsied (B7b) and non-biopsied horns (B7nb) were conducted for every transcript after biopsy.

Chapter 3 Impact of endometrial biopsy on endometrial health

Table 3.3 Sequences of primers for bovine cytokines and GAPDH in quantitative real-time polymerase chain reaction (qRT-PCR).

Accession Number Bp To Seq Reference NCBI L CACCCTCAAGATTGTCAGCA 492 GADPH 249 59.8 BC102589 Primer designed R TTGAGCTCAGGGATGACCTT 740

L TGACCCGCAGAGAGGAAATA 963 TGFβ 250 59.7 NM001166068 (Sugawara et al., 2010) R TTTTCTGTGGAGCTGAAGCA 1212

L CAGCTGCAGATTTCTCACCA 218 IL1β 250 59.8 M37211 (Konnai et al., 2003) R CCAGGGATTTTTGCTCTCTG 467 bp= base pair, T0 :Annealing temperature, Seq: Sequence

Chapter 3 Impact of endometrial biopsy on endometrial health

Table 3.4 Customised RT2 bovine inflammatory array (Qiagen Cat#: CAPB-13580,)

Ref. seq # Official Full Name 1. VEGFA NM_174216 Vascular endothelial growth factor A

2. CSF2 NM_174027 Colony stimulating factor 2 (granulocyte-macrophage)

3. IL1R1 NM_001206735 Interleukin 1 receptor, type I

4. IL6 NM_173923 Interleukin 6 (interferon, beta 2)

5. TNF NM_173966 Tumor necrosis factor

6. SPP1 NM_174187 Secreted phosphoprotein 1 (osteopontin)

7. TGFB1 NM_001166068 Transforming growth factor, beta 1

prostaglandin-endoperoxide synthase 2 (prostaglandin G/H synthase and

8. PTGS2 NM_174445 cyclooxygenase)

9. GAPDH NM_001034034 Glyceraldehyde-3-phosphate dehydrogenase

10.BGDC SA_00137 Cow Genomic DNA Contamination 11. RTC SA_00104 Reverse Transcription Control 12. PPC SA_00103 Positive PCR Control

Chapter 3 Impact of endometrial biopsy on endometrial health

3.2.3 Results

a) Experiment 1:

IL1β and TGFβ were up-regulated in heifers 3 days after biopsy (B4-T7) compared to non-biopsied heifers. At 10 days after biopsy (B7-O-T7), IL1β was down-regulated and TGFβ continued to be up-regulated compared to non-biopsied heifers.

IL1β expression was not different between biopsied and non-biopsied horns in individual animals at either 3 (B4-T7:P=0.125) or 10 days (B7-O-T7, P=0.8) after biopsy. However, IL1β expression was significantly higher in the biopsied horn at 3 compared to 10 days after biopsy (B4-T7 vs. B7-O-T7 P=0.031) (Figure 3.4).

Figure 3.4 Fold change in endometrial gene expression of IL1β in biopsied and non-biopsied horns at 3 days (B4-T7) and 10 days (B7-O-T7) after biopsy compared to non-biopsied heifers. BH: Biopsied horn, NBH; Non-biopsied horn. B4 (Biopsy at Day 4 post-oestrus), B7: (Biopsy at Day 7 post-oestrus); O: (Oestrus); T7: (Endometrial post-mortem tissue at Day 7 post-oestrus).

Chapter 3 Impact of endometrial biopsy on endometrial health TGFβ expression did not differ between the biopsied and non-biopsied uterine horns for individual animals at either 3 days (B4-T7 samples P=0.3) or 10 days (B7-O-T7 samples P=0.62) after biopsy (Figure 3.5B). TGFβ expression was also no different in biopsied horns at either 3 or 10 days after biopsy (P=0.22) (Figure 3.5).

Figure 3.5 Fold change in endometrial gene expression of TGFß in biopsied and non-biopsied horns in heifers biopsied 3 days (B4-T7) or 10 days (B7-O-T7) previously. BH: Biopsied horn, NBH; Non-biopsied horn. B4 (Biopsy at Day 4 post-oestrus), B7: (Biopsy at Day 7 post-oestrus); O: (Oestrus); T7: (Endometrial post mortem tissue at Day 7 post-oestrus).

Chapter 3 Impact of endometrial biopsy on endometrial health b) Experiment 2:

RNA samples from 4 heifers taken 10 days after their initial biopsy (B7-O-T7) were discarded because the quality of RNA was lower than manufacturer’s recommendations. RNA samples from two heifers taken 3 days after initial biopsy (B4-B7) were not included as insufficient samples were collected (not possible to pass the biopsy device through the cervix at Day 7).

Using RT2 Profiler PCR Arrays, the inflammatory gene expression in endometrial biopsies from beef heifers sampled at Day 4 (B4) or metoestrus and at Day 7 (B7) or early dioestrus were compared (Figure 3.6). Expression of RNA from endometrial genes encoding molecules involved in the inflammatory response between heifers biopsied at Day 4 (B4) and heifers biopsied at Day 7 (B7) post-oestrus was not significantly different (P>0.05).

Figure 3.6 Comparison (mean ±SE) of endometrial gene expression at Day 4 (b4) and at Day 7 (b7). Gene expression were normalised against internal control GAPDH expressed as ΔCt., CSF2: Colony stimulating factor 2; IL1R1: Interleukin 1 receptor, Type 1, IL6: Interleukin 6,: PTGS2: Prostaglandin endoperoxidase synthase 2; SPP1: Osteopontin; TGFB1: Transforming growth factor beta 1; TNF: Tumor necrosis factor; VEGFA: Vascular endothelial growth factor A Comparison (mean ±SE) of endometrial gene expression of beef heifers at Day 4 (b4) and at Day 7 (b7).

Chapter 3 Impact of endometrial biopsy on endometrial health To assess the change in gene expression over time, the effect of the biopsy was assessed by comparing the initial endometrial gene expression at Day 7 (B7) of the previous cycle with the gene expression in the biopsied (T7b) and non-biopsied (T7nb) horns, 10 days after the procedure (Figure 3.7). VEGFA was up-regulated in both biopsied and non-biopsied horns whilst ILR1 and IL6 were up-regulated only in the biopsied horn. CSF2, TNF, SPP1, TGFB1 and PTGS2 were down-regulated in both horns and IL6 was down-regulated in the non-biopsied horn (Figure 3.8). Expression of RNA for inflammatory cytokines did not differ statistically between groups (P>0.05).

Figure 3.7 Gene expression of endometrial pro-inflammatory and wound healing cytokines (mean±SE) from biopsies collected at Day 7 post-oestrus and biopsied (T7b) and non-biopsied horn (T7nb) at Day 7 post-oestrus. Gene expression was normalised against internal control GAPDH expressed as ΔCt

Chapter 3 Impact of endometrial biopsy on endometrial health

Figure 3.8 Fold change expression of inflammatory genes between biopsied (T7b) and non-biopsied horn (T7nb) at Day 7 post-oestrus, 10 days after injury. Relative gene expression for every transcript was calculated using the 2- ΔΔ Ct using GAPDH as housekeeping gene. Each bar represents the mean +/− SEM of six animals.

Similarly, to assess the change in gene expression over time the effect of EB was assessed by comparing gene expression at Day 4 (B4) and 3 days later in the biopsied (B7b) and non-biopsied horns (B7nb) (Figure 3.9). VEGFA and CSF2 were down-regulated in both biopsied and non- biopsied horn whereas ILIR, IL6, TNF, SPP1 and PTGS2 were up-regulated in biopsied and non- biopsied horns. TGFβ showed up-regulation in the biopsied horn and down-regulation in the non- biopsied horn (Figure 3.10). These differences in gene expression were not statistically significant (P>0. 05).

Chapter 3 Impact of endometrial biopsy on endometrial health

Figure 3.9 Comparison of inflammatory gene expression at Day 4 (mean±SE) post-oestrus and biopsied (B7b) and non-biopsied horn at Day 7 post-oestrus. Gene expression was normalised against internal control GAPDH expressed as ΔCt

Chapter 3 Impact of endometrial biopsy on endometrial health

4.00 B4-B7

3.50

3.00

2.50

2.00 b7b

b7nb Fold change Fold

1.50

1.00

0.50

0.00 VEGFA CSF2 IL1R1 IL6 TNF SPP1 TGFB1 PTGS2

Figure 3.10 Fold change expression of inflammatory genes between biopsied (B7b) and non- biopsied horn (B7nb) at Day 7 post-oestrus, 3 days after injury. Relative gene expression for every transcript was calculated using the 2- ΔΔ Ct using GAPDH as housekeeping gene. Each bar represents the mean +/− SEM of five animals.

Comparison of mean gene expression between biopsied and non-biopsied horns in B4-B7 and B7- O-T7 is summarised in Figure 3.11. Overall, the assessment 3 days (B4-B7) after biopsy showed up-regulation of IL6, TNF, SPP1 and PTGS2 and down-regulation of VEGF and CSF2 in both biopsied and non-biopsied horns. IL1R and TGFβ were up-regulated in biopsied horn and TGFβ was down-regulated in the non-biopsied horn. The gene expression profile assess 10 days after biopsy (B7-O-T7) showed up-regulation of VEGF and down-regulation of CSF2, TNF, SPP1,

Chapter 3 Impact of endometrial biopsy on endometrial health TGFB and PTGS2 in both horns. Both, IL1R and IL6 were up-regulated in the biopsied horn and down-regulated in the non-biopsied horn.

Figure 3.11 Comparison of gene expression (mean±SE) in biopsied and non-biopsied horns in B4- B7 and B7-O-T7. B4b (Biopsied horn at Day 4 post-oestrus) B4nb (Non-Biopsied horn at Day 4 post-oestrus) T7b (Post-mortem uterine tissue biopsied at Day 7 post-oestrus) T7nb (Post-mortem uterine tissue non-biopsied at Day 7 post-oestrus).

Chapter 3 Impact of endometrial biopsy on endometrial health 3.2.4 Discussion

This study aimed to evaluate the long (10 d) and short (3 d) term endometrial expression of targeted inflammatory genes after a biopsy performed either one cycle prior to (10 d) or during the ongoing cycle (3d) post-oestrus. In these experiments, the gene expression of a previously biopsied uterine horn was compared with that of 1) previously non-biopsied control heifers, 2) the contralateral non- biopsied horn and 3) subsequent gene expression within the same biopsied horn and non-biopsied horn. Changes to the expression of acute inflammatory (IL1α/β, TNFα, IL6, SPP1, CSF2, PTGS2) and wound healing genes (TGFβ, VEGF) were assessed after the mechanical injury caused by EB.

In Experiment 1, we found no significant difference in gene expression between the biopsied and non-biopsied horns in either 3 days (B4-T7) or 10 days (B7-O-T7) heifers. These results suggest that regardless of the injured horn, an endometrial sample can be recovered from any horn and the gene expression is representative of endometrial health. Our findings in gene expression agreed with previous studies conducted by Katagiri and Takahashi (2004) and Fischer et al. (2010) which did not report significant differences between RNA expression of inflammatory genes within the same horn or nearby areas or between the different regions (ipsilateral or contralateral horn to the CL) in postpartum cows with endometritis.

In this study, we found significant, although moderate, up-regulation of the pro-inflammatory gene IL1β (1 to 2 fold increase, P= 0.031) at 3 days compared to 10 days after biopsy. This up-regulation in the injured horn can be explained by the prompt and acute release of pro-inflammatory cytokines after the mechanical injury that activates the sterile inflammatory signalling cascade (Lukens et al., 2012). IL1β stimulate neutrophil and monocyte diapedesis from intact blood vessels promoting the phagocytosis of PAMPs and DAMPs. In sterile inflammation DAMPs released from injured tissue, stimulate IL1β by interacting with IL-1R initiating the innate immunity and host responses to tissue injury (Newton and Dixit, 2012). Based on the preliminary results, the histological assessment did reveal the absence of PMN (neutrophils) and the presence of mononuclear cells in the epithelium and submucosa suggesting that the resolution of the inflammatory process, 3 days after biopsy, is mostly mediated by cytokines.

Based on these preliminary findings, in Experiment 2 we assessed the individual response after EB using a customised inflammatory array to incorporate more pro-inflammatory and wound healing genes to characterise the long and short-term response. As part of this study, we had intended to include samples collected from the 3 days (B4-T7) heifers in Experiment 1. However, the RNA

Chapter 3 Impact of endometrial biopsy on endometrial health quality of these samples was found to be inadequate due to inappropriate storage of the samples to maintain RNA integrity.

Initially, we compared the endometrial gene expression at Day 4 with Day 7 post-oestrus. Expression of pro-inflammatory and wound healing genes was similar in endometrial tissue collected during metoestrus (Day 4) and early dioestrus (Day 7) (Figure 3.6).The expression of some cytokines, growth factors and their receptors by endometrial cells are regulated by ovarian steroidal hormones (Robertson, 2007, Guzeloglu-Kayisli et al., 2009). Within the targeted transcripts, an angiogenic cytokine such as VEGF is up-regulated during oestrus, enabling higher endometrial blood flow, angiogenesis and vascularization as required for endometrial growth during the proliferation and differentiation of epithelial and stromal cells (Tasaki et al., 2010). Likewise, a pro-inflammatory cytokine mediator PTGS2 is up-regulated during late luteal phase as a precursor for PGF2 synthesis from arachidonic acid (AA), playing a role in the initiation of luteolysis (Skarzynski et al., 2003). Other cytokines such as TNFα, IL6 (Fischer et al., 2010) and IL1β (Paula- Lopes et al., 1999) are not dependent on steroidal hormones, being expressed throughout the oestrous cycle. Although we did neither assess P4 concentrations at Day 4 nor at Day 7, the uterine environment in metoestrus (Day 4) or early dioestrus (Day 7) is regulated by rising concentrations of P4 secreted from the CL (Beltman et al., 2009a, Miyamoto et al., 2009, Forde et al., 2011a) . Therefore, from our results, we can suggest that the targeted genes are not mediated by P4 during metoestrus and early dioestrus. Plasma P4 concentrations gradually increase with the formation of the CL, reaching a maximum value in the mid-dioestrus phase between Day 10 to 14 of oestrous cycle (Kastelic et al., 1990, Siqueira et al., 2009). Larger variations in expression of immune genes have been observed in endometrial transcriptome profile from cows in mid-dioestrus (Day 12) compared to early metoestrus (Day 3.5) (Mitko et al., 2008). However, in this case, the difference in P4 concentration was not sufficient to stimulate changes in gene expression of these targeted genes. This statement is a speculation with no evidence that needs to be further clarified.

The assessment of the inflammatory gene expression either 3 days (Figure 3.9; B4-B7) or 10 days (Figure 3.7, B7-O-T7) after EB, showed no significant differences between the initial inflammatory gene expression and the subsequent response in both the biopsied and non-biopsied horns. After this assessment we compared the gene expression of the biopsied horn (b) with the non-biopsied horn (nb) (Figure 3.8 Figure 3.10). In 10 days (B7-O-T7) heifers, there was a non-significant trend in up- regulation of healing cytokines VEGF and pro-inflammatory cytokines such as IL1R and IL6 in the biopsied horn. Likewise, in 3 days (B4-B7) heifers no significant differences in inflammatory genes expression were observed in biopsied and non-biopsied horns. The gene expression profile observed followed the known acute inflammatory response that occurs during the first few days after injury 100

Chapter 3 Impact of endometrial biopsy on endometrial health (Salamonsen, 2003, Eming et al., 2007). Pro-inflammatory cytokines such, IL6, TNF, SPP1 and PTGS2 were up-regulated in both biopsied and non-biopsied horns whereas IL1R and TGFβ were up-regulated only in the biopsied horn. The expression of certain cytokines such as IL1α/β and IL1R which are activated in sterile inflammation, promote the expression of IL6 and TNFα (Lukens et al., 2012) and the latter stimulate the production of endometrial PTGS2 (COX2) which converts arachidonic acid (AA) to prostaglandins (Arosh et al., 2002). During the oestrous cycle, the maximal of COX2 mRNA between Days 16 and 18 corresponds to the time of luteolysis of the oestrous cycle. In this experiment, PTGS2 was upregulated at Day 7 post-oestrus. Whether the upregulation affects the CL function needs to be further elucidated.

This, positive feedback in pro-inflammatory cytokines expression (IL6, TNF, SPP1, PTGS2 IL1R and TGFβ) in 3 days (B4-T7) heifers was observed in this study and this cytokine-mediated response might explain the mild histological lesions observed in heifers biopsied during the ongoing cycle (B4-T7) compared to heifers biopsied one cycle earlier (B7-O-T7) in Study 1 (Histological response to endometrial biopsy).This could be partially explained by the fact that the concomitant administration of cloprostenol in heifers biopsied at Day 7 post-oestrus, induced oestrus 48-72 h after administration. The increase in uterine blood supply during oestrus, allows a physiological migration of PMN (especially neutrophils) to clear the uterus of cellular debris (Ohtani et al., 1993, Subandrio and Noakes, 1997). In mice, it has been demonstrated that endometrial repair is significantly retarded in the absence of neutrophils (Kaitu'u-Lino et al., 2007). Thus, the induction of oestrus may have exacerbated the positive influx of neutrophils in the injured tissue increasing the release of pro-inflammatory cytokines, chemokines and growth factors to enhance endometrial healing in 10 days (B7-O-T7) heifers. Furthermore, despite the histology failing to identify endometritis in biopsied heifers in Study 1, gene expression showed up-regulation of pro- inflammatory and wound healing cytokines between 3 to 10 days after injury in both biopsied and non-biopsied horns. These results suggest that within this time range (3 to 10 days), the endometrial tissue is still undergoing the wound healing process after biopsy.

In Study 1 of this chapter, we concluded that the histological assessment of biopsied and non- biopsied horns showed no morphological differences indicative of endometritis after the injury caused by biopsy. Chapwanya et al., (2009) reported that the severity of uterine inflammation mirrors the molecular expression of inflammatory genes. The results from Experiment 1 suggest that collecting a single EB from either the injured or non-injured horn is indicative of endometrial health. In addition, the uterus from heifers assessed 3 days after biopsy showed up-regulation of pro-inflammatory IL1β in the biopsied horn compared to the biopsied horn from heifers assessed 10 days after biopsy. 101

Chapter 3 Impact of endometrial biopsy on endometrial health In conclusion, these gene expression results support the hypothesis that the trauma caused by EB did not induce significantly molecular changes in inflammatory gene expression by the time of embryo arrival. These results indicate that performing a biopsy as close to the day of ET (Day 7) during ongoing metoestrus (Day 4) or biopsying during the cycle prior to ET, by the time of transfer the endometrium will be healed, thus not affecting embryo development and survival

102

Chapter 3 Impact of endometrial biopsy on endometrial health

3.3 Ultrasonographic changes in uterine horns after endometrial biopsy

Abstract

Transrectal ultrasound is a non-invasive tool to visualise the reproductive tract in farm animals to evaluate physiological and pathological conditions. Uterine fluid can be detected by ultrasound during normal oestrous cycle or in pathological uterine conditions such as subclinical endometritis. This study aimed to detect the presence of uterine fluid after performing endometrial biopsy (EB). Syncronised Droughmaster heifers (n=17) were randomly selected for EB at Day 4 (Group 1 B4 n=6) at Day 7 (Group 2 2XB7; n=8) or non-biopsy (Control, n=3). Heifers were biopsied once from the horn ipsilateral to the CL (Group 1), or in both uterine horns (Group 2). Uteri were scanned before and after the biopsy procedure. Ultrasound assessments prior to biopsy were performed on the day of Cue-Mate removal (Day -7) and then either at Day +4 (Group 1) or Day +7 (Group 2). Ultrasound assessments after biopsy were performed until Day +7 in Group 1 and from Day +9 to Day +13 and Day +16 in Group 2. Two experienced observers assessed the images obtained from the uterine horns per heifer for the presence (+/-) and estimation of the luminal fluid (scale 0-4; ≥ 2 being considered positive). Normally distributed data were analysed using Wilcoxon Sign Rank Test. Overall, the results showed after the initial biopsy was collected at Day 4 or at Day 7, luminal fluid (score ≥ 2) was detected in 54-77% (7/13 or 10/13, respectively) of biopsied heifers 48 to 72 h after biopsy (P<0.05). This accumulation of fluid in uterine horns declined in the subsequent exams. At Day 13, fluid was still accumulated in 2XB7 heifers in 25-37.5% (2/8 or 3/8, respectively). Both observers showed significant differences in fluid accumulation between Day +7 and Day+9 (P<0.05), but only OBS 1 found significant difference in fluid accumulation between D+9 and D+13 (P=0.003). In conclusion, this study showed that endometrial biopsy induced the accumulation of luminal fluid 48-72 h after the procedure. Endometrial repair via the innate immune system occurs rapidly so that luminal fluid by the end of luteal phase is almost cleared. Whether this fluid will affect embryo development and attachment needs to be further investigated.

103

Chapter 3 Impact of endometrial biopsy on endometrial health 3.3.1 Introduction

After mechanical injury or trauma, the immune system responds with rapid onset inflammatory reaction to restore functionality of the damaged tissue (Salamonsen, 2003, Weiss et al., 2009). This inflammatory reaction is characterised by acute changes in vascular flow accompanied by recruitment and migration of inflammatory cells with the subsequent release of cytokines, chemokines and growth factors needed for wound healing (Li et al., 2007). After injury, the blood flow increases as does the vascular permeability resulting in the release of protein-rich fluid (exudate) into the extravascular tissue (Ward et al., 2010). Transrectal ultrasound has been widely used in farm animals as a non-invasive tool to visualise the reproductive tract to evaluate various physiological and pathological conditions (Pierson and Ginther, 1988). Several studies have assessed the accumulation of uterine fluid by ultrasound during the oestrous cycle (Fissore et al., 1986, Kastelic et al., 1991) and postpartum (Mateus et al., 2002, Kasimanickam et al., 2004). For example, during the oestrous cycle, there is limited uterine fluid in dioestrus but this increases several folds during the periovulatory period and when the cow enters oestrus (Barlund et al., 2008). In non-bred heifers, Kastelic et al. (1991) found that the uterine fluid content was about 0.2 ml from Days 10 to Days 16, whereas at Day 18 it increased to up to 3.5 ml. In postpartum cattle and mares that have recently been bred, the presence of fluid in the uterine lumen has been associated with bacterial growth and is considered a positive sign for the diagnoses of subclinical endometritis (Troedsson, 1999, Barlund et al., 2008).

In cycling non-bred heifers, little is known about the endometrial repair following uterine biopsy on endometrial function the vascular responses caused after EB on endometrial function. The specific objective of the first part of this study was to detect the presence of uterine fluid after performing a biopsy at Day 4 post-oestrus (B4) and reassessing the uterus by the time of embryo arrival at Day 7 (Day 0= Day of oestrus).Since we are aiming to apply the model of collecting an EB during the cycle prior to ET, as explained in previous studies, two biopsies (one per horn) need to be collected on the day of ET (Day 7 post-oestrus, Day 0=Day of oestrus). Thus, in the second part of this study, the presence of luminal fluid after EB was performed in both horns at Day 7 was assessed during the luteal phase (Day 7 to Day 16).

104

Chapter 3 Impact of endometrial biopsy on endometrial health 3.3.2 Material and Methods

All the animal procedures were approved by the Animal Welfare Unit, UQ Research & Innovation, under permit AE02430. The experiments were carried out at the beef unit of the University of Queensland (Brisbane, QLD) at Pinjarra Hills.

The procedures for animal management, reproductive assessment for heifer’s selection, synchronisation protocols and biopsy procedure are described previously in Chapter 2 (2.1.2).

3.3.2.1 Experimental design

a) Experiment 1 Uterine ultrasound assessment after endometrial biopsy

Selected heifers were synchronised using the protocol described in Chapter 2 (2.1.2.2). Synchronised Droughtmaster heifers were selected (n=17) by simple randomization for EB (Suresh, 2011). In Group 1, only the uterine horn ipsilateral to the CL was biopsied at Day 4 post-oestrus (n=6; B4), then uterine fluid accumulation was assessed at Day 7 post-oestrus, 3 days after initial biopsy. In Group 2, both uterine horns were biopsied at Day 7 post-oestrus (n=8; 2xB7). Control heifers (n=3) were synchronised, but were not biopsied. Uteri were scanned before and after the biopsy procedure. Ultrasound assessments prior to biopsy were performed on the day of Cue-Mate removal (Day -7) and then either at Day +4 (Group 1) or Day +7 (Group 2). Ultrasound assessments after biopsy were performed until Day +7 in Group 1 and from Day +9 to Day +13 and Day +16 in Group 2. Heifers were not scanned after Day +16 to avoid the potentially confounding presence of uterine fluid from the follicular phase. Control heifers were scanned at Day -7, Day +7, Day+9, Day +13, Day +16 (Figure 3.12).

105

Chapter 3 Impact of endometrial biopsy on endometrial health

3.3.2.2 Uterine fluid assessment

Transrectal ultrasonography was performed using a SonoSite ultrasound M-Turbo (Washington USA) and a 7.5 MHz linear transducer. The transducer was placed into the rectum over the uterus, beginning scanning at the uterine bifurcation, and continuing towards the tip of each uterine horn. During scanning there was no distinction between right or left horn and the images were assessed for presence of uterine fluid regardless of the horn. Two to four images or videos were obtained from the uterine horns per heifer, and were subsequently assessed for the presence (+/-) and

106

Chapter 3 Impact of endometrial biopsy on endometrial health estimation of the luminal fluid (scale 0-4) by two experienced ultrasonographers (JH, GBH), blinded to animal ID, group and day of biopsy. A YES/NO score was given to the presence of luminal uterine fluid using a score system (0 to 4): 0: no fluid, 1: 1-2 mm fluid, 2: 2-3 mm fluid, 3: 3-4 mm fluid, 4: >5 mm fluid (Figure 3.13). The scores given by the two observers were averaged. This scoring system was described and validated by Pierson and Ginther (1987).

Figure 3.13 Ultrasonographic images of uterine horns collected during the study. Intraluminal non- echogenic (black) fluid collections represent categories (0-4) after EB.

3.3.2.3 Statistical analysis

A reliability analysis using the Kappa statistic (value between-1 and 1; k) was performed to determine consistency between the reviewers of the ultrasonograms for detection of uterine fluid accumulation. The scores given to the images were averaged per heifer and between observers (mean ± SE). The confidence interval was 95% and P- value <0.05 for significance. Statistical analyses were carried out in R 3.2.1 (R Core Team, 2015).

107

Chapter 3 Impact of endometrial biopsy on endometrial health 3.3.3 Results

The ultrasound assessment of one animal was discarded and not included in the study because no images were available to assess the uterine content prior to biopsy. Therefore, ultrasound images from a total of 16 heifers were assessed for presence (YES) or absence (NO) of uterine fluid. Analysing the scores given, the observers showed excellent agreement in the score given to the ultrasonograms κ=0.75 (0.60-0.89).

The ultrasound assessment in B4 heifers showed that after the first biopsy was taken for OBS1 4/5 (80%) heifers showed score ≥ 2 at Day 7 while for OBS2 1/5 (20%) heifers showed accumulation of uterine fluid (Table 3.5).

Table 3.5 Individual mean score given by two observers for assessment of accumulation of uterine fluid in heifers biopsied at Day 4 (Group 1, B4 n=5). Heifers were scanned prior to biopsy the day of Cue-Mate removal (Day -7) and then the day of biopsy (Day+4, before the procedure). After biopsy, the uterus was scanned at Day +7. Score system used from 0-4: 0 Uterine fluid absent, 1: 1- 2 mm uterine fluid; 2: 2-3 mm uterine fluid; 3: 3-4 mm uterine fluid; 4: > 5 mm uterine fluid. (-) Negative for scores 0 and 1; (+) Positive for scores ≥2

Before biopsy After biopsy

ID D-7 D+4 D+7

OBS1 0 0 2 3308 OBS2 0 0 0

OBS1 0 1 2 3313 OBS2 0 0 1

OBS1 1 0 0 B4 3315 (n=5) OBS2 1 0 0

OBS1 0.5 0 2 3328 OBS2 0.5 0 0

OBS1 2 1 2.5 3333 OBS2 2 0 2.5

108

Chapter 3 Impact of endometrial biopsy on endometrial health In Group 2, OBS1 found luminal fluid (≥2mm) in 1/8 (12.5%) prior to biopsy of both horns at Day 7. At Day 9, 6/8 (75%) heifers accumulated fluid at score ≥ 2. At Day 13 and Day 16 of scanning, 2/8 (25%) heifers continued to have luminal fluid. OBS2 reported the same findings as OBS1 in accumulation of luminal fluid before biopsy at Day -7 and after biopsy at D+9 (fluid ≥2mm). At Day 13 OBS2 found 3/8 (37.5%) continued with accumulation of luminal fluid and at Day 16 the fluid was persistent in 2/8 (25%) heifers (Table 3.6).

Table 3.6 Individual mean score given by two observers for assessment of accumulation of uterine fluid in heifers biopsied at Day 7 (Group 2, 2xB7 n=8) in both horns. Heifers were scanned prior to biopsy the day of Cue-Mate removal (Day -7) and then the day of biopsy (Day+7, before the procedure). After biopsy, the uterus was scanned at Day +9, Day +13, Day +16. Score system used from 0-4: 0 Uterine fluid absent, 1: 1-2 mm uterine fluid; 2: 2-3 mm uterine fluid; 3: 3-4 mm uterine fluid; 4: > 5 mm uterine fluid. (-) Negative for scores 0 and 1; (+) Positive for scores≥ 2

Before biopsy After biopsy

ID D-7 D+7 D+9 D+13 D+16 OBS1 0 1 2 1 1 3305 OBS2 0 0 2 2.5 1.3 OBS1 0 0 1.6 1 1 3311 OBS2 0 0 1.7 2 1 OBS1 0 1 2.5 0.5 0.3 3312 OBS2 0 0 2 1.5 0 OBS1 1 1 2 0.4 0.7 3332 2xB7 OBS2 1 1 2 0 0.3 (n=8) OBS1 1 0 4 0.5 0 Wini OBS2 1 0 4 0.5 1.7 OBS1 3.2 1.8 2 2 2.6 3189 OBS2 3.1 1.7 2 2 3.2 OBS1 0.7 0.7 1 1 1.2 3244 OBS2 0.7 0.7 1 1 0.8 OBS1 0 1.8 2 2 2 3252 OBS2 0 1.8 2 2 2.3

109

Chapter 3 Impact of endometrial biopsy on endometrial health In non-biopsied control heifers, OBS 1 found 2/3 (66.6%) heifers had accumulation of uterine fluid with score ≥ 2 on the day of intravaginal CIDR removal (Day -7). Luminal fluid was not present later that cycle. For OBS2 1/3 (33.3%) heifers had accumulation of luminal fluid the day of CIDR removal. One heifer showed fluid accumulation at Day 13 that was still persistent at Day 16 (

Table 3.7).

Table 3.7:Individual mean score given by two observers for assessment of accumulation of uterine fluid in non-biopsied heifers (Control n=3).Heifers were scanned prior to biopsy the day of Cue- Mate removal (Day -7) and then the uterus was scanned at Day +7, Day +9, Day +13, Day +16. Score system used from 0-4: 0 Uterine fluid absent, 1: 1-2 mm uterine fluid; 2: 2-3 mm uterine fluid; 3: 3-4 mm uterine fluid; 4: > 5 mm uterine fluid. (-) Negative for scores 0 and 1; (+) Positive for scores ≥2

ID D-7 D+7 D+9 D+13 D+16 OBS1 2 1 1 1 0.3 Trudi OBS2 2 0 1 0 0

Control OBS1 2 1 1 1.5 0.7 3223 (n=3) OBS2 1 1 1 1.5 0.7 OBS1 0 1.7 1 1 1 3352 OBS2 0 1.7 1 2 2

Overall, regardless of the day of biopsy 3-4 heifers (18.7-25% n=16) showed accumulation of uterine fluid (score ≥ 2) before the biopsy was collected. After the initial biopsy was collected at Day 4 (n=5) or at Day 7 (n=8), luminal fluid (score ≥ 2) was detected in 10/13 (77%) 48-72 h after biopsy for OBS1. For OBS, 2 fluid accumulation was detected in 7/13 (53.8%) 48-72 h after biopsy. This accumulation of fluid in uterine horns declined in the subsequent exams. According to OBS1, the accumulation of fluid (score ≥ 2) at Day + 13 and at Day +16 persisted in 2/8 (25%) of 2xB7 heifers from Group 2. For OBS2 fluid accumulation was still at Day +13 in (37.5%) and at Day 16 in 2/8 (25%) heifers.

The score mean of fluid accumulation by two observers is showed in Table 3.8. Significant differences in fluid accumulation was found between in biopsied heifers between Day +7 and Day+9 for OBS 1 (P=0.009) and OBS2 (P=0.01) (Figure 3.14 Figure 3.15). The comparison of fluid accumulation between D+9 and D+13 showed significant difference for OBS1 (P=0.003) but not for OBS2 (P=0.58).

110

Chapter 3 Impact of endometrial biopsy on endometrial health

Table 3.8 Score mean (±SD) assessment of accumulation of uterine fluid in heifers biopsied at Day 4 (Group 1 B4, n=5); at Day 7 in both horns (Group 2 2XB7, n=8); and non-biopsied heifers (Control, n=3) given by two observers (OBS1, OBS2). (B4). Heifers were scanned prior to biopsy the day of Cue-Mate removal (Day -7) and then the day of biopsy (Day+4, before the procedure). After biopsy, the uterus was scanned at Day +7. 2XB7 heifers were scanned prior to biopsy the day of Cue-Mate removal (Day -7) and then the day of biopsy (Day+7, before the procedure). After biopsy, the uterus was scanned at Day +9, Day +13, Day +16. Control heifers were scanned prior to biopsy the day of Cue-Mate removal (Day -7) and then the uterus was scanned at Day +7, Day +9, Day +13, Day +16. Score system used from 0-4: 0 Uterine fluid absent, 1: 1-2 mm uterine fluid; 2: 2-3 mm uterine fluid; 3: 3-4 mm uterine fluid; 4: > 5 mm uterine fluid. P-value <0.05 indicate significant difference in scores before and after biopsy (Group 1: D+4 vs. D+7; Group 2: D7 vs. D9).

D-7 D+4 D+7 D+9 D+13 D+16 P value OBS1 0.7 0.4 1.7 0.09

B4 (±0.83) (±0.54) (±0.97) (n=5) OBS2 0.7 0 0.7 0.37

(±0.83) (0) (±1.09)

OBS1 0.65 0.81 2.12 1.1 0.98 0.009 2XB7 (±1.05) (±0.71) (±0.81) (±0.61) (±0.88) (n=8) OBS2 0.82 0.73 2.09 1.42 1.51 0.01

(±1.1) (±0.78) (±0.91) (±0.93) (±0.98) OBS1 1.33 1.23 1 1.16 0.66 1 Control (±1.15) (±0.4) (0) (±0.28) (±0.35) (n=3) OBS2 1 0.88 1 1.16 0.88 0.06

(±1) (±0.83) (0) (±1.04) (±1.01)

111

Chapter 3 Impact of endometrial biopsy on endometrial health

Figure 3.14 Comparison of scores (±SD) given by Observer 1 for the assessment of accumulation of uterine fluid in heifers biopsied at Day 4, Day 7 and non-biopsied (control).* P< 0.05

Figure 3.15 Comparison of scores (±SD) given by Observer 1 for the assessment of accumulation of uterine fluid in heifers biopsied at Day 4, Day 7 and non-biopsied (control).* P< 0.05.

112

Chapter 3 Impact of endometrial biopsy on endometrial health 3.3.4 Discussion

This study aimed to evaluate the presence of ultrasonographically detectable uterine luminal fluid after endometrial biopsy performed either once at Day +4 or twice on Day +7. The use of ultrasonography to evaluate the uterine cavity for fluid accumulation prior to ET was previously reported in cases of subclinical endometritis (Kasimanickam et al., 2004, Barlund et al., 2008). This study demonstrated that uterine fluid accumulation can be detected within the first 48-72 h after EB (P<0.05). One possible explanation for accumulation of luminal fluid after biopsy is that occur as a consequence of extravasation of inflammatory exudate as part of the healing process. Another possible explanation is that during the procedure and the mechanical injury had induced pathological conditions. In cows (Gilbert, 2011), mares (Koblischke et al., 2008) and women (Boomsma et al., 2009b), ultrasound monitoring is routinely used to detect fluid in the uterine lumen before undergoing ET. The presence of uterine fluid in the uterine cavity may be a symptom of subclinical endometritis in cattle (Kasimanickam et al., 2004, Barański et al., 2012) or a postbreeding inflammatory response in mares (Watson, 2000) or a reflux of hydrosalpinx fluid into the uterine lumen in women (Sharara et al., 1996) all of which might have a deleterious effect on embryo development and survival, decreasing the pregnancy rates after AI/ET by affecting attachment (Maischberger et al., 2008, He et al., 2010, Brodzki et al., 2015).

One possible deleterious mechanism may arise from the presence of cytokines and inflammatory factors within the uterine fluid (Brodzki et al., 2015) as described in Study 2 (Inflammatory gene response to endometrial biopsy). Because attachment of embryos occurs in women within a week after ovulation, it has been suggested that the presence of luminal fluid forms a physical layer with adverse effects on cell proliferation or causes interference with apposition and attachment stages of the implantation process (Chien et al., 2002, Griffiths et al., 2002, He et al., 2010). In cows, the attachment process occurs after the second week of gestation, indicating that embryonic losses could be more strongly associated with the adverse embryotoxic environment that the embryo faces during its development within the uterus rather than a physical barrier. In this study, uterine fluid accumulation after biopsy did not persist beyond Day 13 of oestrous cycle in more than 70% of the biopsied heifers, suggesting that by the time of embryo attachment, the uterine fluid is cleared and if embryo losses occur they are most likely due to earlier effects on embryo development during blastocyst or hatching stages. However, this statement was not investigated in this experiment and no data is presented to support this, therefore further research is needed for confirmation.

Maiden heifers should have an intact sterile uterus, free of pathogens (Fischer-Tenhagen et al., 2012). In this study the detection of almost 20% of heifers with accumulation of uterine fluid before 113

Chapter 3 Impact of endometrial biopsy on endometrial health performing EB led us to speculate about the source of this fluid. After injury, In pathological conditions such as mucometra or hydrometra, which are caused by hypersecretion of endometrial glands or by failures in uterine drainage, are characterised by accumulation of sterile non-echogenic luminal fluid (Watson, 2000). However, these conditions are rare in cattle (Summers, 1974). It is possible that previous infections/trauma might have affected the uterine environment. For instance, the synchronisation protocol with intravaginal inserts and the subsequent oestrus induction with the opening of the cervical canal may have facilitated the entry of ascending bacteria that colonize the uterus and induce a subtle inflammatory reaction. Fischer-Tenhagen et al. (2012) reported that after oestrus synchronisation with intravaginal progesterone inserts, heifers (7/22; 32%), were detected with pyogenic bacteria and PMN cells in endometrial cytology collected on Day 7 and the day of breeding before AI. However, we used only maiden heifers that had not been previously mated or inseminated; thus, we assumed that the uterine cavity was sterile by the time of intravaginal P4 supplementation. Another possible explanation for the accumulation of uterine fluid before biopsy is related to the prevalence of infectious diseases that affect the reproductive tract (Campylobacter foetus subspecies veneralis, Tritrichomonas foetus, Ureaplasma diversum, Bovine virus diarrhoea virus.-BVDV) that might be accompanied with endometritis affecting non-bred heifers has been reported prevalent in Australia (Burns et al., 2010, Argue et al., 2013) and elsewhere (Cardoso et al., 2000). These hypotheses could have been confirmed by collecting an EB not only for the molecular assessment of inflammatory reaction as we previously showed (Study 2, Chapter 3) but also for bacterial culture and histology. This was not done as there was no evidence the animals were unhealthy and further biopsy sampling could have compromised the questions we were asking. Therefore, further studies are needed to clarify these findings.

In conclusion, this study showed that endometrial biopsy induced the accumulation of luminal fluid in more than 55% of heifers sampled 48-72 h after the procedure. Endometrial repair via the innate immune system occurs rapidly so that luminal fluid was detected in fewer than 20% of heifers by the end of the luteal phase. Whether this fluid will affect embryo development and attachment needs to be further investigated.

114

Chapter 4 Impact of endometrial biopsy on corpus luteum function

CHAPTER 4. INVESTIGATION OF THE IMPACT OF ENDOMETRIAL BIOPSY ON CORPUS LUTEUM FUNCTION

115

Chapter 4 Impact of endometrial biopsy on corpus luteum function Abstract

Uterine trauma stimulates the release of pro-inflammatory cytokines, chemokines and inflammatory mediators such as prostaglandins, accompanied by recruitment of immune cells to the site of injury. The release of endogenous pulses of PGF2α induces structural and functional regression of the CL, decreasing the progesterone (P4) levels. Since PGF2α has short half-life due to its metabolism in lungs and endometrium, the main indicator of PGF2α release into circulation is the assessment of its main metabolite, 13,14-dihydro-15-keto-PGF2α (PGFM). This study aims to evaluate the effect of uterine trauma (uterine manipulation UM, and endometrial biopsy EB) on corpus luteum lifespan. In addition, compare the plasma PGFM concentration before and after the uterine trauma. In Experiment 1, pre-synchronised Droughtmaster heifers (n=9; 25.8 ± 0.4 m, LW 391 ± 21.5 kg) were randomly selected either at Day 4 (Group 1, UM4=5) or at Day 7 (Group 2, UM7=4) post-oestrus for transrectal uterine manipulation and transcervical passage of the biopsy Storz® device without sample collection. For P4 analysis, blood samples were taken from the jugular or coccygeal veins throughout the on days: +3, +4, +6, +7, +9, +11, +13, +16 and +19 (Day 0=Synchronised oestrus). For PGFM analysis samples were measured before UM at -24, -4, 0 h (Hour 0= UM either at Day 4 or Day 7). After UM, samples were assessed at +6 and +24 h. In Experiment 2, tropically adapted Charbray heifers (n=13, 2-3 years old, LW 439.8±17.2) were randomly allocated into two groups: Biopsy at Day 4 (B4=6) or Biopsy at Day 7 (B7=7) post-oestrus. For P4 analysis blood samples were taken throughout the on days: +3, +4, +5, +6, +7, +8, +9, +10, +11, +13, +16 +17, +19 and +20 (Day 0=Synchronised oestrus). Samples for PGFM analysis samples were obtained both before biopsy (-24, -12, -4, 0 h) and after biopsy (+6, +18, +24, +36, +48, +60 and +84 h) in cows biopsied at Day 4 or Day 7. Progesterone was assessed by Liquid Chromatography Mass Spectrometry (LCMS). Concentrations of PGFM were determined using an enzyme immunoassay DetectX® kit (Arbor Assays, Ann Arbor, MI, USA). The intra and inter-assay coefficients of variations were 3.7% and 12.8% respectively. A series of generalized additive mixed effect models (GAMMs) were created to quantify P4 differences for UM and EB. Due to large variation in circulating PGFM, response data to UM or EB was calculated by taking hour 0 as the baseline and the value for each sample was converted into a proportion of the baseline. The results confirmed that neither UM, nor EB were predicted to induce luteolysis, thus affecting CL lifespan and both experiments showed higher concentrations of P4 when either UM or EB were performed at Day 4 compared to UM and EB at Day 7 post-oestrus. These results can be explained by differences that exist in PGF2α responsiveness between the CL from early in the cycle at metoestrus and the CL later in early dioestrus. Whether lower P4 profiles affect embryo development needs to be further elucidated.

116

Chapter 4 Impact of endometrial biopsy on corpus luteum function 4.1 Introduction

The corpus luteum (CL) is a temporary gland that secretes progesterone (P4), which is essential to regulation of the oestrous cycle, providing a uterine environment for establishment and maintenance of pregnancy (Schams and Berisha, 2004). CL function is assessed by the measurement of P4 in systemic blood or milk. In ruminants, a systemic P4 concentration threshold of above 1 ng/ml is suggestive of a functional CL, whilst concentrations below 1 ng/ml indicates a non-functional or a regressing CL (Chebel et al., 2006). The decrease in P4 is a consequence of structural and functional luteal regression of the CL, known as luteolysis, that is induced by the pulsatile release of prostaglandin F2-alpha (PGF2α) from the endometrium into the ovarian vein (Ginther et al. 2009). PGF2α is a member of a family of hormone-like molecules known as lipid mediators, which mediate a wide variety of physiological functions. PGF2α stimulates the contraction of smooth muscle in the uterus and bronchi and modulates vasoconstriction. Within the CL, endometrial PGF2α induces complex changes in local blood flow, apoptosis of steroidogenic luteal cells, up- regulation of inflammatory cells and vasoactive peptides and down-regulation of luteotropic receptors (Miyamoto et al., 2010). PGF2α is metabolised in the lungs and has a short half-life (Thatcher et al., 1986). Hence its metabolite 13,14-dihydro-15-keto-PGF2α (PGFM) which has a much longer half-life in the systemic circulation is used as a surrogate method of assessment of

PGF2α levels (Ginther et al., 2010). In cows undergoing luteolysis, it has been observed that PGF2α is released from the uterus over a 24 h period in five to eight episodes each of 1 to 5 h duration (Mann and Lamming, 2006). Hence, PGFM concentration after pulsation ranges between 250-600 pg/ml (Kotwica et al., 1998). More specifically, Ginther et al. (2010) characterised the mean values of PGFM during pre-luteolysis, luteolysis and post-luteolysis in dairy heifers that ranged from 127±27, 434±53 and 220±52 pg/ml, respectively. The release of endometrial PGF2α is also triggered by uterine trauma which is accompanied by mobilization of immune cells and the release of pro-inflammatory cytokines, chemokines, growth factors and other prostanoids (i.e. PGE, thromboxane A2) to the site of trauma (Scenna et al., 2005, Ricciotti and FitzGerald, 2011). Uterine trauma with the subsequent release of endogenous PGF2α may result from reproductive procedures involving uterine manipulation (UM) (Wann and Randel, 1990, Velez et al., 1991), such as passage and placement of devices in the uterine lumen through the cervix for artificial insemination (AI) embryo transfer (ET) (Scenna et al., 2005, Koblischke et al., 2008) or the collection of endometrial biopsy (EB) (Baker et al., 1981, Boos et al., 1996). This stimulation of PGF2α release may disrupt the including early CL regression or insufficient endometrial conditioning.

117

Chapter 4 Impact of endometrial biopsy on corpus luteum function There are species differences in the sensitivity and response of CL to PGF2α. In mares, trauma caused by handling the uterine tract after performing ET or EB during metoestrus or dioestrus, induced the release of luteolytic pulses of PGF2α from the endometrium, subsequently shortening the oestrous cycle (Baker et al., 1981, Watson et al., 1987, Gilbert, 1989, Koblischke et al., 2008). In cattle, although uterine trauma induces the release of pulses of PGF2α, this stimulus is unable to affect the CL lifespan (Scenna et al., 2005, Lopes et al., 2015, Aguiar et al., 2013). This variation in the CL response between species is thought to be due to differences in the rate of catabolism of PGF2α in the lungs. Mares are reported to have a lower rate of PGF2α catabolism which increases the half-life of PGF2α compared to cows (Weems et al., 2006). Thus, the systemic luteolytic dose of PGF2α in mares is one fifth that of cows (Ginther, 2012). In cattle, rapid metabolism of PGF2α by the lungs suggests that the uterine trauma caused by external procedures like transrectal uterine manipulation or more uterine invasive techniques such as EB, may not affect the CL lifespan. Endometrial biopsy (EB) is a useful diagnostic tool to obtain samples which can be utilised to identify and predict uterine function and receptivity by examining morphology, biochemistry and gene expression in cattle (Salilew-Wondim et al., 2010, Binelli et al., 2015). The application of EB to select superior recipients with an optimal expression of endometrial genes prior to ET should increase the proportion of pregnant recipients. In Chapter 3, it was demonstrated that performing an EB during metoestrus (Day +4) did not appear to induce sufficient endometrial inflammatory reaction ET three days later on Day +7 to impact uterine conditions to a level that would jeopardise embryo development or survival.

Embryo development within the uterus relies both on the endometrial environment and on maintenance of P4 secretion from the CL (Bazer et al., 2008). Thus, one aim of this study was to evaluate the functionality of the CL (based on P4 concentrations) subsequent to collection of a biopsy from the horn ipsilateral to CL on Day +4 after oestrus. In cattle, there is a substantial economic benefit in developing a test to predict the probability of pregnancy through assessment of endometrial receptivity. An endometrium with the highest probability of initiating and maintaining a pregnancy should express a different set of biological markers to an endometrium with a lower probability of pregnancy success. This study seeks to identify markers predictive of pregnancy which are expressed close to the day of embryo arrival and also are expressed from the uterine horn in which the embryo will develop (ipsilateral to the CL). In women, endometrial receptivity has been assessed either within a few days prior to the day of ET (Huang et al., 2011) or on the day of ET (Boomsma et al., 2009a). To compare the endometrial transcriptome of cows that would later be determined to be either pregnant or not pregnant, Binelli et al. (2015) collected endometrial biopsies on Day 6 after AI from the horn contralateral to the CL. 118

Chapter 4 Impact of endometrial biopsy on corpus luteum function Subsequently, they identified 216 differentially expressed genes (DEG) associated with pregnancy and non-pregnancy. The authors focused on expression of a subset of 9 genes they believed were indicative of an endometrium more highly receptive to pregnancy. Our study differs in that we will assess endometrial receptivity at Day 7 (the standard day of ET) and in the same and similar location to where the embryo will develop. We believe the timing and location of the EB will increase the specificity of detection of endometrial receptivity in cows destined as recipients for ET. In addition, the study will assess the effect of EB at Day 7 on CL function in cattle. While most of the studies have shown CL lifespan is not affected by either single or repeated endometrial biopsies during an oestrous cycle (Mann and Lamming, 1994, Boos et al., 1996, Katagiri and Takahashi, 2004) or during early pregnancy (Meikle et al., 2005, Pugliesi et al., 2014, Martin et al., 2015) one study showed oestrous cycles may be shortened after performing two endometrial biopsies in the same cycle (Zaayer and Van der Horst, 1986). While it appears that biopsy does not have a substantial negative effect on oestrous cycle length or conceptus survival there is release of endometrial PGF2α after the procedure. Martin et al., (2015) found that on the day of luteolysis (Day 18), PGFM concentrations in biopsied heifers were higher compared to non-biopsied controls (83.4 ±16.3 vs. 55.4 ±16.4, respectively P<0.05). Some studies have administrated cyclooxygenase inhibitors before procedures such as ET or EB to suppress endogenous release of PGF2α synthesis in response to uterine manipulation and the potential subsequent induction of luteolysis (Scenna et al., 2005, Aguiar et al., 2013, Pugliesi et al., 2014, Martin et al., 2015, Lopes et al., 2015).Cyclooxygenase enzymes (COX 1 and COX2) are responsible for the synthesis of prostaglandins from arachidonic acid (AA) that contribute to the acute inflammatory response. Cyclooxygenase inhibitors (non-steroidal anti-inflammatory drug NSAIDs, Flunixin Meglumine, Meloxicam) competitive bind and inactivate COXs site preventing catalysis of A.A to prostaglandins (Ricciotti and FitzGerald, 2011).Overall, these studies suggested that the administration of COX inhibitors at the time of the procedure had shown no effect on subsequent CL lifespan or oestrous cycle length (Martin et al., 2015) or conceptus survival (Geary et al., 2010, Aguiar et al., 2013, Pugliesi et al., 2014).

In cows, in the normal cycle endometrial PGF2α is transported by vascular countercurrent exchange from the uterine vein into the ovarian artery ipsilateral to the ovary that contains the CL and once there induces luteolysis (Ginther, 2012). Thus, the uterine horn where the biopsy is taken from, might affect the entry of PGF2α into the CL. In some studies that had concluded that oestrus length is not affected after biopsy, there is lack of information about the site of endometrial collection (Mann and Lamming, 1994, Katagiri and Takahashi, 2004) while other studies had collected biopsies from the horn contralateral to the CL avoided PGF2α reaching the functional CL (Pugliesi 119

Chapter 4 Impact of endometrial biopsy on corpus luteum function et al., 2014, Martin et al., 2015). Martin et al., (2015) showed administration of COX inhibitors prior to repeated EB of Nellore cows at days zero (ovulation), five, nine, thirteen and nineteen from the horn contralateral to the CL, did not change oestrus length. These studies show conclusively that biopsy from the horn contralateral to CL does not induce luteolysis.

The responsiveness of the CL to PGF2α varies according to the stage of the oestrous cycle. If prostaglandin is administered during the first five days after ovulation (metoestrus) the CL is refractory and the injected prostaglandin fails to induce luteolysis (Colazo et al., 2002). As the CL becomes functional there is a transition period from refractoriness to receptiveness to PGF2α during early dioestrus (Days 5 to 7) (Watts and Fuquay, 1985, Wenzinger and Bleul, 2012). The response to PGF2α increases in middle dioestrus (Days 8-11) reaching the maximum response during late dioestrus (Days 12-15) (Watts and Fuquay, 1985). Endometrial biopsies taken at different stages of the oestrous cycle in Holstein cows, at Days 1, 8, 15 and 19 from the horn ipsilateral to the CL did not show any significant effect on oestrus length (Boos et al. 1996). Thus, when assessing the effect of biopsy on CL function, it is important to identify on which day and stage of the oestrous cycle that the biopsy was taken. There is a need to clarify the effect on CL function from an EB taken from the horn ipsilateral to CL, without administration of COX inhibitors, and at stages of the oestrous cycle where the CL is responsiveness to PGF2α is known.

We hypothesised that the EB performed during metoestrus or early dioestrus will not induce luteolysis affecting the oestrous cycle length. This study aims to evaluate the effect of uterine trauma (transrectal uterine manipulation and EB) on plasma P4 concentration during the oestrous cycle; in addition, compare the plasma PGFM concentration before and after the uterine trauma. The specific objectives will assess the effects of 1) uterine manipulation and 2) Endometrial biopsy on systemic P4 and PGFM concentrations.

120

Chapter 4 Impact of endometrial biopsy on corpus luteum function 4.2 Material and Methods

4.2.1 Animals and management

The experiments were carried out at the beef unit of the University of Queensland (Brisbane, QLD) from Dec-2014-Jun 2015. Heifers grazed improved tropical pastures with water ad libitum. All the animal procedures were approved by the Animal Welfare Unit, UQ Research & Innovation, under permit AE02430.

The procedures for animal management, reproductive assessment for heifer’s selection, synchronisation protocols and biopsy procedure are described previously in Chapter 2 (Animals and management, Synchronisation of oestrus, Endometrial biopsy).

4.2.2 Synchronisation of oestrus:

Heifers (n=24) were selected to synchronise oestrus according to an optimised protocol (Butler et al., 2011). This protocol is already described in detail in Chapter 2 (Synchronisation of oestrus) Briefly, at Day -9 oestradiol benzoate (1mg, Cidirol®, Genetics Aust, Australia) was administered by intramuscular (IM) injection and an intravaginal P4 device (0.786g Cue-Mate®, Bioniche Animal Health lab, Australia) was inserted. At Day -2 the Cue-Mate was removed and Cloprostenol (Estromil®, 250 µg Ilium, Australia) was also administered IM. At Day -1, oestradiol benzoate (1mg) was administered IM. Day 0 (oestrus) was determined by twice daily observation for signs of heat. Only animals that showed signs of oestrus after synchronisation were enrolled.

4.2.3 Experimental design

a) Experiment 1: Effect of uterine manipulation without biopsy collection on P4 and PGFM profile

Nine Droughtmaster heifers (25.8 ± 0.4 m, LW 391 ± 21.5 kg) were synchronised following the protocol described above. Two groups of heifers were randomly selected either at Day 4 (Group 1, UM4=5) or at Day 7 (Group 2, UM7=4) post-oestrus for transrectal uterine manipulation and transcervical passage of the biopsy Storz® device without biopsy collection (Day 0 is the day of synchronised oestrus).

121

Chapter 4 Impact of endometrial biopsy on corpus luteum function b) Experiment 2: Effect of endometrial biopsy performed on metoestrus or dioestrus on P4 and PGFM profile

Tropically adapted non-pregnant Charbray heifers (n=13, 50% B.Iindicus 50% B. Taurus) (2-3 years old, LW 439.8±17.2) were selected. Heifers grazed improved tropical pastures with water ad libitum. At the time of commencement of the treatment, heifers were randomly allocated into two groups: Biopsy at Day 4 (B4=6) or Biopsy at Day 7 (B7=7) post-oestrus (Day 0 is the day of oestrus). EB was sampled as described in Chapter 2 (Endometrial biopsy).

4.2.4 Blood sampling

Blood samples for determination of P4 and PGFM were collected from either the jugular vein or coccygeal vein into 10-ml collection tubes (lithium heparin plasma separation, Vacuette) following the standard operating procedures (SOP ATT 015) approved by the Animal Ethics Committee. Blood samples were maintained in ice water and centrifuged at 2000g for 10 min. The recovered plasma was aliquoted in Eppendorf tubes and stored at -20o prior to assessment of P4 or PGFM.

- Experiment 1: For P4 analysis blood samples were taken on days: +3, +4, +6, +7, +9, +11, +13, +16 and +19 (Day 0=Synchronised oestrus). For PGFM analysis samples were measured before UM at -24, -4, 0 h (Hour 0= UM either at Day 4 or Day 7). After UM, samples were assessed at +6 and +24 h. - Experiment 2: For P4 analysis blood samples were taken throughout the on days: +3, +4, +5, +6, +7, +8, +9, +10, +11, +13, +16 +17, +19 and +20 (Day 0=Synchronised oestrus). Samples for PGFM analysis samples were obtained both before biopsy (-24, -12, -4, 0 h) and after biopsy (+6, +18, +24, +36, +48, +60 and +84 h) in cows biopsied at Day 4 or Day 7.

4.2.5 Progesterone assessment

Progesterone was assessed by Liquid Chromatography Mass Spectrometry (LCMS). The methods for sample preparation, standard solutions, extraction procedure and analytical conditions are described in Appendix C (Progesterone assessment).

4.2.6 PGFM assessment

Concentrations of PGFM were determined in unextracted plasma running duplicate samples using an enzyme immunoassay DetectX® kit (Arbor Assays, Ann Arbor, MI, USA) with a minimum detection limit of 46.2 pg/ml. The methods for sample preparation, standard solutions and assay 122

Chapter 4 Impact of endometrial biopsy on corpus luteum function validation are described in Appendix D .The intra and inter-assay coefficients of variations were 3.7% and 12.8% respectively.

4.2.7 Statistical analysis

Progesterone and PGFM concentrations are expressed as mean± S. E. Statistical analyses were carried out in R 3.2.1 (R Core Team, 2015).

A series of generalized additive mixed effect models (GAMMs) were created to quantify P4 differences for UM (equation 1) and EB (equation 2). The non-linear effect of day of oestrus (DOE) on P4 values was taken into account by incorporating DOE as an additive component in the models. Since multiple measurements of P4 were taken for each individual animal, heifers’ ID was included as a random component. GAMMs were fitted to the data using the mgcv package in R (Wood, 2006). The variables were modelled with a Gaussian distribution and an identity function as:

푃4푗푘 = 훽0 + 푠(퐷푂퐸)푀4푗 × 푡푥푀4푗 + 푠(퐷푂퐸)푀7푗 × 푡푥푀7푗 + 휀푘 (1)

Where:

- 훽0 is the intercept.

- 푠(퐷푂퐸)푀4 represents the smoothing function for DOE with an interaction with the

factor “manipulation at day 4” (푡푥푀4), while 푠(퐷푂퐸)푀7 represents the smoothing function

for DOE with an interaction with the factor “manipulation at day 7” (푡푥푀7). The variable

푡푥푀4푗 is 1 if an observation is from a heifer that underwent manipulation at day 4 and 0

otherwise. The variable 푡푥푀7푗 is 1 if an observation is from a heifer that underwent manipulation at day 7 and 0 otherwise. - 휀 represents the random intercept for heifer k. -

푃4푗푘 = 훽0 + 푠(퐷푂퐸)퐵4푗 × 푡푥퐵4푗 + 푠(퐷푂퐸)퐵7푗 × 푡푥퐵7푗 + 휀푘 (2)

Where:

- 훽0 is the intercept.

- 푠(퐷푂퐸)퐵4 represents the smoothing function for DOE with an interaction with the

factor “biopsy at day 4” (푡푥퐵4), while 푠(퐷푂퐸)퐵7 represents the smoothing function for DOE

with an interaction with the factor “biopsy at day 7” (푡푥퐵7). The variable 푡푥퐵4푗 is 1 if an 123

Chapter 4 Impact of endometrial biopsy on corpus luteum function

observation is from a heifer that was biopsied at day 4 and 0 else. The variable 푡푥퐵7푗 is 1 if an observation is from a heifer that was biopsied at day 7 and 0 else. - 휀 represents the random intercept for heifer k.

P-values were used to assess the significance of the variables. A significance level of α = 0.05 was used. Akaike’s Information Criteria (AIC) were used to select the most parsimonious models between the full models and a series of sub-models where the variables were removed one at the time. Diagnostic plots of residuals were inspected to assess the overall model fit. Residuals were tested for overdispersion.

Due to large variation in circulating PGFM concentrations between heifers (Figure 4.1), PGFM response data to UM or EB was calculated by taking hour 0 as the baseline and the value for each sample was converted into a proportion of the baseline. Proportions for each treatment were plotted as a function of time (h) before and after the baseline to assess the trend of the data.

Figure 4.1 Individual variations in plasma PGFM (pg/ml) concentration in biopsied heifers# 9 and # 11 from

124

Chapter 4 Impact of endometrial biopsy on corpus luteum function 4.3 Results

4.3.1 Experiment 1: Transrectal manipulation of uterine tract and placement of biopsy device within the uterine horns.

a) Progesterone concentration after uterine tract manipulation and biopsy device

Plasma P4 concentrations on Days +3, + 4,+ 6,+ 7, +9, +11,+13, +16 and +19 are shown in Figure 4.2. The results showed that neither manipulation at Day 4 (UM4) nor at Day 7 (UM7) induced luteolysis. However, the comparison of P4 profiles showed higher P4 profile in UM4 heifers than UM7 heifers throughout the days of oestrous cycle. This mean a P4 level above 1 ng/ml was reached by UM4 at Day 3 post-oestrus whilst for UM7 it was reached at Day 4. Plasma P4 concentration reached a maximum level at Day 11 in both UM4 (6.54 ± 1.21 ng/ml) and UM7 (3.81±3.58 ng/ml). Both UM4 and UM7 showed a plateau level of P4 between Day 7 to 13 that ranged between 5-6.5 ng/ml for UM4 and 2.5-3.81 ng/ml for UM7. At Day 19, plasma P4 levels in both UM4 and UM7 decreased to basal concentrations (1.44 ± 0.72 and 0.4± 0.14 in UM4 and UM7, respectively) (Figure 4.2a).

AIC and p-values indicated that all the variables tested in the GAMMs were statistically significant and contributed to the model. Plots of residuals did not show any distinctive pattern. No overdispersion was found (Table 4.1). The fitted curves confirmed that P4 profiles in UM4 showed higher concentrations than UM7 throughout the oestrous cycle and this difference could be due to the effect of early manipulation (Table 4.2b).

Table 4.1 Parameter estimates for the GAMMs specified in Equation (1).

Model #1 P4 after uterine manipulation

Estimate s.e. t-value p-value

훽0 3.012 0.498 6.049 6.02e-08

Smoothers E. df R. df F-stats p-value

s(DOE):txUM4 3.641 3.641 17.064 1.05e-09

s(DOE):txUM7 2.774 2.774 6.486 0.00122

125

Chapter 4 Impact of endometrial biopsy on corpus luteum function

Figure 4.2 a) P4 (ng/ml) measured from Day 3 to 20 of oestrous cycle (n=9). Five heifers (in red) were subjected to a UM at Day+4 of oestrous cycle, while remaining four heifers (in blue) were subjected to a UM at Day 7 of .b) Observed and fitted values for the model described in section 4.1.2.6 showing differences in P4 between treatments (UM4 and UM7). P4 for each treatment is plotted as a function of DOE (Day of oestrous). Each treatment is represented by a different colour. Observed values are plotted as dots. Fitted values are plotted as solid lines and represent the effect of DOE on P4 for each treatment. The shaded areas represented the 95 % point wise confidence intervals.

126

Chapter 4 Impact of endometrial biopsy on corpus luteum function

b) PGFM concentration after uterine tract manipulation and biopsy device

The mean plasma PGFM concentration after UM ranged from 69 to 1328.2 pg/ml. It was found that the average value (mean±SE) for PGFM in manipulated heifers was greater after the procedure than before (422.9±26.2 vs. 435.3±32.3 pg/mL, respectively). Regardless of the day of manipulation (Figure 4.3), there was no variability in PGFM concentration after UM. The comparison of treatments (Figure 4.3b) showed no differences in PGFM between UM4 and UM7 heifers before the biopsy procedure was performed. However, in UM7 heifers PGFM concentration increased 6 h post-manipulation compared to UM4 heifers in which the basal level remained unchanged.

Figure 4.3 PGFM (mean ± SE) measured from -24 h before UM to + 24 h after manipulation (n=9) a) Boxplot of proportions of PGFM before and after manipulation (UM4 and UM7 combined) On each box, the dot is the median, the base of the box the 25th and the top of the box the 75th percentiles, the whiskers extend to the most extreme data points not considered outliers, and any outliers are plotted individually.b) Comparison of PGFM concentration in UM4 and UM7 before

127

Chapter 4 Impact of endometrial biopsy on corpus luteum function and after UM. Five heifers (in red) were subjected to a UM at Day+4 of oestrous cycle (UM4), while remaining four heifers (in blue) were subjected to a UM at Day +7 of oestrous cycle (UM7).

128

Chapter 4 Impact of endometrial biopsy on corpus luteum function 4.3.2 Experiment 2: Endometrial biopsy at Day +4 (B4) or Day +7 (B7) post-oestrus.

a) Effect of endometrial biopsy on P4 concentration in B4 and B7 heifers

Two heifers (2/7; 28.5%) were excluded from the analysis in the B7 group (n=7) because the P4 concentration did not exceed threshold values above 1 ng/ml until Day 8 of the oestrous cycle. Neither EB at Day 4 (n=6, B4) nor at Day 7 (n=5, B7) induced luteolysis after the procedure. However, P4 concentration was higher in B4 than B7 heifers from Day 7 until Day 13 of oestrous cycle. The mean P4 level above 1 ng/ml was reached in B4 group (n=6) at Day 3 post-oestrus. Plasma P4 concentration increased to a maximum level at Day 7 in B4 (6.28 ± 1.53 ng/ml) and Day 8 in B7 (4.47 ± 2.71 ng/ml). Subsequently in both groups, P4 decreases until Day 10 (3.22± 1.67 and 2.42±1.21 ng/ml in B4 and B7, respectively), starting a fluctuation pattern that lasted until Day 16. From Day 16 onwards, plasma P4 level decreased in both groups reaching the lowest level at Day 19 (2.24±0.88 and 1.0±0.6 4 in B4 and B7, respectively). Plasma P4 levels beyond Day 19 were expected to decrease below 1 ng/ml when Cl regression was complete (Figure 4.4a).

AIC and p-values indicated that all the variables tested in the GAMMs were statistically significant and contributed to the model. Plots of residuals did not show any distinctive pattern. No overdispersion was found Table 4.2. The fitted curves confirmed that P4 profiles in the B4 group showed higher concentrations than in the B7 group throughout the oestrous cycle and this difference could be due to the effect of biopsy (Figure 4.4b).

Table 4.2 Parameter estimates for the GAMMs specified in Equation 2.

Model #2 – P4 Biopsy

Covariate Estimate s.e. t-value p-value

훽0 3.2732 0.2709 12.08 < 2e-16

Smoothers E. df R. df F-stats p-value s(DOE):txB4 6.426 6.426 10.29 4.95e-10 s(DOE):txB7 2.383 2.383 7.88 0.00025

129

Chapter 4 Impact of endometrial biopsy on corpus luteum function

Figure 4.4 a) P4 (ng/ml) measured from Day 3 to 20 of oestrous cycle (n=11). Six heifers (in red) were subjected to EB at Day+4 of oestrous cycle (B4), while remaining five heifers (in blue) were subjected to EB at Day 7 of oestrous cycle (B7). b) Observed and fitted values for the model described in section 4.1.2.6 showing differences in P4 between treatments (B4 and B7). P4 for each treatment is plotted as a function of DOE (Day of oestrus). Each treatment is represented by a different colour. Observed values are plotted as dots. Fitted values are plotted as solid lines and represent the effect of DOE on P4 for each treatment. The shaded areas represented the 95 % point wise confidence intervals.

130

Chapter 4 Impact of endometrial biopsy on corpus luteum function b) Effect of endometrial biopsy on PGFM concentration in B4 and B7 heifers

The mean plasma PGFM concentration after EB ranged from 54.2 to 1317.7 pg/ml. The average value (mean±SE) for PGFM in biopsied heifers was greater before than after the biopsy (432±40.8 vs. 392±36.3 pg/mL, respectively) (Figure 4.5). Prior to the biopsy, PGFM levels were similar between B7 and B4 heifers (Figure 4.5a). At 6 h post-biopsy however, PGFM concentration increased in B7 heifers compared to B4 heifers. B7 heifers continued to exhibit higher PGFM concentrations than B4 heifers between 24-48 h after the biopsy (Figure 4.5b).

Figure 4.5 PGFM (pg/ml) measured from -24 h before EB to + 84 h after EB (n=11) a) Boxplot of proportions of PGFM before and after biopsy (B4 and B7 combined). On each box, the dot is the median, the edges of the box are the 25th and 75th percentiles, the whiskers extend to the most extreme data points not considered outliers, and outliers are plotted individually.b) Comparison of PGFM concentration in B4 and B7 before and after EB. Six heifers (in red) were subjected to EB at Day+4 of oestrous cycle (B4), while remaining five heifers (in blue) were subjected to a UM at Day +7 of oestrous cycle (B7).

131

Chapter 4 Impact of endometrial biopsy on corpus luteum function 4.4 Discussion

The aim of this study was to investigate the effect of uterine manipulation (UM) and endometrial biopsy (EB) performed during metoestrus or early dioestrus on CL function and thus the oestrous cycle. The main findings of the study were: (1) According to previous studies conducted neither UM, nor EB were predicted to induce luteolysis, thus affecting CL lifespan and (2) both experiments showed higher concentrations of P4 when either UM or EB were performed at Day 4 compared to UM and EB at Day 7 post-oestrus. Physiologically, during late dioestrus, in the non-pregnant uterus the accumulated membrane phospholipids are converted into arachidonic acid (AA) by the activation of phospholipase A2 (PLA) which in turn is converted into PGF2α by the enzymes cyclooxygenase (COX-1 and COX-2) PGF2α secretion is a key regulator of the cyclical regression of the CL and thus P4 secretion during the oestrous cycle (Ginther et al., 2010). The endometrial PGF2α is transported from the uterine vein into the ovarian artery ipsilateral to the ovary containing the CL through a vascular counter current exchange mechanism, provoking a reduction in luteal blood flow and, inducing luteolysis (Arosh et al., 2002, Acosta et al., 2002). The release of chemical inflammatory mediators such as PGF2α from the uterine endometrium can also be triggered by excessive uterine handling during AI/ET or by mechanical injury from an EB, the (Velez et al., 1991, Aguiar et al., 2013, Martin et al., 2015).

The normal pattern of P4 concentration for an active and functional CL has been described in B. Indicus cattle. According to Figueiredo et al. (1997), P4 concentration increased steadily from the second day post-ovulation and reached a plateau at approximately Days 8 to 16, remaining elevated until luteolysis, which occurred between Days 16 and 19 of the oestrous cycle (Sartori and Barros, 2011). Martin et al. (2008) characterised the circulating P4 during the normal oestrous cycle in Nellore heifers. P4 increased from low concentrations at Day 0 (0.09 ± 0.13 ng/ml) and went up to concentrations above 1 ng/ml during early dioestrus at Day 5 (1.92 ± 1.17 ng/ml), reaching the maximum plateau from Day 9 (3.20 ± 1.46 ng/ml) to Day 13 (3.30 ± 1.41 ng/ml). These elevated concentrations remained until luteolysis occurred at Day 19 (2.41 ± 0.2.55 ng/ml) when a sharp decline to basal concentrations occurs. This is accompanied to a return to the normal oestrous cycle.

In the first experiment, we assessed P4 levels and CL functionality following UM (insertion of the biopsy device through the cervix into the uterine horn ipsilateral to CL but without collecting a sample). Studies in cattle using ET procedures had demonstrated that the mechanical stimulus and

132

Chapter 4 Impact of endometrial biopsy on corpus luteum function trauma originating from intrauterine devices, triggers the release of PGF2α but without affecting CL function (Scenna et al., 2005, Lopes et al., 2015). Scenna et al., (2005) reported that concentrations of PGF2α increased 1 to 4 h after performing ET (235 pg/ml 1h before ET vs. 245-250 pg/ml 1 to 4 h after ET, P < 0.001, respectively) although these values did not induce luteolysis. In our study, we also found that the mean PGFM concentrations were higher after UM (422.9±26.2 vs. 435.3±32.3 pg/mL, respectively) and that the CL was functional C and secreting P4. Interestingly, PGFM was elevated in the UM7 6 h after the manipulation which could be explained by the greater manipulation that is required to pass the device through a cervix that is closed tighter under P4 influence in early dioestrus (Day 7) as compared to the more relaxed cervix found in metoestrus (Day +4) when the P4 is still rising (Bonafos et al., 1995). However, further experiments that include more animals will help us to evaluate this observation.

As shown in Figure 4.2, manipulation and passing the catheter through the cervix did not have any effect on the CL activity since the P4 levels in both UM4 and UM7 animals, followed the normal pattern described in the natural oestrous cycle. In both groups, the mean P4 concentrations stayed above 1 ng/ml after manipulation suggesting an active CL during dioestrus, declining by Day 19 to levels below 1 ng/ml after luteolysis occurred. However, the P4 concentration in UM4 was higher than P4 in UM7 heifers. Therefore, the results of this experiment led us to conclude that UM with the endometrial device inserted but not taking the sample during metoestrus or early dioestrus had minimal or nil effects on CL activity.

During the second experiment, we wanted to assess whether collecting uterine tissue from the horn ipsilateral to the CL during metoestrus or early dioestrus affected the CL function. In cattle, previous studies had collected biopsies from the horn contralateral to CL to ensure a local release of endometrial PGF2α hence avoiding that PGF2α reach the opposite non-biopsied horn containing the CL (Pugliesi et al., 2014, Martin et al., 2015). In our study, we collected the biopsy from the horn ipsilateral at two different stages of CL development after oestrus: at Day 4 when the CL is still refractory to the luteolytic action of PGF2α and at Day 7 when the CL is mature and responsive to the effect of PGF2α (Tsai and Wiltbank, 1998, Acosta et al., 2002). As expected, the CL stayed active after biopsy throughout the oestrous cycle as shown by the P4 pattern during the oestrous cycle in both EB4 and EB7 heifers (Figure 4.4). These findings confirm previous studies (Mann and Lamming, 1994, Boos et al., 1996, Katagiri and Takahashi, 2004) that determined that the collection of endometrial biopsies during oestrous cycle do not affect CL lifespan hence oestrus length in cattle. This lack of luteolytic response could be explained by the low release of PGF2α after EB. Martin et al. (2015) showed that biopsied animals had higher concentrations of PGFM the day of luteolysis 133

Chapter 4 Impact of endometrial biopsy on corpus luteum function (P4 <1 ng/ml) compared to non-biopsied heifers (83.4 ±16.3 vs. 55.4±16.4 pg/ml; P<0.05). These luteolytic PGFM levels described by Martin, are lower compared to those reported by others which ranged between 250-600 pg/ml and which could induce luteolysis (Kotwica et al., 1998, Ginther et al., 2010). Unexpectedly, in this study the mean PGFM concentration before biopsy was higher than after biopsy (432±40.8 vs. 392±36.3 pg/mL, respectively) but these levels were unable to induce luteolysis in either B4 or B7 heifers. The differences in PGFM before and after the biopsy could be explained by the presence of outliers (at -24 h) that shifted the mean PGFM concentrations especially when a low number of animals were used. In addition, we also found high individual variation in the concentration of PGFM between animals before and after the procedures (from <100 pg/ml to 1300 pg/ml) (Figure 4.1). This variation in PGFM levels compared to observations reported from other studies, could be explained by differences in uterine health at the time of assessment, the uterine trauma (single or multiple biopsies), the frequency of sampling and the method of analysis (RIA or ELISA) (Wann and Randel, 1990, Velez et al., 1991, Mann and Lamming, 1994, Mateus et al., 2003). Likewise, it is difficult to determine from these results (Figure 4.3) whether PGFM secretion followed pulsatile release as reported in the literature (Ginther et al., 2009) because the blood sampling should have been done at a shorter intervals to being able to detect the pulsatile pattern and perhaps some peaks were missed in these experiments thus underestimating the mean values. Similarly, as described above in the manipulation experiment, the PGFM response in B7 heifers tended to be higher than B4 heifers, with increasing levels 6 h post-biopsy and thus exhibiting higher PGFM concentrations in the hours subsequent to the biopsy compared to B4 heifers (18 to 60 h after). We did not establish any subjective measurement to determine whether the biopsy was difficult to perform or not and thus more or less likely to cause trauma. However, we can conclude from both experiments that the uterine trauma caused by reproductive tract manipulation and the device traversing through the cervix to collect EB at Day 7 post-oestrus, is more difficult, and requires more manipulation that might induce trauma as reflected by PGFM concentrations. However, these PGFM concentrations still were unable to induce luteolysis.

There are two noteworthy results that arose from these experiments. Firstly, the statistical model showed that even though P4 concentration followed the normal cycling pattern throughout the oestrous cycle as seen in Figure 4.2b and Figure 4.4b, the shape of the curve in EB4 and UM4 heifers showed higher concentrations of P4 compared to EB7 and UM7 heifers. Secondly, the P4 concentrations in both biopsied groups (B4 and B7) showed a sharp decrease from Days 7-8 to Day 10, without showing a plateau phase (Figueiredo et al., 1997, Sartori and Barros, 2011), followed by fluctuations of P4 between Day 11-to Day 16. 134

Chapter 4 Impact of endometrial biopsy on corpus luteum function

The first effect can be explained by differences that exist in PGF2α responsiveness between the CL from early in the cycle at metoestrus and the CL later in early dioestrus. The metoestrus CL has been shown to be non-responsive to exogenous administration of PGF2α as demonstrated by the lack of intraluteal reduction in luteal blood flow in the early CL (Miyamoto et al., 2005) since CL volume and P4 concentration increased in parallel in the following days, without affecting the secretory capacity of luteal cells (Acosta et al., 2002). Beltman et al. (2009b) found that injecting single doses of PGF2α on Days +3, +3.5 and + 4 reduced the P4 concentrations, without affecting the CL weights, suggesting that low doses of P4 during metoestrus decreases the P4 output. Large luteal cells and endothelial cells in capillaries of the CL have prostaglandin receptors (Colazo et al., 2002), thus we suggest that the PGF2α secreted after biopsy or manipulation affected the secretory capacity of luteal cells in early 4 days-cycle CL, but this assumption requires further research.

However, at Day 7 post-oestrus the CL is responsive to PGF2α. In cattle, the secretion of PGF2α after EB as evidenced in this experiment and other studies (Mann and Lamming, 1994, Boos et al., 1996, Martin et al., 2015) is not high enough to induce luteolysis, thus shortening the oestrous cycle length. In our study, the amount of PGF2α secreted after biopsy may not have affected the lifespan of Day 7 CL, but possibly the secretory capacity decreased. In other words, the CL stayed active throughout the oestrous cycle, but apparently it wasn’t fully functional. Different studies confirmed that reduced doses of PGF2α injected at Day 7 induced partial luteolysis (Colazo et al., 2002). This partial luteolysis is accompanied with apoptosis of steroidogenic luteal cells, that decreases the synthesis of P4 in luteal tissue with the subsequent secretion of sub-optimal systemic concentrations of P4 (Juengel et al., 2000).

The explanation for the drop in P4 from Days 8 to 10 observed in these experiments and the subsequent fluctuation during dioestrus is unclear. Ginther et al. (2009) described that the luteolytic pulse of PGF2α is followed by a transient decrease of P4 that rebounds but does not return to the concentration prior to the pulse of PGF2α. Although in these experiments the results showed similar concentrations of PGFM before and after uterine trauma caused by biopsy, the early drop in P4 from Day 8 to 10 (Figure 4.4b) might confirm that non-luteolytic levels of PGF2α were secreted after EB. This statement is supported by Colazo et al., (2002) who found that after administering reduced doses of PGF2α on Day 7, plasma progesterone concentrations declined by 24 h but then recovered to pre-treatment values by 48-72 h after treatment. In this experiment, the concentration of P4 started a fluctuation period without recovering the maximum values in the subsequent days. Further studies are required to clarify this observation. 135

Chapter 4 Impact of endometrial biopsy on corpus luteum function

In conclusion, the mechanical stimulus and the injury caused after UM and passing the biopsy Storz® device through the uterine cervix with further collection of EB in the horn ipsilateral to CL, do not induce sufficient release of PGF2α to induce luteolysis. However, some effects caused by the uterine trauma (manipulation or biopsy) seems to be evident in animals in which uterine trauma occurred during early dioestrus (Day +7) since P4 secretion tends to be lower compared to animals when the trauma was caused during metoestrus (Day +4). Whether lower P4 profiles affect embryo development needs to be further elucidated in subsequent experiments with increased number of experimental animals.

136

Chapter 5 General Discussion

CHAPTER 5. GENERAL DISCUSSION

137

Chapter 5 General Discussion Within this thesis, three models of endometrial biopsy (EB) were investigated to identify, before embryo transfer (ET), markers of endometrial receptivity in recipients more likely to become pregnant. The EB was collected from the horn ipsilateral to the CL, 3-4 cm after the uterine bifurcation using the Storz® device. Day 7 post-oestrus was selected as the day of assessment because it is the day when most ETs are performed. Knowing the uterine environment that the embryo will face upon entering the uterus would be indicative of endometrial receptivity.

Three models of EB were developed and assessed. Model 1 involved the collection of EB at Day 7 of the oestrous cycle prior to the cycle when ET would be performed. Model 2 involved the collection of EB on Day 4 of the oestrous cycle in which ET would be performed. Lastly, Model 3 involved the collection of EB on the same day when ET would be performed. This thesis investigated the subsequent effects of EB on uterine health and CL function through morphological and molecular changes in uterine tissue, uterine ultrasound, serum progesterone (P4) and PGFM (13, 14-dihydro-15-keto-PGF2α) measurements. Previously no direct comparison of these models has been made and hence comparative assessment of the timing and impact of alteration of the timing between EB and when embryo transfer would occur has been possible. The outcome of this research would be a protocol for sampling which does not adversely affect the success of subsequent embryo transfer.

The general discussion will initially provide a description of the main contributions that this research adds to the current knowledge followed by analysis and description of the main advantages and disadvantages of collecting EB in each Model and at the end some description of the main limitations that this research had.

In this thesis, we are the first to describe the effectiveness of endometrial biopsy using the Storz® device (Karl Storz® 10366L) to collect tissue in beef heifers for transcriptome analysis. These findings were published in a Technical note (Appendix A) where the features of the device and the sample recovered are described (Ramirez-Garzon et al., 2017). The semi-rigid, slim, stainless steel Storz® device is designed for human bronchoscopy. The length (55cm) is similar to conventional Yomann Hauptner devices used in obtaining uterine samples in adult cows and mares. We reported that 86% of crossbreed beef heifers were successfully biopsied using this device (n= 44/51). Similar results were reported by Bonnett et al. (1991b) using a large-forceps Hauptner device in postpartum cows. One of the advantages of the Storz ® device is the tissue recovered weighed 30-40 mg which is lower than the 100 mg reported by Katagiri and Takahashi (2004) and 300-500 mg reported by Mann and Lamming (1994) using larger alligator-jaw devices. Moreover, the size of the 138

Chapter 5 General Discussion tissue recovered (2.5 mm) with the round cup, was lower compared to Kevorkian® alligator-jaw device (10 x 5 x 3 mm Düvel et al. (2014) and with the traditional Yomman Hauptner device (0.4 x 0.8 cm Mann and Lamming (1994), suggesting that the Storz® device results in less tissue damage. From this study, we concluded that the Storz® device was adequate to recover uterine tissue in beef heifers and the sample recovered provided sufficient good-quality tissue to perform transcriptome analysis.

We systematically reviewed the literature assessing the effects of uterine sampling procedures on the potential of cattle becoming pregnant. Analysis demonstrated that the pregnancy rates were similar between biopsied and non-biopsied animals, regardless of the method of sampling and it was not possible to make comparison between studies due to inconsistencies in the study design. These inconsistencies included variation in sampling conditions (type of animal, age and reproductive status) and the interval between sampling and breeding. More powerful studies with larger number of animals and consistent study design are needed to assess the effect endometrial sampling procedures on pregnancy rates in cattle.

Finally, this thesis describes the effects of EB collected during different stages of the oestrous cycle by assessing the endometrial and ovarian function at the day of ET (Day 7 post-oestrus) and the subsequent days throughout the same oestrous cycle. In Model 1, we found that performing EB at Day 7 post-oestrus followed by injection of PGF2α during the cycle prior to ET, the assessment of uterine health showed that the endometrium is already healed, 10 days after EB. In Model 2, we found that performing EB at Day 4 post-oestrus during the cycle of ET, the morphological changes, the expression of pro-inflammatory genes and the accumulation of uterine fluid assessed at Day 7 post-oestrus, indicate that the healing process is still ongoing, three days after injury. The CL function after EB from the horn ipsilateral to the CL, showed that the oestrous cycle length was not shortened and P4 concentrations remained above 1 ng/ml up to Day 16. Finally, in Model 3, we found that performing EB at Day 7 post-oestrus, accumulation of uterine fluid was found 48-72 h after EB. The fluid persists beyond Day 13 and tends to disappear by Day 16 around the time of maternal recognition of pregnancy (MRP). Although the oestrous cycle length was not shortened and the CL was active during the cycle with P4 levels above 1 ng/ml before luteolysis, the results suggest that CL function was impaired because the P4 output was reduced in heifers biopsied at Day 7 compared to heifers biopsied at Day 4.

While all the proposed three models allowed the assessment of endometrial receptivity, the main advantages and disadvantages in each model are:

139

Chapter 5 General Discussion The few morphological and molecular changes observed in Model 1, might be attributed to the combination of the small sample size retrieved by the Storz® device, accompanied by the administration of cloprostenol the day of biopsy and the subsequent increase in uterine blood supply after oestrus induction (Bollwein et al., 2000), This allows a physiological migration of PMN (especially neutrophils) to clear the uterus of cellular debris and necrotic and dead cells (Ohtani et al., 1993, Shaham-Albalancy et al., 1997, Roelofs et al., 2010). Thus, the induction of oestrus with the associated rise in oestrogen levels may have exacerbated the positive influx of neutrophils into the injured tissue increasing the release of pro-inflammatory cytokines, chemokines and growth factors to enhance endometrial healing. The results of this study confirm the hypothesis that EB performed one cycle before ET will not affect the subsequent uterine health by the day of ET, suggesting that this model of biopsying the uterus can be used to predict endometrial receptivity without affecting uterine health and the subsequent reproductive performance. However, the main disadvantage of using this model is that the sample is collected in a different cycle and large variation in endometrial gene expression exists between cycles due to changes in periovulatory follicle size and the subsequent secretion of steroidal hormones to prime the uterus (Mesquita et al., 2014). Therefore, the use of EB one cycle early offers the possibility of selecting, discarding and treating recipients based on the diagnosis of uterine health given by molecular and histological assessments, hence improving the likelihood of pregnancy before transferring the embryo. However, the development of endometrial arrays using this Model needs further investigation.

The results in Model 2 suggest that when an EB is performed during metoestrus there is no evidence of infectious endometritis subsequently on the day of ET. However, the sterile endometrial inflammatory response induced after EB and the subsequent release of immune cells and pro- inflammatory cytokines, chemokines and growth factors required for healing the tissue before embryo arrival could jeopardise embryo development and survival. Differences in gene expression exist between the two uterine horns during the oestrous cycle (Bauersachs et al., 2005, Hayashi et al., 2017). Therefore, for development of endometrial arrays, collecting biopsies at Day 4 post- oestrus from the horn ipsilateral to CL, the endometrial gene expression is more representative of endometrial receptivity. In this Model, the EB is performed when the reproductive tract is under the influence of rising levels of P4 secreted from the CL. In cattle, P4 reduces the immune response in the uterus hence the uterus becomes more susceptible to infections (Herath et al., 2006). P4 reduces the immune response in the uterus by lowering the expression of cytotoxic T-lymphocyte type I (Th-1), moving the response towards the less damaging lymphocyte type 2 (Th-2) response, to avoid the rejection of the allograft embryo. More important, as shown in this study, 48 hr after EB there is accumulation of uterine fluid within the uterine horns. Thus, the presence of cytokines and

140

Chapter 5 General Discussion inflammatory factors within the uterine fluid the day of transfer may have a deleterious effect on embryo development (Hill and Gilbert, 2008). Although Pugliesi et al. (2014) collected endometrial biopsies using the Hauptner device six days after insemination from the horn contralateral to the CL accompanied by a simultaneous injection of broad spectrum antibiotics and COX inhibitors, they found lower pregnancy rates at Day 30 in biopsied heifers than non-biopsied heifers. This suggests that even when collecting a biopsy from the contralateral horn, there is still a chance that triggering an inflammatory response could compromise embryo development.

In conclusion, it seems for the purpose of designing endometrial receptivity arrays from endometrial tissue taken from the horn ipsilateral to CL, within the same cycle as ET would be more representative of the uterine environment experienced by the transferred embryo rather than samples recovered one cycle earlier. However, there are possibly still unknown impacts of endometrial injury on embryo survival which need further investigation and for which bovine endometrial assays would be helpful.

Finally, the results of Model 3 suggest that for the purpose of designing endometrial receptivity arrays, recovering endometrial tissue on the day of ET from the horn ipsilateral to CL, would be the most representative of the uterine environment compared to samples recovered one cycle early or during metoestrus. However, embryo survival might be compromised by the presence of cytokines and inflammatory factors within the uterine fluid observed after EB which may have been shown to have a deleterious effect on embryo development (Hill and Gilbert, 2008). In humans and rodents, excessive uterine fluid affects implantation (Ruan et al., 2014) however, it is not clear in cattle whether the clearance of excessive uterine fluid affects embryo attachment. Since biopsied heifers are at the dioestrus stage, with high levels of systemic P4, there is no myometrial contractility to stimulate the fluid removal (Bonafos et al., 1995), thus the clearance of this fluid is through absorption by epithelial cells (Ruan et al., 2014) which could interfere with MRP and appositional attachment. However, this hypothesis needs further investigation. Similarly, although my results confirmed the hypothesis that EB performed at Day 7 post-oestrus will not affect the CL lifespan, CL functionality is impaired throughout the oestrous cycle due to the lower level of P4 found in heifers biopsied at Day 7 post-oestrus compared to heifers biopsied at Day 4 post-oestrus. Since the role of P4 during the second week of gestation is to regulate embryo elongation (Betteridge and Fléchon, 1988) and it is also required for a successful maternal recognition of pregnancy, reduction of P4 output from the CL results in changes to the endometrial transcriptome which in turn is responsible for a delay in conceptus elongation (Clemente et al., 2009, Forde et al., 2012). In conclusion, this Model could be the most accurate and indicative of endometrial receptivity but embryo survival may be highly compromised. 141

Chapter 5 General Discussion One of the main limitations of this thesis is the low number of animals used in the experiments. Further studies with larger number of animals are needed to verify the preliminary effects of sampling on pregnancy rates in recipients after ET described in this thesis. For example, whether the accumulation of uterine fluid after EB will affect embryo development, growth and elongation would be clarified by further investigation.

Likewise, the ultrasound assessment demonstrated that uterine fluid accumulation can be detected within the first 48-72 h after EB. This accumulation of uterine fluid can be associated with subclinical endometritis (SCE) (Barlund et al., 2008). However, detection of uterine fluid by ultrasonography alone is not a reliable method for diagnosis of SCE, and requires confirmation by another test such as endometrial cytology (Sheldon et al., 2009, de Boer et al., 2014). Although the accumulation of uterine fluid was detected at Day 7 post-oestrus, cytology samples were not collected at the same time as manipulation of the reproductive tract and passing the CB device through the cervix has the potential to induce a major inflammatory reaction. Since we were interested in assessing the response of EB by itself, collection of another sample could have been a confounding factor. Therefore, the use of non-invasive procedures to assess uterine blood flow such as using Doppler ultrasound could have been used to confirm SCE. However, based on the histological results it is considered very unlikely that the fluid accumulated in the uterine lumen corresponded to infectious endometritis since there was no evidence of PMN either in the epithelium or the submucosa. Thus, the most probable explanation for accumulation of fluid is extravasation of inflammatory non-infectious origin.

In comparing the effects of EB and UM, we found contradictory results on PGFM concentrations. In the UM experiment, we found higher mean PGFM concentrations after UM than before (422.9±26.2 vs. 435.3±32.3 pg/mL, respectively) which was the expected response. However, the mean PGFM concentrations in UM heifers were unexpectedly similar to mean PGFM concentrations before and after EB using the Storz® device (432±40.8 vs. 392±36.3 pg/mL, respectively) where 30-40 mg of endometrial tissue was retrieved. Although in this study, the PGFM concentrations in both experiments were within the luteolytic range, none of the heifers showed luteolysis after UM or EB. This variation in PGFM levels compared to other studies could be explained by the frequency of sampling. Ginther et al. (2009) described the pulsatile release of PGF2α to induce luteolysis by assessing plasma samples hourly. In this study, to assess luteolytic pulses the blood sampling frequency for PGFM analysis should have been done more frequently and not at the 4-12 hr intervals, because some peaks could have been missed with the subsequent underestimation of PGFM concentrations.

142

Chapter 5 General Discussion The main finding of this study was the development of an adequate model from which endometrial receptivity in cattle could be predicted using endometrial tissue obtained by biopsy. The model is reliable because we used beef heifers as experimental units, whose uterine environment has not been affected by previous breeding, gestation, calving, milk production or metabolic diseases. In addition, endometrial biopsy using Storz® device was used consistently as a method to recover material that subsequently can be used to perform endometrial transcriptome analysis and to assess uterine health. Moreover, this model demonstrated that the uterine environment was not affected after EB and the subsequent reproductive performance would not be compromised. Finally, the time point established to assess the uterine environment was always Day 7 which is the time around embryo arrival or transfer.

143

Chapter 5 General Discussion -

5.1 Future studies

A major conclusion of this thesis was that collecting endometrial samples in tropically adapted beef heifers is feasible using the Storz® device. Results of Chapter 2 confirmed that the uterine tissue collected by this device provides high-quality material to perform further molecular analysis for endometrial receptivity.

Another future application of the endometrial biopsy approach is to improve the diagnosis of pathological conditions that have been associated with altered endometrial receptivity in cases of unexplained infertility. In ET, the selection of recipients based on uterine health diagnosed by histopathology collected during the cycle before ET, could be used an alternative to select heifers with improved receptivity and thus increase the herd pregnancy rates.

Results of Chapter 3 suggest that collecting EB one cycle before ET and oestrus induction (Model 1) is the best approach to assess endometrial receptivity in terms of uterine health. As shown in studies 3.1 and 3.2, the morphological assessment and inflammatory gene expression suggest that by the time of embryo arrival at Day 7 of the following cycle, the uterus is healed. Studies in women have demonstrated that little variation in endometrial gene expression exists between consecutive menstrual cycles (Ruiz-Alonso et al., 2013, Díaz- Gimeno et al., 2013). Thus, further studies should be directed to profile the endometrial transcriptome during consecutive oestrous cycles using uterine tissue to characterise bovine endometrial receptivity.

Endometrial biopsy during the ongoing cycle would be more representative of endometrial receptivity than biopsy from previous cycle. However, in Chapter 3 we concluded that the morphological changes, the trend in up-regulation of pro-inflammatory cytokines (P>0.05) and the accumulation of uterine fluid suggest a mild-localised and non-infectious inflammatory uterine environment that may affect further embryo development by altering interaction with the uterus. Studies in cattle had shown the embryotoxic effects of pro- inflammatory cytokines in uterine fluid in infectious endometritis (Scenna et al., 2004, Hill and Gilbert, 2008). The inflammatory cascade is triggered by pathogen associated molecular pattern (PAMPs) such as bacterial lipopolysaccharides (LPS) with subsequent cellular response and the release of cytokines, chemokines and growth factors. To date little is known about the effects of the sterile inflammatory reaction after endometrial trauma on embryo development and pregnancy success in cattle. In women, the effect of induced mechanical 144

Chapter 5 General Discussion trauma on endometrial tissue (biopsy, scratch/hysteroscopy), improves endometrial receptivity (Potdar et al., 2012). Likewise, after tissue damage the endometrial cells detect damage-associated molecules (DAMPs) released from injured cells, triggering a sterile inflammatory response (Healy et al., 2014). To date, the molecular and cellular repair mechanism reported in women after endometrial injury, have not been studied in cattle. These studies are warranted to elucidate whether the cytotoxic effects of pro-inflammatory cytokines will affect embryo development in sterile inflammation.

We also conclude in Chapter 4 that biopsy during early dioestrus affected CL functionality but not the CL lifespan. It was unclear from this thesis the reason why P4 concentrations in biopsied heifers at dioestrus were lower than heifers biopsied during metoestrus. Further studies should be directed at evaluating CL functionality throughout oestrous cycle utilizing Color Dopler ultrasound, luteal gene expression and PGFM assessment. In addition, the effect of low P4 concentration after biopsy on the interaction between conceptus and uterine environment should be assessed for impact on embryo elongation, maternal recognition of pregnancy and pregnancy success. Ultimately, resolution of the reason for lower levels of P4 might require histological examination of the CL and especially the steroid producing luteal cells to detect any loss of cells which might result in lower P4 secretion.

145

Appendix References

ABEYGUNAWARDENA, H. & DEMATAWEWA, C. M. B. 2004. Pre-pubertal and postpartum anestrus in tropical Zebu cattle. Animal Reproduction Science, 82–83, 373-387. ACOSTA, T. J., HAYASHI, K. G., OHTANI, M. & MIYAMOTO, A. 2003. Local changes in blood flow within the preovulatory follicle wall and early corpus luteum in cows. Reproduction, 125, 759-767. ACOSTA, T. J., YOSHIZAWA, N., OHTANI, M. & MIYAMOTO, A. 2002. Local changes in blood flow within the early and midcycle corpus luteum after prostaglandin F2α injection in the cow. Biology of Reproduction, 66, 651-658. AGUIAR, T. S., ARAÚJO, C. V., TIRLONI, R. R. & MARTINS, L. R. 2013. Effect of Meloxicam on pregnancy rate of recipient heifers following transfer of In Vitro produced embryos. Reproduction in Domestic Animals, 48, 984-988. AHMADI, M. R., NAZIFI, S. & GHAISARI, H. R. 2006. Comparison of hormonal changes of estrous cycle with cytology of cervical mucosa and hematological parameters in dairy heifers. Comparative Clinical Pathology, 15, 94-97. AHMADI, M. R., TAFTI, A. K., NAZIFI, S. & GHAISARI, H. R. 2005. The comparative evaluation of uterine and cervical mucosa cytology with endometrial histopathology in cows. Comparative Clinical Pathology, 14, 90-94. ALMOG, B., SHALOM-PAZ, E., DUFORT, D. & TULANDI, T. 2010. Promoting implantation by local injury to the endometrium. Fertility and Sterility, 94, 2026-2029. ALTMÄE, S., MARTÍNEZ-CONEJERO, J. A., SALUMETS, A., SIMÓN, C., HORCAJADAS, J. A. & STAVREUS-EVERS, A. 2010. Endometrial gene expression analysis at the time of embryo implantation in women with unexplained infertility. MHR: Basic science of reproductive medicine, 16, 178-187. ARCHBALD, L. F., TSAI, I. F., THATCHER, W. W., TRAN, T., WOLFSDORF, K. & RISCO, C. 1998. Use of plasma concentrations of 13,14-dihydro,15-keto-PGF2 alpha (PGFM) in the diagnosis of sub-clinical endometritis and its relationship to fertility in the postpartum dairy cow. Theriogenology, 49, 1425-1436. ARGUE, B., CHOUSALKAR, K. K. & CHENOWETH, P. J. 2013. Presence of Ureaplasma diversum in the Australian cattle population. Australian Veterinary Journal, 91, 99-101. AROSH, J. A., PARENT, J., CHAPDELAINE, P., SIROIS, J. & FORTIER, M. A. 2002. Expression of Cyclooxygenases 1 and 2 and Prostaglandin E Synthase in Bovine Endometrial Tissue During the Estrous Cycle. Biology of Reproduction, 67, 161-169. AZAWI, O. I., OMRAN, S. N. & HADAD, J. J. 2008. A sudy of endometritis causing repeat breeding of cycling Iraqi buffalo cows. Reproduction in Domestic Animals, 43, 735-743. BACHA, W. J. & BACHA, L. M. 2011. Color atlas of veterinary histology, Chichester, West Sussex, Wiley-Blackwell. BAKER, C. B., NEWTON, D. I., MATHER, E. C. & OXENDER, W. D. 1981. Luteolysis in mares after endometrial biopsy. American Journal Veterinary Research, 42, 1816-1818. BALL, B., SHIN, S., PATTEN, V., LEIN, D. & WOODS, G. 1988. Use of a low-volume uterine flush for microbiologic and cytologic examination of the mare's endometrium. Theriogenology, 29, 1269-1283. BARAŃSKI, W., PODHALICZ-DZIĘGIELEWSKA, M., ZDUŃCZYK, S. & JANOWSKI, T. 2012. The diagnosis and prevalence of subclinical endometritis in cows evaluated by different cytologic thresholds. Theriogenology, 78, 1939-1947. BARASH, A., DEKEL, N., FIELDUST, S., SEGAL, I., SCHECHTMAN, E. & GRANOT, I. 2003. Local injury to the endometrium doubles the incidence of successful pregnancies in patients undergoing in vitro fertilization. Fertility and Sterility, 79, 1317-1322.

146

Appendix BARLUND, C. S., CARRUTHERS, T. D., WALDNER, C. L. & PALMER, C. W. 2008. A comparison of diagnostic techniques for postpartum endometritis in dairy cattle. Theriogenology, 69, 714-723. BAUERSACHS, S., ULBRICH, S. E., GROSS, K., SCHMIDT, S. E., MEYER, H. H., EINSPANIER, R., WENIGERKIND, H., VERMEHREN, M., BLUM, H., SINOWATZ, F. & WOLF, E. 2005. Gene expression profiling of bovine endometrium during the oestrous cycle: detection of molecular pathways involved in functional changes. Journal of Molecular Endocrinology, 34, 889-908. BAZER, F. W., BURGHARDT, R. C., JOHNSON, G. A., SPENCER, T. E. & WU, G. 2008. Interferons and progesterone for establishment and maintenance of pregnancy: interactions among novel cell signaling pathways. Reproductive Biology, 8, 179-211. BAZER, F. W., SONG, G., KIM, J., DUNLAP, K. A., SATTERFIELD, M. C., JOHNSON, G. A., BURGHARDT, R. C. & WU, G. 2012. Uterine biology in pigs and sheep. Journal Animal Science Biotechnology, 3, 1-23. BAZER, F. W., THATCHER, W. W., HANSEN, P. J., MIRANDO, M. A., OTT, T. L. & PLANTE, C. 1991. Physiological mechanisms of pregnancy recognition in ruminants. Journal of Reproduction and Fertility. Supplemment, 43, 39-47. BECHER, B., TUGUES, S. & GRETER, M. 2016. GM-CSF: From Growth Factor to Central Mediator of Tissue Inflammation. Immunity, 45, 963-973. BELTMAN, M. E., LONERGAN, P., DISKIN, M. G., ROCHE, J. F. & CROWE, M. A. 2009a. Effect of progesterone supplementation in the first week post conception on embryo survival in beef heifers. Theriogenology, 71, 1173-1179. BELTMAN, M. E., MULLEN, M. P., ELIA, G., HILLIARD, M., DISKIN, M. G., EVANS, A. C. & CROWE, M. A. 2014. Global proteomic characterization of uterine histotroph recovered from beef heifers yielding good quality and degenerate day 7 embryos. Domestic Animal Endocrinology, 46, 49-57. BELTMAN, M. E., ROCHE, J. F., LONERGAN, P., FORDE, N. & CROWE, M. A. 2009b. Evaluation of models to induce low progesterone during the early luteal phase in cattle. Theriogenology, 72, 986-992. BERENDT, F. J., FRÖHLICH, T., SCHMIDT, S. E., REICHENBACH, H. D., WOLF, E. & ARNOLD, G. J. 2005. Holistic differential analysis of embryo‐induced alterations in the proteome of bovine endometrium in the preattachment period. Proteomics, 5, 2551-2560. BERISHA, B., MEYER, H. H. & SCHAMS, D. 2010. Effect of prostaglandin F2 alpha on local luteotropic and angiogenic factors during induced functional luteolysis in the bovine corpus luteum. Biology of Reproduction, 82, 940-947. BERLANGA, O., BRADSHAW, H. B., VILELLA-MITJANA, F., GARRIDO-GOMEZ, T. & SIMON, C. 2011. How endometrial secretomics can help in predicting implantation. Placenta, 32, S271-5. BESENFELDER, U., HAVLICEK, V., KUZMANY, A. & BREM, G. 2010. Endoscopic approaches to manage in vitro and in vivo embryo development: Use of the bovine oviduct. Theriogenology, 73, 768-776. BETTERIDGE, K. J. & FLÉCHON, J. E. 1988. The anatomy and physiology of pre-attachment bovine embryos. Theriogenology, 29, 155-187. BINELLI, M., SCOLARI, S. C., PUGLIESI, G., VAN HOECK, V., GONELLA-DIAZA, A. M., ANDRADE, S. C. S., GASPARIN, G. R. & COUTINHO, L. L. 2015. The Transcriptome Signature of the Receptive Bovine Uterus Determined at Early Gestation. PLoS ONE, 10, e0122874. BLANCHARD, T. L., GARCIA, M. C., KINTNER, L. D. & KENNEY, R. M. 1987. Investigation of the representativeness of a single endometrial sample and the use of trichrome staining to aid in the detection of endometrial fibrosis in the mare. Theriogenology, 28, 445-450.

147

Appendix BLOCK, J., HANSEN, P. J., LOUREIRO, B. & BONILLA, L. 2011. Improving post-transfer survival of bovine embryos produced in vitro: Actions of insulin-like growth factor-1, colony stimulating factor-2 and hyaluronan. Theriogenology, 76, 1602-1609. BÓ, G. A., BARUSELLI, P. S. & MARTıŃ EZ, M. F. 2003. Pattern and manipulation of follicular development in Bos indicus cattle. Animal Reproduction Science, 78, 307-326. BOLLWEIN, H., MEYER, H. H. D., MAIERL, J., WEBER, F., BAUMGARTNER, U. & STOLLA, R. 2000. Transrectal doppler sonography of uterine blood flow in cows during the estrous cycle. Theriogenology, 53, 1541-1552. BONAFOS, L. D., KOT, K. & GINTHER, O. J. 1995. Physical characteristics of the uterus during the bovine estrous cycle and early pregnancy. Theriogenology, 43, 713-721. BONDURANT, R. H. 1999. Inflammation in the bovine female reproductive tract. Journal of Animal Science, 77 Suppl 2, 101-110. BONNETT, B. N. & MARTIN, S. W. 1995. Path analysis of peripartum and postpartum events, rectal palpation findings, endometrial biopsy results and reproductive performance in Holstein-Friesian dairy cows. Preventive Veterinary Medicine, 21, 279-288. BONNETT, B. N., MARTIN, S. W., GANNON, V. P., MILLER, R. B. & ETHERINGTON, W. G. 1991a. Endometrial biopsy in Holstein-Friesian dairy cows. III. Bacteriological analysis and correlations with histological findings. Canadian Journal of Veterinary Research, 55, 168- 173. BONNETT, B. N., MILLER, R. B., ETHERINGTON, W. G., MARTIN, S. W. & JOHNSON, W. H. 1991b. Endometrial biopsy in Holstein-Friesian dairy cows. I. Technique, histological criteria and results. Canadian Journal of Veterinary Research, 55, 155-161. BOOMSMA, C. M., KAVELAARS, A., EIJKEMANS, M. J., LENTJES, E. G., FAUSER, B. C., HEIJNEN, C. J. & MACKLON, N. S. 2009a. Endometrial secretion analysis identifies a cytokine profile predictive of pregnancy in IVF. Human Reproduction, 24, 1427-1435. BOOMSMA, C. M., KAVELAARS, A., EIJKEMANS, M. J. C., AMAROUCHI, K., TEKLENBURG, G., GUTKNECHT, D., FAUSER, B., HEIJNEN, C. J. & MACKLON, N. S. 2009b. Cytokine profiling in endometrial secretions: a non-invasive window on endometrial receptivity. Reproductive BioMedicine Online, 18, 85-94. BOOS, A., MEYER, W., SCHWARZ, R. & GRUNERT, E. 1996. Immunohistochemical assessment of oestrogen receptor and progesterone receptor distribution in biopsy samples of the bovine endometrium collected throughout the oestrous cycle. Animal Reproduction Science, 44, 11-121. BOURKE, M., MILLS, J. N. & BARNES, A. L. 1997. Collection of endometrial cells in the mare. Australian Veterinary Journal, 75, 755-758. BRINSKO, S. P., VARNER, D. D. & BLANCHARD, T. L. 1991. The effect of uterine lavage performed four hours post insemination on pregnancy rate in mares. Theriogenology, 35, 1111-1119. BRODZKI, P., KOSTRO, K., KRAKOWSKI, L. & MARCZUK, J. 2015. Inflammatory cytokine and acute phase protein concentrations in the peripheral blood and uterine washings of cows with subclinical endometritis in the late postpartum period. Veterinary Research Communications, 39, 143-149. BROOK, D. 1993. Uterine cytology. Equine reproduction. Lea & Febiger, Malvern, Pennsylvania. BROOKS, K., BURNS, G. & SPENCER, T. E. 2014. Conceptus elongation in ruminants: roles of progesterone, prostaglandin, interferon tau and cortisol. Journal of Animal Science and Biotechnology, 5, 53. BURNS, B. M., FORDYCE, G. & HOLROYD, R. G. 2010. A review of factors that impact on the capacity of beef cattle females to conceive, maintain a pregnancy and wean a calf— Implications for reproductive efficiency in northern Australia. Animal Reproduction Science, 122, 1-22. BUTLER, S. A. A., PHILLIPS, N. J., BOE-HANSEN, G. B., BO, G. A., BURNS, B. M., DAWSON, K. & MCGOWAN, M. R. 2011. Ovarian responses in Bos indicus heifers 148

Appendix treated to synchronise ovulation with intravaginal progesterone releasing devices, oestradiol benzoate, prostaglandin F2α and equine chorionic gonadotrophin. Animal Reproduction Science, 129, 118-126. CARDOSO, M. V., SCARCELLI, E., GRASSO, L. L. M. P. S., TEIXEIRA, S. R. & GENOVEZ, M. É. 2000. Ureaplasma diversum and reproductive disorder in Brazilian cows and heifers; first report. Animal Reproduction Science, 63, 137-143. CHAPWANYA, A., MEADE, K. G., DOHERTY, M. L., CALLANAN, J. J., MEE, J. F. & O’FARRELLY, C. 2009. Histopathological and molecular evaluation of Holstein-Friesian cows postpartum: Toward an improved understanding of uterine innate immunity. Theriogenology, 71, 1396-1407. CHAPWANYA, A., MEADE, K. G., NARCIANDI, F., STANLEY, P., MEE, J. F., DOHERTY, M. L., CALLANAN, J. J. & O’FARRELLY, C. 2010. Endometrial biopsy: a valuable clinical and research tool in bovine reproduction. Theriogenology, 73, 988-994. CHEBEL, R. C., SANTOS, J. E. P., CERRI, R. L. A., RUTIGLIANO, H. M. & BRUNO, R. G. S. 2006. Reproduction in Dairy Cows Following Progesterone Insert Presynchronization and Resynchronization Protocols. Journal of Dairy Science, 89, 4205-4219. CHEN, G. Y. & NUNEZ, G. 2010. Sterile inflammation: sensing and reacting to damage. Nature Reviews Immunology, 10, 826-837. CHEONG, S. H., NYDAM, D. V., GALVAO, K. N., CROSIER, B. M. & GILBERT, R. O. 2012. Effects of diagnostic low-volume uterine lavage shortly before first service on reproductive performance, culling and milk production. Theriogenology, 77, 1217-1222. CHIEN, L.-W., AU, H.-K., XIAO, J. & TZENG, C.-R. 2002. Fluid accumulation within the uterine cavity reduces pregnancy rates in women undergoing IVF. Human Reproduction, 17, 351- 356. CHOE, C., PARK, J.-W., KIM, E.-S., LEE, S.-G., PARK, S.-Y., LEE, J.-S., CHO, M.-J., KANG, K. R., HAN, J. & KANG, D. 2010. Proteomic analysis of differentially expressed proteins in bovine endometrium with endometritis. The Korean Journal of Physiology & Pharmacology, 14, 205-212. CLEMENTE, M., FUENTE, J. D. L., FAIR, T., NAIB, A. A., GUTIERREZ-ADAN, A., ROCHE, J. F., RIZOS, D. & LONERGAN, P. 2009. Progesterone and conceptus elongation in cattle: a direct effect on the embryo or an indirect effect via the endometrium? Reproduction, 138, 507-517. COCCHIA, N., PACIELLO, O., AULETTA, L., UCCELLO, V., SILVESTRO, L., MALLARDO, K., PARAGGIO, G. & PASOLINI, M. P. 2012. Comparison of the cytobrush, cottonswab, and low-volume uterine flush techniques to evaluate endometrial cytology for diagnosing endometritis in chronically infertile mares. Theriogenology, 77, 89-98. COLAZO, M. G., MARTıŃ EZ, M. F., KASTELIC, J. P. & MAPLETOFT, R. J. 2002. Effects of dose and route of administration of cloprostenol on luteolysis, estrus and ovulation in beef heifers. Animal Reproduction Science, 72, 47-62. DE BOER, M. W., LEBLANC, S. J., DUBUC, J., MEIER, S., HEUWIESER, W., ARLT, S., GILBERT, R. O. & MCDOUGALL, S. 2014. Invited review: Systematic review of diagnostic tests for reproductive-tract infection and inflammation in dairy cows. Journal of Dairy Science, 97, 3983-3999. DEBERTOLIS, L., MARI, G., MERLO, B., MERBACH, S., SCHOON, H. A., IACONO, E. & BOLLWEIN, H. 2016. Effects of induced endometritis on uterine blood flow in cows as evaluated by transrectal Doppler sonography. Journal of Veterinary Science, 17, 189-197. DEGUILLAUME, L., GEFFRÉ, A., DESQUILBET, L., DIZIEN, A., THOUMIRE, S., VORNIÈRE, C., CONSTANT, F., FOURNIER, R. & CHASTANT-MAILLARD, S. 2012. Effect of endocervical inflammation on days to conception in dairy cows. Journal of Dairy Science, 95, 1776-1783. DEL VECCHIO, R. P., MATSAS, D. J., INZANA, T. J., SPONENBERG, D. P. & LEWIS, G. S. 1992. Effect of intrauterine bacterial infusions and subsequent endometritis on prostaglandin 149

Appendix F2 alpha metabolite concentrations in postpartum beef cows. Journal of Animal Science, 70, 3158-3162. DHALIWAL, G. S., MURRAY, R. D. & WOLDEHIWET, Z. 2001. Some aspects of immunology of the bovine uterus related to treatments for endometritis. Animal Reproduction Science, 67, 135-152. DÍAZ-GIMENO, P., RUIZ-ALONSO, M., BLESA, D., BOSCH, N., MARTÍNEZ-CONEJERO, J. A., ALAMÁ, P., GARRIDO, N., PELLICER, A. & SIMÓN, C. 2013. The accuracy and reproducibility of the endometrial receptivity array is superior to histology as a diagnostic method for endometrial receptivity. Fertility and Sterility, 99, 508-517. DIEGELMANN, R. F. & EVANS, M. C. 2004. Wound healing: an overview of acute, fibrotic and delayed healing. Front Biosci, 9, 283-289. DISKIN, M., WATERS, S., PARR, M. & KENNY, D. 2016. Pregnancy losses in cattle: potential for improvement. Reproduction, Fertility and Development, 28, 83-93. DISKIN, M. G., PARR, M. H. & MORRIS, D. G. 2011. Embryo death in cattle: an update. Reproduction, Fertility and Development, 24, 244-251. DISKIN, M. G. & SREENAN, J. M. 1980. Fertilization and embryonic mortality rates in beef heifers after artificial insemination. Journal of Reproduction and Fertility, 59, 463-468. DORNIAK, P., BAZER, F. W. & SPENCER, T. E. 2013. PHYSIOLOGY AND ENDOCRINOLOGY SYMPOSIUM: Biological role of interferon tau in endometrial function and conceptus elongation12. Journal of Animal Science, 91, 1627-1638. DRENNAN, W. G. & MACPHERSON, J. W. 1966. The reproductive tract of bovine slaughter heifers (a biometrical study). Canadian Journal of Comparative Medicine and Veterinary Science, 30, 224-227. DUBUC, J., DUFFIELD, T. F., LESLIE, K. E., WALTON, J. S. & LEBLANC, S. J. 2010. Definitions and diagnosis of postpartum endometritis in dairy cows. Journal of Dairy Science, 93, 5225-5233. DUNNE, L. D., DISKIN, M. G. & SREENAN, J. M. 2000. Embryo and foetal loss in beef heifers between day 14 of gestation and full term. Animal Reproduction Science, 58, 39-44. DÜVEL, A., MAAß, J., HEPPELMANN, M., HUSSEN, J., KOY, M., PIECHOTTA, M., SANDRA, O., SMITH, D. G. E., SHELDON, I. M., DIEUZY-LABAYE, I., ZIEGER, P. & SCHUBERTH, H. J. 2014. Peripheral blood leukocytes of cows with subclinical endometritis show an altered cellular composition and gene expression. Theriogenology, 81, 906-917. EMING, S. A., KRIEG, T. & DAVIDSON, J. M. 2007. Inflammation in wound repair: molecular and cellular mechanisms. The Journal of Investigative Dermatoly, 127, 514-525. ETHERINGTON, W. G., MARTIN, S. W., BONNETT, B., JOHNSON, W. H., MILLER, R. B., SAVAGE, N. C., WALTON, J. S. & MONTGOMERY, M. E. 1988. Reproductive performance of dairy cows following treatment with cloprostenol 26 and/or 40 days postpartum: a field trial. Theriogenology, 29, 565-575. EURELL, J. A. & FRAPPIER, B. L. 2013. Dellmann's Textbook of Veterinary Histology. 6 ed. Hoboken: Wiley. FAULKNER, S., ELIA, G., O'BOYLE, P., DUNN, M. & MORRIS, D. 2013. Composition of the bovine uterine proteome is associated with stage of cycle and concentration of systemic progesterone. Proteomics, 13, 3333-3353. FIALA, S., ESMERALDINO, A., JOBIM, M., GARBADE, P., WOLF, C., RICHTER, G., GREGORY, R. & MATTOS, R. 2010. Endometrial fibrotic changes. Is one biopsy enough to diagnose degenerative changes. Animal Reproduction Science, 121, S89-S90. FIGUEIREDO, R. A., BARROS, C. M., PINHEIRO, O. L. & SOLER, J. M. P. 1997. Ovarian follicular dynamics in nelore breed (Bos indicus) cattle. Theriogenology, 47, 1489-1505. FISCHER-TENHAGEN, C., VON KRUEGER, X. & HEUWIESER, W. 2012. Short communication: Evaluation of vaginal discharge following treatment with a progesterone insert. Journal of Dairy Science, 95, 4447-4451. 150

Appendix FISCHER, C., DRILLICH, M., ODAU, S., HEUWIESER, W., EINSPANIER, R. & GABLER, C. 2010. Selected pro-inflammatory factor transcripts in bovine endometrial epithelial cells are regulated during the oestrous cycle and elevated in case of subclinical or clinical endometritis. Reproduction, Fertility and Development, 22, 818-829. FISSORE, R. A., EDMONDSON, A. J., PASHEN, R. L. & BONDURANT, R. H. 1986. The use of ultrasonography for the study of the bovine reproductive tract. II. Non-pregnant, pregnant and pathological conditions of the uterus. Animal Reproduction Science, 12, 167-177. FLEIGE, S. & PFAFFL, M. W. 2006. RNA integrity and the effect on the real-time qRT-PCR performance. Molecular Aspects of Medicine, 27, 126-139. FORDE, N., BELTMAN, M. E., DUFFY, G. B., DUFFY, P., MEHTA, J. P., Ó'GAORA, P., ROCHE, J. F., LONERGAN, P. & CROWE, M. A. 2011a. Changes in the Endometrial Transcriptome During the Bovine Estrous Cycle: Effect of Low Circulating Progesterone and Consequences for Conceptus Elongation1. Biology of Reproduction, 84, 266-278. FORDE, N., BELTMAN, M. E., LONERGAN, P., DISKIN, M., ROCHE, J. F. & CROWE, M. A. 2011b. Oestrous cycles in Bos taurus cattle. Animal Reproduction Science, 124, 163-169. FORDE, N., CARTER, F., SPENCER, T. E., BAZER, F. W., SANDRA, O., MANSOURI-ATTIA, N., OKUMU, L. A., MCGETTIGAN, P. A., MEHTA, J. P., MCBRIDE, R., O'GAORA, P., ROCHE, J. F. & LONERGAN, P. 2011c. Conceptus-Induced Changes in the Endometrial Transcriptome: How Soon Does the Cow Know She Is Pregnant? Biology of Reproduction, 85, 144-156. FORDE, N., MCGETTIGAN, P., BROWNE, J., CARTER, F., LOFTUS, B. J., ROCHE, J. F., LONERGAN, P., CROWE, M. A. & EVANS, A. C. O. 2009. Pregnancy-Specific Endometrial Gene Expression at the Initiation of Conceptus Elongation in Cattle. Biology of Reproduction, 81, 303-303. FORDE, N., MEHTA, J. P., MINTEN, M., CROWE, M. A., ROCHE, J. F., SPENCER, T. E. & LONERGAN, P. 2012. Effects of low progesterone on the endometrial transcriptome in cattle. Biology of Reproduction, 87, 124. GABLER, C., DRILLICH, M., FISCHER, C., HOLDER, C., HEUWIESER, W. & EINSPANIER, R. 2009. Endometrial expression of selected transcripts involved in prostaglandin synthesis in cows with endometritis. Theriogenology, 71, 993-1004. GABLER, C., FISCHER, C., DRILLICH, M., EINSPANIER, R. & HEUWIESER, W. 2010. Time- dependent mRNA expression of selected pro-inflammatory factors in the endometrium of primiparous cows postpartum. Reproductive Biology and Endocrinology, 8, 152. GALVAO, K. N., SANTOS, N. R., GALVAO, J. S. & GILBERT, R. O. 2011. Association between endometritis and endometrial cytokine expression in postpartum Holstein cows. Theriogenology, 76, 290-299. GEARY, T. 2005. Management strategies to reduce embryonic loss. In: NEBRASKA, U. O., ed. Range Beef Cow Symposium 2005 Rapid City, South Dakota. University of Nebraska, 69- 78. GEARY, T. W., ANSOTEGUI, R. P., MACNEIL, M. D., ROBERTS, A. J. & WATERMAN, R. C. 2010. Effects of flunixin meglumine on pregnancy establishment in beef cattle1. Journal of Animal Science, 88, 943-949. GEARY, T. W., BURNS, G. W., MORAES, J. G. N., MOSS, J. I., DENICOL, A. C., DOBBS, K. B., ORTEGA, M. S., HANSEN, P. J., WEHRMAN, M. E., NEIBERGS, H., O'NEIL, E., BEHURA, S. & SPENCER, T. E. 2016. Identification of Beef Heifers with Superior Uterine Capacity for Pregnancy. Biology of Reproduction, 95, 47, 1-12. GHANEM, N., SALILEW-WONDIM, D., GAD, A., TESFAYE, D., PHATSARA, C., THOLEN, E., LOOFT, C., SCHELLANDER, K. & HOELKER, M. 2011. Bovine blastocysts with developmental competence to term share similar expression of developmentally important genes although derived from different culture environments. Reproduction, 142, 551-564.

151

Appendix GHASEMI, F., GONZALEZ-CANO, P., GRIEBEL, P. J. & PALMER, C. 2012. Proinflammatory cytokine gene expression in endometrial cytobrush samples harvested from cows with and without subclinical endometritis. Theriogenology, 78, 1538-1547. GILBERT, R. O. 1989. The effect of endometrial biopsy on ovulatory interval in mares. Theriogenology, 32, 451-454. GILBERT, R. O. 2011. The effects of endometritis on the establishment of pregnancy in cattle. Reproduction, Fertility and Development, 24, 252-257. GILBERT, R. O., SHIN, S. T., GUARD, C. L., ERB, H. N. & FRAJBLAT, M. 2005. Prevalence of endometritis and its effects on reproductive performance of dairy cows. Theriogenology, 64, 1879-1888. GINTHER, O. J. 2012. The end of the tour de force of the corpus luteum in mares. Theriogenology, 77, 1042-1049. GINTHER, O. J., ARAUJO, R. R., PALHÃO, M. P., RODRIGUES, B. L. & BEG, M. A. 2009. Necessity of Sequential Pulses of Prostaglandin F2alpha for Complete Physiologic Luteolysis in Cattle. Biology of Reproduction, 80, 641-648. GINTHER, O. J., GASTAL, E. L., GASTAL, M. O., UTT, M. D. & BEG, M. A. 2007. Luteal blood flow and progesterone production in mares. Animal Reproduction Science, 99, 213- 220. GINTHER, O. J., KNOPF, L. & KASTELIC, J. P. 1989. Temporal associations among ovarian events in cattle during oestrous cycles with two and three follicular waves. Journal of Reproduction and Fertility, 87, 223-230. GINTHER, O. J., SHRESTHA, H. K., FUENZALIDA, M. J., SHAHIDUZZAMAN, A. K. M. & BEG, M. A. 2010. Characteristics of Pulses of 13,14-Dihydro-15-Keto-Prostaglandin F2alpha Before, During, and after Spontaneous Luteolysis and Temporal Intrapulse Relationships with Progesterone Concentrations in Cattle. Biology of Reproduction, 82, 1049-1056. GNAINSKY, Y., GRANOT, I., ALDO, P. B., BARASH, A., OR, Y., SCHECHTMAN, E., MOR, G. & DEKEL, N. 2010. Local injury of the endometrium induces an inflammatory response that promotes successful implantation Fertility and Sterility, 94, 2030-2036. GODKIN, J. D. & DORE, J. J. 1998. Transforming growth factor beta and the endometrium. Reviews of Reproduction, 3, 1-6. GOSHEN, T., GALON, N., ARAZI, A. & SHPIGEL, N. 2012. The effect of uterine biopsy on reproductive performance of dairy cattle: A case control study. Israel Journal of Veterinary Medicine, 67, 34-38. GRANOT, I., GNAINSKY, Y. & DEKEL, N. 2012. Endometrial inflammation and effect on implantation improvement and pregnancy outcome. Reproduction, 144, 661-668. GRAY, C. A., BURGHARDT, R. C., JOHNSON, G. A., BAZER, F. W. & SPENCER, T. E. 2002. Evidence that absence of endometrial gland secretions in uterine gland knockout ewes compromises conceptus survival and elongation. Reproduction, 124, 289-300 GRAY, C. A., TAYLOR, K. M., RAMSEY, W. S., HILL, J. R., BAZER, F. W., BARTOL, F. F. & SPENCER, T. E. 2001. Endometrial glands are required for preimplantation conceptus elongation and survival. Biology of Reproduction, 64, 1608-1613. GRIFFITHS, A. N., WATERMEYER, S. R. & KLENTZERIS, L. D. 2002. Fluid Within the Endometrial Cavity in an IVF Cycle—A Novel Approach to Its Management. Journal of Assisted Reproduction and Genetics, 19, 298-301. GUZELOGLU-KAYISLI, O., KAYISLI, U. A. & TAYLOR, H. S. 2009. The Role of Growth Factors and Cytokines during Implantation: Endocrine and Paracrine Interactions. Seminars in Reproductive Medicine, 27, 62-79. HACKETT, A. J., DURNFORD, R., MAPLETOFT, R. J. & MARCUS, G. J. 1993. Location and status of embryos in the genital tract of superovulated cows 4 to 6 days after insemination. Theriogenology, 40, 1147-1153.

152

Appendix HANNAN, N. J., NIE, G., RAINZCUK, A., ROMBAUTS, L. J. F. & SALAMONSEN, L. A. 2012. Uterine Lavage or Aspirate: Which View of the Intrauterine Environment? Reproductive Sciences, 19, 1125-1132. HANSEN, P. J. 2009. Effects of heat stress on mammalian reproduction. Philosophical Transactions of the Royal Society B: Biological Sciences, 364, 3341-3350. HANSEN, P. J. 2011. The immunology of early pregnancy in farm animals. Reproduction in Domestic Animals, 46 Suppl 3, 18-30. HARTMANN, D., ROHKOHL, J., MERBACH, S., HEILKENBRINKER, T., KLINDWORTH, H. P., SCHOON, H. A. & HOEDEMAKER, M. 2016. Prevalence of cervicitis in dairy cows and its effect on reproduction. Theriogenology, 85, 247-253. HAYASHI, K., HOSOE, M., KIZAKI, K., FUJII, S., KANAHARA, H., TAKAHASHI, T. & SAKUMOTO, R. 2017. Differential gene expression profiling of endometrium during the mid-luteal phase of the estrous cycle between a repeat breeder (RB) and non-RB cows. Reproductive Biology and Endocrinology : RB&E, 15, 20. HE, R.-H., GAO, H.-J., LI, Y.-Q. & ZHU, X.-M. 2010. The associated factors to endometrial cavity fluid and the relevant impact on the IVF-ET outcome. Reproductive Biology and Endocrinology : RB&E, 8, 46-46. HEALY, L. L., CRONIN, J. G. & SHELDON, I. M. 2014. Endometrial cells sense and react to tissue damage during infection of the bovine endometrium via interleukin 1. Scientific Reports, 4, 7060. HEPPELMANN, M., BRÖMMLING, A., ULBRICH, S. E., WEINERT, M., PIECHOTTA, M., WRENZYCKI, C., MERBACH, S., SCHOON, H. A., HOEDEMAKER, M. & BOLLWEIN, H. 2015. Effect of suppression of postpartum ovulation on endometrial inflammation in dairy cows. Theriogenology, 84, 155-162. HERATH, S., DOBSON, H., BRYANT, C. E. & SHELDON, I. M. 2006. Use of the cow as a large animal model of uterine infection and immunity. Journal of Reproductive Immunology, 69, 13-22. HERATH, S., LILLY, S. T., SANTOS, N. R., GILBERT, R. O., GOETZE, L., BRYANT, C. E., WHITE, J. O., CRONIN, J. & SHELDON, I. M. 2009. Expression of genes associated with immunity in the endometrium of cattle with disparate postpartum uterine disease and fertility. Reprod Biol Endocrinol, 7, 55. HILL, J. & GILBERT, R. 2008. Reduced quality of bovine embryos cultured in media conditioned by exposure to an inflamed endometrium. Australian Veterinary Journal, 86, 312-316. HOWARD, H. J. & BRITT, J. H. 1990. Prostaglandin F-2α causes regression of an hCG-induced corpus luteum before Day 5 of its lifespan in cattle. Journal of Reproduction and Fertility, 90, 245-253. HUANG, S. Y., WANG, C. J., SOONG, Y. K., WANG, H. S., WANG, M. L., LIN, C. Y. & CHANG, C. L. 2011. Site-specific endometrial injury improves implantation and pregnancy in patients with repeated implantation failures. Reproductive Biology and Endocrinology, 9, 140. HUMBLOT, P. 2001. Use of pregnancy specific proteins and progesterone assays to monitor pregnancy and determine the timing, frequencies and sources of embryonic mortality in ruminants. Theriogenology, 56, 1417-1433. IGWEBUIKE, U. M. 2006. Trophoblast cells of ruminant placentas--A minireview. Animal Reproduction Science, 93, 185-198. INSKEEP, E. K. & DAILEY, R. A. 2005. Embryonic Death in Cattle. Veterinary Clinics of North America: Food Animal Practice, 21, 437-461. JAUREGUIBERRY, M., GIULIODORI, M. J., MANG, A. V., MADOZ, L. V., POTHMANN, H., DRILLICH, M. & DE LA SOTA, R. L. 2017. Short communication: Repeat breeder cows with fluid in the uterine lumen had poorer fertility. Journal of Dairy Science, 100, 3083- 3085.

153

Appendix JUENGEL, J. L., HAWORTH, J. D., ROLLYSON, M. K., SILVA, P. J., SAWYER, H. R. & NISWENDER, G. D. 2000. Effect of Dose of Prostaglandin F2α on Steroidogenic Components and Oligonucleosomes in Ovine Luteal Tissue. Biology of Reproduction, 62, 1047-1051. KAITU'U-LINO, T. J., MORISON, N. B. & SALAMONSEN, L. A. 2007. Neutrophil depletion retards endometrial repair in a mouse model. Cell Tissue Res, 328, 197-206. KANNAKI, T. R., SHANMUGAM, M. & VERMA, P. C. 2011. Toll-like receptors and their role in animal reproduction. Animal Reproduction Science, 125, 1-12. KARIMZADEH, M. A., AYAZI ROZBAHANI, M. & TABIBNEJAD, N. 2009. Endometrial local injury improves the pregnancy rate among recurrent implantation failure patients undergoing in vitro fertilisation/intra cytoplasmic sperm injection: A randomised clinical trial. Australian and New Zealand Journal of Obstetrics and Gynaecology, 49, 677-680. KASIMANICKAM, R., DUFFIELD, T. F., FOSTER, R. A., GARTLEY, C. J., LESLIE, K. E., WALTON, J. S. & JOHNSON, W. H. 2004. Endometrial cytology and ultrasonography for the detection of subclinical endometritis in postpartum dairy cows. Theriogenology, 62, 9- 23. KASIMANICKAM, R., DUFFIELD, T. F., FOSTER, R. A., GARTLEY, C. J., LESLIE, K. E., WALTON, J. S. & JOHNSON, W. H. 2005a. A comparison of the cytobrush and uterine lavage techniques to evaluate endometrial cytology in clinically normal postpartum dairy cows. The Canadian Veterinary Journal, 46, 255-259. KASIMANICKAM, R., DUFFIELD, T. F., FOSTER, R. A., GARTLEY, C. J., LESLIE, K. E., WALTON, J. S. & JOHNSON, W. H. 2005b. A comparison of the cytobrush and uterine lavage techniques to evaluate endometrial cytology in clinically normal postpartum dairy cows. Canadian Veterinary Journal, 46, 255-259. KASIMANICKAM, R., DUFFIELD, T. F., FOSTER, R. A., GARTLEY, C. J., LESLIE, K. E., WALTON, J. S. & JOHNSON, W. H. 2005c. The effect of a single administration of cephapirin or cloprostenol on the reproductive performance of dairy cows with subclinical endometritis. Theriogenology, 63, 818-830. KASTELIC, J., BERGFELT, D. & GINTHER, O. 1991. Ultrasonic detection of the conceptus and characterization of intrauterine fluid on days 10 to 22 in heifers. Theriogenology, 35, 569- 581. KASTELIC, J. P., BERGFELT, D. R. & GINTHER, O. J. 1990. Relationship between ultrasonic assessment of the corpus luteum and plasma progesterone concentration in heifers. Theriogenology, 33, 1269-1278. KATAGIRI, S. & TAKAHASHI, Y. 2004. Changes in EGF concentrations during estrous cycle in bovine endometrium and their alterations in repeat breeder cows. Theriogenology, 62, 103- 112. KAUFMANN, T. B., DRILLICH, M., TENHAGEN, B. A., FORDERUNG, D. & HEUWIESER, W. 2009. Prevalence of bovine subclinical endometritis 4h after insemination and its effects on first service conception rate. Theriogenology, 71, 385-391. KENNEY, R. M. 1978. Cyclic and pathologic changes of the mare endometrium as detected by biopsy, with a note on early embryonic death. Journal of the American Veterinary Medical Association, 172, 241-262. KOBLISCHKE, P., KINDAHL, H., BUDIK, S., AURICH, J., PALM, F., WALTER, I., KOLODZIEJEK, J., NOWOTNY, N., HOPPEN, H. O. & AURICH, C. 2008. Embryo transfer induces a subclinical endometritis in recipient mares which can be prevented by treatment with non-steroid anti-inflammatory drugs. Theriogenology, 70, 1147-1158. KONO, H., ONDA, A. & YANAGIDA, T. 2014. Molecular determinants of sterile inflammation. Current Opinion in Immunology, 26, 147-156. KOTWICA, J., SKARZYNSKI, D., JAROSZEWSKI, J., WILLIAMS, G., L. & BOGACKI, M. 1998. Uterine secretion of prostaglandin F2α stimulated by different doses of oxytocin and

154

Appendix released spontaneously during luteolysis in cattle. Reproduction Nutrition Development, 38, 217-226. LAMBERTZ, C., VÖLKER, D., JANOWITZ, U. & GAULY, M. 2014. Evaluation of vaginal discharge with the Metricheck device and the relationship to reproductive performance in postpartum dairy cows. Animal Science Journal, 85, 848-852. LARSON, S. F., BUTLER, W. R. & CURRIE, W. B. 1997. Reduced Fertility Associated with Low Progesterone Postbreeding and Increased Milk Urea Nitrogen in Lactating Cows. Journal of Dairy Science, 80, 1288-1295. LEBLANC, S. J., DUFFIELD, T. F., LESLIE, K. E., BATEMAN, K. G., KEEFE, G. P., WALTON, J. S. & JOHNSON, W. H. 2002. Defining and diagnosing postpartum clinical endometritis and its impact on reproductive performance in dairy cows. Journal of Dairy Science, 85, 2223-2236. LEDGARD, A. M., LEE, R. S. F. & PETERSON, A. J. 2009. Bovine endometrial legumain and TIMP-2 regulation in response to presence of a conceptus. Molecular Reproduction and Development, 76, 65-74. LEE, J., MCCRACKEN, J. A., BANU, S. K., RODRIGUEZ, R., NITHY, T. K. & AROSH, J. A. 2010. Transport of prostaglandin F(2alpha) pulses from the uterus to the ovary at the time of luteolysis in ruminants is regulated by prostaglandin transporter-mediated mechanisms. Endocrinology, 151, 3326-35. LEESE, H. J. 2003. What does an embryo need? Human Fertility, 6, 180-185. LEUTERT, C., SUTHAR, V. & HEUWIESER, W. 2013. Evaluation of transrectal examination of cervical diameter by palpation in dairy cows. Journal of Dairy Science, 96, 1063-1070. LEUTERT, C., VON KRUEGER, X., PLONTZKE, J. & HEUWIESER, W. 2012. Evaluation of vaginoscopy for the diagnosis of clinical endometritis in dairy cows. Journal of Dairy Science, 95, 206-212. LI, J., CHEN, J. & KIRSNER, R. 2007. Pathophysiology of acute wound healing. Clinics in Dermatology, 25, 9-18. LINTON, J. K. & SERTICH, P. L. 2016. The impact of low-volume uterine lavage on endometrial biopsy classification. Theriogenology, 86, 1004-1007. LONERGAN, P., FAIR, T., FORDE, N. & RIZOS, D. 2016. Embryo development in dairy cattle. Theriogenology, 86, 270-277. LOPES, L. M. J., BALBINOT, M., FONSECA, B. A., DE ARAÚJO, C. V. & MARTINS, L. R. 2015. Pregnancy rates and serum 13,14-dihydro-15-keto-PGF2α concentrations in recipient Nelore heifers treated with meloxicam after the transfer of in vitro–produced embryos. Theriogenology, 84, 553-558. LOPEZ-HELGUERA, I., LOPEZ-GATIUS, F. & GARCIA-ISPIERTO, I. 2012. The influence of genital tract status in postpartum period on the subsequent reproductive performance in high producing dairy cows. Theriogenology, 77, 1334-1342. LU, S., PENG, H., ZHANG, H., ZHANG, L., CAO, Q., LI, R., ZHANG, Y., YAN, L., DUAN, E. & QIAO, J. 2013. Excessive intrauterine fluid cause aberrant implantation and pregnancy outcome in mice. PLOS ONE, 8, e78446. LUCY, M. C. 2001. Reproductive Loss in High-Producing Dairy Cattle: Where Will It End? Journal of Dairy Science, 84, 1277-1293. LUKENS, J. R., GROSS, J. M. & KANNEGANTI, T.-D. 2012. IL-1 family cytokines trigger sterile inflammatory disease. Frontiers in Immunology, 3, 315. MACHADO, V. S., KNAUER, W. A., BICALHO, M. L. S., OIKONOMOU, G., GILBERT, R. O. & BICALHO, R. C. 2012. A novel diagnostic technique to determine uterine health of Holstein cows at 35 days postpartum. Journal of Dairy Science, 95, 1349-1357. MADOZ, L. V., GIULIODORI, M. J., JAUREGUIBERRY, M., PLÖNTZKE, J., DRILLICH, M. & DE LA SOTA, R. L. 2013. The relationship between endometrial cytology during estrous cycle and cutoff points for the diagnosis of subclinical endometritis in grazing dairy cows. Journal of Dairy Science, 96, 4333-4339. 155

Appendix MADOZ, L. V., GIULIODORI, M. J., MIGLIORISI, A. L., JAUREGUIBERRY, M. & DE LA SOTA, R. L. 2014. Endometrial cytology, biopsy, and bacteriology for the diagnosis of subclinical endometritis in grazing dairy cows. Journal of Dairy Science, 97, 195-201. MAHAJAN, N. 2015. Endometrial receptivity array: Clinical application. Journal of Human Reproductive Sciences, 8, 121-129. MAISCHBERGER, E., IRWIN, J. A., CARRINGTON, S. D. & DUGGAN, V. E. 2008. Equine post-breeding endometritis: A review. Irish Veterinary Journal, 61, 163-168. MANN, G. E. & LAMMING, G. E. 1994. Use of repeated biopsies to monitor endometrial oxytocin receptors in the cow. The Veterinary Record, 135, 403-405. MANN, G. E. & LAMMING, G. E. 2001. Relationship between maternal endocrine environment, early embryo development and inhibition of the luteolytic mechanism in cows. Reproduction, 121, 175-80. MANN, G. E. & LAMMING, G. E. 2006. Timing of prostaglandin F2α release episodes and oxytocin receptor development during luteolysis in the cow. Animal Reproduction Science, 93, 328-336. MARTAL, J., CHÊNE, N., CAMOUS, S., HUYNH, L., LANTIER, F., HERMIER, P., L'HARIDON, R., CHARPIGNY, G., CHARLIER, M. & CHAOUAT, G. 1997. Recent developments and potentialities for reducing embryo mortality in ruminants: the role of IFNτ and other cytokines in early pregnancy. Reproduction, Fertility and Development, 9, 355-380. MARTIN, I., MEMBRIVE, C. M. B., VETTORATO, L. F., BINELLI, M., OBA, E. & FERREIRA, J. C. P. 2015. Plasma PGFM and progesterone concentrations, luteolysis moment and estrous cycle length in Nelore cows submitted to uterine biopsies. Arquivo Brasileiro de Medicina Veterinária e Zootecnia, 67, 1210-1216. MARTIN, I., TORRES NETO, R., OBA, E., BURATINI JR, J., BINELLI, M., LAUFER- AMORIM, R. & FERREIRA, J. C. P. 2008. Immunohistochemical detection of receptors for oestrogen and progesterone in endometrial glands and stroma during the oestrous cycle in Nelore (Bos taurus indicus) cows. Reproduction in Domestic Animals, 43, 415-421. MATEUS, L., LOPES DA COSTA, L., BERNARDO, F. & ROBALO SILVA, J. 2002. Influence of Puerperal Uterine Infection on Uterine Involution and Postpartum Ovarian Activity in Dairy Cows. Reproduction in Domestic Animals, 37, 31-35. MATEUS, L., LOPES DA COSTA, L., DINIZ, P. & ZIECIK, A. J. 2003. Relationship between endotoxin and prostaglandin (PGE2 and PGFM) concentrations and ovarian function in dairy cows with puerperal endometritis. Animal Reproduction Science, 76, 143-154. MCDOUGALL, S. 2005. Gross abnormalities, bacteriology and histological lesions of uteri of dairy cows failing to conceive or maintain pregnancy. New Zealand Veterinary Journal, 53, 253- 256. MCDOUGALL, S., MACAULAY, R. & COMPTON, C. 2007. Association between endometritis diagnosis using a novel intravaginal device and reproductive performance in dairy cattle. Animal Reproduction Science, 99, 9-23. MCENTEE, K. 1990. Chapter 7 - The Uterus: Anatomy. In: MCENTEE, K. (ed.) Reproductive Pathology of Domestic Mammals. San Diego: Academic Press. MCNEILL, R. E., DISKIN, M. G., SREENAN, J. M. & MORRIS, D. G. 2006. Associations between milk progesterone concentration on different days and with embryo survival during the early luteal phase in dairy cows. Theriogenology, 65, 1435-1441. MEIKLE, A., CAVESTANY, D., SAHLIN, L., THATCHER, W., GAROFALO, E., KINDAHL, H. & FORSBERG, M. 2005. Effect of Pregnancy on Endometrial Sex Steroid Receptors and on Prostaglandin F2α Release after Uterine Biopsy in Heifers. In: MAKKAR, H. S. & VILJOEN, G. (eds.) Applications of Gene-Based Technologies for Improving Animal Production and Health in Developing Countries. Springer Netherlands.

156

Appendix MEIRA, E. B., JR., HENRIQUES, L. C., SA, L. R. & GREGORY, L. 2012. Comparison of ultrasonography and histopathology for the diagnosis of endometritis in Holstein-Friesian cows. Journal of Dairy Science, 95, 6969-6973. MESQUITA, F. S., PUGLIESI, G., SCOLARI, S. C., FRANÇA, M. R., RAMOS, R. S., OLIVEIRA, M., PAPA, P. C., BRESSAN, F. F., MEIRELLES, F. V., SILVA, L. A., NOGUEIRA, G. P., MEMBRIVE, C. M. B. & BINELLI, M. 2014. Manipulation of the periovulatory sex steroidal milieu affects endometrial but not luteal gene expression in early diestrus Nelore cows. Theriogenology, 81, 861-869. MESSIER, S., HIGGINS, R., COUTURE, Y. & MORIN, M. 1984. Comparison of swabbing and biopsy for studying the flora of the bovine uterus. The Canadian Veterinary Journal, 25, 283-288. MIRAVET-VALENCIANO, J. A., RINCON-BERTOLIN, A., VILELLA, F. & SIMON, C. 2015. Understanding and improving endometrial receptivity. Current Opinion in Obstetrics & Gynecoly, 27, 187-192. MITKO, K., ULBRICH, S. E. & WENIGERKIND, H. 2008. Dynamic changes in messenger RNA profiles of bovine endometrium during the oestrous cycle: Focus on mammalian embryogenomics. Reproduction, 135, 225-240. MIYAMOTO, A., SHIRASUNA, K. & SASAHARA, K. 2009. Local regulation of corpus luteum development and regression in the cow: Impact of angiogenic and vasoactive factors. Domestic Animal Endocrinology, 37, 159-169. MIYAMOTO, A., SHIRASUNA, K., SHIMIZU, T., BOLLWEIN, H. & SCHAMS, D. 2010. Regulation of corpus luteum development and maintenance: specific roles of angiogenesis and action of prostaglandin F2alpha. Society of Reproduction and Fertility Supplement, 67, 289-304. MIYAMOTO, A., SHIRASUNA, K., WIJAYAGUNAWARDANE, M. P. B., WATANABE, S., HAYASHI, M., YAMAMOTO, D., MATSUI, M. & ACOSTA, T. J. 2005. Blood flow: A key regulatory component of corpus luteum function in the cow. Domestic Animal Endocrinology, 29, 329-339. MOROTTI, F., SANCHES, B. V., PONTES, J. H., BASSO, A. C., SIQUEIRA, E. R., LISBOA, L. A. & SENEDA, M. M. 2014. Pregnancy rate and birth rate of calves from a large-scale IVF program using reverse-sorted semen in Bos indicus, Bos indicus-taurus, and Bos taurus cattle. Theriogenology, 81, 696-701. MULLEN, M. P., ELIA, G., HILLIARD, M., PARR, M. H., DISKIN, M. G., EVANS, A. C. O. & CROWE, M. A. 2012. Proteomic characterization of histotroph during the preimplantation phase of the estrous cycle in cattle. Journal of Proteome Research, 11, 3004-3018. MULLINS, K. J. & SAACKE, R. G. 1989. Study of the functional anatomy of bovine cervical mucosa with special reference to mucus secretion and sperm transport. The Anatomical Record, 225, 106-117. MUNOZ, M., CORRALES, F. J., CAAMANO, J. N., DIEZ, C., TRIGAL, B., MORA, M. I., MARTIN, D., CARROCERA, S. & GOMEZ, E. 2012. Proteome of the early embryo- maternal dialogue in the cattle uterus. Journal of Proteome Research, 11, 751-766. MURUGAVEL, K., YÁNIZ, J., SANTOLARIA, P., LÓPEZ-BÉJAR, M. & LÓPEZ-GATIUS, F. 2003. Prostaglandin based estrus synchronization in postpartum dairy cows: an update. JOURNAL OF APPLIED RESEARCH IN VETERINARY MEDICINE, 1, 51-65. NEWTON, K. & DIXIT, V. M. 2012. Signaling in Innate Immunity and Inflammation. Cold Spring Harbor Perspectives in Biology, 4, a006049. O’HARA, L., FORDE, N., CARTER, F., RIZOS, D., MAILLO, V., EALY, A. D., KELLY, A. K., RODRIGUEZ, P., ISAKA, N., EVANS, A. C. O. & LONERGAN, P. 2014. Paradoxical effect of supplementary progesterone between Day 3 and Day 7 on corpus luteum function and conceptus development in cattle. Reproduction, Fertility and Development, 26, 328-336. OHTANI, S., OKUDA, K., NISHIMURA, K. & MOHRI, S. 1993. Histological changes in bovine endometrium during the estrous cycle. Theriogenology, 39, 1033-1042. 157

Appendix OKUDA, K., MIYAMOTO, Y. & SKARZYNSKI, D. J. 2002. Regulation of endometrial prostaglandin F2α synthesis during luteolysis and early pregnancy in cattle. Domestic Animal Endocrinology, 23, 255-264. OLIVEIRA, B. M., ARRUDA, R. P., THOME, H. E., MATURANA FILHO, M., OLIVEIRA, G., GUIMARAES, C., NICHI, M., SILVA, L. A. & CELEGHINI, E. C. 2014. Fertility and uterine hemodynamic in cows after artificial insemination with semen assessed by fluorescent probes. Theriogenology, 82, 767-772. OVERBECK, W., WITTE, T. S. & HEUWIESER, W. 2011. Comparison of three diagnostic methods to identify subclinical endometritis in mares. Theriogenology, 75, 1311-1318. PASCOTTINI, O. B., DINI, P., HOSTENS, M., DUCATELLE, R. & OPSOMER, G. 2015. A novel cytologic sampling technique to diagnose subclinical endometritis and comparison of staining methods for endometrial cytology samples in dairy cows. Theriogenology, 84, 1438-1446. PAULA-LOPES, F. F., DE MORAES, A. A. & HANSEN, P. J. 1999. The presence of interleukin- 1beta in the bovine reproductive tract. Journal of Interferon & Cytokine Research, 19, 279- 285. PETER, A. T., LEVINE, H., DROST, M. & BERGFELT, D. R. 2009. Compilation of classical and contemporary terminology used to describe morphological aspects of ovarian dynamics in cattle. Theriogenology, 71, 1343-1357. PIERSON, R. A. & GINTHER, O. J. 1987. Ultrasonographic appearance of the bovine uterus during the estrous cycle. Journal of the American Veterinary Medical Association, 190, 995- 1001. PIERSON, R. A. & GINTHER, O. J. 1988. Ultrasonic imaging of the ovaries and uterus in cattle. Theriogenology, 29, 21-37. PLONTZKE, J., MADOZ, L. V., DE LA SOTA, R. L., DRILLICH, M. & HEUWIESER, W. 2010. Subclinical endometritis and its impact on reproductive performance in grazing dairy cattle in Argentina. Animal Reproduction Science, 122, 52-57. POLAT, B., CENGIZ, M., CANNAZIK, O., COLAK, A., ORUC, E., ALTUN, S., SALAR, S. & BASTAN, A. 2015. Endometrial echotexture variables in postpartum cows with subclinical endometritis. Animal Reproduction Science, 155, 50-55. PONTES, J. H. F., NONATO-JUNIOR, I., SANCHES, B. V., ERENO-JUNIOR, J. C., UVO, S., BARREIROS, T. R. R., OLIVEIRA, J. A., HASLER, J. F. & SENEDA, M. M. 2009. Comparison of embryo yield and pregnancy rate between in vivo and in vitro methods in the same Nelore (Bos indicus) donor cows. Theriogenology, 71, 690-697. POTDAR, N., GELBAYA, T. & NARDO, L. G. 2012. Endometrial injury to overcome recurrent embryo implantation failure: a systematic review and meta-analysis. Reproductive BioMedicine Online, 25, 561-571. PRINS, J. R., GOMEZ-LOPEZ, N. & ROBERTSON, S. A. 2012. Interleukin-6 in pregnancy and gestational disorders. Journal of Reproductive Immunology, 95, 1-14. PUGLIESI, G., SCOLARI, S. C., MESQUITA, F. S., MATURANA FILHO, M., ARAUJO, E. R., CARDOSO, D., SALES, J. N., MARTIN, I., SA FILHO, M., BERTAN, C. M. & BINELLI, M. 2014. Impact of probing the reproductive tract during early pregnancy on fertility of beef cows. Reproduction in Domestic Animals, 49, e35-9. RAMIREZ-GARZON, O., SATAKE, N., LYONS, R. E., HILL, J., HOLLAND, M. K. & MCGOWAN, M. 2017. Endometrial biopsy in Bos indicus beef heifers. Reproduction in Domestic Animals, 1-3 doi.org/10.1111/rda.12944. RHOADS, M. L., MEYER, J. P., LAMBERSON, W. R., KEISLER, D. H. & LUCY, M. C. 2008. Uterine and hepatic gene expression in relation to days postpartum, estrus, and pregnancy in postpartum dairy cows. Journal of Dairy Science, 91, 140-150. RIBEIRO, E. S., SANTOS, J. E. P. & THATCHER, W. W. 2016. Role of lipids on elongation of the preimplantation conceptus in ruminants. Reproduction, 152, R115-R126.

158

Appendix RICCIOTTI, E. & FITZGERALD, G. 2011. Prostaglandins and Inflammation. Arteriosclerosis, thrombosis, and vascular biology, 31, 986-1000. RICKETTS, S. W. 1975. The technique and clinical application of endometrial biopsy in the mare. Equine Veterinary Journal, 7, 102-108. ROBERTS, M., CHEN, Y., EZASHI, T. & WALKER, A. 2008. Interferons and the maternal- conceptus dialog in mammals. Seminars in Cell & Developmental Biology, 19, 170 - 177. ROBERTSON, S. A. 2007. GM-CSF regulation of embryo development and pregnancy. Cytokine & Growth Factor Reviews, 18, 287-298. ROBINSON, R. S., HAMMOND, A. J., WATHES, D. C., HUNTER, M. G. & MANN, G. E. 2008. Corpus Luteum–Endometrium–Embryo Interactions in the Dairy Cow: Underlying Mechanisms and Clinical Relevance. Reproduction in Domestic Animals, 43, 104-112. ROBINSON, R. S., MANN, G. E., LAMMING, G. E. & WATHES, D. C. 2001. Expression of oxytocin, oestrogen and progesterone receptors in uterine biopsy samples throughout the oestrous cycle and early pregnancy in cows. Reproduction, 122, 965-79. ROCK, K. L., LATZ, E., ONTIVEROS, F. & KONO, H. 2010. The sterile inflammatory response. Annual review of immunology, 28, 321-342. ROELOFS, J., LOPEZ-GATIUS, F., HUNTER, R. H., VAN EERDENBURG, F. J. & HANZEN, C. 2010. When is a cow in estrus? Clinical and practical aspects. Theriogenology, 74, 327- 344. RUAN, Y. C., CHEN, H. & CHAN, H. C. 2014. Ion channels in the endometrium: regulation of endometrial receptivity and embryo implantation. Human Reproduction Update, 20, 517- 529. RUIZ-ALONSO, M., BLESA, D., DÍAZ-GIMENO, P., GÓMEZ, E., FERNÁNDEZ-SÁNCHEZ, M., CARRANZA, F., CARRERA, J., VILELLA, F., PELLICER, A. & SIMÓN, C. 2013. The endometrial receptivity array for diagnosis and personalized embryo transfer as a treatment for patients with repeated implantation failure. Fertility and Sterility, 100, 818- 824. SALAMONSEN, L. A. 2003. Tissue injury and repair in the female human reproductive tract. Reproduction, 125, 301-311 SALAMONSEN, L. A., EDGELL, T., ROMBAUTS, L. J., STEPHENS, A. N., ROBERTSON, D. M., RAINCZUK, A., NIE, G. & HANNAN, N. J. 2013. Proteomics of the human endometrium and uterine fluid: a pathway to biomarker discovery. Fertility and sterility, 99, 1086-1092. SALILEW-WONDIM, D., HOLKER, M., RINGS, F., GHANEM, N., ULAS-CINAR, M., PEIPPO, J., THOLEN, E., LOOFT, C., SCHELLANDER, K. & TESFAYE, D. 2010. Bovine pretransfer endometrium and embryo transcriptome fingerprints as predictors of pregnancy success after embryo transfer. Physiological Genomics, 42, 201-218. SALILEW-WONDIM, D., SCHELLANDER, K., HOELKER, M. & TESFAYE, D. 2012. Oviductal, endometrial and embryonic gene expression patterns as molecular clues for pregnancy establishment. Animal Reproduction Science, 134, 9-18. SAMUELSON, D. A. 2007. Textbook of veterinary histology, St. Louis, Mo, Saunders Elsevier. SANTOS, J. E. P., THATCHER, W. W., CHEBEL, R. C., CERRI, R. L. A. & GALVÃO, K. N. 2004. The effect of embryonic death rates in cattle on the efficacy of estrus synchronization programs. Animal Reproduction Science, 82–83, 513-535. SARTORI, R. & BARROS, C. M. 2011. Reproductive cycles in Bos indicus cattle. Animal Reproduction Science, 124, 244-250. SARTORI, R., HAUGHIAN, J. M., SHAVER, R. D., ROSA, G. J. & WILTBANK, M. C. 2004. Comparison of ovarian function and circulating steroids in estrous cycles of Holstein heifers and lactating cows. Journal of Dairy Science, 87, 905-920. SARTORI, R., SARTOR-BERGFELT, R., MERTENS, S. A., GUENTHER, J. N., PARRISH, J. J. & WILTBANK, M. C. 2002. Fertilization and Early Embryonic Development in Heifers and

159

Appendix Lactating Cows in Summer and Lactating and Dry Cows in Winter. Journal of Dairy Science, 85, 2803-2812. ŠAVC, M., DUANE, M., O'GRADY, L. E., SOMERS, J. R. & BELTMAN, M. E. 2016. Uterine disease and its effect on subsequent reproductive performance of dairy cattle: a comparison of two cow-side diagnostic methods. Theriogenology, 86, 1983-1988. SAVIO, J. D., BOLAND, M. P. & ROCHE, J. F. 1990. Development of dominant follicles and length of ovarian cycles in post-partum dairy cows. Journal of Reproduction and Fertility, 88, 581-591. SCENNA, F. N., EDWARDS, J. L., ROHRBACH, N. R., HOCKETT, M. E., SAXTON, A. M. & SCHRICK, F. N. 2004. Detrimental effects of prostaglandin F2α on preimplantation bovine embryos. Prostaglandins & Other Lipid Mediators, 73, 215-226. SCENNA, F. N., HOCKETT, M. E., TOWNS, T. M., SAXTON, A. M., ROHRBACH, N. R., WEHRMAN, M. E. & SCHRICK, F. N. 2005. Influence of a prostaglandin synthesis inhibitor administered at embryo transfer on pregnancy rates of recipient cows. Prostaglandins& Other Lipid Mediators, 78, 38-45. SCHAMS, D. & BERISHA, B. 2004. Regulation of corpus luteum function in cattle--an overview. Reproduction in Domestic Animals, 39, 241-251. SCHMAUDER, S., WEBER, F., KIOSSIS, E. & BOLLWEIN, H. 2008. Cyclic changes in endometrial echotexture of cows using a computer-assisted program for the analysis of first- and second-order grey level statistics of B-Mode ultrasound images. Animal Reproduction Science, 106, 153-61. SCHNEIDER, C. P., SCHWACHA, M. G. & CHAUDRY, I. H. 2004. The role of interleukin-10 in the regulation of the systemic inflammatory response following trauma-hemorrhage. Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease, 1689, 22-32. SCHWARZ, K. R. L., PIRES, P. R. L., ADONA, P. R., CÂMARA DE BEM, T. H. & LEAL, C. L. V. 2008. Influence of nitric oxide during maturation on bovine oocyte meiosis and embryo development in vitro. Reproduction, Fertility and Development, 20, 529-536. SCULLY, S., EVANS, A. C. O., CARTER, F., DUFFY, P., LONERGAN, P. & CROWE, M. A. 2015. Ultrasound monitoring of blood flow and echotexture of the corpus luteum and uterus during early pregnancy of beef heifers. Theriogenology, 83, 449-458. SENS, A. & HEUWIESER, W. 2013. Presence of Escherichia coli, Trueperella pyogenes, alpha- hemolytic streptococci, and coagulase-negative staphylococci and prevalence of subclinical endometritis. Journal of Dairy Science, 96, 6347-6354. SHAHAM-ALBALANCY, A., NYSKA, A., KAIM, M., ROSENBERG, M., FOLMAN, Y. & WOLFENSON, D. 1997. Delayed effect of progesterone on endometrial morphology in dairy cows. Animal Reproduction Science, 48, 159-174. SHARARA, F. I., SCOTT, J. R. T., MARUT, E. L. & QUEENAN, J. J. T. 1996. In-vitro fertilization outcome in women with hydrosalpinx*. Human Reproduction, 11, 526-530. SHELDON, I. M. & DOBSON, H. 2004. Postpartum uterine health in cattle. Animal Reproduction Science, 82–83, 295-306. SHELDON, I. M., LEWIS, G. S., LEBLANC, S. & GILBERT, R. O. 2006. Defining postpartum uterine disease in cattle. Theriogenology, 65, 1516-1530. SHELDON, I. M., PRICE, S. B., CRONIN, J., GILBERT, R. O. & GADSBY, J. E. 2009. Mechanisms of infertility associated with clinical and subclinical endometritis in high producing dairy cattle. Reproduction in Domestic Animals, 44 Suppl 3, 1-9. SHELDON, I. M., WILLIAMS, E. J., MILLER, A. N. A., NASH, D. M. & HERATH, S. 2008. Uterine diseases in cattle after parturition. Veterinary Journal (London, England : 1997), 176, 115-121. SHELDON, P. E. & COPPENGER, C. J. 1977. A rapid radioimmunoassay for serum testosterone. Steroids, 30, 149-57. SHEN, H., KREISEL, D. & GOLDSTEIN, D. R. 2013. Processes of Sterile Inflammation. Journal of Immunology, 191, 2857-2863. 160

Appendix SINGH, J., MURRAY, R. D., MSHELIA, G. & WOLDEHIWET, Z. 2008. The immune status of the bovine uterus during the peripartum period. The Veterinary Journal, 175, 301-309. SINGH, R. B., SHARMA, R. D., SINGH, N. & SINGH, G. B. 1983. Bio-histopathological studies of endometrium in repeat breeding buffaloes (Bubalus bubalis). Theriogenology, 19, 151- 157. SIQUEIRA, L. G. B., TORRES, C. A. A., AMORIM, L. S., SOUZA, E. D., CAMARGO, L. S. A., FERNANDES, C. A. C. & VIANA, J. H. M. 2009. Interrelationships among morphology, echotexture, and function of the bovine corpus luteum during the estrous cycle. Animal Reproduction Science, 115, 18-28. SKARZYNSKI, D. J., BAH, M. M., DEPTULA, K. M., WOCLAWEK-POTOCKA, I., KORZEKWA, A., SHIBAYA, M., PILAWSKI, W. & OKUDA, K. 2003. Roles of tumor necrosis factor-alpha of the estrous cycle in cattle: an in vivo study. Biology of Reproduction, 69, 1907-1913. SKJERVEN, O. 1956. Endometrial biopsy studies in reproductively normal cattle; clinical, histochemical and histological observations during the estrous cycle. Acta Endocrinolica Supplementum 22, 1-101. SNIDER, T. A., SEPOY, C. & HOLYOAK, G. R. 2011. Equine endometrial biopsy reviewed: Observation, interpretation, and application of histopathologic data. Theriogenology, 75, 1567-1581. SOTO, P., NATZKE, R. P. & HANSEN, P. J. 2003. Identification of possible mediators of embryonic mortality caused by mastitis: actions of lipopolysaccharide, prostaglandin F2alpha, and the nitric oxide generator, sodium nitroprusside dihydrate, on oocyte maturation and embryonic development in cattle. American Journal of Reproductive Immunology, 50, 263-272. SPENCER, T. E. 2013. Early pregnancy: Concepts, challenges, and potential solutions. Animal Frontiers, 3, 48-55. SPENCER, T. E., BURGHARDT, R. C., JOHNSON, G. A. & BAZER, F. W. 2004a. Conceptus signals for establishment and maintenance of pregnancy. Animal Reproduction Science, 82- 83, 537-550. SPENCER, T. E., DUNLAP, K. A. & FILANT, J. 2012. Comparative developmental biology of the uterus: Insights into mechanisms and developmental disruption. Molecular and Cellular Endocrinology, 354, 34-53. SPENCER, T. E., JOHNSON, G. A., BAZER, F. W. & BURGHARDT, R. C. 2004b. Implantation mechanisms: insights from the sheep. Reproduction, 128, 657-668. STRONGE, A. J. H., SREENAN, J. M., DISKIN, M. G., MEE, J. F., KENNY, D. A. & MORRIS, D. G. 2005. Post-insemination milk progesterone concentration and embryo survival in dairy cows. Theriogenology, 64, 1212-1224. STUDER, E. & MORROW, D. A. 1978. Postpartum evaluation of bovine reproductive potential: comparison of findings from genital tract examination per rectum, uterine culture, and endometrial biopsy. Journal of the American Veterinary Medical Association, 172, 489-494. SUBANDRIO, A. & NOAKES, D. 1997. Neutrophil migration into the uterine lumen of the cow: the influence of endogenous and exogenous sex steroid hormones using two intrauterine chemoattractants. Theriogenology, 47, 825-35. SUGAWARA, K., KIZAKI, K., HERATH, C., HASEGAWA, Y. & HASHIZUME, K. 2010. Transforming growth factor beta family expression at the bovine feto-maternal interface. Reproductive Biology Endocrinology, 8, 120. SUMMERS, P. 1974. An abattoir study of the genital pathology of cows in Northern Australia. Australian Veterinary Journal, 50, 403-406. SURESH, K. P. 2011. An overview of randomization techniques: An unbiased assessment of outcome in clinical research. Journal of Human Reproductive Sciences, 4, 8-11. TASAKI, Y., NISHIMURA, R., SHIBAYA, M., LEE, H.-Y., ACOSTA, T. J. & OKUDA, K. 2010. Expression of VEGF and its receptors in the bovine endometrium throughout the estrous 161

Appendix cycle: effects of VEGF on prostaglandin production in endometrial cells. Journal of Reproduction and Development, 56, 223-229. THATCHER, W. W., TERQUI, M., THIMONIER, J. & MAULEON, P. 1986. Effect of estradiol- 17β on peripheral plasma concentration of 15-keto,14-dihydro PGF2α and luteolysis in cyclic cattle. Prostaglandins, 31, 745-756. THOME, H. E., DE ARRUDA, R. P., DE OLIVEIRA, B. M., MATURANA FILHO, M., DE OLIVEIRA, G. C., GUIMARAES CDE, F., DE CARVALHO BALIEIRO, J. C., AZEDO, M. R., POGLIANI, F. C. & CELEGHINI, E. C. 2016. Uterine lavage is efficient to recover endometrial cytology sample and does not interfere with fertility rate after artificial insemination in cows. Theriogenology, 85, 1549-1554. THOMPSON, M. R., KAMINSKI, J. J., KURT-JONES, E. A. & FITZGERALD, K. A. 2011. Pattern Recognition Receptors and the Innate Immune Response to Viral Infection. Viruses, 3, 920-940. TIZARD, I. R. 2012. Veterinary immunology, Elsevier Health Sciences. TROEDSSON, M. H. 1999. Uterine clearance and resistance to persistent endometritis in the mare. Theriogenology, 52, 461-471. TSAI, S. J. & WILTBANK, M. C. 1998. Prostaglandin F2alpha regulates distinct physiological changes in early and mid-cycle bovine corpora lutea. Biology of Reproduction, 58, 346-352. ULBRICH, S. E., GROEBNER, A. E. & BAUERSACHS, S. 2013. Transcriptional profiling to address molecular determinants of endometrial receptivity – Lessons from studies in livestock species. Methods, 59, 108-115. UNTERGASSER, A., NIJVEEN, H., RAO, X., BISSELING, T., GEURTS, R. & LEUNISSEN, J. A. M. 2007. Primer3Plus, an enhanced web interface to Primer3. Nucleic Acids Research, 35, W71-W74. VAN DER GAAST, M. H., BEIER-HELLWIG, K., FAUSER, B. C., BEIER, H. M. & MACKLON, N. S. 2003. Endometrial secretion aspiration prior to embryo transfer does not reduce implantation rates. Reprod Biomed Online, 7, 105-9. VELAZQUEZ, M. A., PARRILLA, I., VAN SOOM, A., VERBERCKMOES, S., KUES, W. & NIEMANN, H. 2010. Sampling techniques for oviductal and uterine luminal fluid in cattle. Theriogenology, 73, 758-767. VELEZ, J. S., RANDEL, R. D. & NEUENDORFF, D. A. 1991. Effect of uterine manipulation on postpartum fertility and plasma 13,14-dihydro-15-keto-prostaglandin F2α in Brahman cows and first-calf heifers. Theriogenology, 36, 987-998. VERBERCKMOES, S., VAN SOOM, A., DE PAUW, I., DEWULF, J., VERVAET, C. & DE KRUIF, A. 2004. Assessment of a new utero-tubal junction insemination device in dairy cattle. Theriogenology, 61, 103-115. WALLACE, L. D., BREINER, C. A., BREINER, R. A., SPELL, A. R., CARTER, J. A., LAMB, G. C. & STEVENSON, J. S. 2011. Administration of human chorionic gonadotropin at embryo transfer induced ovulation of a first wave dominant follicle, and increased progesterone and transfer pregnancy rates. Theriogenology, 75, 1506-1515. WALSH, S. W., WILLIAMS, E. J. & EVANS, A. C. O. 2011. A review of the causes of poor fertility in high milk producing dairy cows. Animal Reproduction Science, 123, 127-138. WANG, C. K., ROBINSON, R. S., FLINT, A. P. F. & MANN, G. E. 2007. Quantitative analysis of changes in endometrial gland morphology during the bovine oestrous cycle and their association with progesterone levels. Reproduction, 134, 365-371. WANN, R. A. & RANDEL, R. D. 1990. Effect of uterine manipulation 35 days after parturition on plasma concentrations of 13, 14-dihydro-15-keto prostaglandin F2 alpha in multiparous and primiparous Brahman cows. Journal of Animal Science, 68, 1389-1394. WARD, P. A., SERHAN, C. N., AYOUB, S. S. & GILROY, D. W. 2010. Fundamentals of Inflammation, New York, UNITED STATES, Cambridge University Press.

162

Appendix WATSON, E. & SERTICH, P. 1992. Effect of repeated collection of multiple endometrial biopsy specimens on subsequent pregnancy in mares. Journal of the American Veterinary Medical Association, 201, 438-440. WATSON, E., STOKES, C., DAVID, J., BOURNE, F. & RICKETTS, S. 1987. Concentrations of uterine luminal prostaglandins in mares with acute and persistent endometritis. Equine veterinary journal, 19, 31-37. WATSON, E. D. 2000. Post-breeding endometritis in the mare. Animal Reproduction Science, 60– 61, 221-232. WATTS, T. L. & FUQUAY, J. W. 1985. Response and fertility of dairy heifers following injection with prostaglandin F2α during early, middle or late diestrus. Theriogenology, 23, 655-661. WEEMS, C. W., WEEMS, Y. S. & RANDEL, R. D. 2006. Prostaglandins and reproduction in female farm animals. The Veterinary Journal, 171, 206-228. WEHREND, A., FAILING, K. & BOSTEDT, H. 2003. Cervimetry and ultrasonographic observations of the cervix regression in dairy cows during the first 10 days post partum. J Vet Med A Physiol Pathol Clin Med, 50, 470-3. WEISS, G., GOLDSMITH, L. T., TAYLOR, R. N., BELLET, D. & TAYLOR, H. S. 2009. Inflammation in Reproductive Disorders. Reproductive Sciences 16, 216-229. WENZINGER, B. & BLEUL, U. 2012. Effect of a prostaglandin F2α analogue on the cyclic corpus luteum during its refractory period in cows. BMC Veterinary Research, 8, 220. WESTERMANN, S., DRILLICH, M., KAUFMANN, T. B., MADOZ, L. V. & HEUWIESER, W. 2010. A clinical approach to determine false positive findings of clinical endometritis by vaginoscopy by the use of uterine bacteriology and cytology in dairy cows. Theriogenology, 74, 1248-1255. WILLIAMS, B. L., SENGER, P. L. & OBERG, J. L. 1987. Influence of cornual insemination on endometrial damage and microbial flora in the bovine uterus. Journal of Animal Science, 65, 212-216. WILLIAMS, E. J., FISCHER, D. P., PFEIFFER, D. U., ENGLAND, G. C. W., NOAKES, D. E., DOBSON, H. & SHELDON, I. M. 2005. Clinical evaluation of postpartum vaginal mucus reflects uterine bacterial infection and the immune response in cattle. Theriogenology, 63, 102-117. WILTBANK, M. C., BAEZ, G. M., GARCIA-GUERRA, A., TOLEDO, M. Z., MONTEIRO, P. L. J., MELO, L. F., OCHOA, J. C., SANTOS, J. E. P. & SARTORI, R. 2016. Pivotal periods for pregnancy loss during the first trimester of gestation in lactating dairy cows. Theriogenology, 86, 239-253. WILTBANK, M. C., SOUZA, A. H., CARVALHO, P. D., BENDER, R. W. & NASCIMENTO, A. B. 2011. Improving fertility to timed artificial insemination by manipulation of circulating progesterone concentrations in lactating dairy cattle. Reproduction, Fertility and Development, 24, 238-243. WILTBANK, M. C., SOUZA, A. H., CARVALHO, P. D., CUNHA, A. P., GIORDANO, J. O., FRICKE, P. M., BAEZ, G. M. & DISKIN, M. G. 2014. Physiological and practical effects of progesterone on reproduction in dairy cattle. Animal, 8 Suppl 1, 70-81. WOOD, S. 2006. Generalized additive models: an introduction with R, London, UK, Chapman and Hall/CRC. WRIGHT, J. M. 1981. Non-surgical embryo transfer in cattle embryo-recipient interactions. Theriogenology, 15, 43-56. YAMASHITA, H., KAMADA, D., SHIRASUNA, K., MATSUI, M., SHIMIZU, T., KIDA, K., BERISHA, B., SCHAMS, D. & MIYAMOTO, A. 2008. Effect of local neutralization of basic fibroblast growth factor or vascular endothelial growth factor by a specific antibody on the development of the corpus luteum in the cow. Molecular Reproduction and Development, 75, 1449-1456. ZAAYER, D. & VAN DER HORST, C. J. G. 1986. Non-fertility in cows: Treatment with PGF and investigation of uterine biopsies. Cytobios, 45, 55-70. 163

Appendix Appendix A. Technical note

Ramirez-Garzon O, Satake N, Lyons RE, Hill J, Holland MK, McGowan M. Endometrial biopsy in Bos indicus beef heifers. Reprod Dom Anim. 2017;00:1–3. https://doi.org/10.1111/rda.12944.

Appendix B. Abstract conference

Ramirez-Garzon O., Satake N., Lyons R. E., Palmieri C., Hill J., Gallego-Lopez C., Holland M. K., McGowan M. (2016) 63 Effect of endometrial biopsy on uterine health of tropically adapted beef cattle. Reproduction, Fertility and Development 29, 139-139.https://doi.org/10.1071/RDv29n1Ab63

162

Appendix

Appendix C. Progesterone assessment

a. Instrumentation

The tandem LC-MS/MS (LC/MS/MS) system was a Shimadzu Nexera™ uHPLC connected to a triple quadrupole mass spectrometer, Shimadzu LCMS 8050 (Shimadzu Corporation, Tokyo, Japan) with the following configuration: Nexera™ uHPLC: two LC-30AD chromatographic pumps, CTO-

30A column oven, SIL-30AC autosampler and DGU-20A5 degasser and LCMS 8050: a DUIS-ESI source, Argon gas (BOC, QLD, Australia) into the collision cell and a nitrogen generator (NM32L, Peak Scientific Australia Pty. Ltd., VIC, Australia).

b. Data processing software

Mass transition of Progesterone (P4) was m/z 315.20 < 109.15 for quantitation, with m/z 315.20 <

97.20 as confirmation. Deuterated-progesterone (D9P4; CDN Isotopes, Inc Pointe-Claire, Quebec, Canada) was used as an internal standard, MRM transition was m/z 324.20 < 113.25 for quantitation and with m/z 324.20 < 113.25 and m/z 324.20 < 287.10 for confirmation. Quantification was performed with Skyline-daily software (Beta version, Macross Lab Software, Seattle, WA, USA).

c. Analytical Conditions

Chromatographic separation was performed using a Reverse phase Biphenyl column (Kinetex 1.7 µm Biphenyl 100Å, size 50 x 2.1 mm; Phenomenex Inc., Lane Cove, NSW, Australia) equipped with a SecurityGuard™ ULTRA Guard Cartridge (Phenomenex Inc., Lane Cove, NSW, Australia). The mobile phases were UHPLC-grade water for pump A and 100% Methanol (Optima® LC/MS solvents, Thermo Fisher Scientific Inc, Australia) for pump B. The column temperature was 40 ºC.

The gradient elution was a binary gradient from 55% of mobile phase B to 95 % over 3.5 min then mobile phase B was held at 95% to 4.00 min elapse time before rapidly reducing mobile phase B back to 55 % at 4.01 min and maintained at 55% until pump pressures were returned to stable initial column pressure. The total chromatographic separation was carried out over 7 min. Total flow rate of 400 µL/min between pump A and B was maintained.

d. Instrument conditions

DUIS-ESI source was in the positive ion mode, desolvation temperature at 250 ºC, heating block at 400 ºC, the gas flow at 2 L/min, with the drying gas flow of 10 L/min and the capillary voltage set 163

Appendix at 4500 V. The collision gas pressure at 270 kPa and collision energy was -12 V for both the analyte and internal standards. Dwell time for the transition was set at 100 ms.

e. Standards, quality control standards and sample preparation

Stock standard solutions (100 µg/mL) and working solutions of P4 and Deuterated P4 (D9P4) were prepared in methanol. Pooled plasma of three bull plasma samples was filtered through activated charcoal (Sheldon and Coppenger, 1977) and used as sample matrix (blank) with the analyte of interest. Briefly, 5 mg/mL of activated charcoal was mixed with pooled bull plasma and agitated at 250 rpm, at 37 ºC for 2 h, centrifuged at 20,000g, the supernatant was then filtered using a sterile 0.2 µm syringe-filter to remove the remaining activated charcoal particles.

Instrument method: Linear analytical calibration of six standards into activated charcoal-filtered plasma matrix with concentration ranges from 0.5 to 50 ng/ml. For quality control (QC) samples, blank plasma matrix was spiked with 2.5 ng/mL for low, 10 ng/mL for mid-range and 20 ng/mL upper prior to protein precipitation. Each sample was spiked with 2.5 ng/mL internal standard prior to protein precipitation.

Plasma samples were prepared by removing protein and phospholipids before steroid extraction. Protein was precipitated by mixing plasma to cold 1% formic acid in acetonitrile at a ratio of 1:4, vortexing for 1 min, x2 centrifugation at 20 000 g for 10 min, samples were rested at 4 ºC for 30 min between centrifugation steps. The precipitate was loaded into Phospholipids Removal plates (Phree™, Phenomenex, Torrance, CA, USA) to remove remaining protein and phospholipids and extract steroids by solid phase extraction, as per manufacturer’s instructions. The elutant was collected and evaporated to dryness using vacuum concentrator (Speedvac) for 2-4 hr. The samples were reconstituted in methanol (50 µl) and transferred to HPLC vials and 1 µL sample was injected into the uHPLC column. QC samples were assessed for every 20 samples test to assess instrument performance.

f. Results

Calibration assessments showed a linear response of r2 > 0.9919. Across 3 batches the average slopes and intercepts were 0.99563 ± 0.02 and 698,541 ± 6.93 (mean ± SEM), respectively. Instrument accuracy was calculated to be 91.9%, 89.8%, 95.1%, 94.2%, 99.6% and 99.1% over the six calibration points of the 0.5 (LLOQ) – 25 (HLOQ) ng/mL range (Figure C1, C2).LOD without matrix 0.391 ng/mL. Instrument QC CV all <15%. Extraction efficiency: Batch1=87.3%±18.1; Batch2=79.92 %±15.6; Batch3=82.0%±16.7

164

Appendix

Figure C1: Representative calibration curve for Progesterone assessment by LCMS.

165

Appendix

Figure C2: Instrument stability across the 3 QC concentrations all less than 15%. QC is assess after a set of every 20 samples (so batch of 100 there are 6 points) b) at the retention time stability.

166

Appendix

Appendix D. PGFM assessment

Plasma PGFM concentrations were determined using an enzyme immunoassay kit (DetectX® 13,14, Dihydro-15-keto-Prostaglandin F2alpha (PGFM), Catalogue #K022 Arbor Assays®, Ann Arbor, Michigan, USA) as per the manufacturer’s instructions without modification. The antiserum used in this assay (rabbit polyclonal antibody) is reported to cross-react 100% with 13,14-dihydro- 15-keto-Prostaglandin F2α (PGFM) 1.5% with PGEM and 0% with all other metabolites tested. The sensitivity of the PGFM kit reported was 46.2 pg/ml.

a. Assay validation

A pool of plasma samples (200µl) was serially diluted two-fold in assay buffer Catalogue #X067 Arbor Assays®, USA) (neat, 1:2, 1:4, 1:8, 1:16, 1:32, 1:64) and analysed for PGFM as per the kit’s instructions (Table D1). The percent binding of samples were plotted against the percent binding of the standard curve, aligning the neat sample with the standard sample (400 pg/ml) that most close matched its percent binding. The validation tests showed parallelism among serial dilutions against the standard curve as shown in (Figure D2). This procedure also indicated that the average plasma PGFM levels for this species were too low to analyse at a 1:8 dilution as recommended by the manufacturer. To ensure that samples could be analysed at a 1:2 dilution (within the detectable limit of the assay) instead without interference with plasma constituents that might interfere with antibody binding, a recovery test was performed.

b. Recovery test

To ensure there was limited interference by plasma constituents to the binding of plasma PGFM to the EIA antibody a recovery test was performed. A pooled sample was aliquoted (75 µl per tube) and spiked with three different levels PGFM standards (100, 200 and 400 pg/ml) at a 1:2 dilution and analysed with the original un-spiked pooled sample as per the kit’s instructions (Table D2).

167

Appendix

Table D1: Serial dilution of plasma samples and standards for PGFM assay validation.

120

100

60 Standards

% % Binding 40 Serial Pool

0 3200 1600 800 400 200 100 50 0 Standard Concentrations pg/ml

Figure D1: Parallelism of serial plasma dilutions (% binding) of PGFM against standard curve.

168

Appendix Table D2: The reported PGFM concentrations of the pooled plasma sample and spiked aliquots and the recovery calculation of expected versus reported concentrations. The individual percentage of recovered PGFM per spiked aliquot and overage average PGFM recovery are reported

Amount Expected = (Pool concentration/2) + (Amount Spike/2); % Recovery = (Concentration)/Amount expected; Average overall recovery = % recovery/3

Standard preparation The PGFM standard solutions were made by serial dilutions in assay buffer (Catalogue #X067 Arbor Assays®, USA) of PGFM stock solution (Catalogue #C085) as per the manufacturer’s recommendations. Standards (50 µl) were run in duplicates in the following concentrations 3200, 1600, 800, 400, 200, 100 and 50 pg/ml on every EIA.

Sample preparation Plasma samples (150 µl) were thawed, gently mixed and diluted 1:2 in assay buffer (150 µl). Immediately prior to analysis, sample dilutions were vortexed and then analysed (50 µl per well in duplicate) as per the manufacturer’s recommendations. The intra and inter-assay coefficients of variations were 3.7% and 12.8% respectively.

169