Using Transcriptomic and Proteomic Analysis to Characterize the Biological Basis of Puberty in Brahman Heifers

Thesis Title: Using transcriptomic and proteomic analysis to characterize the biological basis of puberty in Brahman heifers

Loan To Nguyen

Bachelor of Biotechnology

Master of Biotechnology

A thesis submitted for the degree of Doctor of Philosophy at The University of Queensland in 2018 School of Chemistry and Molecular Biosciences

Abstract: The beef industry in Northern Australia is facing increased demands for efficient herd productivity. Brahman cattle, a breed of the Bos indicus sub-species, can withstand hostile conditions in northern production systems (Burrow 2012). However, there is a large variability in the age at onset of puberty in Bos indicus and Bos taurus crossbred heifers (Warnick 1963; Wythe 1970; Reynolds 1973). Brahman heifers are older and heavier at puberty in comparison with Hereford, Angus, Holstein and Jersey heifers, which are all Bos taurus breeds (Baker et al. 1989). Late onset of puberty in Bos indicus can lead to a decrease of lifetime productivity (Lesmeister et al. 1973; Johnston et al. 2013). Therefore, reducing the age at puberty to increase cattle productivity is a major aim for Bos indicus breeders.

In heifers, puberty is initiated when the hypothalamus-pituitary-gonad axis loses its sensitivity to negative feedback effects of steroid hormones, allowing an increase in gonadotropin-releasing hormone (GnRH) secretion from GnRH neurons in the hypothalamus. The increase then stimulates the secretion of gonadotropins luteinizing hormones (LH) and follicle stimulating hormone (FSH). FSH and LH then regulate gonadal development (Day et al., 1984; Day et al., 1987; Schillo et al., 1992; Day, 1998; Gasser et al., 2006). However, the process is controlled by multiple factors and the complex interactions between environmental and genetic factors regulating the process is only now coming to light. Therefore, the aim of this PhD project was to provide a transcriptomic and proteomic analyses of important tissues related to heifer puberty. In this work, I targeted the hypothalamus, the pituitary gland, the liver and the adipose tissue from Brahman heifers that were either pre- or post-pubertal.

The transcriptomic analysis using an RNA sequencing technique provided reliable sequence reads. An average of 60 million paired-end reads was produced from each sample, of which 60 – 70% were properly mapped to the Bos taurus reference genome assembly UMD3.1. Sequence data of these transcriptome data sets are available through the Functional Annotation of Animal Genomes project (http://data.faang.org/home). The study identified 392, 284, 452 and 525 differentially expressed (DE) genes among pre- and post-pubertal heifers in the hypothalamus, pituitary gland, liver and adipose tissue, respectively (P-value < 0.05). A regulatory impact factor metric identified key regulators (transcription factors, TF) for each studied tissue. A list of TF that were also DE genes associated with Brahman heifer puberty was revealed: BDNF, STAT6, PBX2, PBRM1, E2F8, NFAT5, SIX5, ZBTB38, ZNF605, HES7, HEY2, RORC, PPARG, PITX1, NFE2, HOXB5, and zinc finger family genes. The DE and key TF genes were mainly assigned to three pathways: Wnt, TGF-beta and PPAR signalling (P-

value < 0.05, corrected). These results support our knowledge about the complex nature of puberty, a trait influenced by numerous genes organized in functional pathways.

The project also investigated the protein abundance differences between pre- and post- pubertal heifers using the same tissues. The hypothalamus and pituitary gland appeared as tissues having the highest number of differentially abundant (DA) proteins after puberty (adjusted P-value < 0.01). These valuable proteome datasets are publicly available via The PRoteomics IDEntifications (PRIDE) database (https://www.ebi.ac.uk/pride/archive/login). The functional analysis of DA proteins showed numerous enriched pathways (FDR < 0.05) that might regulate the occurrence of puberty, such as GnRH signalling, estrogen signalling pathway, steroid hormone biosynthesis, axon guidance, glutamatergic synapse, GABAergic synapse, oxytocin signalling pathway, calcium signalling pathway, PI3K-AKT signalling pathway, PPAR signalling pathway, ribosome and specific amino acid metabolisms. The candidate proteins revealed in this study will be useful for understanding the molecular mechanisms of puberty in Bos indicus breeds.

When compared, proteins and their corresponding transcripts expression levels in the four studied tissues the Person correlations were weak: 0.05 in the Hypothalamus (P-value = 0.16), 0.006 in pituitary gland (P-value = 0.8), 0.13 in liver (P-value = 0.001) and -0.06 in the adipose tissue (P-value = 0.2). The non-significant global correlation between mRNA levels and protein abundance presented in this Thesis confirmed that the transcriptome data is not sufficient to predict or speculate regarding the corresponding proteome data when samples represent tissues. It is possible that in single cell analyses, the transcriptome would match the proteome better, but this data was not available. In other words, changes in mRNA expression might provide a limited understanding of changes in the expression of proteins in complex systems.

In summary, these are the first studies that performed transcriptomic and proteomic analyses of reproductive tissues (hypothalamus and pituitary) and metabolic-related tissues (liver and adipose tissue) in pre- and post-pubertal Brahman heifers. The results presented in this Thesis provide an important step forward in our efforts to understand the molecular biology underlying puberty. This knowledge can bring the research one step closer towards the goal of enhancing genetic improvement predictions for puberty in Bos indicus cattle.

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.

III

Publications during candidature

Peer-reviewed papers M. R. S. Fortes, L. T. Nguyen, L.R. Porto Neto, A. Reverter, S. S. Moore, S. A. Lehnert, M. G. Thomas, (2016) Polymorphism and genes associated with puberty in heifers. Theriogenology 86: 333-339 M. R. S. Fortes, L. T. Nguyen, Weller, M.D.C.A., Cánovas. A., Islas-Trejo, A., Porto- Neto, L.R., Reverter, A., Lehnert, S.A., Boe-Hansen, G.B., Thomas, M.G., Medrano, J.F. and Moore, S.S. (2016) Transcriptome analyses identify five transcription factors differentially expressed in the hypothalamus of post-versus pre-pubertal Brahman heifers. Journal of Animal Science, 94(9):3693-3702 L.T. Nguyen, Reverter. A., Cánovas. A., Venus. B., Islas-Trejo. A., Porto-Neto. L.R., Lehnert. S.A., Medrano. J.F., Moore. S.S. and M.R. S. Fortes, (2017) Global differential gene expression in the pituitary gland and the ovaries of pre- and postpubertal Brahman heifers. Journal of Animal Science, 95:599–615 M.M. Dias, A. Cánovas, C. Mantilla-Rojas, D.G. Riley, P. Luna-Nevarez, S.J. Coleman, S.E. Speidel, R.M. Enns, A. Islas-Trejo, J.F. Medrano,S.S. Moore, M.R.S. Fortes, L.T. Nguyen, B. Venus, I.S.D.P. Diaz,F.R.P. Souza, L.F.S. Fonseca, F. Baldi, L.G. Albuquerque, M.G. Thomas and H.N. Oliveira, (2017) SNP detection using RNA- sequences of candidate genes associated with puberty in cattle. Genetics and molecular research: GMR 16(1)

L. T. Nguyen, A. Reverter, A. Cánovas, B. Venus, S. T. Anderson, A. Islas-Trejo, M. M. Dias, N.F. Crawford, S. A. Lehnert, J. F. Medrano, M. G. Thomas, S. S. Moore and M. R. S. Fortes, (2018) STAT6, PBX2, and PBRM1 emerge as predicted regulators of 452 differentially expressed genes associated with Puberty in Brahman Heifers. Frontiers in Genetics 9(87).

L. T. Nguyen, L. F. Zacchi, B. L. Schulz, S. S. Moore and M. R. S. Fortes, (2018) Adipose tissue proteomic analyses to study puberty in Brahman heifers. Journal of Animal Science, 96(6): 2392–2398.

M. R. S. Fortes, L. F. Zacchi, L. T. Nguyen, F. Raidan, M. M. D. C. A. Weller, J. C. J. Ying, A. Reverter, J. P. A. Rego, G. B. Boe- Hansen, L. R. Porto-Neto, S. A. Lehnert, A. Canovas, B. L. Schulz, A. Islas-Trejo, J. F. Medrano, M. G. Thomas, S. S. Moore,

(2018) Pre and post-puberty expression of genes and proteins in the uterus of Bos indicus heifers: the luteal phase effect post-puberty. Animal Genetics. http://dx.doi.org/10.1111/age.12721.

P. A. S. Fonseca, S. Id-Lahoucine, A. Reverter, J. F. Medrano, M. R. S. Fortes, J. Casellas, F. Miglior, L. Brito, M. R. S. Carvalho, F. S. Schenkel, L. T. Nguyen, L. R. Porto-Neto, M. G. Thomas and A. Cánovas, (2018) Combining multi-OMICs information to identify key-regulator genes for pleiotropic effect on fertility and production traits in beef cattle. Plos One, Accepted manuscript.

Conference papers

L.T. Nguyen, G.M.F.D Camargo, R.E. Lyons, S.A. Lehnert and M.R.S. Fortes (2015) Effects of TEX11 and AR polymorphisms on reproduction and growth traits in Australian beef cattle. 21th Conference of the Association for the Advancement of Animal Breeding and Genetics, Victoria, Australia M. R. S. Fortes, F. Almughlliq, L. T. Nguyen, L.R. Porto Neto, S. A. Lehnert (2015) Non-synonimous polymorphism in HELB is associated with male and female reproductive traits in cattle. 21th Conference of the Association for the Advancement of Animal Breeding and Genetics, Victoria, Australia. L.T. Nguyen, M.R.S. Fortes and A. Reverter (2017) A pipeline for the analysis of multi-omics data. 22th Conference of the Association for the Advancement of Animal Breeding and Genetics 22:289-292, Townville, Australia, 2 July – 5 July. P. A. S. Fonseca, S. Id-Lahoucine, A. Reverter, J. F. Medrano, M. R. S. Fortes, J. Casellas, F. Miglior, L. Brito, M. R. S. Carvalho, F. S. Schenkel, L. T. Nguyen, L. R. Porto-Neto, M. G. Thomas and A. Cánovas, (2018) Identification of pleiotropic key- regulatory genes related to economically relevant traits in beef cattle through systems biology approach. Proceedings of the World Congress on Genetics Applied to Livestock Production, 11.736, Auckland, New Zealand, 11 Feb – 16 Feb.

Conference presentations (underline denotes poster presentation, “*” denotes oral presentation)

L.T. Nguyen*, G.M.F.D Camargo, R.E. Lyons, S.A. Lehnert and M.R.S. Fortes (2015) Effects of TEX11 and AR polymorphisms on reproduction and growth traits in Australian beef cattle. 21th Conference of the Association for the Advancement of Animal Breeding and Genetics, Victoria, Australia L.T. Nguyen*, M.R.S. Fortes and A. Reverter (2017) A pipeline for the analysis of multi-omics data. 22th Conference of the Association for the Advancement of Animal Breeding and Genetics 22: 289-292, Townville, Australia, 2 July – 5 July. L.T. Nguyen, A. Reverter, A. Cánovas, B. Venus, A. Islas-Trejo, S.A. Lehnert, J.F. Medrano, S.S. Moore and M.R.S. Fortes (2016) Liver transcriptome from pre- versus post-pubertal Brahman heifers. 35th Conference for the International Society for Animal Genetics (ISAG).

Publications included in this thesis

M. R. S. Fortes, L. T. Nguyen, Weller, M.D.C.A., Cánovas. A., Islas-Trejo, A., Porto-Neto, L.R., Reverter, A., Lehnert, S.A., Boe-Hansen, G.B., Thomas, M.G., Medrano, J.F. and Moore, S.S. (2016) Transcriptome analyses identify five transcription factors differentially expressed in the hypothalamus of post-versus pre-pubertal Brahman heifers. Journal of Animal Science, 94(9):3693-3702 - incorporated as Chapter 1.1

Contributor Statement of contribution Author: M. R. S. Fortes Carried out animal trial (100%) Collected samples (60%) Designed experiments (40%) Analysed the data (40%) Interpreted the results (50%)

Provided figures and tables (60%) Wrote the paper (100%) Author: Loan To Nguyen (Candidate) Designed experiments (10%) Analysed the data (40%) Interpreted the results (30%) Provided figures and tables (40%) Wrote and Edited the paper (50%) Author: Weller, M.D.C.A. Interpreted the results (10%) Edited the paper (5%) Author: Cánovas. A. Performed the assembly (100%) Edited the paper (5%) Author: Islas-Trejo Performed RNA sequencing (100%) Edited the paper (5%) Author: Porto-Neto, L.R. Collected samples (40%) Designed experiments (10%) Edited the paper (5%) Author: Reverter, A. Analysed the data (20%) Edited the paper (5%) Author: Lehnert, S.A. Designed experiments (10%) Edited the paper (5%) Author: Boe-Hansen, G.B. Designed experiments (10%) Edited the paper (5%) Author: Thomas, M.G. Designed experiments (10%) Edited the paper (5%) Author: Medrano, J.F. Supervised library preparation and RNA sequencing (100%) Edited the paper (5%) Author: Moore, S.S. Designed experiments (10%) Edited the paper (5%)

VII

Nguyen. L.T., Reverter. A., Cánovas. A., Venus. B., Islas-Trejo. A., Porto-Neto. L.R., Lehnert. S.A., Medrano. J.F., Moore. S.S. and M.R. S. Fortes, (2017) Global differential gene expression in the pituitary gland and the ovaries of pre- and post-pubertal Brahman heifers. Journal of Animal Science, 95:599–615 - incorporated as Chapter 1.2

Contributor Statement of contribution Author: Loan To Nguyen (Candidate) Designed experiments (20%) Analysed the data (60%) Interpreted the results (100%) Provided figures and tables (100%) Wrote the paper (100%) Author: Reverter, A. Analysed the data (20%) Edited the paper (20%) Author: Venus. B. Extracted RNA samples (100%) Author: Cánovas. A. Performed the assembly (100%) Author: Islas-Trejo, A. Performed RNA sequencing (100%) Author: Porto-Neto, L.R. Collected samples (40%) Designed experiments (10%) Edited the paper (5%) Author: Lehnert, S.A. Designed experiments (10%) Edited the paper (5%) Author: Medrano, J.F. Supervised library preparation and RNA sequencing (100%) Edited the paper (5%) Author: Moore, S.S. Designed experiments (20%) Edited paper (5%) Author: M. R. S. Fortes Carried out animal trial (100%) Collected samples (60%) Designed experiments (40%) Analysed the data (20%) Edited paper (60%)

VIII

Contributor Statement of contribution Author: Loan To Nguyen (Candidate) Designed experiments (20%) Analysed the data (60%) Interpreted the results (100%) Provided figures and tables (100%) Wrote the paper (100%) Author: Reverter, A. Analysed the data (20%) Edited paper (20%) Author: Cáfnovas, A. Performed the assembly (100%) Edited paper (5%) Author: Venus, B. Extracted RNA samples (100%) Author: Anderson, S. T. Measured hormone levels (100%) Edited paper (5%) Author: Islas-Trejo, A. Performed RNA sequencing (100%) Author: Dias, M. M. Performed quality control of raw data (50%) Author: Crawford, N.F. Performed quality control of raw data (50%) Author: Lehnert, S.A. Designed experiments (10%) Edited the paper (5%) Author: Medrano, J.F. Supervised library preparation and RNA sequencing (100%) Edited the paper (5%) Author: Thomas, M.G. Designed experiments (10%) Edited the paper (5%) Author: Moore, S.S. Designed experiments (10%) Edited paper (5%) Author: M. R. S. Fortes Carried out animal trial (100%) Collected samples (100%) Designed experiments (50%)

Analysed the data (20%) Edited paper (50%)

L. T. Nguyen. L. F. Zacchi, B. L. Schulz, S. S. Moore and M. R. S. Fortes, (2018) Liver proteomics to study the onset of puberty in Brahman heifers. Proceedings of the World Congress on Genetics Applied to Livestock Production, 11. 256, Auckland, New Zealand, 11 Feb – 16 Feb – incorporated as Chapter 2.3 Contributor Statement of contribution Author: Loan To Nguyen (Candidate) Designed experiments (40%) Extracted protein (100%) Analysed the data (80%) Interpreted the results (100%) Provided figures and tables (100%) Wrote the paper (100%) Author: Zacchi, L. F. Designed experiments (10%) Performed mass spectrometry (100%) Analysed the data (20%) Author: Schulz, B. L. Designed experiments (10%) Author: Moore, S.S. Edited paper (10%) Author: M. R. S. Fortes Designed experiments (40%) Edited paper (90%)

L. T. Nguyen. L. F. Zacchi, B. L. Schulz, S. S. Moore and M. R. S. Fortes, (2018) Adipose tissue proteomic analyses to study puberty in Brahman heifers. Journal of Animal Science, 96 (6), 2392–2398, https://doi.org/10.1093/jas/sky128 – incorporated as Chapter 2.4

Contributor Statement of contribution Author: Loan To Nguyen (Candidate) Designed experiments (40%) Extracted protein (100%) Analysed the data (80%) Interpreted the results (100%) Provided figures and tables (100%)

Wrote the paper (100%) Author: Zacchi, L. F Designed experiments (10%) Performed mass spectrometry (100%) Analysed the data (20%) Edited paper (10%) Author: Schulz, B. L Designed experiments (10%) Edited paper (10%) Author: Moore, S.S. Designed experiments (10%) Edited paper (10%) Author: M. R. S. Fortes Designed experiments (30%) Edited paper (70%)

Manuscripts included in this thesis No manuscripts included

Contributions by others to the thesis

The conception and design of this research project, as well as, analysis and interpretation of data were achieved through discussions and consultations with the advisory team and collaborators as listed below Fortes, M. R.S. and Moore, S.S assisted substantially in the conception and design of the thesis as a whole Fortes, M. R.S. and Moore, S.S assisted with editing and revision of the thesis document and associated publications Venus, B. contributed to laboratory works in Part 1 Reverter, A. assisted with the statistical analysis of data presented in Part 1 Zacchi, L.F., and Schulz, B.L. contributed to data normalization presented in Part 2

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

None

Research involving Human or Animal Subjects Management, handling, and euthanasia of animals were approved by the Animal Ethics Committee of The University of Queensland, Production and Companion Animal group (certificate number QAAFI/279/12). A copy of animal ethics approval certificate was included in the thesis appendix.

XII

Acknowledgements

Foremost, I would like to show my heartfelt appreciation to my principal supervisor, Dr. Marina Fortes, for being such a supportive advisor, for her ongoing guidance, encouragement, immense knowledge and for the opportunities that she has given me in attending conferences, workshops as well as in connecting me to the network of talented experts. She is a wonderful supervisor who has helped and made me a better scientist. I have always felt privileged to be one of her students. Dr. Marina has been and will continue to be one of the most influential people in my research career.

I would like to send my sincere thanks to my associate supervisor, Professor Stephen Moore, for being present and assisting throughout my research degree. Thanks for his instruction, suggestions, encouragement and insightful comments.

I would also like to send my grateful thanks to Dr. Antonio Reverter for his assistance in statistical analyses of my data. I admire him for his profound knowledge of statistics and bioinformatics. No matter what the question, he always had the answers with a big broad smile.

I would also like to acknowledge my committee members, Professor Bernard Carroll, Associate Professor Joe Rothnagel, Dr. Benjamin Schulz for their helpful comments and suggestions. Their invaluable comments at each milestone helped in getting my thesis to the form that I am now submitting.

Heartfelt thanks go to Professor Milton G Thomas for taking me as a visiting scholar at his group in Colorado State University, United States. I will never forget his support for providing me the opportunity to visit his research group.

I would like to take this opportunity to thank the University of Queensland for giving me a chance to come to this esteemed institute and pursue my PhD. Moreover, I owe a debt of gratitude to the School of Chemistry and Molecular Biosciences and Australian Mathematical Sciences Institute for support funds which helped me to attend numerous conferences and workshops.

XIII

My gratitude also extends to staff at the Department of Animal Science, University of California, Davis and Centre for Genetic Improvement of Livestock, University of Guelph for their help in performing RNA sequencing and assembling my research samples; to Ms. Bronwyn Venus in Queensland Alliance for Agriculture & Food Innovation for her support and assistance in laboratory; to Dr. Lucia Zacchi in School of Chemistry and Molecular Biosciences for her support with proteomics; to Dr. Sigrid Lehnert in CSRIO for her advice during my study as well as her support for my future career; and to colleagues at Vietnam National University of Agriculture, for their encouragement.

I am also grateful to my friends Dr. Nguyen Nhiep, Ms. Huong Pham, Ms. Thu Vu, Dr. Cuong Chu, Ms. Hiep Tran, Ms. Linh Nguyen and Ms. Stephlina D’Cunha, for their help, support and continuous encouragement during this time.

For my parents (Mr. Ngu Nguyen and Mrs. Quy Nguyen), my sister (Ms. Linh Thuy Nguyen) and my nephew (Duc Minh Nguyen), thanks for their constant and unconditional support. Thank you so much for listening to me every night and sharing happiness and sadness with me during my studying. For my nephew, you are my sunshine. Whenever I did not have any motivation and hope, your smile woke me up.

Finally, I would like to express my gratitude to the most important person in my life – my husband, Hoang Ngo. He has been a constant source of inspiration and strength. Thanks for understanding my absence during these years as well as encouraging me in my academic pursuit. Without your encouragement, I could never have completed this journey.

XIV

Financial support This research was funded by the start-up grant awarded to Professor Stephen Moore. Financial support for this project was also provided by University of Queensland in the form of an UQ International Scholarship (UQI).

Keywords Puberty, RNA sequencing, Brahman heifers, Liquid chromatography mass spectrometry, Differentially expressed genes, Transcription factor, Gene co-expression network, PRIDE database, Protein-protein interaction network, FAANG

Australian and New Zealand Standard Research Classifications (ANZSRC) ANZSRC code: 070206 Animal Reproduction, 55% ANZSRC code: 060405 Gene Expression, 45%

Fields of Research (FoR) Classification FoR code: 0604 Genetics, 60% FoR code: 0702 Animal Production, 40%

XVI

Table of Contents Abstract ...... I Declaration by author ...... III Publications during candidature ...... IV Publications included in this thesis ...... VI Manuscripts included in this thesis ...... XI Contributions by others to the thesis ...... XII Statement of parts of the thesis submitted to qualify for the award of another degree ...... XII Acknowledgements ...... XIII Financial support ...... XV Keywords ...... XVI Australian and New Zealand standard research classifications (ANZSRC) ...... XVI Fields of research (for) classification ...... XVI List of figures ...... XIX List of tables ...... XXI List of abbreviations used in the thesis ...... XXII SECTION I. INTRODUCTION, REVIEW OF LITERATURE, AND OBJECTIVES ...... 1 INTRODUCTION ...... 2 REVIEW OF LITERATURE ...... 5 I.1. Introduction to age at puberty ...... 5 I.2. Age at puberty in Bos indicus and their crosses ...... 6 I.3. Mechanism controlling puberty ...... 7 I.4. Studying the animal transcriptome and proteome ...... 20 OBJECTIVES ...... 29 SECTION II. RESEARCH CHAPTERS ...... 30 PART 1 – TRANSCRIPTOMIC STUDIES ...... 31 Chapter 1.1. Transcriptome analyses identify five transcription factors differentially expressed in the hypothalamus of post-versus pre-pubertal Brahman heifers ...... 32 Chapter 1.2. Global differential gene expression in the pituitary gland of pre- and post- pubertal Brahman heifers ...... 48 Chapter 1.3. STAT6, PBX2 and PBRM1 emerge as predicted regulators of 452 differentially expressed genes associated with puberty in Brahman heifers ...... 61 Chapter 1.4. Global transcriptome analysis of adipose tissue identified differentially expressed genes related to puberty in Brahman heifers ...... 79 PART 2 – PROTEOMIC STUDIES ...... 95 Chapter 2.1. Hypothalamic proteomics revealed differentially abundant proteins and pathways for puberty process in Brahman heifers ...... 96

XVII

Chapter 2.2. Pituitary proteomics revealed differentially abundant proteins and pathways related to puberty in Brahman heifers...... 113 Chapter 2.3. Liver proteomics to study the onset of puberty in Brahman heifers ...... 125 Chapter 2.4. Adipose tissue proteomic analyses to study puberty in Brahman heifers 137 SECTION III. FINAL DISCUSSION AND FUTURE DIRECTIONS ...... 148 III. 1. Research overview ...... 149 III.2. Future directions ...... 157 III.3. Conclusion ...... 160 LIST OF REFERENCES ...... 161 APPENDICES ...... 210

XVIII

List of Figures SECTION I Figure I.1 Illustration of the Hypothalamus-Pituitary-Gonads axis. Figure I.2 A schematic depicted regulatory signals of the onset of puberty (Pineda et al., 2010). Figure I.3 A schematic illustration neuropeptide regulation of GnRH neuronal function (Ojeda et al., 2010a). Figure I.4 A schematic diagram presenting potential roles in the Hypothalamus-Pituitary- Ovary axis of several genes and their biological mechanism for age at menarche (Perry et al., 2014a). Figure I.5 The hypothetical of transcriptional control puberty by repressor and activator, adapted from Ojeda et al. (2010a). Figure I.6 Farmed animal transcriptomic studies published in NCBI between 2006 and 2016. Figure I.7 The fundamental proteomic workflow (Almeida et al., 2015). SECTION II Figure 1.1.1 Pipeline for analysis of the transcriptomic dataset. Figure 1.1.2 Volcano plots reveal genes that significantly differ between pre- vs. post- pubertal heifers in the hypothalamus. Figure 1.1.3 Co-expression network in the hypothalamus of post and pre-pubertal Bos indicus heifers.

Figure 1.1.4 Adiponectin KEGG pathway with differentially expressed genes are marked by red stars (from analyses using DAVID (Dennis et al., 2003)).

Figure 1.2.1 Volcano plots reveal genes that significantly differ between pre- vs. post- pubertal heifers in the pituitary gland.

Figure 1.2.2 Predicted co-expression gene networks comprising pre- and post-pubertal data in the pituitary gland.

Figure 1.3.1 Volcano plot of differentially expressed genes (n = 452) in liver between pre- vs post-pubertal heifers.

Figure 1.3.2 Liver gene co-expression network constructed by PCIT in pre- and post- puberty Brahman heifers.

XIX

Figure 1.3.3 Subset of the liver co-expression network showing the best trio of TF and its predicted target genes in pre- and post-puberty Brahman heifers.

Figure 1.4.1 Plot of differentially expressed genes (n = 525) in adipose tissue samples between pre- vs post-pubertal heifers.

Figure 1.4.2 Adipose tissue gene co-expression network constructed by the partial correlation and information theory algorithm in pubertal Brahman heifers.

Figure 2.1.1 Schematic overview of the experimental workflow.

Figure 2.1.2 Gonadotropin-releasing hormone (GnRH) signalling pathway containing differentially abundant proteins in the hypothalamus of pre- and post-pubertal Brahman heifers.

Figure 2.1.3 Protein-protein interaction network in the hypothalamic tissue.

Figure 2.2.1 KEGG pathway analysis of differentially abundant proteins in the pituitary gland of pubertal Brahman heifers.

Figure 2.2.2 Protein-protein interaction network of differentially abundant proteins between pre- and post-pubertal heifers in the pituitary gland.

Figure 2.3.1 Volcano plot of differentially abundant proteins in the liver between pre- vs post-pubertal heifers.

Figure 2.3.2 Scatterplots with associated correlation coefficient (r) for mRNA and protein expression in the liver.

Figure 2.3.3 The interaction network of 275 differentially abundant proteins between pre- and post-pubertal Brahman heifers.

Figure 2.4 Protein-protein interaction network in the adipose tissue.

Section III

Figure III.1 Venn diagram of expressed mRNAs and proteins as well as differentially expressed (DE) mRNA and differentially abundant (DA) protein in each dataset

Figure III.2 Number of proteins identified, quantified and differential abundance in four studied tissues.

Figure III.3 Venn diagram of transcription factors identified by regulatory impact factor metric in the liver (LIV), the hypothalamus (HYP), adipose tissue (FAT) and pituitary (PIT).

List of Tables SECTION I Table I.1 Location of metabolic regulators in Hypothalamus and Pituitary of several species. SECTION II Table 1.1 List of five transcription factors (TF) were significant in two analyses: differential gene expression (DE) and regulatory impact factor metrics (RIF1 and RIF2, P-value < 0.05). Table 1.2 Genes in common in pituitary transcriptome profiles between Brahman and Brangus heifers. Table 1.3 The reads per kilobase per million mapped read (RPKM) values for differentially expressed genes in liver between pre- vs. post-pubertal Brahman heifers (|FC| 3, P 0.01).

Table 1.4 Genes that were differentially expressed (DE) and are transcription factors (TF), their connectivity in the co-expression network and fold changes (FC). Table 2.1.1 List of common differentially expressed genes and abundant proteins in two hypothalamus studies Table 2.1.2 Affected KEGG pathways and differentially abundant proteins at the onset of heifer puberty Table 2.2.1 Highly significant differentially abundant proteins (|Fold change| > 1 and P- value < 0.001) in comparison between pre- and post-pubertal Brahman heifers in the pituitary gland. Table 2.2.2 List of common differentially expressed genes and abundant proteins in transcriptome and proteome approaches in the pituitary gland (P-value < 0.05) Table 2.3 Differentially expressed genes and differentially abundant proteins between pre- and post-pubertal Brahman heifers according to both liver transcriptome and proteome studies. Table 2.4.1 Highly significant differentially abundant proteins in comparison between pre- and post-pubertal Brahman heifers in abdominal adipose tissue. Table 2.4.2 Differentially expressed genes and differentially abundant proteins between pre- and post-pubertal Brahman heifers according to both adipose tissue transcriptome and proteome studies.

XXI

List of Abbreviations used in the thesis

BP Biological process BW Body weight CC Cellular component CL Corpus luteum CS Condition score DA Differentially abundant DE Differentially expressed FC Fold change FDR False discovery rate FSH Follicle-stimulating hormone GnRH Gonadotropin-releasing hormone GO Gene ontology HPG Hypothalamus - pituitary - ovary IDA Information-dependent acquisition KEGG Kyoto Encyclopedia of Genes and Genomes LC - ESI -MS/MS Liquid chromatography-electrospray ionization-tandem mass spectrometry LH Luteinizing hormone MF Molecular function PCIT Partial correlation and information theory PCR Polymerase chain reaction PPI Protein-protein interaction QC Quality control RIF Regulatory impact factor RIN RNA integrity number RNA Ribonucleic acid RPKM Reads per kilobase per million mapped reads SWATH Sequential window acquisition of all theoretical mass spectra TF Transcription factor

XXII

SECTION I. INTRODUCTION, REVIEW OF LITERATURE, AND OBJECTIVES

INTRODUCTION

The beef cattle industry is an important contributor of the Australian economy, accounting for approximately 23% of the total gross value of farm production and approximately 22% of total value of farm export income in 2015-16 (Ashton et al., 2016). The Northern region, one of two main beef regions in Australia, encompasses the Northern Territory, northern Western Australia and Queensland. The farm cash incomes from the Northern region was recorded higher than that of the Southern region in both 2014-15 and 2015-16 (Ashton et al., 2016). In addition, Queensland maintains the largest proportion of total beef herd in Australia with around 45% followed by New Sale Wale with over 20% (Ashton et al., 2016). The Northern region primarily includes Bos indicus breeds such as Brahman and their crosses such as Santa Gertrudis and Droughtmaster (Gleeson et al., 2012). Puberty is a pivotal physiological milestone in the reproductive life of a heifer (Robinson &Shelton, 1977). However, the reproductive performance of female cattle in the Northern region is lower than that of the Southern region. Evidence to date supports that puberty in Bos indicus cattle ranges from 10 to 40 months of age in comparison with a range from 7 to 13 months of age in temperate Bos taurus (Lunstra &Echternkamp, 1982; Lunstra &Cundiff, 2003; Abeygunawardena &Dematawewa, 2004; Brito et al., 2004; Bortolussi et al., 2005; Shamay et al., 2005). Calving first as 3 year-olds then leads female cattle in northern Australia to generate 3 or 4 calves over the next 4 to 6 years (Gleeson et al., 2012). An observation of lifetime productivity by 5 years after first calving reported that heifers calved at 2-year-olds produced 0.6 more calve than those calved at 3-year-olds (Meaker et al., 1980). The lower productivity of northern beef cattle limits profitability and rate of genetic gain. The decrease of profitability in terms of trade due to later puberty in the northern beef cattle industry has been reported (Gleeson et al., 2012; McLean et al., 2013). Therefore, selection early pubertal heifer cattle would help to improve the whole herd reproductive performance and hence profitability. In heifers, puberty is triggered when the hypothalamus-pituitary-gonad (HPG) axis loses its sensitivity to negative feedback effects of steroid hormones, allowing an increase in gonadotropin-releasing hormone (GnRH) secretion from GnRH neurons in the hypothalamus. This increase then stimulates the secretion of gonadotropins luteinizing hormones (LH) and follicle stimulating hormone (FSH). FSH and LH then regulate gonadal development (Day et al., 1984; Day et al., 1987; Schillo et al., 1992; Day, 1998; Gasser et al., 2006). However, the

pubertal process is controlled by multiple factors and the complex interactions between environmental and genetic factors regulating this process has just commencing to be understood. The change in pulsatile secretion of GnRH depends on trans-synaptic and glial inputs to the GnRH neuronal network (reviewed in (Lomniczi et al., 2015c). The trans-synaptic input involves neuronal excitatory and inhibitory inputs, the glial input is almost all excitatory (Prevot, 2002). The glial cell utilizes growth factors or small molecules for cell-cell signalling to regulate GnRH release (Roth et al., 2006; Ojeda et al., 2008). In addition, each level of HPG axis is adversely influenced by inadequate nutrition. For instance, under-nutrition can decrease the secretion of LH in pre-pubertal heifers and thus impact on the occurrence of puberty (Schillo et al., 1992; Amstalden et al., 2000). Heifers’ off-spring maintained on a low energy diet attain puberty 19 days later than those on a high energy diet (Corah et al., 1975). Bos taurus straight-bred heifers raised on a low-level nutrient attained puberty 191 days and 148 days later than those developed on a high-level nutrient and their crosses, respectively (Wiltbank et al., 1969). Though inadequate nutrition has been proposed to affect the HPG axis, the precise physiological mechanisms underlying the interaction between nutrition and puberty remain poorly understood. Further studies are needed to fully investigate the metabolic regulators involved in bovine puberty. To date, genomics has provided an opportunity to characterize genes, markers and pathways related to puberty. Several genes such as G protein-coupled receptor 54 (GPR54), kisspeptin (KISS1), tachykinin 3 (TAC3), FSH receptor and cytochrome P450, family 3, subfamily A, polypeptide 4 (CYP3A4) were identified in which mutations associated with pubertal failure or known to be involved in the regulation of gonadotropin release (Aittomaki et al., 1995; de Roux et al., 2003; Seminara et al., 2003; Xin et al., 2005; Lapatto et al., 2007; Topaloglu et al., 2009). Genome-wide association studies (GWAS) reported an association between mutations in Lin28 homolog B (LIN28B) and early menarche (He et al., 2009; Ong et al., 2009; Perry et al., 2009b; Sulem et al., 2009). In heifers, GWAS also reported 350 markers associated with age at first calving and age at puberty (Hu et al., 2016). Nevertheless, none of these markers hold a master position to synchronise glial and neuronal networks involved in the puberty onset. It then has been suggested that puberty is controlled by regulatory gene networks comprising multiple functional modules (Ojeda et al., 2006). Due to the rapid development of high-throughput technologies, more comprehensive techniques have been developed on different levels, from genes to proteins. Omics technologies including transcriptomic and proteomics have become of mainstream applications to better

understand animal physiology and production. Transcriptomics has been utilizing to identify the differentially expressed genes between different experimental conditions. Down- and up- regulation of genes provided a better understanding of molecular mechanism in which animal response toward contrasting conditions. Review of literature from 2006 to 2016 noted that publications of the bovine transcriptome accounted for over 38% of total transcriptomic publications in farm animals (Parreira &de Sousa Araújo, 2018). Although transcriptomic approaches have successfully studied transcriptomic profiles of multi-tissue in Brangus heifers under pubertal conditions, the differences in age at puberty between Brangus and Brahman suggested a whole transcriptomic study of Brahman heifers. Furthermore, the expression of genes is regulated at the post-transcriptional and translational levels, thereby proteome analyses could provide an insight into molecular mechanisms. Taken all together, transcriptomic and proteomic analyses will aid a more complete understanding of mechanisms underlying puberty transition in Bos indicus sub-species. This is the first study that has performed transcriptomic and proteomic analyses of reproductive tissues (hypothalamus and pituitary) and metabolic-related tissues (liver and adipose tissue) in Brahman heifers. Basically, the aim of this project was to reveal the differences in pubertal-related transcript and protein expression patterns. These omics combined with bioinformatics tools generated large datasets for puberty in Bos indicus sub- species. Key genes and proteins that can regulate the maturation of the HPG axis were identified. Along with differential expression analyses, the gene-gene and protein-protein interaction using network approach provide clusters of co-regulated genes or proteins. Furthermore, the possible biological significance of the identified genes and proteins were further elucidated using functional enrichment analyses of gene ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. These results could provide a great promise in supporting to heighten reproductive performance in Bos indicus cattle.

REVIEW OF LITERATURE

I.1. Introduction to age at puberty

Age at puberty is a significant production trait because it is an important factor to determine lifetime productivity in beef heifers (Schillo et al., 1992). It indicates the initiation of an animal’s reproductive life and influences generation interval, and thus it may affect the rate of genetic gain as well as herd productivity. In short, heifers that attained puberty earlier can conceive earlier in the first breeding season with obvious consequences for the next seasons: calving early in the season allows more time for re-conception. Lesmeister et al., (1973) suggested that beef cows which reached puberty earlier calved earlier throughout their reproductive lives than those that attained puberty later. Moreover, calves born from earlier puberty heifers grew considerably faster from birth to weaning than those in the late puberty group. Consistent with these ideas, early pubertal heifers (first calving at two-year-olds) generated 0.6 more calves than those calving first as three-year-olds, when 5 years of productive life were observed (Meaker et al., 1980). Likewise, Luna-Nevarez et al., (2010) noted that early-maturing cows had improved fertility according to two indicators: younger age at first calving and shorter calving intervals. Calving intervals are largely influenced by postpartum anoestrus, a period of reproductive quiescence after calving. Post-partum anoestrus was genetically correlated to age at puberty, in 3-year-old (0.34) and 4-year-old (0.74) Brahman heifers (Johnston et al 2010). These genetic correlations again confirm that early pubertal heifers are likely to have shorter calving intervals because they will have shorter post-partum anoestrus. As a result, selecting for early pubertal heifers is a targeted approach for decreasing the generation interval and also enhancing fertility.

Age at puberty in heifers has been determined by observations of the first behavioural oestrous, which is followed by ovulation and first corpus luteum (CL) development (Romano et al., 2007; Johnston et al., 2009). Ovulation and first CL can be observed via ultrasound scanning or estimated by plasma progesterone values above 1ng/ml (Yelich et al., 1995). The heritability estimates of age at puberty ranged from 0.1 to 0.67 (Cammack et al., 2009). The measuring of plasma progesterone concentration or detecting CL using ultrasound scanning are challenging, time-consuming and costly methods for routine on-farm use. In fact, the detection of CL using ultrasound image requires an experienced technician, quite often a veterinarian. Determination of plasma progesterone concentrations requires blood collection and laboratory analyses, and thus it is quite expensive and typically only conducted for research

purpose. From a research point of view, these methods are very useful in investigating puberty and further elucidating its molecular basis, including genetic and metabolic factors. Further understanding of pubertal development is the necessary backbone for future innovation in terms of cattle management, selection, husbandry and breeding practices that could improve herd fertility.

I.2. Age at puberty in Bos indicus and their crosses

Zebu cattle (Bos indicus breeds), which dominate herds in tropical and sub-tropical areas such as India, Northern Australia, and Brazil, are known to attain puberty at an older age than European cattle (Bos taurus breeds) (Laster et al., 1979; Martin et al., 1992; Lunstra &Cundiff, 2003). On average, the age at puberty in zebu cattle ranges from 10 to 40 months in comparison with that of 7 to 13 months in taurine temperate cattle (Lunstra &Echternkamp, 1982; Lunstra &Cundiff, 2003; Abeygunawardena &Dematawewa, 2004; Brito et al., 2004; Shamay et al., 2005). The average age at puberty among Brahman, Angus, and Angus x Brahman heifers was 27.2, 14.4 and 15.3 months, respectively (Reynolds, 1973). This was also confirmed by the report of Lopez et al., (2006) that noted Angus heifers attained puberty around 1 month earlier than Brangus heifers (3/8 Brahman x 5/8 Angus), and both Angus and Brangus cows reach puberty much earlier than purebred zebu heifers (Lopez et al., 2006). Although Bos indicus cattle are older age at puberty in comparison with Bos taurus, Bos indicus breeds are extensively utilized in the tropical and sub-tropical regions due to their adaption to harsh conditions (high temperature, seasonally poor pasture quality and parasites) in tropical environments. The Brahman is a Zebu breed that originated in the United States from cross-breeding of a mixture of at least four Bos indicus breeds from India and Brazil (Sanders, 1980). Due to the grading up process, Zebu cattle also contained Bos taurus content (Porto Neto et al., 2013). Brahman cattle are suited to harsh environmental conditions in tropical regions because of their resistance to tick, worms, heat and nutritional stresses (Frisch, 1987). The breed has lower inherent voluntary feed intake as well as lower relative maintenance requirements in comparison with Bos taurus (Frisch &Vercoe, 1977). The adaption to the tropical and sub- tropical environments makes Bos indicus breeds feasible options for production in these harsh environments. To minimize problems related to puberty of Zebu cattle, the increase of knowledge of physiological mechanisms underlying puberty in Bos indicus breeds has been useful to develop strategies for animal selection and breeding Zebu cattle that attain puberty earlier.

The above reasons encourage studies for understanding genes, proteins and pathways involving in puberty process in Brahman cattle, the predominant sub-species in northern Australia. Advances in genetic technologies will help to understand genetic contributions in pubertal Brahman heifers. Results from this study could contribute to functional information into genomic selection and breeding programs for early pubertal Brahman cattle.

I.3. Mechanism controlling puberty

The onset of puberty comprises a series of complex interactions between genetics and environmental factors. These processes are initiated with the release of gonadotropin-releasing hormone (GnRH) from the hypothalamus, which then stimulates the production and release of follicle stimulating hormone (FSH) and luteinizing hormone (LH) from the anterior pituitary gland. The FSH and LH secretion then enhance ovarian follicles development and estradiol production (Bliss et al., 2010). Importantly, the reproductive endocrine axis is not a unidirectional pathway. The estradiol hormones produced by the gonads also affect the hypothalamus and pituitary hormones through feedback mechanisms (Figure I.1). Although the Hypothalamus-Pituitary-Ovary (HPO) axis has been well documented, knowledge about the precise physiological mechanisms that control the initiation of female puberty is incomplete. It is widely accepted the GnRH release is the trigger for puberty, but the mechanisms that lead to GnRH release in heifers are less understood. Understanding factors that influence puberty and the mechanisms that mediate puberty may lead to the advance of new management strategies for improving fertility in heifers.

Figure I.1 Illustration of the Hypothalamus-Pituitary-Gonads axis (adapted from Schillo et al., 1992)

There are various complex steps in the puberty process. The process seems to involve neuropeptide, metabolic, genetics and environmental regulation. The environment in which cows are maintained can affect the onset of puberty. Feed intake, dietary energy restriction, metabolites and metabolic hormones, which are indicative of nutritional status, may influence the onset of puberty (Cammack et al., 2009). For instance, Insulin-like growth factor I (IGF1) and leptin serum levels have been reported to be positively associated with age at puberty (Garcia et al., 2002; Garcia, 2003; Daftary &Gore, 2005). The variation of age at puberty among sub-species of cattle and breeds also indicates that genetics is another important factor contributing to age at puberty. These factors that contribute to puberty are summarized in Figure I.2 and discussed in following paragraphs.

Figure I.2 A schematic depicted regulatory signals of the onset of puberty (Pineda et al., 2010) I.3.1. Neuropeptide regulation of puberty The initiation of puberty requires an increase in the pulsatile release of GnRH from the hypothalamus. This change depends on trans-synaptic and glial inputs to the GnRH neuronal network (reviewed in (Lomniczi et al., 2015c). While the trans-synaptic input involves neuronal excitatory and inhibitory inputs, the glial input is almost all excitatory (Prevot, 2002). Figure I.3 illustrates the regulation of trans-synaptic systems and glial systems as well as the components that control of GnRH release.

The excitatory trans-synaptic regulation of puberty is not only provided by neurons that use amino acids such as glutamate but also neurons that use peptides such as kisspeptin and neurokinin B as excitatory neurotransmitters or neuromodulators (Ojeda et al., 2006; Oakley et al., 2009; d'Anglemont de Tassigny &Colledge, 2010). The kisspeptin (KISS1) and its receptor (GPR54) have been known as powerful stimulators of GnRH release and thus hold a

pivotal role on central regulation of the HP axis (Seminara et al., 2003; Shahab et al., 2005; Kauffman et al., 2007; Oakley et al., 2009; Topaloglu et al., 2012). Expression of KISS1 and KISS1R/GPR54 increase near the expected time of puberty (Navarro et al., 2004; Castellano et al., 2005; Shahab et al., 2005). Most importantly, the association between inactivating mutations of GPR54 gene and lack of puberty attainment and hypogonadotropic hypogonadism in mice and humans is a testament to the important role played by KISS1/GPR54 in the reproductive system (de Roux et al., 2003). Kisspeptin neurons express estrogen receptor α (ERα) and participate in the feedback from gonads that impact on GnRH neurons (Hashizume et al., 2010). Evidence suggests that a decrease of negative feedback of estradiol on KISS1 expression is related to an increase of pulsatile release of LH in sheep during sexual maturation (Redmond et al., 2011). In ovariectomized (OVX) cows, Kisspeptin can stimulate LH and GH releases (Whitlock et al., 2008). Studies in human also noted that an inactivating mutation of TAC3 (which encodes neurokinin B) and TACR3 (neurokinin B receptor) leads to pubertal failure (Topaloglu et al., 2008). In short, these neuropeptides regulate puberty.

The trans-synaptic inhibitory control of GnRH release comprises three different neuronal subsets, including GABAergic neurons, opiatergic neurons and RFamide-related peptide gene (RFRP) neurons (Terasawa &Fernandez, 2001; Ebling &Luckman, 2008; Gibson et al., 2008; Ducret et al., 2009; Tsutsui et al., 2010). Among these three neuronal subsets, the inhibitory action of RFPR neurons is the simplest since it uses two peptides (RFRP1 and RFRP2) acting on a single receptor (GPR147 or GPR74), which is expressed in GnRH neurons (Hinuma et al., 2000; Ducret et al., 2009). The action of GABAergic neurons is more complex. The GABAergic neurons can inhibit the secretion of GnRH indirectly by acting on neurons connected to the GnRH neuronal network or can stimulate GnRH neurons directly by activation of GABAA receptors (Terasawa &Fernandez, 2001; DeFazio et al., 2002; Moenter &DeFazio, 2005; Ojeda &Skinner, 2006). The inhibition of opiatergic neurons may be exerted indirectly on the functional stimulatory neurons such as KNDy neurons (which produced kisspeptin, neurokinin B and dynorphin) that are part of the GnRH neuronal network or directly on GnRH neurons (Dudas &Merchenthaler, 2006; Navarro et al., 2009). Direct and indirect, excitatory and inhibitory inputs constitute the complex neuronal signalling driving GnRH neuronal development and hormonal release (Figure I.3.).

In addition to neurons, the initiation of puberty by GnRH release is also driven by glial cells (Ojeda et al., 2010b). Glial cells contribute to the puberty process via two mechanisms. One mechanism consists of growth factors such as the transforming growth factor-beta (TGFβ),

the epidermal growth factor (EGF) family, basic fibroblast growth factor (bFGF) and insulin- like growth factor 1 (IGF1) (Prevot, 2002; Mahesh et al., 2006). These growth factors released by glial cells stimulate the release of GnRH or to promote the differentiation and survival of GnRH neurons (Ojeda et al., 2010b). The other mechanism utilizes three systems including synaptic cell adhesion molecule 1 (SynCAM1), receptor-like protein tyrosine phosphatase-β (RPTPβ) and the neural cell adhesion molecule NCAM for plastic rearrangements of glia-to- GnRH neurons adhesiveness (Parkash &Kaur, 2005; Parent et al., 2007; Sandau et al., 2011a; Sandau et al., 2011b).

The control centre of appetite in the hypothalamus has been indicated as another pathway influencing pubertal onset (Cardoso et al., 2014b; Cardoso et al., 2015). These pathways include orexigenic peptides such as neuropeptide Y (NPY) and agouti-related peptide (AgRP) and anorexigenic peptides such as proopiomelanocortin (POMC) and cocaine- and amphetamine-regulated transcript (CART). A study by Wu and colleagues (2012) explored that ablation in NPY/AgRP of obese and infertile ob/ob mice restored fertility in both sexes and resulted in a prolonged period of food intake reduction. Other studies also reported the inhibitory effect of NPY on the secretion of GnRH/LH in ewe and ovariectomized rats (Barker- Gibb et al., 1995; Xu et al., 2009). The POMC/CART neurons express LEPR, NPYR and insulin receptor (IR) suggesting its ability to sense central and peripheral signals involved in metabolism and nutrition (Roa, 2013).

Figure I.3 A schematic illustration neuropeptide regulation of GnRH neuronal function (Ojeda et al., 2010a)

I.3.2. Nutrition and metabolic regulation of puberty

Reproduction and growth are energy intensive processes that require a high level of energy input. Body energy reserves are a key determinant for pubertal onset. In cattle, the proportion of mature body weight at puberty is highly conserved at about 55% and 60% of the mature weight in Bos taurus and in Bos indicus, respectively (Freetly et al., 2011). To manage heifers for proper growth as well as normal pubertal development, heifers need adequate and balanced diet (including vitamins and essential nutrients) from birth to puberty. In other words, nutrition-hormone interaction during all phases of a growth plays a critical role in the prediction and control of metabolic condition and pubertal development. Thus, many studies have focused on nutritional changes and their impact on age at puberty, as discussed in this section. Nutritional regulation of reproductive neuroendocrine axis and pubertal development was demonstrated in a series of studies. For example, beef heifers maintained on low energy diet that limited weight gain reached puberty later than those fed high energy diet for greater weight gain (Rathbone et al., 2001; Nogueira, 2004). Furthermore, the offspring of heifers fed a low energy diet approached puberty 19 days later than those maintained on high energy diet (Corah et al., 1975). Insufficient nutrition considerably delays puberty in both Bos indicus and Bos taurus cattle (Abeygunawardena &Dematawewa, 2004). Management strategies to “boost” heifer nutrition can lead to earlier puberty onset. The current hypothesis is that one of the ways in which nutrition affects the initiation of puberty is by influencing the increase of LH release. Heifers fed adequate energy diet for growth presented increased LH secretion and attained puberty earlier; in comparison to heifers fed a restricted diet that failed to present increased LH pulse frequency (Day et al., 1986). Likewise, dietary energy restriction inhibited the pre-pubertal increase in LH secretion in ovariectomized, ovary intact and estradiol-treated heifers (Kurz et al., 1990). In addition, studies in female sheep have shown that growth-restricted conditions can inhibit GnRH pulse secretion (I'Anson et al., 2000). Metabolic signalling molecules like metabolic hormones (e.g., insulin, IGF1, GH, leptin, adipokines and ghrelin) and amino acids, glucose and fatty acids are the proposed mediators between nutrition and GnRH release (Schneider &Wade, 1989, 1990; Boukhliq &Martin, 1997; He et al., 1999; Rodriguez et al., 1999; Bruning et al., 2000; Tanaka et al., 2000; Oltmanns et al., 2001; Tena-Sempere, 2008; Hiney et al., 2009; Hiney et al., 2010). Regulatory activity of metabolic hormones and changes in nutritional status can affect kisspeptin, NPY and POMC neuronal networks that contribute to GnRH neuron inputs (Amstalden et al., 2014). There is a reduction of hypothalamic KISS1 expression in restricted

feeding and fasting animals such as rodents, nonhuman primates and sheep (Castellano et al., 2005; Luque et al., 2007; Backholer et al., 2010b; Wahab et al., 2011). Further evidence for the link between metabolic hormones (IGF1 and LEP) and neuropeptides (KISS1, NPY and POMC neurons) that work together to influence GnRH signalling will be discussed in next sections. I.3.2.1. IGF1 as a metabolic signal for the onset of puberty The liver’s most noticeable modulation of nutritional effects on cow reproductive efficiency occurs via the synthesis and secretion of insulin growth factor I (IGF1) (Hiney et al., 1996; Zulu et al., 2002a; Zulu et al., 2002b). Zulu et al (2002a) noted that either the level of nutrient intake or body condition could affect liver function and IGF1 secretion. Particularly, dairy cattle with higher body condition at the age of calving present a deeper and longer negative energy balance period, compromised liver function, low level of serum IGF1 and consequently suppressed reproductive efficiencies such as inactive ovaries or persistent corpus luteum (Zulu et al., 2002a). The role of IGF1, secreted mainly by the liver, in cattle reproduction will be further explored.

The expression of IGF1 receptor in several reproductive tissues (Table I.1) suggested its role in the reproductive system (Daftary &Gore, 2005). Several studies reported that plasma IGF1 concentration is related to nutrient intake and body condition (Yelich et al., 1996; Zulu et al., 2002a) and low serum IGF1 concentration is correlated with delayed puberty (Granger et al., 1989) and with an extended postpartum interval in heifers (Rutter et al., 1989). The association between serum concentrations of IGF1 with body weight were established in heifers with the estimated correlation ranging from 0.88 to 0.92 (Lammers et al., 1999; Lacau- Mengido et al., 2000). Several publications reported the increase of plasma IGF1 concentration in heifers, rodents, humans and even in cloned calves as puberty approached (David J et al., 1987; Jones et al., 1991; Adam et al., 1995; Hiney et al., 1996; Yelich et al., 1996; Garcia et al., 2002; Govoni et al., 2002; Maciel et al., 2004). Though, Garcia et al. (2003) stated the decrease of circulating IGF1 concentration in the last 10 weeks preceding puberty in beef heifers of a 20-week sampling period; the metabolic status, sampling interval, duration of sampling before puberty and breed differences should be taken into account for this contrasting result (in comparison to the above-cited work). Remarkably, the concentration of IGF1 in blood is heritable in cattle, with estimate ranging from 0.23 to 0.52 (Davis &Simmen, 1997; Grochowska et al., 2001; Davis &Simmen, 2006; Swali &Wathes, 2006). Moreover, circulating IGF1 has been utilized as a predictor of reproductive success due to its association

with age at first calving (Yilmaz et al., 2006; Brickell, 2007), twin ovulations (Echternkamp et al., 2004) and conception rate to first service (Patton et al., 2007). However, recently, there is no clear report of utilizing circulating IGF1 as an accurate indicator of cattle puberty in commercial operations. All in all, serum IGF1 level is highly variable before puberty and its action on reproductive function requires further investigation.

Table I.2 Location of metabolic regulators in Hypothalamus and Pituitary of several species

Species Pituitary Hypothalamus Reference Sheep Ob-Rb Ob-Rb (Dyer et al., 1997; Adam et al., 2000; IGF1, IGF1R Iqbal et al., 2000) IGFBPs

Cattle Ob-Ra & Ob-Rb Ob-Rb (Ren et al., 2002; Chelikani et al., 2003; AdipoR1 & AdipoR2 AdipoR1 & AdipoR2 Yonekura et al., 2003; Tabandeh et al., 2011) Pig Ob-Rb Ob-Rb (Lin et al., 2000; Lin et al., 2001; Xu et AdipoR1 & AdipoR2 al., 2004; Kiezun et al., 2013) IGF1R IGF1R Human AdipoR1 & AdipoR2 AdipoR1 & AdipoR2 (Psilopanagioti et al., 2009; Michalakis Ob-Rb &Segars, 2010) Rat Ob Ob (Goodyer et al., 1984; Werther et al.,

Ob-Rb & Ob-Ra Ob-Rb 1989; Guan et al., 1997; Zamorano et al., AdipoQ IGF1R 1997; Junko et al., 1999; Jin et al., 2000; AdipoR1 & AdipoR2 Morash et al., 2001; Yonekura et al., IGF1R 2003; Malagón et al., 2006; Francisca et al., 2007)

Mouse Ob Ob-Rb (Jin et al., 2000; Wilkinson et al., 2007) Ob-Rb AdipoR1 & AdipoR2 Note: Ob: Leptin gene; Ob-Ra: Leptin receptor short isoform; Ob-Rb: Leptin receptor long isoform; IGF1: insulin- like growth factor 1; IGF1R: insulin-like growth factor; IGFBPs: insulin-like growth factor binding protein, AdipoQ: Adiponectin, AdipoR1 and AdipoR2: adiponectin receptor

The circulating IGF1 concentration before puberty is affected by nutritional status (Granger et al., 1989). Dietary restriction in heifers was used as a model to investigate the nutritional effects on reproductive performance mediated via IGF1; well-fed pre-pubertal heifers attained puberty at an earlier age in comparison with feed-restricted animals (Schillo et al., 1992; Macdonald et al., 2005). Likewise, the association between growth performance and the rapid rise in IGF1 concentration was also observed in feed-restricted pre-pubertal heifers (Yambayamba et al., 1996). Moreover, the frequency of LH pulses was decreased with the decline in serum IGF1 levels in feed-restricted pre-pubertal heifers (Yelich et al., 1996; Amstalden et al., 2000). In addition, with the decrease of serum IGF1 and insulin, undernutrition also leads to a decline in circulating IGF binding proteins (IGFBPs) concentration (Webb et al., 1999). Due to the fact that IGFBPs transport and enhance the half-life of IGF1,

lower concentrations of IGFBPs caused by under-nutrition limit the IGFs availability and effects. Accumulated data suggest that IGF1 and IGFBPs are pivotal metabolic indicators involved in the initiation of puberty in heifers.

The potential sites of action and the precise mechanisms by which IGF1 influences puberty should be further elucidated. Several reports indicated that IGF1 may directly influence hypothalamic function (Hiney et al., 1991; Zhen et al., 1997; Anderson et al., 1999) and pituitary (Soldani et al., 1995; Wilson, 1995). For example, in vitro studies in the hypothalamic GT1 and NLT cells reported that GnRH cell lines are receptive to IGF1 (Ochoa et al., 1997; Kim et al., 1998; Anderson et al., 1999). A study in pre-pubertal female rats also suggested that IGF1 can activate the expression of the KISS1 gene in the brain (Hiney et al., 2009). Furthermore, the intraventricular injection of IGF1 to peri-pubertal or immature juvenile rat raises plasma LH levels (Hiney et al., 1996). The IGF1 was also noted to modulate LH release in pigs (Whitley et al., 1995). However, knowledge and confirmation of the effect of IGF1 on LH release in ruminants are still limited. Research has shown increased IGF1 concentration during the estrous phase in goats (Hashizume et al., 2000). Another study demonstrated a possible link between IGF1 concentration and E2 during estrus in cattle. In fact, IGF1 increases GnRH-stimulated LH release in cultured bovine AP cells, and the interaction between IGF1 and estradiol-17beta enhances the response to GnRH (Hashizume et al., 2002), the first finding to demonstrate the stimulatory effects of IGF1 on LH secretion in cattle.

In addition to the above evidence, the expression of hepatic IGF1 mRNA and blood IGF1 were increased with the activation of hepatic estrogen receptor alpha (ER) (Della Torre et al., 2011). Liver ER plays a pivotal role in the regulation of lipid metabolism and gluconeogenesis (Qiu et al., 2017). Recently, a study in mice hypothesised that liver ER is a critical sensor of nutrient availability which drives liver metabolism as well as reporting liver functions to the CNS (Benedusi et al., 2017). As a result, the liver has been shown as a key organ contributing to the integration of metabolic and reproductive activities. The growing evidence supported by this study targets the liver as a metabolic organ involving in the puberty process.

I.3.2.2. Leptin as a metabolic signal for the onset of puberty Leptin, an adipokine, secreted mainly by adipose tissue, is a key signalling molecule that links nutritional status to reproductive function. The expression of leptin (LEP/OB) in reproductive tissues was summarised in Table I.1. To date, many reports have shown an

increase in serum leptin concentration during puberty in heifers (Garcia et al., 2002), pigs (Qian et al., 1999) mice (Barash et al., 1996; Chehab et al., 1997), and humans (Matkovic et al., 1997). Evidence suggests that heifers that attained puberty showed an increase in serum LEP level and LEP gene expression, serum IGF1 level and body weight (Garcia et al., 2002). In fact, body weight is highly correlated with timing of the initiation of puberty and serum LEP concentration (R=0.85) (Garcia et al., 2002; Williams et al., 2002). In the contrast, short-term (48h) fasting decreased synthesis and secretion of LEP and frequency of LH pulses (Amstalden et al., 2000). Fasted beef heifers administrated exogenous LEP are capable of increasing LH concentrations (Maciel et al., 2004; Zieba et al., 2004). Furthermore, lack of LEP or leptin receptor (LEPR/OBR) can cause animals to fail to attain the initiation of puberty as well as stay sexually immature during their entire lives (Swerdloff, 1978). Early-maturing Bos indicus heifers have greater LEP expression by adipose tissue and lesser NPY receptor expression in the hypothalamus at a lower body weight (Vaiciunas et al., 2008). The known effect of LEP over puberty is the underpinning reason for targeting AT in this study.

It has been proposed that LEP acts on GnRH neurons directly, but until now, little co- expression between LEP receptor and GnRH neuron has been observed (Quennell et al., 2009). Moreover, the detection of leptin-induced STAT3 phosphorylation was not observed in GnRH neurons but was found in the anteroventral periventricular nucleus (AVPV) (Quennell et al., 2009). The AVPV region contains neurons (such as kisspeptin) activated during the pre- ovulatory surge (Smith et al., 2006). Studies in neuron-specific leptin receptor deleted mice found that fertility was not affected. About 70- 80% of knockout mice in this study can produce offspring (McMinn et al., 2005). Contrary to this study, Quennell et al. (2009) reported that most female and male neuron-specific leptin receptor knockout mice were completely infertile. Lastly, a single-cell RT-PCR was utilized to confirm the mRNA expression of LEPR in GnRH neurons, however, no GnRH neurons contained mRNA level of LEPR. In short, the effect of leptin on GnRH release is probably complex and indirect and likely modulated via a different pathway.

KISS1 neurons have been suggested as mediators for integrating metabolic signals to GnRH neurons. There is supporting evidence for this hypothesis. The expression of LEP, STAT3 and KISS1 neurons were detected within the arcuate nucleus (ARC) (Qiu et al., 2011; Quennell et al., 2011). For instance, the KISS1 expression in the ARC region is significantly less in ob/ob mice deficient leptin compared to wild-type mice (Smith et al., 2006; Quennell et al., 2011). Unfortunately, i.c.v injection of LEP into lean sheep for 72h rescued LH pulses

(Backholer et al., 2010b), but KISS1 expression was only partially restored (Backholer et al., 2010a). Similarly, rats treated leptin did not fully reverse the lactation-induced reduction in the expression of KISS1 mRNA (Xu et al., 2009). In addition, a study reported no impaired fertility in mice with a deletion of LEPR in KISS1 neurons (Donato et al., 2011). Although LEPR is identified in a sub-population of KISS1 neurons, the possible indirect mechanisms of LEP acting on KISS1 neuron remains to be clarified.

Quennell et al. (2009) stated that LEP exerts its effect on forebrain neurons which are essential for regulation of fertility and energy balance. Collective evidence suggests that NPY/AgRP and POMC/CART neurons are key mediators for LEP action upon GnRH neurons (Sahu, 1998; Iqbal et al., 2001; Czaja et al., 2002; Luque et al., 2007; Wu et al., 2012; Roa, 2013). To be more specific, NPY regulates the secretion of GnRH and LH in rodents (Kalra, 1993), ewes (Morrison et al., 2003), pigs (Barb et al., 2006) and cows (Thomas et al., 1999). Central administration of NPY can stimulate appetite in the pig (Barb et al., 2001) and ewe (Miner, 1992), which was reversed by intracerebroventricular injection of LEP in pig (Barb et al., 2006). AgRP peptides have been known to inhibit puberty and hold a pivotal role in the effect of leptin in female puberty (Sheffer-Babila et al., 2013). During negative metabolic conditions, the suppression of CART expression could lead to the inhibition of reproductive axis (True et al., 2013). The mRNA expression of POMC is decreased in leptin-deficient ob/ob mice and the administration of LEP can rescue this expression in control mice (Thornton et al., 1997). Furthermore, loss of LEP signalling in POMC neurons can impact on body weight homeostasis and induce synapsis plasticity (Balthasar et al., 2004; Chun &Jo, 2010).

Previous studies also noted that POMC/CART and NPY/AgRP cells can communicate with KISS1 cells (Backholer et al., 2010b). In vitro immortalized hypothalamic N6 cells treated with NPY induced KISS1 expression (Luque et al., 2007). In the ovine ARC, NPY fibers and POMC fibers were in close apposition with 13 to 30% and 30 to 40% of kisspeptin cells, respectively. This study also indicated that kisspeptin fibers were seen in close apposition to approximately 7% of NPY cells and 20% of POMC cells in ewes. In OVX ewes, i.c.v infusion of kisspeptin increased NPY mRNA expression, decreased POMC mRNA expression in the ARC and elevated LH concentration (Backholer et al., 2010b). However, the fertility of ob/ob mouse deficient NPY with LEP treatment was only partially restored indicating other neuronal system impacts on LEP action (Erickson et al., 1996). Consequently, other metabolic factors may have a crucial role in regulating the kisspeptin system to stimulate gonadotropin release.

The above data provides evidence to suggest direct and/or indirect function of IGF1 and LEP on the initiation of puberty. However, the precise mechanisms or other metabolic factors contributing to the integration of metabolism and reproduction need further elucidate. Targeting tissues related to metabolism such the liver and adipose tissue with modern research tools is likely to enhance our knowledge about the link between nutrition and the puberty onset.

I.3.3. Season affecting puberty

Cattle are not typically considered as a seasonal breeder, unlike horses or sheep. There is, however, evidence supporting seasonal variation in cattle reproductive performance (Hansen, 1985; Stahringer et al., 1990). Season seems to affect the reproductive function in Zebu cattle more than in Bos taurus breeds (Hansen, 1985; Rhodes et al., 1995; Tatman et al., 2004). Heifers born in spring reached puberty at a later age than those born in autumn (Rathbone et al., 2001). This study was in agreement with the findings of studies carried on Angus x Holstein heifers (Schillo et al., 1982; Schillo et al., 1983), while another study pointed out the opposite result (Peter &Ball, 1987). In addition, a few reports noted that heifers born early in spring were heavier and later at puberty than heifers born late in spring (Arije &Wiltbank, 1971; Swiersta et al., 1977). The initiation of puberty in heifers was delayed in the winter (Grass et al., 1982). All in all, most cows reached puberty during autumn, summer or subsequent spring. It is unclear also if season is an isolated effect or confounded with the availability and quality of feedstuffs, since pasture quality/quantity can vary with seasons (Van Beest et al., 2010). I.3.4. Genetic factors affecting puberty The variation in age at puberty between groups or breeds is determined by genetic factors and environmental factors in which genetic factors accounted for 50-80% in humans (Parent et al., 2003; Hodges &Palmert, 2007; Gajdos et al., 2010). In cattle, the heritability for age at puberty was estimated within the range of 0.1 to 0.6 (Cammack et al., 2009). Higher heritability estimates, closer to 0.6, were encountered in breeds that were not selected for this trait such as the Bos indicus breeds like Brahman (Johnston et al., 2009) and Nellore cattle (Nogueira, 2004). It has been widely accepted that the initiation of puberty is mediated by genetic networks which comprised various genes with small effects of each gene (Ojeda et al., 2006). Several genetic loci associated with age of puberty have been identified. Mutations influencing different genes such as GPR54 (de Roux et al., 2003; Seminara et al., 2003), GNRHR (Bedecarrats &Kaiser, 2007), tachykinin 3 (TAC3) and its receptor TACR3 (Topaloglu

et al., 2009) and KISS1 (Lapatto et al., 2007) lead to pubertal failure in mammals. Genes involved in IGF1 pathway (IGF1R, IGFBP2, IGFBP4, PIK3R1, EIF2AK3, IRS1 and GSK3B) were associated with heifer puberty in Brahman and Tropical Composite (Fortes et al., 2013b). The association between IGF1 and age of puberty in Angus male cattle was reported (Liron et al., 2012). In the Nellore breed, a SNP in FSHR was associated with early puberty phenotype (Milazzotto et al., 2008). A study of 1689 non-precocious and precocious Nellore heifers revealed that SNPs located within the PPP3CA and FABP4 genes that showed significant association with sexual precocity (Dias et al., 2015). A GWAS in human reported 106 genomic loci associated with age at menarche (Perry et al., 2014a). Figure I.4 illustrates several genes and their biological mechanisms for age at menarche in the HPO axis. In heifers, approximately 350 markers associated with age at first calving and age at puberty were retrieved (Hu et al., 2016). A study in Nellore cows revealed the association between U6 spliceosomal RNA and age at first calving (Nascimento et al., 2016). A comparison between genes involved in human puberty and gene located within cattle QTL associated with heifer puberty showed few common genes such as CHST8, GABRA1, LEP and PROP1.

Figure I.4. A schematic diagram presenting potential roles in Hypothalamus-Pituitary-Ovary axis of several genes and their biological mechanism for age at menarche (Perry et al., 2014a)

The precise mechanism underlying pubertal development perhaps can be explained by a hypothesis that at an early stage, puberty somehow is mediated by transcriptional/post- transcriptional repression genes that suppress activation of pubertal genes, and therefore, inhibit the pubertal process. From this hypothesis, it can be predicted that at early pubertal

development, the level of puberty-delaying gene expression is high, while the expression of puberty-inducing genes is minimal. This trend will be reversed with the increase of inducing (or activating) genes’ expression on the onset of puberty (Figure I.5).

Figure I.5 The hypothetical of transcriptional control puberty by repressor and activator, (adapted from Ojeda et al. (2010)) Ojeda et al., (2010, 2013) demonstrated that the Polycomb group (PcG) gene and lin- 28 homolog B (LIN28B) may function to inhibit the activation of puberty-inducing genes including KISS1, TAC3, TFT1 (transcription termination factor, RNA polymerase I), EAP1 (Enhanced At Puberty1) and OCT2. In addition, zinc-finger genes (ZNF) were also determined as transcriptional repressors (Urrutia, 2003; Vogel et al., 2006). Regulatory networks for puberty have been constructed using candidate genes within QTL regions associated with puberty in cattle (Fortes et al., 2010a; Fortes et al., 2011; Fortes et al., 2016b). The study by Fortes et al. (2011) suggested five transcription factors - HIVEP3 (human immunodeficiency virus type I enhancer binding protein 3), TOX (thymocyte selection-associated high mobility group box), EYA1 (EYA transcriptional coactivator and phosphatase 1), NCOA2 (nuclear receptor coactivator 2), and ZFHX4 (zinc finger homeobox 4) as regulators of bovine puberty, hubs in a regulatory network predicted in 2 tropical breeds of beef cattle. This study also suggested genes involved in cell adhesion, axon guidance, ErbB signalling and glutamate activity may affect GnRH secretion and thus impact on the onset of puberty. Another regulatory network proposed for cattle puberty from differential gene expression in hypothalamus and pituitary gland identified 5 key regulators: PITX2, FOXA1, DACH2, PROP1 and SIX6 (Cánovas et al., 2014a). Mammalian pubertal development is thus controlled at the transcriptional/post-transcriptional level by groups of genes and transcription factors (TF). The

relative importance of each gene or TF may depend on the species or breed under investigation.

To our knowledge, there is no study of gene networks involving in the puberty onset in Brahman heifers. Thus, continued work to identify candidate transcripts and proteins controlling puberty in Brahman, a pre-dominant Bos indicus breed in Australia, may provide not only specific biomarkers and pathways involved but also important foundation and hypothesis for the design of the next experiments which can help selection program for optimizing age at puberty.

I.4. Studying the animal transcriptome and proteome

As mentioned in Section I.3.4, both candidate gene approaches and GWAS have provided multiple loci as well as markers associated with heifer puberty. Because genomic regions associated with specific traits identified by GWAS comprise many genes, the determination of the causative gene or mutation underlying identified regions is relatively difficult. However, biology involves the interaction between diverse levels, from genes, through transcripts and proteins, and then to metabolites. Biological research targets not only molecules but also how these molecules interact. Omics technologies including transcriptomics and proteomics have become of mainstream applications to better understand animal physiology and production. Using these techniques, the functional status of a molecule can be identified. Basic concepts and applications of these two omics technologies in cattle will be explained in this section.

I.4.1. Animal transcriptome

One objective of the biological research is to detect all molecules as well as their interactions with others in biological processes such as growth and development. Genes may be expressed in specific tissues or only at certain physiological state in the animal life cycle making such determinations difficult. In addition, many genes have unknown or partially understood functions. Since the role of RNA is the key intermediate between genome and proteome, the identification of transcripts as well as the quantification of gene expression has usually been used in biological research. With its important advantages (discovery, sensitivity and range of expression), RNA sequencing (RNA-Seq) are now increasingly using for transcriptomic studies (Han et al., 2015). This technique was first applied in model organisms including Arabidopsis (Lister et al., 2008), yeast (Nagalakshmi et al., 2008) and mouse (Mortazavi et al., 2008), but has rapidly increased its popularity to a number of other organisms including human (Sultan et al., 2008) and bovine (Huang &Khatib, 2010). RNA-seq was

applied to study the transcription of genes into mRNA at a particular time, in a particular cell or tissue and in a particular state. Up- or down-regulated genes can lead to different levels of proteins as well as metabolites that result in phenotypic changes. Thus, improving our knowledge of the regulation of particular genes could aid insight into the biological function.

I.4.1.1. RNA sequencing workflow

A crucial pre-requisite for expression studies is to generate data which can elucidate the biological questions. This can be achieved by defining a suitable experimental design. Generally, expression studies compare expression levels between two experimental conditions. In the disease example, the two experimental conditions would be the healthy and diseased. Other experimental settings could include high production versus low production, high reproduction versus low reproduction or various time points as conditions. Using bioinformatics tools, the differentially expressed (DE) transcripts between contrasting experimental conditions and their biological functions can be revealed. In more detail, the RNA-seq technique is started by the conversion of mRNA to short cDNA with an oligonucleotide adapter attached to one or both ends. Using next-generation sequencing (NGS), each individual molecule is sequenced from one end, termed single-end sequencing, or both ends, termed pair-end sequencing. Millions of short sequence reads are typically ranged from 30 to 400 base pairs (bp) depending on the sequencing technology used. At present, several NGS platforms have been applied for RNA-seq including the Illumina GA IIx and HiSeq, Applied Biosystems SOLiD, single molecule real-time PCR PacBio, Roche 454 Life Science and nanopore technology-driven portable device MinION (Metzker, 2010; Chu &Corey, 2012; Han et al., 2015). Following sequencing, the raw sequencing data serves as starting material for differential expression and functional biology analyses. The quality of raw reads is first checked to guarantee the success of subsequent analyses. The raw reads that pass the quality control are either aligned to a reference genome or reference transcripts or assembled de novo when a reference genome is not available. The mapped reads of each sample are then indexed into different categories including exon-level, gene-level or transcript-level. The data are subsequently normalized and differentially expressed genes are assessed. Different tools for quality control, mapping and data normalization were recently reviewed (Conesa et al., 2016). Once the list of DE genes is generated, these candidate genes are often ranked based on statistical criteria to reflect their significance. Typically, genes having the smallest p-value are often considered as significant genes. From a biological perspective, genes interact with each

other in several ways. For example, the puberty process comprises many genes in which each gene has small effect. A small expression variation of a particular gene in which its protein holds a catalytic role may have remarkable physiological consequences. Therefore, in addition to DE analyses, transcriptomic studies analyse biological pathways and gene-gene interactions of these DE genes. Genes that are differentially co-expressed between different experimental conditions or groups can be regulators, and thus can explain the differences between phenotypes (Kostka &Spang, 2004; Hu et al., 2009; Amar et al., 2013). (Hudson et al., 2009). Taken all together, transcriptomic studies provide significant advances on the knowledge of molecular mechanisms underlying animal responses toward a specific state.

I.4.1.2. Transcriptome studies in cattle production and reproduction

Transcriptome analysis technologies are extensively utilized in farm animal studies. A survey in 2016 indicated the increase of 6-fold in the number of publications on animal transcriptomics since 2010 (Parreira &de Sousa Araújo, 2018). This report also mentioned that more than 38% of publications on animal transcriptomics (249 out of 644) provided knowledge of bovine physiology. Figure I.6 presents the number of publications on several farmed animal transcriptomics found in NCBI from 2006 to 2016. The high number of publications on cattle transcriptome highlight the importance of this species in animal production.

Figure I.6 Farmed animal transcriptomic studies published in NCBI between 2006 and 2016 To date, there have been a considerable number of studies of the bovine transcriptome. Studies have, for example, focused on muscle pathways affected by the production system (Guo et al., 2015; Tizioto et al., 2016) or fat deposition to improve meat quality (Wang et al., 2009a; Sumner-Thomson et al., 2011; Jin et al., 2012; Huang et al., 2017). The differences in

the transcriptome profile of residual feed intake were also revealed (Weber et al., 2016). In Nellore cattle, for instance, 8 key regulators for feed efficiency were identified (Alexandre et al., 2015). Using differential gene co-expression network approach, co-expressed genes were noted to associate with lipid metabolism and insulin responses. Another study in Nellore cattle proposed that hepatic metabolic adaptations to oxidative stress was a new pathway controlling feed efficiency (Tizioto et al., 2015). In a nutshell, transcriptomic studies of a single tissue for cattle production traits has been well established. Another area where transcriptomics is utilized is to compare animals that differ in reproduction and development (Vigneault et al., 2009; Cánovas et al., 2014a; Scolari et al., 2016; Baufeld et al., 2017; Capra et al., 2017; Gonella-Diaza et al., 2017). RNA-seq studies provided knowledge of bovine embryo transcriptome (Vigneault et al., 2009; Scolari et al., 2016) and as well used to identify DE genes associated with bovine ovulation rate (Kamalludin et al., 2017). Differentially expressed genes in bovine oviduct, matured oocytes, bovine endometrium, pre-implantation embryos and early bovine pregnancy have also reported using microarray studies (Ishiwata et al., 2003; Ushizawa et al., 2004; Misirlioglu et al., 2006; Minten et al., 2013; Bentley et al., 2015; Maillo et al., 2015; Maillo et al., 2016). These studies provided knowledge of metabolic mechanisms underlying studied phenotype and then could help develop strategies for animal selection. To our knowledge, only the study by Cánovas et al. (2014) was focused on differences in the transcriptome between pre- and post-pubertal state in Brangus heifers. This study provided a list of 1,515 differentially expressed (DE) genes and 943 tissue-specific (TS) genes within 17,832 genes expressed across 8 studied tissues (hypothalamus, pituitary gland, ovary, uterus, endometrium, muscle, liver and adipose). In addition, five transcription factors were suggested as key transcriptional regulators involving in puberty onset including PITX2, DACH2, FOXA1, SIX6, and PROP1. According to Cánovas et al. (2014), the largest overall number of differential gene expression among pre- and post-pubertal Brangus heifers was observed in adipose tissue (n= 316). A large number of differentially expressed genes indicated the potential link between adipose tissue and onset of puberty. In terms of a puberty related co-expression gene network in Brangus heifers, the study from Cánovas et al. (2014) noted that while liver experienced an abundance of connections to co-expression networks (40%), adipose tissue presented a low percentage of connections to co-expression networks (2%). This study also showed that the liver had the largest disappearance of connections via the onset of puberty, whereas the hypothalamus had the largest number of connections upon puberty. The above study supported

knowledge about genes, their interactions at the onset of puberty in Brangus cattle (3/8 Brahman, Bos indicus x 5/8 Angus, Bos taurus) with the limitation of animals (n =8). Therefore, other gene products have not identified yet could be involved in the onset of puberty in Bos indicus (Brahman) cattle suggesting another trial to confirm these data. As mentioned above in the Section I.2, decreasing the age at puberty to increase Bos indicus cow productivity is a worthwhile goal for management and breeding. Puberty is a complex process which involves many genes of small effect. To understanding how animals reach puberty, all levels of physiology can be examined using omics technologies, focusing on capturing the variation or difference at all biology levels. In general, the aim of the first part of this thesis is to determine whether there is any difference in gene expression between pre- and post-pubertal Brahman heifers using transcriptomic profiles of reproductive tissues (hypothalamus and pituitary gland) and tissue related to metabolism (liver and adipose).

I.4.2. Animal proteome

It is well established how information is transferred from DNA via mRNA to proteins. When comparing the number of genes to the number of proteins, it is clear that the number of coding genes is significantly lower than the number of proteins. Neither does the abundance of any particular mRNA directly relate to the abundance of the protein product due to different turnover rates. Genomic studies based on short-read DNA sequencing are thus often inadequate to fully discover the variety and abundance of protein products. Transcriptomic studies have helped to provide an integrated view of gene expression and phenotypes; however, these studies do not include information about mRNA splicing variants, post-transcriptional and post-translational modifications or protein synthesis and degradation. Further, the correlation between mRNA and protein abundance in complex samples is not well correlated (Anderson &Seilhamer, 1997; Gygi et al., 1999; Koussounadis et al., 2015). Several factors can influence the correlation between measured mRNA and protein levels including biological factors (post- transcriptional and post-translational processes) and methodological factors (failure to detect splice variant or post-translational modification, noise and experimental errors) (Maier et al., 2009; Kelemen et al., 2013; Yang et al., 2016b). In addition, both mRNA and protein expression information were required for better understanding of the control of gene network (Hatzimanikatis &Lee, 1999). As a result, the study of the protein expression in a biological sample will provide an integrated vision of the cellular processes in a given cell state. Proteomics provides tools to directly assess and assemble biomolecules that are responsible for a function of biological systems. In farm animals, it has been applied to solve scientific

problems such as animal health, growth rate, production and reproduction (Doherty et al., 2004; Picard et al., 2010; Martins et al., 2012; Souza et al., 2012).

I.4.2.1. Proteomics workflow

Mass spectrometry (MS) has been used to identify and quantify of proteins in any biological system. This technique provides both qualitative and quantitative information of biomolecules after their conversion to ions (Hachey &Chaurand, 2004; Guerrera &Kleiner, 2005; Han et al., 2008; Lindemann et al., 2017; Vidova &Spacil, 2017). Methods based on the MS couples MS instruments with multidimensional-electrophoretic (gel-based) or multidimensional-chromatographic nonelectrophoretic (gel-free) methods for separation and then identification of differentially expressed proteins (Figure I.7).

The gel-based method was started with the separation of protein mixtures by 2D difference gel electrophoresis (DIGE) or two-dimensional (2D) polyacrylamide gel electrophoresis (PAGE) which was followed by digestion in gel of visible protein spots and then analysis by MS (Henzel et al., 1993; Shevchenko et al., 1996). The identification of protein is based on peptide mapping of observed peptide mass information with protein database (Yates et al., 1993). However, the gel-based MS techniques are of limited utilise for highly alkaline, acidic and hydrophobic proteins due to finite limits to the isoelectric point, molecular weight range and hydrophobicity (Sato et al., 2002; Ferguson &Smith, 2003). Further, incomplete separation can result in overlapping spots which then influence the identification of protein using MS (Ferguson &Smith, 2003). The most significant disadvantage of gel-based methods is the difficulty in identifying gene products with low copy number such as TF which may play important roles in gene expression (Ferguson &Smith, 2003).

The coupling of liquid chromatography to mass spectrometry (LC-MS) is a gel-free approach which provides a powerful analytical technique for identification and quantification of proteins with high accuracy and reproducibility (Aebersold &Mann, 2003; Han et al., 2008; Walther &Mann, 2010). The combination techniques start with protease digestion (e.g., trypsin) of protein mixture to generate peptides. Extracted peptides are then separated using LC. Liquid chromatography separates peptides based on physical properties. After separation in the chromatography column, peptides are ionized and sprayed into the mass spectrometer which directly connected with a high-performance liquid chromatography column. Atmospheric pressure chemical ionization (APCI) and electrospray ionization (ESI) has been used as two standard equipment for ionization of peptides (Fenn et al., 1989; Ho et al., 2003;

Korfmacher, 2005). When charged peptides enter the mass spectrometer, the mass-to-charge ration (m/z) of all peptides is measured (MS1). Subsequently, the specific or abundant peptide ions are fragmented and m/z of these fragment ions are detected (MS2). Typically, quadrupole ion trap (QIT), triple-quadrupole (TQ), quadrupole time-of-flight (QTOF) and Fourier transform ion-cyclotron resonance (FTICR) were different types of mass analysers (Cotter, 1992; McLuckey et al., 1994a, b; Henchman &Steel, 1998). The m/z value and peptide fragment ion pattern of each peptide ion facilitate the confident of peptide identification in studied samples. The identified peptide sequences are then mapped to the protein database of studied organisms, and the signal intensities of fragment and peptide ions can be utilized for quantification. Subsequent function annotation using high-throughput bioinformatics tools is needed to understanding how proteome explains the phenotype.

Figure I.7 The fundamental proteomic workflow (Almeida et al., 2015)

I.4.2.2. Proteome studies in cattle production and reproduction

Given the economic importance of dairy farming, most bovine proteomic studies have been conducted to characterize bovine milk proteome or to identify differential protein expression during bovine mastitis (Baeker et al., 2002; Hogarth et al., 2004; O’Donnell et al., 2004; Boehmer et al., 2010a; Boehmer et al., 2010b; Danielsen et al., 2010; Boehmer, 2011). To a lesser extent, the proteome of the bovine mammary gland has also been applied (Beddek et al., 2008; Peng et al., 2008). The effect of peri-pubertal feeding on bovine mammary gland development was studied (Daniels et al., 2006). The experiment was conducted with two dietary treatment on heifers and reported 07 proteins as differentially abundance in mammary

glands of heifers with 200 and 350 kg of body weight. Mammary gland tissue of cows during lactation (day 120 and 151) was used to characterize enzyme involving in metabolic pathways for milk molecules synthesis (Beddek et al., 2008). Other bovine mammary proteomic study aimed to identify proteins affecting lipid metabolism (Peng et al., 2008). Proteomics studies in pregnant and lactating cows have been applied to identify differential protein expression in blood serum, foetal membranes, uterine tissue, liver, allantoic and amniotic fluid. (Pyo et al., 2003; Kim et al., 2005; Lippolis &Reinhardt, 2005; Cairoli et al., 2006; Klisch et al., 2006; Riding et al., 2008; Kuhla et al., 2009; Kim et al., 2010; Li, 2012; Munoz et al., 2012; Yang et al., 2012b; Kurpińska et al., 2014; Lee et al., 2015; Rawat et al., 2016). A proteomic study in uterine fluid collected on d7 or d15 of the oestrous cycle in beef heifers aimed to identify differential uterine proteins expressions as well as to demonstrate the relationship between uterine proteins and progesterone concentration (Faulkner, 2012; Faulkner et al., 2013). Studies confirmed the association between uterine protein changes and systemic progesterone concentration as well as the stage of the oestrous cycle. Bovine pregnancy-associated proteins were also reported by Pyo et al. (2003), Lee et al. (2015) and Rawat et al. (2016). A proteomic study on the uterine fluid of pregnant cows analysed the interactions between embryo and mother at the early stage of pregnancy (Munoz et al., 2012). Another uterine proteomic study conducted in infertile cows and identified a potential biomarker (a protein of mass lower than 10 kDa) of repeat breeding (Minhas V. &H.M., 2008). The liver is a metabolic organ involved in numerous metabolism process. The first bovine liver proteome maps were published by (Talamo et al., 2003) and (D'Ambrosio et al., 2005). 58 proteins and 71 proteins involved in energy generation and metabolism of lipids, carbohydrates, amino acids as well as in polypeptide folding, synthesis and structure have been reported by Talamo et al. (2003) and D'Ambrosio et al. (2005), respectively. A liver proteomic study in ketotic and normal cows provided insight into differential protein expressions as well as pathogenic effects of ketosis (Xu &Wang, 2008; Xu et al., 2008). A global proteomic study of fatty liver in dairy cows revealed novel proteins which associated with signalling for energy homeostasis (Kuhla et al., 2009). Mammary gland and liver metabolisms of lactating cows were confirmed by the 2DE-DIGE (Rawson et al., 2012). Rawson et al. (2012) noted that enzymes required for gluconeogenesis, urea synthesis and -oxidation were abundant in the liver than in mammary gland. Proteomic studies in adipose tissue in farm animals were reviewed by (Sauerwein et al., 2014). For example, several proteomic studies compared differential protein expression at

different fat depots in cattle (Rajesh et al., 2010; Romao et al., 2013). Romao et al. (2013) noted that differentially expressed proteins associated with lipid metabolism and energy production were more abundant in visceral fat than subcutaneous fat. Another adipose tissue proteomic study in dairy cows identified novel biomarkers associated with metabolic status and insulin resistance that could be utilized to define high-yielding dairy cows adaptation to peripartum metabolic stress (Zachut, 2015). It is likely that most of adipose proteomic studies aimed to understand proteins related to intramuscular fat deposition for marbling issue (Zhang et al., 2010; Zhao et al., 2010; Shen et al., 2012; Mao et al., 2016). In addition, one study revealed age effect on protein changes in bovine subcutaneous adipose tissue (Romao et al., 2014). All in all, compared with proteomic studies of mammary and liver in cattle production and reproduction, those of adipose tissue have been scant. Hypothalamus and pituitary are central to the reproduction system, but to date, proteomic studies in these two tissues have focused on protein changes in response to feed restriction in cattle (Kuhla et al., 2007; Kuhla et al., 2010). Recently, proteomic data of differentially expressed proteins (P-value < 0.05) between less fertile and control cow were published (Zachut et al., 2016a; Zachut et al., 2016b). The differentially expressed proteins included ITIH1, SERPINA1, TIMP2, C8A, HSPG2, F2, COL1A2 and IL1RAP. To date, most of reproductive proteomic studies have been conducted in human and mice. Further, human and mouse proteomic data are extensively deposited in repositories such as Peptide Atlas (http://www.peptideatlas.org/ repository), Global Proteome Machine Database (GPMDB, http://gpmdb.thegpm.org) and The PRIDE proteomics identifications (http://www. ebi.ac.uk/pride). With regard to the bovine proteome, only GPMDB and Peptide Atlas cover large collection of peptides and proteins (more than 20000 peptides). These data retrieved mainly from milk and mammary tissues. However, human materials used for proteomics are limited. In mice models, their sex organs are quite tiny. Moreover, species differences exist between mice and human. Agricultural or farm animals have larger reproductive organs and have more similar reproductive physiology with human. As a result, the study of changes in cattle reproduction on the protein levels can result not only in better understanding of the reproductive physiology of the animal but also could provide information for human reproductive physiology.

OBJECTIVES

It is evident from literature review that puberty is a complex trait, modulated by many genes and interacting factors. So far, the transcriptomic and proteomic studies conducted in cattle have identified candidate genes or proteins associated with meat quality, muscle growth, feed efficiency, lactation or milk production. Regarding beef cows in northern Australia, there has been no direct study using reproductive tissues (hypothalamus and pituitary) and metabolism-related tissues (liver and adipose tissue) to characterize the transcriptome and proteome of pre- and post-pubertal Zebu heifers. Studies conducted with reproductive tissues aimed to understand gene and protein expression changes in response to feed restriction. Omics studies of metabolism-related tissues mainly targeted meat quality traits. Therefore, there was a need to investigate these tissues in the context of puberty, an important component of beef cattle production in Australia. In short, the research described in this Thesis was the first to reveal the differences in transcript and protein expression patterns in pre- and post-pubertal Brahman heifers. We hypothesised that the transcriptome and the proteome would change with the onset of puberty. Transcriptomic and proteomic analyses, along with enrichment analyses for GO terms, KEGG pathways and the prediction of interaction networks were carried towards the objectives listed here: To identify differentially expressed (DE) genes and differentially abundant (DA) proteins in reproductive tissues (hypothalamus and pituitary) and metabolism- related tissues (liver and adipose tissue) between pre- and post-pubertal Brahman heifers. To predict gene co-expression networks for differentially expressed genes and regulatory transcription factors. To visualize protein-protein interaction networks for a deeper understanding of the changes in the proteome related to puberty. To estimate the correlation between mRNA expression and protein abundance To provide data of expressed transcripts and proteins into public repositories (FAANG and Pride) for future research aiming to improve the annotation of the bovine reference genome. The results of the transcriptome and proteome analyses formed this Thesis, divided into two parts: 1) Transcriptomic studies and 2) Proteomic studies.

SECTION II. RESEARCH CHAPTERS

PART 1 – TRANSCRIPTOMIC STUDIES

Chapter 1.1. Transcriptome analyses identify five transcription factors differentially expressed in the hypothalamus of post-versus pre-pubertal Brahman heifers

Manuscript published in Journal of Animal Science., Volume 94, Issue 9, September 2016, Pages 3693–3702

http://dx.doi.org/10.2527/jas.2016-0471

Abstract

Puberty onset is a developmental process influenced by genetic determinants, environment and nutrition. Mutations and regulatory gene networks constitute the molecular basis for the genetic determinants of puberty onset. The emerging knowledge of these genetic determinants presents opportunities for innovation in the breeding of early pubertal cattle. This initial chapter presents new data on hypothalamic gene expression related to puberty in Bos indicus (Brahman) in age and weight-matched heifers. Six post-pubertal heifers were compared to six pre-pubertal heifers using the whole-genome RNA-seq methodology for quantification of global gene expression in the hypothalamus. Five transcription factors (TF) with potential regulatory roles in the hypothalamus were identified in this experiment: E2F8, NFAT5, SIX5, ZBTB38, and ZNF605. These TF genes were significantly differentially expressed in the hypothalamus of post- versus pre-pubertal heifers and were also identified as significant according to the applied regulatory impact factor metric (P-value < 0.05). Two of these five TF, ZBTB38, and ZNF605, coded for zinc finger proteins, belonging to a gene family previously reported to have a central regulatory role in mammalian puberty. The SIX5 gene belongs to the SIX family implicated in transcriptional regulation of gonadotrope gene expression. Tumor-related genes such as E2F8 and NFAT5 are known to affect basic cellular processes that are relevant in both cancer and developmental processes. Mutations in NFAT5 were associated with puberty in humans. Mutations in these TF, together with other genetic determinants discovered previously could be used in genomic selection to predict the genetic merit of cattle (i.e. the likelihood of the offspring presenting earlier than average puberty for Brahman). Knowledge of key mutations involved in genetic traits is an advantage for genomic prediction because it can increase its accuracy. Keywords: gene expression, puberty, Bos indicus, transcription factors, hypothalamus

1.1.1. Introduction

Puberty is a developmental process regulated by interacting genes. Gene interactions associated with puberty are described as a regulatory gene network comprised of functional modules (Ojeda et al., 2006). The hypothalamus plays a pivotal role in the central control of reproduction in mammals. The hypothalamic release of pulsatile GnRH is considered the trigger for mammalian puberty because it initiates the release of LH and FSH; hormones required for gonadal activity and gametogenesis. In turn, pulsatile GnRH release is regulated

by a gene network, interacting signals and pathways relevant to achieving puberty (Ojeda et al., 2010a). Transcriptional regulation and epigenetic mechanisms are known to influence puberty progression (Lomniczi et al., 2015a; Lomniczi et al., 2015b). For example, the acetylation of the KISS1 promoter region can contribute to its expression and influence GnRH release (Tomikawa et al., 2012). The existence of these complex trans-synaptic pathways overlaid by regulation of gene expression makes the task of understanding which genes regulate puberty particularly challenging.

Phenotypes controlled by multiple genes, such as age at puberty, are by definition quantitative traits and can be targeted in breeding programs (Hill, 2010). Breeding cattle for younger age at puberty is desirable in Bos indicus cattle, which are generally older at puberty when compared with Bos taurus (Laster et al., 1976; Abeygunawardena &Dematawewa, 2004; Brito et al., 2004; Johnston et al., 2009). Efforts to breed cattle for early puberty are already part of breeding programs (MacGregor &Casey, 1999; Cammack et al., 2009; Johnston, 2014). Knowledge of specific genes, mutations and gene networks can be used to enhance breeding programs that use genomic selection (Snelling et al., 2013; Fortes et al., 2014; Pérez-Enciso et al., 2015). In this context, data on hypothalamic gene expression of post-pubertal versus pre- pubertal Bos indicus heifers were utilized to construct a co-expression gene network. This study has hypothesized that it is possible to identify differentially expressed genes in the comparison between post-pubertal and pre-pubertal heifers and to predict interactions between these genes and their regulators.

1.1.2. Materials and Methods

Animals and Samples

Management, handling, and euthanasia of animals were approved by the Animal Ethics Committee of The University of Queensland, Production and Companion Animal group (certificate number QAAFI/279/12). Twelve young Brahman heifers with clear phenotypic characteristics of Bos indicus cattle were sourced in October 2012 from 2 commercial herds in Queensland, Australia. They were unrelated heifers born during the wet season of 2011/2012 (<250 kg BW). Heifers were managed at the Gatton Campus beef cattle facilities of the University of Queensland (Gatton, QLD, Australia), where they grazed together in a pasture system.

The aim was to collect samples for comparison of 6 post-pubertal heifers, at the luteal phase, with 6 pre-pubertal heifers, which had not experienced a luteal phase, that were of

similar age and weight. Heifers were examined every 2 weeks for observation of pubertal development, from October 2012 to May 2013. Rectal palpation was performed and ovarian activity was observed using ultrasonography (HS-2000 (VET); Honda Electronics Co., Ltd., Aichi, Japan). Pubertal status was defined by the first presence of a corpus luteum (CL) observed with ultrasound. Timing of tissue harvest was based on the date of first CL observation. Heifers were euthanized 23 d, on average, after observation of the first CL. When a CL was observed, the post-pubertal heifer was identified and then paired with a pre-pubertal heifer, which was randomly chosen from the remaining heifers and processed on the same day. The heifers were weighed and body condition was scored (5-point scale) prior to tissue harvest. Heifers were euthanized in pairs by stunning with a non-penetrating captive bolt followed by exsanguination for the post-pubertal heifers on the luteal phase of their second oestrus cycle. The non-penetrating captive bolt methodology was used in preference to a penetrating bolt as it served to protect the integrity of the hypothalamus tissue, as previously noted (Cánovas et al., 2014a). BW and CS of all euthanized heifers were recorded so that we could compare the means and standard deviations of the pre- and post-pubertal groups using a t-test statistic.

The presence of a CL was confirmed on the ovary at tissue harvest, and a blood sample was collected for progesterone analysis. Progesterone concentrations in plasma extracts were measured by RIA at the Animal Endocrinology Laboratory of the University of Queensland (Brisbane, QLD, Australia). Progesterone concentrations in hexane extracts of the plasma samples were measured by RIA as previously described (Curlewis et al., 1985) except that progesterone antiserum C-9817 (Bioquest Ltd., North Ryde, NSW, Australia) was used. Extraction efficiency was 75% and the values reported herein were not corrected for these losses. The sensitivity of the assay was 0.1 ng/mL and the intra- and inter-assay CV was 5.0%.

After euthanasia, hypothalamic tissue harvest was approximately 1 cm3 of tissue ranging from the preoptic region to the arcuate nucleus, verified by anatomical landmarks such as the crossing of the optic chiasm and the mammillary bodies as previously described (Cánovas et al., 2014a). Samples were preserved by snap freezing in liquid nitrogen and kept at −80°C until RNA extraction. In total, 12 hypothalamic (from 6 post-pubertal and 6 pre- pubertal heifers) were separately processed for RNA extraction and sequencing.

RNA extraction and sequencing

Prior to RNA extraction, hypothalamic tissue was pulverized under liquid nitrogen and homogenized to form a uniform sample representative of the whole organ. Total RNA was

isolated from 25 mg of the homogenized samples from post- and pre-pubertal heifers using the RNeasy mini kit (QIAGEN Pty Ltd., Victoria, Australia) combined with Trizol and its recommended methodologies (Life Technologies Inc., California, USA). Total RNA was resuspended in ribonuclease-free ultrapure water and stored at -80ºC until further use. Ribonucleic acid concentrations were measured by Nanodrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Wilmington, DE) with an optimal 260/280 ratio between 1.8 and 2.1. Intact 28S and 18S rRNA subunit integrity was assessed with an Agilent Bioanalyzer (RIN ranges from 6.9 to 7.4); traces did not show DNA presence.

Library preparation and sequencing

From total RNA, TruSeq RNA sample preparation kit (Illumina, California, USA) was used for library preparation. The kit protocol includes messenger RNA (mRNA) purification using poly T oligo-attached magnetic beads. Purified mRNA was then fragmented and converted to cDNA, which was double-stranded, ligated to adapters and amplified with PCR to create the libraries (all performed as per kit protocol, following manufacturer’s instructions). Libraries were multiplex, 6 libraries per lane, and paired-end sequenced with an Illumina HiSeq 2000 analyzer. Sequence fragments were mapped to the annotated bovine reference genome (UMD3.1; release annotation 77; ftp://ftp.ensembl.org/pub/release-77/genbank/bos_taurus/) using the CLC Genomics Workbench software (CLC Bio, Aarhus, Denmark) with its default parameters for alignment, quality control and calculation of gene expression levels. The algorithm underpinning the CLC software is ERANGE (Mortazavi et al., 2008). In short, for the assembly procedure, the sequences were mapped to the reference genome accounting for a maximum of two gaps or mismatches in each sequence. Quality control (QC) analysis was performed using procedures described previously with the CLC Genomics Workbench software (Cánovas et al., 2013). This tool assesses sequence quality indicators based on the FastQC-project (http:// www.bioinformatics.babraham.ac.uk/projects/fastqc/). Quality was measured taking into account sequence-read lengths and base-coverage, nucleotide contributions and base ambiguities, quality scores as emitted by the base caller and overrepresented sequences (Cánovas et al., 2014b). All the samples analysed passed all the QC parameters having the same length (100 bp), 100% coverage in all bases, 25% of A, T, G and C nucleotide contributions, 50% GC on base content and less than 0.1% over-represented sequences, indicating a very good quality. Data were normalized by calculating the ‘reads per kilo base per million mapped reads’ (RPKM) for each gene (Mortazavi et al., 2008). To select expressed genes a threshold of RPKM ≥ 0.2 was used (Wickramasinghe et al., 2012). Only

genes with average RPKM ≥ 0.2 in at least one tissue were considered expressed and had its data used in subsequent analyses (Wickramasinghe et al., 2012; Cánovas et al., 2014a).

Differential gene expression

Genes differentially expressed (DE) in the hypothalamus of post-pubertal versus pre- pubertal heifers were identified with analyses of the RPKM values. Base-2 log transformation of the RPKM values was performed to avoid bias DE, especially for genes with fewer reads (Bullard et al., 2010). Base-2 log transformed RPKM values were then normalized using mixed models to increase the sensitivity to detect differential expression and co-expression. This normalization approach for transformed RPKM values was previously described (Reverter et al., 2005; Cánovas et al., 2014a).

In more detail, differential gene expression after puberty was calculated using a mixed model: �� = � + �� + �� + �� + �� where log 2-transformed RPKM (Yijkpt) was modelled as a function of the fixed effect of the i library (Li) and of the random effects of gene (Gj) and the interaction of gene x animal x physiological state x tissue (GAPTjkpt) for the i library and the j gene of the k animal (12 levels) in the p physiological state (with 2 levels) from the t tissue (with 7 levels). Finally, eijkpt represents the random residual term. It was possible to fit tissue as a random interaction with gene and animal because this study is part of a larger experiment where 7 tissue samples were evaluated per animal (hypothalamus, pituitary, ovaries, uterus, muscle, adipose and liver). Similarly, the library effect was not confounded with animal as there were 7 libraries per animal (one for each tissue). Solutions to these mixed models were estimated using VCE6 software (ftp://ftp.tzv.fal.de/pub/vce6/). Normalized expression values for each gene in each sample were estimated from linear combinations of the mixed model solutions for library, gene, animal and tissue. Normalized expression values were subjected to t-test to compare the average expression in post- versus pre-pubertal heifers and identify DE genes using as a threshold for significance of P-value < 0.05. This P-value threshold was considered in the context of the harsh normalization performed and the subsequent analyses that used the DE gene list as a starting point for further scrutiny.

Gene network prediction and key regulators

To construct a co-expression gene network, normalized gene expression values were used as input to the partial correlation and information theory (PCIT) algorithm of (Reverter &Chan, 2008). A partial correlation between two genes is the correlation between this particular pair of genes that is independent of a third gene. In brief, PCIT is a data-driven

approach that explores all correlations between possible triplets of genes before determining which pair-wise correlations are significant. First, correlations for all expressed transcription factors (TF) and all DE genes were estimated. Then, correlations deemed significant formed the connections between genes in the network, which was further pruned to contain only DE genes and top-ranking TF from the regulatory impact factor metric (RIF) described below (Reverter &Chan, 2008). The gene network was then visualized and analysed using Cytoscape (Shannon et al., 2003).

To identify regulatory elements, the AnimalTFDB bovine database (http://bioinfo.life.hust.edu.cn/AnimalTFDB/download_index?tr=Bos_taurus), which includes classification and annotation of genome-wide TF, transcription co-factors and chromatin remodeling factors were utilized. A RIF metric (Reverter et al., 2010) was applied to identify key regulators among hypothalamic expressed TF. The RIF was explored as two alternative measures: RIF1 and RIF2, calculated from the number of DE genes and the predicted interactions between TF and target DE genes as described previously (Hudson et al., 2009; Reverter et al., 2010). The two alternative measures of RIF explored were computed as follows:

�=n�� 1 ̂ 2 RIF1� = ∑ �̂� x �� x DW�� n�� =1

�=n�� 1 2 = ∑ PIF� x DW�� n�� =1 and

�=n�� 1 2 2 RIF2� = ∑ [(�1� x �1��) − (�2� � �2��) ] n�� =1

th where nde is the number of DE genes; �̂� is the estimated average expression of the j DE ̂ gene, averaged across the two conditions being contrasted (post- and pre-puberty); �� is the differential expression of the jth DE gene; and DW is the differential wiring (or connectivity) th th between the i TF and the j DE gene, and computed from the difference between r1ij and r2ij, the co-expression correlation between the ith TF and the jth DE gene in conditions 1 (post- puberty) and 2 (pre-puberty):

DW�� = �1�� − �2��

The calculation of RIF1 uses the concept of phenotype impact factor (PIF) defined for each DE gene and computed from the product of its average expression and its differential expression (Reverter et al., 2010).

In essence, the first metric, RIF1, captures TF showing differential connectivity to DE genes between the two pubertal states (i.e. qualitative changes in predicted TF-target gene interactions). The alternative RIF2 focuses on TF showing evidence as predictors of change in abundance of DE genes between post- and pre-pubertal heifers (i.e. quantitative changes in predicted TF-target gene interactions). Both RIF1 and RIF2 identified key TF from the DE genes. For comparisons between both RIF1 and RIF2, and across datasets, RIF measures were transformed to a z-score by subtracting the mean and dividing by the standard deviation (SD). Using a nominal P-value < 0.05, a TF was deemed as a key TF if either of the two RIF scores was higher than 1.96 standard deviation.

Functional enrichment analyses

Three lists of genes associated to puberty in B. indicus heifer emerged from above- mentioned analyses: 1) DE genes and regulatory elements; 2) Genes and TF that formed the predicted network and 3) Top ranking TF. These lists of genes were used as target gene lists (one at the time) to compare with a background gene list formed by all genes expressed in the hypothalamus. The target vs background lists comparison was carried out by uploading these lists into DAVID, used to perform functional enrichment analyses (Huang da et al., 2009). With DAVID, genes were annotated in terms of their known function, gene ontologies and pathways. The output of DAVID analyses is overrepresented pathways or ontologies associated with each target gene list. Significant results after Benjamini-Hochberg correction for multiple testing are reported.

1.1.3. Results

At tissue harvest, average serum concentration of progesterone was 0.4 ± 0.2 ng/mL for pre-pubertal heifers and 2.0 ± 0.7 ng/mL for post-pubertal heifers. This range of progesterone levels could be considered normal for the luteal phase of pubertal Bos indicus heifers, which are lower than that observed for cows or Bos taurus cattle (Sartori &Barros, 2011). The observation of 2 consecutive progesterone values higher than 1 ng/mL has been used as a criterion for puberty achievement in heifers (Lopez et al., 2006; Shirley et al., 2006). No significant difference in BW or BCS between post-pubertal and pre-pubertal heifers was

observed at euthanasia. Body weight averages were 363 (SD 38.62) and 338 kg (SD 54.17; P- value = 0.38), and BCS (scale 1–5) averages were 3.75 (SD 0.41) and 3.5 (SD 0.44; P-value = 0.18).

The RNA-Seq data from pre- vs. post-pubertal Brahman heifers were aligned and filtered as described in the methods section. The working pipeline for transcriptomic analysis is shown in Figure 1.1.1. An average of 61 million sequence reads was obtained for each individual sample in the hypothalamus. Approximately, 57% of the reads were categorized as mapped reads to the bovine reference sequence. Of the 24,596 annotated B. taurus genes; 16,810 genes were found expressed in the hypothalamus samples (RPKM ≥ 0.2), which accounted for 68% of the known genome. These mapped sequences were used in subsequent analyses. Sequence data is available through the Functional Annotation of Animal Genomes project (http://data.faang.org/home).

Figure 1.1.1 Pipeline for analysis of transcriptomic dataset

Identification of differentially expressed genes Just over 2% of expressed genes were differentially expressed in the post- versus pre- pubertal heifer comparison. Genes highly DE (|FC| >2 and P-value < 0.01) in the hypothalamus

are shown in Figure 1.1.2. Of the 392 DE genes, 173 were downregulated and 219 were upregulated after puberty. More information about DE genes in the hypothalamus of pre- and post-pubertal Bos indicus heifers can be found in Supplementary Table S1 in (Fortes et al., 2016a). This result represented the difference in gene expression between heifers at their second luteal phase (post-puberty) and heifers who never experienced a luteinic phase (pre- puberty). Therefore, gene expression differences in post- versus pre-pubertal could be influenced by the presence versus absence of progesterone feedback to the hypothalamus and reflect physiological changes due to progesterone signalling as well as those that are involved in bringing about puberty onset. Among the DE genes, 29 were transcription factors (TF), 7 were transcription co-factors and 4 were chromatin remodelling factors. These elements were further investigated in network analyses and taking into account the RIF metrics obtained for TF.

Figure 1.1.2 Volcano plots reveal genes that significantly differ between pre- vs. post-pubertal heifers in the hypothalamus. The x-axis represents the fold change and the y-axis represents statistical significance for each gene. Red indicates genes that significantly differ (P-value < 0.05) between 2 groups. Genes symbols are provided for genes with a |fold change| ≥ 2 and P-value ≤ 0.01. Identification of transcription factors

The RIF metric was applied to all 1,085 TF expressed in the hypothalamus, DE or not. These analyses identified 111 significant TF (P-value < 0.05). This TF list was significantly

enriched for two pathways: prostate cancer (P-value = 3.4x10-4) and pathways in cancer (P- value = 1.1x10-3). Of the 111 top-ranking TF, 29 were genes that code for TF of the zinc finger family: 15 ZNF genes and 14 others, including uncharacterized proteins with zinc finger core characteristics, as per AnimalTFDB. Top ranking TF in the hypothalamus of post- versus pre- pubertal Brahman heifers was presented in Supplementary Table S2 in (Fortes et al., 2016a). Five TF were significant according to both RIF and DE analyses (Table 1.1).

Table 1.1 List of five transcription factors (TF) were significant in two analyses: differential gene expression (DE) and regulatory impact factor metrics (RIF1 and RIF2, P-value < 0.05). TF ENSB Tag Description RIF1 RIF2 DE1 FC2 P (DE) E2F8 ENSBTAG00000017446 E2F 3.10 0.84 2.25 1.44 0.01 transcription factor 8 NFAT5 ENSBTAG00000013412 Nuclear factor -0.38 -2.00 -0.37 0.75 0.04 of activated T- cells 5, tonicity- responsive SIX5 ENSBTAG00000013346 SIX homeobox -0.19 1.96 0.28 3.82 0.05 5 ZBTB38 ENSBTAG00000040061 Zinc finger and 0.49 -2.31 -0.15 0.91 0.03 BTB domain containing 38 ZNF605 ENSBTAG00000005240 Zinc finger -2.44 0.34 -0.29 0.56 0.03 protein 605 1DE values represent the difference in expression values between post- and pre-pubertal heifers average gene expression values (normalized base-2 log transformed RPKM values). 2FC represents the fold change in gene expression (normalized base-2 log transformed RPKM values) between post- and pre-pubertal heifers.

Gene co-expression network A co-expression network was predicted from RNA-seq data, illustrating the complex nature of hypothalamic activity associated with post- versus pre-puberty differences. First, the network formed by correlations between all expressed TF and DE genes were analysed. Then the network that contained only DE genes, DE regulatory elements and top-ranking TF from the RIF analyses were examined (Figure 1.1.3). The genes in the network were not significantly enriched for particular functional pathways when interrogated using DAVID software. The functional gene annotation analysis nevertheless detected six genes in the network which participate in the cross-talk between

adiponectin, leptin and insulin signalling pathways: TNFRSF1A, RELA, NFKB1, PRKAG2, CPT1C, and LEPR (Figure 1.1.4). All of these genes were upregulated after puberty with the exception of CPT1C.

Figure 1.1.3 Co-expression network in the hypothalamus of post and pre-pubertal Bos indicus heifers. Predicted interactions between differentially expressed genes and top-ranking transcription factors. The colour gradient illustrates network degrees, from low (green) to medium (yellow) and high (red), based on the number of predicted interactions for each gene.

Figure 1.1.4 Adiponectin KEGG pathway with differentially expressed genes are marked by red stars (from analyses using DAVID (Dennis et al., 2003)).

1.1.4. Discussion

The number of genes expressed in the hypothalamus was 16,810 (genes with RPKM values greater than 0.2 on average in hypothalamus samples). Of note, the number of genes deemed expressed and its splice variants would vary with the use of newer and increasingly annotated versions of the bovine reference genome, which should be available in the foreseeable future (Andersson et al., 2015). To confirm “typical” hypothalamic gene expression from this list of expressed genes is challenging. A recent study of female cattle tissues found that less than 1% of annotated genes presented a restricted tissue-specific expression (McGettigan et al., 2016). The 10 most hypothalamic specific genes, according to McGettigan et al. (2016), were expressed in all samples of this study (RPKM > 0.2 in each sample) and did not differ between post- and pre-puberty. These 10 genes, including the most specific myellin proteolipid protein, might be considered constitutively expressed in hypothalamus since they could be confirmed by others and were independent of developmental status in Brahman heifers samples (P-value > 0.05). Of the 111 hypothalamus-expressed TF that were ranked most highly in the RIF analysis, 29 were genes from the zinc finger family: 15 ZNF genes and 14 others, including uncharacterized proteins. This result seems to agree with reported findings regarding the central role of zinc finger family members in the regulation of puberty onset (Lomniczi et al., 2015b). Given the tested contrast, the identified zinc finger molecules might play a role in hypothalamic response to progesterone feedback in heifers. Specifically, it is tempting to hypothesize that ZBTB38 and ZNF605 are downregulated after puberty as results of either pubertal development as a whole or simply as a result of progesterone feedback. In either case, the role for these two high-ranking and DE zinc fingers in heifer reproductive biology should be further investigated. This chapter provides new evidence for physiological mechanisms associated with Bos indicus puberty. The studied heifers formed a contrast between hypothalamus samples under progesterone influence post-puberty (second estrus cycle) and hypothalamus samples from heifers that had never experienced a luteal phase (pre-puberty). Given this contrast, it is likely that DE genes could be influenced by puberty onset and the presence versus absence of progesterone feedback in the hypothalamus. Progesterone feedback is an important step in pubertal development and peri-pubertal heifers tend to experience short luteal phases in the first cycle (Atkins et al., 2013). Other contrasts should be explored in future experiments for detailing gene expression profiles at additional time points. Future experiments could compare

samples from post-pubertal heifers at all phases of the oestrus cycle, compare gene expression from one estrous phase to another and compare each phase with pre-pubertal gene expression. These other contrasts would allow differentiating progesterone influence on gene expression from post- versus pre-puberty differences in the absence of progesterone influence. The only other RNA-seq study of heifer puberty investigated a contrast similar to the one presented here in terms of progesterone influence. Cánovas et al (2014a) reported 275 DE genes in the hypothalamus of pre-pubertal Brangus (3/8 Brahman x 5/8 Angus) heifers with low levels of serum progesterone (0.5 ± 0.3 ng/mL) compared to post-pubertal heifers with higher levels corresponding to CL presence at euthanasia (7.1±1.0 ng/mL). However, the pre-pubertal heifers in that study were much younger and smaller than the age-matched heifers in the current study. Comparing the Brangus heifers study to the current results of Brahman heifers, only two genes were DE in both breeds: BARX2 and VAX1. Breed differences, including Bos taurus and Bos indicus genetics and experimental design (euthanasias for post- and pre-puberty were not matched in the Brangus experiment) could contribute to these contrasting results. However, encouraging confirmation in terms of key TF emerged from both studies since 43 of 111 (~39%) top-ranking TF in Brahman were also deemed regulators in the Brangus network: BSX, DLX1, DLX5, DMRT2, E2F3, E2F7, EGR3, EGR4, ETS1, ETV6, FOXA1, FOXA2, HAND2, HOXD9, INSM1, IRX2, LHX5, LHX9, MESP2, NFAT5, NFKB1, NR1H4, NR5A2, OVOL1, OVOL2, PAX3, PAX7, POU4F2, POU4F3, PPARG, PROP1, RAX, SHOX, SIX3, SOX5, SP5, TAL1, TCF21, TFCP2L1, TSC22D3, TTF1, USF2 and WT1. Note that NFAT5 is one of the five DE and top-ranking TF. Understanding the specific roles of these potential regulators of heifer puberty required further investigation. For some of these genes, or at least gene families, previous reports provide clues as to their function in the context of puberty mechanisms. This is the case for PPARG and PROP1, genes of the SIX and E2F families, as well as some tumour related genes all alleged to participate in the cross-talk upstream of GnRH release, as discussed in the following paragraphs. Previous work derived gene networks for cattle puberty from genome-wide association studies (GWAS) using polymorphism co-association to predict gene interactions (Fortes et al., 2010a; Fortes et al., 2011). Some of the key regulators identified in those studies and postulated here were the same: PROP1 and PPARG. Despite being previously identified as key regulators of heifer puberty and having significant RIF scores (supplementary Table S2 in Fortes et al., 2016a), these TF were not central nodes in the co-expression network reported here because they were only connected to 7 or fewer genes. Central nodes in the network were connected to up to 54 genes.

Genes from the SIX family determine DNA binding specificity, mediate protein-protein interactions, were implicated in developmental processes and in the maintenance of differentiated tissue states (Boucher et al., 2000). In this study, SIX5 was DE and a top-ranking TF. In Cánovas et al (2014), SIX6 was deemed a key regulator. In both studies, SIX3 was identified as a potential regulator of heifer puberty. Although SIX5 did not affect fertility in knockout mice (Klesert et al., 2000), it is possible that members of the SIX family contribute to pubertal development. Knockdown and knockout experiments of SIX6 support roles for SIX3 and SIX6 in transcriptional regulation of gonadotrope gene expression and SIX3 and SIX6 have been shown to compensate functionally for each other (Xie et al., 2015). Compensatory roles and results presented herein might suggest that SIX5, a top TF in Brahmans, is more important in a Bos indicus background and could compensate for SIX6 detected only in Brangus. Meanwhile, SIX3 may play a role in both Bos indicus and Bos taurus cattle. The cross-talk between adiponectin, leptin and insulin pathways in the hypothalamus has long been discussed as an underlying mechanism that links nutritional status with puberty (Coope et al., 2008; Cardoso et al., 2015). In the current study, six DE genes including TNFRSF1A, RELA, NFKB1, PRKAG2, CPT1C, and LEPR were known to belong to all three pathways. It is likely that these genes are important factors in energy homeostasis, in the context of heifer puberty. Out of the six, only NFKB1 had been identified by the previous study (Cánovas et al., 2014a). It is noteworthy that these genes are also known to modulate NPY neurons postulated to affect GnRH release in association to leptin signalling in heifers (Cardoso et al., 2014b). Alleles of SIX6 and LIN28B associated with increased age at puberty in girls were also associated with taller adult height (Perry et al., 2014b). A confirmed mutation in PLAG1 associated with age at puberty in heifers follows the same pattern: the allele associated with late puberty increases height and weight (Karim et al., 2011; Littlejohn et al., 2012; Nishimura et al., 2012; Fortes et al., 2013a; Utsunomiya et al., 2013; Saatchi et al., 2014). It is conceivable that SIX5, SIX6 and SIX3 may present another example of genes which link growth and energy homeostasis to timing of puberty in cattle. Genes of the SIX family might act similarly to leptin and adiponectin: providing a permissive link between energy homeostasis and GnRH release (Amstalden et al., 2014; Cardoso et al., 2015). Elucidation of SIX pathways and how it may fit into the cross-talk upstream of GnRH release requires further investigation. Certain tumor-related genes are thought to be puberty-delaying genes, transcriptional repressors central to the regulatory network that controls puberty onset (Roth et al., 2007; Ojeda et al., 2010a). Genes from the E2F family might fall into this “tumor-related” category of

important TF for puberty. They are known to affect cell cycle and progression; basic processes that are relevant in both cancer and developmental processes (Christensen et al., 2005). The E2F8 and NFAT5 genes may fall into this category, as these rank within the five DE and top- ranking TF and have been shown to play roles in cancer development (Jauliac et al., 2002; Chen et al., 2009a). However, from current data E2F8 had higher levels of expression post- puberty, while NFAT5 had higher levels pre-puberty and so it is likely that their effects on puberty are in opposite directions. First described as a tonicity-responsive TF crucial to kidney function, NFAT5 is also expressed in brain and testicular tissues, however, the roles it may play in those tissues are largely unknown (Trama et al., 2000; Lopez-Rodriguez et al., 2004). It is worth noting that NFAT5 is the only one of the five DE and top-ranking TF which was identified by the previous Brangus heifer hypothalamus study (Cánovas et al., 2014a). Further, mutations in NFAT5 were associated with age at menarche (puberty) in women (Chen et al., 2012).

1.1.5. Conclusion

Knowledge of key regulators, specific genes, mutations, gene networks and pathways can be used to enhance genomic approaches for selective breeding (Snelling et al., 2013; Fortes et al., 2014). In this context, data presented herein regarding hypothalamic gene expression of post- versus pre-pubertal Bos indicus heifers could have practical implications. The regulators for which this study proposes a potential role in puberty should be mined for mutations which could be tested for their effect on age at puberty in cattle. Mutations in these regulators, together with other genetic determinants discovered in previous GWAS could form the basis for DNA diagnostic-tools predicting early puberty onset. Knowledge of biologically relevant mutations is an advantage for genomic selection: increasing the accuracy and supporting the prediction of phenotypes across breed (Snelling et al., 2013; Pérez-Enciso et al., 2015).

Chapter 1.2. Global differential gene expression in the pituitary gland of pre- and post- pubertal Brahman heifers

Manuscript published in Journal of Animal Science., Volume 95, Issue 2, February 2017, Pages 599–615

http://dx.doi.org/10.2527/jas.2016.0921

Abstract

To understand genes, pathways, and networks related to puberty, this project continued to characterize the transcriptome of another reproductive tissue: the pituitary gland. Samples were harvested from pre- and post-pubertal Brahman heifers (same age group). Brahman heifers (Bos indicus) are older at puberty compared with Bos taurus, a productivity issue. With RNA sequencing, this study identified differentially expressed (DE) genes and important transcription factors (TF) and predicted co-expression networks. The number of DE genes detected in the pituitary gland was 284 (P-value < 0.05), and VWC2L was the most DE gene (fold change = 4.12, P-value = 0.01). The gene VWC2L promotes bone mineralization through transforming growth factor-β (TGFβ) signalling. Further studies of the link between bone mineralization and puberty could target VWC2L. Pathways previously related to puberty such as TGFβ signalling (P-value = 6.71 × 10−5), Wnt signalling (P-value = 4.1 × 10−2), and peroxisome proliferator-activated receptor (PPAR) signalling (P-value = 4.84 × 10−2) were enriched in pituitary gland transcriptomic data set. Recent work of hypothalamic gene expression also pointed to zinc fingers as TF for bovine puberty. Although some zinc fingers may be ubiquitously expressed, the identification of DE genes in common across tissues points to key regulators of puberty. The hypothalamus and pituitary gland had eight DE genes in common. This study confirmed the complexity of puberty and suggested further investigation of genes that code zinc fingers.

Keywords: Bos indicus, gene expression, gene networks, pituitary gland, transcription factors, zinc fingers

1.2.1. Introduction

Puberty is a complex process initiated by the altered biphasic release of GnRH. The release of GnRH stimulates the secretion of LH and FSH from the pituitary gland. Via bloodstream, pituitary gland hormones bind to receptors in the gonads and stimulate ovulation and the production of the gonadal steroid hormones including progesterone, oestrogen, activins and inhibins (Day et al., 1984; Day et al., 1987; Schillo et al., 1992; Day, 1998; Gasser et al., 2006; Bliss et al., 2010). Gonadal steroids such as estrogen and progesterone are released to the bloodstream and then inhibit or stimulate the release of GnRH in a cyclic manner related to ovarian dynamics during the oestrus cycle (Bliss et al., 2010): this is a complex feedback loop that influences the progression of puberty. Estrogen also modulates LH and FSH release

via its receptors (Gharib et al., 1987; Shupnik et al., 1988, 1989), and both ER and ER are observed in the hypothalamus, pituitary gland and ovaries (Hatoya et al., 2003). Although these events are similar in Bos indicus and Bos taurus cattle, they are initiated later in Bos indicus. Puberty in B. indicus included an increase in LH release while FSH release did not change (Nogueira et al., 2003) and these hormonal patterns were similar in Bos taurus breeds (Evans et al., 1994). Furthermore, the study of 17β-estradiol negative feedback on LH release noted a decreased feedback at an older age in Bos indicus heifers than that of Bos taurus heifers (Rodrigues et al., 2002). The Bos indicus cattle are better adapted to tropical regions, and the difference of region adaptation seems to have resulted in reproductive function differences in the two breeds (Mezzadra et al., 1993; Wolfenson et al., 2000; Nogueira, 2004). However, knowledge of mechanisms controlling the onset of puberty in Bos indicus is still limited. Ribonucleic acid sequencing is an emerging technology to analyse global differential expression (Mortazavi et al., 2008; Wang et al., 2009b). This study used RNA sequencing of the pituitary gland to identify differentially expressed (DE) genes and pathways related to puberty in age- and weight-matched Brahman heifers. The difference of gene expression before and after pubertal development between Brangus heifers (Canovas et al., 2014a) and Brahman heifers was also reported.

1.2.2. Materials and methods

Animals and samples All procedures, including animal handling and euthanizing, were approved by the Animal Ethics Committee of The University of Queensland, Production and Companion Animal group (certificate number QAAFI/279/12). Twelve Brahman heifers used in this current work were the same as those utilized in the hypothalamic transcriptomic study (Chapter 1.1). Observation of luteal development was performed using ovarian ultrasound scans. Six pre-pubertal Brahman heifers (mean ± SD; body weight = 338 ± 54.17 kg, condition score = 3.5 ± 0.44 and average daily gain = 0.76 ± 0.1 kg/d) were randomly chosen from the group of animals that had never ovulated. Six post-pubertal Brahman heifers (mean ± SD; body weight = 363 ± 38.62 kg, condition score = 3.75 ± 0.41 and average daily gain = 0.89 ± 0.1 kg/d) were euthanized on the luteal phase of their second cycle. Tissue harvesting was executed as fast as possible (within 20 min after euthanasia) to preserve RNA quality. The pituitary gland tissues from each heifer were snap-frozen in liquid nitrogen after harvesting and then kept at -800C until RNA extraction.

Ribonucleic acid extraction Prior to RNA extraction, 12 pituitary glands were pulverized under liquid nitrogen and homogenized to form a uniform sample representative of the whole organ. Total RNA was isolated from fragmented frozen tissues (approximately 25 mg) using a combination of RNeasy and TRIzol methods. The quality of the total RNA was evaluated using the RNA integrity number value in an Agilent Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA). Values for RNA integrity number ranged from 8.2 to 8.9, indicating good quality of RNA samples. High-quality RNA samples were sent to the University of California, Davis, for library preparation and sequencing. Library preparation and sequencing

A TruSeq RNA sample preparation kit (Illumina Inc., San Diego, CA) was used to first purify the RNA; mRNA purification used poly T oligo-attached magnetic beads. Purified mRNA was then fragmented and converted to cDNA, which was double-stranded, ligated to adapters, and amplified with PCR to create the libraries (all performed per kit protocol, following manufacturer’s instructions). Libraries were sequenced with an Illumina HiSeq analyzer (Illumina Inc., San Diego, CA).

The bovine reference genome, also known as the UMD3.1 assembly (release 77; ftp://ftp.ensembl.org/pub/release-77/genbank/bos_taurus), was used as the reference genome for sequence assembly, which was performed using the CLC workbench software (CLC bio, Aarhus, Denmark). This software was also used for quality control of sequence data and calculation of reads per kilobase per million mapped reads (RPKM) per gene as described previously (Mortazavi et al., 2008). Only transcripts with RPKM ≥ 0.2 were considered expressed genes; otherwise, they were discarded from further analyses (Wickramasinghe et al., 2012; Cánovas et al., 2014a). Uniform RPKM values across tissues may serve to identify ubiquitous genes or tissue-specific genes, but these analyses should be performed across all studied tissues and are beyond the current study, which is focused on pituitary gland expression.

Differential gene expression

Differentially expressed (DE) genes in the pituitary gland of post- versus pre-pubertal heifers were identified based on normalized RPKM values. This study fitted a mixed model to the base-2 log transformed RPKM values as previously detailed (Chapter 1.1). In this mixed model, the library was fitted as a fixed effect, the main effect of genes was fitted as a random

effect, and tissue × gene × animal was fitted as a random interaction. This study is part of a larger experiment where seven tissues were sampled per animal (hypothalamus, pituitary gland, liver, ovary, uterus, fat and muscle). The library effect was not confounded with animal as there were 7 libraries per animal (1 for each tissue). The VCE6 software (ftp://ftp.tzv.fal.de/pub/vce6; accessed June 15, 2015) was performed to estimate variance components and to obtain solutions of the mixed model as described previously (Chapter 1.1). A t-test was used to test the hypothesis that the differential expression in post- versus pre- pubertal heifers was significant, and the P-value of <0.05 was used as the threshold to determine DE genes. This nominal P-value of <0.05 threshold was used in context with the strict normalization performed and the subsequent analyses, which used the DE genes for further scrutiny.

Key regulators and gene network prediction

To determine the regulators in pituitary transcriptome data set, the AnimalTFDB bovine database (http://bioinfo.life.hust.edu.cn/AnimalTFDB/download_index?tr=Bos_taurus) was used. A regulatory impact factor (RIF) metric was performed to identify key regulators among TF expressed in the pituitary gland as described previously (Hudson et al., 2009; Reverter et al., 2010). Using a P-value cutoff of 0.05, a TF was considered a top-ranking regulator if either of the two RIF scores were higher than 1.96 SD units.

The partial correlation and information theory (PCIT) algorithm was performed as described previously to detect the association between DE genes and all expressed TF in a co- expression gene network (Reverter &Chan, 2008). Using Cytoscape (Shannon et al., 2003), co-expression networks that combined pre- and post-pubertal data to predict gene interactions the pituitary gland, were visualized and analysed.

Functional enrichment analyses

Gene ontology (GO) enrichment and pathways analysis were performed with three types of gene lists: 1) DE genes from the pituitary gland, 2) top-ranking TF from the pituitary gland, and 3) genes and TF that formed the predicted networks for the pituitary gland. These lists of genes were used as target gene lists (one at a time) to compare with a background gene list formed by all genes expressed in these tissues. The annotation terms enriched for these lists of genes were identified using a DAVID functional annotation chart (Dennis et al., 2003; Huang da et al., 2009). Significant enrichment results after Benjamini– Hochberg correction for multiple testing are reported.

1.2.3. Results

An average of 45 million sequence reads was obtained for each individual sample from the pituitary gland. Approximately, 61% of the reads were categorized as mapped reads to the Bos taurus genome reference. These mapped sequences were then subsequently utilized for down-stream analyses. Sequence data is available through the Functional Annotation of Animal Genomes project (http://data.faang.org/home). Identification of differentially expressed gene The number of DE genes detected in the pituitary gland tissues was 284. Of these 284 DE genes, 165 genes were down-regulated and 119 genes were up-regulated after puberty. Supplementary Table S1 in Nguyen et al. (2017) provided more information on these DE genes. Among the pituitary gland DE genes, there were 12 TF, 3 transcription co-factors and 1 chromatin remodelling factors observed. Genes highly differential expressed (|FC| 3 and P- value 0.01) in the pituitary are shown in Figure 1.2.1. For example, VWC2L (part of the TGF pathway) was among the DE genes with higher FC and its potential role in puberty is further explored in the discussion section.

Figure 1.2.1 Volcano plots reveal genes that significantly differ between pre- vs. post-pubertal heifers in the pituitary gland. The x-axis represents the fold change and the y-axis represents statistical significance for each gene. Red indicates genes that significantly differ (P-value < 0.05) between 2 groups. Gene symbols are provided for genes with a |fold change| ≥ 3 and P-value ≤ 0.01.

Enrichment analyses of GO terms were performed using the 284 DE genes as the target gene list and the genes expressed in the pituitary gland (16,927 genes) as a background list. In the pituitary gland, no GO terms or pathways were enriched after correction for multiple testing. Only one keyword term approached significance “secreted”, with 22 genes associated with this term (P-value = 0.06) (keywords were defined by the SwissProt/UniProt and Protein

Information Resource). The enrichment for the term “secreted” should not be surprising considering that this is an active gland. Using the RIF metrics, this study identified 119 significant top-ranking TF (P-value < 0.05) in the pituitary gland. List of these 119 TF, their symbols, description and RIF scores were presented in Supplementary Table S3 in Nguyen et al. (2017). Of these 119 TF, 27 genes coded for TF of the zinc finger family. The transcription factors GATAD2B, BARHL1, SP5 and ZNF8 had the highest scores in RIF1 (>3 SD units). All of these high scoring TF were less expressed in post-puberty heifers (P-value < 0.05). Pathway analysis for 119 TF suggested these TF were enriched for 7 pathways, all related to cancer, including endometrial cancer and leukaemia. Co-expression networks were predicted with PCIT using the expression values for DE genes and all expressed TF in the pituitary gland (1,085 TF). The predicted network from pituitary gland results had 7,403 connections for 1,284 genes. (Figure 1.2.2). In this network, LIM homeobox 1 (LHX1), Forkhead box F1 (FOXF1) and a novel gene (uncharacterized protein - ENSBTAG00000033642) were the only three genes that were simultaneously top- ranking TF (RIF score > 1.96) and DE (P-value < 0.05). Tools in DAVID functional annotation chart (Huang da et al., 2009) were used to identify enriched pathways from the genes in the predicted pituitary gland network. Many pathways related to reproduction were identified including TGFβ signalling (P-value = 6.71x10-5), Wnt signalling (P-value = 4.1x10-2), and PPAR signalling (P-value = 4.84x10-2, corrected P- values). Supplementary Table S5 presented in Nguyen et al. (2017) provided further information of these enriched pathways. The link between these pathways and reproduction is further deliberated in the discussion section.

Figure 1.2.2 Predicted co-expression gene networks comprising pre- and post-pubertal data in the pituitary gland. Each node represents a gene or a transcription factor (TF). Nodes represented as triangles are TF and ellipses are all other genes. Edges represent significant predicted interaction between nodes. Node colour indicates the number of connections of a specific gene in the network. The

colour spectrum ranges from green (low) to red (high) number of connections. Node size indicates the relative amount of expression in the post-puberty sample.

Comparing the pituitary gland transcriptome between Brahman (current study) and Brangus heifers (Cánovas et al., 2014a), 11 genes were DE in both breeds (P-value < 0.05): GPX6, DMBX1, IYD, C4BPA, MAG, LHX1, EXPI, LPO, SAA3, P2RY8 and a novel gene (uncharacterized protein). Out of these 11, only 4 genes had the same expression pattern in both Brahman and Brangus, similarly up- or down-regulated after puberty and these were: GPX6, IYD, EXPI, and SAA3. The other 7 genes were reversed in their expression pattern: instead of up-regulated, they were down-regulated in post-puberty heifers or vice-versa. These genes had a |FC| > 1.87 in both breeds, except only for EXPI, which had a FC = -1.27 in Brahman heifers and a FC = -0.85 in Brangus heifers. Out of 7 genes that were different with regards to the expression pattern in two breeds, 5 were down-regulated in Brangus and up- regulated in Brahman heifers post-puberty (Table 1.2). In addition, 8 genes were also revealed as DE genes in the hypothalamic transcriptome (Chapter 1.1) and in the pituitary gland transcriptome (current chapter) of pre- and post-pubertal Brahman heifers Brahman heifers. Again, the IYD genes were identified as common DE genes in the hypothalamus and the pituitary gland in Brahman heifers. Table 1.2 Genes in common in pituitary transcriptome profiles between Brahman and Brangus heifers ENSB tag1 Gene2 FC3_BRAN** FC_BRAH*

ENSBTAG00000017044 MAG 2.605 -3.443 ENSBTAG00000033562 LHX1 1.873 -2.093 ENSBTAG00000009625 DMBX1 -2.289 2.072 ENSBTAG00000047839 P2RY8 -1.749 2.689 ENSBTAG00000009876 C4BPA -2.251 3.005 ENSBTAG00000012780 LPO -2.016 3.262 ENSBTAG00000031825 Uncharacterized protein -2.543 3.532 ENSBTAG00000012023 GPX6 -2.469 -3.264 ENSBTAG00000025260 EXPI -1.274 -0.856 ENSBTAG00000015935 IYD 2.241 2.451 ENSBTAG00000022396 SAA3 3.789 3.124

1ENSB tag = Ensembl gene identifier according to http://uswest.ensembl.org/index.html. 2Gene symbol related to the ENSB tag. 3FC = fold change ** P-value < 0.01 * P-value < 0.05

1.2.4. Discussion

The pituitary gland had 284 DE genes in Brahman heifers compared to 292 DE genes in Brangus heifers. Only 11 genes were DE in both breeds (P-value < 0.05): GPX6, DMBX1, IYD, C4BPA, MAG, LHX1, EXPI, LPO, SAA3, P2RY8 and a novel gene (uncharacterized protein). Out of 11 genes, 4 had the same expression pattern in both Brahman and Brangus: GPX6, IYD, EXPI, and SAA3. In pigs, GPX6 was proposed as a peripheral blood marker for early pregnancy detection (Shen et al., 2014). The early pregnancy scenario may be similar to post-puberty conditions in the luteal phase experienced by the Brahman and Brangus heifers. The gene IYD was also observed as a DE gene in the hypothalamus of Brahman heifers (Chapter 1.1) as well as the pituitary gland of Brahman (current study) and Brangus heifers (Cánovas et al., 2014a). Known to be involved in thyroid hormone biosynthesis, IYD’s role in reproductive tissues is less clear. The gene EXPI expressed in the uterus of mice was related to pregnancy success (Nuno-Ayala et al., 2012). Its expression in the pituitary gland is reported here and in the Brangus study (Cánovas et al., 2014a), but no further literature is available to illustrate EXPI potential role in this tissue. Genes of the SAA family, like SSA3, were expressed in human pituitary gland (Urieli-Shoval et al., 1998), although its known role is in the immune system. Despite differences in the experimental design and number of studied animals in two experiments, expression of these common DE genes was supported by both studies and reversed DE genes could serve as a starting point to elucidate the differences in pubertal development between two breeds. The most DE genes (|FC| > 4) in the pituitary gland of Brahman heifers were VWC2L (FC = 4.12 and P-value = 0.01) and an uncharacterized protein (ENSBTAG00000036258 or SMIM10L1, FC = 4.09 and P-value = 0.01). The gene VWC2L is known to be expressed in the brain, but its function is better understood in bones, where its expression was associated with osteoblast matrix mineralization (Ohyama et al., 2012). The effects of VWC2L might be through members of the transforming growth factor-beta (TGF) superfamily, such as the BMP it antagonizes (Miwa et al., 2009). In rats, TGFβ1 induces the expression of GnRH (Srivastava et al., 2014). There is strong evidence for TGFβ superfamily signalling in the pituitary gland (Qian et al., 1996; Bilezikjian et al., 2006). In humans, an association between bone mineralization and pubertal development has been observed (Eastell, 2005). A functional role for VWC2L in cattle puberty, potentially via TGFβ signalling in the pituitary gland, is a new hypothesis emerging from this current study. Further investigation into VWC2L may help to elucidate the mechanisms that link bone mineralization and pubertal development in cattle and

serve for comparisons to other mammals. The role that SMIM10L1, small integral membrane protein 10 like 1, may play in puberty is difficult to speculate given the scarcity of data on this uncharacterized protein. Lomniczi et al. (2015b) reported zinc finger genes as a sub-network controlling the onset of female puberty. In the studied Brahman heifers, zinc fingers were DE and/or important TF in the hypothalamus (Chapter 1.1) and the pituitary gland (as reported herein). The possibility of zinc finger expression being ubiquitous in bovine tissues was considered. However, each of the investigated tissues had its specific zinc fingers that were DE and/or key TF (not all were the same across tissues). Most DE genes were tissue specific: hypothalamus and pituitary gland had 8 DE genes in common. Zinc fingers emerged as important TF across two tissues of reproductive axis. According to the RIF metrics, 26% (hypothalamus) and 22% (pituitary gland) of TF identified as important regulators of puberty in Brahman heifers coded for genes of the zinc finger family. Among key zinc finger TF, two were the same for the pituitary gland and the hypothalamus (Chapter 1.1): ZNF791 and ZNF576. The potential role of genes belonging to the zinc finger family in the pubertal process was proposed by the previous GWAS: associations between single nucleotide polymorphisms located near ZNF462 and ZNF483 and age of menarche were identified in women (Perry et al., 2009b; Elks et al., 2010; Chen et al., 2012; Demerath et al., 2013). Further, studies in male and female mice indicated that there was an increase of MKRN3/ZNF127 mRNA levels pre- puberty, a reduction in levels immediately before puberty and low levels after puberty (Abreu et al., 2013). Decreased expression of zinc finger genes including GATAD1 and ZNF573 in peri-puberty female monkeys was reported (Lomniczi et al., 2015b). In fact, GATAD1 was suggested to have a dual role in both contributing to GnRH release at the occurrence of puberty and silencing GnRH pulse generator at the infantile-early juvenile transition. This study also noted that the GATAD1 gene can repress directly the transcription of two puberty activating genes, KISS1 and TAC3. The downstream effect of KISS1 and TAC3 leads to an impact on GnRH release on the hypothalamic level. Zinc fingers pathways to impact on pituitary gland function are less clear. Pathway analysis using the list of pituitary gland DE genes, key TF and genes in the predicted network of the pituitary gland revealed enrichment for TGFβ signalling pathway, Wnt signalling pathway, PPAR signalling pathway. These pathways’ function and enriched genes are discussed in the following paragraphs.

The TGF superfamily signalling plays a pivotal role in the regulation of cell differentiation, growth, morphogenesis, tissue homeostasis and regeneration (Massague, 2012). Transforming growth factor-beta (TGF), member of TGF superfamily, can increase GnRH gene expression as well as GnRH release (Prevot, 2002; Mahesh et al., 2006; Ojeda et al., 2010a). Genes of this superfamily pathway detected in the pituitary gland network include SMAD6, SMAD4, TFDP1, SP1, SMAD2, SMAD7, PITX2, SMAD3, E2F4, FOXJ1, SMAD1, ID1, ID3, ID2, MYC and CHRD. Among these detected genes, only CHRD was observed as down-regulated after puberty (FC = -0.25, P-value = 0.03). Chordin, or CHRD, is an antagonist of bone morphogenetic proteins (BMPs) activity (Larrain et al., 2000; Oelgeschlager et al., 2000). Bone morphogenetic proteins play significant roles in follicular development (Juengel et al., 2004; Hussein et al., 2005) and a variety of physiological processes such as embryonic development, tissue homeostasis, apoptosis, proliferation, migration, differentiation and bone formation (Urist, 1965; Wu &Hill, 2009; Wagner et al., 2010). More specifically to pituitary gland function, BMP receptors, SMAD4 and SMAD3 were related to signalling for FSH transcription (Rejon et al., 2013; Fortin et al., 2014). Therefore, TGF superfamily genes could be involved in regulation of FSH secretion in pubertal Brahman heifers. Genes in the pituitary gland network involved in the Wnt signalling pathway included TP53, SMAD genes, RACI and NFAT5, among others. Tumour antigen p53 (TP53, RIF2 = 2.81) is associated with pituitary gland cell turnover, a process linked to the pituitary gland plasticity that is important for pubertal development (Garcia-Lavandeira et al., 2015). Additionally, the tumor-related genes were suggested as puberty-delaying genes (Roth et al., 2007; Ojeda et al., 2010a). Note that SMAD genes belong to both Wnt and TGF signalling pathways. Interestingly, RAC1 was reported to regulate several reproductive events, including embryo implantation (Grewal et al., 2008; Nicola et al., 2008), meiotic spindle stability and anchoring in mammalian oocytes (Halet &Carroll, 2007) and embryonic epithelial morphogenesis (He et al., 2010). In addition, RAC1 can stimulate STAT3 activation and therefore affect the transcription of many genes (Faruqi et al., 2001; Park et al., 2004; Kawashima et al., 2006). The deletion of STAT3 reduced POMC and raised AgRP and NPY levels, resulting in obesity, hyperphagia, thermal dysregulation and infertility (Gao et al., 2004). In pituitary gland cells, RAC1 is thought to take part in the intracellular signalling that responds to GnRH secretion from Hypothalamus (Kraus et al., 2001). Perhaps this pituitary gland response to GnRH signalling explains the negative FC observed for RAC1, as progesterone levels in the post-puberty heifers create a context of negative feedback to GnRH

production, with consequences in the pituitary gland. The transcription factor NFAT5 present in the pituitary gland network was also considered a key TF in the hypothalamus of Brahman heifers (Chapter 1.1). This TF was also identified in the Brangus study of heifer puberty (Cánovas et al., 2014a). Mutations in NFAT5 were associated with puberty in women (Chen et al., 2012). These pituitary gland results and mentioned gene function warrant further investigation of TP53, SMAD genes, RAC1 and NFAT5 in the context of cattle puberty. Genes in the pituitary gland network that were annotated to the PPAR signalling pathway were AQP7, CYP4A22, ILK, RXRA, RXRB, PPARG, NR1H3, PPARA, RXRG and PPARD. The TF PPARG was previously identified as a key regulator of cattle puberty, in a network prediction from genome-wide association data (Fortes et al., 2010a). It is easy to speculate that PPAR signalling is important for puberty: because of its relevance to fat metabolism and the well-known interplay between nutrition, metabolism and puberty in cattle (Cardoso et al., 2014a; Cardoso et al., 2015). Three among the PPAR signalling genes were DE AQP7 had a high fold change (FC = 3.83, P-value = 0.0009). The gene AQP7 is a member of a small family that facilitates rapid passive movement of water (Huang et al., 2006). Water movement across cell membranes is important in processes underlying reproduction and 11 isoforms of the AQP family were reported as expressed in ovaries, testis and embryos (Huang et al., 2006). This study revealed up-regulation of AQP7 expression in the pituitary gland of post-puberty Brahman heifers. The expression of AQP genes can be mediated by steroid sex hormones (Jablonski et al., 2003; Richard et al., 2003; Lindsay &Murphy, 2004; He et al., 2006). High progesterone levels in post-puberty heifers contrasting with low levels in the pre-puberty group may influence DE results for AQP7 in the pituitary gland. Further research is required to understand the specific role of AQP7 in cattle pituitary gland. The gene CYP4A22, another PPAR signalling gene up-regulated in post-puberty heifers, belongs to a cytochrome P450 family that is important for steroid synthesis (Pikuleva &Waterman, 2013). Pituitary gland expression of a cytochrome P450 family member suggests this aromatase would be related to the mechanisms of action of gonadal steroids on pituitary gland differentiation and secretion, as shown by immunohistochemistry in rats (Carretero et al., 2003). The gene CYP4A22 could be part of the pathways in the pituitary gland that respond to progesterone signalling, linking pituitary gland function to ovarian feedback mechanisms established with puberty. The third DE gene in the PPAR signalling pathway was ILK, known as an essential upstream regulator of Akt activation (Troussard et al., 2003). Activation of Akt is relevant for puberty: KISS1 expression is mediated by an IGF1/Akt/mTOR pathway (Hiney et al., 2010;

Srivastava et al., 2016). The expression of KISS1, fundamental for GnRH release at puberty, was modulated by mTOR in pre-puberty female rats and mTOR may also be a link between leptin signalling and GnRH release (Roa et al., 2009). Evidence for the role of KISS1 pathway directly influencing pituitary gland activity is mounting (Gahete et al., 2016). The fact that ILK was DE in the pituitary gland may serve as further evidence for this direct influence of KISS1 pathway on pituitary gland function (not only via Hypothalamic signalling and GnRH release). Our results and the fact that ILK is essential for PPAR signalling and Akt regulation (which in turn is important for KISS1 expression) raise the possibility that ILK is another molecular link between energy metabolism and the onset of puberty in cattle.

1.2.5. Conclusion

Both Brahman and Brangus transcriptomes highlight the importance of zinc fingers for cattle puberty in reproductive tissues. Interactions between zinc finger genes and predicted target genes need validation, beyond co-expression results. Interactions between zinc finger genes and targets considered as puberty activating genes are of particular interest to further unveil the proposed mechanisms of puberty (Lomniczi et al., 2015b). Pre-puberty, zinc finger genes are proposed to inhibit puberty activating genes. When the expression of zinc fingers decreased after puberty, it could result in the increase of puberty activating genes contributing to puberty onset. In summary, reported RNA sequencing data from pre- versus post-puberty Brahman heifers in the pituitary gland were used to identify key TF regulators of DE genes and to predict co-expression networks. Reported results corroborate some of the known biologies of puberty, recapitulating genes and pathways involved in GnRH secretion. Tissue-specific zinc finger TF seemed particularly important regulators of gene expression in the hypothalamus-pituitary axis of pre- versus post-puberty heifers. Furthermore, novel genes were implicated in pubertal development, which require additional research for their characterization.

Chapter 1.3. STAT6, PBX2 and PBRM1 emerge as predicted regulators of 452 differentially expressed genes associated with puberty in Brahman heifers

Manuscript published in Frontiers in Genetics., Volume 9, Issue 87, March 2018

https://doi.org/10.3389/fgene.2018.00087

Abstract published in Journal of Animal Science., Volume 94, Issue suppl_4, September 2016, Pages 20–21

https://doi.org/10.1093/ansci/94.supplement4.20

Abstract

The liver plays a central role in metabolism and produces important hormones. Hepatic estrogen receptors and the release of insulin-like growth factor one (IGF1) are critical links between liver function and the reproductive system. However, the role of the liver in pubertal development is not fully understood. To explore this question, this study applied transcriptomic analyses to liver samples of pre- and post-pubertal Brahman heifers and identified differentially expressed (DE) genes and genes encoding transcription factors (TF). Differential expression of genes suggests potential biological mechanisms and pathways linking liver function to puberty. The analyses identified 452 DE genes and 82 TF with a significant contribution to differential gene expression by using a regulatory impact factor metric. Brain-derived neurotrophic factor (BDNF) was observed as the most down-regulated gene (P-value = 0.003) in post-pubertal heifers and was proposed to influence pubertal development in Brahman heifers. Additionally, co-expression network analysis provided evidence for three TF as key regulators of liver function during pubertal development: STAT6, PBX2 and PBRM1. Pathway enrichment analysis identified TGF- and Wnt signalling pathways as significant annotation terms for the list of DE genes and TF in the co-expression network. Molecular information regarding genes and pathways described in this work are important to further our understanding of puberty onset in Brahman heifers. Keywords: Bos indicus, puberty, gene expression, RNA sequencing, gene network, liver

1.3.1. Introduction

The beef industry in Northern Australia is facing an increased demand for improved herd productivity. Brahman cattle, a breed of the Bos indicus sub-species, can withstand hot- humid conditions but enter puberty at an older age in comparison with Bos taurus (Johnston et al., 2009). Late onset of puberty in Bos indicus predicts a decrease in lifetime productivity (Lesmeister et al., 1973; Johnston et al., 2013). Therefore, reducing the age at puberty to increase Bos indicus cow productivity is a worthwhile goal for management and breeding.

Reproduction is an energy-intensive process that is likely to require the specific involvement of the liver. The physiological mechanisms controlling energy balance are closely linked to fertility, to minimise the risk that pregnancy and lactation coincide with periods of nutritional stress (Mircea et al., 2007). While in placental mammals, the hypothalamus- pituitary-ovaries axis takes precedence in the integration of metabolic and reproductive status, there are evidences for the involvement of the liver in this process. In the postpartum cow, it is

known that the metabolic stress associated with transition is linked to impaired liver function and delayed ovulation (Montagner et al., 2016). In mice, it has been shown that the hepatic synthesis of IGF-1 is regulated by amino acid-dependent activation of ERα in the liver (Della Torre et al., 2011). Recently, a study using female mice having liver-specific ablation of ERα reported that liver ERα was essential for the modulation of AgRP neurons in the coordination of reproductive function and energy metabolism (Benedusi et al., 2017). Additionally, an association between single nucleotide polymorphisms in genes of the IGF1 signalling pathways and age at puberty in Brahman cattle was observed (Fortes et al., 2013a).

Several studies investigated the change in hepatic mRNA expression of genes encoding proteins that participate in various processes including growth hormone signalling, liver lipoprotein assembly, ureagenesis and gluconeogenesis (Loor, 2010). A few studies have utilized microarray technology to evaluate hepatic metabolic adaptations to dairy cow throughout pregnancy, transition period, early lactation and mid-lactation (Herath et al., 2004; Loor et al., 2005; Loor, 2010; McCarthy et al., 2010; Akbar et al., 2013). A microarray study of the effect of pregnancy and diet in liver gene expression revealed specific hepatic adaptations of beef cows to different nutritional environments. For example, the study found clear evidence of gluconeogenesis in the liver of pregnant cows during limited forage availability (Laporta et al., 2014). Using candidate gene approach or transcriptomics investigating genomic peripartal adaptions in dairy cows provided insights into physiological function and genetics of key tissues.

A further study has evaluated the liver transcriptome during puberty onset in Brangus heifers (3/8 Brahman; Bos indicus x 5/8 Angus; Bos taurus) (Cánovas et al., 2014a). This study has used RNA sequencing, which is a more sensitive transcriptome profiling method than microarray. Sequencing RNA is capable of detecting not only expression differences in the most highly expressed metabolic genes, but also in regulatory genes (Marioni et al., 2008). In this chapter, I describe the use of RNA sequencing to evaluated mRNA expression, regulatory factors and potential biological pathways that occur in the liver related to pubertal development in Brahman heifers; a different population that is predominantly Bos indicus. The heifers were used in two studies that reported transcriptomics of the Hypothalamus-Pituitary axis, with no observation of liver function (Chapter 1.1 and 1.2). Molecular information of key regulators and pathways in the liver may reveal mechanisms involved in puberty onset and energy metabolism. To access IGF1 signalling in this context, this study also reports hormonal

measurements. This information may contribute to future approaches for reducing the age at puberty of Bos indicus cattle used in tropical beef production systems.

1.3.2. Materials and methods

Ethics statement

The Animal Ethics Committee of the University of Queensland, Production and Companion Animal group approved this experiment (certificate number QAAFI/279/12). Animals and samples As above mentioned, heifers used in this Chapter were the same as those utilized in two previous studies (Chapter 1.1 and 1.2). Briefly, this study performed ultrasound observations of pubertal development every fortnight from October 2012 to May 2013 in order to define pubertal status among heifers (Johnston et al., 2009). Euthanasia occurred 15 days after the observation of the first corpus luteum (CL), with samples collected in the next estrous cycle. Six post-pubertal heifers were euthanized during the luteal phase of their second estrous cycle, confirmed by the observation of the second CL at euthanasia. Serum progesterone concentrations were measured to confirm a functional CL in post- pubertal heifers (2.0 ± 0.7 ng/mL, mean ± SE). Pre-puberty heifers were randomly selected from the group that had never ovulated (plasma progesterone concentration 0.4 ± 0.2 ng/mL, mean ± SE) and paired with post-pubertal animals in slaughter day. Therefore, on each slaughter day, two heifers were euthanatized, one pre- and one post-puberty.

Circulating IGF1 concentrations were measured using a commercial radioimmunoassay kit (10IGF100 Kit; Bioclone, Sydney, NSW, Australia). The method included an acid-ethanol extraction to remove IGF1 binding proteins and measure total IGF1. All samples were analysed within a single assay kit as previously described (Dahlanuddin et al., 2014). The assay sensitivity was 0.2 ng/ml, and the within-assay coefficient of variation was 2.5%.

After slaughter, liver tissue harvest was conducted as fast as possible to preserve RNA integrity. The entire liver was removed from the animal and 3 samples of 1 cm3 were dissected from the liver and snap frozen in liquid nitrogen. Samples were stored at -80ºC until RNA extraction. In total, 12 liver samples were processed separately for RNA extraction and sequencing. Ribonucleic Acid extraction Total RNA extraction was performed using the combination of RNeasy mini kit (QIAGEN Pty Ltd., Victoria, Australia) and TRIzol (Life Technologies Inc., California, USA) methods as previously described (Chapter 1.1). Quality of the total RNA was evaluated using

the RNA integrity number (RIN) measured with an Agilent Bioanalyzer 2100 (Agilent Technologies). Values of RIN range from 7.3 to 8.5, which indicated good quality of the RNA samples, which were sent to the University of California, Davis, for library preparation and sequencing. Library preparation and sequencing mRNA was purified, fragmented, and used to synthesize cDNA, as described in (Cánovas et al., 2010). Briefly, after ligation of the adapters to the ends of double-stranded cDNA fragments, PCR was conducted to create the final cDNA library. Sequencing libraries were constructed with the TruSeq RNA sample preparation kit (Illumina, San Diego, CA, United States). RNA sequencing was conducted with a HiSeq 2000 Sequencer Analyzer (Illumina, San Diego, CA, United States). Quality control was performed using procedures described previously (Cánovas et al., 2013) using the application NGS quality control of CLC Bio Genomic Workbench software (CLC Bio, Aarhus, Denmark). All samples passed all the parameters indicating a very good quality. Sequence pair-end reads (100 bp) were assembled against the annotated bovine genome (release 77; ftp://ftp.ensembl.org/pub/release-77/genbank/bos_taurus/). The “reads per kilobase per million mapped reads” (RPKM = total exon reads/mapped reads in millions × exon length in kb) was calculated for data normalization (Mortazavi et al., 2008). A threshold of RPKM ≥ 0.2 was utilized to annotated expressed genes (Wickramasinghe et al., 2012). Normalization and transformation data were performed using CLCBio Genomic Workbench software (CLC Bio, Aarhus, Denmark) to transform the expression data from negative binomial distribution to normal distribution. Identification of differentially expressed genes

Because genes with low counts can be easily biased without transformation, the base-2 log transformed RPKM values were used. This study normalized the log-transformed RPKM values using mixed model equations to increase the sensitivity to detect differential expression and co-expression. This normalization approach for transformed RPKM values was previously described (Reverter et al., 2005; Cánovas et al., 2014a). The study was part of a large experiment where 7 tissues were sampled per animal (hypothalamus, pituitary, liver, ovary, uterus, muscle and adipose tissue). Therefore, differential gene expression across tissues was estimated using a mixed model where base-2 log transformed RPKM was modelled as a function of the fixed effect of the library and of the random effects of gene and the interaction between gene x animal x physiological state x tissue. A t-test was used to test the hypothesis

that the differential expression in post- vs. pre-pubertal heifers was significant. With the strict normalization performed, this study then used P-value < 0.05 as the threshold to determine DE genes. This seemingly not very stringent nominal P-value was used in context with the strict normalization performed and the subsequent analyses, for which differential expression is one of many criteria under scrutiny.

Identification of key gene regulators

To determine gene regulators from the data, the study mined the AnimalTFDB bovine database (http://bioinfo.life.hust.edu.cn/AnimalTFDB/download_index?tr=Bos_taurus), which comprises the classification and annotation of animal genomes for transcription factors (TFs), chromatin remodelling factors, and transcription co-factors. Among the annotated TF for Bos taurus, 1,085 were expressed in the liver and further filtered for significance in terms of co-expression with DE genes, using regulatory impact factor (RIF) metrics (Hudson et al., 2009; Reverter et al., 2010). A TF was considered as a key regulator if either of the two RIF scores was higher than 1.96 of the standard deviation, equivalent to a P-value level of at least 0.05.

Gene network prediction

The partial correlation and information theory (PCIT) algorithm was performed to detect the association between genes in a co-expression gene network as described previously (Reverter &Chan, 2008). The co-expression network predicted for the liver data was then visualized with Cytoscape (Shannon et al., 2003). From the large predicted network, the study explored the subnetwork deemed to have biological significance for puberty trait. The subnetwork was used to identify the best trio TF that spanned most of the network topology with minimum redundancy. Specifically, an information lossless approach (Reverter and Fortes, 2013) that explored the 59,640 possible trios among 82 available TFs in the network was used to identify the best TF trio.

Functional enrichment analysis

For enriched pathways and gene expression patterns, the Database for Annotation, Visualization, and Integrated Discovery (DAVID, http://david.abcc.ncifcrf.gov) was utilized (Dennis et al., 2003; Huang da et al., 2009). In the study, the queried gene lists included genes and TF that formed the predicted gene network. Significant gene ontologies and pathways after Benjamini-Hochberg correction for multiple testing are reported.

1.3.3. Results

The liver transcriptome data passed quality control performed with CLC Genomics Workbench (CLC Bio, Aarhus, Denmark). Each individual sample had about 63 million sequence reads. Previous studies demonstrated that approximately 30 million reads are sufficient to detect more than 90% of annotated genes in mammalian genomes (Wang et al., 2011; Lee et al., 2013; Singh et al., 2017). The relatively high number of sequence reads generated in this transcriptome study indicates that liver transcriptomic data are adequate for the identification of DE genes. The number of unique reads and RPKM of each gene per physiological state are provided in Supplementary Table S1 in Nguyen et al. (2018). Sequence data is available through the Functional Annotation of Animal Genomes project (http://data.faang.org/home).

Identification of differentially expressed gene

A total of 16,539 transcripts (RPKM 0.2) were detected in both groups (pre- and post- puberty). A t-test of log-transformed data identified 452 DE genes (including 57 novel genes), of which 253 were up-regulated and 199 were down-regulated post-puberty (P-value < 0.05). Ten genes showed a 3-fold change (FC) difference in expression levels and P-value < 0.01 between pre- and post-puberty heifers (Table 1.3). Figure 1.3.1 shows a volcano plot of log2 FC versus –log10 P-values for pre- versus post-puberty gene expression. The gene annotation, FC and P-value of 452 DE genes are presented in Supplementary Table S2 in Nguyen et al. (2018). Significant DE genes were useful for understanding the biological mechanisms in the liver underlying puberty onset in Brahman cattle.

Figure 1.3.1 Volcano plot of differentially expressed genes (n = 452) in liver between pre- vs post- pubertal heifers. The X-axis represents the fold change while the y-axis represents statistical

significance for each gene. Red dots indicate genes that differ significantly (P-value < 0.05) between the two groups. Genes plotted in the left portion of the graph were expressed at a lower level in post- pubertal liver, and gene in the right-hand portion had higher expression levels post-puberty. Gene symbols are provided for genes with a Fold Change |FC| 3 and P-value 0.01.

Table 1.3 The reads per kilobase per million mapped read (RPKM) values for differentially expressed genes in liver between pre- vs. post-pubertal Brahman heifers (|FC| 3, P 0.01)

ENSB Tag1 Symbol2 RPKM_PRE3 RPKM_POST4 FC5 ENSBTAG00000043414 snoR38 1.415 6.750 5.335 ENSBTAG00000044882 Novel gene 0.546 4.877 4.330 ENSBTAG00000011660 MSMB 4.309 7.658 3.349 ENSBTAG00000042447 SNORD49 4.599 0.457 -4.142 ENSBTAG00000030124 Novel gene 6.815 2.908 -3.907 ENSBTAG00000008134 BDNF 5.890 2.022 -3.868 ENSBTAG00000017502 RIMKLA 6.187 2.928 -3.259 ENSBTAG00000045577 MCCD1 6.804 3.690 -3.114 ENSBTAG00000004657 FBLL1 4.978 1.947 -3.031 ENSBTAG00000033173 BHLHE22 6.871 3.865 -3.007 1ENSB Tag: Ensembl gene identifier according to www.ensembl.org; 2Gene: gene symbol related to the ENSB Tag; 3RPKM_PRE: The RPKM in pre-pubertal heifers (average); 4RPKM_POST: The RPKM in post-pubertal heifers (average); 5FC: Fold change (RPKM _POST minus RPKM_PRE).

Insulin-like growth factor 1 (IGF1) is the major hormone secreted by the liver and is known to increase during puberty. In the current study, the circulating IGF1 concentrations differed between pre- and post-pubertal heifers at euthanasia (P-value = 0.008) with the average serum IGF1 levels were 159.3 25.5 ng/ml at pre-puberty and 203.2 31.1 ng/ml at post-pubertal heifers. Although, RNA-seq analysis showed an increase in IGF-1 mRNA levels (2.01 0.17 versus 2.35 0.19) after puberty in the liver, the result was not significant (P- value = 0.222).

Identification of key gene regulators

From AnimalTFDB Bovine database, 1,085 TF that were expressed in the liver were retrieved. Using regulatory impact factor (RIF) metrics, these known TF were filtered for those most consistently associated with DE genes from this study, identifying 82 TF (P < 0.05). Supplementary Table S3 summarizing relevant data for these TF: RIF results, expression levels

and its description is presented in Nguyen et al. (2018). Of the 82 TF, 19 genes (23%) coded for TF of the zinc finger family (ZNF). Further, out of the 82 TF, 2 TF (SOX13 and BHLHE22), were themselves identified as DE genes.

Predicted gene co-expression network and sub-network

Partial correlation information theory (PCIT) algorithm determined significant partial correlations between DE and TF. A predicted gene co-expression network with 1,408 nodes representing genes and a total of 8,330 edges, which account for the predicted interactions was constructed (Figure 1.3.2). In order to identify potential regulators of the predicted gene co- expression network, the study focused on 82 TF contained in the network. After exploring all the TF trios, the top trio which spanned most of the network topology with highest connectivity (a total of 59,640 possible connections) and minimum redundancy was identified, including the signal transducer and activator of transcription 6 (STAT6), PBX homeobox 2 (PBX2) and polybromo 1 (PBRM1). Figure 1.3.3 presents the connections between the top trio of TF and their potential targets.

Figure 1.3.2 Liver gene co-expression network constructed by PCIT in pre- and post-puberty Brahman heifers. The entire network comprises 1,408 nodes (or genes) and 8,330 interactions. The colour spectrum ranges from green to red for low and high number of connection, respectively.

Figure 1.3.3 Subset of the liver co-expression network showing the best trio of TF and its predicted target genes in pre- and post-puberty Brahman heifers. Each node represents a gene. Nodes represented as Triangles are TF and other coding sequences are represented as ellipses. Edges represent significant interaction between nodes. Node colour indicates the number of connections of a specific node in the network. The colour spectrum ranges from green to red for low and high number of connections.

Functional enrichment analysis of target genes involved in gene co-expression network

Functional analysis using DAVID (Dennis et al., 2003; Huang da et al., 2009) allowed identification of biological functions overrepresented in nodes involved in the liver co- expression network. Results showed that 91 GO terms (49.7%, 91/183) were significantly enriched in the biological process category, 8 GO terms (53.3%, 8/15) were significantly enriched in the molecular function category, and 21 GO terms (75%, 21/28) were significantly enriched in the cellular component category. In the biological process category, GO terms were related to liver development, gonad development, immune system development and muscle organ development. Most importantly, many of the enriched GO terms were closely related to reproduction, including reproductive developmental process, reproductive structure development and response to protein stimulus. In addition, the molecular function term associated with steroid hormone receptor activity was also enriched in liver co-expression network. Pathway analyses revealed 10 significantly enriched pathways (47.6%, 10/21). Among these overrepresented pathways identified, TGF- signalling (adjusted P = 1.7x10-4)

and Wnt signalling (adjusted P = 7.4x10-4) pathways were observed. Supplementary Table S4 provides the full list of enriched GO terms and pathways, discovered using all genes in the network as the target dataset. Supplementary Table S4 can be found at Nguyen et al. (2018).

1.3.4. Discussion

Reducing the age at puberty to increase cattle productivity is a major aim for Bos indicus breeders. Although the Hypothalamus-Pituitary-Ovarian axis is central to reproduction, the involvement of the liver in controlling energy balance and affecting reproduction was reported before (Fontana &Della Torre, 2016; Montagner et al., 2016). IGF1 seems to be an important link between liver function and puberty onset (Akers et al., 2005). Although the post-pubertal liver samples were collected from animals with significantly higher progesterone levels, the study found no direct evidence of increased synthesis of liver enzymes involved in the metabolism of steroid hormones. IGF1 transcripts were not among the list of DE genes, although this endocrine signal from the liver is known to increase leading up to puberty and serum IGF1 was increased at post-pubertal Brahman heifers (current study). The bioavailability of and circulating half-life of IGF1 is determined by insulin-like growth factor binding proteins (IGFBPs) and these may be an important consideration in puberty. It should be noted that most assays measure total IGF1, after extraction procedures that remove IGFBPs, like results presented in this current study. Very few studies determine the small (<1% total), free bioactive fraction of IGF1 and/or concentrations of IGFBPs. Results herein suggest that circulating IGF1 concentrations are influenced by multiple factors beyond IGF1 gene expression.

In liver, 452 genes were DE between pre- and post-pubertal Brahman heifers (this study). Previously, 288 genes were DE between pre- and post-pubertal Brangus heifers (Cánovas et al., 2014a). In Brangus heifers, liver DE genes contributed an abundant number of connections to the co-expression network and had the largest disappearance of connections after puberty. In short, network topology suggests that the liver warrants further scrutiny in Brangus heifers that was beyond the scope of the original publication (Cánovas et al., 2014a). Here, gene ontology and pathway enrichment analyses for both lists of DE genes, from the Brangus study and the current Brahman data were performed. No ontologies were the same across breeds. Only 10 DE genes were the same across breeds and these are discussed further below. In short, biological differences between Brangus and Brahman heifers seem evident from the contrasting results in these transcriptomics studies.

The most DE genes (|FC| > 3 and P-value ≤ 0.01) in the liver of Brahman heifers were snoR38, snorD49, MSMB, RIMKLA, MCCD1, FBLL1, BHLHE22, BDNF and two uncharacterized proteins (ENSBTAG00000044882 and ENSBTAG00000030124). These emerging candidate genes are discussed in the following paragraphs.

Small nucleolar RNA R38 (snoR38) and small nucleolar RNA SNORD49 (snorD49) are non-coding RNAs functioning in modifications of other small nuclear RNA (Matera et al., 2007). The snoRNA families are essential for major biological processes such as mRNA splicing and protein translation (Matera et al., 2007). There is limited evidence for the involvement of these snoRNA with puberty. The first deletion animal model of another snoRNA gene (snorD116) in mice revealed a role in the growth and feeding regulation for the snoRNA family (Ding et al., 2008). The highest and lowest mRNA levels after puberty of snoR38 (FC = 5.33) and snorD49 (FC = -4.12) warrant further studies to understand the role that these non-coding RNAs play in liver function and puberty.

The gene β-microseminoprotein (MSMB) plays an important role in semen quality and fertilization (Anahi Franchi et al., 2008). Not restricted to male tissues, MSMB protein was also identified in porcine corpus luteum (Tanaka et al., 1995) and its gene expression was identified in human female reproductive tissues (Baijal-Gupta et al., 2000). Importantly, MSMB influences FSH secretion from the pituitary gland, impacting on ovarian function (Thakur et al., 1981; Sheth et al., 1984; Frankenberg et al., 2011). It remains to be explored if liver production of MSMB achieves the pituitary signalling in growing heifers.

Very little is known about mitochondrial coiled-coil domain 1 (MCCD1) and fibrillarin like 1 (FBLL1) function in the liver or with relation to puberty onset. One study in humans identified high expression levels of MCCD1 in foetal liver (Semple et al., 2003). In cattle, MCCD1 was DE in both RNA sequencing studies of puberty, this current study and the study by Cánovas and colleagues (2014a). Its liver function merits further investigation.

The gene RIMKLA is involved in alanine, aspartate and glutamate metabolism; as per KEGG pathway annotation (Kanehisa et al., 2017). Notably, glutamate and aspartate are major metabolic fuels for nutrient metabolism and oxidative defence (Yao et al., 2008; Brasse-Lagnel et al., 2009). It seems coherent that RIMKLA would be relevant for the liver metabolic function and perhaps it is another link between energy metabolism and reproduction to be explored.

The gene BHLHE22 (FC = -3.00 and P-value ≤ 0.01) was revealed as the most down- regulated DE gene after puberty. This gene is also a significant TF (RIF2 score of -2.88). The

BHLHE22 gene belongs to basic helix-loop-helix (bHLH) family and plays a significant role in cell proliferation and differentiation of several developmental pathways as well as cell fate determination (Lee et al., 1995; Ma et al., 1996; Farah et al., 2000; Xu et al., 2002). Further, an in vitro transfection assay used in mice showed that BHLHB5 (an alias of BHLHE22) strongly inhibits the expression of the human PAX6 promoter (Xu et al., 2002). The PAX6 promoter is known as a TF involved in embryonic development and neurulation (Callaerts et al., 1997). A PAX6 mutation was associated with isolated GH deficiency (Guerra-Junior et al., 2008). BHLHE22 has been described as a transcriptional repressor of insulin expression in pancreatic -cells (Peyton et al., 1996; Melkman-Zehavi et al., 2011). Insulin can mediate follicular growth in cattle (Webb et al., 2004), stimulate GnRH release from the hypothalamus in combination with glucose (Arias et al., 1992), and may also facilitate IGF1 synthesis and secretion by the liver (Keisler &Lucy, 1996; Webb et al., 2004). The role of insulin in the regulation of lipid, glucose, protein homeostasis and energy balance (Saltiel &Kahn, 2001; Liu &Barrett, 2002; Obici &Rossetti, 2003) suggests a link between insulin and the reproductive axis. In the liver study, BHLHE22 was the most down-regulated gene, with lower expression in post-pubertal heifers. Lower expression of BHLHE22 could mean decreased repression of insulin expression leading to increased GH and GnRH stimulus via IGF1 signalling. Therefore, liver produced BHLHE22 could impact on animal growth and pubertal development.

The BDNF gene is related to the neural development and peripheral metabolism (Binder &Scharfman, 2004; Pedersen et al., 2009). In the brain, BDNF can suppress GABAergic synaptic transmission by acute down-regulation of GABAA receptors and thus can affect GnRH release (Henneberger et al., 2002). Previous studies suggested BDNF as a key component of the hypothalamic pathway controlling energy homeostasis and body weight (Wisse &Schwartz, 2003; Xu et al., 2003; Jo &Chua, 2013). A genome-wide association studies (GWAS) in humans found BDNF to be related to timing of puberty and body mass index (Perry et al., 2014b). Further, estrogen-BDNF-NPY has been noted as an important tri-molecular cascade in understanding the hormonal regulation in the hippocampus (Scharfman &MacLusky, 2006). It is unclear whether BDNF is able to cross the blood-brain barrier. Some researchers have found evidence for a link between central BDNF and peripheral BDNF (Poduslo &Curran, 1996; Pan et al., 1998; Rasmussen et al., 2009; Seifert et al., 2010), whereas others have argued that it does not cross the blood-brain barrier (Pardridge et al., 1998; Kyeremanteng et al., 2012). Hence, if further research can prove BDNF delivery across the

blood-brain barrier, it is possible that BDNF produced in the liver may have endocrine effects in the brain.

Brain-derived neurotrophic factor in the liver, similarly to skeletal muscle, results in increase of AMP-activated protein kinase (AMPK) and its downstream target acetyl coenzyme A carboxylase (ACC), inhibiting fatty acid synthesis and enhancing fatty acid oxidation (Matthews et al., 2009; Pedersen et al., 2009; Genzer et al., 2017). A study in humans suggested that fatty acid oxidation is higher in children than adults (Kostyak et al., 2007). Estrogen was also cited to regulate hepatic fatty acid oxidation (O’Sullivan, 2012). In the liver, there is little information of precise mechanisms in which estrogen reduces fatty acid oxidation. Our study results led us to hypothesize that the interaction between estrogen and BDNF in fatty acid oxidation in the liver, contributing to metabolic changes that can regulate puberty onset.

Comparing the liver transcriptional profile between Brahman heifer study (this Chapter) and a study by Cánovas et al. (2014a) in Brangus heifers, 10 genes (including a novel gene) were DE in both populations (P < 0.05). Five genes, MCCD1, ADGRF2, BEX2, PDZD7 and LRRC46 had a |FC| > 1 in both breeds. The expression of these genes was up-regulated in Brangus heifers and down-regulated in Brahman heifers. The Brangus study involved 8 heifers greatly differing in age and weight whereas our Brahman study used 12 heifers that were age- and weight-matched. Further, in the absence of a reference genome of Bos indicus, Bos taurus reference genome was utilized for sequence assembly, and so the divergence between Bos taurus and Bos indicus genomes can impact results. The significant difference in expression levels and patterns of these 5 DE genes between two breeds warrants further studies. The candidate gene MCCD1 and its limited literature were discussed above. Similarly, PDZD7 and LRRC46 roles in puberty and liver function cannot be speculated from current knowledge. The remaining two genes, ADGRF2 and BEX2, are discussed below.

The expression of adhesion G protein-coupled receptor F2 (ADGRF2, alias GPR111), in reproductive tissues and lung, was reported (Fredriksson et al., 2002). This study was the first to report mRNA expression of ADGRF2 in the liver of Bos indicus heifers. It is intriguing to suggest that ADGRF2 could be another link between liver function and puberty because G protein-coupled receptors have been associated with GnRH regulation (Noel &Kaiser, 2011).

The brain expressed X-linked 2 (BEX2) was observed as a DE gene in the adipose tissue of pre- and post-pubertal Brangus heifers (Cánovas et al., 2014a). The BEX2 gene is highly expressed in the human embryonic brain and have a regulatory role in embryonic development

(Han et al., 2005). A study of mice liver gene expression revealed a strong expression of BEX2 in stem/progenitor cells (Ito et al., 2014). Further, BEX2 is a downstream molecule of the mammalian target of rapamycin (mTOR) signalling pathway (Hu et al., 2015) that can regulate lipogenesis and ketogenesis in liver (Laplante &Sabatini, 2012). The mTOR pathway is also a known regulator of ovarian activity (Lu et al., 2017). In short, BEX2 was DE in two studies of pubertal heifers, two different breeds, and thus it merits further investigation. It is possible that this is a liver signal with an impact on ovarian activity.

Transcription factors play a key role in controlling gene expression, but their expression levels are often low and not detected as DE (Vaquerizas et al., 2009). The interactions between TF are important for tissue remodelling and temporal changes in gene expression (Ravasi et al., 2010). Differential expression analyses overlook vital changes in regulatory information. Adding an analysis focused on identifying key TF could help to understand the gene regulation processes under investigation (i.e. puberty). Previously, TF in the zinc finger (ZNF) family were found to be DE and/or important TF in the transcriptomic profile of hypothalamus, the pituitary gland in Brahman heifers undergoing puberty (Chapter 1.1 and 1.2). These studies noted that 26% of top ranking TF from the hypothalamus and 22% from the pituitary gland top-ranking TF coded for ZNF family members in the same Brahman heifers (Chapter 1.1 and 2.2). Likewise, this current study revealed that 23% of TF identified by RIF analysis of liver transcriptome data belong to the ZNF family.

The potential role of ZNF genes in the puberty process was suggested by several studies. A mouse study found that a mutation in regulator of sex-limitation (RSL), one of the Kruppel-associated box zinc finger proteins (KRAB-ZFP) genes, can impact reproduction by regulating expression patterns of target genes in liver (Krebs &Robins, 2010). In addition, ZNF genes have been implicated in the epigenetic control of transcription in the female primate hypothalamus around puberty (Lomniczi et al., 2015b). Previous GWAS in women reported the association between single nucleotide polymorphism located near ZNF462 and ZNF483 and age of menarche, which is the age of puberty in girls (Perry et al., 2009b; Elks et al., 2010; Chen et al., 2012; Demerath et al., 2013). Expression of ZNF127 was increased pre-puberty and decreased immediately before puberty (Abreu et al., 2013). Study of female monkeys also reported a decrease of ZNF573 mRNA levels in peripubertal animals (Lomniczi et al., 2015b). This study contributes to the growing body of evidence that supports ZNF genes can influence puberty onset, a developmental role which may extend to tissues and organs outside of the reproductive axis.

In the sub-network, the trio of TF that spanned most of the network topology with minimum redundancy and highest connectivity was STAT6, PBX2 and PBRM1. Previous evidence suggested these TF have important roles in the liver and reproductive function. Specifically, the STAT6 locus on BTA5 has been described as a QTL associated with reproduction (Kappes et al., 2000; Allan et al., 2009; Kim et al., 2009; Luna-Nevarez et al., 2011; Hawken et al., 2012a). Further, this gene was identified as a key TF in a gene network constructed using GWAS results of first service conception in Brangus heifers (Fortes et al., 2012). Other studies noted the association between STAT6 gene and age at first corpus luteum, an indicator of puberty onset in Brahman and Tropical Composite heifers (Fortes et al., 2010a; Fortes et al., 2011). This study supported the potential role of STAT6 in puberty onset in Brahman heifers.

The PBX2 gene has a role in the development of ovarian follicles (Ota et al., 2008). Pbx2-Prep1 complexes repress HNF1α-mediated activation of the UDP Glucuronosyltransferase Family 2 Member B17 (UGT2B17) promoter in liver cells (Gregory &Mackenzie, 2002). The UGT2B17 gene, a sex steroid-metabolizing gene, has been associated with male infertility and impaired spermatogenesis (Plaseska-Karanfilska et al., 2012). Hepatocyte nuclear factor 1 α (HNF-1α) is a homeodomain-containing transcription factor that regulates liver-specific gene transcription (Mendel &Crabtree, 1991) and was suggested to control development and metabolism in a HNF-1α-null mouse study (Pontoglio et al., 1996). In liver, HNF-1α regulates the expression of glucocorticoid receptor (GR), IGF1, STAT5 and other GH-responsive genes (Lee et al., 1998; Lin et al., 2008). In sub-network of liver predicted gene co-expression, PBX2 and STAT6 were also connected to STAT5 (RIF2 score of 2.13) suggesting that the interaction between these TF could contribute to the regulation of growth, liver metabolism and puberty onset.

Finally, the gene PBRM1 seems to play a role in metabolic and immune system regulation, pertinent to liver expression. The gene PBMR1 was described as a repressor of interleukin 10 (IL-10) transcription; an anti-inflammatory cytokine involved in metabolic syndrome (Mallat et al., 1999; Calcaterra et al., 2009; Wurster et al., 2012). Calcaterra et al. (2009) study showed high levels of IL-10 in serum samples of obese children. Further, IL-10 was proposed to be involved in the inflammatory network of metabolic syndrome in correlation with adiponectin (Bottner et al., 2004; Nishida et al., 2007). Of note, adiponectin plays a significant role in energy homeostasis (Lee &Shao, 2014). In the liver, adiponectin can activate glucose transport as well as enhances insulin sensitivity (Berg et al., 2001; Combs et al., 2001;

Ye &Scherer, 2013). A study of Holstein cows reported an association between follicular growth and the change in adiponectin and its receptor expression (Tabandeh et al., 2010). The role of PBRM1 as a regulator of heifer puberty needs further investigation, but it is possible that it acts through adiponectin signalling.

After the identification of DE genes and TF, GO and pathway analysis was performed to better understand the biological function of these genes in the context of puberty. Information about gene co-expression, enriched GO and pathways facilitates the interpretation of RNA-Seq results. Based on the GO analysis of 1,408 nodes from the liver co-expression network, multiple biological processes were affected. GO terms “reproductive developmental process” and “reproductive structure development” were significantly enriched and are logical in the context of puberty. Steroid hormone receptor activity and steroid binding were expected GO terms as the liver is the principal site of steroid hormone metabolism.

This work observed TGF-β signalling (P = 6.4x10-6) and Wnt signalling (P = 3.8x10-5) pathways among the enriched pathways. Both pathways were also enriched in pre- vs post- pubertal results from the pituitary gland of Brahman heifers (Chapter 1.2). Of note, transforming growth factor-beta (TGF-) superfamily signalling plays a pivotal role in the regulation of cell differentiation, growth, morphogenesis, tissue homeostasis and regeneration (Massague, 2012). In neural tissue, TGF-1 one member of the TGF- superfamily, can increase GnRH gene expression as well as GnRH release (Prevot, 2002; Mahesh et al., 2006). Expression and release of GnRH are pivotal for puberty. The Wnt signalling pathway is an important physiological regulator of embryonic and liver development as well as mammalian hepatic metabolism (McLin et al., 2007; Marfil et al., 2010; Sethi &Vidal-Puig, 2010; Liu et al., 2011). Results from functional enrichment analyses provide evidence of pathways that are relevant to both liver metabolism and reproductive function. These pathways may point to some of the links between liver and reproductive function in Bos indicus cattle.

1.3.5. Conclusion

The study successfully exploited RNA-seq data to identify the transcriptomic differences in between pre- and post-pubertal Brahman heifers. Previously, liver transcriptomics in Bos indicus bulls and steers identified DE genes related to feed efficiency (Alexandre et al., 2015; Tizioto et al., 2015). This work is the first attempt to demonstrate the molecular mechanisms of puberty in the liver of Brahman heifers. In the study, 452 DE genes were identified, many of which are closely related to reproductive developmental process,

reproductive structure development, steroid hormone receptor activity and steroid binding. In liver, TGF-β signalling and Wnt signalling genes may play a role in reproductive function. Moreover, the genes, BDNF, STAT6, PBX2, and PBRM1 might impact on the regulation of growth, liver metabolism and puberty onset. As BDNF and estrogen can regulate fatty acid oxidation, this study reasoned that BDNF and estrogen signalling may interact. This interaction can contribute to metabolic changes that can regulate the occurrence of puberty in Brahman heifers. Further studies are warranted to determine the function of these candidate genes. Findings presented in this Chapter provide useful information for understanding molecular mechanisms in the liver that may influence puberty onset of Brahman heifers.

Chapter 1.4. Global transcriptome analysis of adipose tissue identified differentially expressed genes related to puberty in Brahman heifers

Abstract

The adipose tissue regulates a wide range of physiological processes including metabolism, appetite, body weight, inflammation and reproduction. Its ability to modulate reproduction via cross-talk between leptin and the hypothalamic-pituitary-gonadal axis has been well established. To unravel the potential underlying mechanism governing integration between adipose function and puberty, transcriptome from adipose tissue of pre- and post- pubertal Brahman heifers was profiled. Upon comparison, 525 differentially expressed (DE) genes between pre- and post-puberty were identified. Among these DE genes, several key genes were also revealed in adipose transcriptome profile of Brangus heifers in another study. The gene complement C8 alpha chain is proposed as a candidate gene involved in delay of puberty in Brahman heifers, considering our results and literature. DE genes in adipose tissues were enriched for several biological processes such as immune systems, cytokine activity and peptidase inhibitor activity. Wnt and TGF signalling with its related genes, such as SMADs and LEF1, seem to play important roles in puberty, according to pathway enrichment analyses (adjusted P-value < 0.005). Using regulatory impact factor metric, seven transcription factors (TF) with potential regulatory roles in the adipose tissue were identified in this study: HES7, HEY2, RORC, PPARG, PITX1, NFE2 and HOXB5 (P-value < 0.05). These TF genes were significantly DE in the adipose tissue of pre- versus post-pubertal heifers and were also identified as nodes having high connections (more than 20 connections) in gene co-expression network. Additionally, this study did not find a difference in serum leptin concentration between pre- and post-pubertal Brahman heifers. The leptin mRNA expression was not detected in pubertal Brahman heifers. We propose that leptin signalling might not be a key component in the interaction between adipose tissue and puberty in Bos indicus cattle, which is different from the established role of leptin signalling in other mammals. Instead, other regulatory genes and networks seemed more relevant in our model. In summary, this study analysed and discussed the adipose tissue transcriptome of pre- versus post-pubertal Brahman heifers.

Keywords: Adipose tissue, RNA sequencing, C8A, Wnt and TGF signalling pathway

1.4.1. Introduction

The role of adipose tissue for energy storage has been well established for many years. Indeed, surplus energy is efficiently deposited in adipose tissue as neutral triglycerides. When

energy is required, triglycerides in adipose tissue are broken down into fatty acids and glycerol through the lipolytic pathway. The released fatty acids and glycerol are then transported in the bloodstream and subsequently infiltrate into the liver, muscle as well as other organs, orchestrating whole-body energy balance and lipid distribution (Frayn, 2002; Sethi &Vidal- Puig, 2007).

Due to its ability to secrete endocrine factors called adipokines, the adipose tissue has been considered as an active endocrine organ, which is involved in complex physiological processes including metabolism, appetite, body weight, inflammation and reproduction (Kusminski et al., 2016). Numerous animal physiology studies have been focused on mechanisms associated with cattle production. These studies aimed to characterize the molecular mechanism underlying lipid mobilization to meet the energy requisites of milk production or fat deposition to improve meat quality (Wang et al., 2009a; Sumner-Thomson et al., 2011; Jin et al., 2012; Huang et al., 2017). However, the ability of the adipose tissue to modulate reproduction, and/or to respond to reproductive hormones, has been somewhat overlooked.

The adipose tissue produces leptin (LEP), which has been identified as a permissive factor for the onset of puberty and its thought to contributed to fertility maintenance (Moschos et al., 2002; Sanchez-Garrido &Tena-Sempere, 2013; Kawwass et al., 2015; Nelson et al., 2017). The presence of the protein coded by the LEP gene and its receptors in mammalian reproductive organs has been summarized in Table I.1 of this thesis. The phosphoinositide 3- kinase (PI3K) signalling, involved in the regulation of reproduction (Aziz et al., 2014; Beymer et al., 2014; Guillermet-Guibert et al., 2015), has been known to mediate leptin and insulin effects on metabolism (Foukas et al., 2006). A study in male mice models also reported adipose tissue-specific PI3K pathway as an integrator of reproductive function and metabolism (Nelson et al., 2017). Another study in mice recently indicated that leptin signalling in neurons that control appetite and reproduction is sufficient for the onset of puberty in mice (Egan et al., 2017). In short, the adipose tissue can integrate energy status with the reproductive system function.

Identifying adipose tissue-specific biological mechanisms creating the integration between energy status and reproduction may help to understand the occurrence of late puberty in Bos indicus cattle. To study the role of the adipose tissue in puberty, RNA sequencing (RNA- Seq) of adipose tissue from pre- and post-pubertal Brahman heifers was utilized. Despite the wide applications of RNA-seq in cattle studies, there has been only one RNA-seq study on

global gene expression changes in the adipose tissue related to heifer puberty and this was carried in Brangus animals (Cánovas et al., 2014a). The study presented in this Chapter is the first transcriptome study in the adipose tissue of Brahman heifers, a predominant breed in tropical regions. The purpose of this study was to understand the difference in expression levels of genes before and after puberty, as well as factors participating in the regulation of gene expression.

1.4.2. Materials and methods All procedures, including animal handling and euthanizing, were approved by the Animal Ethics Committee of The University of Queensland, Production and Companion Animal group (certificate number QAAFI/279/12). A group of 20 heifers was observed via regular ultrasound scanning to identify the occurrence of ovulation, marked by the subsequent presence of a CL. Six pre-pubertal Brahman heifers (mean ± SD; body weight = 338 ± 54.17 kg, condition score = 3.5 ± 0.44 and average daily gain = 0.76 ± 0.1 kg/d) were randomly chosen from the group of animals that had never ovulated. Six post-pubertal Brahman heifers (mean ± SD; body weight = 363 ± 38.62 kg, condition score = 3.75 ± 0.41 and average daily gain = 0.89 ± 0.1 kg/d) were euthanized on the luteal phase of their second cycle.

Serum LEP concentrations were measured using a commercial radioimmunoassay kit (10IGF100 Kit; Bioclone, Sydney, NSW, Australia). The intra-assay coefficient of variation was 4.2% at mean serum LEP concentration of 20.4 ng/mL. The inter-assay coefficient of variation was 6.0% at mean serum LEP concentration of 5.4 ng/mL.

Adipose tissue samples were harvested immediately after slaughter. Twelve abdominal adipose tissue samples, one for each heifer, were snap-frozen in liquid nitrogen and stored at - 800C for further analysis. Total RNA was extracted from adipose tissues samples with a combination of TRIzol and RNeasy methods. The RNA quality was evaluated using an Agilent Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA). RNA integrity number value ranged from 7.0 to 8.5, excepted one control sample that had RIN number lower than 6. Therefore, a total of 11 high-quality RNA samples (RIN > 7) were sent for library preparation and sequencing.

1ug of total RNA was then used for cDNA library construction using the Illumina TrueSeq stranded mRNA Sample Preparation kit. cDNA libraries were then sequenced with an Illumina HiSeq analyser (Illumina Inc., San Diego, CA). Sequencing data were analysed

using CLC Bio Genomic Workbench software (CLC Bio, Aarhus, Denmark). Quality control analysis was performed using procedures described by (Cánovas et al., 2013). Sequence paired- end reads (100 bp) were aligned to the bovine reference genome sequences UMD3.1 (ftp://ftp.ensembl.org/pub/release-77/genbank/bos_taurus/).

Gene expression was normalized by calculating the reads per kilobase per million mapped reads (RPKM) for each sample. Differentially expressed genes between pre- and post- pubertal heifers were identified based on previously described methods (Cánovas et al., 2014a). This study is part of a large experiment where 7 tissues were sampled per animal (hypothalamus, pituitary, liver, ovary, uterus, muscle and adipose tissue). Therefore, differential gene expression across tissues was estimated using a mixed model where base-2 log transformed RPKM was modelled as a function of the fixed effect of the library and of the random effects of gene and the interaction between gene x animal x physiological state x tissue. A differential expressed gene was considered if the gene was at least 2.57 standard deviations from the mean, corresponding to P-value < 0.01.

Significant transcription factors (TF) were identified using the regulatory impact factor (RIF) metrics (Hudson et al., 2009; Reverter et al., 2010). Gene co-expression network was derived using the DE and TS as nodes. The significant co-expression between these nodes was determined using the partial correlation and information theory (PCIT) algorithm (Reverter &Chan, 2008). The enrichment analysis of gene ontologies and pathways were performed using the Database for Annotation, Visualization, and Integrated Discovery (DAVID, http://david.abcc.ncifcrf.gov (Dennis et al., 2003; Huang da et al., 2009). A cut-off of P-value 0.05 (corrected for multiple testing using Benjamini-Hochberg) was set for the enrichment analysis.

1.4.3. Results

An average of 58 million sequence reads was obtained for each individual sample from adipose tissue. Approximately, 74% of the reads were categorized as mapped reads to the bovine reference sequence. The relatively high number of sequence reads generated in adipose transcriptome study were utilized for subsequent analyses.

Leptin is the major hormone secreted by the adipose tissue and is known to increase during puberty. In this study, the mean values of circulating LEP concentrations in serum were not statistically different between pre- and post-pubertal heifers at euthanasia (P-value = 0.35). The average serum LEP levels were 2.67 0.4 ng/ml at pre-puberty and 2.03 0.5 ng/ml at

post-pubertal heifers. In addition, adipose transcriptome did not reveal the expression of leptin mRNA.

Identification of differentially expressed gene RNA sequencing of adipose tissue samples identified 525 differentially expressed (DE) genes (including 100 novel genes), of which 186 were up-regulated and 339 were down- regulated after puberty in Brahman heifers (P-value < 0.01). Ten genes showed a 4-fold change (FC) difference in expression levels with P-value < 0.01 between pre- and post-pubertal heifers (Figure 1.4.1). Further analysis of significant DE genes may help to provide a better understanding of the biological mechanisms in the adipose tissue underlying puberty in Brahman heifers.

We compared the adipose transcriptome profile of pubertal Brangus heifers reported by Cánovas and colleagues (2014a) to our results of Brahman heifers. There were 66 DE genes in common between two studies. However, only 4 DE genes were expressed in the same pattern in the two experiments. The other 62 DE genes were either up-regulated in Brangus and down- regulated in Brahman or vice versa. Most of these DE genes had |FC| higher than 2, indicating, its potential role in puberty onset.

Identification of transcription factors Using RIF metrics, known TF expressed in adipose tissue were filtered for those most consistently associated with DE genes. These metrics identified 52 significant TF (P-value < 0.01). Of the 52 TF, 10 genes (19%) coded for TF of the zinc finger family (ZNF). Two TF, HES6 and NR2F6, were deemed to be key regulators based on their RIF2 scores of -3.00 and 3.13, respectively. The TF genes MSC, JUP and PRDM16 were among top-ranking TF with a RIF1 score of -5.57, -5.31 and -5.22, respectively. Estrogen receptor alpha (ER) and peroxisome proliferator-activated receptor gamma (PPARG) were also identified in this analysis and have been known to play a role in reproduction. Further, out of the 52 TF, 19 TF were themselves identified as DE genes. These DE TF genes are listed in Table 1.4.

Predicted gene co-expression network and sub-network

After retaining gene pairs with a correlation co-expression > |0.90|, the predicted gene co-expression network with 1,684 nodes representing genes and a total of 6,143 edges was constructed (Figure 1.4.2). The four TF having the largest number of connections were AFF4, MGA, ZNF507 and BDP1. The TF AFF4 and MGA had 42 connections with other genes or TF,

which were followed by ZNF507 with 41 connections. The TF BDP1 had 40 connections in the gene co-expression network.

Regarding DE TF genes in the network, out of 19 TF and DE genes, 18 were presented in co-expression networks. More information about its fold change, as well as connections, are reported in Table 1.4. Noteworthy, NFE2, HES7, PITX1, HOXB5 and HEY2 were DE and TF with a high number of connections (more than 20 connections). These genes were co-expressed with more than 20 DE genes or DE TF genes in the adipose gene co-expression network (Table 1.4). Additionally, three DE TF genes: HES7, HEY2 and RORC had some common connections and their increased expression levels were associated with increased mRNA levels of their connection nodes, post-puberty. Among these 3 DE TF genes, the HES7 was listed as an important regulator in RIF metric analysis (RIF1 score of 2.75) and was linked to the important TF PPARG, intensively studied for its roles in energy homeostasis. On the other hand, the reduction of PITX1, HOXB5 and NFE2 were correlated with other down-regulated DE genes. PITX1 was co-expressed with 28 down-regulated genes after puberty (FC < -2.3) and with C8B, also down-regulated but at a lower fold change (FC = -1.77). As a result, due to differential expression and high network connectivity, HES7, HEY2, RORC, PPARG, PITX1, NFE2 and HOXB5 seem to be key regulators in the adipose tissue of pubertal Brahman heifers.

Figure 1.4.1 Plot of differentially expressed genes (n = 525) in adipose tissue samples between pre- vs post-pubertal heifers. The X-axis represents the log2 of average expression while the y-axis represents log2 fold change for each gene. Red dots indicate genes that differ significantly (P-value < 0.01) between the two groups. Genes plotted in the up portion of the graph were expressed at a higher

level in post-pubertal adipose tissue, and genes in the down portion had lower expression levels post- puberty. Gene symbols are provided for genes with a Fold Change |FC| 4 and P-value 0.01.

Table 1.4 Genes that were differentially expressed (DE) and are transcription factors (TF), their connectivity in the co-expression network and fold changes (FC).

Ensembl tag ID DE TF Network Connections FC ENSBTAG00000001562 NFE2 34 -1.47 ENSBTAG00000012436 HES7 30 1.06 ENSBTAG00000004602 PITX1 28 2.30 ENSBTAG00000045835 HOXB5 24 -1.64 ENSBTAG00000000919 HEY2 20 1.07 ENSBTAG00000001803 FLH5 19 0.85 ENSBTAG00000017405 RORC 17 0.82 ENSBTAG00000009631 ZNF683 15 -1.53 ENSBTAG00000017800 DMRT3 14 -2.35 ENSBTAG00000006844 LEF1 13 -1.39 ENSBTAG00000011003 IKZF3 13 -1.24 ENSBTAG00000011734 ANKRD1 12 -1.99 ENSBTAG00000043953 SPIB 7 -1.82 ENSBTAG00000047499 IRX2 3 3.31 ENSBTAG00000040082 HOXA10 2 -2.12 ENSBTAG00000013880 WWC1 2 -0.95 ENSBTAG00000004159 SIX2 1 -1.36 ENSBTAG00000003650 NR4A2 1 0.80 ENSBTAG00000017409 DLX3 0 1.24 Functional enrichment analysis of DE genes and target genes involved in network

Of those transcripts that were differentially expressed, 309 out of 525 transcripts were annotated with 150 GO terms. The GO terms were distributed in the categories as follow: 72.67% in biological process, 6% in cellular components and 21.33% in molecular function. Results showed that 28 GO terms (25.69%, 28/109) were significantly enriched in the biological process category, 4 GO terms (44.44%, 4/9) were significantly enriched in the cellular component category, and 6 GO terms (18.75%, 6/32) were significantly enriched in the molecular function category (adjusted P-values < 0.05). In the biological process category, GO terms were related to immune response including acute inflammatory response, immune

response, positive regulation of immune system process, adaptive immune response and activation of immune response (adjusted P-value < 0.05). Further, regulation of response to external stimulus and positive regulation of response to stimulus were among enriched GO biological process terms. The molecular function enriched terms were peptidase inhibitor activity, enzyme inhibitor activity, endopeptidase inhibitor activity, serine-type endopeptidase inhibitor activity, cytokine activity and calcium ion binding (adjusted P-value < 0.05).

Figure 1.4.2 Adipose tissue gene co-expression network constructed by the partial correlation and information theory algorithm in pubertal Brahman heifers. The entire network comprises of 1,684 nodes (or genes) and 6,143 interactions. The colour spectrum ranges from green to red for low and high number of connection, respectively.

The pathway enrichment analyses also helped to annotate DE genes in terms of their function. The KEGG pathway complement and coagulation cascades (adjusted P = 3.2x10-10), systemic lupus erythematosus (adjusted P = 1.7x10-04) and cytokine-cytokine receptor interaction (adjusted P-value = 0.03) were significant.

The list of genes included in the co-expression network was also targeted in functional enrichment analyses. For this list, the results showed that 59 GO terms (30.41%, 59/194) were significant in the biological process category, 30 GO terms (73.17%, 30/41) were significant in the cellular component category, and 26 GO terms (46.43.75%, 26/56) were significant in the molecular function category (P-value < 0.05, corrected). In the biological process category,

GO terms related to regulation of transcription, gene expression as well as metabolic process were abundant. Moreover, genes involved in TGF signalling pathway (P-value = 1.38 x 10-4) and Wnt signalling pathway (P-value = 0.005) were enriched among target genes. Network genes annotated to the TGF signalling pathway were CREBBP, SMAD6, SMAD4, TFDP1, SP1, SMAD2, SMAD7, RBL1, GDF5, RHOA, E2F4, TGFB1, RBL2, THBS4, SMAD1, ID1, ID3, ID2 and MYC. Note that, GDF5 (FC = -0.74) and THBS4 (FC = 1.33) were genes with differential expression levels in our study. Genes involved in Wnt signalling pathways included CREBBP, TBL1X, SMAD4, NFATC2, SMAD2, SOX17, NFATC1, LEF1, CTBP2, SMAD3, CTNNB1, RHOA, TBL1XR1, TP53, CTBP1, NLK, TCF7L1, RUVBL1, JUN, NFATC3, MYC, NFAT5 and PPARD. While LEF1 (FC = -1.39) was DE; SOX17 (RIF1 = -4.75), CTBP1 (RIF1 = -3.35) and RUVBL1 (RIF1 = -4.63) were among top-ranking TF. These pathways and related genes will be discussed further in the discussion session.

1.4.4. Discussion

Bos indicus is the predominant bovine sub-species in tropical climate due to their adaptation to the harsh environment, however, they are older when they enter puberty in comparison with the Bos taurus sub-species. Fat plays a pivotal role in meat quality; therefore, several transcriptome analyses identified DE genes in adipose tissue related to lipid metabolism (Wang et al., 2009a; Sumner-Thomson et al., 2011; Jin et al., 2012; Huang et al., 2017). There was only one previous study that analysed transcriptomic data of adipose tissue related to puberty, which was carried out in Brangus heifers (Cánovas et al., 2014a). This Brangus experiment reported DE genes in a contrast between pre and post-pubertal heifers, similar to our experiment with Brahman heifers. However, with much greater Bos taurus genetic content, the Brangus heifers may present a different physiology and therefore a different gene expression pattern related to puberty. Knowledge of the differences between Bos indicus and Bos taurus breeds in relation to gene expression profiles, affected pathways and key regulators may help to explain the differences in puberty timing between breeds. In other words, we compare our results with the report from Cánovas and colleagues (2014a) in an attempt to elucidate some underpinning reasons for the delayed age at puberty in Bos indicus cattle.

Among 525 DE genes in adipose tissue of Brahman heifers, 66 genes were also identified as DE genes in Brangus heifers (Cánovas et al., 2014a). Common expression of these genes and their differences in expression patterns between two groups proposed their involvement in modulating the onset of puberty in two different breeds. Notably, most of these common DE genes had |FC| higher than 3 in either Brahman or Brangus, accounting for 83%.

Out of 66 common DE genes, 34 DE genes were also observed as abundant proteins in less fertile cows when compared to control cows using proteomic analysis of pre-ovulatory follicular fluids (Zachut et al., 2016a; Zachut et al., 2016b). Of these 34 abundant proteins, only complement C8 alpha chain (C8A) was identified as a differentially abundant protein in low fertility cows (Zachut et al., 2016a; Zachut et al., 2016b). The abundance of C8A was 1.2- fold higher (P-value < 0.03) in the follicular fluid of less fertile cows than in control cows and was proposed to accelerate follicular growth in the less fertile group (Zachut et al., 2016a; Zachut et al., 2016b). Consistent with follicular fluid study in dairy cow, C8A was also identified in higher abundance in pre-human chorionic gonadotropin (pre-hCG) women (Zamah et al., 2015). In fat transcriptome studies of pre- and post-pubertal heifers, C8A was revealed as highly down-regulated DE gene (FC = -3.12, P-value < 0.01) at post-puberty in Brahman heifers (current chapter) and up-regulated DE gene (FC = 3.17, P-value < 0.01) at post-puberty in Brangus heifers (Cánovas et al., 2014a). C8A was noted to contribute to the role of complement factors in the maturation of oocyte (Yoo et al., 2013). The adipose tissue transcriptomic studies in Brahman heifers (current study) and Brangus heifers (Cánovas et al., 2014a), as well as literatures, suggest further investigation of C8A genes in the context of puberty.

Leptin is an important adipokine secreted from adipose tissue. It has been reported that circulating leptin concentration increased as puberty approaches in rodents, humans and heifers (Ahima et al., 1997; Quinton et al., 1999; Garcia et al., 2002; Vaiciunas et al., 2008). However, this current study did not find a difference in serum leptin concentration between pre- and post- pubertal Brahman heifers. In fact, leptin concentration in post-pubertal heifers was lower in compared to pre-pubertal heifers. It must be noted that serum LEP concentration at post- pubertal was recorded 23 days, on average, after the observation of the first CL. Meanwhile, the pre-pubertal heifers had never ovulated. There is the possibility that measurement time of plasma leptin concentration in this study could impact on our results.

Regarding leptin mRNA expression, this current study did not identify its expression in pre- and post-pubertal Brahman heifers. The different of leptin mRNA level between pre- and post-pubertal Brangus heifers was not statistically significant (Cánovas et al. 2014a). Furthermore, Maciel et al. (2004) tested the effect of leptin on the onset of puberty in heifers (Brahman x Hereford) and found no effect of chronical administration of leptin to induce puberty or regulate endocrine characteristics nearing the pubertal time in beef heifer. A study in human also noted that the puberty process did not result from the increase of leptin mRNA

or serum leptin in girls (Kulik-Rechberger et al., 2008). Kulik-Rechberger et al. (2008) reported no significant difference in leptin mRNA and serum in pre- and post-pubertal girls. The lack of a significant effect of leptin on the reproductive process has been reported in hairy sheep (Catunda et al., 2013). A study in Nellore heifers indicated that the concentration of leptin only rose as heifers became older (Ferraz et al., 2018). Supplementary feeding to unfavourable genetic merit for age at first calving in Nellore heifers was not able to decrease age at puberty in this breed (Ferraz et al., 2018). There evidence suggest that leptin as a permissive signal of the metabolic factors allowing puberty initiation (Maciel et al., 2004; Roa et al., 2010). Therefore, identification of other factors in addition to leptin could help to understand the late at puberty onset of Bos indicus breeds.

In addition to genes showing differential expression between pre and post-pubertal animals, the identification of transcription factors and their interaction with DE genes could provide insights into biological mechanisms. The study retrieved information from AnimalTFDB for expressed transcription factors. Interestingly, there were 19 DE genes were also identified as TF. Only the DE HES7 gene was noted as significant TF (RIF1 score of 2.75) in terms of their correlation with DE genes. A study in rat also noted an increase of HES7 expression in visceral adipose tissues in comparison to subcutaneous adipose tissue (Ferrer- Lorente et al., 2014). The expression of HES7 gene was regulated by the Notch signalling pathway (Bessho et al., 2001). Obviously, the Notch signalling pathway has been known to involve in adipocytes developments and differentiation as well as in energy metabolism (Gridley &Kajimura, 2014; Bi &Kuang, 2015; Chartoumpekis et al., 2015; Bi et al., 2016; Shan et al., 2017). The increase of Notch signalling in adipocytes results in ectopic lipid accumulation and insulin resistance in mice (Chartoumpekis et al., 2015). Further, inhibition of Notch signalling decreased leptin mRNA expression in adipose tissue. Treating LEPob mice with γ-secretase inhibitors, an inhibitor of Notch activity, decreased body weight and improved glucose metabolism (Bi et al., 2014). The effect of inhibition of Notch activity not only elicited adipocytes but also targeted other important metabolic organs such as the liver and intestine. The differential expression of HES7 gene after puberty in adipose tissue, its transcriptional repressor function as well as the role of its downstream pathway (Notch signalling) in energy metabolism and obesity made it as a potential candidate to integrating energy metabolism and puberty in Bos indicus heifers.

Consistent with previous studies (Chapter 1.1, 1.2 and 1.3), genes belong to ZNF family were among top-ranking TF in adipose tissue, accounting for 19%. Combining data from 4

chapters, approximately 20% of TF identified by RIF analysis of hypothalamus (Chapter 1.1), pituitary (Chapter 1.2), liver (Chapter 1.3) and adipose (Chapter 1.4) transcriptome belong to the ZNF family. A high proportion of ZNF genes at puberty in multi-tissue transcriptomic studies suggested its potential contribution to the pubertal process in cattle. Noteworthy, Zinc finger genes have been involved in the epigenetic regulation of puberty in female primate (Lomniczi et al., 2015b). Using a system biology approach, ZNF genes were considered as repressors of repressors in puberty network (Lomniczi et al., 2013). A ZNF sub-network of ovary transcriptomic study has been published for Brahman pubertal heifers (Nguyen et al., 2017). This study reported the decrease of ZNF mRNA expressions after puberty. The involvement of ZNF genes in the onset of puberty was also suggested by several GWAS which showed the association between age of menarche and SNPs located near ZNF483, ZNF462 and ZNF131 (Perry et al., 2009b; Elks et al., 2010; Perry et al., 2014a). Other studies also noted the reduction of ZNF expression before puberty (Abreu et al., 2013; Lomniczi et al., 2015b). In addition, makorin ring finger protein 3 (MKRN3), a protein with 4 zinc finger domains, is important for the initiation of puberty (Liu et al., 2017). These findings again confirm ZNF’s function in puberty initiation.

In adipose gene co-expression network in Brahman (current study), HES7, HEY2, RORC, PPARG, PITX1, NFE2 and HOXB5 deems to be key regulators involving in puberty in Brahman heifers. As discussed previously, HES7 is a transcriptional repressor which can influence energy metabolism and obesity. This gene was also co-expressed with PPARG, a key regulator for cattle puberty (Fortes et al., 2010a; Ravasi et al., 2010; Fortes et al., 2011). The HEY2 gene was also involved in Notch signalling and differentially expressed in developing mouse mammary gland (Park et al., 2013). Polymorphisms located in RORC was associated with meat production in beef cattle (Gorlov et al., 2014), showing its relation to fat content. These DE TF genes: HES7, HEY2, RORC and PPARG shared common target genes and their mRNA level were increased at post-puberty, suggesting the up-regulation of pathways for which they were involved.

In addition to these above-mentioned DE TF, PITX1, NFE2 and HOXB5 were decreased expression levels after puberty. The targets (mRNAs) which connected with PITX1, NFE2 and HOXB5 in the network were also down-regulated DE genes. Of note, PITX1 gene has been shown to modulate the expression of LHb, FSHB and GnRH receptor (Quirk et al., 2001; Zakaria et al., 2002; Jeong et al., 2004a). The DE TF, NFE2 is a regulator of the oxidative stress response (Williams et al., 2016). In addition, homeobox genes were noted as

differentially expressed genes in adipose deposition in sheep and as regulators in adipose transcriptome study in Brangus heifers (Cánovas et al., 2014a; Kang et al., 2017). Remarkably, PITX1 and C8A have a common target gene, C8B. As mentioned above, C8A was highly down- regulated at puberty in Brahman (this current study) and up-regulated in Brangus heifers (Cánovas et al., 2014a). The decrease C8A mRNA level after puberty was correlated with 19 down-regulated DE genes in adipose gene-expression network of Brahman heifers. The differential expression and their transcriptional functions in energy metabolism, immune system and adipogenesis suggested the role of these candidate genes in puberty initiation.

The analysis of gene ontology for DE genes revealed numerous enriched GO terms such as immune responses, response to stimulus, catalytic and binding activity, peptidase inhibitor activity as well as cytokine-cytokine receptor interactions. Adipose tissue mainly comprises adipocytes, immune cells and vascular endothelial cells. These components interact with each other and impact on adipose tissue function. Genes were involved in several immune responses, indicating that the immune system plays a role in control reproduction. A study in human reported that reproduction is intimately connected to immune function in women (Abrams &Miller, 2011). Pathway analysis of differential abundance protein in follicular fluids of less fertile cows indicated immune response as the top pathways (Zachut et al., 2016a). The differentially abundant protein (C8A) infertile cow was part of the innate immune system. In addition to this, genes were involved in response to stimulus, confirming that puberty process generates some stress for beef cattle. This is logical because cattle will require to have sufficient energy for puberty. Moreover, genes were also involved in several catalytic activity and binding activity in the molecular function category, demonstrating that pubertal heifers had strong metabolic activity. Moreover, many biologically active peptides have been known to modulate reproductive function. Peptide hormones are metabolized by peptidase, therefore the identification of enriched GO terms related to peptidase inhibitor activity, enzyme inhibitor activity, endopeptidase inhibitor activity, serine-type endopeptidase inhibitor activity (adjusted P < 0.05) suggested the role of these DE genes in modulating puberty system. Further, cytokine-cytokine receptor interaction was detected as a down-regulated pathway in the period from puberty to pregnancy in mouse models (Zhao et al., 2012).

Regarding KEGG analyses of target genes including DE genes and TF, TGF- and Wnt signalling pathways were identified as significant annotation terms. These enriched pathways were also found among DE genes and TF between pre- and post-pubertal Brahman heifers in

the transcriptome studies of the pituitary gland and the liver (Chapter 1.2 and 1.3). More detail about these pathway function and involved genes will be discussed in following paragraphs.

The Wnt signalling is involved in numerous biological processes including tissue regeneration and plasticity as well as embryonic development. In metabolically relevant tissues, the adaptation of these tissues to nutritional challenges is facilitated by Wnt signalling (Sethi &Vidal-Puig, 2010). Transgenic and in vitro studies noted that Wnt signalling can inhibit murine fat differentiation (Ross et al., 2000; Bennett et al., 2002; Longo et al., 2004; Wright et al., 2007; Sethi &Vidal-Puig, 2010). Further, activation of Wnt/-catenin in adipose progenitor cells resulted in adipose tissue degeneration and subsequently increased glucose uptake in muscle via undefined muscle-fat endocrine axis (Papatriantafyllou, 2012). Glucose is a main circulating nutrient for cellular energy balance. Of note, lymphoid enhancer-binding factor (LEF1) has been known as a key TF in the Wnt signalling pathway (Meece et al., 2007). As expected, LEF1 was involved in Wnt signalling pathway in the adipose transcriptomic study of Brahman heifers (current study). The decrease of LEF1 mRNA at puberty suggested the down-regulation of Wnt signalling pathway which then can increase glucose level in the blood. The circulating glucose level will reflect available energy for adequate LH secretion and ovarian stimulation. In addition to LEF1, other TF involved in adipose gene co-expression network of pubertal heifers were also detected in this pathway. The potential link between these TF such as TP53, RAC1 and NFAT5 and puberty has been previously described in Chapter 1.2.

The TGF signalling pathway holds an important role in adipogenesis, lipid and glucose metabolism, insulin resistance and reactive oxygen species production (Tsurutani et al., 2011; Tan et al., 2012). In fact, the TGF pathway transmits its signals via TF called SMADs, with the principal facilitator SMAD3 (Feng &Derynck, 2005). A study in mice noted that the inhibition of TGF signalling can improve adipose function and improved metabolic function (Tan et al., 2012). The involvement of TGF signalling pathway in the pathogenesis of obesity has underscored its effect on adiposity and metabolism (Tan et al., 2012). Excess adiposity may affect timing of puberty onset and hormonal parameters during puberty. More importantly, cross-talk between TGF and Wnt signalling has been reported. SMADs and LEF/-catenin, principal facilitators in these pathways, synergistically regulate a set of shared genes (Willert et al., 2002; Schwartz et al., 2003). This study identified SMADs genes in both TGF and Wnt signalling pathways. Overall, the activation of TGF signalling and genes enriched in this pathway may have an inhibitory effect on puberty initiation.

1.4.5. Conclusion

RNA sequencing technology and bioinformatics methods were applied to identify the adipose transcriptomic difference between pre- and post-pubertal Brahman heifers and explore the molecular mechanisms of puberty. In the study, 525 DE genes were identified, 66 genes of which are also detected in adipose transcriptome profile of Brangus heifers. The decrease of C8A mRNA level after puberty in Brahman was proposed to play a role in the delay of puberty in the breed. Many DE genes were also revealed as key regulators in puberty such as HES7, HEY2, RORC, PPARG, PITX1, NFE2 and HOXB5, suggesting further investigation in the context of puberty for these regulators. In addition, leptin might have a permissive effect on puberty initiation in Brahman. The effect of Wnt and TGF signalling in adipogenesis, glucose uptake and insulin resistance provided the link to nutritional changes and puberty. This study provides useful knowledge of basis molecular mechanism of puberty in Brahman and selecting beef cattle with early pubertal trait.

PART 2 – PROTEOMIC STUDIES

Chapter 2.1. Hypothalamic proteomics revealed differentially abundant proteins and pathways for puberty process in Brahman heifers

Abstract

It is well-established that the initiation of mammalian puberty requires an increase in the pulsatile release of gonadotropin-releasing hormone (GnRH) from the hypothalamus. The hypothalamic GnRH pulse is regulated by multi-level signals, which can be species-specific and reproductive stage-dependent. Bos indicus breeds are better adapted to tropical regions, but they are often older than Bos taurus breeds at puberty. The objective of this study was to measure protein abundance related to puberty in the hypothalamus of Bos indicus cattle, more specifically Brahman heifers. The study selected pre- and post-pubertal heifers according to the observation of corpus luteum. The proteins extracted from hypothalamus samples of pre- and post-pubertal heifers were identified and quantified using mass spectrometry. These proteomic data are publicly available via ProteomeXchange with identifier PXD009620. Among quantified proteins, 275 proteins were differentially abundant (DA) between pre- and post-puberty heifers (adjusted P-value < 1%). The phospholipase C beta 1 (PLCB1) protein that plays a pivotal role in puberty was identified as a DA protein in this study as well as a differentially expressed gene in the hypothalamic transcriptome study (Chapter 1.1). KEGG pathway analyses revealed that highly abundant proteins (adjusted P-value < 0.05) in post- pubertal heifers were significantly enriched for pathways involved in puberty processes such as GnRH, calcium, and oxytocin signalling pathways. Meanwhile, lowly abundant proteins (adjusted P-value < 0.05) at post-puberty were significantly enriched for estrogen signalling pathway, axon guidance and GABAergic synapse. Other pathways like oxidative phosphorylation, the tricarboxylic acid cycle, non-alcoholic fatty liver disease, glycolysis and gluconeogenesis, metabolism of amino acids and carbohydrate metabolisms were among the enriched pathways for the DA proteins in the hypothalamus (adjusted P-value < 0.05). Protein- protein interaction network showed HSP90AA1, MDH2 and ALB as major hubs. The protein HSP90AA1 was identified as down-regulated in pre-pubertal mammary epithelial cells in a previous study. Meanwhile, the protein MDH2 has been known as protein marker for male fertility. In the network, both HSP90AA1 and ALB were interacting with other DA proteins involved in estrogen and oxytocin signalling, suggesting potential roles for these hubs in the onset of puberty in the brain. In summary, this is the first study that analysed and discussed the proteome of the hypothalamic tissue of Brahman heifers related to the process of puberty.

Keywords: puberty, Brahman heifers, PCLB1, GnRH signalling, HSP90AA1, MDH2 and ALB

2.1.1. Introduction

As reviewed in Section I of this Thesis, farm animal biology can be studied through high-throughput omics techniques at different levels. Transcriptomic (i.e., mRNA) was used to study changes in Brahman heifers as described in the first 4 chapters (Chapter 1.1 to 1.4). However, the reported weak correlation between mRNA and protein levels suggests there is benefit from the integration of different levels of evidence (i.e. RNA and proteins) to provide insights into any biological process. Multiple omics data was used to predict an interactive gene network for puberty in Brangus cattle (Cánovas et al., 2014a). Using multiple omics techniques as well as several samples from the same animals to study multiple levels of transcripts and proteins can aid to the understanding of underlying mechanisms.

It is well-established that the initiation of mammalian puberty requires an increase in the pulsatile release of gonadotropin-releasing hormone (GnRH) from the hypothalamus. This sustained increase is influenced by trans-synaptic and glial inputs to the GnRH neuronal network (Lomniczi et al., 2015c). While the trans-synaptic signalling involves neuronal excitatory and inhibitory inputs, the glial input is almost all excitatory (Prevot, 2002). Further, gonadal steroid hormones play a critical role in feedback effects on GnRH secretion in the hypothalamus (Bliss et al., 2010). Ojeda et al. (2009) first proposed a gene network involved in the neuroendocrine regulation of puberty process. Crucial individual components of regulatory systems in this network continue to be clarified. In short, the regulation of the GnRH trigger for puberty is multi-level and multi-factorial, fitting for a complex and polygenic inheritance that was described for age at puberty in cattle and humans (Perry et al., 2009a; Fortes et al., 2010b).

Reproduction traits that describe cow fertility are relevant for beef and dairy production. Age of puberty is one of the reproductive traits measured that has been overlooked by current proteomics studies. Current proteomic studies have focused on the abundance and type of proteins expressed in pregnant and lactating cows (Pyo et al., 2003; Kim et al., 2005; Lippolis &Reinhardt, 2005; Cairoli et al., 2006; Klisch et al., 2006; Riding et al., 2008; Kuhla et al., 2009; Kim et al., 2010; Li, 2012; Munoz et al., 2012; Yang et al., 2012b; Kurpińska et al., 2014; Lee et al., 2015; Rawat et al., 2016). These studies showed differential protein abundance, when pregnant and non-pregnant cows or lactating versus dry cows were compared. The take home message from these previous studies is that circulating progesterone levels impact on protein expression in the uterus and the uterine fluid in cows. While those studies focussed on the uterus or uterine fluid, some studies of hypothalamic proteomics

evaluated the difference in protein abundance related to feed restriction, as oppose to normal diet in cattle (Kuhla et al., 2007; Kuhla et al., 2010). The current study described in this chapter is the first proteomic study focused on hypothalamus and puberty in cattle.

We compared the protein abundance in hypothalamus samples from pre- and post- pubertal Brahman heifers. The dataset presented here renders itself to the investigation of puberty in Bos indicus, a pre-dominant sub-species in tropical and sub-tropical regions worldwide. In more detail, liquid chromatography electrospray ionization tandem mass spectrometric (LC-ESI-MS/MS) was used to identify proteins differentially abundant (DA) in the hypothalamic tissue of pre- and post-pubertal Brahman heifers. Bioinformatics tools were employed to determine Gene Ontology (GO) terms and Kyoto Encyclopedia of Genes and Genome (KEGG) pathways that were overrepresented in the list of DA proteins. Using the same samples and animals from the transcriptome study (Chapter 1.1), we estimated the correlation between mRNA and protein profiles. The DA proteins were also used to predict an interaction network. Network analyses allowed us to identify proteins that were central hubs in the network. Hub proteins are likely to control other proteins and therefore drive puberty.

2.1.2. Materials and methods Animals The twelve (12) Brahman heifers utilized in proteomic experiments were the same as those used in transcriptomic studies. These animals were housed at the Gatton Campus facilities of the University of Queensland. The study was conducted in accordance with the Animal Ethics Committee of the University of Queensland, Production and Companion Animal group (certificate number QAAFI/279/12). There was no difference (P-value > 0.05) in body weight and body condition scores between pre- and post-pubertal heifers. The presence of the first corpus luteum (CL) and concentration of progesterone were used to define pubertal status. Plasma progesterone concentrations were 0.4 0.2 ng/mL and 2.0 0.7 ng/mL in pre- and post-pubertal heifers, respectively. Post-pubertal heifers were euthanized around day 15 after observation of the first CL. On slaughter day, a pre-pubertal heifer was randomly selected and paired with the post-pubertal heifer that was on the luteal phase of the cycle. After slaughter, the hypothalamic tissue samples were harvested and processed as described next.

Protein Sample preparation

The hypothalamic tissue samples were manually homogenized into smaller fragments using hard aluminium foil to protect the tissue and a hammer to “pulverize” the sample over

dry ice bed to preserve the sample and randomize the fragments. This procedure ensured that each sample was representative of the entire hypothalamic tissue. The resulting tissue fragments were transferred into Eppendorf lobind microcentrifuge tubes (Sigma-Aldrich) and lysis solution (7M urea, 2M thiourea, 4%SDS, 10mM DTT and 1mM PMSF) were added. The fragments in solution were sonicated at power level 4 for 10 seconds. Subsequently, the homogenate was vortexed vigorously for 1 hour at 30°C. Following that, 25mM acrylamide was added to the samples and subsequently incubated for 1 hour at 30°C. Next, 5mM DTT was added into samples in order to quench excess acrylamide. Four volumes of methanol: acetone (1:1) were added, and the samples precipitated overnight at -20°C. The precipitates were subsequently dissolved by adding 50mM ammonium acetate, and the protein concentrations were measured with Nanodrop (Thermo Scientific). After measuring concentration, about 100µg of protein were transferred into a 10 kDa Amicon Ultra 0.5 centrifugal filter device (Merck Millipore). This filter device was then inserted into a collection tube and centrifuged at maximum speed for at least 30 minutes. Protein was diluted in 50mM ammonium bicarbonate and then again centrifuged at maximum speed for at least 30 minutes. Trypsin was used as a protease for digestion of protein solution followed by an incubation overnight at 37°C. Peptides were desalted by C-18 Zip-tip (adapted from Millipore procedure) and stored at -20°C until analysis. These procedures were adapted from Tan and colleagues (2014).

Mass Spectrometry and data analysis

The LC-ESI-MS/MS were performed using a Prominence nanoLC system (Shimadzu) and TripleToF 5600 mass spectrometer with a Nanospray III interface (SCIEX) as previously described (Xu et al., 2015). Peptides were separated using a 70 min LC gradient. MS-TOF scan was performed. We performed information-dependent acquisition (IDA) of top peptides for 1 randomly chosen pre-pubertal sample and 1 randomly chosen post-pubertal sample. Subsequently, sequential window acquisition of all theoretical mass spectra (SWATH) was performed on all samples using prepared IDA library for each group. Protein Pilot v5.0.1 (ABSCIEX) was then utilized for peptide identification. Bovine protein database was used for peptide mapping and retrieved from UniProt (www.uniprot.org; 43,813 entries assigned to Bos taurus). Identified peptides with more than 99% confidence and with false discovery rate (FDR) less than 1% were used for subsequent analyses. Ion libraries were used for SWATH analyses. The abundance of proteins and peptides was computed by PeakView v2.1 software (ABSCIEX). Statistical analyses were then performed using MSstats (v2.6) in R (Choi et al., 2014).

100

Correlation analysis The study applied Pearson correlation analysis to calculate the correlation between mRNA expression and protein abundance in the hypothalamus from pre- and post-pubertal Brahman heifers. We estimated the correlation between the log-fold change (FC) values for mRNA and protein expression between the two groups. The correlation coefficient was calculated using the cor() function in the stats R package with the Pearson option. Functional enrichment analysis of DA proteins The GO terms and KEGG pathways related to the DA proteins were identified using STRINGv10 database (http://string-db.org). In general, GO terms comprise of three categories: biological process (BP), cellular component (CC) and molecular function (MF). Over- represented GO terms help to describe the gene function associated with the list of DA proteins. Pathway enrichment analyses can identify metabolic and signal transduction pathways associated with the DA proteins. The list of DA protein between pre- and post-puberty was submitted to Stringv10 database for KEGG pathway analyses. Benjamini-Hochberg method was utilized to correct for multiple tests and a corrected P-value (FDR) 5% was observed to identify significant enrichment GO terms and KEGG pathways. Figure 2.1.1 shows an experimental workflow including sample preparation, LC-ESI-MS/MS analyses and functional enrichment analyses conducted in this work.

Figure 2.1.1 Schematic overview of experimental workflow for proteomics.

101

Protein-protein interaction network inference

We used STRINGv10 (http://string-db.org) to infer a protein-protein interaction network for the list of DA proteins. STRINGv10 uses a collection of known interactions inferred from curated databases, experimental evidence, and predictions from information about coding genes and co-expression to estimate a score for each interaction in its output. The higher the score, the more evidence for the interaction between proteins. Protein interactions with a score > 0.6 were selected for further investigation. Cytoscape (Shannon et al., 2003) was utilized to visualize the protein-protein interaction network and to identify key proteins.

2.1.3. Results

Proteins expressed in the hypothalamus of pre- and post-pubertal Brahman heifers were characterized using LC-ESI-MS/MS and quantitative proteomics methods (SWATH). The hypothalamic data was submitted into Proteome Xchange Consortium via PRIDE partner repository with identifier PXD009620 (Vizcaino et al., 2016). LC-ESI-MS/MS identified 1045 proteins that formed the IDA library. The relative protein quantification was obtained using SWATH analyses. Only proteins having consistent quantitative results in two states were included in further analysis. Quantitative analysis identified 765 proteins (FDR < 0.01). Among these quantified proteins, 275 proteins were DA between pre- and post-puberty (adjusted P- value < 0.01). Out of 275 DA proteins, 120 proteins showed decreased abundance and 155 proteins were increased post-puberty (adjusted P-value < 0.01). Interestingly, there were 8 DA proteins having 1-fold change in abundance between pre- and post-pubertal animals (adjusted P-value < 10-5). The DA proteins were the target of further analyses, such as enrichment analyses and network prediction.

The expression levels of proteins and its corresponding mRNA levels (Chapter 1.1) were investigated. If the matching pair of protein and corresponding mRNA was not identified, the protein was excluded in this analysis. In total, 648 of 765 identified proteins corresponded to mRNA transcripts quantified in the hypothalamic transcriptome data (Chapter 1.1). The correlation between transcript and protein in hypothalamus for 648 proteins was positive but not significant, with correlation coefficient r = 0.05 (P-value = 0.16). The absence of a significant correlation between transcripts and corresponding proteins may due to multiple factors. Factors such as sample processing, technological sensitives, and post-translation regulation of gene expression may play a role in the absence of correlation between transcripts and proteins. The next comparison between transcriptome and proteome data was to investigate

102

if DE genes were also DA proteins. Only eight DE genes were also and DA proteins in the hypothalamus: TUBB2B, OMG, SPTAN1, RAP1A, SSBP1, PLCB1, EIF3J and H2AFX (Table 2.1.1). Six of these eight DA proteins showed the same direction of regulation as their mRNAs.

Table 2.1.1 List of common differentially expressed genes and abundant proteins in two hypothalamus studies

Ensembl ID UniProt ID Gene symbol FC_mRNA FC_protein

ENSBTAG00000004093 Q6B856 TUBB2B -0.203 -0.394

ENSBTAG00000025213 Q0IIH3 OMG 0.199 -0.297

ENSBTAG00000015327 E1BFB0 SPTAN1 -0.131 -0.117

ENSBTAG00000014710 P62833 RAP1A 0.113 0.087

ENSBTAG00000010931 F1N1S0 SSBP1 0.189 0.321

ENSBTAG00000008338 P10894 PLCB1 -0.319 0.779

ENSBTAG00000000359 G8JKV2 EIF3J 0.119 0.845

ENSBTAG00000038047 Q17QG8 H2AFX 0.199 1.026

To understand the difference between pre- and post-pubertal heifers, DA proteins (adjusted P-value < 1%) were characterized by enriched GO terms. A total of 275 DA proteins (adjusted P-value < 1%) were categorized into three functional groups: BP, CC and MF. In BP, 341 significant GO terms (FDR < 0.05) were found. DA proteins were mainly enriched for the BP term “organic substance metabolic process” (31%), “single-organism metabolic process” (27%), “oxidation-reduction process” (14%) and “nervous system development” (12%). As expected, DA proteins were involved in peptide metabolic process and brain development (including development of neurons, glial cells and axons). These DA proteins were also annotated to the terms “response to oxidative stress”, “glutamine metabolic and catabolic process”, “energy derivation by oxidation of organic compounds”, and “regulation of stress- activated MAPK cascade” (FDR < 0.05). There were 116 enriched GO terms found in CC category (FDR < 0.05) and the majority of DA proteins were located in cytoplasmic (51%), and intracellular (45%) space. In regards to MF group, 89 enriched GO terms were identified (FDR < 0.05) and represented by proteins involved in binding (38%) and catalytic activity (34%).

103

The DA proteins were investigated for enriched KEGG pathways (adjusted P-value < 0.01). The analysis revealed 67 enriched pathways. Enriched pathways were estrogen signalling pathway (FDR = 0.002), axon guidance (FDR = 0.006), Glutamatergic synapse (FDR = 0.0001), GABAergic synapse (FDR = 0.005), GnRH signalling pathway (FDR = 0.02), oxytocin signalling pathway (FDR = 0.001), calcium signalling (FDR = 0.01) and PPAR signalling pathway (FDR = 0.05). These enriched pathways could be expected given the known biology of pubertal development, as discussed in the following section (Table 2.1.2). DA proteins involved in GnRH signalling pathway were highlighted in Figure 2.1.2. In addition, DA proteins were also involved in other pathways such as oxidative phosphorylation (FDR = 6.6 x 10-12), the tricarboxylic acid cycle (TCA, FDR = 3.58 x 10-9), non-alcoholic fatty liver disease (NAFLD, FDR = 1.08 x 10-5), glycolysis and gluconeogenesis (FDR < 0.05), thyroid hormone synthesis (FDR = 0.05), metabolism of amino acids and carbohydrate metabolism (FDR < 0.05). The DE gene and corresponding DA protein, PLCB1, was annotated to four enriched pathways: GnRH signalling, estrogen signalling, calcium signalling and oxytocin signalling. In short, pathways related to neuronal-hormonal signalling and metabolism were significant for the DA proteins.

The list of DA proteins was used for interaction network inference. Interaction scores higher than 0.6 underpin the edges of the predicted protein-protein interaction network for the hypothalamus data (Figure 2.1.3). The network contained 181 nodes (proteins) with 486 edges (interactions). Heat shock protein 90 alpha family class A member 1 (HSP90AA1) had the highest number of connections in the network (34 edges), followed by malate dehydrogenase 2 (MDH2) and albumin (ALB) with 23 and 21 connections, respectively. Of note, HSP90AA1 contributes to the estrogen signalling pathway, according to the KEGG pathway analysis (Table 2.1.2). In addition, KRAS protein, which plays a role in GnRH signalling, axon guidance, estrogen signalling and oxytocin signalling interacted with 13 proteins in the network. PLCB1, a DE gene and DA protein had nine connections in the network: GNAQ, GNAI2, PIP4K2A, GNAO1, SYNJ1, GNB2, PRKCG, CAMK2D, CAMK2A. These genes that were hubs in the network and were associated with enriched pathways of biological relevance for puberty may be important drivers of puberty in cattle.

104

Table 2.1.2 Enriched pathways related to differentially abundant proteins in heifer puberty

Pathway name Count False discovery rate Proteins involved GnRH signalling 5 0.02 GNAQ, PLCB1, CAMK2A, pathway CAMK2D, KRAS Calcium signalling 8 0.01 SLC25A6, ATP2B3, GNAQ, pathway PLCB1, PRKCG, CAMK2A, CAMK2D, ATP2B4 Oxytocin signalling 9 0.001 EEF2, PRKCG, CAMK2A, pathway GNAQ, CAMK2D, PLCB1, GNAI2, GNAO1, KRAS Axon guidance 7 0.006 GNAI2, RAC1, KRAS, CDK5, DPYSL2, DPYSL5, CFL1 GABAergic synapse 6 0.004 GAD2, GLUL, PRKCG, GNAO1, GNAI2, GNB2 Estrogen signalling 7 0.002 HSPA8, GNAO1, KRAS, GNAI2, pathway HSP90AA1, PLCB1, GNAQ PPAR signalling pathway 4 0.05 DBI, ILK, APOA2, FABP7 Low and high abundant proteins at post-puberty (adjusted P-value < 0.01) appear in normal and bold type, respectively.

Figure 2.1.2 Gonadotropin-releasing hormone (GnRH) signalling pathway containing differentially abundant proteins in the hypothalamus of pre- and post-pubertal Brahman heifers. Pathway components that represent differentially abundant proteins are in orange. Pathway was adapted from the KEGG database (http://www.genome.jp/kegg-bin/show_pathway?bta04912, accessed: 11/03/2018).

105

Figure 2.1.3 Protein-protein interaction network in the hypothalamic tissue. Each node represents a protein. Edges illustrate the interaction between proteins. Node size represents the level of differential abundance measured for each protein in absolute terms, that is bigger nodes represent higher fold changes. Node colour range from green to red represent low to high connectivity in the network for each specific protein.

2.1.4. Discussion

High-throughput techniques such as RNA sequencing and proteomics analyses provided powerful methods to understand gene expression and regulation. However, no single approach can provide a complete understanding of the fundamental biology underlying complex traits. Analyses at multiple levels are needed to fully interpret complex biological systems. In this study, proteomics using high-throughput LC-ESI-MS/MS revealed a total of 765 proteins (FDR < 0.01) expressed in the hypothalamus. Among these 765 proteins, the corresponding mRNAs for 648 proteins were also identified in the hypothalamic transcriptomic study (Chapter 1.1). However, the correlation between protein abundance and mRNA expression in the hypothalamus was not significant (r = 0.05, P-value = 0.16). To date, reports on the correlation between mRNA and proteins are scarce. Yeast, bacteria and human cancer data reports found limited correlations between mRNA and protein levels (Anderson &Seilhamer, 1997; Gygi et al., 1999; Chen et al., 2002; Griffin et al., 2002; Ghaemmaghami

106

et al., 2003; Greenbaum et al., 2003; Washburn et al., 2003; Tian et al., 2004; Nie et al., 2006; Jayapal et al., 2008; de Sousa Abreu et al., 2009; Haider &Pal, 2013; Bai et al., 2015; Carvalhais et al., 2015; Payne, 2015; Edfors et al., 2016). A weak correlation of 0.05 was reported for 17 DE genes and corresponding protein levels analysed in bovine mammary gland (Dai et al., 2017). The weak correlation between mRNA and protein expression levels may result from methodological constraints and biological factors (translational regulation as well as protein in vivo half-lives) (Maier et al., 2009; Vogel &Marcotte, 2012; Haider &Pal, 2013). Additional data on factors affecting protein levels as well as in-depth studies can help address this conundrum.

In spite of the lack of correlation overall, eight DA proteins were also DE genes in these pubertal Brahman heifers (Chapter 1.1). The DA proteins that were also DE genes were: TUBB2B, OMG, SPTAN1, RAP1A, SSBP1, PLCB1, EIF3J and H2AFX. Only PLCB1 and OMG had different patterns in the two studies. The lower rate of mRNA production and short mRNA half-life could explain the higher rate of proteins abundance in the hypothalamus of Brahman heifers (Yen et al., 2008; Sharova et al., 2009; Eden et al., 2011; Schwanhausser et al., 2011; Vogel &Marcotte, 2012). These DE genes and DA proteins were not listed as DE genes in the hypothalamic transcriptome of pubertal Brangus heifers (Cánovas et al., 2014a). This could be possible because of breed differences and differences in the methods for selecting pre-pubertal heifers. In the Brangus study, the pre-pubertal heifers were different from the post- pubertal heifers in terms of weight (this is not the case for our Brahman heifers). It remains to be confirmed if these genes are relevant for puberty only in the Bos indicus context or not.

Brain development and axon guidance are a prominent feature during pubertal development. Among the eight DA proteins, tubulin beta 2B class IIb (TUBB2B) is known to be involved in neuronal migration and axonal guidance (Bond et al., 1984; Romaniello et al., 2015). A study in female rat noted that the neurotrophic effects of estrogen on the central nervous system at the onset of puberty were partly mediated by the increase of tubulin beta class II (Rogers et al., 1993). Further, the significant increase of TUBB2B and TUBB2A mRNA expression in low progesterone in heifer endometrium was reported (Forde et al., 2012). The hypothalamic transcriptome (Chapter 1.1) and proteome (current study) revealed decreased abundance of TUBB2B at post-puberty, suggesting the correlation between TUBB2B and steroid hormones. High expression of TUBB2B in pre-pubertal Brahman heifers can be proposed to play a part in the negative feedback of ovarian hormones to the central nervous system. This is an emerging hypothesis from our data that requires further research.

107

Another interesting protein among the eight that were DA and DE genes is spectrin alpha nonerythrocytic 1 (SPTAN1). This protein was identified as an upregulated gene in the mammary epithelial cells of pre-pubertal female mice in comparison to post-pubertal mice (Pal et al., 2017). These results in mice seem contrasting to ours, as SPTAN1 was downregulated after puberty in cattle. Species differences or differences in tissue sample, as well as the definition for puberty (and therefore timing of sampling), could explain these contrasting results. Nonetheless, these experiments in mice and cattle suggest that SPTAN1 is a candidate protein for further investigation in the context of pubertal development.

Perhaps the Phospholipase C beta 1 (PLCB1) was the DA protein and DE gene with more evidence of its role in puberty. PLCB1 plays a crucial role in reproductive physiology. Genome-wide association studies in pigs reported the contribution of PLCB1 to growth at puberty onset and in the gonadotropin signalling pathway (Meng et al., 2017). Further, PLCB1 disruption resulted in infertile mice with pleiotropic reproductive defects (Filis et al., 2013). PLCB1 is a key factor modulating GPCR signalling in the control of reproductive physiology in mice (Jiang et al., 1997; Kim et al., 1997; Xie et al., 1999; Bohm et al., 2002; Ballester et al., 2004; Filis et al., 2013). Our results serve as evidence for the role that PLCB1 may have in puberty onset in Bos indicus cattle: it interacts in the predicted network with other genes related to GnRH signalling and other hormonal and neuronal signalling pathways known to be involved in the feedback mechanism driving the activation of hypothalamus – pituitary-gonadal axis (HPG).

The activation of HPG culminates with the pulsatile release of GnRH from the hypothalamus. GnRH triggers the synthesis and secretion of luteinizing hormone (LH) and follicular stimulating hormone (FSH) in the pituitary gland, which stimulates the production of gonadal hormones leading to ovulation (Bliss et al., 2010). Both positive and negative feedback at several levels regulate HPG. One of positive feedback on the activation of GnRH secretion is commenced by GnRH receptor (GnRHR) - G-protein αq subunit (Gαq/11) which induces increased intracellular calcium (Khadra &Li, 2006). In agreement with this positive feedback, our proteomic study in the hypothalamus revealed the up-regulation of calcium signalling pathway (FDR = 0.05) and GnRH signalling pathway (FDR = 0.02). G protein subunit alpha q (GNAQ), phospholipase C beta 1 (PLCB1), calcium/calmodulin-dependent protein kinase II alpha (CAMK2A) and calcium/calmodulin-dependent protein kinase II delta (CAMK2D) showed increased abundance at post-puberty in Brahman heifers and are associated with the up-regulated pathways.

108

Oxytocin signalling was an enriched pathway in our results and so we have recapitulated some of the known biology of puberty. The stimulatory effect of oxytocin on GnRH release, which then modulates the initiation of puberty, was reported in female rats (Parent et al., 2008). Six upregulated DA proteins at post-puberty are involved in oxytocin signalling: CAMK2A, GNAQ, CAMK2D, PLCB1, EEF2 (eukaryotic translation elongation factor 2) and PRKCG (protein kinase C gamma). Three downregulated DA proteins such as G protein subunit alpha i2 (GNAI2), G protein subunit alpha o1 (GNAO1) and KRAS proto- oncogene, GTPase (KRAS) also contribute to oxytocin signalling. The oxytocin receptor is known as a typical G protein couple receptor that activates PLCβ and then leads to intracellular Ca2+ mobilization (Gimpl &Fahrenholz, 2001; Gravati et al., 2010). Intracellular calcium mobilization is relevant in many fundamental cellular activities like gene regulation, apoptosis and cell proliferation (Berridge et al., 2000). In hypothalamic neurons, the pulsatile release of GnRH is depended on voltage-gated calcium influx (Krsmanović et al., 1992). Therefore, it is logical to propose that these DA proteins involved in oxytocin and calcium signalling can contribute to the increase in GnRH release that marks the onset of puberty.

GnRH release is regulated by both excitatory and inhibitory inputs. Apart from excitatory inputs discussed above, the γ -aminobutyric acid (GABA) is known as an inhibitory neurotransmitter related to GnRH secretion in the hypothalamus (Mitsushima et al., 1994; Mitsushima &Kimura, 1997; Han et al., 2004). A study in female monkeys noted that the decrease in tonic GABA inhibition resulted in an increase of glutamate and other excitatory neurotransmitters contributing to the increase of GnRH release (Terasawa &Fernandez, 2001; Ojeda et al., 2010a). However, an indirect excitatory effect of GABA on GnRH neurons has also been reported (DeFazio et al., 2002). In the current study, DA proteins were enriched for GABAergic synapse pathway. Three up-regulated DA proteins (GAD2, GLUL, PRKCG) and 3 downregulated DA proteins (GNAO1, GNAI2, GNB2) were identified in our analyses. These results indicate that specific genes contribute to the regulation of GABAergic input controlling puberty in Brahman heifers.

It is well documented that gonadal steroid hormones play a critical role in feedback regulation of GnRH secretion in the hypothalamus (Glidewell-Kenney et al., 2007; Rance, 2009). In female animals, the feedback of gonadal hormones in the HPG axis are species- specific and different at the various stages of the reproductive cycle. The down-regulation of estrogen to GnRH expression in the hypothalamus occurs during the follicular phase (Zoeller et al., 1988; Petersen et al., 1995; Petersen et al., 2003). The increase of estrogen levels during

109

middle to late of follicular phase is capable of up-regulating the kisspeptin expression which then leads to the significant secretion of GnRH (Evans et al., 1997; Irwig et al., 2004; Herbison, 2008; Clarkson &Herbison, 2009). In addition to estrogen, the complex feedback of progesterone is exerted at all levels of the HPG axis. Progesterone shows negative feedback upon the release of hypothalamic GnRH during the luteal phase and limits the positive effect of estrogen (Harris et al., 1999; Richter et al., 2002; Sleiter et al., 2009). This is the classic role of progesterone during the reproductive cycle. Another role for progesterone has been proposed to explain the peri-pubertal transition period. The effect of progesterone on hypothalamic GnRH secretion is also positive (Everett, 1948; Krey et al., 1973; Rao &Mahesh, 1986; Caraty &Skinner, 1999; Micevych et al., 2003). A study in ewes reported that the enhancement in progesterone levels during the luteal phase is essential for positive feedback of estradiol in stimulating the pre-ovulatory GnRH surge (Caraty &Skinner, 1999). Note that, post-pubertal heifers used in the current study had a higher concentration of progesterone in the blood (P-value < 0.05) and pre-pubertal heifers had never ovulated. Given this contrast, progesterone levels may be the cause of differential abundance for proteins involved in estrogen signalling pathways. The GABAergic and glutamatergic neurons were suggested as target of estradiol (Petersen et al., 2003). As a result, the feedback effect mediated by progesterone and estrogen contribute to the multi-levels system controlling GnRH synthesis and release.

Peroxisome proliferator-activated receptors (PPARs) are ligand-activated nuclear receptors that control many important processes such as energy homeostasis, lipid and glucose metabolism, and immune response (Yang et al., 2008; Wahli &Michalik, 2012). It is well-documented that glucose, fatty acid and fuel sensors are critical factors regulating female reproduction. The effects of PPARs on female reproduction has been reported (Komar, 2005; Froment et al., 2006; Chen et al., 2010; Kowalewski et al., 2011). In the pituitary, peroxisome proliferators-activated receptor gamma (PPARG) was able to repress LH in female mice (Sharma et al., 2011). However, there is a limited amount of work on the effect of PPARs on the hypothalamus of cattle. The current study identified the PPAR signalling pathway as an enriched pathway for the DA proteins. At post-puberty, three lower abundant proteins (DBI, ILK, FABP7) and one higher abundant protein (APOA2) were significantly DA and annotated to the PPAR pathway. Even though these proteins are part of the PPAR pathway, their role in puberty may be discussed in light of their other functional

110

roles (see below). It is possible that these four DA proteins are other links between reproduction and energy metabolism.

The diazepam binding inhibitor (DBI) is able to down-regulate the signal transduction at the gamma-aminobutyric acid, type A receptor (GABAA) in synapses (Gray et al., 1986). Further, DBI was suggested as a suppressor of the effects of estrogen levels in female mice, probably through GABA signalling in the hypothalamus (Dong et al., 2002). In addition, a genome-wide association study discovered a copy number variation that contributed to age at menarche (puberty in humans) mapped to DBI (Liu et al., 2012). In our results, DBI was down-regulated post-puberty, meaning that higher levels of DBI may act to inhibit estrogen effects and serve to delay the maturation of the HPG in pre-pubertal animals.

Another PPAR signalling protein that was downregulated at post-puberty was integrin-linked kinase (ILK) protein. ILK was highly expressed in the placenta during gestation as well as in trophoblasts during the first trimester of pregnancy (Elustondo et al., 2006; Yen et al., 2014). PAT-4, a homolog of ILK was important for ovulation (Xu et al., 2006). The role for ILK in cattle puberty remains unclear, but it is possible that it plays a role in ovulation and it responds negatively to progesterone signalling.

Brain fatty acid binding protein (FABP7) was downregulated in our study, it is part of PPAR signalling and it is important for brain development as well. FABP7 plays an important role in early nervous system development (Bennett et al., 1994; Feng et al., 1994; Kurtz et al., 1994). A recent study in RNA profiling of mouse mammary epithelial cells revealed the differential expression of FABP7 between pre- and pubertal stage (Pal et al., 2017). The emerging hypothesis is that FABP7 maybe another signal integrating energy metabolism with brain development/maturation during puberty in cattle. Its abundance was higher before puberty when it could contribute to brain development and hypothalamic maturation. After puberty, the abundance was lower, which may implicate that FABP7 can be modulated by progesterone signalling. These hypothesis merit further research.

The apolipoprotein A2 (APOA2) is involved in lipid metabolism and body fat distribution (Zaki et al., 2013). APOA2 was upregulated after puberty and is part of the PPAR pathway. Body fat is known to influence puberty: lower body fat delays puberty, higher body fat causes earlier pubertal development (Shalitin &Kiess, 2017). APOA2 maybe another permissive signal, similar to Leptin, for the puberty process. In summary, the current study reported helped to develop a new hypothesis for the roles that proteins in the PPAR pathway

111

may play in cattle puberty. Future work should target DBI, ILK, FABP7, and APOA2 to elucidate how these proteins link metabolism to puberty in Bos indicus cattle.

The interaction network revealed three DA proteins as highly connected hubs: HSP90AA1, MDH2 and ALB. The protein HSP90AA1 was involved in estrogen signalling and it interacted with other proteins of the PPAR and oxytocin pathways (ILK, EEF2 and PRKCG). Recently, a transcriptomic study in pre- versus post-pubertal mammary epithelial cells noted HSP90AA1 as DE gene (Pal et al., 2017). Even though there was little evidence of a relationship between MDH2 and female reproduction, MDH2 was characterised as a protein marker for male fertility (Kwon et al., 2015a; Kwon et al., 2015b). A study in Egyptian boys suggested ALB status involved in each stage of puberty (Cole et al., 1982). The increase of albumin excretion rate in urinary tract was significantly associated with puberty in non-diabetic children and adolescents (Bangstad et al., 1993). Bovine serum albumin – estrogen compounds impacted on GnRH1 neuronal activity (Temple &Wray, 2005). The inferred network predicted interactions between ALB and proteins involved in estrogen (HSPA8) and oxytocin (EEF2) signalling. In summary, previous studies demonstrated the potential link of these hubs proteins and reproductive physiology. Still, it is difficult to speculate their specific roles in cattle puberty beyond suggesting that these central hubs modulate the enriched pathways and key candidate proteins discussed above. These three proteins could be proposed as the new “maestros” in puberty development in Bos indicus cattle, an idea that requires further validation studies.

2.1.5. Conclusion

In this study, DA proteins related to puberty were mainly associated with metabolic pathways, energy metabolism, and nervous system development. In agreement with previous findings of complex feedback effects on GnRH release, the KEGG pathway analysis indicated that these DA proteins are involved in pathways that convey both inhibitory and excitatory inputs for hypothalamic neurons. It is hypothesised that the increase of GnRH release at the hypothalamus of Brahman heifers may be initiated via oxytocin signalling, via PLCB1. The current study opens the possibility to test the DA proteins as candidate regulators for the onset of puberty, especially the three network hubs.

112

Chapter 2.2. Pituitary proteomics revealed differentially abundant proteins and pathways related to puberty in Brahman heifers

113

Abstract

The onset of puberty is controlled by regulatory gene networks formed by multiple functional modules. This study presents the first quantitative mass spectrometry analyses that describe the changes in protein levels in the pituitary gland during pubertal transition in Brahman heifers. It describes potential regulators and pathways driving puberty. There were 186 proteins that were differentially abundant (DA) between pre- and post-pubertal heifers. These data are available via ProteomeXchange with identifier PXD009619. When compared, proteomics and transcriptomic findings for the same tissue samples, 5 DA proteins were identified as differentially expressed genes: IGFBP2, ENO1, CHGB, TAGLN and HIST2H2AC. Except for HIST2H2AC, the other 4 had the same mRNA-protein expression pattern. According to functional enrichment analyses, the DA proteins discovered were annotated to gene ontology terms that are known as relevant to puberty processes. The list of DA proteins was significantly enriched for pathways such as ribosome, complement and coagulation cascades, glycolysis and gluconeogenesis, biosynthesis of amino acids and focal adhesion. The ribosomal proteins may play an important role in regulating heifer puberty in Bos indicus. This study revealed candidate proteins and highlighted some of the molecular mechanisms potentially regulating pituitary function in relation to puberty and the estrous cycle, as post-puberty heifers were euthanized in the luteal phase of the cycle. Differentially abundant proteins may inform about candidate genes and genetic variates related to protein expression affecting heifer puberty in Bos indicus breeds.

Keywords: puberty, Brahman, ribosomal proteins, IGFBP2, ENO1, LC-MS/MS

2.2.1. Introduction

It is well-documented that the puberty onset is initiated by a decrease in estradiol negative feedback on the release of gonadotropin-releasing hormone (GnRH). The increase of GnRH secretion from hypothalamus then stimulates the release of the gonadotrophins; luteinizing hormone (LH) and follicle stimulating hormone (FSH) from the pituitary gland (Rawlings et al., 2003; Romano et al., 2007; Johnston et al., 2009). These hormones then lead to the first ovulation and continued an ovarian cyclic activity (Day et al., 1984; Day et al., 1987; Day, 1998; Bliss et al., 2010). The gonadostat theory suggested that negative feedback of estradiol on gonadotropins secretion by the pituitary gland decreased as puberty approaches in heifers (Schillo et al., 1982; Day et al., 1984; Day et al., 1987). Although, this mechanism is

114

similar between Bos indicus and Bos taurus, the onset of puberty in Bos indicus still occurs later in life. Understanding pituitary mRNA and protein profiles at puberty in Bos indicus could shed light on the underlying mechanisms that delay puberty in this sub-species.

In Part I of this PhD thesis, the pituitary transcriptome (Chapter 1.2) provided a list of candidate genes and suggested potential puberty-related pathways. The proteomic study reported in this chapter offers a new approach to investigate final gene products, protein expression and regulation that cannot be assumed from the transcriptomic studies. In short, the relationship between mRNA expression and protein abundance is not straightforward and is far from a perfect correlation. Although the pituitary gland secretes pubertal-related hormones (LH and FSH) and contributes to the control of puberty processes, there is no previous study of the pituitary proteome at puberty in Bos indicus heifers. A database of a Bos indicus pituitary proteome is not yet available. Therefore, proteomic studies of pre- and post-pubertal Brahman heifers might provide new information about hormonal secretion and protein abundance at the pituitary level.

This study aimed at quantifying proteomic profiles in both pre- and post-pubertal heifers. Profiles pre- and post-puberty were compared to ascertain differentially abundant (DA) proteins. Protein profiles were obtained using liquid chromatography-electrospray ionization- tandem mass spectrometry (LC-ESI-MS/MS). Differentially abundant proteins between pre- post-pubertal heifers in the luteal phase were used to create a target list for enrichment analyses. Enrichment analyses were performed to identify biological processes and pathways. Knowledge of protein patterns and biological processes can provide significant insights into the molecular mechanisms of controlling puberty in cattle.

2.2.2. Materials and methods

Ethics statement

The Animal Ethics Committee of the University of Queensland, Production and Companion Animal group approved this experiment (certificate number QAAFI/279/12). Animals Twelve Brahman heifers used in this current work were the same as those utilized in transcriptomic studies (Chapter 1.1 to 1.4). Animals that were born during the wet season 2011/2012 and purchased from commercial Queensland herds. All heifers had typical phenotypic characteristics of Bos indcicus cattle. The Gatton Campus facilities of the University of Queensland housed the animals during the experiment.

115

Observation of luteal development was performed using ovarian ultrasound scans. When a corpus luteum was observed, the heifer was considered as post-pubertal heifer and assigned to euthanasia after 15 days of the first corpus luteum observation. In the same slaughter day, another heifer that never ovulated was randomly chosen from the group and paired with post-pubertal heifer. In addition, the circulating concentration of progesterone in pre- and post-pubertal heifers were also measured to confirm the pubertal status. Higher concentration of plasma progesterone in the post-pubertal heifers was observed. The pituitary glands (including both anterior and posterior pituitary) were harvested after euthanizing.

Protein Sample preparation

Protein extraction and digestion of heifer pituitary tissues were performed as previously described (Tan et al., 2014). Briefly, pituitary fraction of each sample was homogenized in a lysis buffer (7M urea, 2M thiourea, 4%SDS, 10mM DTT and 1mM PMSF). The samples were sonicated for 10 seconds and vortexed vigorously at 30°C for 1 hour. The mixtures were then added 25mM acrylamide and incubated at 30°C for 1 hour. Excess acrylamide was quenched by adding 5mM DTT. The supernatants were precipitated at -20°C overnight with four volume of methanol: acetone (1:1). After discarding methanol: acetone, and air-drying, the resulting pellets were dissolved in 50 mM ammonium acetate buffer. Protein concentrations were determined using Nanodrop (Thermo Scientific). Next, approximately 100 g of protein was loaded into a 10 kDa Amicon Ultra 0.5 centrifugal filter device (Merck Millipore) and centrifuged at maximum speed for 30 minutes. Protein was then diluted in 50mM ammonium bicarbonate, followed by another centrifugation at maximum speed for 30 minutes. Trypsin was then utilized to digest the protein. Peptides were desalted by C-18 Zip-tip (adapted from Millipore procedure) and stored at -20°C until analysis.

Liquid chromatography electrospray ionization tandem mass spectrometry (LC-ESI-MS/MS) The LC-ESI-MS/MS were performed on a Prominence nanoLC system (Shimadzu) connected to a TripleToF 5600 mass spectrometer, equipped with a Nanospray III interface (SCIEX) as previously described (Xu et al., 2015). A 70 min LC gradient was used to separate peptides. MS-TOF scan was conducted, which followed by information-dependent acquisition (IDA) of top peptides for 2 randomly chosen pre-pubertal samples and 2 randomly chosen post- pubertal samples. Next, sequential window acquisition of all theoretical mass spectra (SWATH) was performed on all samples using prepared IDA library for each group. In order to identify a protein, data analysis was performed using Protein Pilot v5.0.1 (ABSCIEX). Briefly, the Bovine protein database was retrieved from UniProt

116

(www.uniprot.org; 43,813 entries assigned to Bos taurus) and used for peptide mapping. We used the settings as previously described (Nguyen et al., 2018). Identified peptides with more than 99% of confidence and FDR < 1% were utilized for subsequent analyses. The IDA library prepared was then used for SWATH analyses using Peakview v2.1 (SCIEX) with the parameters as previously described (Nguyen et al., 2018). The differentially abundant proteins were identified using MSstats (v2.6) in R software as previously described (Choi et al., 2014; Zacchi &Schulz, 2016). The P-value was adjusted for multiple test correction, with P-value cutoff of 0.01. Correlation analysis The Pearson’s correlation coefficient (r) was used to compute the expression levels of mRNAs and proteins in the pituitary gland from pre- and post-pubertal Brahman heifers. The cor() function in R was used in this calculation. The unmapped mRNA-proteins were not included in this analysis.

Protein-protein interaction network and Enrichment analyses

The STRINGv10 web-server (http://string-db.org) was used to retrieve the gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) annotations for differentially abundant proteins as well as to construct the protein to protein interaction network (Szklarczyk et al., 2014). We applied the Benjamini-Hochberg method for multiple test correction. Significant enrichment is reported when the false discovery rate (FDR) was lower than 0.05 for GO terms and KEGG pathways. The interaction network was visualized in the Cytoscape (Shannon et al., 2003). We removed from the network nodes without connections and interaction score < 0.6 to focus on the proteins that interact with other proteins in the data.

2.2.3. Results

A total of 969 proteins (FDR < 0.01) were identified using Protein Pilot v5.0.1 (ABSCIEX) and 715 of those were quantified (FDR < 0.01) using SWATH analysis. Among quantified proteins, 96 and 90 proteins were significantly decreased and increased in abundance post-puberty, respectively (adjusted P-value < 0.01). We identified 11 highly DA proteins with fold change (FC) 1 or higher in absolute terms and P-value < 0.001 (Table 2.2.1). The mass spectrometry pituitary proteomic data have been submitted into Proteome Xchange Consortium via PRIDE partner repository with identifier PXD009619 (Vizcaino et al., 2016).

117

Table 2.2.1 Highly significant differentially abundant proteins (|Fold change| > 1 and P-value < 0.001) in comparison between pre- and post-pubertal Brahman heifers in the pituitary gland.

Ensembl ID UniProt ID Gene symbol Fold change Adjusted P- value ENSBTAG00000014713 Q0P5D6 RARRES1 -1.845 1.24E-04 ENSBTAG00000009665 F1MRT9 UTRN -1.255 1.69E-04 ENSBTAG00000027446 Q5E9C0 RSU1 -1.180 3.87E-08 Q58DP8 NOLA1 -1.140 3.87E-08 ENSBTAG00000013419 Q3SX02 HTATIP2 -1.079 5.58E-05 ENSBTAG00000017460 A1A4Q2 PRORSD1 -1.034 9.79E-04 ENSBTAG00000003072 P48818 ACADVL -1.019 3.02E-04 ENSBTAG00000006982 Q58DS9 RAB5C 1.310 1.52E-06 ENSBTAG00000008027 P01180 AVP 1.359 0.00E+00 ENSBTAG00000030164 Q32PB9 RPL38 1.448 6.62E-13 ENSBTAG00000031160 F1MLW7 LOC100297192 2.668 2.48E-11

When compared the DA proteins in the pituitary gland with its corresponding mRNA in the same tissue and animals (Chapter 1.2), five were in common including IGFBP2, TAGLN, CHGB, ENO1 and HIST2H2AC. Among these five, all but 1 had the same mRNA and protein expression patterns, the exception being HIST2H2AC. In fact, five proteins were decreased in abundance post-puberty, while mRNA expression of HIST2H2AC was increased post-puberty. Table 2.2.2 lists these common mRNA and proteins and their expression levels post-puberty. Overall, the correlation between protein and mRNA expression was minimal (r = 0.006; P-value = 0.8). This lack of correlation is consistent with only 5 DA proteins which were also DE genes.

Gene ontology analysis of DA proteins was conducted using String v10 web-server (Szklarczyk et al., 2014). GO analysis categorized proteins in three groups: biological process (BP), molecular function (MF) and cellular component (CC). In BP, there were 214 enriched terms identified (FDR < 0.05). Among these enriched BP terms, the majority of DA proteins classified in single-organism cellular process (83 DA proteins), cellular process (81 DA proteins), metabolic process (75 DA proteins) and biological regulation (65 DA proteins). In MF, 39 enriched terms were observed (FDR < 0.05). A large number of proteins with catalytic activity and binding activity were identified. Hydrolases, oxidoreductases and endopeptidase were the main protein classes revealed in this work. Additionally, these proteins were mainly

118

found in organelle (83 DA proteins), membrane-bounded organelle (81 DA proteins), cytoplasm (81 DA proteins), intracellular (74 DA proteins) and extracellular region part (74 DA proteins). These enriched go terms for BP, MF and CC are all broad categories, broad GO terms.

Table 2.2.2 List of common differentially expressed genes and abundant proteins in transcriptome and proteome approaches in the pituitary gland (P-value < 0.05)

Ensembl ID UniProt ID Gene symbol FC_mRNA FC_protein ENSBTAG00000005596 F1N2P8 IGFBP2 -0.615 -0.452 ENSBTAG00000007196 V6F957 TAGLN -0.405 -0.698 ENSBTAG00000011782 P23389 CHGB -0.266 -0.653 ENSBTAG00000013411 F1MB08 ENO1 -0.126 -0.170 ENSBTAG00000032456 F2Z4I6 HIST2H2AC 0.680 -0.338

FC means fold change

To analyse gene function, we also investigated enrichment according to KEGG pathway annotations. A total of 186 DA proteins were assigned to 12 KEGG pathways (FDR < 0.05). Among these enriched pathways, the ribosome pathway represented the largest number of abundant proteins (FDR = 0.0002). Complement and coagulation cascades, glycolysis and gluconeogenesis, biosynthesis of amino acids, focal adhesion were also overrepresented (Figure 2.2.1).

Figure 2.2.1 KEGG pathway analysis of differentially abundant proteins in the pituitary gland of pubertal Brahman heifers.

In order to extend knowledge about the interaction among DA proteins, the protein- protein interaction network was constructed using Stringv10 database and visualized using

119

Cytoscape (Shannon et al., 2003; Szklarczyk et al., 2014). The interaction network comprised 113 nodes with 254 edges. Among these nodes, ribosome-related proteins seem to form a dominant cluster in the network. The protein RPS5 had 17 connections within the network. A common DE gene DA and protein, ENO1, was found in the network with 10 connections. This protein is involved in carbon metabolism, biosynthesis of amino acids and glycolysis/gluconeogenesis identifying by KEGG pathway analysis of DA proteins in the pituitary between pre- and post-pubertal Brahman heifers.

Figure 2.2.2 Protein-protein interaction network of differentially abundant proteins between pre- and post-pubertal heifers in the pituitary gland. Each protein is a node and protein-protein interactions are illustrated as edges that link nodes. The size of the node represents the increased or decreased in protein abundance post-puberty. Node color represents the connectivity of a protein within the network; it is a range from green to red for low to high connectivity. For example, the node for ALB is small and dark red because this protein is lower in abundance post-puberty and highly connected to other proteins in the network.

2.2.4. Discussion

The pituitary transcriptomic analysis identified 284 DE genes, of which 165 were up- regulated and 119 were down-regulated post-puberty in Brahman heifers (Chapter 1.2). Proteomics revealed 186 DA proteins, of which 96 increased in abundance and 90 were decreased in abundance post-puberty. Similar to previous integrated transcriptome and proteome datasets in the hypothalamus (Chapter 2.1), these datasets in the pituitary gland were also divergent. The mRNA-protein correlation was very low (r = 0.006; P-value = 0.8).

120

The findings indicated that only five of 186 DA proteins were also DE genes. Other studies observed this weak correlation between transcriptomic and proteomic data (Anderson &Seilhamer, 1997; Gygi et al., 1999; Chen et al., 2002; Griffin et al., 2002; Ghaemmaghami et al., 2003; Greenbaum et al., 2003; Washburn et al., 2003; Tian et al., 2004; Nie et al., 2006; Nie et al., 2007; Jayapal et al., 2008; de Sousa Abreu et al., 2009; Haider &Pal, 2013; Bai et al., 2015; Carvalhais et al., 2015; Payne, 2015; Edfors et al., 2016). These weak correlations probably result from the difference between two methodologies. Biological factors, such as translational regulation and protein in vivo half-lives, were another aspect that could explain the seldom overlap between proteomic and transcriptomic data (Maier et al., 2009; Vogel &Marcotte, 2012; Haider &Pal, 2013). These differences between transcriptomic and proteomic data emphasize the importance of collecting data at the two different levels in order to elucidate the molecular mechanisms related to the physiological phenomenon.

Five differentially expressed genes in the transcriptome: IGFBP2, TAGLN, CHGB, ENO1 and HIST2H2AC were significantly less abundant post-puberty in the current proteomic analyses. The function of these genes should be further investigated in the context of puberty. Their known roles related to puberty and the estrus cycle are discussed below. Collectively, these common DE genes and DA proteins may have a pivotal role in regulating reproductive function.

IGFBP2 was a predominant IGFBPs synthesized and secreted in the anterior pituitary during pre-ovulatory and early luteal phase in beef cattle (Funston et al., 1995; Roberts et al., 2001). In the anterior pituitary, the expression of IGFBP2 fluctuated with changes in the stage of the estrous cycle in association with serum progesterone (Funston et al., 1995). In addition, estrogen signalling increased IGFBP2 expression in the anterior pituitary of ewes, cattle, pigs and rats (Michels et al., 1993; Clapper et al., 1998; Roberts et al., 2001; Rempel &Clapper, 2002). Roberts et al. (2001) reported that IGFBPs level in the anterior pituitary decreased from the pre-ovulatory to the early luteal development in beef cattle. Consistent with Robert et al. (2001) observations, this current work identified the decrease in mRNA expression (Chapter 1.2) and lower abundance of IGFBP2 (current chapter) in the pituitary gland at a luteal phase in bovine. The release of IGFBP2 in the pituitary gland was stimulated by gonadotropin- releasing hormone as demonstrated by Robert et al. (2001). Also, polymorphisms in IGFBP2 were associated with age at puberty in Brahman cattle (Fortes et al., 2013b).

Chromogranin B (CHGB) belongs to a chromogranin protein family which has been noted to affect the secretion and storage of FSH and LH at different periods of estrous cycle in

121

sheep (Crawford &McNeilly, 2002). Study of granin-gonadotropin interactions in sheep observed a decrease of CHGB mRNA level after the preovulatory LH surge (Crawford &McNeilly, 2002). In agreement with sheep’ study, CHGB mRNA expression decreased at post-puberty in the pituitary transcriptome of Brahman heifers (Chapter 1.2; Table 2.2.2). Additionally, the neuropeptide CHGB was selected as one of fertility-related neuropeptides in the pituitary gland neuropeptidome in pre- and post-pubertal Brangus heifer (DeAtley et al., 2018). Neuropeptidome study of Brangus heifers also reported the higher levels of CHGB neuropeptide products in the pre-pubertal pituitary gland (DeAtley et al., 2018). Using mass spectrometry method, the proteome of the pituitary gland in Brahman heifers (current Chapter) observed a lower abundance of CHGB protein after puberty. These results provided further support of the importance of CHGB in the pituitary glands before and after puberty in cattle.

The DA protein -enolase (ENO1) is known as one of 3 isoforms of glycolytic enzyme that are responsible for catalyzing the conversion of 2-phosphoglycerate to phosphoenolpyruvate (Pancholi, 2001). A study in the ovaries of pre-laying and laying geese suggested that ENO1 may regulate reproduction function in female geese (Ji et al., 2014; Kang et al., 2014). The overexpression of ENO1 in granular cell induced the mRNA expression of FSH and LH receptors (Ji et al., 2014). Activation of these receptors is necessary for the hormonal functioning of pituitary glycoprotein hormones (LH and FSH). Pituitary FSH and LH concentrations were observed to be decreased at puberty in heifers (Desjardins &Hafs, 1968). This Thesis investigated the pituitary gland transcriptome and proteome between pre- and post-pubertal heifers and found lower expression level of ENO1 mRNA and protein post- puberty (Chapter 1.2 and current Chapter). On the contrary, the study of mouse mammary glands using single-cell RNA profiling reported ENO1 as a down-regulated gene in a pre- puberty gene network (Pal et al., 2017). It seems that ENO1 appear to impact female puberty, but its precise function needs to be considered in terms of tissue and species-specific.

The DA protein coded by the DE gene HIST2H2AC, a member of histone 2A family, was mapped to a genetic locus reported to link puberty timing and pubertal height growth in human (Cousminer et al., 2013). In beef cattle, polymorphisms located between 20 and 30 Mb on chromosome 14 were associated with fat deposition, height, weight, IGF1 hormonal levels, and age at puberty in beef cattle (Snelling et al., 2010; Fortes et al., 2013a). A genetic correlation between age at puberty and high height in Brahman heifers was also reported (Vargas et al., 1998). Differential expression of HIST2H2AC at mRNA and proteins level between pre- and post-pubertal heifers proposed its potential regulatory role at puberty.

122

Therefore, this candidate should be mined for mutations that could be tested for their effects on height and puberty in cattle.

Differentially abundant proteins in the hypothalamus of pubertal goats and rats suggested 11 GO terms involved in the onset of puberty (Gao et al., 2018). The study presented herein also identified 6 of these 11 GO terms, including regulation of multicellular organismal process (FDR = 0.002), extracellular region part (FDR = 4.26 x10-42), extracellular matrix (FDR = 1.49 x 10-07), proteinaceous extracellular matrix (FDR = 4.23 x 10-05), calcium ion binding (FDR = 0.01) and extracellular matrix structural constituent (FDR = 0.01). Our findings and the referenced literature suggest the involvement of these GO terms in the regulation of puberty in goats, rats and heifers.

Among 12 enriched pathways, protein digestion and absorption, focal adhesion, and ECM-receptor interaction were also identified as enriched pathways in the hypothalamic tissues in pubertal goats and rats (Gao et al., 2018). Proteins in these pathways had increased abundancy post-puberty in Brahman heifers. These DA proteins clustered together in the interaction network. Mutations in these proteins were associated with muscle disorders and bone density in early puberty (Suuriniemi et al., 2006; Bönnemann, 2011; Kostik et al., 2013). Additionally, a DA protein: AGRN gene was utilized in a comparative genomic hybridization array to reveal differentially methylated regions between high and low-gain body weight in early pubertal heifers (9 months of age) (Alves et al., 2017). Briefly, a study by Alves et al. (2017) started with a hypothesis that high rates of BW gain in heifers during juvenile period affect methylation patterns of genes obtained from the hypothalamus arcuate nucleus region and these alterations may affect age at puberty in heifers. Alves et al. (2018) study then confirmed that DNA methylation was an additional mechanism in which the expression of genes in the hypothalamic regions can regulate growth and pubertal development.

Gao et al. (2018) performed KEGG analysis of long non-coding RNA targets in the hypothalamus of pubertal rats and reported significant enrichment for the ribosome pathway. In our current study, the ribosome pathway was also enriched and the proteins in this pathway were less abundant post-puberty. Further, these ribosomal proteins formed a dominant cluster in the pituitary interaction network reported herein. A study in pubertal Brangus heifers observed the highest expression of the ribosomal protein L39 gene in the pituitary gland (Cánovas et al., 2014a). Further, DA proteins in the interaction network, such as RPS5, RPS3 and RPLP1, were also listed in the networks constructed from endometrial gene expression of low and high fertility heifers (Killeen et al., 2014). Recently, miR-503-3p was proposed as a

123

new repressor of the initiation of puberty in female mice (Tong et al., 2018). The stable overexpression of miR-503-3p in the GT1-7 cell line influenced ribosome biogenesis pathways and resulted in down-regulation of puberty-related genes (Tong et al., 2018). The ribosomal protein RPL22 inhibited the expression of Lin28B, a puberty-related gene (Ong et al., 2009; Rao et al., 2012). The PI3K and MAPK signalling pathways that are involved in puberty were also regulated by a ribosomal protein: RPS7 (Bliss et al., 2010; Acosta-Martínez, 2011; Wang et al., 2013; Nelson et al., 2017; Pal et al., 2017). From the interaction network, there were 7 ribosomal proteins that interacted with UBB protein. Disruption in UBB gene in both male and female mice resulted in infertile animals (Ryu et al., 2008). These collective evidences suggested proteins in the ribosomal pathway are relevant for pubertal development in mammals. Further work could target the DA ribosomal proteins and their targets in the pituitary gland to investigate the molecular mechanisms of puberty in Bos indicus.

Another cluster in the interaction network comprised proteins involved in carbon metabolism, biosynthesis of amino acids and the glycolysis/gluconeogenesis pathway. The protein with most of the connections in this cluster was ENO1, a DA protein and DE gene, as described above. The onset of puberty requires communication between trans-synaptic and glial systems, involving excitatory amino acids and growth factors (Ojeda et al., 2006). Therefore, DA proteins involved in the biosynthesis of amino acids, such as ENO1, ENO2, PHGDH, TPI1 and ALDOC, may contribute to the regulation of puberty in Bos indicus. In addition, glycolysis converts glucose into pyruvate in order to generate ATP (energy) and NADH. Therefore, DA proteins in the pituitary gland may involve in the neuroendocrine and energy production that regulate and provide energy for the pubertal process.

2.2.5. Conclusion

The present study provides a first pituitary proteomic dataset of Bos indicus at luteal phase. The match transcript and protein changes at puberty development serve to identify candidate genes required for the onset of puberty (e.g., IGFBP2, ENO1, CHGB). These proteins are involved in biological processes and pathways associated with female puberty. The abundance of proteins involved in protein digestion and absorption, focal adhesion, and ECM-receptor interaction were significantly increased post-puberty. The decreased abundance of ribosomal proteins post-puberty can be interpreted in the context of these genes’ inhibitory input to puberty-related genes. Proteins related to energy production and amino acid biosynthesis may have a crucial role in the neuroendocrine regulation of the pubertal process, meriting further investigation.

124

Chapter 2.3. Liver proteomics to study the onset of puberty in Brahman heifers

Manuscript published in Proceedings of the World Congress on Genetics Applied to Livestock Production, 11.256 http://www.wcgalp.org/proceedings/2018/liver-proteomics-study-onset-puberty-brahman- heifers

125

Abstract

The liver plays a prominent role in metabolism in mammals, and considerable evidence supports a link between liver function and female fertility in cattle. Hence, it would be helpful to the study of puberty to investigate protein abundance patterns in the liver of Brahman heifers (a Bos indicus breed). We identified 777 proteins in liver and 275 of these were differentially abundant between pre- and post-pubertal heifers. These data are available via ProteomeXchange with identifier PXD009419. Eight differentially abundant proteins were identified as differentially expressed genes in the liver transcriptome. The correlation coefficient (r) between the identified transcript-protein pairs was equal 0.13 (P-value = 0.001). The functional analysis of the differentially abundant proteins showed enriched pathways that might regulate puberty onset, such as steroid hormone biosynthesis, PPAR signalling pathway, lipid metabolism, amino acid metabolisms and oxidative phosphorylation (FDR < 0.05). In conclusion, the proteomics approach was effective in identifying differentially abundant proteins and furthering our understanding of the biological link between liver function and puberty onset.

2.3.1. Introduction

It has been established that reproduction and growth are intensive processes requiring a high level of energy input. The state of body energy reserve is known as a key determinant of the puberty onset in mammals (Parent et al., 2003). In Bos indicus cattle, the proportion of mature body weight at puberty is highly conserved, accounting for 60% of the mature weight (Freetly et al., 2011). To manage heifers for proper growth as well as normal pubertal development, cattle are required to feed on an adequate diet (complete with vitamins and essential nutrients) from birth to puberty. Several studies reported that beef heifers maintained on a low-energy diet that limited weight gain reached puberty later than those fed a high-energy diet for greater weight gain (Rathbone et al., 2001; Nogueira, 2004). Furthermore, the offspring of heifers fed a low-energy diet approached puberty 19 days later than those maintained on a high-energy diet (Corah et al., 1975). Insufficient nutrition considerably delays puberty in both Bos indicus and Bos taurus cattle (Abeygunawardena &Dematawewa, 2004) (Abeygunawardena &Dematawewa, 2004).

The liver, an important metabolic organ, regulates body energy metabolism and contributes metabolic and hormonal signals to the reproductive system (Rui, 2014). Insulin- like growth factor 1 (IGF1), secreted mainly by the liver, has a direct action on the anterior

126

pituitary and gonads, suggesting its role in reproductive physiology (Daftary &Gore, 2005). Hepatic estrogen receptor (ER) has been demonstrated as a sensor of metabolic signals which is capable of tuning energy metabolism to reproductive needs (Della Torre et al., 2011). Indeed, hepatic ER holds a key role in the regulation of gluconeogenesis and lipid metabolism (Qiu et al., 2017). Metabolic fuels (glucose, fatty acids and amino acids), as well as hormones like IGF1, are important metabolic signals to the hypothalamus-pituitary-gonadal axis, which can trigger the onset of puberty. Furthermore, liver X receptor, member of the nuclear receptor family of TF, plays a crucial role in many metabolic pathways and involves in the regulation of female reproduction (Lobaccaro et al., 2013). Endocrine mechanisms that control the onset of puberty have been well characterized, but the precise mechanisms in which metabolic signals contribute to this process is still unclear. The crosstalk between the endocrine signals and the metabolic signals that drive puberty needs further investigation.

The liver transcriptomic analyses of gene expression under pre- and post-pubertal conditions provided a list of differentially expressed genes and relevant pathways (Chapter 1.3). However, the expression of several genes is modulated at both translational and post- translational levels. Studies of transcriptome and proteome in the hypothalamus and the pituitary presented in previous chapters noted the difference in number and correlation between differentially expressed genes and differentially abundant proteins. Hence, the study of liver proteomics might help to further understanding the pathways and hormones that link the nutritional/metabolic signals and the puberty process. This is the first documented application of liver proteomics to puberty transition in Bos indicus cattle.

2.3.2. Materials and methods Experimental procedures were approved by the Animal Ethics Committee of the University of Queensland, Production and Companion Animal group (certificate number QAAFI/279/12). Studied animals were purchased from two commercial Brahman breeders. The purchased heifers were of similar age and had weight less than 250 kg upon arrival to the Gatton Campus facilities of the University of Queensland. During the experiment, these heifers were managed as one cohort as they consumed a pasture-based diet. Pubertal status was defined using the observation of the first corpus luteum (CL). Serum progesterone concentrations were also measured to confirm a functional CL in post-pubertal heifers (2.0 0.7 ng/mL, mean SE). Pre-puberty heifers were randomly selected from the group that had never ovulated (plasma progesterone concentration 0.4 0.2 ng/mL, mean SE) and paired with post-pubertal animals in slaughter days.

127

The entire liver was removed from the animal and 3 samples of 1 cm3 were dissected from the liver and snap frozen in liquid nitrogen. In total, 6 pre- and 6 post-pubertal liver tissues were processed separately for protein identification and quantification as described previously (Chapter 2.1). Briefly, liver proteins were diluted, alkylated, precipitated and digested with trypsin. Peptides were desalted using C18 Ziptips (Millipore) and analysed by liquid chromatography electrospray ionization tandem mass spectrometric (LC-ESI-MS/MS). Peptides were separated in a gradient formed by mobile phase A (1% acetonitrile and 0.1% formic acid) and a mobile phase B (80% acetonitrile and 0.1% formic acid). Tandem MS analysis was conducted using TripleTof 5600 instrument (ABSciex) with a Nanospray III interface. Protein identification was analysed with ProteinPilot (AB SCIEX), searching UniProt database of B. taurus reference. User-defined search parameters were described previously (Nguyen et al., 2018). The ProteinPilot data (false discovery rate (FDR) < 0.01) were then used as ion libraries for sequential window acquisition of all theoretical mass spectra (SWATH) analysis. SWATH was performed on all samples using an IDA library prepared from 1 randomly chosen pre-pubertal sample and 1 randomly chosen post-pubertal sample. The spectra generated from SWATH acquisition and the ion library were imported into PeakView software v2.1 (SCIEX). Statistical analyses were then performed using MSstats (v2.6) in R as previously described (Choi et al., 2014; Zacchi &Schulz, 2016). The correlation between mRNA expression (Chapter 1.3) and protein abundance (current study) was computed using the cor() function in R package with a Pearson option. This was calculated by correlating the log-fold change (FC) values of mRNA and protein levels between pre- and post-pubertal Brahman heifers. Enriched gene ontology (GO) terms and Kyoto Encyclopedia of Genes and Genome (KEGG) pathways of differentially abundant proteins were analysed using STRINGv10 database (http://string-db.org). The interaction among proteins with scores > 0.6 were downloaded from STRINGv10 analysis and then transferred to Cytoscape software (Shannon et al., 2003) for further visualization and identify hubs of the interaction network.

2.3.3. Results

The study employed LC-ESI-MS/MS to investigate differentially abundant (DA) proteins in the liver of pre- and post-pubertal Brahman heifers. A total of 777 proteins were identified (FDR < 1%), and 630 of those were quantified. To identify DA proteins between pre- and post-pubertal period, the ion ratios were determined. The DA proteins were selected according to the adjusted P-value < 0.05. There were 275 DA proteins extracted (159 high

128

abundance and 116 low abundance proteins at post-puberty). Figure 2.3.1 presents a volcano plot of identified proteins in which abundance proteins are shown in red dot. The mass spectrometry proteomics data have been uploaded with the Proteome Xchange Consortium via the PRIDE partner repository with the dataset identifier PXD009419 (Vizcaino et al., 2016).

Figure 2.3.1 Volcano plot of differentially abundant proteins in the liver between pre- vs post- pubertal heifers. The X-axis represents the fold change while the y-axis represents statistical significance for each protein. Red dots indicated proteins that differed significantly (FDR < 0.05) between two groups. Proteins plotted in the left portion of the graph were expressed at a lower level in post-pubertal adipose tissue, and proteins in the right-hand portion had higher expression levels post- puberty. Symbols are provided for proteins with a |Fold Change (FC)| 1 and FDR 10-5. The correlation between protein level and its corresponding mRNA expression (Chapter 1.3) was also characterized. In total, 611 of 630 quantified proteins were mapped to the liver transcriptome dataset (Chapter 1.3) and these mapped proteins-mRNAs were used for calculating the correlation. The ratio of mRNA and protein in the liver for 611 mapped transcripts-proteins was positively correlated with correlation coefficient r = 0.13 (P-value = 0.001, Figure 2.3.2). In addition, 8 DA proteins were also identified as differentially expressed genes in the liver transcriptome. The expression patterns of these proteins were similar to that of mRNA, except for PHB2 and CYP2E1. Table 2.3 shows these genes and proteins that were both DE and DA.

129

Figure 2.3.2 Scatterplots with an associated correlation coefficient (r) for mRNA and protein expression in the liver.

Table 2.3 Differentially expressed genes and differentially abundant proteins between pre- and post-pubertal Brahman heifers according to both liver transcriptome and proteome studies.

UniProt Assession log2FC adj.P Gene symbol FC_mRNA T-test P Q2HJ97 -0.4681 6.41E-05 PHB2 0.2370 2.43E-02 Q95M40 -0.2972 1.30E-13 TXNDC17 -0.2573 1.43E-02 Q2TBV4 -0.1439 4.79E-03 CYP2E1 0.3729 4.97E-02 F1MAV0 -0.1368 9.72E-03 FGB -0.1140 4.23E-02 Q58DM8 0.0878 7.94E-07 ECHS1 0.2906 1.36E-02 Q3ZCJ6 0.1097 4.47E-04 EPHX1 0.2708 2.83E-02 F1MNA6 0.3709 3.00E-04 FTCD 0.2534 3.00E-02 F1N6P2 0.5159 5.89E-07 SURF1 0.1878 2.60E-02 FC means fold change Adj.P means adjusted P-value

To better understand the DA proteins, GO annotation was performed using 275 DA proteins between pre- and post-pubertal heifers. A total of 251 DA proteins were annotated to 293 GO biological process terms, 89 molecular function terms and 95 cellular component terms (FDR < 0.05). Each DA protein was annotated to at least one term. In the cellular component category, the DA proteins were mainly distributed in the cytoplasm (132 DA proteins), cell (127 DA proteins), intracellular (125 DA proteins), organelle (124 DA proteins), membrane

130

(119 proteins) and mitochondrion (57 DA proteins). In the molecular function category, the DA proteins were mainly related to catalytic activity (99 DA proteins), binding (99 DA proteins), oxidoreductase activity (35 DA proteins), carbohydrate derivative binding (32 DA proteins), enzyme regulator activity (13 DA proteins) and transporter activity (13 DA proteins). In the biological process category, the DA proteins were mostly associated with metabolic (122 DA proteins), cellular (112 DA proteins), single-organism (110 DA proteins), response to stimulus (44 DA proteins) and oxidation-reduction process (41 DA proteins).

It is well-established that proteins interact and cooperate to regulate biological mechanisms. Therefore, a KEGG pathway analysis was employed to reveal enriched pathways in the list of DA proteins between pre- and post-puberty in the liver. A total of 55 overrepresented pathways for the set of DA proteins between pre- and post-pubertal heifers were identified. Among these KEGG pathways, most of the pathways (FDR < 0.05) were associated with the amino acid metabolism such as tyrosine, tryptophan, pyruvate, arginine, proline, alanine, aspartate and glutamate. Some DA proteins were mapped for pathways involved in fatty acid degradation and metabolism (FDR < 0.05). Carbohydrate metabolisms were among the enriched pathways. In addition, this study also revealed the enrichment of steroid hormone biosynthesis, PPAR signalling pathway, oxidative phosphorylation and citrate cycle (TCA) pathways.

Out of 275 DA proteins, 194 DA proteins formed the interaction network (interaction scores > 0.6). These proteins (nodes) shared 719 edges in the network (Figure 2.3.3). Among these DA proteins, enoyl-CoA hydratase and 3-hydroxyacyl CoA dehydrogenase (EHHADH) were hubs in the interaction network. These proteins connected with other 39 nodes, following by aldehyde dehydrogenase 2 family member (ALDH2) and aldehyde dehydrogenase 1 family member L1 (ALDH1L1) with 33 and 31 connections, respectively. At post-puberty, EHHADH and ALDH2 experienced low abundance and ALDH1L1 was high abundance.

131

Figure 2.3.3 The interaction network of 275 differentially abundant proteins between pre- and post-pubertal Brahman heifers. Each node illustrates a protein. Edges represent the interaction among proteins. Node size indicates the abundance of proteins (more or less in abundance at post-puberty). Node colour range from green to red to show the gradient of connectivity: low to high connection of a specific protein.

2.3.4. Discussion

The study performed proteomic analyses in order to identify puberty-related proteins in the liver of pre- and post-pubertal Brahman heifers. 275 DA proteins were revealed in the liver proteomic study of pre- and post-pubertal heifers (adjusted P-value < 0.05). Only 8 proteins were identified as DE genes in liver transcriptome (Chapter 1.3). Previous chapters presented in this Thesis demonstrated relatively positive correlation between transcripts and proteins (P-value < 0.1), this study again reported a low correlation (r = 0.13, P-value = 0.001). The detection sensitivities for each approach are different, therefore the comparison between two approaches is difficult. Further, the difference between these two techniques might arise from the post-transcriptional and post-translation modification or protein half-live.

Among 8 DE genes and DA proteins, the depletion of thioredoxin domain containing 17 (TRP14, an alias of TXNDC17) enhanced TNF--induced JNK and p38 MAPK activation (Jeong et al., 2004b). Likewise, depletion of TRP14 also inhibited TNF--induced NF-kB

132

activation (Jeong et al., 2009). In rat ovaries, TNF-alpha played a significant role in the development of follicle and ovulation during puberty (Rice et al., 1998). For ECHS1, several reports described the role of this gene in fatty acid metabolism (Janssen et al., 1997; Ramayo- Caldas et al., 2014). The presence and balance of fatty acids are linked to numerous cellular functions such as oxidative stress, mitochondrial energy production, cytoplasmic and membrane functions which all affect fertility (Wathes et al., 2007; Tremellen, 2008). Moreover, the SURF1 gene plays a significant role in oxidative phosphorylation (Smeitink et al., 2001). The oxidative phosphorylation is a metabolic pathway in the production of adenosine triphosphat (ATP), a molecule supplying energy for metabolism (Smeitink et al., 2001; Zhu et al., 2010; Cao et al., 2015). The association between oxidative stress and female fertility has been previously reported (Agarwal et al., 2005; Ruder et al., 2009). The DE genes identified by suppression subtractive hybridisation analysis among three goat pubertal groups were involved in the oxidative phosphorylation, one of the seven most-enriched KEGG pathways (Cao et al., 2015).

It is logical given that DA proteins in the liver of pubertal Brahman heifers enriched in metabolic pathways that can produce energy for reproduction. Enriched pathways such as fatty acid metabolism and degradation, glycolysis, gluconeogenesis as well as amino acid metabolisms (e.g., tryptophan, tyrosine, arginine, proline, valine, glycine and lycine) were abundant. The breakdown of fat and carbohydrates by the TCA cycle and glycolysis is derived by oxidative phosphorylation inside mitochondria and subsequently generate adenosine triphosphate (ATP) for supporting reproduction (Berg et al., 2002). Of note, oxidative phosphorylation and TCA cycle are essential for mouse ovarian follicle development (Wycherley et al., 2005). This study revealed the increase in abundance of proteins involved in oxidative phosphorylation at post-puberty, resulting in the increase of ATP production which can provide energy for the puberty process.

Study of urine metabonomics data of Chinese girls noted that abnormal metabolism of aromatic amino acid like tryptophan and tyrosine might have a close correlation with central precocious puberty by inhibiting hypothalamus-pituitary-adrenal (HPA) axis and activating hypothalamus-pituitary-gonadal (HPG) axis (Yang et al., 2012a). This study also demonstrated that two pathways: tyrosine metabolism and tryptophan metabolism were upstream of HPG and HPA axis, whereas arginine and proline metabolism was downstream of these axes (Yang et al., 2012a). In the current study, the proteins involved in arginine and proline metabolism including ALDH2, GLUD1, GOT1, GOT2, ASL, HOGA1 and AGMAT were decreased in

133

abundance at post-puberty. Regarding tryptophan and tyrosine metabolism, both up-regulated and down-regulated proteins were identified in these pathways. Although it is hard to determine how these proteins interact and modulate these pathways, the enrichment of the up- and downstream pathways of HPG and HPA axis suggest that cattle puberty onset may be influenced by similar metabolic factors, suggested in cited human studies.

The PPAR signalling pathway was also identified in the enrichment analyses. Peroxisome proliferator-activated receptors (PPARs) are members of the ligand-activated nuclear receptor family and play an important role in energy homeostasis (Yang et al., 2008). Among three isotypes of PPARs, PPAR can regulate the expression of the leptin gene that is a metabolic hormone affecting reproduction in heifers (Garcia et al., 2002). The mRNA expression of PPAR decreased during luteinisation and the formation of corpus luteum in rat mature follicles (Komar et al., 2001). Further, the activation of PPAR can inhibit the production of progesterone and 17β-estradiol in porcine corpus luteum of early pregnancy (Kurzynska et al., 2014). PPAR was noted as a negative regulator of macrophages which regulated mice luteal development during embryo implantation (Ricote et al., 1998; Care et al., 2013). Proteins involved in this pathway (SLC27A2, ACADL, FABP1, EHHADH, SCP2, APOA1, DBI and CYP27A1) were experienced a decrease in abundance at luteal phase. In the current study, post-pubertal heifers were at the luteal phase and had a higher concentration of progesterone in comparison with pre-pubertal heifers. Collectively, the high level of proteins involved in hepatic PPAR signalling pathway at follicular phase may exert a negative effect on corpus luteum development and inhibit puberty process.

The liver is the site of steroid hormone inactivation, and we also discovered the steroid hormone biosynthesis as a significantly enriched KEGG pathway in our liver proteomics (FDR = 0.01). Differentially abundant proteins in this pathway were HSD11B1, HSD17B6, UGT1A1, CYP2D14, CYP2E1, CYP1A1, CYP2C18, LOC511498 and a novel protein. These DA proteins were decreased in abundance post-puberty, except for HSD17B6 and CYP1A. The decrease of HSD11B1 expression level after puberty was also detected in monkeys (Stute et al., 2012). Further, HSD11B1 protein is responsible for glucocorticoids regeneration (Draper et al., 2003) and the glucocorticoids secretion level was positively correlated with the age at puberty onset in humans (Shi et al., 2011). Therefore, the down-regulation of HSD11B1 in post-pubertal heifers seen in the current study was consistent with other studies and suggested its role in B. indicus puberty.

134

Understanding all direct and indirect interactions between proteins is an important prerequisite for enhancing our knowledge about molecular mechanism and protein function. Currently, protein associations and interactions were annotated at different levels in online resources (Salwinski et al., 2004; McDowall et al., 2009; Kerrien et al., 2012; Licata et al., 2012; Schaefer et al., 2012; Chatr-Aryamontri et al., 2013; Zhang et al., 2013; Baspinar et al., 2014; Schmitt et al., 2014). Among available online database, the STRINGv10 database (Search Tool for the Retrieval of Interacting Genes/Proteins) housed more than 200 million annotated interactions of 5 million proteins from more than 2031 organisms (Szklarczyk et al., 2014). The STRING basic interaction unit is the functional association. These interactions were derived from known experimental interactions, curated databases or predicted de novo by co- expression analysis as well as using genomic information or transferred observed interactions from one organism to other organisms using pre-computed orthology relations (Szklarczyk et al., 2014). Therefore, the DA proteins were submitted to STRINGv10 database to create a protein-protein interaction (PPI) network.

In the PPI network, EHHADH was identified as a hub with 39 connections. EHHADH was enriched in the PPAR signalling pathway, fatty acid metabolism and degradation, amino acid metabolism (e.g., tryptophan, valine, leucine and isoleucine), according to KEGG pathway analysis. Endometrial microarray study in early luteal phase endometrial revealed differential expression of EHHADH in human (Humaidan et al., 2012). Further, EHHADH was involved in network generated from endometrial gene expression data during the mid-luteal phase from low and high fertility heifers (Killeen et al., 2014). Among connections with EHHADH, the study found the connection between EHHADH and SDHA, a protein involved in the TCA cycle and oxidative phosphorylation. Of note, SDHA is known as a tumor suppressor gene and sub-network of tumor suppressor genes was hypothesized to coordinate the activity of neuroendocrine networks involved in the pubertal process (Ojeda et al., 2006; Burnichon et al., 2010). The interaction between a DE gene and DA protein: ECHS1 and EHHADH was also identified. Collectively, EHHADH might play a crucial role in liver metabolism and integrate the link between metabolic status and reproduction in cattle. Further investigation to elucidate this hypothesis is warranted.

2.3.5. Conclusion

In conclusion, the liver proteome provided a list of 275 candidate proteins from pre- and post-pubertal Brahman heifers. In comparison with liver transcriptome, there were 8 genes in common and the positive correlation between the mRNA-protein pair was observed (P-value

135

= 0.001). The functional analysis of DA proteins showed numerous enriched metabolic pathways (FDR < 0.05) that generate energy demands for the occurrence of puberty, such as fatty acid metabolism, carbohydrate metabolism, amino acid metabolism, oxidative phosphorylation and TCA cycle. Another pathway: PPAR signalling pathway that might integrate metabolic status and reproduction at luteal development was suggested. The inactivation of steroid hormones by the liver at the follicular phase can contribute to the initiation of puberty in heifers. Although additional work is warranted to elucidate the precise regulatory mechanisms of identified proteins and pathways, the results of the present work may indicate the potential role of EHHADH and PPAR signalling in the integration between metabolic signals and reproduction in Bos indicus breeds.

136

Chapter 2.4. Adipose tissue proteomic analyses to study puberty in Brahman heifers

Manuscript published in Journal of Animal Science., Volume 96, Issue 6, June 2018,

Pages 2392–2398

https://doi.org/10.1093/jas/sky128

137

Abstract

Adipose tissue has been recognized as an active endocrine organ which can modulate numerous physiological processes such as metabolism, appetite, immunity and reproduction. The aim of this study was to look for differentially abundant (DA) proteins and their biological functions in the abdominal adipose tissue between pre- and post-pubertal Brahman heifers. Twelve Brahman heifers were divided into two groups and paired on slaughter day. Pre-pubertal heifers had never ovulated and post-pubertal heifers were slaughtered on the luteal phase of their second estrous cycle. After ensuring the occurrence of puberty in post- pubertal heifers, abdominal adipose tissue samples were collected. Mass spectrometry proteomic analysis identified 646 proteins and revealed that 171 proteins showed differential abundance in adipose tissue between the pre- and post-puberty groups (adjusted P-value < 0.05). Data are available via ProteomeXchange with identifier PXD009452. Using a list of 51 highly DA proteins as target (adjusted P-value < 10-5), we found 14 enriched pathways. The metabolism of glucose, lipids and amino acids in the adipose tissue mainly participated in oxidation and energy supply for heifers when puberty occurred. Our study also revealed the DA proteins were enriched for estrogen signalling and PI3K-Akt signalling pathways, which are known integrators of metabolism and reproduction. These results suggest new candidate proteins that may contribute to a better understanding of the signalling mechanisms that relate adipose tissue function to puberty. Protein-protein interaction network analysis identified four hub proteins that had the highest degrees of connection: PGK1, ALDH5A1, EEF2 and LDHB. Highly connected proteins are likely to influence the functions of all DA proteins identified, directly or indirectly.

Keywords: Bos indicus, Brahman, estrogen signalling, heifers, proteomics, puberty

2.4.1. Introduction

Age of puberty is a heritable trait that determines the ability of a heifer to become pregnant in the first breeding season. Age of puberty is also correlated with rebreeding and with the likelihood to remain in the herd in subsequent years (Schillo et al., 1992; Perry &Cushman, 2013). Puberty impacts on heifer lifetime production. For example, Lesmeister et al., (1973) suggested that beef cows which reached puberty earlier calved earlier throughout their reproductive lives, compared to those that attained puberty later. Consistent with these ideas, early pubertal heifers (first calving as 2-year-olds) generated 0.6 more calves than those

138

calving first as 3-year-olds, over the first 5 years of productive life (Meaker et al., 1980). Therefore, selection for early age at puberty may improve cattle profitability and longevity.

In order to achieve early pubertal heifers, maintaining adequate nutritional management is crucial. Studies have reported that beef heifers maintained on low energy diets that limit weight gain reached puberty later than those fed high energy diets for greater weight gain (Rathbone et al., 2001; Nogueira, 2004). Insufficient nutrition considerably delays puberty in both Bos indicus and Bos taurus cattle (Abeygunawardena &Dematawewa, 2004). One of the ways in which nutrition affects the initiation of puberty is by influencing the increase of luteinizing hormone (LH) release. Heifers fed an adequate energy diet for growth presented increased LH secretion and attained puberty, while heifers fed a restricted diet failed to present increased LH pulse frequency and did not achieve puberty (Day et al., 1986). Likewise, dietary energy restriction inhibited the pre-pubertal increase of LH secretion in ovariectomized, ovary intact, and estradiol-treated heifers (Kurz et al., 1990). The role of nutrition and metabolism in pubertal development underpins the hypothesis that the adipose tissue impacts the onset of puberty (Schillo et al., 1992; Kinder et al., 1995; Hiney et al., 1996). This role merits further investigation.

Targeting adipose tissue, which can modulate complex physiological process including metabolism, appetite, body weight, inflammation and reproduction (Kusminski et al., 2016), with mass spectrometry proteomics is expected to enhance our knowledge about the links between nutrition, adipose tissue signalling, the onset of puberty, and the pulsatile LH secretion. Here, we present results from differential abundance analysis using mass spectrometry proteomics on adipose tissue of six pre- and six post-pubertal Brahman heifers. Our aim was to identify differentially abundant (DA) proteins and understand their biological functions in relation to the onset of puberty.

2.4.2. Materials and methods

Animals and samples

Samples were collected from twelve commercial Brahman heifers of similar age. These heifers were handled and managed as per approval of the Animal Ethics Committee of the University of Queensland, Production and Companion Animal group (certificate number QAAFI/279/12).

Pubertal status was defined using the observation of first corpus luteum (CL) with ultrasound (Johnston et al., 2009). Pre-pubertal heifers were paired and slaughtered with post-

139

pubertal heifers. Pre-pubertal heifers had never ovulated. Post-pubertal heifers were slaughtered on the luteal phase of their second estrous cycle. Experimental set-up took approximately nine months until the last post-pubertal heifer to achieve puberty.

Serum progesterone concentration for the pre-pubertal heifers were 0.4 ± 0.2 ng/mL and for the post-pubertal heifers were 2.0 ± 0.7 ng/mL. Associated data collected during tissue harvest of these two groups were body weights (BW) and condition score (CS). Body weights and CS were not different between the groups as previously reported (Fortes et al., 2016a; Nguyen et al., 2017).

Protein sample preparation, identification and quantification

Abdominal adipose tissues of six pre- and six post-pubertal Brahman heifers were dissected into small fragments. Samples were processed essentially as previously described (Tan et al., 2014). Each fraction was diluted in lysis buffer (6M guanidine, 10mM DTT and 50mM Tris pH 8) and sonicated at power level 4 for 10 seconds. After homogenization in a QIAshredder homogenizer (QIAGEN Pty. Ltd., Melbourne, VIC, Australia) at 4 °C for 10 min, the samples were vortexed vigorously at 30°C for 1 h. Next, 25mM acrylamide was added to samples, followed by another incubation at 30°C for 1 h. Excess acrylamide was quenched in the samples by the addition of 5mM DTT, and samples were precipitated with 4 volumes of methanol: acetone (1:1) at -20°C overnight. After dissolving the precipitate in 50mM ammonium acetate, the protein concentration was measured using Nanodrop (Thermo Scientific). Approximately 50 µg of protein was diluted to a final volume of 100 µL and digested by addition of 1µg of trypsin (Sigma) overnight at 37°C. Samples were kept at −20°C until mass spectrometry.

Mass spectrometry and data analysis

Peptides were desalted using C18 Ziptips (Millipore), resuspended in 9.25% acetonitrile and 0.1% formic acid and analysed by liquid chromatography electrospray ionization tandem mass spectrometry (LC-ESI-MS/MS) using a Prominence nanoLC system (Shimadzu) and TripleTof 5600 mass spectrometer with a Nanospray III interface (SCIEX) essentially as described (Bailey et al., 2012). In order to build the fragment ion library for SWATH (Sequential window acquisition of all theoretical mass spectra), we performed an Information Dependent Analysis (IDA) on 2 randomly chosen pre-pubertal sample and 2 randomly chosen post-pubertal sample. Fifteen l of each sample was injected. Samples were separated using reversed-phase chromatography on a Shimadzu Prominence nanoLC system.

140

Using a flow rate of 30 µl/min, samples were desalted on an Agilent C18 trap (0.3 x 5 mm, 5 µm) for 3 min, followed by separation on a Vydac Everest C18 (300 A, 5 µm, 150 mm x 150 µm) column at a flow rate of 1 µl/min. A gradient of 10 to 60% buffer B over 45 min, where buffer A = 1 % acetonitrile/0.1% Formic acid and buffer B = 80% acetonitrile/0.1% formic acid was used to separate peptides. Eluted peptides were directly analysed on a TripleTof 5600 instrument (ABSciex) using a Nanospray III interface. Gas and voltage settings were adjusted as required. MS TOF scan across 350-1800 m/z was performed for 0.5 s followed by information dependent acquisition of up to 20 peptides with intensity greater than 100 counts, across 100-1800 m/z (0.05 sec per spectra) using collision energy (CE) of 40 +/- 15 V. For SWATH analyses, MS scans across 350-1800 m/z were performed (0.5 sec), followed by high sensitivity data-independent acquisition mode using 26 m/z isolation windows for 0.095 s, across 400-1250 m/z. Collision energy values for SWATH samples were automatically assigned by Analyst software based on m/z mass windows.

The IDA files were analysed with Protein Pilot v5.0.1 (SCIEX). The database used by ProteinPilot to identify peptides and proteins was downloaded from UniProt (www.uniprot.org) on March 28, 2016, and contained a total of 43,813 entries assigned to B. taurus, including 6,870 from Swiss-Prot and 36,948 from TrEMBL. The search in ProteinPilot was done using the 4 IDA files described above with the following parameters: (1) Sample type: identification; (2) Cysteine alkylation: acrylamide; (3) Digestion: (4) Trypsin; Instrument: TripleTOF 5600; (5) Special Factors: None; (6) Species: Bos taurus; (7) ID focus: biological modifications; (8) Database: Bos taurus database as described above; (9) Search effort: thorough; (10) the false discovery rate (FDR) analysis: Yes; and (11) User modified parameter files: No. The IDA library prepared was then used for the analysis of the SWATH acquisition runs using PeakView v2.1 (SCIEX). The SWATH acquisition runs were obtained by injecting 2 l of sample for all samples (6 pre- and 6 post-pubertal samples). The parameters used for analysis in Peakview were as follows: 6 peptides/protein; 6 transitions/peptide; 99% peptide confidence threshold; 1% FDR; exclude modified peptides and shared peptides; 6 min extraction window; and 75 ppm extraction window width. Statistical analyses were performed using MSstats (v2.6) in R (Choi et al., 2014) as previously described (Zacchi &Schulz, 2016).

Correlation analysis The Pearson’s correlation coefficient (r) was used to compute the expression levels of mRNAs and proteins in the adipose tissue from pre- and post-pubertal Brahman heifers. The

141

cor() function in R was used in this calculation. The unmapped mRNA-proteins were not included in this analysis.

Functional enrichment analyses

The list of differentially abundant (DA) proteins was analysed with Bos taurus protein database in the STRINGv10 database (http://string-db.org) for functional enrichment analyses. Adjusted P-values for multiple testing were reported. The protein-protein interaction score was downloaded from the STRINGv10 database and then transferred to the Cytoscape software (Shannon et al., 2003) for further visualization and scrutiny of the protein-protein interaction network derived from the adipose tissue proteome data. Network analyses identified proteins that were hubs.

2.4.3. Results

A total of 646 proteins were identified in the adipose tissue of Brahman heifers (FDR < 0.01). Using relative quantitative SWATH analysis, 497 proteins were quantified. The mass spectrometry proteomics data have been uploaded with the Proteome Xchange Consortium via the PRIDE partner repository with the dataset identifier PXD009452 (Vizcaino et al., 2016). Among quantified proteins, 171 proteins had different levels in the comparison between pre- and post-pubertal heifers (adjusted P-value < 0.05). Further information of these differentially abundant (DA) proteins can be found at Supplementary Table 1 in (Nguyen et al., 2018). Of these DA proteins, 9 proteins were significantly different with a fold change |FC| > 1 and adjusted P-value < 10-5 (Table 2.4.1). The proteins galectin 1 (LGALS1), protein kinase cAMP- dependent type II regulatory subunit alpha (PRKAR2A) and cofilin 1 (CFL1) were less abundance at post-puberty, whereas transketolase (TKT), VPS35, retromer complex component (VPS35), extracellular matrix protein 1 (ECM1), vacuolar protein sorting 4 homolog A (VPS4A), clusterin (CLU) and hemoglobin, beta (HBB) were more abundance at post-puberty.

The correlation between protein level and its corresponding mRNA expression (Chapter 1.4) was also characterized. In total, 461 out of 497 quantified proteins were mapped to the adipose tissue transcriptome dataset (Chapter 1.4) and these mapped proteins-mRNAs were used for calculating the correlation. The ratio of mRNA and protein in the adipose tissue for 461 mapped transcripts-proteins was negatively correlated with correlation coefficient r = -0.06 (P-value = 0.2). This lack of correlation could be expected since only 07 DA proteins were also differentially expressed (DE) genes in adipose tissue transcriptome (Chapter 1.4).

142

The expression patterns of these proteins were different from that of mRNA, except for C3, CKM and a novel gene (ENSBTAG00000006864). Table 2.4.2 shows these genes and proteins that were both DE and DA.

Table 2.4.1 Highly significant differentially abundant proteins in comparison between pre- and post-pubertal Brahman heifers in abdominal adipose tissue.

Ensembl ID UniProt ID Gene symbol Fold change Adjusted P- value ENSBTAG00000015089 P11116 LGALS1 -1.38 9.4 x 10-06 ENSBTAG00000014205 P00515 PRKAR2A -1.11 1.36 x 10-09 ENSBTAG00000021455 B0JYL8 CFL1 -1.05 3.21 x 10-08 ENSBTAG00000003758 Q6B855 TKT 1.07 8.1 x 10-07 ENSBTAG00000002493 Q2HJG5 VPS35 1.08 4.6 x 10-05 ENSBTAG00000003806 A5PJT7 ECM1 1.08 1.0 x 10-07 ENSBTAG00000001659 Q2HJB1 VPS4A 1.35 8.5 x 10-08 ENSBTAG00000005574 P17697 CLU 1.44 7.9 x 10-05 ENSBTAG00000038748 D4QBB3 HBB 1.64 9.8 x 10-12

Table 2.4.2 Differentially expressed genes and differentially abundant proteins between pre- and post-pubertal Brahman heifers according to both adipose tissue transcriptome and proteome studies.

UniProt Assession Gene symbol log2FC* FC_mRNA* Q3MHN2 C9 0.17 -2.78 Q2KJF1 A1BG 0.17 -3.43 A0A0F6QNP7 C3 0.22 2.16 Q2KIU3 MGC137211 0.28 -2.68 Q2KIF2 LRG1 0.40 -1.22 Q9XSC6 CKM 0.49 1.78 F1MVK1 Uncharacterized protein 0.93 1.02 FC means fold change * P-value < 0.01

Functional enrichment analysis was carried out to provide an overview of gene ontology categories and pathways of 51 DA proteins between pre- and post-pubertal heifers (adjusted P-value < 10-5). These 51 DA proteins were assigned to 45 enriched GO terms (adjusted P-value < 0.05), wherein 15 terms were categorized into biological process, 21 into molecular function and 18 into cellular component. The oxidation-reduction process, lipid

143

metabolic process, and fatty acid metabolic process were enriched among the biological process GO terms (adjusted P-value < 0.05). Proteins involved in these three biological processes GO terms increased abundance after puberty. Pathway analysis revealed 14 enriched pathways (adjusted P-value < 0.05), including glucose and amino acid metabolism as well as estrogen signalling pathway and PI3K-Akt signalling pathway. These last two signalling pathways contribute to the process of puberty as discussed in the following section.

From the protein-protein interaction network, hubs of the interaction network that showed a high degree of connectivity were identified (Figure 2.4). The protein with the highest number of connections (degree = 12) was phosphoglycerate kinase 1 (PGK1). This hub was followed by 3 other proteins that had the same number of connections (degree = 11): aldehyde dehydrogenase 5 family member A1 (ALDH5A1), eukaryotic translation elongation factor 2 (EEF2) and lactate dehydrogenase B (LDHB). Among these hubs, PGK1 and LDHB were annotated to the glycolysis pathway. Figure 2.4 shows the interaction network among DA proteins of Brahman heifers.

Figure 2.4 Protein-protein interaction network in the adipose tissue. Each node represents a protein. Protein nodes which are enlarged represent the high abundance (more or less abundance at post- puberty). Node colour range from green to red for low to high connection of a specific protein.

2.4.4. Discussion

Adipose tissue has been considered as an active endocrine organ with diverse functions including immune responses, glucose metabolism, bone metabolism and reproduction (Kusminski et al., 2016). However, relatively little is known about the association between

144

proteins secreted by the adipose tissues and the metabolic status of pubertal cattle. Adipose proteomic studies were focused on molecular mechanisms underlying fat accumulation or fat deposition for milk production and meat quality (Zhao et al., 2010; Cho et al., 2016; Ceciliani et al., 2018). Our study was the first to screen the DA proteins of the adipose tissue of Brahman heifers between pre- and post-puberty. This proteomic dataset from Brahman heifers adds novel information on abundant proteins in the adipose tissue between pre- and post-puberty. Comprehensive analysis of 171 DA proteins will help to investigate the molecular mechanisms involved in puberty onset of Bos indicus sub-species, a globally important subspecies for tropical systems.

Functional enrichment analyses provided GO and pathway annotation to the reported abundant proteins. Six proteins with increased abundance post-puberty were involved in the metabolic process of lipid and fatty acid. These proteins were ALDH1A1, enoyl-CoA hydratase, short chain 1 (ECHS1), Acyl-CoA synthetase medium-chain family member 1 (ACSM1), Acyl-CoA synthetase family member 2 (ACSF2), fatty acid synthase (FASN) and ATP synthase, H+ transporting, mitochondrial F1 complex, beta polypeptide (ATP5B). It is possible that post-pubertal heifers need to mobilize body fat to provide sufficient energy for cycling and in preparation for reproduction. Seven proteins with increased abundance post- puberty including ECHS1, ALDH1A1, LDHB, FASN, succinate-CoA ligase alpha subunit (SUCLG1), peroxiredoxin 1 (PRDX1) and cytochrome c oxidase subunit 5A (COX5A) were implicated in the oxidation-reduction process, suggesting the role of redox pathways in the control of female reproduction. Altering the oxidation-reduction status may influence signalling pathways, transcription factors and epigenetic mechanisms and impact on oocyte and embryo quality (Agarwal et al., 2008). Moreover, the high representation of catalytic activity and binding activity among DA proteins suggest that pubertal heifers had strong metabolic activity and enhanced biological reactions. Proteins annotated to these catalytic and binding activities had increased expression levels after puberty, during the luteal phase. Future work using more animals, targeting all phases of the estrous cycle, and with deeper proteome coverage will further our understanding of this process. Experiments relating nutrient restriction during puberty to changes in the proteome may also help in clarifying the biological changes that occur in heifers as a result of puberty.

The wide range of enriched pathways in this study indicated that the function of adipose tissue in relation to puberty is complex. It seems that numerous metabolic pathways working together in the adipose tissue are both influenced by puberty and signal feedback to this

145

process. Some of the pathways mapped by the proteins that increased in abundance after puberty were involved in glycolysis, butanoate metabolism, propanoate metabolism, pyruvate metabolism and biosynthesis of amino acids. The increased expression level of these proteins fuelled the upregulation of these pathways associated with energy metabolism. It is possible that the upregulation of these pathways relates to mobilizing sufficient energy for pubertal development and subsequent reproductive capability.

Proteins involved in estrogen signalling pathway included two up-regulated proteins: phosphoenolpyruvate carboxykinase 2, mitochondrial (PCK2) and heat shock protein 90 alpha family class B member 1 (HSP90AB1) and three down-regulated proteins: collagen type VI alpha 2 chain (COL6A2), collagen type VI alpha 3 chain (COL6A3) and collagen type I alpha 1 chain (COL1A1). Estrogen signalling is an important factor in gonadostat theory. The positive and negative feedback of gonadal steroids on the secretion of GnRH and gonadotropin that are necessary for ovulation has been known for decades (Docke &Dorner, 1965; Kulin et al., 1969; Clarke &Cummins, 1984; Yasin et al., 1995; Moenter et al., 2009; Mayer et al., 2010). In adipose tissue, estrogens were recognized as regulators of lipid metabolism and glucose metabolism (Kim et al., 2014). Estrogen signalling is an expected link between the reproductive axis and the adipose tissue.

Proteins involved in the PI3K-Akt signalling pathway included an up-regulated protein (heat shock 70kDa protein 1A [HSPA1A]) and two down-regulated proteins (G protein subunit alpha o1 [GNAO1] and HSP90AB1). The PI3K-Akt pathway is known as an integrator of reproductive and metabolic functions (Acosta-Martínez, 2011). In fact, this pathway is activated by growth factors and hormones such as estrogen receptor alpha (ER), insulin receptor (IR), insulin-like growth factor I (IGF1) and leptin receptor (LepR) that are critical for GnRH release, puberty and sexual behaviour (Etgen &Acosta-Martinez, 2003; Morrison et al., 2005; Morton et al., 2005; Gelling et al., 2006; Mayer et al., 2010). Further, estrogen and progesterone are known to modulate the expression or activity of members of the PI3K-Akt signalling pathway such as p85, IRS-1 and IRS-2 (Acosta-Martínez, 2011). Recently, a study in male mice noted that deletion of a catalytic subunit of PI3K (p110) in the adipose tissue led to the delay of puberty onset (Nelson et al., 2017). Our results suggest that PI3K-Akt signalling is important in cattle puberty as well.

From the protein-protein interaction network, this study identified four hub proteins that had the highest degrees of connection: PGK1, ALDH5A1, EEF2 and LDHB. Among

146

these hub proteins, PGK1 and LDHB are involved in glycolysis. In particular, PGK1 is a glycolytic enzyme (Zhao et al., 2016), whereas LDHB is involved in a post-glycolysis process. LDHB encodes an enzyme that can catalyze the interconversion of pyruvate and lactate with concomitant interconversion of NADH and NAD+. The 2 other hub proteins, ALDH5A1 and EEF2, also play a role in metabolic pathways. The protein ALDH5A1 is a member of aldehyde dehydrogenases family that is known to diminish oxidative stress (Vasiliou et al., 2004; Chen et al., 2009b; Singh et al., 2013). The link between female infertility and oxidative stress suggested that an increase of ALDH5A1 abundance at puberty in Brahman heifers may reduce the risk of oxidative stress and help animals to attained puberty (Ruder et al., 2009; Gupta et al., 2014). Lastly, the protein EEF2, a member of G- protein superfamily, has been known as an essential factor for protein translation and synthesis (Kaul et al., 2011; Hekman et al., 2012). These four protein hubs should be further investigated in the context of cattle puberty.

2.4.5. Conclusion

Reproductive function, enabled through the processes of puberty, is an energy intense activity. It cannot be denied that fatty acids, AA, purines and pyrimidines as well as metabolites are essential compounds required for the final stage of oocyte maturation. The mature bovine cumulus-oocyte complex consumes twice as much oxygen, glucose and pyruvate as the immature cumulus-oocyte complex (Sutton et al., 2003a). The maturing cumulus-oocyte complex uses glucose for energy production and numerous other cellular processes (Sutton et al., 2003b). Energy for the reproductive function is likely mobilized from peripheral tissues, such as abdominal fat. Consistent with this, this study has identified candidate proteins that are increased in abundance in post-puberty adipose tissue that are key for fatty acid metabolism, lipid metabolism, and glycolysis/gluconeogenesis. Further investigation of our candidate proteins is necessary to fully understand their roles in the process of puberty in Bos indicus heifers.

147

SECTION III. FINAL DISCUSSION AND FUTURE DIRECTIONS

148

III. 1. Research overview

The multitude of DE genes and DA proteins discovered across four tissues (Chapters 1.1 to 2.4) fit the expectations for a complex biological process that impacts in the whole body. Our evidence also fits the emerging inheritance model suggested by genomic approaches that described a complex genetic architecture for puberty (Liu et al., 2009; Fortes et al., 2010b; Nascimento et al., 2016). Puberty is controlled by complex regulatory gene networks comprising multiple functional modules (Ojeda et al., 2006). It is a challenge to identify the impact of each gene or protein on pubertal development, given this scenario where we are probably dealing with multiple genes of small effects. In this cases, a systems biology approach, with network inference and analyses might be a better fit. Network analyses have successfully identified master regulators of complex processes that can be the target for gene manipulation or selection.

Due to the complexity of the puberty process, it is important to enhance our knowledge of the biological mechanisms involved, including gene and protein interactions. Until today, genomics has opened the opportunity for the characterization of genes, genetic markers and pathways related to puberty. In heifers, GWAS reported 350 genetic markers associated with age at first calving and age at puberty (Hu et al., 2016). These GWAS point to a complex genetic architecture of puberty that is probably also influenced by transcriptional and post- transcriptional regulation of gene expression levels. The analyses of complementary “omics” approached, beyond GWAS, has merit in the context of understanding the biological mechanisms of complex traits such as puberty. Transcriptomic profile in Bos taurus heifers reported genes associated with the puberty process (Cánovas et al., 2014a). The relative importance for puberty of each gene or transcript may depend on the species or breed under investigation. Differences between breeds were observed in GWAS that targeted age at puberty (Hawken et al., 2012b) and so the current understanding is that different breeds may carry different mutations that affect puberty. In this context, the regulatory networks need to be investigated within and between breeds to enhance our understanding of the process at the species level.

To better understand the molecular basis of puberty in Bos indicus cattle, a predominant sub-species in sub-tropical regions, transcriptome profiles between pre- and post-pubertal Brahman heifers have been investigated using RNA-sequencing technology, as presented in the above results chapters. The mRNA profiles revealed numerous DE protein-coding mRNAs, which helped to enhance knowledge of biological pathways activated (or down-regulated)

149

post-puberty in Bos indicus. DE genes were identified in each tissue investigated, confirming that puberty is a whole-body event. However, transcriptome profiles do not accurately reflect the protein abundance (Gygi et al., 1999; de Sousa Abreu et al., 2009; Carvalhais et al., 2015; Payne, 2015; Edfors et al., 2016). Therefore, proteome analysis was needed to characterize DA proteins. Proteome analyses in this context complemented transcriptome profiles and previous GWAS efforts. A handful of DE genes were also DA proteins, reinforcing the evidence for these candidate molecules as key parts of the molecular mechanisms that drive puberty in Bos indicus.

Reproductive tissues (hypothalamus and pituitary) and metabolism-related tissues (liver and adipose tissues) were targeted with transcriptomic and proteomic analyses. In this Thesis, RNA-seq and LC-MS/MS were used to profile mRNA and protein expression within the same animals (n = 12), experimental conditions (pre- and post-puberty) and tissues. This study suggested an overall lack of correlation between transcriptome and proteome. Future research using these and similar datasets will improve our understanding of post-transcriptional regulation as one of the mechanisms that erodes the transcript to protein correlation.

Research chapters grouped as “Part 1” presented and discussed transcriptome profiles, while “Part 2” detailed findings of proteome profiles for each tissue. In this final section, the overview of experimental findings and future directions are outlined.

III.1.1. RNA-sequencing and mass spectrometry data between pre- and post-pubertal Brahman heifers

In Part 1 of this thesis, I presented global gene expressions of 4 tissues including the hypothalamus (Chapter 1.1), the pituitary gland (Chapter 1.2), the liver (Chapter 1.3) and the adipose tissue (Chapter 1.4), in a comparison between pre- and post-pubertal heifers using RNA-sequencing. Each transcriptome dataset provided reliable sequence reads in which each individual sample had about 60 million sequence reads. Despite the absence of a Bos indicus reference genome, these datasets provided 60 to 70% mapped reads. Of the 24,596 annotated Bos taurus genes; 16,978 genes (including 1723 novel genes) were found expressed across studied tissues (RPKM ≥ 0.2), which accounted for 69% of the known genome (Figure III.1). Previous studies noted that approximately 30 million reads are sufficient to detect more than 90% of annotated genes in mammalian genomes (Wang et al., 2011; Lee et al., 2013; Singh et al., 2017). These mapped sequences can represent known and novel isoforms that are expressed in heifer key tissues related to puberty and facilitated subsequent analyses including detection

150

of DE genes and construction of gene co-expression networks (Chapter 1.1 to 1.4). These valuable datasets of RNA-sequencing were made publicly available at the Gene Expression Omnibus (GEO; https://www.ncbi.nlm.nih.gov/geo/; accession number GSE111339).

Together with transcriptome analysis, proteome analysis is a pivotal step for understanding the relevant biological processes. Understanding the proteome profiles between pre- and post-pubertal conditions have provided new information into the molecular mechanisms underlying puberty in Bos indicus. The second goal of this Thesis was to examine bovine multi-tissues proteome and to generate a list of high-confidence proteins. To my knowledge, research chapters presented in this Thesis were the first multi-tissue proteome profiles of pre- and post-puberty in Bos indicus cattle. These valuable data of mass spectrometry were made publicly available via The PRoteomics IDEntifications (PRIDE) database (https://www.ebi.ac.uk/pride/archive). Unique accession number of proteomics data are PXD009620 (Chapter 1.1), PXD009619 (Chapter 1.2), PXD009419 (Chapter 1.3) and PXD009452 (Chapter 1.4).

Figure III.1 Venn diagram of expressed mRNAs and proteins as well as differentially expressed (DE) mRNA and differentially abundant (DA) protein in each dataset. Information of this figure was retrieved from data reported all research chapters presented in Part 1 and Part 2.

Combining transcriptomic and proteomic analyses of reproductive tissues and pubertal- related tissues in Bos indicus is necessary to interpret the complex molecular mechanisms controlling protein expression in the puberty process. At first, the quantified proteins obtained in multi-tissue proteome analyses were compared with their corresponding transcripts from transcriptome analyses. Figure III.1 summarizes the comparison between expressed genes and

151

proteins in each tissue as well as overlapped proteins - transcripts. Although more than 84% of proteins were mapped to the transcriptome data, the weak correlations between transcriptome and proteome data of 4 studied tissues were observed. When protein abundance was compared with corresponding transcript expression levels the Person correlation was weak: 0.05 (P-value = 0.16; for Hypothalamus; Chapter 1.1 and 2.1), 0.006 (P-value = 0.8; for pituitary gland Chapter 1.2 and 2.2), 0.13 (P-value = 0.001; for liver Chapter 1.3 and 2.3) and -0.06 (P-value = 0.2; for adipose tissue Chapter 1.4 and 2.4). The weak correlation between mRNA levels and protein abundance was also demonstrated in bovine mammary gland studies (r = 0.05, P-value < 0.05) (Dai et al., 2017). These weak correlations between mRNA and protein data are consistent with analyses of yeast and human transcriptome and proteome analyses (Anderson &Seilhamer, 1997; Gygi et al., 1999; Chen et al., 2002; Griffin et al., 2002; Ghaemmaghami et al., 2003; Greenbaum et al., 2003; Washburn et al., 2003; Tian et al., 2004; Nie et al., 2006; Jayapal et al., 2008; de Sousa Abreu et al., 2009; Haider &Pal, 2013; Bai et al., 2015; Carvalhais et al., 2015; Payne, 2015; Edfors et al., 2016).

The discrepancy between transcriptome and proteome analyses might arise due to the experimental errors in terms of data extraction for each approach (Maier et al., 2009). These errors might bring noise that can impact the correlation between mRNA expression and protein abundance. The normalization of our transcriptome data was performed with mixed model equations accounting for factors including tissue, gene and animals. Meanwhile, the normalization of protein abundance was conducted using the linear mixed model in MSstats (v2.6) in R with consideration of proteins (Choi et al., 2014). Further, the identification of transcripts was performed by mapping sequence reads to the complete Bos taurus reference genome which accounted for a maximum of two gaps or mismatches in each sequence. Whereas, peptide identification performed in this Thesis was searched against a Bos taurus protein database (UniProt) which comprised only 6870 reviewed entries from Swiss-Prot and 36,948 unreviewed entries from TrEMBL. Identified peptides were only those presented in the search database. Another limitation to consider when comparing these two “omics” datasets is the presence of isoforms sharing identical peptide fragments.

However, and perhaps more important than methodological issues, the relationship between mRNAs and proteins is not straightforward due to biological factors. For example, translational efficiency, folding, alternative splicing, secretion, modification, degradation as well as transport and localization of mRNA and proteins all amount to create a disparity between mRNA expression and protein abundance in the same sample (Bitton et al., 2008;

152

Manzoni et al., 2016). A study in 1989 reported that mRNA half-life is about 10 – 20h whereas it is about 48-72h in eukaryotic proteins (Hargrove &Schmidt, 1989). These above-mentioned issues need to be considered when compared mRNA and protein expression. Taken together, the poor global correlation between mRNAs and proteins presented in this Thesis confirmed that the transcriptome data is different from and needs the complementary information from the corresponding proteome data. In short, changes in mRNA expression might provide a limited understanding of changes in the expression of proteins.

III.1.2. Differential expression of transcripts and proteins between pre- and post-pubertal heifers

As already pointed out, the goals of this Thesis were to analyse transcriptome and proteome of biologically important representative tissues and to generate high confidence lists of candidate mRNA and proteins associated with puberty. Research chapters grouped as “Part 1” revealed DE genes between pre- and post-pubertal heifers. The number of expressed and differentially expressed genes in each tissue was summarized and presented in Figure III.1. In terms of proteome analyses, research chapters grouped as “Part 2” also provided the lists of identified, quantified and DA proteins as summarized in Figure III.1 and III.2. From the sets of DE genes and DA proteins in each tissue, corresponding mRNA and protein pairs were identified (Figure III.1 and III.2). In all tissues, the number of genes that coded for DE mRNAs that were also DA proteins was small, and always a minority. These common DE genes and DA proteins represented high-confidence candidates for further studies of puberty process in Bos indicus.

Figure III.2 Number of proteins identified, quantified, differential abundance and common differentially expressed genes and proteins in four studied tissues. Information of this figure was retrieved from data reported in chapters 2.1, 2.2, 2.3 and 2.4 of Part 2.

153

An example of a gene that codes for a DE transcript and a DA protein in hypothalamus “omics” studies is PLCB1. The contribution of PLCB1 to growth at puberty onset as well as in the gonadotropin signalling pathway was reported in genome-wide association studies in pig (Meng et al., 2017). Another example of a gene that was DE and coded for a DA protein was IGFBP2 in the pituitary gland, which was previously reported to affect female puberty, noted as predominant protein secreted in the pituitary gland during pre-ovulatory and early luteal phase in beef cattle (Funston et al., 1995; Roberts et al., 2001). Similarly, the DE gene CHGB codes for the corresponding protein that was DA and it belongs to a chromogranin protein family that affects the secretion and storage of FSH and LH at different periods of the estrous cycle in sheep (Crawford &McNeilly, 2002). ENO1, another DE and DA gene in our studies was a down-regulated gene in a pre-puberty gene network in mice (Pal et al., 2017).

Despite their potential relevance, the genes that were DE and DA at the same time are the exception to the rule. The majority of the transcriptome data is not well correlated with the proteome data. This weak (or absent) correlations point to the necessity of enhancing our understanding of transcription and translation regulatory mechanisms. These regulatory mechanisms are complex and multifactorial. In this thesis, I focused on the role of transcription factors as the metrics and the databases for TF and TFBS are available for the bovine species and this availability facilitates the analyses. Future work could explore the present datasets to analyse the role of non-coding RNA and other regulatory molecules.

III.1.3. Potential regulators of puberty process in Brahman heifers

Using the regulatory impact factor (RIF) metric described by Reverter et al. (2010), this Thesis provided lists of regulatory transcription factors (TF) expressed in each studied tissue. A Venn diagram of TF in each tissue and TF deemed relevant across tissues is presented (Figure III.3). Among these TF, estrogen receptor beta (ER) and transcription factor CP2 like 1 (TFCP2L1) were identified as important TF in the hypothalamus, the pituitary and the liver. Whereas, NFKB1 was identified as an important TF in the hypothalamus, the pituitary and the adipose tissue. These “key” TF were also listed as potential regulators in the gene co- expression network derived from pre- and post-pubertal Brangus heifers (Cánovas et al., 2014a). The fact that this Thesis studies and the work in Brangus heifers identified the same TF suggests their role in cattle puberty regardless of the breed under investigation. Future work could target specifically these TF to verify their role in cattle puberty by using gene editing methodologies for example.

154

Figure III.3 Venn diagram of transcription factors identified by regulatory impact factor metric in the liver (LIV), the hypothalamus (HYP), adipose tissue (FAT) and pituitary (PIT). Information of this figure was retrieved from research chapters reported in Part 1 of this Thesis.

This Thesis also highlighted the importance of zinc finger TF in cattle puberty. According to the RIF metrics, 26 (hypothalamus; Chapter 1.1), 22 (pituitary gland; Chapter 1.2), 23 (liver; Chapter 1.3) and 19% (adipose tissue; Chapter 1.4) of TF identified as important regulators of puberty in Brahman heifers coded for genes of the zinc finger family. A ZNF sub- network of ovary transcriptomic study has been published for Brahman pubertal heifers (Nguyen et al., 2017). Likewise, ZNF genes formed a sub-network controlling puberty in mice (Lomniczi et al., 2013). The expression of ZNF mRNA level before puberty was noted to be decreased (Abreu et al., 2013; Lomniczi et al., 2015b; Nguyen et al., 2017). Noteworthy, Zinc finger genes have been involved in the epigenetic regulation of puberty in female primate (Lomniczi et al., 2015b). GWAS also reported the association between SNPs located near ZNF genes and age of menarche in women (Perry et al., 2009b; Elks et al., 2010; Perry et al., 2014a). In addition, makorin ring finger protein 3 (MKRN3), a protein with 4 zinc finger domains, is important for the initiation of puberty (Liu et al., 2017). These evidence and our findings confirmed ZNF’s function in the puberty initiation and suggest that these ZNF mechanisms in puberty are conserved across mammals.

III.1.4. Biological significance of transcripts and proteins identified by transcriptomic and proteomic analyses

To understand better that biological function of DE genes and DA proteins in these pre- and post-puberty comparisons, I applied functional enrichment analyses. Lists of DA proteins showed overrepresentation for numerous GO terms and pathways using Stringv10 web-based tool (Snel et al., 2000; Szklarczyk et al., 2014). Whereas, only DE genes in adipose tissue were

155

significantly enriched for several GO terms and pathways, using DAVID (Huang da et al., 2009). Again, it can be assumed that if the analysis was performed solely on a single approach (mRNA or protein), we would have a less complete picture of the molecular mechanisms of puberty in cattle and might reach different conclusions.

Functional enrichment analyses of DE genes and DA proteins in the adipose tissue revealed terms in common between these target lists. The terms in common in the molecular function category of GO terms were: enzyme inhibitor activity, endopeptidase inhibitor activity and calcium ion binding (Chapter 1.4 and 2.4). These terms were also found in functional analyses of DA proteins generated from the pituitary (Chapter 2.2) and the liver proteome (Chapter 2.3). Calcium plays a crucial role in neuronal energy metabolism, synaptic signalling processes and neurotransmission (Brini et al., 2014). For example, the binding of calcium to a presynaptic calcium sensor triggers neurotransmitters and hormone release in pituitary gland cells (Mason et al., 1988). Calcium ion binding was also enriched in the list of DA proteins in the hypothalamus of pre- and post-pubertal goats and rats (Gao et al., 2018), confirming its involvement in puberty process in mammals. In short, calcium signalling is fundamental to brain function during puberty.

The KEGG pathway analysis in the hypothalamus found that up-regulated proteins at post-puberty were enriched in GnRH, calcium and oxytocin signalling pathways, whereas down-regulated proteins were found in estrogen signalling pathway, axon guidance and GABAergic synapse (Chapter 2.1). These findings are in agreement with the complex feedback effects on GnRH release that trigger puberty (Bliss et al., 2010). Moreover, pathways such as oxidative phosphorylation, the tricarboxylic acid cycle, glycolysis and gluconeogenesis, metabolism of amino acids and carbohydrate metabolisms were among enriched pathways for the DA proteins in the hypothalamus (Chapter 2.1), the pituitary (Chapter 2.2), the liver (Chapter 2.3) and the adipose tissue (Chapter 2.4). Of note, oxidative phosphorylation and tricarboxylic acid cycle are essential for mouse ovarian follicle development (Wycherley et al., 2005). The activation of these pathways at puberty referred to the increase of ATP production, which can provide energy for pubertal development. In addition, the down-regulation of steroid hormone biosynthesis in the liver (Chapter 2.3) at post- puberty noted the involvement of the liver as a site of steroid hormone inactivation, which then affects the initiation of puberty in cattle.

Noteworthy, although the correlation between mRNA expression and protein abundance was weak, functional enrichment analysis showed similarities in several terms and

156

pathways between two approaches as well as across studied tissues, suggesting the similarity at functional level despite specific gene activation in each analysis. It seems that terms and pathways related to cellular processes, energy metabolism or cell cycles were the most cohesive. Whereas, the signalling pathways appeared to have a high level of inter-pathway communications. Other omics studies in dairy cows, pigs and human also reported the weak correlation between mRNA-proteins pairs and the similarities at the functional level (Zhao et al., 2009; Murgiano et al., 2010; Yang et al., 2016a; Dai et al., 2017). Through integrated transcriptome and proteome analyses, this Thesis identified important biological process and pathways that directly or indirectly regulate the puberty process in Bos indicus.

III.2. Future directions

The findings presented in this PhD Thesis provide useful lists of potential pubertal- related candidate genes and proteins as well as information of relevant TF and molecular mechanisms associated with puberty in cattle. These lists and the derived networks are a starting point for future experimental approaches that might dissect with more detail the role of each gene or protein as well as to investigate if the predicted TF and networks can be validated. In this section, we discuss possible avenues for future investigation.

In this thesis work, the identification of transcripts and proteins was based on the available Bos taurus databases. Bovine Bos taurus and Bos indicus sub-species diverged about 330,000 years ago and so their genome sequences contain important differences (Elsik et al., 2009; Porto-Neto et al., 2013; Porto-Neto et al., 2014). Genomic differences and divergence can impact these Thesis findings. In the absence of a Bos indicus reference genome, transcriptome or proteome, this work used the Bos taurus reference and databases and therefore it depends on the levels of homology between the two sub-species. The upcoming release of the first draft genome for Bos indicus will provide opportunities to improve the transcriptome and proteome data generated from this work. Using a Bos indicus genome, data produced by this work can serve for more refined comparisons between Bos indicus and Bos taurus gene expression and protein levels. Further, the Bos indicus reference genome is likely to use the RNA sequencing data presented herein to annotate specific isoforms and SNPs that might be unique to this tropically adapted sub-species. The raw datasets generated from multi-tissues transcriptome and proteome analyses of this study are available on public platforms, which then will no doubt provide untapped information for further studies by the scientific community. For instance, transcriptome datasets that are available via the Functional Annotation of Animal Genomes (FAANG) consortium (http://data.faang.org/home) can be

157

used in future studies for the detection of alternatively spliced isoforms, identification of RNA editing sites or identification of expressed variations. Beyond DE genes, there are thousands of genes that are stably expressed in various tissues or stages, regardless of environmental effects. Therefore, another possible further utilization of these datasets is to identify reference genes for qPCR. These reference genes will be then applied in gene expression normalization as suggested by (Zhao et al., 2017). In reverse, tissue-specific data can be mined to identify tissue- specific biomarkers, such as genes that are specifically expressed in one tissue and not in others. Likewise, proteomic data stored in the PRIDE database (https://www.ebi.ac.uk/pride/archive) can be extracted, reused, reanalysed or reprocessed in the context of a Bos indicus reference for new discoveries. The shared data can be also utilized by computational biologists in order to explore and develop bioinformatics tools for data analyses in this growing field of science.

The combining of transcriptome and proteome analyses presented herein revealed potential candidate genes and proteins involved in the puberty process. Experimental validations using methods such as qPCR or Western blots could provide additional confirmation for the DE genes and DA proteins; although close correlations between RNA-seq and qPCR are often reported (Griffith et al., 2010; Fang &Cui, 2011; Shi &He, 2014; Wu et al., 2014; Everaert et al., 2017). Therefore, functional validations and studies for the candidate genes are potentially more valuable than methodological confirmation of the DE/DA status.

Important TF identified by regulatory impact factor metric and involved in gene co- expression networks in this Thesis warrant additional studies for their interaction with puberty- related candidate genes. For example, follow-up studies of ZNF genes are needed to elucidate whether or not ZNF genes repress puberty-activating genes at pre-puberty heifers and examine their interactions in pubertal gene regulatory networks. Additionally, the CRISPR/cas9 can be used to create knock out and knock in animals or cell lines in order to investigate the function of candidate TF for puberty in cattle. The CRISPR/cas9 has been successfully applied to validate candidate genes for male infertility in knock out and knock in mice models (Kherraf et al., 2018). This technique also used to generate knock out mice in order to confirm a new repressor (miR-505-3p) of the onset of puberty in female mice (Tong et al., 2018). It is an approach with great potential to prove or disregard the hypothesis that emerged from this Thesis.

Another possible way to investigate the regulation of gene expression is obtained via the combined analyses between mRNA and other types of RNAs (microRNAs and long non- coding RNAs). In recent years, there are several studies supported the role of microRNAs and

158

long non-coding RNAs (lncRNAs) in regulation of gene expression during puberty (Elks et al., 2010; Sangiao-Alvarellos et al., 2013; Geach, 2016; Han et al., 2016; Messina et al., 2016; Gao et al., 2017; Gao et al., 2018). Adding these regulators into gene co-expression networks can provide a holistic view of gene expression dynamics. Any interesting findings would then require further functional validations at the phenotypic level that could be done by performing knockout experiments or screening in vitro interactions between candidate regulators and other factors (genes or transcription factors).

Biology involves the interaction between diverse levels, from genes, through transcripts and proteins, and then to metabolites. In real life, effects on one level impact on other levels creating a multi-factorial “cause” for observed phenotypes. Therefore, combining multiple omics approaches (transcriptome – proteome – metabolomics) aids in providing a better insight into relevant biological systems compared to studying biology by using a single approach. There were no previous studies combining transcriptome, proteome and metabolomics analyses in the context of puberty in Bos indicus.

Gene expression is also controlled by loci interaction. In other words, epistasis is an important factor controlling gene expression and should be not overlooked when studying complex traits like puberty (Carlborg &Haley, 2004). In human, the association between candidate genes and age at menarche were characterized using epistasis analyses (Pyun et al., 2014). In addition, the effect of SNP – SNP interactions between candidate genes (CYP19A1 and ESR1, FSHR and CYP19A1 as well as HSD17B4 and TG) on premature ovarian failure was reported (Kim et al., 2011a; Kim et al., 2011b; Pyun et al., 2012). With the rapid development of bioinformatics tools, one further important step of this work should be combining the QTL at the transcriptional level (eQTL), translational level (tQTL) and phenotypic QTL (pQTL) in order to enhance our knowledge on underpinning mechanisms that regulate puberty process.

Another application of multi-tissue omics analyses of pre- and post-puberty presented herein is to use DE genes/DA proteins as well as key regulators to enhance genome approaches for selective breeding. A study in Holstein dairy cow demonstrated that DE genes in endometrium and corpus luteum between low and high fertility cow were enriched for variants related to fertility (Moore et al., 2016). Therefore, the differential expression data identified when compared pre- and post-pubertal stage of biologically important representative tissues in this Thesis serves as the first material for discovering potential candidate mutations affecting puberty in beef cattle. The use of these strong candidate genes will help to decrease the number of single nucleotide polymorphism (SNP) tested and to provide potential causative mutations

159

into SNP arrays. SNPs identified within protein-coding regions as well as promoter regions of these candidate genes could be then tested for their association with puberty in male and female cattle. Including potential causative mutations might increase the accuracy of genomic selection for complex traits like puberty, and thus can accelerate the genetic improvement rate in cattle puberty (Snelling et al., 2013; Pérez-Enciso et al., 2015).

III.3. Conclusion

In conclusion, this Thesis provides a list of DE genes and DA proteins that involve in several biologically important pathways associated with puberty. Predictions about important TF and gene networks were also gained from this work. The large datasets created confirmed that puberty onset comprises coordinated activities of regulatory networks structured in functional modules. The combined results of transcriptomic and proteomic analyses presented in this study could be summarized as a list of potential regulators associated with female puberty: TUBB2B, OMG, SPTAN1, RAP1A, SSBP1, PLCB1, EIF3J, H2AFX, IGFBP2, TAGLN, CHGB, ENO1, HIST2H2AC, PHB2, TXNDC17, CYP2E1, FGB, ECHS1, EPHX1, FTCD, SURF1, C9, A1BG, C3, MGC137211, LRG1, CKM and a novel gene (ENSBTAG00000006864). Understanding the specific roles of these regulators will provide insights into our knowledge of cattle puberty. Further works focussing these potential regulators might help to discover pubertal-activating or -inhibiting candidate genes as well as potential causative mutations which can aid advantage for genome selection.

160

List of References

Abeygunawardena, H., & C. M. Dematawewa. (2004). Pre-pubertal and postpartum anestrus in tropical Zebu cattle. Anim Reprod Sci 82-83:373-387. Abrams, E. T., & E. M. Miller. (2011). The roles of the immune system in women's reproduction: evolutionary constraints and life history trade-offs. Am J Phys Anthropol 146 Suppl 53:134-154. Abreu, A. P., A. Dauber, D. B. Macedo, S. D. Noel, V. N. Brito, J. C. Gill, P. Cukier, I. R. Thompson, V. M. Navarro, P. C. Gagliardi et al. (2013). Central precocious puberty caused by mutations in the imprinted gene MKRN3. N Engl J Med 368(26):2467-2475. Acosta-Martínez, M. (2011). PI3K: An Attractive Candidate for the Central Integration of Metabolism and Reproduction. Frontiers in Endocrinology 2:110. Adam, C. L., T. S. Gadd, P. A. Findlay, & D. C. Wathes. (2000). IGF-I stimulation of luteinizing hormone secretion, IGF-binding proteins (IGFBPs) and expression of mRNAs for IGFs, IGF receptors and IGFBPs in the ovine pituitary gland. Journal of Endocrinology 166(2):247-254. Adam, C. L., C. E. Kyle, P. Young, & T. Atkinson. (1995). Effect of nutritional growth restriction on timing of reproductive development and plasma concentrations of insulin-like growth factor-1 and growth hormone in male red deer ( Cervus elaphus) reared in constant photoperiod. Animal Science 61(01):155-160. Aebersold, R., & M. Mann. (2003). Mass spectrometry-based proteomics. Nature 422(6928):198-207. Agarwal, A., S. Gupta, L. Sekhon, & R. Shah. (2008). Redox considerations in female reproductive function and assisted reproduction: from molecular mechanisms to health implications. Antioxid Redox Signal 10(8):1375-1403. Agarwal, A., S. Gupta, & R. K. Sharma. (2005). Role of oxidative stress in female reproduction. Reproductive biology and endocrinology : RB&E 3:28-28. Ahima, R. S., J. Dushay, S. N. Flier, D. Prabakaran, & J. S. Flier. (1997). Leptin accelerates the onset of puberty in normal female mice. The Journal of Clinical Investigation 99(3):391-395. Aittomaki, K., J. L. Lucena, P. Pakarinen, P. Sistonen, J. Tapanainen, J. Gromoll, R. Kaskikari, E. M. Sankila, H. Lehvaslaiho, A. R. Engel et al. (1995). Mutation in the follicle- stimulating hormone receptor gene causes hereditary hypergonadotropic ovarian failure. Cell 82(6):959-968. Akbar, H., M. Bionaz, D. B. Carlson, S. L. Rodriguez-Zas, R. E. Everts, H. A. Lewin, J. K. Drackley, & J. J. Loor. (2013). Feed restriction, but not l-carnitine infusion, alters the liver transcriptome by inhibiting sterol synthesis and mitochondrial oxidative phosphorylation and increasing gluconeogenesis in mid-lactation dairy cows. J Dairy Sci 96(4):2201-2213. Akers, R. M., S. E. Ellis, & S. D. Berry. (2005). Ovarian and IGF-I axis control of mammary development in prepubertal heifers. Domest Anim Endocrinol 29(2):259-267. Alexandre, P. A., L. J. A. Kogelman, M. H. A. Santana, D. Passarelli, L. H. Pulz, P. Fantinato- Neto, P. L. Silva, P. R. Leme, R. F. Strefezzi, L. L. Coutinho et al. (2015). Liver transcriptomic networks reveal main biological processes associated with feed efficiency in beef cattle. BMC Genomics 16(1):1073. Allan, M. F., L. A. Kuehn, R. A. Cushman, W. M. Snelling, S. E. Echternkamp, & R. M. Thallman. (2009). Confirmation of quantitative trait loci using a low-density single nucleotide polymorphism map for twinning and ovulation rate on bovine chromosome 5. J Anim Sci 87(1):46-56.

161

Almeida, A. M., A. Bassols, E. Bendixen, M. Bhide, F. Ceciliani, S. Cristobal, P. D. Eckersall, K. Hollung, F. Lisacek, G. Mazzucchelli et al. (2015). Animal board invited review: advances in proteomics for animal and food sciences. Animal 9(1):1-17. Alves, B. R., R. C. Cardoso, R. Doan, Y. Zhang, S. V. Dindot, G. L. Williams, & M. Amstalden. (2017). Nutritional programming of accelerated puberty in heifers: alterations in DNA methylation in the arcuate nucleus. Biol Reprod 96(1):174-184. Amar, D., H. Safer, & R. Shamir. (2013). Dissection of Regulatory Networks that Are Altered in Disease via Differential Co-expression. PLOS Computational Biology 9(3):e1002955. Amstalden, M., R. C. Cardoso, B. R. Alves, & G. L. Williams. (2014). Reproduction Symposium: hypothalamic neuropeptides and the nutritional programming of puberty in heifers. J Anim Sci 92(8):3211-3222. Amstalden, M., M. R. Garcia, S. W. Williams, R. L. Stanko, S. E. Nizielski, C. D. Morrison, D. H. Keisler, & G. L. Williams. (2000). Leptin Gene Expression, Circulating Leptin, and Luteinizing Hormone Pulsatility Are Acutely Responsive to Short-Term Fasting in Prepubertal Heifers: Relationships to Circulating Insulin and Insulin-Like Growth Factor I. Biology of Reproduction 63(1):127-133. Anahi Franchi, N., C. Avendano, R. I. Molina, A. D. Tissera, C. A. Maldonado, S. Oehninger, & C. E. Coronel. (2008). beta-Microseminoprotein in human spermatozoa and its potential role in male fertility. Reproduction 136(2):157-166. Anderson, L., & J. Seilhamer. (1997). A comparison of selected mRNA and protein abundances in human liver. Electrophoresis 18(3-4):533-537. Anderson, R. A., I. H. Zwain, A. Arroyo, P. L. Mellon, & S. S. Yen. (1999). The insulin-like growth factor system in the GT1-7 GnRH neuronal cell line. Neuroendocrinology 70(5):353-359. Andersson, L., A. L. Archibald, C. D. Bottema, R. Brauning, S. C. Burgess, D. W. Burt, E. Casas, H. H. Cheng, L. Clarke, C. Couldrey et al. (2015). Coordinated international action to accelerate genome-to-phenome with FAANG, the Functional Annotation of Animal Genomes project. Genome Biology 16 Arias, P., M. Rodriguez, B. Szwarcfarb, I. R. Sinay, & J. A. Moguilevsky. (1992). Effect of insulin on LHRH release by perifused hypothalamic fragments. Neuroendocrinology 56(3):415-418. Arije, G. F., & J. N. Wiltbank. (1971). Age and weight at puberty in Hereford heifers. J. Anim. Sci 33:401. Ashton, D., M. Oliver, & H. Valle. 2016. Australian Beef: Financial performance of beef farms, 2013-14 to 2015-16, Canberra. Atkins, J. A., K. G. Pohler, & M. F. Smith. (2013). Physiology and Endocrinology of Puberty in Heifers. Veterinary Clinics of North America: Food Animal Practice 29(3):479-492. Aziz, R., M. Beymer, A. L. Negron, A. Newshan, G. Yu, B. Rosati, D. McKinnon, M. Fukuda, R. Z. Lin, C. Mayer et al. (2014). Galanin-like peptide (GALP) neurone-specific phosphoinositide 3-kinase signalling regulates GALP mRNA levels in the hypothalamus of males and luteinising hormone levels in both sexes. J Neuroendocrinol 26(7):426-438. Backholer, K., M. Bowden, K. Gamber, C. Bjorbaek, J. Iqbal, & I. J. Clarke. (2010a). Melanocortins mimic the effects of leptin to restore reproductive function in lean hypogonadotropic ewes. Neuroendocrinology 91(1):27-40. Backholer, K., J. T. Smith, A. Rao, A. Pereira, J. Iqbal, S. Ogawa, Q. Li, & I. J. Clarke. (2010b). Kisspeptin cells in the ewe brain respond to leptin and communicate with neuropeptide Y and proopiomelanocortin cells. Endocrinology 151(5):2233-2243.

162

Baeker, R., S. Haebel, K. Schlatterer, & B. Schlatterer. (2002). Lipocalin-type prostaglandin D synthase in milk: a new biomarker for bovine mastitis. Prostaglandins Other Lipid Mediat 67(1):75-88. Bai, Y., S. Wang, H. Zhong, Q. Yang, F. Zhang, Z. Zhuang, J. Yuan, X. Nie, & S. Wang. (2015). Integrative analyses reveal transcriptome-proteome correlation in biological pathways and secondary metabolism clusters in A. flavus in response to temperature. Scientific Reports 5:14582. Baijal-Gupta, M., M. W. Clarke, M. A. Finkelman, C. M. McLachlin, & V. K. Han. (2000). Prostatic secretory protein (PSP94) expression in human female reproductive tissues, breast and in endometrial cancer cell lines. J Endocrinol 165(2):425-433. Bailey, U.-M., M. F. Jamaluddin, & B. L. Schulz. (2012). Analysis of Congenital Disorder of Glycosylation-Id in a Yeast Model System Shows Diverse Site-Specific Under- glycosylation of Glycoproteins. Journal of Proteome Research 11(11):5376-5383. Ballester, M., J. Molist, M. Lopez-Bejar, A. Sanchez, J. Santalo, J. M. Folch, & E. Ibanez. (2004). Disruption of the mouse phospholipase C-beta1 gene in a beta-lactoglobulin transgenic line affects viability, growth, and fertility in mice. Gene 341:279-289. Balthasar, N., R. Coppari, J. McMinn, S. M. Liu, C. E. Lee, V. Tang, C. D. Kenny, R. A. McGovern, S. C. Chua, J. K. Elmquist et al. (2004). Leptin Receptor Signaling in POMC Neurons Is Required for Normal Body Weight Homeostasis. Neuron 42(6):983- 991. Bangstad, H. J., K. Dahl-Jorgensen, P. Kjaersgaard, K. Mevold, & K. F. Hanssen. (1993). Urinary albumin excretion rate and puberty in non-diabetic children and adolescents. Acta Paediatr 82(10):857-862. Barash, I. A., C. C. Cheung, D. S. Weigle, H. Ren, E. B. Kabigting, J. L. Kuijper, D. K. Clifton, & R. A. Steiner. (1996). Leptin is a metabolic signal to the reproductive system. Endocrinology 137(7):3144-3147. Barb, C. R., G. J. Hausman, & K. L. Houseknecht. (2001). Biology of leptin in the pig. Domest Anim Endocrinol 21(4):297-317. Barb, C. R., R. R. Kraeling, G. B. Rampacek, & G. J. Hausman. (2006). The role of neuropeptide Y and interaction with leptin in regulating feed intake and luteinizing hormone and growth hormone secretion in the pig. Reproduction 131(6):1127-1135. Barker-Gibb, M. L., C. J. Scott, J. H. Boublik, & I. J. Clarke. (1995). The role of neuropeptide Y (NPY) in the control of LH secretion in the ewe with respect to season, NPY receptor subtype and the site of action in the hypothalamus. J Endocrinol 147(3):565-579. Baspinar, A., E. Cukuroglu, R. Nussinov, O. Keskin, & A. Gursoy. (2014). PRISM: a web server and repository for prediction of protein-protein interactions and modeling their 3D complexes. Nucleic Acids Res 42(Web Server issue):W285-289. Baufeld, A., D. Koczan, & J. Vanselow. (2017). Induction of altered gene expression profiles in cultured bovine granulosa cells at high cell density. Reproductive Biology and Endocrinology : RB&E 15:3. Beddek, A. J., P. Rawson, L. Peng, R. Snell, K. Lehnert, H. E. Ward, & T. W. Jordan. (2008). Profiling the metabolic proteome of bovine mammary tissue. Proteomics 8(7):1502- 1515. Bedecarrats, G. Y., & U. B. Kaiser. (2007). Mutations in the human gonadotropin-releasing hormone receptor: insights into receptor biology and function. Semin Reprod Med 25(5):368-378. Benedusi, V., S. Della Torre, N. Mitro, D. Caruso, A. Oberto, C. Tronel, C. Meda, & A. Maggi. (2017). Liver ERalpha regulates AgRP neuronal activity in the arcuate nucleus of female mice. Sci Rep 7(1):1194.

163

Bennett, C. N., S. E. Ross, K. A. Longo, L. Bajnok, N. Hemati, K. W. Johnson, S. D. Harrison, & O. A. MacDougald. (2002). Regulation of Wnt signaling during adipogenesis. J Biol Chem 277(34):30998-31004. Bennett, E., K. L. Stenvers, P. K. Lund, & B. Popko. (1994). Cloning and characterization of a cDNA encoding a novel fatty acid binding protein from rat brain. J Neurochem 63(5):1616-1624. Bentley, P. A., E. H. Wall, G. E. Dahl, & T. B. McFadden. (2015). Responses of the mammary transcriptome of dairy cows to altered photoperiod during late gestation. Physiological Genomics 47(10):488-499. Berg, A. H., T. P. Combs, X. Du, M. Brownlee, & P. E. Scherer. (2001). The adipocyte-secreted protein Acrp30 enhances hepatic insulin action. Nat Med 7(8):947-953. Berg, J. M., J. L. Tymoczko, & L. Stryer. 2002. Oxidative Phosphorylation. In: W. H. Freeman, editor, Biochemistry, New York. Berridge, M. J., P. Lipp, & M. D. Bootman. (2000). The versatility and universality of calcium signalling. Nat Rev Mol Cell Biol 1(1):11-21. Bessho, Y., G. Miyoshi, R. Sakata, & R. Kageyama. (2001). Hes7: a bHLH-type repressor gene regulated by Notch and expressed in the presomitic mesoderm. Genes Cells 6(2):175- 185. Beymer, M., A. L. Negron, G. Yu, S. Wu, C. Mayer, R. Z. Lin, U. Boehm, & M. Acosta- Martinez. (2014). Kisspeptin cell-specific PI3K signaling regulates hypothalamic kisspeptin expression and participates in the regulation of female fertility. Am J Physiol Endocrinol Metab 307(11):E969-982. Bi, P., & S. Kuang. (2015). Notch signaling as a novel regulator of metabolism. Trends Endocrinol Metab 26(5):248-255. Bi, P., T. Shan, W. Liu, F. Yue, X. Yang, X.-R. Liang, J. Wang, J. Li, N. Carlesso, X. Liu et al. (2014). Notch signaling regulates adipose browning and energy metabolism. Nature medicine 20(8):911-918. Bi, P., F. Yue, A. Karki, B. Castro, S. E. Wirbisky, C. Wang, A. Durkes, B. D. Elzey, O. M. Andrisani, C. A. Bidwell et al. (2016). Notch activation drives adipocyte dedifferentiation and tumorigenic transformation in mice. J Exp Med 213(10):2019- 2037. Bilezikjian, L. M., A. L. Blount, C. J. Donaldson, & W. W. Vale. (2006). Pituitary actions of ligands of the TGF-beta family: activins and inhibins. Reproduction 132(2):207-215. Binder, D. K., & H. E. Scharfman. (2004). Brain-derived Neurotrophic Factor. Growth factors (Chur, Switzerland) 22(3):123-131. Bitton, D. A., M. J. Okoniewski, Y. Connolly, & C. J. Miller. (2008). Exon level integration of proteomics and microarray data. BMC Bioinformatics 9:118. Bliss, S. P., A. M. Navratil, J. Xie, & M. S. Roberson. (2010). GnRH signaling, the gonadotrope and endocrine control of fertility. Front Neuroendocrinol 31(3):322-340. Boehmer, J. L. (2011). Proteomic Analyses of Host and Pathogen Responses during Bovine Mastitis. Journal of Mammary Gland Biology and Neoplasia 16(4):323-338. Boehmer, J. L., J. A. DeGrasse, M. A. McFarland, E. A. Tall, K. J. Shefcheck, J. L. Ward, & D. D. Bannerman. (2010a). The proteomic advantage: label-free quantification of proteins expressed in bovine milk during experimentally induced coliform mastitis. Vet Immunol Immunopathol 138(4):252-266. Boehmer, J. L., J. L. Ward, R. R. Peters, K. J. Shefcheck, M. A. McFarland, & D. D. Bannerman. (2010b). Proteomic analysis of the temporal expression of bovine milk proteins during coliform mastitis and label-free relative quantification. J Dairy Sci 93(2):593-603.

164

Bohm, D., H. Schwegler, L. Kotthaus, K. Nayernia, M. Rickmann, M. Kohler, J. Rosenbusch, W. Engel, G. Flugge, & P. Burfeind. (2002). Disruption of PLC-beta 1-mediated signal transduction in mutant mice causes age-dependent hippocampal mossy fiber sprouting and neurodegeneration. Mol Cell Neurosci 21(4):584-601. Bond, J. F., G. S. Robinson, & S. R. Farmer. (1984). Differential expression of two neural cell- specific beta-tubulin mRNAs during rat brain development. Mol Cell Biol 4(7):1313- 1319. Bönnemann, C. G. (2011). The collagen VI-related myopathies: Ullrich congenital muscular dystrophy and Bethlem myopathy. Handbook of clinical neurology 101:81-96. Bortolussi, G., J. G. McIvor, J. J. Hodgkinson, S. G. Coffey, & C. R. Holmes. (2005). The northern Australian beef industry, a snapshot. 2. Breeding herd performance and management. Australian Journal of Experimental Agriculture 45(9):1075-1091. Bottner, A., J. Kratzsch, G. Muller, T. M. Kapellen, S. Bluher, E. Keller, M. Bluher, & W. Kiess. (2004). Gender differences of adiponectin levels develop during the progression of puberty and are related to serum androgen levels. J Clin Endocrinol Metab 89(8):4053-4061. Boucher, C. A., C. L. Winchester, G. M. Hamilton, A. D. Winter, K. J. Johnson, & M. E. S. Bailey. (2000). Structure, mapping and expression of the human gene encoding the homeodomain protein, SIX2. Gene 247(1-2):145-151. Boukhliq, R., & G. B. Martin. (1997). Administration of fatty acids and gonadotrophin secretion in the mature ram. Anim Reprod Sci 49(2-3):143-159. Brasse-Lagnel, C., A. Lavoinne, & A. Husson. (2009). Control of mammalian gene expression by amino acids, especially glutamine. Febs j 276(7):1826-1844. Brickell, J. S. B., N. Cheng, Z. Wathes, D.C. . (2007). Influence of plasma IGF-1 concentration and body weight at 6 months on age at first calving in dairy heifers on commercial farms. Reprod Domest Anim 42:118. Brini, M., T. Calì, D. Ottolini, & E. Carafoli. (2014). Neuronal calcium signaling: function and dysfunction. Cellular and Molecular Life Sciences 71(15):2787-2814. Brito, L. F., A. E. Silva, M. M. Unanian, M. A. Dode, R. T. Barbosa, & J. P. Kastelic. (2004). Sexual development in early- and late-maturing Bos indicus and Bos indicus x Bos taurus crossbred bulls in Brazil. Theriogenology 62(7):1198-1217. Bruning, J. C., D. Gautam, D. J. Burks, J. Gillette, M. Schubert, P. C. Orban, R. Klein, W. Krone, D. Muller-Wieland, & C. R. Kahn. (2000). Role of brain insulin receptor in control of body weight and reproduction. Science 289(5487):2122-2125. Bullard, J. H., E. Purdom, K. D. Hansen, & S. Dudoit. (2010). Evaluation of statistical methods for normalization and differential expression in mRNA-Seq experiments. BMC Bioinformatics 11(1):94. Burnichon, N., J. J. Briere, R. Libe, L. Vescovo, J. Riviere, F. Tissier, E. Jouanno, X. Jeunemaitre, P. Benit, A. Tzagoloff et al. (2010). SDHA is a tumor suppressor gene causing paraganglioma. Hum Mol Genet 19(15):3011-3020. Cairoli, F., M. Battocchio, M. C. Veronesi, D. Brambilla, F. Conserva, I. Eberini, R. Wait, & E. Gianazza. (2006). Serum protein pattern during cow pregnancy: Acute-phase proteins increase in the peripartum period. Electrophoresis 27(8):1617-1625. Calcaterra, V., M. De Amici, C. Klersy, C. Torre, V. Brizzi, F. Scaglia, M. Albanesi, R. Albertini, B. Allais, & D. Larizza. (2009). Adiponectin, IL-10 and metabolic syndrome in obese children and adolescents. Acta Biomed 80(2):117-123. Callaerts, P., G. Halder, & W. J. Gehring. (1997). PAX-6 in development and evolution. Annu Rev Neurosci 20:483-532. Cammack, K. M., M. G. Thomas, & R. M. Enns. (2009). Review: Reproductive Traits and Their Heritabilities in Beef Cattle. The Professional Animal Scientist 25(5):517-528.

165

Cánovas, A., A. Reverter, K. L. DeAtley, R. L. Ashley, M. L. Colgrave, M. R. S. Fortes, A. Islas-Trejo, S. Lehnert, L. Porto-Neto, G. Rincón et al. (2014a). Multi-Tissue Omics Analyses Reveal Molecular Regulatory Networks for Puberty in Composite Beef Cattle. PLOS ONE 9(7):e102551. Cánovas, A., G. Rincon, C. Bevilacqua, A. Islas-Trejo, P. Brenaut, R. C. Hovey, M. Boutinaud, C. Morgenthaler, M. K. VanKlompenberg, P. Martin et al. (2014b). Comparison of five different RNA sources to examine the lactating bovine mammary gland transcriptome using RNA-Sequencing. Sci Rep 4:5297. Cánovas, A., G. Rincon, A. Islas-Trejo, R. Jimenez-Flores, A. Laubscher, & J. F. Medrano. (2013). RNA sequencing to study gene expression and single nucleotide polymorphism variation associated with citrate content in cow milk. J Dairy Sci 96(4):2637-2648. Cánovas, A., G. Rincon, A. Islas-Trejo, S. Wickramasinghe, & J. F. Medrano. (2010). SNP discovery in the bovine milk transcriptome using RNA-Seq technology. Mamm Genome 21(11-12):592-598. Cao, G. L., T. Feng, M. X. Chu, R. Di, Y. L. Zhang, D. W. Huang, Q. Y. Liu, W. P. Hu, & X. Y. Wang. (2015). Subtraction suppressive hybridisation analysis of differentially expressed genes associated with puberty in the goat hypothalamus. Reprod Fertil Dev Capra, E., F. Turri, B. Lazzari, P. Cremonesi, T. M. Gliozzi, I. Fojadelli, A. Stella, & F. Pizzi. (2017). Small RNA sequencing of cryopreserved semen from single bull revealed altered miRNAs and piRNAs expression between High- and Low-motile sperm populations. BMC Genomics 18(1):14. Caraty, A., & D. C. Skinner. (1999). Progesterone priming is essential for the full expression of the positive feedback effect of estradiol in inducing the preovulatory gonadotropin- releasing hormone surge in the ewe. Endocrinology 140(1):165-170. Cardoso, R. C., B. R. Alves, L. D. Prezotto, J. F. Thorson, L. O. Tedeschi, D. H. Keisler, C. S. Park, M. Amstalden, & G. L. Williams. (2014a). Use of a stair-step compensatory gain nutritional regimen to program the onset of puberty in beef heifers. J Anim Sci 92(7):2942-2949. Cardoso, R. C., B. R. Alves, S. M. Sharpton, G. L. Williams, & M. Amstalden. (2015). Nutritional Programming of Accelerated Puberty in Heifers: Involvement of Pro- Opiomelanocortin Neurones in the Arcuate Nucleus. J Neuroendocrinol 27(8):647-657. Cardoso, R. C., B. R. C. Alves, L. D. Prezotto, J. F. Thorson, L. O. Tedeschi, D. H. Keisler, M. Amstalden, & G. L. Williams. (2014b). Reciprocal changes in leptin and NPY during nutritional acceleration of puberty in heifers. Journal of Endocrinology 223(3):289-298. Care, A. S., K. R. Diener, M. J. Jasper, H. M. Brown, W. V. Ingman, & S. A. Robertson. (2013). Macrophages regulate corpus luteum development during embryo implantation in mice. J Clin Invest 123(8):3472-3487. Carlborg, Ö., & C. S. Haley. (2004). Epistasis: too often neglected in complex trait studies? Nature Reviews Genetics 5:618. Carretero, J., G. Vazquez, M. Rubio, E. Blanco, J. A. Juanes, E. Perez, D. Burks, & R. Vazquez. (2003). Postnatal differentiation of the immunohistochemical expression of aromatase P450 in the rat pituitary gland. Histol Histopathol 18(2):419-423. Carvalhais, V., A. França, G. B. Pier, M. Vilanova, N. Cerca, & R. Vitorino. (2015). Comparative proteomic and transcriptomic profile of Staphylococcus epidermidis biofilms grown in glucose-enriched medium. Talanta 132:705-712. Castellano, J. M., V. M. Navarro, R. Fernandez-Fernandez, R. Nogueiras, S. Tovar, J. Roa, M. J. Vazquez, E. Vigo, F. F. Casanueva, E. Aguilar et al. (2005). Changes in hypothalamic KiSS-1 system and restoration of pubertal activation of the reproductive axis by kisspeptin in undernutrition. Endocrinology 146(9):3917-3925.

166

Catunda, A. G. V., I. C. S. Lima, G. C. Bandeira, C. R. F. Gadelha, E. S. Pereira, C. S. B. Salmito-Vanderley, A. A. Araújo, G. A. Martins, & A. C. N. Campos. (2013). Blood leptin, insulin and glucose concentrations in hair sheep raised in a tropical climate. Small Ruminant Research 114(2):272-279. Ceciliani, F., C. Lecchi, J. Bazile, & M. Bonnet. 2018. Proteomics Research in the Adipose Tissue. In: A. M. de Almeida, D. Eckersall and I. Miller, editors, Proteomics in Domestic Animals: from Farm to Systems Biology. Springer International Publishing, Cham. p. 233-254. Chartoumpekis, D. V., D. L. Palliyaguru, N. Wakabayashi, N. K. Khoo, G. Schoiswohl, R. M. O'Doherty, & T. W. Kensler. (2015). Notch intracellular domain overexpression in adipocytes confers lipodystrophy in mice. Mol Metab 4(7):543-550. Chatr-Aryamontri, A., B. J. Breitkreutz, S. Heinicke, L. Boucher, A. Winter, C. Stark, J. Nixon, L. Ramage, N. Kolas, L. O'Donnell et al. (2013). The BioGRID interaction database: 2013 update. Nucleic Acids Res 41(Database issue):D816-823. Chehab, F. F., K. Mounzih, R. Lu, & M. E. Lim. (1997). Early onset of reproductive function in normal female mice treated with leptin. Science 275(5296):88-90. Chelikani, P. K., D. R. Glimm, & J. J. Kennelly. (2003). Short Communication: Tissue Distribution of Leptin and Leptin Receptor mRNA in the Bovine. Journal of Dairy Science 86(7):2369-2372. Chen, C. T., L. Fernandez-Rhodes, R. G. Brzyski, C. S. Carlson, Z. Chen, G. Heiss, K. E. North, N. F. Woods, A. Rajkovic, C. Kooperberg et al. (2012). Replication of loci influencing ages at menarche and menopause in Hispanic women: the Women's Health Initiative SHARe Study. Hum Mol Genet 21(6):1419-1432. Chen, G., T. G. Gharib, C.-C. Huang, J. M. G. Taylor, D. E. Misek, S. L. R. Kardia, T. J. Giordano, M. D. Iannettoni, M. B. Orringer, S. M. Hanash et al. (2002). Discordant Protein and mRNA Expression in Lung Adenocarcinomas. Molecular & Cellular Proteomics 1(4):304-313. Chen, H.-Z., S.-Y. Tsai, & G. Leone. (2009a). Emerging roles of E2Fs in cancer: an exit from cell cycle control. Nature reviews: Cancer 9(11):785-797. Chen, L. R., S. C. Lee, Y. P. Lin, Y. L. Hsieh, Y. L. Chen, J. R. Yang, J. F. Liou, C. F. Chen, Y. P. Lee, & Y. L. Shiue. (2010). Prostaglandin-D synthetase induces transcription of the LH beta subunit in the primary culture of chicken anterior pituitary cells via the PPAR signaling pathway. Theriogenology 73(3):367-382. Chen, Y., G. Mehta, & V. Vasiliou. (2009b). Antioxidant defenses in the ocular surface. Ocul Surf 7(4):176-185. Cho, J. H., J. Y. Jeong, R. H. Lee, M. N. Park, S.-H. Kim, S.-M. Park, J.-C. Shin, Y.-J. Jeon, J.-H. Shim, N.-J. Choi et al. (2016). Regional Differences of Proteins Expressing in Adipose Depots Isolated from Cows, Steers and Bulls as Identified by a Proteomic Approach. Asian-Australasian Journal of Animal Sciences 29(8):1197-1206. Choi, M., C. Y. Chang, T. Clough, D. Broudy, T. Killeen, T. MacLean, & O. Vitek. (2014). MSstats: an R package for statistical analysis of quantitative mass spectrometry-based proteomic experiments. Bioinformatics 30(17):2524-2526. Christensen, J., P. Cloos, U. Toftegaard, D. Klinkenberg, A. P. Bracken, E. Trinh, M. Heeran, L. Di Stefano, & K. Helin. (2005). Characterization of E2F8, a novel E2F-like cell- cycle regulated repressor of E2F-activated transcription. Nucleic Acids Research 33(17):5458-5470. Chu, Y., & D. R. Corey. (2012). RNA Sequencing: Platform Selection, Experimental Design, and Data Interpretation. Nucleic Acid Therapeutics 22(4):271-274. Chun, S. K., & Y.-H. Jo. (2010). Loss of Leptin Receptors on Hypothalamic POMC Neurons Alters Synaptic Inhibition. Journal of Neurophysiology 104(5):2321-2328.

167

Clapper, J. A., J. L. Snyder, A. J. Roberts, D. L. Hamernik, & G. E. Moss. (1998). Estradiol increases relative amounts of insulin-like growth factor binding protein (IGFBP)-3 in serum and expression of IGFBP-2 in anterior pituitaries of ewes. Biol Reprod 59(1):124-130. Clarke, I. J., & J. T. Cummins. (1984). Direct pituitary effects of estrogen and progesterone on gonadotropin secretion in the ovariectomized ewe. Neuroendocrinology 39(3):267-274. Clarkson, J., & A. E. Herbison. (2009). Oestrogen, kisspeptin, GPR54 and the pre-ovulatory luteinising hormone surge. J Neuroendocrinol 21(4):305-311. Cole, T. J., S. I. Salem, A. S. Hafez, O. M. Galal, & A. Massoud. (1982). Plasma albumin, parasitic infection and pubertal development in Egyptian boys. Trans R Soc Trop Med Hyg 76(1):17-20. Combs, T. P., A. H. Berg, S. Obici, P. E. Scherer, & L. Rossetti. (2001). Endogenous glucose production is inhibited by the adipose-derived protein Acrp30. J Clin Invest 108(12):1875-1881. Conesa, A., P. Madrigal, S. Tarazona, D. Gomez-Cabrero, A. Cervera, A. McPherson, M. W. Szcześniak, D. J. Gaffney, L. L. Elo, X. Zhang et al. (2016). A survey of best practices for RNA-seq data analysis. Genome Biology 17(1):13. Coope, A., M. Milanski, E. P. Araujo, M. Tambascia, M. J. Saad, B. Geloneze, & L. A. Velloso. (2008). AdipoR1 mediates the anorexigenic and insulin/leptin-like actions of adiponectin in the hypothalamus. FEBS Lett 582(10):1471-1476. Corah, L. R., T. G. Dunn, & C. C. Kaltenbach. (1975). Influence of prepartum nutrition on the reproductive performance of beef females and the performance of their progeny. J Anim Sci 41(3):819-824. Cotter, R. J. (1992). Time-of-flight mass spectrometry for the structural analysis of biological molecules. Analytical chemistry 64(21):1027A-1039A. Cousminer, D. L., D. J. Berry, N. J. Timpson, W. Ang, E. Thiering, E. M. Byrne, H. R. Taal, V. Huikari, J. P. Bradfield, M. Kerkhof et al. (2013). Genome-wide association and longitudinal analyses reveal genetic loci linking pubertal height growth, pubertal timing and childhood adiposity. Hum Mol Genet 22(13):2735-2747. Crawford, J., & A. McNeilly. (2002). Co-localisation of gonadotrophins and granins in gonadotrophs at different stages of the oestrous cycle in sheep. Journal of Endocrinology 174(2):179-194. Curlewis, J. D., M. Axelson, & G. M. Stone. (1985). Identification of the major steroids in ovarian and adrenal venous plasma of the brush-tail possum (Trichosurus vulpecula) and changes in the peripheral plasma levels of oestradiol and progesterone during the reproductive cycle. J Endocrinol 105(1):53-62. Czaja, K., M. Lakomy, W. Sienkiewicz, J. Kaleczyc, Z. Pidsudko, C. R. Barb, G. B. Rampacek, & R. R. Kraeling. (2002). Distribution of neurons containing leptin receptors in the hypothalamus of the pig. Biochem Biophys Res Commun 298(3):333-337. D'Ambrosio, C., S. Arena, F. Talamo, L. Ledda, G. Renzone, L. Ferrara, & A. Scaloni. (2005). Comparative proteomic analysis of mammalian animal tissues and body fluids: bovine proteome database. J Chromatogr B Analyt Technol Biomed Life Sci 815(1-2):157- 168. d'Anglemont de Tassigny, X., & W. H. Colledge. (2010). The role of kisspeptin signaling in reproduction. Physiology (Bethesda) 25(4):207-217. Daftary, S. S., & A. C. Gore. (2005). IGF-1 in the brain as a regulator of reproductive neuroendocrine function. Experimental Biology and Medicine 230(5):292-306. Dahlanuddin, B. S. Ningsih, D. P. Poppi, S. T. Anderson, & S. P. Quigley. (2014). Long-term growth of male and female Bali cattle fed Sesbania grandiflora. Animal Production Science 54(10):1615-1619.

168

Dai, W., Q. Chen, Q. Wang, R. R. White, J. Liu, & H. Liu. (2017). Complementary transcriptomic and proteomic analyses reveal regulatory mechanisms of milk protein production in dairy cows consuming different forages. Scientific Reports 7:44234. Daniels, K. M., K. E. Webb, M. L. McGilliard, M. J. Meyer, M. E. Van Amburgh, & R. M. Akers. (2006). Effects of Body Weight and Nutrition on Mammary Protein Expression Profiles in Holstein Heifers. Journal of Dairy Science 89(11):4276-4288. Danielsen, M., M. C. Codrea, K. L. Ingvartsen, N. C. Friggens, E. Bendixen, & C. M. Rontved. (2010). Quantitative milk proteomics--host responses to lipopolysaccharide-mediated inflammation of bovine mammary gland. Proteomics 10(12):2240-2249. David J, H., S. Jennifer A, S. Carolyn D, & B. Robert C. (1987). Hormonal Regulation of the Peripubertal Surge of Insulin-Like Growth Factor-I in the Rat. Endocrinology 120(2):491-496. Davis, M. E., & R. C. Simmen. (1997). Genetic parameter estimates for serum insulin-like growth factor I concentration and performance traits in Angus beef cattle. J Anim Sci 75(2):317-324. Davis, M. E., & R. C. Simmen. (2006). Genetic parameter estimates for serum insulin-like growth factor I concentrations, and body weight and weight gains in Angus beef cattle divergently selected for serum insulin-like growth factor I concentration. J Anim Sci 84(9):2299-2308. Day, M. L., Anderson, L. H. (1998). Current concepts on the control of puberty in cattle. Journal of Animal Science 76 (Suppl. 3):1-15. Day, M. L., K. Imakawa, M. Garcia-Winder, D. D. Zalesky, B. D. Schanbacher, R. J. Kittok, & J. E. Kinder. (1984). Endocrine mechanisms of puberty in heifers: estradiol negative feedback regulation of luteinizing hormone secretion. Biol Reprod 31(2):332-341. Day, M. L., K. Imakawa, P. L. Wolfe, R. J. Kittok, & J. E. Kinder. (1987). Endocrine mechanisms of puberty in heifers. Role of hypothalamo-pituitary estradiol receptors in the negative feedback of estradiol on luteinizing hormone secretion. Biol Reprod 37(5):1054-1065. Day, M. L., K. Imakawa, D. D. Zalesky, R. J. Kittok, & J. E. Kinder. (1986). Effects of restriction of dietary energy intake during the prepubertal period on secretion of luteinizing hormone and responsiveness of the pituitary to luteinizing hormone- releasing hormone in heifers. J Anim Sci 62(6):1641-1648. de Roux, N., E. Genin, J. C. Carel, F. Matsuda, J. L. Chaussain, & E. Milgrom. (2003). Hypogonadotropic hypogonadism due to loss of function of the KiSS1-derived peptide receptor GPR54. Proc Natl Acad Sci U S A 100(19):10972-10976. de Sousa Abreu, R., L. O. Penalva, E. M. Marcotte, & C. Vogel. (2009). Global signatures of protein and mRNA expression levels. Mol Biosyst 5(12):1512-1526. DeAtley, K. L., M. L. Colgrave, A. Cánovas, G. Wijffels, R. L. Ashley, G. A. Silver, G. Rincon, J. F. Medrano, A. Islas-Trejo, M. R. S. Fortes et al. (2018). Neuropeptidome of the Hypothalamus and Pituitary Gland of Indicine × Taurine Heifers: Evidence of Differential Neuropeptide Processing in the Pituitary Gland before and after Puberty. Journal of Proteome Research 17(5):1852-1865. DeFazio, R. A., S. Heger, S. R. Ojeda, & S. M. Moenter. (2002). Activation of A-type gamma- aminobutyric acid receptors excites gonadotropin-releasing hormone neurons. Mol Endocrinol 16(12):2872-2891. Della Torre, S., G. Rando, C. Meda, A. Stell, P. Chambon, A. Krust, C. Ibarra, P. Magni, P. Ciana, & A. Maggi. (2011). Amino acid-dependent activation of liver estrogen receptor alpha integrates metabolic and reproductive functions via IGF-1. Cell Metab 13(2):205- 214.

169

Demerath, E. W., C. T. Liu, N. Franceschini, G. Chen, J. R. Palmer, E. N. Smith, C. T. Chen, C. B. Ambrosone, A. M. Arnold, E. V. Bandera et al. (2013). Genome-wide association study of age at menarche in African-American women. Hum Mol Genet 22(16):3329- 3346. Dennis, G., B. T. Sherman, D. A. Hosack, J. Yang, W. Gao, H. C. Lane, & R. A. Lempicki. (2003). DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biology 4(9):R60. Desjardins, C., & H. D. Hafs. (1968). Levels of pituitary FSH and LH in heifers from birth through puberty. J Anim Sci 27(2):472-477. Dias, M. M., F. R. Souza, L. Takada, F. L. Feitosa, R. B. Costa, I. D. Diaz, D. F. Cardoso, R. L. Tonussi, F. Baldi, L. G. Albuquerque et al. (2015). Study of lipid metabolism-related genes as candidate genes of sexual precocity in Nellore cattle. Genet Mol Res 14(1):234-243. Ding, F., H. H. Li, S. Zhang, N. M. Solomon, S. A. Camper, P. Cohen, & U. Francke. (2008). SnoRNA Snord116 (Pwcr1/MBII-85) Deletion Causes Growth Deficiency and Hyperphagia in Mice. PLOS ONE 3(3):e1709. Docke, F., & G. Dorner. (1965). The mechanism of the induction of ovulation by oestrogens. J Endocrinol 33(3):491-499. Doherty, M. K., L. McLean, J. R. Hayter, J. M. Pratt, D. H. Robertson, A. El-Shafei, S. J. Gaskell, & R. J. Beynon. (2004). The proteome of chicken skeletal muscle: changes in soluble protein expression during growth in a layer strain. Proteomics 4(7):2082-2093. Donato, J., Jr., R. M. Cravo, R. Frazao, L. Gautron, M. M. Scott, J. Lachey, I. A. Castro, L. O. Margatho, S. Lee, C. Lee et al. (2011). Leptin's effect on puberty in mice is relayed by the ventral premammillary nucleus and does not require signaling in Kiss1 neurons. J Clin Invest 121(1):355-368. Dong, E., K. Matsumoto, & H. Watanabe. (2002). Diazepam binding inhibitor (DBI) reduces testosterone and estradiol levels in vivo. Life Sci 70(11):1317-1323. Draper, N., E. A. Walker, I. J. Bujalska, J. W. Tomlinson, S. M. Chalder, W. Arlt, G. G. Lavery, O. Bedendo, D. W. Ray, I. Laing et al. (2003). Mutations in the genes encoding 11beta- hydroxysteroid dehydrogenase type 1 and hexose-6-phosphate dehydrogenase interact to cause cortisone reductase deficiency. Nat Genet 34(4):434-439. Ducret, E., G. M. Anderson, & A. E. Herbison. (2009). RFamide-related peptide-3, a mammalian gonadotropin-inhibitory hormone ortholog, regulates gonadotropin- releasing hormone neuron firing in the mouse. Endocrinology 150(6):2799-2804. Dudas, B., & I. Merchenthaler. (2006). Three-dimensional representation of the neurotransmitter systems of the human hypothalamus: inputs of the gonadotrophin hormone-releasing hormone neuronal system. J Neuroendocrinol 18(2):79-95. Dyer, C. J., J. M. Simmons, R. L. Matteri, & D. H. Keisler. (1997). Leptin receptor mRNA is expressed in ewe anterior pituitary and adipose tissues and is differentially expressed in hypothalamic regions of well-fed and feed-restricted ewes. Domestic animal endocrinology 14(2):119-128. Eastell, R. (2005). Role of oestrogen in the regulation of bone turnover at the menarche. J Endocrinol 185(2):223-234. Ebling, F. J., & S. M. Luckman. (2008). RFAmide-related peptide: another sexy peptide? Endocrinology 149(3):899-901. Echternkamp, S. E., A. J. Roberts, D. D. Lunstra, T. Wise, & L. J. Spicer. (2004). Ovarian follicular development in cattle selected for twin ovulations and births. J Anim Sci 82(2):459-471.

170

Eden, E., N. Geva-Zatorsky, I. Issaeva, A. Cohen, E. Dekel, T. Danon, L. Cohen, A. Mayo, & U. Alon. (2011). Proteome half-life dynamics in living human cells. Science 331(6018):764-768. Edfors, F., F. Danielsson, B. M. Hallström, L. Käll, E. Lundberg, F. Pontén, B. Forsström, & M. Uhlén. (2016). Gene‐specific correlation of RNA and protein levels in human cells and tissues. Molecular Systems Biology 12(10):883. Egan, O. K., M. A. Inglis, & G. M. Anderson. (2017). Leptin Signaling in AgRP Neurons Modulates Puberty Onset and Adult Fertility in Mice. J Neurosci 37(14):3875-3886. Elks, C. E., J. R. Perry, P. Sulem, D. I. Chasman, N. Franceschini, C. He, K. L. Lunetta, J. A. Visser, E. M. Byrne, D. L. Cousminer et al. (2010). Thirty new loci for age at menarche identified by a meta-analysis of genome-wide association studies. Nat Genet 42(12):1077-1085. Elsik, C. G., R. L. Tellam, & K. C. Worley. (2009). The Genome Sequence of Taurine Cattle: A Window to Ruminant Biology and Evolution. Science 324(5926):522-528. Elustondo, P. A., G. E. Hannigan, I. Caniggia, & D. J. MacPhee. (2006). Integrin-linked kinase (ILK) is highly expressed in first trimester human chorionic villi and regulates migration of a human cytotrophoblast-derived cell line. Biol Reprod 74(5):959-968. Erickson, J. C., G. Hollopeter, & R. D. Palmiter. (1996). Attenuation of the Obesity Syndrome of ob/ob Mice by the Loss of Neuropeptide Y. Science 274(5293):1704-1707. Etgen, A. M., & M. Acosta-Martinez. (2003). Participation of growth factor signal transduction pathways in estradiol facilitation of female reproductive behavior. Endocrinology 144(9):3828-3835. Evans, A. C., G. P. Adams, & N. C. Rawlings. (1994). Endocrine and ovarian follicular changes leading up to the first ovulation in prepubertal heifers. J Reprod Fertil 100(1):187-194. Evans, N. P., G. E. Dahl, V. Padmanabhan, L. A. Thrun, & F. J. Karsch. (1997). Estradiol requirements for induction and maintenance of the gonadotropin-releasing hormone surge: implications for neuroendocrine processing of the estradiol signal. Endocrinology 138(12):5408-5414. Everaert, C., M. Luypaert, J. L. V. Maag, Q. X. Cheng, M. E. Dinger, J. Hellemans, & P. Mestdagh. (2017). Benchmarking of RNA-sequencing analysis workflows using whole-transcriptome RT-qPCR expression data. Scientific Reports 7:1559. Everett, J. W. (1948). Progesterone and estrogen in the experimental control of ovulation time and other features of the estrous cycle in the rat. Endocrinology 43(6):389-405. Fang, Z., & X. Cui. (2011). Design and validation issues in RNA-seq experiments. Briefings in bioinformatics 12(3):280-287. Farah, M. H., J. M. Olson, H. B. Sucic, R. I. Hume, S. J. Tapscott, & D. L. Turner. (2000). Generation of neurons by transient expression of neural bHLH proteins in mammalian cells. Development 127(4):693-702. Faruqi, T. R., D. Gomez, X. R. Bustelo, D. Bar-Sagi, & N. C. Reich. (2001). Rac1 mediates STAT3 activation by autocrine IL-6. Proc Natl Acad Sci U S A 98(16):9014-9019. Faulkner, S. 2012. The Effect of Stage of Cycle and Steroid Environment on the Bovine Uterine Proteome and Differences from Plasma, University College Dublin. Faulkner, S., G. Elia, O. B. P, M. Dunn, & D. Morris. (2013). Composition of the bovine uterine proteome is associated with stage of cycle and concentration of systemic progesterone. Proteomics 13(22):3333-3353. Feng, L., M. E. Hatten, & N. Heintz. (1994). Brain lipid-binding protein (BLBP): a novel signaling system in the developing mammalian CNS. Neuron 12(4):895-908. Feng, X. H., & R. Derynck. (2005). Specificity and versatility in tgf-beta signaling through Smads. Annu Rev Cell Dev Biol 21:659-693.

171

Fenn, J. B., M. Mann, C. K. Meng, S. F. Wong, & C. M. Whitehouse. (1989). Electrospray ionization for mass spectrometry of large biomolecules. Science 246(4926):64-71. Ferguson, P. L., & R. D. Smith. (2003). Proteome analysis by mass spectrometry. Annual review of biophysics and biomolecular structure 32(1):399-424. Ferraz, M. V. C., A. V. Pires, M. H. Santos, R. G. Silva, G. B. Oliveira, D. M. Polizel, M. V. Biehl, R. Sartori, & G. P. Nogueira. (2018). A combination of nutrition and genetics is able to reduce age at puberty in Nelore heifers to below 18 months. Animal 12(3):569- 574. Ferrer-Lorente, R., M. T. Bejar, & L. Badimon. (2014). Notch Signaling Pathway Activation in Normal and Hyperglycemic Rats Differs in the Stem Cells of Visceral and Subcutaneous Adipose Tissue. Stem Cells and Development 23(24):3034-3048. Filis, P., P. C. Kind, & N. Spears. (2013). Implantation failure in mice with a disruption in Phospholipase C beta 1 gene: lack of embryonic attachment, aberrant steroid hormone signalling and defective endocannabinoid metabolism. Molecular Human Reproduction 19(5):290-301. Fontana, R., & S. Della Torre. (2016). The Deep Correlation between Energy Metabolism and Reproduction: A View on the Effects of Nutrition for Women Fertility. Nutrients 8(2):87. Forde, N., J. P. Mehta, M. Minten, M. A. Crowe, J. F. Roche, T. E. Spencer, & P. Lonergan. (2012). Effects of low progesterone on the endometrial transcriptome in cattle. Biol Reprod 87(5):124. Fortes, M. R., K. Kemper, S. Sasazaki, A. Reverter, J. E. Pryce, W. Barendse, R. Bunch, R. McCulloch, B. Harrison, S. Bolormaa et al. (2013a). Evidence for pleiotropism and recent selection in the PLAG1 region in Australian Beef cattle. Anim Genet 44(6):636- 647. Fortes, M. R., Y. Li, E. Collis, Y. Zhang, & R. J. Hawken. (2013b). The IGF1 pathway genes and their association with age of puberty in cattle. Anim Genet 44(1):91-95. Fortes, M. R., L. T. Nguyen, M. M. Weller, A. Canovas, A. Islas-Trejo, L. R. Porto-Neto, A. Reverter, S. A. Lehnert, G. B. Boe-Hansen, M. G. Thomas et al. (2016a). Transcriptome analyses identify five transcription factors differentially expressed in the hypothalamus of post- versus prepubertal Brahman heifers. J Anim Sci 94(9):3693-3702. Fortes, M. R., A. Reverter, S. H. Nagaraj, Y. Zhang, N. N. Jonsson, W. Barris, S. Lehnert, G. B. Boe-Hansen, & R. J. Hawken. (2011). A single nucleotide polymorphism-derived regulatory gene network underlying puberty in 2 tropical breeds of beef cattle. J Anim Sci 89(6):1669-1683. Fortes, M. R., A. Reverter, Y. Zhang, E. Collis, S. H. Nagaraj, N. N. Jonsson, K. C. Prayaga, W. Barris, & R. J. Hawken. (2010a). Association weight matrix for the genetic dissection of puberty in beef cattle. Proc Natl Acad Sci U S A 107 Fortes, M. R., W. M. Snelling, A. Reverter, S. H. Nagaraj, S. A. Lehnert, R. J. Hawken, K. L. DeAtley, S. O. Peters, G. A. Silver, G. Rincon et al. (2012). Gene network analyses of first service conception in Brangus heifers: use of genome and trait associations, hypothalamic-transcriptome information, and transcription factors. J Anim Sci 90(9):2894-2906. Fortes, M. R. S., L. T. Nguyen, L. R. Porto Neto, A. Reverter, S. S. Moore, S. A. Lehnert, & M. G. Thomas. (2016b). Polymorphisms and genes associated with puberty in heifers. Theriogenology 86(1):333-339. Fortes, M. R. S., L. R. Porto-Neto, K. L. DeAtley, A. Reverter, M. G. Thomas, S. S. Moore, S. A. Lehnert, & W. M. Snelling. (2014). Genetic Markers in Transcription Factors of Differentially Expressed Genes Associated with Post-partum Anoestrus Predict

172

Pregnancy Outcome in an Independent Population of Beef Cattle. In: World Congress of Genetics Applied to Livestock Production, Vancouver, Canada Fortes, M. R. S., A. Reverter, Y. Zhang, E. Collis, S. H. Nagaraj, N. N. Jonsson, K. C. Prayaga, W. Barris, & R. J. Hawken. (2010b). Association weight matrix for the genetic dissection of puberty in beef cattle. Proceedings of the National Academy of Sciences of the United States of America 107(31):13642-13647. Fortin, J., U. Boehm, C.-X. Deng, M. Treier, & D. J. Bernard. (2014). Follicle-stimulating hormone synthesis and fertility depend on SMAD4 and FOXL2. The FASEB Journal 28(8):3396-3410. Foukas, L. C., M. Claret, W. Pearce, K. Okkenhaug, S. Meek, E. Peskett, S. Sancho, A. J. Smith, D. J. Withers, & B. Vanhaesebroeck. (2006). Critical role for the p110alpha phosphoinositide-3-OH kinase in growth and metabolic regulation. Nature 441(7091):366-370. Francisca, R.-P., J. M.-F. Antonio, T. Sulay, P. Leonor, T.-S. Manuel, D. Carlos, P. C. Justo, & M. M. María. (2007). Regulation of Pituitary Cell Function by Adiponectin. Endocrinology 148(1):401-410. Frankenberg, S., J. Fenelon, B. Dopheide, G. Shaw, & M. B. Renfree. (2011). A novel MSMB- related microprotein in the postovulatory egg coats of marsupials. BMC Evolutionary Biology 11:373-373. Frayn, K. N. (2002). Adipose tissue as a buffer for daily lipid flux. Diabetologia 45(9):1201- 1210. Fredriksson, R., M. C. Lagerström, P. J. Höglund, & H. B. Schiöth. (2002). Novel human G protein-coupled receptors with long N-terminals containing GPS domains and Ser/Thr- rich regions. FEBS Letters 531(3):407-414. Freetly, H. C., L. A. Kuehn, & L. V. Cundiff. (2011). Growth curves of crossbred cows sired by Hereford, Angus, Belgian Blue, Brahman, Boran, and Tuli bulls, and the fraction of mature body weight and height at puberty. J Anim Sci 89(8):2373-2379. Frisch, J. E., & J. E. Vercoe. (1977). Food intake, eating rate, weight gains, metabolic rate and efficiency of feed utilization in Bos taurus and Bos indicus crossbred cattle. Animal Science 25(3):343-358. Frisch, R. E. (1987). Body fat, menarche, fitness and fertility. Hum Reprod 2(6):521-533. Froment, P., F. Gizard, D. Defever, B. Staels, J. Dupont, & P. Monget. (2006). Peroxisome proliferator-activated receptors in reproductive tissues: from gametogenesis to parturition. Journal of Endocrinology 189(2):199-209. Funston, R. N., G. E. Moss, & A. J. Roberts. (1995). Insulin-like growth factor-I (IGF-I) and IGF-binding proteins in bovine sera and pituitaries at different stages of the estrous cycle. Endocrinology 136(1):62-68. Gahete, M. D., M. C. Vázquez-Borrego, A. J. Martínez-Fuentes, M. Tena-Sempere, J. P. Castaño, & R. M. Luque. (2016). Role of the Kiss1/Kiss1r system in the regulation of pituitary cell function. Molecular and Cellular Endocrinology Gajdos, Z. K. Z., K. D. Henderson, J. N. Hirschhorn, & M. R. Palmert. (2010). Genetic Determinants of Pubertal Timing in the General Population. Molecular and cellular endocrinology 324(1-2):21-29. Gao, Q., M. J. Wolfgang, S. Neschen, K. Morino, T. L. Horvath, G. I. Shulman, & X. Y. Fu. (2004). Disruption of neural signal transducer and activator of transcription 3 causes obesity, diabetes, infertility, and thermal dysregulation. Proc Natl Acad Sci U S A 101(13):4661-4666. Gao, X., J. Ye, C. Yang, L. Luo, Y. Liu, J. Ding, Y. Zhang, Y. Ling, W. Huang, X. Zhang et al. (2018). RNA-seq analysis of lncRNA-controlled developmental gene expression during puberty in goat & rat. BMC Genetics 19:19.

173

Gao, X., J. Ye, C. Yang, K. Zhang, X. Li, L. Luo, J. Ding, Y. Li, H. Cao, Y. Ling et al. (2017). Screening and evaluating of long noncoding RNAs in the puberty of goats. BMC Genomics 18:164. Garcia, M. R. 2003. Developmental, dietary, and hormonal regulation of leptin synthesis, secretion, and action in cattle. Ph.D., Texas A&M University, Ann Arbor. Garcia, M. R., M. Amstalden, S. W. Williams, R. L. Stanko, C. D. Morrison, D. H. Keisler, S. E. Nizielski, & G. L. Williams. (2002). Serum leptin and its adipose gene expression during pubertal development, the estrous cycle, and different seasons in cattle. J Anim Sci 80(8):2158-2167. Garcia-Lavandeira, M., E. Diaz-Rodriguez, D. Bahar, A. R. Garcia-Rendueles, J. S. Rodrigues, C. Dieguez, & C. V. Alvarez. (2015). Pituitary Cell Turnover: From Adult Stem Cell Recruitment through Differentiation to Death. Neuroendocrinology 101(3):175-192. Gasser, C. L., D. E. Grum, M. L. Mussard, F. L. Fluharty, J. E. Kinder, & M. L. Day. (2006). Induction of precocious puberty in heifers I: enhanced secretion of luteinizing hormone. J Anim Sci 84(8):2035-2041. Geach, T. (2016). microRNAs regulate puberty timing. Nature Reviews Endocrinology 12:372. Gelling, R. W., G. J. Morton, C. D. Morrison, K. D. Niswender, M. G. Myers, Jr., C. J. Rhodes, & M. W. Schwartz. (2006). Insulin action in the brain contributes to glucose lowering during insulin treatment of diabetes. Cell Metab 3(1):67-73. Genzer, Y., N. Chapnik, & O. Froy. (2017). Effect of brain-derived neurotrophic factor (BDNF) on hepatocyte metabolism. The International Journal of Biochemistry & Cell Biology 88(Supplement C):69-74. Ghaemmaghami, S., W. K. Huh, K. Bower, R. W. Howson, A. Belle, N. Dephoure, E. K. O'Shea, & J. S. Weissman. (2003). Global analysis of protein expression in yeast. Nature 425(6959):737-741. Gharib, S. D., M. E. Wierman, T. M. Badger, & W. W. Chin. (1987). Sex steroid hormone regulation of follicle-stimulating hormone subunit messenger ribonucleic acid (mRNA) levels in the rat. J Clin Invest 80(2):294-299. Gibson, E. M., S. A. Humber, S. Jain, W. P. Williams, 3rd, S. Zhao, G. E. Bentley, K. Tsutsui, & L. J. Kriegsfeld. (2008). Alterations in RFamide-related peptide expression are coordinated with the preovulatory luteinizing hormone surge. Endocrinology 149(10):4958-4969. Gimpl, G., & F. Fahrenholz. (2001). The oxytocin receptor system: structure, function, and regulation. Physiol Rev 81(2):629-683. Gleeson, T., P. Martin, & C. Mifsud. 2012. Northern Australian beef industry: Assessment of risks and opportunities, Canberra. Glidewell-Kenney, C., L. A. Hurley, L. Pfaff, J. Weiss, J. E. Levine, & J. L. Jameson. (2007). Nonclassical estrogen receptor alpha signaling mediates negative feedback in the female mouse reproductive axis. Proc Natl Acad Sci U S A 104(19):8173-8177. Gonella-Diaza, A. M., S. C. da Silva Andrade, M. Sponchiado, G. Pugliesi, F. S. Mesquita, V. Van Hoeck, R. de Francisco Strefezzi, G. R. Gasparin, L. L. Coutinho, & M. Binelli. (2017). Oviductal transcriptional profiling of a bovine fertility model by next- generation sequencing. Genom Data 13:27-29. Goodyer, C. G., L. De Stephano, W. H. Lai, H. J. Guyda, & B. I. Posner. (1984). Characterization of insulin-like growth factor receptors in rat anterior pituitary, hypothalamus, and brain. Endocrinology 114(4):1187-1195. Gorlov, I., A. Fedunin, D. Randelin, & G. Sulimova. (2014). Polymorphisms of bGH, RORC, and DGAT1 genes in Russian beef cattle breeds. Russian journal of genetics 50(12):1302-1307.

174

Govoni, K. E., X. C. Tian, G. W. Kazmer, M. Taneja, B. P. Enright, A. L. Rivard, X. Yang, & S. A. Zinn. (2002). Age-related changes of the somatotropic axis in cloned Holstein calves. Biol Reprod 66(5):1293-1298. Granger, A. L., W. E. Wyatt, W. M. Craig, D. L. Thompson, Jr., & F. G. Hembry. (1989). Effects of breed and wintering diet on growth, puberty and plasma concentrations of growth hormone and insulin-like growth factor 1 in heifers. Domest Anim Endocrinol 6(3):253-262. Grass, J. A., P. J. Hansen, J. J. Rutledge, & E. R. Hauser. (1982). Genotype x environmental interactions on reproductive traits of bovine females. I. Age at puberty asinfluenced by breed, breed of sire, dietary regimen and season. J.Anim. Sci 55:1441. Gravati, M., M. Busnelli, E. Bulgheroni, A. Reversi, P. Spaiardi, M. Parenti, M. Toselli, & B. Chini. (2010). Dual modulation of inward rectifier potassium currents in olfactory neuronal cells by promiscuous G protein coupling of the oxytocin receptor. Journal of Neurochemistry 114(5):1424-1435. Gray, P. W., D. Glaister, P. H. Seeburg, A. Guidotti, & E. Costa. (1986). Cloning and expression of cDNA for human diazepam binding inhibitor, a natural ligand of an allosteric regulatory site of the gamma-aminobutyric acid type A receptor. Proceedings of the National Academy of Sciences 83(19):7547-7551. Greenbaum, D., C. Colangelo, K. Williams, & M. Gerstein. (2003). Comparing protein abundance and mRNA expression levels on a genomic scale. Genome Biology 4(9):117. Gregory, P. A., & P. I. Mackenzie. (2002). The homeodomain Pbx2-Prep1 complex modulates hepatocyte nuclear factor 1alpha-mediated activation of the UDP- glucuronosyltransferase 2B17 gene. Mol Pharmacol 62(1):154-161. Grewal, S., J. G. Carver, A. J. Ridley, & H. J. Mardon. (2008). Implantation of the human embryo requires Rac1-dependent endometrial stromal cell migration. Proc Natl Acad Sci U S A 105(42):16189-16194. Gridley, T., & S. Kajimura. (2014). Lightening up a notch: Notch regulation of energy metabolism. Nat Med 20(8):811-812. Griffin, T. J., S. P. Gygi, T. Ideker, B. Rist, J. Eng, L. Hood, & R. Aebersold. (2002). Complementary profiling of gene expression at the transcriptome and proteome levels in Saccharomyces cerevisiae. Mol Cell Proteomics 1(4):323-333. Griffith, M., O. L. Griffith, J. Mwenifumbo, R. Goya, A. S. Morrissy, R. D. Morin, R. Corbett, M. J. Tang, Y.-C. Hou, & T. J. Pugh. (2010). Alternative expression analysis by RNA sequencing. Nature methods 7(10):843. Grochowska, R., P. Sorensen, L. Zwierzchowski, M. Snochowski, & P. Lovendahl. (2001). Genetic variation in stimulated GH release and in IGF-I of young dairy cattle and their associations with the leucine/valine polymorphism in the GH gene. J Anim Sci 79(2):470-476. Guan, X. M., J. F. Hess, H. Yu, P. J. Hey, & L. H. van der Ploeg. (1997). Differential expression of mRNA for leptin receptor isoforms in the rat brain. Mol Cell Endocrinol 133(1):1- 7. Guerra-Junior, G., A. M. Spinola-Castro, A. A. Siviero-Miachon, R. G. Nogueira, S. H. Lemos- Marini, L. F. D'Souza-Li, P. C. Silva, E. S. Franca, F. C. Soardi, & M. P. Mello. (2008). Absence of mutations in Pax6 gene in three cases of morning glory syndrome associated with isolated growth hormone deficiency. Arq Bras Endocrinol Metabol 52(8):1221-1227. Guerrera, I. C., & O. Kleiner. (2005). Application of mass spectrometry in proteomics. Biosci Rep 25(1-2):71-93.

175

Guillermet-Guibert, J., L. B. Smith, G. Halet, M. A. Whitehead, W. Pearce, D. Rebourcet, K. Leon, P. Crepieux, G. Nock, M. Stromstedt et al. (2015). Novel Role for p110beta PI 3-Kinase in Male Fertility through Regulation of Androgen Receptor Activity in Sertoli Cells. PLoS Genet 11(7):e1005304. Guo, B., P. L. Greenwood, L. M. Cafe, G. Zhou, W. Zhang, & B. P. Dalrymple. (2015). Transcriptome analysis of cattle muscle identifies potential markers for skeletal muscle growth rate and major cell types. BMC Genomics 16(1):177. Gupta, S., J. Ghulmiyyah, R. Sharma, J. Halabi, & A. Agarwal. (2014). Power of proteomics in linking oxidative stress and female infertility. Biomed Res Int 2014:916212. Gygi, S. P., Y. Rochon, B. R. Franza, & R. Aebersold. (1999). Correlation between protein and mRNA abundance in yeast. Mol Cell Biol 19(3):1720-1730. Hachey, D. L., & P. Chaurand. (2004). Proteomics in reproductive medicine: the technology for separation and identification of proteins. J Reprod Immunol 63(1):61-73. Haider, S., & R. Pal. (2013). Integrated analysis of transcriptomic and proteomic data. Curr Genomics 14(2):91-110. Halet, G., & J. Carroll. (2007). Rac activity is polarized and regulates meiotic spindle stability and anchoring in mammalian oocytes. Dev Cell 12(2):309-317. Han, C., H. Liu, J. Liu, K. Yin, Y. Xie, X. Shen, Y. Wang, J. Yuan, B. Qiang, Y. J. Liu et al. (2005). Human Bex2 interacts with LMO2 and regulates the transcriptional activity of a novel DNA-binding complex. Nucleic Acids Res 33(20):6555-6565. Han, S. K., M. G. Todman, & A. E. Herbison. (2004). Endogenous GABA release inhibits the firing of adult gonadotropin-releasing hormone neurons. Endocrinology 145(2):495- 499. Han, W., Y. Zhu, Y. Su, G. Li, L. Qu, H. Zhang, K. Wang, J. Zou, & H. Liu. (2016). High- Throughput Sequencing Reveals Circulating miRNAs as Potential Biomarkers for Measuring Puberty Onset in Chicken (Gallus gallus). PLOS ONE 11(5):e0154958. Han, X., A. Aslanian, & J. R. Yates. (2008). Mass Spectrometry for Proteomics. Current opinion in chemical biology 12(5):483-490. Han, Y., S. Gao, K. Muegge, W. Zhang, & B. Zhou. (2015). Advanced Applications of RNA Sequencing and Challenges. Bioinformatics and Biology Insights 9(Suppl 1):29-46. Hansen, P. J. (1985). Seasonal modulation of puberty and the postpartum anestrus in cattle: A review. Livestock Production Science 12(4):309-327. Hargrove, J. L., & F. H. Schmidt. (1989). The role of mRNA and protein stability in gene expression. Faseb j 3(12):2360-2370. Harris, T. G., S. Dye, J. E. Robinson, D. C. Skinner, & N. P. Evans. (1999). Progesterone can block transmission of the estradiol-induced signal for luteinizing hormone surge generation during a specific period of time immediately after activation of the gonadotropin-releasing hormone surge-generating system. Endocrinology 140(2):827- 834. Hashizume, T., A. Kumahara, M. Fujino, & K. Okada. (2002). Insulin-like growth factor I enhances gonadotropin-releasing hormone-stimulated luteinizing hormone release from bovine anterior pituitary cells. Animal Reproduction Science 70(1–2):13-21. Hashizume, T., K. Ohtsuki, & N. Matsumoto. (2000). Plasma insulin-like growth factor-I concentrations increase during the estrous phase in goats. Domest Anim Endocrinol 18(2):253-263. Hashizume, T., H. Saito, T. Sawada, T. Yaegashi, A. A. Ezzat, K. Sawai, & T. Yamashita. (2010). Characteristics of stimulation of gonadotropin secretion by kisspeptin-10 in female goats. Anim Reprod Sci 118(1):37-41. Hatoya, S., R. Torii, D. Kumagai, K. Sugiura, N. Kawate, H. Tamada, T. Sawada, & T. Inaba. (2003). Expression of estrogen receptor alpha and beta genes in the mediobasal

176

hypothalamus, pituitary and ovary during the canine estrous cycle. Neurosci Lett 347(2):131-135. Hatzimanikatis, V., & K. H. Lee. (1999). Dynamical Analysis of Gene Networks Requires Both mRNA and Protein Expression Information. Metabolic Engineering 1(4):275- 281. Hawken, R. J., Y. D. Zhang, M. R. Fortes, E. Collis, W. C. Barris, N. J. Corbet, P. J. Williams, G. Fordyce, R. G. Holroyd, J. R. Walkley et al. (2012a). Genome-wide association studies of female reproduction in tropically adapted beef cattle. J Anim Sci 90(5):1398- 1410. Hawken, R. J., Y. D. Zhang, M. R. S. Fortes, E. Collis, W. C. Barris, N. J. Corbet, P. J. Williams, G. Fordyce, R. G. Holroyd, J. R. W. Walkley et al. (2012b). Genome-wide association studies of female reproduction in tropically adapted beef cattle. Journal of Animal Science 90(5):1398-1410. He, C., P. Kraft, C. Chen, J. E. Buring, G. Paré, S. E. Hankinson, S. J. Chanock, P. M. Ridker, D. J. Hunter, & D. I. Chasman. (2009). Genome-wide association studies identify loci associated with age at menarche and age at natural menopause. Nature Genetics 41:724. He, D., T. Funabashi, A. Sano, T. Uemura, H. Minaguchi, & F. Kimura. (1999). Effects of glucose and related substrates on the recovery of the electrical activity of gonadotropin- releasing hormone pulse generator which is decreased by insulin-induced hypoglycemia in the estrogen-primed ovariectomized rat. Brain Res 820(1-2):71-76. He, R.-H., J.-Z. Sheng, Q. Luo, F. Jin, B. Wang, Y.-L. Qian, C.-Y. Zhou, X. Sheng, & H.-F. Huang. (2006). Aquaporin-2 expression in human endometrium correlates with serum ovarian steroid hormones. Life Sciences 79(5):423-429. He, X., J. Liu, Y. Qi, C. Brakebusch, A. Chrostek-Grashoff, D. Edgar, P. D. Yurchenco, S. A. Corbett, S. F. Lowry, A. M. Graham et al. (2010). Rac1 is essential for basement membrane-dependent epiblast survival. Mol Cell Biol 30(14):3569-3581. Hekman, K. E., G. Y. Yu, C. D. Brown, H. Zhu, X. Du, K. Gervin, D. E. Undlien, A. Peterson, G. Stevanin, H. B. Clark et al. (2012). A conserved eEF2 coding variant in SCA26 leads to loss of translational fidelity and increased susceptibility to proteostatic insult. Hum Mol Genet 21(26):5472-5483. Henchman, M., & C. Steel. (1998). Understanding the quadrupole mass filter through computer simulation. Journal of chemical education 75(8):1049. Henneberger, C., R. Juttner, T. Rothe, & R. Grantyn. (2002). Postsynaptic action of BDNF on GABAergic synaptic transmission in the superficial layers of the mouse superior colliculus. J Neurophysiol 88(2):595-603. Henzel, W. J., T. M. Billeci, J. T. Stults, S. C. Wong, C. Grimley, & C. Watanabe. (1993). Identifying proteins from two-dimensional gels by molecular mass searching of peptide fragments in protein sequence databases. Proceedings of the National Academy of Sciences 90(11):5011-5015. Herath, C. B., S. Shiojima, H. Ishiwata, S. Katsuma, T. Kadowaki, K. Ushizawa, K. Imai, T. Takahashi, A. Hirasawa, G. Tsujimoto et al. (2004). Pregnancy-associated changes in genome-wide gene expression profiles in the liver of cow throughout pregnancy. Biochem Biophys Res Commun 313(3):666-680. Herbison, A. E. (2008). Estrogen positive feedback to gonadotropin-releasing hormone (GnRH) neurons in the rodent: the case for the rostral periventricular area of the third ventricle (RP3V). Brain Res Rev 57(2):277-287. Hill, W. G. (2010). Understanding and using quantitative genetic variation. Philosophical Transactions of the Royal Society B-Biological Sciences 365(1537):73-85.

177

Hiney, J. K., S. R. Ojeda, & W. L. Dees. (1991). Insulin-like growth factor I: a possible metabolic signal involved in the regulation of female puberty. Neuroendocrinology 54(4):420-423. Hiney, J. K., V. Srivastava, C. L. Nyberg, S. R. Ojeda, & W. L. Dees. (1996). Insulin-like growth factor I of peripheral origin acts centrally to accelerate the initiation of female puberty. Endocrinology 137(9):3717-3728. Hiney, J. K., V. K. Srivastava, & W. Les Dees. (2010). Insulin-like growth factor-1 stimulation of hypothalamic KiSS-1 gene expression is mediated by Akt: effect of alcohol. Neuroscience 166(2):625-632. Hiney, J. K., V. K. Srivastava, M. D. Pine, & W. Les Dees. (2009). Insulin-like growth factor- I activates KiSS-1 gene expression in the brain of the prepubertal female rat. Endocrinology 150(1):376-384. Hinuma, S., Y. Shintani, S. Fukusumi, N. Iijima, Y. Matsumoto, M. Hosoya, R. Fujii, T. Watanabe, K. Kikuchi, Y. Terao et al. (2000). New neuropeptides containing carboxy- terminal RFamide and their receptor in mammals. Nat Cell Biol 2(10):703-708. Ho, C. S., C. W. K. Lam, M. H. M. Chan, R. C. K. Cheung, L. K. Law, L. C. W. Lit, K. F. Ng, M. W. M. Suen, & H. L. Tai. (2003). Electrospray Ionisation Mass Spectrometry: Principles and Clinical Applications. The Clinical Biochemist Reviews 24(1):3-12. Hodges, C. A., & M. R. Palmert. 2007. Genetic Regulation of the Variation in Pubertal Timing. In: O. H. Pescovitz and E. C. Walvoord, editors, When Puberty is Precocious: Scientific and Clinical Aspects. Humana Press, Totowa, NJ. p. 83-102. Hogarth, C. J., J. L. Fitzpatrick, A. M. Nolan, F. J. Young, A. Pitt, & P. D. Eckersall. (2004). Differential protein composition of bovine whey: a comparison of whey from healthy animals and from those with clinical mastitis. Proteomics 4(7):2094-2100. Hu, R., X. Qiu, G. Glazko, L. Klebanov, & A. Yakovlev. (2009). Detecting intergene correlation changes in microarray analysis: a new approach to gene selection. BMC Bioinformatics 10(1):20. Hu, Z., Y. Wang, F. Huang, R. Chen, C. Li, F. Wang, J. Goto, D. J. Kwiatkowski, J. Wdzieczak- Bakala, P. Tu et al. (2015). Brain-expressed X-linked 2 is pivotal for hyperactive mechanistic target of rapamycin (mTOR)-mediated tumorigenesis. J Biol Chem 290(42):25756-25765. Hu, Z.-L., C. A. Park, & J. M. Reecy. (2016). Developmental progress and current status of the Animal QTLdb. Nucleic Acids Research 44(Database issue):D827-D833. Huang da, W., B. T. Sherman, & R. A. Lempicki. (2009). Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 4(1):44-57. Huang, H.-F., R.-H. He, C.-C. Sun, Y. Zhang, Q.-X. Meng, & Y.-Y. Ma. (2006). Function of aquaporins in female and male reproductive systems. Human Reproduction Update 12(6):785-795. Huang, W., Y. Guo, W. Du, X. Zhang, A. Li, & X. Miao. (2017). Global transcriptome analysis identifies differentially expressed genes related to lipid metabolism in Wagyu and Holstein cattle. Scientific Reports 7:5278. Hudson, N. J., A. Reverter, & B. P. Dalrymple. (2009). A Differential Wiring Analysis of Expression Data Correctly Identifies the Gene Containing the Causal Mutation. PLOS Computational Biology 5(5):e1000382. Humaidan, P., I. Van Vaerenbergh, C. Bourgain, B. Alsbjerg, C. Blockeel, F. Schuit, L. Van Lommel, P. Devroey, & H. Fatemi. (2012). Endometrial gene expression in the early luteal phase is impacted by mode of triggering final oocyte maturation in recFSH stimulated and GnRH antagonist co-treated IVF cycles. Hum Reprod 27(11):3259- 3272.

178

Hussein, T. S., D. A. Froiland, F. Amato, J. G. Thompson, & R. B. Gilchrist. (2005). Oocytes prevent cumulus cell apoptosis by maintaining a morphogenic paracrine gradient of bone morphogenetic proteins. Journal of Cell Science 118(22):5257-5268. I'Anson, H., J. M. Manning, C. G. Herbosa, J. Pelt, C. R. Friedman, R. I. Wood, D. C. Bucholtz, & D. L. Foster. (2000). Central inhibition of gonadotropin-releasing hormone secretion in the growth-restricted hypogonadotropic female sheep. Endocrinology 141(2):520- 527. Iqbal, J., S. Pompolo, R. V. Considine, & I. J. Clarke. (2000). Localization of leptin receptor- like immunoreactivity in the corticotropes, somatotropes, and gonadotropes in the ovine anterior pituitary. Endocrinology 141(4):1515-1520. Iqbal, J., S. Pompolo, T. Murakami, E. Grouzmann, T. Sakurai, B. Meister, & I. J. Clarke. (2001). Immunohistochemical characterization of localization of long-form leptin receptor (OB-Rb) in neurochemically defined cells in the ovine hypothalamus. Brain Res 920(1-2):55-64. Irwig, M. S., G. S. Fraley, J. T. Smith, B. V. Acohido, S. M. Popa, M. J. Cunningham, M. L. Gottsch, D. K. Clifton, & R. A. Steiner. (2004). Kisspeptin activation of gonadotropin releasing hormone neurons and regulation of KiSS-1 mRNA in the male rat. Neuroendocrinology 80(4):264-272. Ishiwata, H., S. Katsuma, K. Kizaki, O. V. Patel, H. Nakano, T. Takahashi, K. Imai, A. Hirasawa, S. Shiojima, H. Ikawa et al. (2003). Characterization of gene expression profiles in early bovine pregnancy using a custom cDNA microarray. Mol Reprod Dev 65(1):9-18. Ito, K., S. Yamazaki, R. Yamamoto, Y. Tajima, A. Yanagida, T. Kobayashi, M. Kato-Itoh, S. Kakuta, Y. Iwakura, H. Nakauchi et al. (2014). Gene targeting study reveals unexpected expression of Brain-expressed X-linked 2 in endocrine and tissue Stem/Progenitor cells in Mice. Journal of Biological Chemistry 289(43):29892-29911. Jablonski, E. M., N. A. McConnell, F. M. Hughes, & Y. M. Huet-Hudson. (2003). Estrogen Regulation of Aquaporins in the Mouse Uterus: Potential Roles in Uterine Water Movement. Biology of Reproduction 69(5):1481-1487. Janssen, U., E. M. Davis, M. M. Le Beau, & W. Stoffel. (1997). Human mitochondrial enoyl- CoA hydratase gene (ECHS1): structural organization and assignment to chromosome 10q26.2-q26.3. Genomics 40(3):470-475. Jauliac, S., C. Lopez-Rodriguez, L. M. Shaw, L. F. Brown, A. Rao, & A. Toker. (2002). The role of NFAT transcription factors in integrin-mediated carcinoma invasion. Nature Cell Biology 4(7):540-544. Jayapal, K. P., R. J. Philp, Y. J. Kok, M. G. Yap, D. H. Sherman, T. J. Griffin, & W. S. Hu. (2008). Uncovering genes with divergent mRNA-protein dynamics in Streptomyces coelicolor. PLoS One 3(5):e2097. Jeong, K. H., W. W. Chin, & U. B. Kaiser. (2004a). Essential role of the homeodomain for pituitary homeobox 1 activation of mouse gonadotropin-releasing hormone receptor gene expression through interactions with c-Jun and DNA. Mol Cell Biol 24(14):6127- 6139. Jeong, W., T. S. Chang, E. S. Boja, H. M. Fales, & S. G. Rhee. (2004b). Roles of TRP14, a thioredoxin-related protein in tumor necrosis factor-alpha signaling pathways. J Biol Chem 279(5):3151-3159. Jeong, W., Y. Jung, H. Kim, S. J. Park, & S. G. Rhee. (2009). Thioredoxin-related protein 14, a new member of the thioredoxin family with disulfide reductase activity: Implication in the redox regulation of TNF-α signaling. Free Radical Biology and Medicine 47(9):1294-1303.

179

Ji, H., H. Yang, J. Wang, J. Guo, Z. Hu, H. Zhang, F. Kong, L. Zhen, L. Guo, & J. Liu. (2014). Effect of adenovirus-mediated up-regulation of alpha-enolase gene products on follicle- stimulating hormone receptor mRNA and luteinizing hormone receptor mRNA of granular cells from goose F1 follicles. Res Vet Sci 96(3):526-532. Jiang, H., Y. Kuang, Y. Wu, W. Xie, M. I. Simon, & D. Wu. (1997). Roles of phospholipase C beta2 in chemoattractant-elicited responses. Proc Natl Acad Sci U S A 94(15):7971- 7975. Jin, L., S. Zhang, B. G. Burguera, M. E. Couce, R. Y. Osamura, E. Kulig, & R. V. Lloyd. (2000). Leptin and leptin receptor expression in rat and mouse pituitary cells. Endocrinology 141(1):333-339. Jin, W., E. N. Olson, S. S. Moore, J. A. Basarab, U. Basu, & L. L. Guan. (2012). Transcriptome analysis of subcutaneous adipose tissues in beef cattle using 3' digital gene expression- tag profiling. J Anim Sci 90(1):171-183. Jo, Y.-H., & S. C. Chua. (2013). The Brain–Liver Connection Between BDNF and Glucose Control. Diabetes 62(5):1367-1368. Johnston, D. J. (2014). Genetic Improvement of Reproduction in Beef Cattle. In: 10th World Congress of Genetics Applied to Livestock Production, Canada Johnston, D. J., S. A. Barwick, N. J. Corbet, G. Fordyce, R. G. Holroyd, P. J. Williams, & H. M. Burrow. (2009). Genetics of heifer puberty in two tropical beef genotypes in northern Australia and associations with heifer- and steer-production traits. Animal Production Science 49(6):399-412. Johnston, D. J., S. A. Barwick, G. Fordyce, R. G. Holroyd, P. J. Williams, N. J. Corbet, & T. Grant. (2013). Genetics of early and lifetime annual reproductive performance in cows of two tropical beef genotypes in northern Australia. Animal Production Science 54(1):1-15. Jones, E. J., J. D. Armstrong, & R. W. Harvey. (1991). Changes in metabolites, metabolic hormones, and luteinizing hormone before puberty in Angus, Braford, Charolais, and Simmental heifers. J Anim Sci 69(4):1607-1615. Juengel, J. L., K. J. Bodensteiner, D. A. Heath, N. L. Hudson, C. L. Moeller, P. Smith, S. M. Galloway, G. H. Davis, H. R. Sawyer, & K. P. McNatty. (2004). Physiology of GDF9 and BMP15 signalling molecules. Animal Reproduction Science 82–83:447-460. Junko, M., Y. Ichiro, T. Yoshihiro, M. Takashi, I. Kazunori, S. Kenji, & K. Yasuhiro. (1999). Developmental Changes in Long-Form Leptin Receptor Expression and Localization in Rat Brain. Endocrinology 140(11):5233-5238. Kalra, S. P. (1993). Mandatory neuropeptide-steroid signaling for the preovulatory luteinizing hormone-releasing hormone discharge. Endocr Rev 14(5):507-538. Kamalludin, M. H., A. G. Guerra, M. Wiltbank, & B. W. Kirkpatrick. (2017). Trio, a novel high fecundity allele: I. Transcriptome analysis of granulosa cells from carriers and non-carriers of a major gene for bovine ovulation rate. Biol Reprod Kanehisa, M., M. Furumichi, M. Tanabe, Y. Sato, & K. Morishima. (2017). KEGG: new perspectives on genomes, pathways, diseases and drugs. Nucleic Acids Res 45(D1):D353-d361. Kang, B., D. M. Jiang, L. Bai, H. He, & R. Ma. (2014). Molecular characterisation and expression profiling of the ENO1 gene in the ovarian follicle of the Sichuan white goose. Mol Biol Rep 41(4):1927-1935. Kang, D., G. Zhou, S. Zhou, J. Zeng, X. Wang, Y. Jiang, Y. Yang, & Y. Chen. (2017). Comparative transcriptome analysis reveals potentially novel roles of Homeobox genes in adipose deposition in fat-tailed sheep. Scientific Reports 7(1):14491.

180

Kappes, S. M., G. L. Bennett, J. W. Keele, S. E. Echternkamp, K. E. Gregory, & R. M. Thallman. (2000). Initial results of genomic scans for ovulation rate in a cattle population selected for increased twinning rate. J Anim Sci 78(12):3053-3059. Karim, L., H. Takeda, L. Lin, T. Druet, J. A. C. Arias, D. Baurain, N. Cambisano, S. R. Davis, F. Farnir, B. Grisart et al. (2011). Variants modulating the expression of a chromosome domain encompassing PLAG1 influence bovine stature. Nat Genet 43 Kauffman, A. S., D. K. Clifton, & R. A. Steiner. (2007). Emerging ideas about kisspeptin- GPR54 signaling in the neuroendocrine regulation of reproduction. Trends Neurosci 30(10):504-511. Kaul, G., G. Pattan, & T. Rafeequi. (2011). Eukaryotic elongation factor-2 (eEF2): its regulation and peptide chain elongation. Cell Biochemistry and Function 29(3):227- 234. Kawashima, T., Y. C. Bao, Y. Nomura, Y. Moon, Y. Tonozuka, Y. Minoshima, T. Hatori, A. Tsuchiya, M. Kiyono, T. Nosaka et al. (2006). Rac1 and a GTPase-activating protein, MgcRacGAP, are required for nuclear translocation of STAT transcription factors. J Cell Biol 175(6):937-946. Kawwass, J. F., R. Summer, & C. B. Kallen. (2015). Direct effects of leptin and adiponectin on peripheral reproductive tissues: a critical review. Mol Hum Reprod 21(8):617-632. Keisler, D. H., & M. C. Lucy. (1996). Perception and interpretation of the effects of undernutrition on reproduction. J.Anim. Sci 74(suppl_3):1-17. Kelemen, O., P. Convertini, Z. Zhang, Y. Wen, M. Shen, M. Falaleeva, & S. Stamm. (2013). Function of alternative splicing. Gene 514(1):1-30. Kerrien, S., B. Aranda, L. Breuza, A. Bridge, F. Broackes-Carter, C. Chen, M. Duesbury, M. Dumousseau, M. Feuermann, U. Hinz et al. (2012). The IntAct molecular interaction database in 2012. Nucleic Acids Res 40(Database issue):D841-846. Khadra, A., & Y.-X. Li. (2006). A Model for the Pulsatile Secretion of Gonadotropin-Releasing Hormone from Synchronized Hypothalamic Neurons. Biophysical Journal 91(1):74- 83. Kherraf, Z.-E., B. Conne, A. Amiri-Yekta, M. C. Kent, C. Coutton, J. Escoffier, S. Nef, C. Arnoult, & P. F. Ray. (2018). Creation of knock out and knock in mice by CRISPR/Cas9 to validate candidate genes for human male infertility, interest, difficulties and feasibility. Molecular and Cellular Endocrinology 468:70-80. Kiezun, M., A. Maleszka, N. Smolinska, A. Nitkiewicz, & T. Kaminski. (2013). Expression of adiponectin receptors 1 (AdipoR1) and 2 (AdipoR2) in the porcine pituitary during the oestrous cycle. Reproductive Biology and Endocrinology : RB&E 11:18-18. Killeen, A. P., D. G. Morris, D. A. Kenny, M. P. Mullen, M. G. Diskin, & S. M. Waters. (2014). Global gene expression in endometrium of high and low fertility heifers during the mid- luteal phase of the estrous cycle. BMC Genomics 15:234. Kim, D., K. S. Jun, S. B. Lee, N. G. Kang, D. S. Min, Y. H. Kim, S. H. Ryu, P. G. Suh, & H. S. Shin. (1997). Phospholipase C isozymes selectively couple to specific neurotransmitter receptors. Nature 389(6648):290-293. Kim, E. S., X. Shi, O. Cobanoglu, K. Weigel, P. J. Berger, & B. W. Kirkpatrick. (2009). Refined mapping of twinning-rate quantitative trait loci on bovine chromosome 5 and analysis of insulin-like growth factor-1 as a positional candidate gene. J Anim Sci 87(3):835-843. Kim, H. R., R. X. Han, J. T. Yoon, C. S. Park, & D. I. Jin. (2010). A two‐dimensional electrophoresis reference map for the bovine placenta during late pregnancy. Proteomics 10(3):564-573.

181

Kim, H. R., J. K. Kang, J. T. Yoon, H. H. Seong, J. K. Jung, H. M. Lee, C. Sik Park, & D. I. Jin. (2005). Protein profiles of bovine placenta derived from somatic cell nuclear transfer. Proteomics 5(16):4264-4273. Kim, J. H., H. T. Cho, & Y. J. Kim. (2014). The role of estrogen in adipose tissue metabolism: insights into glucose homeostasis regulation. Endocr J 61(11):1055-1067. Kim, M. L., S. Yuhua, & C. G. Andrea. (1998). Insulin-Like Growth Factor-I Effects on Gonadotropin-Releasing Hormone Biosynthesis in GT1–7 Cells. Endocrinology 139(3):1125-1132. Kim, S., J. A. Pyun, D. H. Cha, J. J. Ko, & K. Kwack. (2011a). Epistasis between FSHR and CYP19A1 polymorphisms is associated with premature ovarian failure. Fertil Steril 95(8):2585-2588. Kim, S., J. A. Pyun, H. Kang, J. Kim, D. H. Cha, & K. Kwack. (2011b). Epistasis between CYP19A1 and ESR1 polymorphisms is associated with premature ovarian failure. Fertil Steril 95(1):353-356. Kinder, J. E., E. G. Bergfeld, M. E. Wehrman, K. E. Peters, & F. N. Kojima. (1995). Endocrine basis for puberty in heifers and ewes. J Reprod Fertil Suppl 49:393-407. Klesert, T. R., D. H. Cho, J. I. Clark, J. Maylie, J. Adelman, L. Snider, E. C. Yuen, P. Soriano, & S. J. Tapscott. (2000). Mice deficient in Six5 develop cataracts: implications for myotonic dystrophy. Nature Genetics 25(1):105-109. Klisch, K., A. Boos, M. Friedrich, K. Herzog, M. Feldmann, N. Sousa, J.-F. Beckers, R. Leiser, & G. Schuler. (2006). The glycosylation of pregnancy-associated glycoproteins and prolactin-related protein-I in bovine binucleate trophoblast giant cells changes before parturition. Reproduction 132(5):791-798. Komar, C. M. (2005). Peroxisome proliferator-activated receptors (PPARs) and ovarian function – implications for regulating steroidogenesis, differentiation, and tissue remodeling. Reproductive Biology and Endocrinology 3(1):41. Komar, C. M., O. Braissant, W. Wahli, & T. E. Curry, Jr. (2001). Expression and localization of PPARs in the rat ovary during follicular development and the periovulatory period. Endocrinology 142(11):4831-4838. Korfmacher, W. A. (2005). Principles and applications of LC-MS in new drug discovery. Drug Discov Today 10(20):1357-1367. Kostik, M. M., A. M. Smirnov, G. S. Demin, M. M. Mnuskina, L. A. Scheplyagina, & V. I. Larionova. (2013). Genetic polymorphisms of collagen type I α1 chain (COL1A1) gene increase the frequency of low bone mineral density in the subgroup of children with juvenile idiopathic arthritis. The EPMA Journal 4(1):15-15. Kostka, D., & R. Spang. (2004). Finding disease specific alterations in the co-expression of genes. Bioinformatics 20 Suppl 1:i194-199. Kostyak, J. C., P. Kris-Etherton, D. Bagshaw, J. P. DeLany, & P. A. Farrell. (2007). Relative fat oxidation is higher in children than adults. Nutrition Journal 6:19-19. Koussounadis, A., S. P. Langdon, I. H. Um, D. J. Harrison, & V. A. Smith. (2015). Relationship between differentially expressed mRNA and mRNA-protein correlations in a xenograft model system. Scientific Reports 5:10775. Kowalewski, M. P., A. Meyer, B. Hoffmann, S. Aslan, & A. Boos. (2011). Expression and functional implications of Peroxisome Proliferator—Activated Receptor Gamma (PPARγ) in canine reproductive tissues during normal pregnancy and parturition and at antiprogestin induced abortion. Theriogenology 75(5):877-886. Kraus, S., Z. Naor, & R. Seger. (2001). Intracellular Signaling Pathways Mediated by the Gonadotropin-Releasing Hormone (GnRH) Receptor. Archives of Medical Research 32(6):499-509.

182

Krebs, C. J., & D. M. Robins. (2010). A Pair of Mouse KRAB Zinc Finger Proteins Modulates Multiple Indicators of Female Reproduction. Biology of Reproduction 82(4):662-668. Krey, L. C., L. Tyrey, & J. W. Everett. (1973). The estrogen-induced advance in the cyclic LH surge in the rat: dependency on ovarian progesterone secretion. Endocrinology 93(2):385-390. Krsmanović, L., S. Stojilković, F. Merelli, S. M. Dufour, M. A. Virmani, & K. J. Catt. (1992). Calcium signaling and episodic secretion of gonadotropin-releasing hormone in hypothalamic neurons. Proceedings of the National Academy of Sciences 89(18):8462- 8466. Kuhla, B., D. Albrecht, R. Bruckmaier, T. Viergutz, G. Nurnberg, & C. C. Metges. (2010). Proteome and radioimmunoassay analyses of pituitary hormones and proteins in response to feed restriction of dairy cows. Proteomics 10(24):4491-4500. Kuhla, B., S. Kuhla, P. E. Rudolph, D. Albrecht, & C. C. Metges. (2007). Proteomics analysis of hypothalamic response to energy restriction in dairy cows. Proteomics 7(19):3602- 3617. Kuhla, B. r., D. Albrecht, S. Kuhla, & C. C. Metges. (2009). Proteome analysis of fatty liver in feed-deprived dairy cows reveals interaction of fuel sensing, calcium, fatty acid, and glycogen metabolism. Physiological genomics 37(2):88-98. Kulik-Rechberger, B., R. Tarkowski, D. A. Magoffin, P. Osemlak, T. Rechberger, & A. J. Jakimiuk. (2008). Leptin gene expression in subcutaneous adipose tissue in girls before and during puberty. European Journal of Obstetrics & Gynecology and Reproductive Biology 136(2):210-214. Kulin, H. E., M. M. Grumbach, & S. L. Kaplan. (1969). Changing Sensitivity of the Pubertal Gonadal Hypothalamic Feedback Mechanism in Man. Science 166(3908):1012-1013. Kurpińska, A., A. Jarosz, & W. Skrzypczak. (2014). Proteomic studies in pregnant and lactating cows. A review. J Anim Feed Sci 23:203-211. Kurtz, A., A. Zimmer, F. Schnutgen, G. Bruning, F. Spener, & T. Muller. (1994). The expression pattern of a novel gene encoding brain-fatty acid binding protein correlates with neuronal and glial cell development. Development 120(9):2637-2649. Kurz, S. G., R. M. Dyer, Y. Hu, M. D. Wright, & M. L. Day. (1990). Regulation of luteinizing hormone secretion in prepubertal heifers fed an energy-deficient diet. Biology of Reproduction 43:450 - 456. Kurzynska, A., M. Bogacki, K. Chojnowska, & I. Bogacka. (2014). Peroxisome proliferator activated receptor ligands affect progesterone and 17beta-estradiol secretion by porcine corpus luteum during early pregnancy. J Physiol Pharmacol 65(5):709-717. Kusminski, C. M., P. E. Bickel, & P. E. Scherer. (2016). Targeting adipose tissue in the treatment of obesity-associated diabetes. Nat Rev Drug Discov 15(9):639-660. Kwon, W.-S., M. S. Rahman, D.-Y. Ryu, Y.-J. Park, & M.-G. Pang. (2015a). Increased male fertility using fertility-related biomarkers. Scientific Reports 5:15654. Kwon, W. S., M. S. Rahman, J. S. Lee, S. J. Yoon, Y. J. Park, & M. G. Pang. (2015b). Discovery of predictive biomarkers for litter size in boar spermatozoa. Mol Cell Proteomics 14(5):1230-1240. Kyeremanteng, C., J. James, J. Mackay, & Z. Merali. (2012). A study of brain and serum brain- derived neurotrophic factor protein in Wistar and Wistar-Kyoto rat strains after electroconvulsive stimulus. Pharmacopsychiatry 45(6):244-249. Lacau-Mengido, I. M., M. E. Mejia, G. S. Diaz-Torga, A. Gonzalez Iglesias, N. Formia, C. Libertun, & D. Becu-Villalobos. (2000). Endocrine studies in ivermectin-treated heifers from birth to puberty. J Anim Sci 78(4):817-824.

183

Lammers, B. P., A. J. Heinrichs, & R. S. Kensinger. (1999). The effects of accelerated growth rates and estrogen implants in prepubertal Holstein heifers on growth, feed efficiency, and blood parameters. J Dairy Sci 82(8):1746-1752. Lapatto, R., J. C. Pallais, D. Zhang, Y. M. Chan, A. Mahan, F. Cerrato, W. W. Le, G. E. Hoffman, & S. B. Seminara. (2007). Kiss1-/- mice exhibit more variable hypogonadism than Gpr54-/- mice. Endocrinology 148(10):4927-4936. Laplante, M., & D. M. Sabatini. (2012). mTOR signaling in growth control and disease. Cell 149(2):274-293. Laporta, J., G. J. Rosa, H. Naya, & M. Carriquiry. (2014). Liver functional genomics in beef cows on grazing systems: novel genes and pathways revealed. Physiol Genomics 46(4):138-147. Larrain, J., D. Bachiller, B. Lu, E. Agius, S. Piccolo, & E. M. De Robertis. (2000). BMP- binding modules in chordin: a model for signalling regulation in the extracellular space. Development 127(4):821-830. Laster, D. B., G. M. Smith, L. V. Cundiff, & K. E. Gregory. (1979). Characterization of biological types of cattle (Cycle II) . II. Postweaning growth and puberty of heifers. J Anim Sci 48(3):500-508. Laster, D. B., G. M. Smith, & K. E. Gregory. (1976). Characterization of biological types of cattle IV. Postweaning growth and puberty of heifers. J Anim Sci 43(1):63-70. Lee, B., & J. Shao. (2014). Adiponectin and energy homeostasis. Reviews in Endocrine and Metabolic Disorders 15(2):149-156. Lee, H.-J., M. Jang, H. Kim, W. Kwak, W. Park, J. Y. Hwang, C.-K. Lee, G. W. Jang, M. N. Park, H.-C. Kim et al. (2013). Comparative Transcriptome Analysis of Adipose Tissues Reveals that ECM-Receptor Interaction Is Involved in the Depot-Specific Adipogenesis in Cattle. PLoS ONE 8(6):e66267. Lee, J. E., S. M. Hollenberg, L. Snider, D. L. Turner, N. Lipnick, & H. Weintraub. (1995). Conversion of Xenopus ectoderm into neurons by neuroD, a basic helix-loop-helix protein. Science 268(5212):836-844. Lee, J. E., J. Y. Lee, H. R. Kim, H. Y. Shin, T. Lin, & D. I. Jin. (2015). Proteomic Analysis of Bovine Pregnancy-specific Serum Proteins by 2D Fluorescence Difference Gel Electrophoresis. Asian-Australasian Journal of Animal Sciences 28(6):788-795. Lee, Y. H., B. Sauer, & F. J. Gonzalez. (1998). Laron dwarfism and non-insulin-dependent diabetes mellitus in the Hnf-1alpha knockout mouse. Mol Cell Biol 18(5):3059-3068. Lesmeister, J. L., P. J. Burfening, & R. L. Blackwell. (1973). Date of First Calving in Beef Cows and Subsequent Calf production. Journal of Animal Science 36(1):1-6. Li, S. (2012). Preliminary study on serum proteomics in periparturient dairy cows. Agricultural Science Research Paper. Animal Husbandry and Veterinary Medicine:121. Licata, L., L. Briganti, D. Peluso, L. Perfetto, M. Iannuccelli, E. Galeota, F. Sacco, A. Palma, A. P. Nardozza, E. Santonico et al. (2012). MINT, the molecular interaction database: 2012 update. Nucleic Acids Res 40(Database issue):D857-861. Lin, J., C. R. Barb, R. L. Matteri, R. R. Kraeling, X. Chen, R. J. Meinersmann, & G. B. Rampacek. (2000). Long form leptin receptor mRNA expression in the brain, pituitary, and other tissues in the pig. Domest Anim Endocrinol 19(1):53-61. Lin, J., C. Richard Barb, R. R. Kraeling, & G. B. Rampacek. (2001). Developmental changes in the long form leptin receptor and related neuropeptide gene expression in the pig brain. Biol Reprod 64(6):1614-1618. Lin, W.-Y., Y.-J. Hu, & Y.-H. Lee. (2008). Hepatocyte nuclear factor-1α regulates glucocorticoid receptor expression to control postnatal body growth. American Journal of Physiology - Gastrointestinal and Liver Physiology 295(3):G542-G551.

184

Lindemann, C., N. Thomanek, F. Hundt, T. Lerari, H. E. Meyer, D. Wolters, & K. Marcus. (2017). Strategies in relative and absolute quantitative mass spectrometry based proteomics. Biological chemistry 398(5-6):687-699. Lindsay, L. A., & C. R. Murphy. (2004). Redistribution of aquaporins in uterine epithelial cells at the time of implantation in the rat. Acta Histochemica 106(4):299-307. Lippolis, J. D., & T. A. Reinhardt. (2005). Proteomic survey of bovine neutrophils. Veterinary immunology and immunopathology 103(1-2):53-65. Liron, J. P., A. J. Prando, M. E. Fernandez, M. V. Ripoli, A. Rogberg-Munoz, D. E. Goszczynski, D. M. Posik, P. Peral-Garcia, A. Baldo, & G. Giovambattista. (2012). Association between GNRHR, LHR and IGF1 polymorphisms and timing of puberty in male Angus cattle. BMC Genet 13:26. Littlejohn, M., T. Grala, K. Sanders, C. Walker, G. Waghorn, K. Macdonald, W. Coppieters, M. Georges, R. Spelman, E. Hillerton et al. (2012). Genetic variation in PLAG1 associates with early life body weight and peripubertal weight and growth in Bos taurus. Anim Genet 43 Liu, H., M. M. Fergusson, J. J. Wu, Rovira, II, J. Liu, O. Gavrilova, T. Lu, J. Bao, D. Han, M. N. Sack et al. (2011). Wnt signaling regulates hepatic metabolism. Sci Signal 4(158):ra6. Liu, H., X. Kong, & F. Chen. (2017). Mkrn3 functions as a novel ubiquitin E3 ligase to inhibit Nptx1 during puberty initiation. Oncotarget 8(49):85102-85109. Liu, Y.-Z., J. Li, R. Pan, H. Shen, Q. Tian, Y. Zhou, Y.-J. Liu, & H.-W. Deng. (2012). Genome- Wide Copy Number Variation Association Analyses for Age at Menarche. The Journal of Clinical Endocrinology and Metabolism 97(11):E2133-E2139. Liu, Y. Z., Y. F. Guo, L. Wang, L. J. Tan, X. G. Liu, Y. F. Pei, H. Yan, D. H. Xiong, F. Y. Deng, N. Yu et al. (2009). Genome-wide association analyses identify SPOCK as a key novel gene underlying age at menarche. PLoS Genet 5(3):e1000420. Liu, Z., & E. J. Barrett. (2002). Human protein metabolism: its measurement and regulation. Am J Physiol Endocrinol Metab 283(6):E1105-1112. Lobaccaro, J. M., D. Gallot, S. Lumbroso, & K. Mouzat. (2013). Liver X Receptors and female reproduction: when cholesterol meets fertility! J Endocrinol Invest 36(1):55-60. Lomniczi, A., J. M. Castellano, H. Wright, B. Selcuk, K. Sonmez, & S. R. Ojeda. 2015a. Gene networks, epigenetics and the control of female puberty, Brain crosstalk in puberty and adolescence. Springer. p. 97-119. Lomniczi, A., H. Wright, J. M. Castellano, V. Matagne, C. A. Toro, S. Ramaswamy, T. M. Plant, & S. R. Ojeda. (2015b). Epigenetic regulation of puberty via Zinc finger protein- mediated transcriptional repression. Nat Commun 6 Lomniczi, A., H. Wright, J. M. Castellano, K. Sonmez, & S. R. Ojeda. (2013). A system biology approach to identify regulatory pathways underlying the neuroendocrine control of female puberty in rats and nonhuman primates. Horm Behav 64(2):175-186. Lomniczi, A., H. Wright, & S. R. Ojeda. (2015c). Epigenetic regulation of female puberty. Frontiers in Neuroendocrinology 36:90-107. Longo, K. A., W. S. Wright, S. Kang, I. Gerin, S. H. Chiang, P. C. Lucas, M. R. Opp, & O. A. MacDougald. (2004). Wnt10b inhibits development of white and brown adipose tissues. J Biol Chem 279(34):35503-35509. Loor, J. J. (2010). Genomics of metabolic adaptations in the peripartal cow. Animal 4(7):1110- 1139. Loor, J. J., H. M. Dann, R. E. Everts, R. Oliveira, C. A. Green, N. A. Guretzky, S. L. Rodriguez- Zas, H. A. Lewin, & J. K. Drackley. (2005). Temporal gene expression profiling of liver from periparturient dairy cows reveals complex adaptive mechanisms in hepatic function. Physiol Genomics 23(2):217-226.

185

Lopez, R., M. G. Thomas, D. M. Hallford, D. H. Keisler, G. A. Silver, B. S. Obeidat, M. D. Garcia, & C. R. Krehbiel. (2006). Case Study: Metabolic Hormone Profiles and Evaluation of Associations of Metabolic Hormones with Body Fat and Reproductive Characteristics of Angus, Brangus, and Brahman Heifers. The Professional Animal Scientist 22(3):273-282. Lopez-Rodriguez, C., C. L. Antos, J. M. Shelton, J. A. Richardson, F. M. Lin, T. I. Novobrantseva, R. T. Bronson, P. Igarashi, A. Rao, & E. N. Olson. (2004). Loss of NFAT5 results in renal atrophy and lack of tonicity-responsive gene expression. Proceedings of the National Academy of Sciences of the United States of America 101(8):2392-2397. Lu, X., S. Guo, Y. Cheng, J. H. Kim, Y. Feng, & Y. Feng. (2017). Stimulation of ovarian follicle growth after AMPK inhibition. Reproduction 153(5):683-694. Luna-Nevarez, P., G. Rincon, J. F. Medrano, D. G. Riley, C. C. Chase, Jr., S. W. Coleman, D. M. Vanleeuwen, K. L. DeAtley, A. Islas-Trejo, G. A. Silver et al. (2011). Single nucleotide polymorphisms in the growth hormone-insulin-like growth factor axis in straightbred and crossbred Angus, Brahman, and Romosinuano heifers: population genetic analyses and association of genotypes with reproductive phenotypes. J Anim Sci 89(4):926-934. Lunstra, D. D., & L. V. Cundiff. (2003). Growth and pubertal development in Brahman-, Boran-, Tuli-, Belgian Blue-, Hereford- and Angus-sired F1 bulls. J Anim Sci 81(6):1414-1426. Lunstra, D. D., & S. E. Echternkamp. (1982). Puberty in beef bulls: acrosome morphology and semen quality in bulls of different breeds. J Anim Sci 55(3):638-648. Luque, R. M., R. D. Kineman, & M. Tena-Sempere. (2007). Regulation of hypothalamic expression of KiSS-1 and GPR54 genes by metabolic factors: analyses using mouse models and a cell line. Endocrinology 148(10):4601-4611. Ma, Q., C. Kintner, & D. J. Anderson. (1996). Identification of neurogenin, a vertebrate neuronal determination gene. Cell 87(1):43-52. Macdonald, K. A., J. W. Penno, A. M. Bryant, & J. R. Roche. (2005). Effect of feeding level pre- and post-puberty and body weight at first calving on growth, milk production, and fertility in grazing dairy cows. J Dairy Sci 88(9):3363-3375. MacGregor, R. G., & N. H. Casey. (1999). Evaluation of calving interval and calving date as measures of reproductive performance in a beef herd. Livestock Production Science 57(2):181-191. Maciel, M. N., D. A. Zieba, M. Amstalden, D. H. Keisler, J. P. Neves, & G. L. Williams. (2004). Chronic administration of recombinant ovine leptin in growing beef heifers: effects on secretion of LH, metabolic hormones, and timing of puberty. J Anim Sci 82(10):2930-2936. Mahesh, V. B., K. M. Dhandapani, & D. W. Brann. (2006). Role of astrocytes in reproduction and neuroprotection. Mol Cell Endocrinol 246(1-2):1-9. Maier, T., M. Güell, & L. Serrano. (2009). Correlation of mRNA and protein in complex biological samples. FEBS Letters 583(24):3966-3973. Maillo, V., C. de Frutos, P. O'Gaora, N. Forde, G. W. Burns, T. E. Spencer, A. Gutierrez-Adan, P. Lonergan, & D. Rizos. (2016). Spatial differences in gene expression in the bovine oviduct. Reproduction 152(1):37-46. Maillo, V., P. O. Gaora, N. Forde, U. Besenfelder, V. Havlicek, G. W. Burns, T. E. Spencer, A. Gutierrez-Adan, P. Lonergan, & D. Rizos. (2015). Oviduct-Embryo Interactions in Cattle: Two-Way Traffic or a One-Way Street? Biol Reprod 92(6):144.

186

Malagón, M. M., F. Rodríguez-Pacheco, A. J. Martínez-Fuentes, S. Tovar, L. Pinilla, M. Tena- Sempere, C. Diéguez, & J. P. Castaño. (2006). Regulation of pituitary cell function by the adipokine adiponectin. Neuroendocrinology 27 Mallat, Z., C. Heymes, J. Ohan, E. Faggin, G. Leseche, & A. Tedgui. (1999). Expression of interleukin-10 in advanced human atherosclerotic plaques: relation to inducible nitric oxide synthase expression and cell death. Arterioscler Thromb Vasc Biol 19(3):611- 616. Manzoni, C., D. A. Kia, J. Vandrovcova, J. Hardy, N. W. Wood, P. A. Lewis, & R. Ferrari. (2016). Genome, transcriptome and proteome: the rise of omics data and their integration in biomedical sciences. Briefings in bioinformatics:bbw114. Mao, Y., D. L. Hopkins, Y. Zhang, P. Li, L. Zhu, P. Dong, R. Liang, J. Dai, X. Wang, & X. Luo. (2016). Beef quality with different intramuscular fat content and proteomic analysis using isobaric tag for relative and absolute quantitation of differentially expressed proteins. Meat Science 118:96-102. Marfil, V., M. Moya, C. E. Pierreux, J. V. Castell, F. P. Lemaigre, F. X. Real, & R. Bort. (2010). Interaction between Hhex and SOX13 Modulates Wnt/TCF Activity. Journal of Biological Chemistry 285(8):5726-5737. Marioni, J. C., C. E. Mason, S. M. Mane, M. Stephens, & Y. Gilad. (2008). RNA-seq: an assessment of technical reproducibility and comparison with gene expression arrays. Genome Res 18(9):1509-1517. Martin, L. C., J. S. Brinks, R. M. Bourdon, & L. V. Cundiff. (1992). Genetic effects on beef heifer puberty and subsequent reproduction. J Anim Sci 70(12):4006-4017. Martins, R. P., M. Collado-Romero, M. Martinez-Gomariz, A. Carvajal, C. Gil, C. Lucena, A. Moreno, & J. J. Garrido. (2012). Proteomic analysis of porcine mesenteric lymph-nodes after Salmonella typhimurium infection. J Proteomics 75(14):4457-4470. Mason, W. T., S. R. Rawlings, P. Cobbett, S. K. Sikdar, R. Zorec, S. N. Akerman, C. D. Benham, M. J. Berridge, T. Cheek, & R. B. Moreton. (1988). Control of secretion in anterior pituitary cells--linking ion channels, messengers and exocytosis. J Exp Biol 139:287-316. Massague, J. (2012). TGF[beta] signalling in context. Nat Rev Mol Cell Biol 13(10):616-630. Matera, A. G., R. M. Terns, & M. P. Terns. (2007). Non-coding RNAs: lessons from the small nuclear and small nucleolar RNAs. Nature Reviews Molecular Cell Biology 8:209. Matkovic, V., J. Z. Ilich, M. Skugor, N. E. Badenhop, P. Goel, A. Clairmont, D. Klisovic, R. W. Nahhas, & J. D. Landoll. (1997). Leptin is inversely related to age at menarche in human females. J Clin Endocrinol Metab 82(10):3239-3245. Matthews, V. B., M. B. Astrom, M. H. Chan, C. R. Bruce, K. S. Krabbe, O. Prelovsek, T. Akerstrom, C. Yfanti, C. Broholm, O. H. Mortensen et al. (2009). Brain-derived neurotrophic factor is produced by skeletal muscle cells in response to contraction and enhances fat oxidation via activation of AMP-activated protein kinase. Diabetologia 52(7):1409-1418. Mayer, C., M. Acosta-Martinez, S. L. Dubois, A. Wolfe, S. Radovick, U. Boehm, & J. E. Levine. (2010). Timing and completion of puberty in female mice depend on estrogen receptor α-signaling in kisspeptin neurons. Proc Natl Acad Sci USA 107(52):22693- 22698. McCarthy, S. D., S. M. Waters, D. A. Kenny, M. G. Diskin, R. Fitzpatrick, J. Patton, D. C. Wathes, & D. G. Morris. (2010). Negative energy balance and hepatic gene expression patterns in high-yielding dairy cows during the early postpartum period: a global approach. Physiol Genomics 42a(3):188-199. McDowall, M. D., M. S. Scott, & G. J. Barton. (2009). PIPs: human protein-protein interaction prediction database. Nucleic Acids Res 37(Database issue):D651-656.

187

McGettigan, P. A., J. A. Browne, S. D. Carrington, M. A. Crowe, T. Fair, N. Forde, B. J. Loftus, A. Lohan, P. Lonergan, K. Pluta et al. (2016). Fertility and genomics: comparison of gene expression in contrasting reproductive tissues of female cattle. Reproduction Fertility and Development 28(1-2):11-24. McLean, I., P. Holmes, & D. Counsell. 2013. Final Report: The Northern beef report, 2013 Northern beef situation analysis (B.COM.0348), Sydney. McLin, V. A., S. A. Rankin, & A. M. Zorn. (2007). Repression of Wnt/β-catenin signaling in the anterior endoderm is essential for liver and pancreas development. Development 134(12):2207. McLuckey, S. A., G. J. Van Berkel, D. E. Goeringer, & G. L. Glish. (1994a). Ion trap mass spectrometry of externally generated ions. Analytical Chemistry 66(13):689A-696A. McLuckey, S. A., G. J. Van Berkel, D. E. Goeringer, & G. L. Glish. (1994b). Ion trap mass spectrometry using high-pressure ionization. Analytical chemistry 66(14):737A-743A. McMinn, J. E., S. M. Liu, H. Liu, I. Dragatsis, P. Dietrich, T. Ludwig, C. N. Boozer, & S. C. Chua, Jr. (2005). Neuronal deletion of Lepr elicits diabesity in mice without affecting cold tolerance or fertility. Am J Physiol Endocrinol Metab 289(3):E403-411. Meaker, H. J., T. P. N. Coetsee, & A. W. Lishman. (1980). The effects of age at first calving on the productive and reproductive performance of beef cows. South African Journal of Animal Science 10(2):105-113. Meece, A., S. He, & C. Keefer. (2007). 272 expression of Lef1 transcription factor and its novel isoforms in pre-implantation embryos. Reproduction, Fertility and Development 19(1):252-252. Melkman-Zehavi, T., R. Oren, S. Kredo-Russo, T. Shapira, A. D. Mandelbaum, N. Rivkin, T. Nir, K. A. Lennox, M. A. Behlke, Y. Dor et al. (2011). miRNAs control insulin content in pancreatic β-cells via downregulation of transcriptional repressors. The EMBO Journal 30(5):835-845. Mendel, D. B., & G. R. Crabtree. (1991). HNF-1, a member of a novel class of dimerizing homeodomain proteins. J Biol Chem 266(2):677-680. Meng, Q., K. Wang, X. Liu, H. Zhou, L. Xu, Z. Wang, & M. Fang. (2017). Identification of growth trait related genes in a Yorkshire purebred pig population by genome-wide association studies. Asian-Australasian Journal of Animal Sciences 30(4):462-469. Messina, A., F. Langlet, K. Chachlaki, J. Roa, S. Rasika, N. Jouy, S. Gallet, F. Gaytan, J. Parkash, M. Tena-Sempere et al. (2016). A microRNA switch regulates the rise in hypothalamic GnRH production before puberty. Nat Neurosci 19(6):835-844. Metzker, M. L. (2010). Sequencing technologies - the next generation. Nat Rev Genet 11(1):31-46. Mezzadra, C., A. Homse, D. Sampedro, & R. Alberio. (1993). Pubertal traits and seasonal variation of the sexual activity in Brahman, Hereford and crossbred heifers. Theriogenology 40(5):987-996. Micevych, P., K. Sinchak, R. H. Mills, L. Tao, P. LaPolt, & J. K. Lu. (2003). The luteinizing hormone surge is preceded by an estrogen-induced increase of hypothalamic progesterone in ovariectomized and adrenalectomized rats. Neuroendocrinology 78(1):29-35. Michalakis, K. G., & J. H. Segars. (2010). The role of adiponectin in reproduction: from polycystic ovary syndrome to assisted reproduction. Fertility and Sterility 94(6):1949- 1957. Michels, K. M., W. H. Lee, A. Seltzer, J. M. Saavedra, & C. A. Bondy. (1993). Up-regulation of pituitary [125I]insulin-like growth factor-I (IGF-I) binding and IGF binding protein- 2 and IGF-I gene expression by estrogen. Endocrinology 132(1):23-29.

188

Milazzotto, M. P., P. Rahal, M. Nichi, T. Miranda-Neto, L. A. Teixeira, J. B. S. Ferraz, J. P. Eler, F. Campagnari, & J. F. Garcia. (2008). New molecular variants of hypothalamus– pituitary–gonad axis genes and their association with early puberty phenotype in Bos taurus indicus (Nellore). Livestock Science 114(2):274-279. Miner, J. L. (1992). Recent advances in the central control of intake in ruminants. J Anim Sci 70(4):1283-1289. Minhas V., & S. H.M. (2008). Detection Of A Uterine Secretory Protein Unique To Uninfected Repeat Breeder (Infertile) Cows By SDS – PAGE. Internet Journal of Veterinary Medicine 4(2):1-4. Minten, M. A., T. R. Bilby, R. G. S. Bruno, C. C. Allen, C. A. Madsen, Z. Wang, J. E. Sawyer, A. Tibary, H. L. Neibergs, T. W. Geary et al. (2013). Effects of Fertility on Gene Expression and Function of the Bovine Endometrium. PLOS ONE 8(8):e69444. Mircea, C. N., M. E. Lujan, & R. A. Pierson. (2007). Metabolic fuel and clinical implications for female reproduction. J Obstet Gynaecol Can 29(11):887-902. Misirlioglu, M., G. P. Page, H. Sagirkaya, A. Kaya, J. J. Parrish, N. L. First, & E. Memili. (2006). Dynamics of global transcriptome in bovine matured oocytes and preimplantation embryos. Proceedings of the National Academy of Sciences 103(50):18905-18910. Mitsushima, D., D. L. Hei, & E. Terasawa. (1994). gamma-Aminobutyric acid is an inhibitory neurotransmitter restricting the release of luteinizing hormone-releasing hormone before the onset of puberty. Proc Natl Acad Sci U S A 91(1):395-399. Mitsushima, D., & F. Kimura. (1997). The maturation of GABA(A) receptor-mediated control of luteinizing hormone secretion in immature male rats. Brain Res 748(1-2):258-262. Miwa, H., A. Miyake, Y. Kouta, A. Shimada, Y. Yamashita, Y. Nakayama, H. Yamauchi, M. Konishi, & N. Itoh. (2009). A novel neural-specific BMP antagonist, Brorin-like, of the Chordin family. FEBS Lett 583(22):3643-3648. Moenter, S. M., Z. Chu, & C. A. Christian. (2009). Neurobiological mechanisms underlying oestradiol negative and positive feedback regulation of gonadotropin-releasing hormone (GnRH) neurones. Journal of neuroendocrinology 21(4):327-333. Moenter, S. M., & R. A. DeFazio. (2005). Endogenous gamma-aminobutyric acid can excite gonadotropin-releasing hormone neurons. Endocrinology 146(12):5374-5379. Montagner, P., A. R. T. Krause, E. Schwegler, M. M. Weschenfelder, V. R. Rabassa, A. Schneider, R. A. Pereira, C. C. Brauner, F. A. B. Del Pino, F. M. Gonçalves et al. (2016). Reduction of liver function delays resumption of postpartum ovarian activity and alters the synthesis of acute phase proteins in dairy cows. Research in Veterinary Science 106:84-88. Moore, S. G., J. E. Pryce, B. J. Hayes, A. J. Chamberlain, K. E. Kemper, D. P. Berry, M. McCabe, P. Cormican, P. Lonergan, T. Fair et al. (2016). Differentially Expressed Genes in Endometrium and Corpus Luteum of Holstein Cows Selected for High and Low Fertility Are Enriched for Sequence Variants Associated with Fertility. Biol Reprod 94(1):19. Moran, C., J. F. Quirke, & J. F. Roche. (1989). Puberty in heifers: A Review. Animal Reproduction Science 18(1–3):167-182. Morash, B., D. Wilkinson, P. Murphy, E. Ur, & M. Wilkinson. (2001). Developmental regulation of leptin gene expression in rat brain and pituitary. Molecular and Cellular Endocrinology 185(1–2):151-159. Morrison, C. D., J. A. Daniel, J. H. Hampton, P. R. Buff, T. M. McShane, M. G. Thomas, & D. H. Keisler. (2003). Luteinizing hormone and growth hormone secretion in ewes infused intracerebroventricularly with neuropeptide Y. Domest Anim Endocrinol 24(1):69-80.

189

Morrison, C. D., G. J. Morton, K. D. Niswender, R. W. Gelling, & M. W. Schwartz. (2005). Leptin inhibits hypothalamic Npy and Agrp gene expression via a mechanism that requires phosphatidylinositol 3-OH-kinase signaling. Am J Physiol Endocrinol Metab 289(6):E1051-1057. Mortazavi, A., B. A. Williams, K. McCue, L. Schaeffer, & B. Wold. (2008). Mapping and quantifying mammalian transcriptomes by RNA-Seq. Nat Methods 5(7):621-628. Morton, G. J., R. W. Gelling, K. D. Niswender, C. D. Morrison, C. J. Rhodes, & M. W. Schwartz. (2005). Leptin regulates insulin sensitivity via phosphatidylinositol-3-OH kinase signaling in mediobasal hypothalamic neurons. Cell Metab 2(6):411-420. Moschos, S., J. L. Chan, & C. S. Mantzoros. (2002). Leptin and reproduction: a review. Fertil Steril 77(3):433-444. Munoz, M., F. J. Corrales, J. N. Caamano, C. Diez, B. Trigal, M. I. Mora, D. Martin, S. Carrocera, & E. Gomez. (2012). Proteome of the early embryo-maternal dialogue in the cattle uterus. J Proteome Res 11(2):751-766. Murgiano, L., A. D’Alessandro, M. G. Egidi, A. Crisa, G. Prosperini, A. M. Timperio, A. Valentini, & L. Zolla. (2010). Proteomics and transcriptomics investigation on longissimus muscles in Large White and Casertana pig breeds. Journal of proteome research 9(12):6450-6466. Nascimento, A. V., M. C. Matos, L. O. Seno, A. R. Romero, J. F. Garcia, & A. B. Grisolia. (2016). Genome wide association study on early puberty in Bos indicus. Genet Mol Res 15(1) Navarro, V. M., J. M. Castellano, R. Fernandez-Fernandez, M. L. Barreiro, J. Roa, J. E. Sanchez-Criado, E. Aguilar, C. Dieguez, L. Pinilla, & M. Tena-Sempere. (2004). Developmental and hormonally regulated messenger ribonucleic acid expression of KiSS-1 and its putative receptor, GPR54, in rat hypothalamus and potent luteinizing hormone-releasing activity of KiSS-1 peptide. Endocrinology 145(10):4565-4574. Navarro, V. M., M. L. Gottsch, C. Chavkin, H. Okamura, D. K. Clifton, & R. A. Steiner. (2009). Regulation of gonadotropin-releasing hormone secretion by kisspeptin/dynorphin/neurokinin B neurons in the arcuate nucleus of the mouse. J Neurosci 29(38):11859-11866. Nelson, V. L., A. L. Negron, I. Reid, J. A. Thomas, L. Yang, R. Z. Lin, & M. Acosta-Martinez. (2017). Loss of PI3K p110alpha in the Adipose Tissue Results in Infertility and Delayed Puberty Onset in Male Mice. Biomed Res Int 2017:3756089. Nguyen, L. T., A. Reverter, A. Cánovas, B. K. Venus, A. Islas-Trejo, L. R. Porto-Neto, S. A. Lehnert, J. F. Medrano, S. S. Moore, & M. R. S. Fortes. (2017). Global differential gene expression in the pituitary gland and the ovaries of pre- and post-pubertal Brahman heifers. Journal Animal Science 95:599-615. Nguyen, L. T., L. F. Zacchi, B. L. Schulz, S. S. Moore, & M. R. S. Fortes. (2018). Adipose tissue proteomic analyses to study puberty in Brahman heifers. J Anim Sci Nicola, C., P. K. Lala, & C. Chakraborty. (2008). Prostaglandin E2-mediated migration of human trophoblast requires RAC1 and CDC42. Biol Reprod 78(6):976-982. Nie, L., G. Wu, D. E. Culley, J. C. Scholten, & W. Zhang. (2007). Integrative analysis of transcriptomic and proteomic data: challenges, solutions and applications. Crit Rev Biotechnol 27(2):63-75. Nie, L., G. Wu, & W. Zhang. (2006). Correlation between mRNA and protein abundance in Desulfovibrio vulgaris: A multiple regression to identify sources of variations. Biochemical and Biophysical Research Communications 339(2):603-610. Nishida, M., T. Moriyama, Y. Sugita, & K. Yamauchi-Takihara. (2007). Interleukin-10 associates with adiponectin predominantly in subjects with metabolic syndrome. Circ J 71(8):1234-1238.

190

Nishimura, S., T. Watanabe, K. Mizoshita, K. Tatsuda, T. Fujita, N. Watanabe, Y. Sugimoto, & A. Takasuga. (2012). Genome-wide association study identified three major QTL for carcass weight including the PLAG1-CHCHD7 QTN for stature in Japanese Black cattle. BMC Genetics 13(1):1-11. Noel, S. D., & U. B. Kaiser. (2011). G protein-coupled receptors involved in GnRH regulation: molecular insights from human disease. Mol Cell Endocrinol 346(1-2):91-101. Nogueira, G. P. (2004). Puberty in South American Bos indicus (Zebu) cattle. Anim Reprod Sci 82-83:361-372. Nogueira, G. P., R. F. S. de Lucia, F. V. Pereira, & P. D. Cirilo. (2003). Precocious fertility in Nelore heifers. Biol. Reprod. 68:382. Nuno-Ayala, M., N. Guillen, C. Arnal, J. M. Lou-Bonafonte, A. de Martino, J. A. Garcia-de- Jalon, S. Gascon, L. Osaba, J. Osada, & M. A. Navarro. (2012). Cystathionine beta- synthase deficiency causes infertility by impairing decidualization and gene expression networks in uterus implantation sites. Physiol Genomics 44(14):702-716. O’Donnell, R., J. W. Holland, H. C. Deeth, & P. Alewood. (2004). Milk proteomics. International Dairy Journal 14(12):1013-1023. O’Sullivan, A. J. 2012. Fat Storage in Women: From Puberty to the Menopause. In: V. R. Preedy, editor, Handbook of Growth and Growth Monitoring in Health and Disease. Springer New York, New York, NY. p. 1087-1101. Oakley, A. E., D. K. Clifton, & R. A. Steiner. (2009). Kisspeptin signaling in the brain. Endocr Rev 30(6):713-743. Obici, S., & L. Rossetti. (2003). Minireview: nutrient sensing and the regulation of insulin action and energy balance. Endocrinology 144(12):5172-5178. Ochoa, A., C. Domenzain, C. Clapp, & G. Martinez de la Escalera. (1997). Differential effects of basic fibroblast growth factor, epidermal growth factor, transforming growth factor- alpha, and insulin-like growth factor-I on a hypothalamic gonadotropin-releasing hormone neuronal cell line. J Neurosci Res 49(6):739-749. Oelgeschlager, M., J. Larrain, D. Geissert, & E. M. De Robertis. (2000). The evolutionarily conserved BMP-binding protein Twisted gastrulation promotes BMP signalling. Nature 405(6788):757-763. Ohyama, Y., M. Katafuchi, A. Almehmadi, S. Venkitapathi, H. Jaha, J. Ehrenman, J. Morcos, R. Aljamaan, & Y. Mochida. (2012). Modulation of matrix mineralization by Vwc2- like protein and its novel splicing isoforms. Biochemical and Biophysical Research Communications 418(1):12-16. Ojeda, S. R., C. Dubay, A. Lomniczi, G. Kaidar, V. Matagne, U. S. Sandau, & G. A. Dissen. (2010a). Gene networks and the neuroendocrine regulation of puberty. Mol Cell Endocrinol 324(1-2):3-11. Ojeda, S. R., A. Lomniczi, C. Mastronardi, S. Heger, C. Roth, A. S. Parent, V. Matagne, & A. E. Mungenast. (2006). Minireview: the neuroendocrine regulation of puberty: is the time ripe for a systems biology approach? Endocrinology 147(3):1166-1174. Ojeda, S. R., A. Lomniczi, & U. Sandau. (2010b). Contribution of glial-neuronal interactions to the neuroendocrine control of female puberty. Eur J Neurosci 32(12):2003-2010. Ojeda, S. R., A. Lomniczi, & U. S. Sandau. (2008). Glial-gonadotrophin hormone (GnRH) neurone interactions in the median eminence and the control of GnRH secretion. J Neuroendocrinol 20(6):732-742. Ojeda, S. R., & M. K. Skinner. 2006. Puberty in the rat, Knobil and Neill's Physiology of Reproduction. Elsevier Inc. p. 2061-2126. Oltmanns, K. M., B. Fruehwald-Schultes, W. Kern, J. Born, H. L. Fehm, & A. Peters. (2001). Hypoglycemia, but not insulin, acutely decreases LH and T secretion in men. J Clin Endocrinol Metab 86(10):4913-4919.

191

Ong, K. K., C. E. Elks, S. Li, J. H. Zhao, J. a. Luan, L. B. Andersen, S. A. Bingham, S. Brage, G. D. Smith, U. Ekelund et al. (2009). Genetic variation in LIN28B is associated with the timing of puberty. Nature Genetics 41:729. Ota, T., H. Asahina, S.-H. Park, Q. Huang, T. Minegishi, N. Auersperg, & P. C. Leung. (2008). HOX cofactors expression and regulation in the human ovary. Reproductive Biology and Endocrinology 6(1):49. Pal, B., Y. Chen, F. Vaillant, P. Jamieson, L. Gordon, A. C. Rios, S. Wilcox, N. Fu, K. H. Liu, F. C. Jackling et al. (2017). Construction of developmental lineage relationships in the mouse mammary gland by single-cell RNA profiling. Nat Commun 8(1):1627. Pan, W., W. A. Banks, M. B. Fasold, J. Bluth, & A. J. Kastin. (1998). Transport of brain- derived neurotrophic factor across the blood-brain barrier. Neuropharmacology 37(12):1553-1561. Pancholi, V. (2001). Multifunctional alpha-enolase: its role in diseases. Cell Mol Life Sci 58(7):902-920. Papatriantafyllou, M. (2012). WNT chews the fat with glucose uptake. Nature Reviews Molecular Cell Biology 13:339. Pardridge, W. M., D. Wu, & T. Sakane. (1998). Combined Use of Carboxyl-Directed Protein Pegylation and Vector-Mediated Blood-Brain Barrier Drug Delivery System Optimizes Brain Uptake of Brain-Derived Neurotrophic Factor Following Intravenous Administration. Pharmaceutical Research 15(4):576-582. Parent, A.-S., G. Rasier, V. Matagne, A. Lomniczi, M.-C. Lebrethon, A. Gérard, S. R. Ojeda, & J.-P. Bourguignon. (2008). Oxytocin Facilitates Female Sexual Maturation through a Glia-to-Neuron Signaling Pathway. Endocrinology 149(3):1358-1365. Parent, A. S., A. E. Mungenast, A. Lomniczi, U. S. Sandau, E. Peles, M. A. Bosch, O. K. Ronnekleiv, & S. R. Ojeda. (2007). A contactin-receptor-like protein tyrosine phosphatase beta complex mediates adhesive communication between astroglial cells and gonadotrophin-releasing hormone neurones. J Neuroendocrinol 19(11):847-859. Parent, A. S., G. Teilmann, A. Juul, N. E. Skakkebaek, J. Toppari, & J. P. Bourguignon. (2003). The timing of normal puberty and the age limits of sexual precocity: variations around the world, secular trends, and changes after migration. Endocr Rev 24(5):668-693. Park, E. J., K. A. Ji, S. B. Jeon, W. H. Choi, I. O. Han, H. J. You, J. H. Kim, I. Jou, & E. H. Joe. (2004). Rac1 contributes to maximal activation of STAT1 and STAT3 in IFN- gamma-stimulated rat astrocytes. J Immunol 173(9):5697-5703. Park, J. P., A. Raafat, J. A. Feltracco, W. M. Blanding, & B. W. Booth. (2013). Differential gene expression in nuclear label-retaining cells in the developing mouse mammary gland. Stem Cells Dev 22(8):1297-1306. Parkash, J., & G. Kaur. (2005). Neuronal-glial plasticity in gonadotropin-releasing hormone release in adult female rats: role of the polysialylated form of the neural cell adhesion molecule. J Endocrinol 186(2):397-409. Parreira, J. R., & S. de Sousa Araújo. 2018. Studying the Animal Transcriptome: State of the Art and Challenges in the Context of Animal and Veterinary Sciences. In: d. A. A., E. D. and M. I., editors, Proteomics in Domestic Animals: from Farm to Systems Biology. Springer, Cham. p. 421-446. Patton, J., D. A. Kenny, S. McNamara, J. F. Mee, F. P. O'Mara, M. G. Diskin, & J. J. Murphy. (2007). Relationships among milk production, energy balance, plasma analytes, and reproduction in Holstein-Friesian cows. J Dairy Sci 90(2):649-658. Payne, S. H. (2015). The utility of protein and mRNA correlation. Trends Biochem Sci 40(1):1- 3. Pedersen, B. K., M. Pedersen, K. S. Krabbe, H. Bruunsgaard, V. B. Matthews, & M. A. Febbraio. (2009). Role of exercise-induced brain-derived neurotrophic factor

192

production in the regulation of energy homeostasis in mammals. Exp Physiol 94(12):1153-1160. Peng, L., P. Rawson, D. McLauchlan, K. Lehnert, R. Snell, & T. W. Jordan. (2008). Proteomic analysis of microsomes from lactating bovine mammary gland. J Proteome Res 7(4):1427-1432. Pérez-Enciso, M., J. C. Rincón, & A. Legarra. (2015). Sequence- vs. chip-assisted genomic selection: accurate biological information is advised. Genetics Selection Evolution 47(1):1-14. Perry, G. A., & R. Cushman. (2013). Effect of age at puberty/conception date on cow longevity. Vet Clin North Am Food Anim Pract 29(3):579-590. Perry, J. R., L. Stolk, N. Franceschini, K. L. Lunetta, G. Zhai, P. F. McArdle, A. V. Smith, T. Aspelund, S. Bandinelli, E. Boerwinkle et al. (2009a). Meta-analysis of genome-wide association data identifies two loci influencing age at menarche. Nat Genet Perry, J. R. B., F. Day, C. E. Elks, P. Sulem, D. J. Thompson, T. Ferreira, C. He, D. I. Chasman, T. Esko, G. Thorleifsson et al. (2014a). Parent-of-origin specific allelic associations among 106 genomic loci for age at menarche. Nature 514(7520):92-97. Perry, J. R. B., F. Day, C. E. Elks, P. Sulem, D. J. Thompson, T. Ferreira, C. Y. He, D. I. Chasman, T. Esko, G. Thorleifsson et al. (2014b). Parent-of-origin-specific allelic associations among 106 genomic loci for age at menarche. Nature 514(7520):92-+. Perry, J. R. B., L. Stolk, N. Franceschini, K. L. Lunetta, G. Zhai, P. F. McArdle, A. V. Smith, T. Aspelund, S. Bandinelli, E. Boerwinkle et al. (2009b). Meta-analysis of genome- wide association data identifies two loci influencing age at menarche. Nature genetics 41(6):648-650. Peter, A. R., & P. J. H. Ball. 1987. Reproduction in cattle. 2nd Ed ed. Blackwell Science, Germany. Petersen, S. L., S. McCrone, M. Keller, & S. Shores. (1995). Effects of estrogen and progesterone on luteinizing hormone-releasing hormone messenger ribonucleic acid levels: consideration of temporal and neuroanatomical variables. Endocrinology 136(8):3604-3610. Petersen, S. L., E. N. Ottem, & C. D. Carpenter. (2003). Direct and indirect regulation of gonadotropin-releasing hormone neurons by estradiol. Biol Reprod 69(6):1771-1778. Peyton, M., C. M. Stellrecht, F. J. Naya, H. P. Huang, P. J. Samora, & M. J. Tsai. (1996). BETA3, a novel helix-loop-helix protein, can act as a negative regulator of BETA2 and MyoD-responsive genes. Molecular and Cellular Biology 16(2):626-633. Picard, B., C. Berri, L. Lefaucheur, C. Molette, T. Sayd, & C. Terlouw. (2010). Skeletal muscle proteomics in livestock production. Briefings in functional genomics 9(3):259-278. Pikuleva, I. A., & M. R. Waterman. (2013). Cytochromes P450: Roles in Diseases. Journal of Biological Chemistry 288(24):17091-17098. Pineda, R., E. Aguilar, L. Pinilla, & M. Tena-Sempere. 2010. Chapter 5 - Physiological Roles of the Kisspeptin/GPR54 System in the Neuroendocrine Control of Reproduction. In: L. Martini, editor, Progress in Brain Research No. 181. Elsevier. p. 55-77. Plaseska-Karanfilska, D., P. Noveski, T. Plaseski, I. Maleva, S. Madjunkova, & Z. Moneva. (2012). Genetic Causes of Male Infertility. Balkan Journal of Medical Genetics : BJMG 15(Suppl):31-34. Poduslo, J. F., & G. L. Curran. (1996). Permeability at the blood-brain and blood-nerve barriers of the neurotrophic factors: NGF, CNTF, NT-3, BDNF. Molecular Brain Research 36(2):280-286. Pontoglio, M., J. Barra, M. Hadchouel, A. Doyen, C. Kress, J. P. Bach, C. Babinet, & M. Yaniv. (1996). Hepatocyte nuclear factor 1 inactivation results in hepatic dysfunction, phenylketonuria, and renal Fanconi syndrome. Cell 84(4):575-585.

193

Porto Neto, L. R., S. A. Lehnert, M. R. S. Fortes, M. Kelly, & A. Reverter. (2013). Population Stratification and Breed Composition of Australian Tropically Adapted Cattle. In: Proceedings of the Association for the Advancement of Animal Breeding and Genetics Napier, New Zealand. p 147-150. Porto-Neto, L. R., A. Reverter, K. C. Prayaga, E. K. F. Chan, D. J. Johnston, R. J. Hawken, G. Fordyce, J. F. Garcia, T. S. Sonstegard, S. Bolormaa et al. (2014). The Genetic Architecture of Climatic Adaptation of Tropical Cattle. PLOS ONE 9(11):e113284. Porto-Neto, L. R., T. S. Sonstegard, G. E. Liu, D. M. Bickhart, M. V. Da Silva, M. A. Machado, Y. T. Utsunomiya, J. F. Garcia, C. Gondro, & C. P. Van Tassell. (2013). Genomic divergence of zebu and taurine cattle identified through high-density SNP genotyping. BMC Genomics 14(1):876. Prevot, V. (2002). Glial-neuronal-endothelial interactions are involved in the control of GnRH secretion. J Neuroendocrinol 14(3):247-255. Psilopanagioti, A., H. Papadaki, E. F. Kranioti, T. K. Alexandrides, & J. N. Varakis. (2009). Expression of adiponectin and adiponectin receptors in human pituitary gland and brain. Neuroendocrinology 89(1):38-47. Pyo, J., S. I. Hwang, J. Oh, S. J. Lee, S. C. Kang, J. S. Kim, & J. Lim. (2003). Characterization of a bovine pregnancy-associated protein using two-dimensional gel electrophoresis, N-terminal sequencing and mass spectrometry. Proteomics 3(12):2420-2427. Pyun, J. A., S. Kim, D. H. Cha, J. J. Ko, & K. Kwack. (2012). Epistasis between the HSD17B4 and TG polymorphisms is associated with premature ovarian failure. Fertil Steril 97(4):968-973. Pyun, J. A., S. Kim, N. H. Cho, I. Koh, J. Y. Lee, C. Shin, & K. Kwack. (2014). Genome-wide association studies and epistasis analyses of candidate genes related to age at menarche and age at natural menopause in a Korean population. Menopause 21(5):522-529. Qian, H., C. R. Barb, M. M. Compton, G. J. Hausman, M. J. Azain, R. R. Kraeling, & C. A. Baile. (1999). Leptin mRNA expression and serum leptin concentrations as influenced by age, weight, and estradiol in pigs. Domest Anim Endocrinol 16(2):135-143. Qian, X., L. Jin, J. P. Grande, & R. V. Lloyd. (1996). Transforming growth factor-beta and p27 expression in pituitary cells. Endocrinology 137(7):3051-3060. Qiu, J., Y. Fang, M. A. Bosch, O. K. Ronnekleiv, & M. J. Kelly. (2011). Guinea pig kisspeptin neurons are depolarized by leptin via activation of TRPC channels. Endocrinology 152(4):1503-1514. Qiu, S., J. T. Vazquez, E. Boulger, H. Liu, P. Xue, M. A. Hussain, & A. Wolfe. (2017). Hepatic estrogen receptor α is critical for regulation of gluconeogenesis and lipid metabolism in males. Scientific Reports 7(1):1661. Quennell, J. H., C. S. Howell, J. Roa, R. A. Augustine, D. R. Grattan, & G. M. Anderson. (2011). Leptin deficiency and diet-induced obesity reduce hypothalamic kisspeptin expression in mice. Endocrinology 152(4):1541-1550. Quennell, J. H., A. C. Mulligan, A. Tups, X. Liu, S. J. Phipps, C. J. Kemp, A. E. Herbison, D. R. Grattan, & G. M. Anderson. (2009). Leptin indirectly regulates gonadotropin- releasing hormone neuronal function. Endocrinology 150(6):2805-2812. Quinton, N. D., R. F. Smith, P. E. Clayton, M. S. Gill, S. Shalet, S. K. Justice, S. A. Simon, S. Walters, M. C. Postel-Vinay, A. I. Blakemore et al. (1999). Leptin binding activity changes with age: the link between leptin and puberty. J Clin Endocrinol Metab 84(7):2336-2341. Quirk, C. C., K. L. Lozada, R. A. Keri, & J. H. Nilson. (2001). A single Pitx1 binding site is essential for activity of the LHβ promoter in transgenic mice. Molecular Endocrinology 15(5):734-746.

194

Rajesh, R. V., G. N. Heo, M. R. Park, J. S. Nam, N. K. Kim, D. Yoon, T. H. Kim, & H. J. Lee. (2010). Proteomic analysis of bovine omental, subcutaneous and intramuscular preadipocytes during in vitro adipogenic differentiation. Comp Biochem Physiol Part D Genomics Proteomics 5(3):234-244. Ramayo-Caldas, Y., M. Ballester, M. R. Fortes, A. Esteve-Codina, A. Castelló, J. L. Noguera, A. I. Fernández, M. Pérez-Enciso, A. Reverter, & J. M. Folch. (2014). From SNP co- association to RNA co-expression: Novel insights into gene networks for intramuscular fatty acid composition in porcine. BMC Genomics 15(1):1-15. Rance, N. E. (2009). Menopause and the human hypothalamus: evidence for the role of kisspeptin/neurokinin B neurons in the regulation of estrogen negative feedback. Peptides 30(1):111-122. Rao, I. M., & V. B. Mahesh. (1986). Role of progesterone in the modulation of the preovulatory surge of gonadotropins and ovulation in the pregnant mare's serum gonadotropin- primed immature rat and the adult rat. Biol Reprod 35(5):1154-1161. Rao, S., S.-Y. Lee, A. Gutierrez, J. Perrigoue, R. J. Thapa, Z. Tu, J. R. Jeffers, M. Rhodes, S. Anderson, T. Oravecz et al. (2012). Inactivation of ribosomal protein L22 promotes transformation by induction of the stemness factor, Lin28B. Blood 120(18):3764-3773. Rasmussen, P., P. Brassard, H. Adser, M. V. Pedersen, L. Leick, E. Hart, N. H. Secher, B. K. Pedersen, & H. Pilegaard. (2009). Evidence for a release of brain-derived neurotrophic factor from the brain during exercise. Exp Physiol 94(10):1062-1069. Rathbone, M. J., J. E. Kinder, K. Fike, F. Kojima, D. Clopton, C. R. Ogle, & C. R. Bunt. (2001). Recent advances in bovine reproductive endocrinology and physiology and their impact on drug delivery system design for the control of the estrous cycle in cattle. Advanced Drug Delivery Reviews 50(3):277-320. Ravasi, T., H. Suzuki, C. V. Cannistraci, S. Katayama, V. B. Bajic, K. Tan, A. Akalin, S. Schmeier, M. Kanamori-Katayama, N. Bertin et al. (2010). An atlas of combinatorial transcriptional regulation in mouse and man. Cell 140(5):744-752. Rawat, P., S. Bathla, R. Baithalu, M. L. Yadav, S. Kumar, S. A. Ali, A. Tiwari, M. Lotfan, J. Naru, M. Jena et al. (2016). Identification of potential protein biomarkers for early detection of pregnancy in cow urine using 2D DIGE and label free quantitation. Clinical Proteomics 13(1):15. Rawlings, N. C., A. C. Evans, A. Honaramooz, & P. M. Bartlewski. (2003). Antral follicle growth and endocrine changes in prepubertal cattle, sheep and goats. Anim Reprod Sci 78(3-4):259-270. Rawson, P., C. Stockum, L. Peng, B. Manivannan, K. Lehnert, H. E. Ward, S. D. Berry, S. R. Davis, R. G. Snell, D. McLauchlan et al. (2012). Metabolic proteomics of the liver and mammary gland during lactation. J Proteomics 75(14):4429-4435. Redmond, J. S., G. M. Baez-Sandoval, K. M. Spell, T. E. Spencer, C. A. Lents, G. L. Williams, & M. Amstalden. (2011). Developmental changes in hypothalamic Kiss1 expression during activation of the pulsatile release of luteinising hormone in maturing ewe lambs. J Neuroendocrinol 23(9):815-822. Rejon, C. A., M. A. Hancock, Y. N. Li, T. B. Thompson, T. E. Hébert, & D. J. Bernard. (2013). Activins bind and signal via bone morphogenetic protein receptor type II (BMPR2) in immortalized gonadotrope-like cells. Cellular Signalling 25(12):2717-2726. Rempel, L. A., & J. A. Clapper. (2002). Administration of estradiol-17beta increases anterior pituitary IGF-I and relative amounts of serum and anterior pituitary IGF-binding proteins in barrows. J Anim Sci 80(1):214-224. Ren, M. Q., J. Wegner, O. Bellmann, G. A. Brockmann, F. Schneider, F. Teuscher, & K. Ender. (2002). Comparing mRNA levels of genes encoding leptin, leptin receptor, and

195

lipoprotein lipase between dairy and beef cattle. Domest Anim Endocrinol 23(3):371- 381. Reverter, A., W. Barris, S. McWilliam, K. A. Byrne, Y. H. Wang, S. H. Tan, N. Hudson, & B. P. Dalrymple. (2005). Validation of alternative methods of data normalization in gene co-expression studies. Bioinformatics 21(7):1112-1120. Reverter, A., & E. K. Chan. (2008). Combining partial correlation and an information theory approach to the reversed engineering of gene co-expression networks. Bioinformatics 24(21):2491-2497. Reverter, A., N. J. Hudson, S. H. Nagaraj, M. Perez-Enciso, & B. P. Dalrymple. (2010). Regulatory impact factors: unraveling the transcriptional regulation of complex traits from expression data. Bioinformatics 26(7):896-904. Reynolds, W. L. 1973. Reproduction of Brahman, Angus, Africaoder, and their crosses at Jeanerette, Louisiana. Rhodes, F. M., G. De'ath, & K. W. Entwistle. (1995). Animal and temporal effects on ovarian follicular dynamics in Brahman heifers. Animal Reproduction Science 38(4):265-277. Rice, V. M., S. D. Limback, K. F. Roby, & P. F. Terranova. (1998). Changes in circulating and ovarian concentrations of bioactive tumour necrosis factor alpha during the first ovulation at puberty in rats and in gonadotrophin-treated immature rats. J Reprod Fertil 113(2):337-341. Richard, C., J. Gao, N. Brown, & J. Reese. (2003). Aquaporin Water Channel Genes Are Differentially Expressed and Regulated by Ovarian Steroids during the Periimplantation Period in the Mouse. Endocrinology 144(4):1533-1541. Richter, T. A., J. E. Robinson, & N. P. Evans. (2002). Progesterone blocks the estradiol- stimulated luteinizing hormone surge by disrupting activation in response to a stimulatory estradiol signal in the ewe. Biol Reprod 67(1):119-125. Ricote, M., A. C. Li, T. M. Willson, C. J. Kelly, & C. K. Glass. (1998). The peroxisome proliferator-activated receptor-gamma is a negative regulator of macrophage activation. Nature 391(6662):79-82. Riding, G. A., A. Jones, M. K. Holland, J. R. Hill, & S. A. Lehnert. (2008). Proteomic analysis of bovine conceptus fluids during early pregnancy. Proteomics 8(1):160-177. Roa, J. (2013). Role of GnRH Neurons and Their Neuronal Afferents as Key Integrators between Food Intake Regulatory Signals and the Control of Reproduction. International Journal of Endocrinology 2013:10. Roa, J., D. Garcia-Galiano, J. M. Castellano, F. Gaytan, L. Pinilla, & M. Tena-Sempere. (2010). Metabolic control of puberty onset: new players, new mechanisms. Mol Cell Endocrinol 324(1-2):87-94. Roa, J., D. Garcia-Galiano, L. Varela, M. A. Sanchez-Garrido, R. Pineda, J. M. Castellano, F. Ruiz-Pino, M. Romero, E. Aguilar, M. Lopez et al. (2009). The mammalian target of rapamycin as novel central regulator of puberty onset via modulation of hypothalamic Kiss1 system. Endocrinology 150(11):5016-5026. Roberts, A. J., R. N. Funston, & G. E. Moss. (2001). Insulin-like growth factor binding proteins in the bovine anterior pituitary. Endocrine 14(3):399-406. Robinson, T., & J. Shelton. 1977. Reproduction in cattle. 3rd ed. N.Y. Academic Press, New York. Rodrigues, H. D., J. E. Kinder, & L. A. Fitzpatrick. (2002). Estradiol regulation of luteinizing hormone secretion in heifers of two breed types that reach puberty at different ages. Biol Reprod 66(3):603-609. Rodriguez, M., P. Arias, D. Refojo, C. Feleder, & J. Moguilevsky. (1999). Arrest of pulsatile luteinizing hormone (LH) secretion during insulin-induced hypoglycemia (IIH):

196

improvement by intrahypothalamic perfusion with glucose. Exp Clin Endocrinol Diabetes 107(4):257-261. Rogers, L. C., I. de Boer, M.-P. Junier, & S. R. Ojeda. (1993). Estradiol Increases Neural- Specific Class II-β-Tubulin mRNA Levels in the Developing Female Hypothalamus by Regulating mRNA Stability. Molecular and Cellular Neuroscience 4(5):424-431. Romaniello, R., F. Arrigoni, M. T. Bassi, & R. Borgatti. (2015). Mutations in alpha- and beta- tubulin encoding genes: implications in brain malformations. Brain Dev 37(3):273-280. Romano, M. A., V. H. Barnabe, J. P. Kastelic, C. A. de Oliveira, & R. M. Romano. (2007). Follicular dynamics in heifers during pre-pubertal and pubertal period kept under two levels of dietary energy intake. Reprod Domest Anim 42(6):616-622. Romao, J. M., M. L. He, T. A. McAllister, & L. L. Guan. (2014). Effect of age on bovine subcutaneous fat proteome: molecular mechanisms of physiological variations during beef cattle growth. J Anim Sci 92(8):3316-3327. Romao, J. M., W. Jin, M. He, T. McAllister, & L. L. Guan. (2013). Elucidation of Molecular Mechanisms of Physiological Variations between Bovine Subcutaneous and Visceral Fat Depots under Different Nutritional Regimes. PLOS ONE 8(12):e83211. Ross, S. E., N. Hemati, K. A. Longo, C. N. Bennett, P. C. Lucas, R. L. Erickson, & O. A. MacDougald. (2000). Inhibition of adipogenesis by Wnt signaling. Science 289(5481):950-953. Roth, C. L., C. Mastronardi, A. Lomniczi, H. Wright, R. Cabrera, A. E. Mungenast, S. Heger, H. Jung, C. Dubay, & S. R. Ojeda. (2007). Expression of a tumor-related gene network increases in the mammalian hypothalamus at the time of female puberty. Endocrinology 148(11):5147-5161. Roth, C. L., A. L. McCormack, A. Lomniczi, A. E. Mungenast, & S. R. Ojeda. (2006). Quantitative proteomics identifies a change in glial glutamate metabolism at the time of female puberty. Mol Cell Endocrinol 254-255:51-59. Ruder, E. H., T. J. Hartman, & M. B. Goldman. (2009). Impact of oxidative stress on female fertility. Curr Opin Obstet Gynecol 21(3):219-222. Rui, L. (2014). Energy metabolism in the liver. Compr Physiol 4(1):177-197. Rutter, L. M., R. Snopek, & J. G. Manns. (1989). Serum concentration of IGF-I in post-partum beef cows. Journal Animal Science 67:2060-2066. Ryu, K. Y., J. C. Garza, X. Y. Lu, G. S. Barsh, & R. R. Kopito. (2008). Hypothalamic neurodegeneration and adult-onset obesity in mice lacking the Ubb polyubiquitin gene. Proc Natl Acad Sci U S A 105(10):4016-4021. Saatchi, M., R. D. Schnabel, J. F. Taylor, & D. J. Garrick. (2014). Large-effect pleiotropic or closely linked QTL segregate within and across ten US cattle breeds. BMC Genomics 15(1):442. Sahu, A. (1998). Evidence suggesting that galanin (GAL), melanin-concentrating hormone (MCH), neurotensin (NT), proopiomelanocortin (POMC) and neuropeptide Y (NPY) are targets of leptin signaling in the hypothalamus. Endocrinology 139(2):795-798. Saltiel, A. R., & C. R. Kahn. (2001). Insulin signalling and the regulation of glucose and lipid metabolism. Nature 414(6865):799-806. Salwinski, L., C. S. Miller, A. J. Smith, F. K. Pettit, J. U. Bowie, & D. Eisenberg. (2004). The Database of Interacting Proteins: 2004 update. Nucleic Acids Res 32(Database issue):D449-451. Sanchez-Garrido, M. A., & M. Tena-Sempere. (2013). Metabolic control of puberty: Roles of leptin and kisspeptins. Hormones and Behavior 64(2):187-194. Sandau, U. S., A. E. Mungenast, Z. Alderman, S. P. Sardi, A. I. Fogel, B. Taylor, A. S. Parent, T. Biederer, G. Corfas, & S. R. Ojeda. (2011a). SynCAM1, a synaptic adhesion

197

molecule, is expressed in astrocytes and contributes to erbB4 receptor-mediated control of female sexual development. Endocrinology 152(6):2364-2376. Sandau, U. S., A. E. Mungenast, J. McCarthy, T. Biederer, G. Corfas, & S. R. Ojeda. (2011b). The synaptic cell adhesion molecule, SynCAM1, mediates astrocyte-to-astrocyte and astrocyte-to-GnRH neuron adhesiveness in the mouse hypothalamus. Endocrinology 152(6):2353-2363. Sanders, J. O. (1980). History and Development of Zebu Cattle in the United States. J Anim Sci 50:1188-1200. Sangiao-Alvarellos, S., M. Manfredi-Lozano, F. Ruiz-Pino, V. M. Navarro, M. A. Sanchez- Garrido, S. Leon, C. Dieguez, F. Cordido, V. Matagne, G. A. Dissen et al. (2013). Changes in hypothalamic expression of the Lin28/let-7 system and related microRNAs during postnatal maturation and after experimental manipulations of puberty. Endocrinology 154(2):942-955. Sartori, R., & C. M. Barros. (2011). Reproductive cycles in Bos indicus cattle. Anim Reprod Sci 124(3-4):244-250. Sato, K., T. Iwasaki, K. Sakakibara, S. Itakura, & Y. Fukami. (2002). Towards the molecular dissection of fertilization signaling: Our functional genomic/proteomic strategies. Proteomics 2(9):1079-1089. Sauerwein, H., E. Bendixen, L. Restelli, & F. Ceciliani. (2014). The adipose tissue in farm animals: a proteomic approach. Curr Protein Pept Sci 15(2):146-155. Schaefer, M. H., J. F. Fontaine, A. Vinayagam, P. Porras, E. E. Wanker, & M. A. Andrade- Navarro. (2012). HIPPIE: Integrating protein interaction networks with experiment based quality scores. PLoS One 7(2):e31826. Scharfman, H. E., & N. J. MacLusky. (2006). Estrogen and brain-derived neurotrophic factor (BDNF) in hippocampus: complexity of steroid hormone-growth factor interactions in the adult CNS. Frontiers in neuroendocrinology 27(4):415-435. Schillo, K. K., D. J. Dierschke, & E. R. Hauser. (1982). Influences of month of birth and age on patterns of luteinizing hormone secretion in prepubertal heifers. Theriogenology 18:593. Schillo, K. K., J. B. Hall, & S. M. Hileman. (1992). Effects of nutrition and season on the onset of puberty in the beef heifer. J Anim Sci 70(12):3994-4005. Schillo, K. K., P. J. Hansen, L. A. Kamwanja, D. J. Dierschke, & E. R. Hauser. (1983). Influence of season on sexual development in heifers: age at puberty asrelated to growth and serum concentrations of gonadotropins, prolactin, thyroxine and progesterone. Biol. Reprod 28:329. Schmitt, T., C. Ogris, & E. L. Sonnhammer. (2014). FunCoup 3.0: database of genome-wide functional coupling networks. Nucleic Acids Res 42(Database issue):D380-388. Schneider, J. E., & G. N. Wade. (1989). Availability of metabolic fuels controls estrous cyclicity of Syrian hamsters. Science 244(4910):1326-1328. Schneider, J. E., & G. N. Wade. (1990). Decreased availability of metabolic fuels induces anestrus in golden hamsters. Am J Physiol 258(3 Pt 2):R750-755. Schwanhausser, B., D. Busse, N. Li, G. Dittmar, J. Schuchhardt, J. Wolf, W. Chen, & M. Selbach. (2011). Global quantification of mammalian gene expression control. Nature 473(7347):337-342. Schwartz, D. R., R. Wu, S. L. Kardia, A. M. Levin, C. C. Huang, K. A. Shedden, R. Kuick, D. E. Misek, S. M. Hanash, J. M. Taylor et al. (2003). Novel candidate targets of beta- catenin/T-cell factor signaling identified by gene expression profiling of ovarian endometrioid adenocarcinomas. Cancer Res 63(11):2913-2922.

198

Scolari, S. C., G. Pugliesi, R. F. Strefezzi, S. C. Andrade, L. L. Coutinho, & M. Binelli. (2016). Dynamic remodeling of endometrial extracellular matrix regulates embryo receptivity in cattle. Reproduction Seifert, T., P. Brassard, M. Wissenberg, P. Rasmussen, P. Nordby, B. Stallknecht, H. Adser, A. H. Jakobsen, H. Pilegaard, H. B. Nielsen et al. (2010). Endurance training enhances BDNF release from the human brain. American Journal of Physiology - Regulatory, Integrative and Comparative Physiology 298(2):R372. Seminara, S. B., S. Messager, E. E. Chatzidaki, R. R. Thresher, J. S. Acierno, Jr., J. K. Shagoury, Y. Bo-Abbas, W. Kuohung, K. M. Schwinof, A. G. Hendrick et al. (2003). The GPR54 gene as a regulator of puberty. N Engl J Med 349(17):1614-1627. Semple, J. I., G. Ribas, G. Hillyard, S. E. Brown, C. M. Sanderson, & R. D. Campbell. (2003). A novel gene encoding a coiled-coil mitochondrial protein located at the telomeric end of the human MHC Class III region. Gene 314:41-54. Sethi, J. K., & A. Vidal-Puig. (2010). Wnt signalling and the control of cellular metabolism. Biochem J 427(1):1-17. Sethi, J. K., & A. J. Vidal-Puig. (2007). Thematic review series: Adipocyte Biology. Adipose tissue function and plasticity orchestrate nutritional adaptation. Journal of Lipid Research 48(6):1253-1262. Shahab, M., C. Mastronardi, S. B. Seminara, W. F. Crowley, S. R. Ojeda, & T. M. Plant. (2005). Increased hypothalamic GPR54 signaling: a potential mechanism for initiation of puberty in primates. Proc Natl Acad Sci U S A 102(6):2129-2134. Shalitin, S., & W. Kiess. (2017). Putative Effects of Obesity on Linear Growth and Puberty. Hormone Research in Paediatrics 88(1):101-110. Shamay, A., D. Werner, U. Moallem, H. Barash, & I. Bruckental. (2005). Effect of nursing management and skeletal size at weaning on puberty, skeletal growth rate, and milk production during first lactation of dairy heifers. J Dairy Sci 88(4):1460-1469. Shan, T., J. Liu, W. Wu, Z. Xu, & Y. Wang. (2017). Roles of Notch Signaling in Adipocyte Progenitor Cells and Mature Adipocytes. Journal of Cellular Physiology 232(6):1258- 1261. Shannon, P., A. Markiel, O. Ozier, N. S. Baliga, J. T. Wang, D. Ramage, N. Amin, B. Schwikowski, & T. Ideker. (2003). Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res 13(11):2498-2504. Sharma, S., P. M. Sharma, D. S. Mistry, R. J. Chang, J. M. Olefsky, P. L. Mellon, & N. J. G. Webster. (2011). PPARG Regulates Gonadotropin-Releasing Hormone Signaling in LbetaT2 Cells In Vitro and Pituitary Gonadotroph Function In Vivo in Mice. Biology of Reproduction 84(3):466-475. Sharova, L. V., A. A. Sharov, T. Nedorezov, Y. Piao, N. Shaik, & M. S. Ko. (2009). Database for mRNA half-life of 19 977 genes obtained by DNA microarray analysis of pluripotent and differentiating mouse embryonic stem cells. DNA Res 16(1):45-58. Sheffer-Babila, S., Y. Sun, D. D. Israel, S.-M. Liu, G. Neal-Perry, & S. C. Chua. 2013. Agouti- related peptide plays a critical role in leptin's effects on female puberty and reproduction. Shen, J., C. Zhou, S. Zhu, W. Shi, M. Hu, X. Fu, C. Wang, Y. Wang, Q. Zhang, & Y. Yu. (2014). Comparative transcriptome analysis reveals early pregnancy-specific genes expressed in peripheral blood of pregnant sows. PLoS One 9(12):e114036. Shen, Y. N., S. H. Kim, D. H. Yoon, H. G. Lee, H. S. Kang, & K. S. Seo. (2012). Proteome Analysis of Bovine Longissimus dorsi Muscle Associated with the Marbling Score. Asian-Australasian Journal of Animal Sciences 25(8):1083-1088.

199

Sheth, A. R., N. Arabatti, M. Carlquist, & H. Jornvall. (1984). Characterization of a polypeptide from human seminal plasma with inhibin (inhibition of FSH secretion)- like activity. FEBS Lett 165(1):11-15. Shevchenko, A., O. N. Jensen, A. V. Podtelejnikov, F. Sagliocco, M. Wilm, O. Vorm, P. Mortensen, A. Shevchenko, H. Boucherie, & M. Mann. (1996). Linking genome and proteome by mass spectrometry: large-scale identification of yeast proteins from two dimensional gels. Proceedings of the National Academy of Sciences 93(25):14440- 14445. Shi, L., S. A. Wudy, A. E. Buyken, C. Maser-Gluth, M. F. Hartmann, & T. Remer. (2011). Prepubertal glucocorticoid status and pubertal timing. J Clin Endocrinol Metab 96(6):E891-898. Shi, Y., & M. He. (2014). Differential gene expression identified by RNA-Seq and qPCR in two sizes of pearl oyster (Pinctada fucata). Gene 538(2):313-322. Shirley, K. L., M. G. Thomas, D. H. Keisler, D. M. Hallford, D. M. Montrose, G. A. Silver, & M. D. Garcia. (2006). A Chihuahuan Desert Brangus Breeding Program: Feed Efficiency, Metabolic Hormones, and Puberty in Heifers Sired by Bulls with Differing Expected Progeny Differences for Growth and Scrotal Circumference. The Professional Animal Scientist 22(1):48-58. Shupnik, M. A., S. D. Gharib, & W. W. Chin. (1988). Estrogen suppresses rat gonadotropin gene transcription in vivo. Endocrinology 122(5):1842-1846. Shupnik, M. A., S. D. Gharib, & W. W. Chin. (1989). Divergent effects of estradiol on gonadotropin gene transcription in pituitary fragments. Mol Endocrinol 3(3):474-480. Singh, P., G. Singh, A. Bhandawat, G. Singh, R. Parmar, R. Seth, & R. K. Sharma. (2017). Spatial transcriptome analysis provides insights of key gene(s) involved in steroidal saponin biosynthesis in medicinally important herb Trillium govanianum. 7:45295. Singh, S., C. Brocker, V. Koppaka, C. Ying, B. Jackson, A. Matsumoto, D. C. Thompson, & V. Vasiliou. (2013). Aldehyde Dehydrogenases in Cellular Responses to Oxidative/electrophilic Stress. Free radical biology & medicine 56:89-101. Sleiter, N., Y. Pang, C. Park, T. H. Horton, J. Dong, P. Thomas, & J. E. Levine. (2009). Progesterone receptor A (PRA) and PRB-independent effects of progesterone on gonadotropin-releasing hormone release. Endocrinology 150(8):3833-3844. Smeitink, J., L. van den Heuvel, & S. DiMauro. (2001). The genetics and pathology of oxidative phosphorylation. Nat Rev Genet 2(5):342-352. Smith, J. T., B. V. Acohido, D. K. Clifton, & R. A. Steiner. (2006). KiSS-1 neurones are direct targets for leptin in the ob/ob mouse. J Neuroendocrinol 18(4):298-303. Snel, B., G. Lehmann, P. Bork, & M. A. Huynen. (2000). STRING: a web-server to retrieve and display the repeatedly occurring neighbourhood of a gene. Nucleic Acids Res 28 Snelling, W. M., M. F. Allan, J. W. Keele, L. A. Kuehn, T. McDaneld, T. P. Smith, T. S. Sonstegard, R. M. Thallman, & G. L. Bennett. (2010). Genome-wide association study of growth in crossbred beef cattle. J Anim Sci 88(3):837-848. Snelling, W. M., R. A. Cushman, J. W. Keele, C. Maltecca, M. G. Thomas, M. R. Fortes, & A. Reverter. (2013). Breeding and Genetics Symposium: networks and pathways to guide genomic selection. J Anim Sci 91(2):537-552. Soldani, R., A. Cagnacci, A. M. Paoletti, S. S. Yen, & G. B. Melis. (1995). Modulation of anterior pituitary luteinizing hormone response to gonadotropin-releasing hormone by insulin-like growth factor I in vitro. Fertil Steril 64(3):634-637. Souza, C. E., J. P. Rego, C. H. Lobo, J. T. Oliveira, F. C. Nogueira, G. B. Domont, M. Fioramonte, F. C. Gozzo, F. B. Moreno, A. C. Monteiro-Moreira et al. (2012). Proteomic analysis of the reproductive tract fluids from tropically-adapted Santa Ines rams. J Proteomics 75(14):4436-4456.

200

Srivastava, V. K., J. K. Hiney, & W. L. Dees. (2014). Actions and interactions of alcohol and transforming growth factor beta1 on prepubertal hypothalamic gonadotropin-releasing hormone. Alcohol Clin Exp Res 38(5):1321-1329. Srivastava, V. K., J. K. Hiney, & W. L. Dees. (2016). Manganese-Stimulated Kisspeptin Is Mediated by the IGF-1/Akt/Mammalian Target of Rapamycin Pathway in the Prepubertal Female Rat. Endocrinology 157(8):3233-3241. Stahringer, R. C., D. A. Neuendorff, & R. D. Randel. (1990). Seasonal variations in characteristics of estrous cycles in pubertal Brahman heifers. Theriogenology 34(2):407-415. Stute, P., S. Sielker, C. E. Wood, T. C. Register, C. J. Lees, F. N. Dewi, J. K. Williams, J. D. Wagner, U. Stefenelli, & J. M. Cline. (2012). Life stage differences in mammary gland gene expression profile in non-human primates. Breast Cancer Res Treat 133(2):617- 634. Sulem, P., D. F. Gudbjartsson, T. Rafnar, H. Holm, E. J. Olafsdottir, G. H. Olafsdottir, T. Jonsson, P. Alexandersen, B. Feenstra, H. A. Boyd et al. (2009). Genome-wide association study identifies sequence variants on 6q21 associated with age at menarche. Nature Genetics 41:734. Sumner-Thomson, J. M., J. L. Vierck, & J. P. McNamara. (2011). Differential expression of genes in adipose tissue of first-lactation dairy cattle. J Dairy Sci 94(1):361-369. Sutton, M. L., P. D. Cetica, M. T. Beconi, K. L. Kind, R. B. Gilchrist, & J. G. Thompson. (2003a). Influence of oocyte-secreted factors and culture duration on the metabolic activity of bovine cumulus cell complexes. Reproduction 126(1):27-34. Sutton, M. L., R. B. Gilchrist, & J. G. Thompson. (2003b). Effects of in-vivo and in-vitro environments on the metabolism of the cumulus-oocyte complex and its influence on oocyte developmental capacity. Hum Reprod Update 9(1):35-48. Suuriniemi, M., V. Kovanen, A. Mahonen, M. Alen, Q. Wang, A. Lyytikainen, & S. Cheng. (2006). COL1A1 Sp1 polymorphism associates with bone density in early puberty. Bone 39(3):591-597. Swali, A., & D. C. Wathes. (2006). Influence of the dam and sire on size at birth and subsequent growth, milk production and fertility in dairy heifers. Theriogenology 66(5):1173- 1184. Swerdloff, R. (1978). Physiological control of puberty. Med Clin North Am 62:351-366. Swiersta, E. E., G. W. Rahnfeld, R. L. Cliplef, & J. H. Strain. (1977). Age and weight at puberty of crossbred heifers sired by Charolais, Limousin and Simmental bulls. Can. J. Anim. sci. 57:697. Szklarczyk, D., A. Franceschini, S. Wyder, K. Forslund, D. Heller, J. Huerte-Cepas, M. Simonovic, A. Roth, A. Santos, K. P. Tsafou et al. (2014). STRING v10: protein– protein interaction networks, integrated over the tree of life. Nucleic Acids Research Tabandeh, M. R., A. Hosseini, M. Saeb, M. Kafi, & S. Saeb. (2010). Changes in the gene expression of adiponectin and adiponectin receptors (AdipoR1 and AdipoR2) in ovarian follicular cells of dairy cow at different stages of development. Theriogenology 73(5):659-669. Tabandeh, M. R., A. Hosseini, P. Sarkoohi, & S. Seifi. 2011. Changes in the gene expression of adiponectin and adiponectin receptors (AdipoR1 and AdipoR2) in pituitary and hypothalamus during the estrous cycle and pregnancy 4th International Congress of Biochemistry and Molecular Biology No. 44. Clinical Biochemistry, Mashhad, Iran. Talamo, F., C. D'Ambrosio, S. Arena, P. Del Vecchio, L. Ledda, G. Zehender, L. Ferrara, & A. Scaloni. (2003). Proteins from bovine tissues and biological fluids: defining a reference electrophoresis map for liver, kidney, muscle, plasma and red blood cells. Proteomics 3(4):440-460.

201

Tan, C. K., H. C. Chong, E. H. P. Tan, & N. S. Tan. (2012). Getting ‘Smad’ about obesity and diabetes. Nutrition &Amp; Diabetes 2:e29. Tan, N. Y., U.-M. Bailey, M. F. Jamaluddin, S. H. B. Mahmud, S. C. Raman, & B. L. Schulz. (2014). Sequence-based protein stabilization in the absence of glycosylation. Nature Communications 5:3099. Tanaka, T., K. Itahana, N. Andoh, T. Takeya, & E. Sato. (1995). Expression of prostatic secretory protein (PSP)-like protein in porcine corpus luteum: isolation and characterization of a new gene encoding PSP94-like protein. Mol Reprod Dev 42(2):149-156. Tanaka, T., S. Nagatani, D. C. Bucholtz, S. Ohkura, H. Tsukamura, K. Maeda, & D. L. Foster. (2000). Central action of insulin regulates pulsatile luteinizing hormone secretion in the diabetic sheep model. Biol Reprod 62(5):1256-1261. Tatman, S. R., D. A. Neuendorff, T. W. Wilson, & R. D. Randel. (2004). Influence of season of birth on growth and reproductive development of Brahman bulls. Theriogenology 62:93-102. Temple, J. L., & S. Wray. (2005). Bovine serum albumin-estrogen compounds differentially alter gonadotropin-releasing hormone-1 neuronal activity. Endocrinology 146(2):558- 563. Tena-Sempere, M. (2008). Ghrelin and reproduction: ghrelin as novel regulator of the gonadotropic axis. Vitam Horm 77:285-300. Terasawa, E., & D. L. Fernandez. (2001). Neurobiological mechanisms of the onset of puberty in primates. Endocr Rev 22(1):111-151. Thakur, A. N., A. Y. Vaze, B. Dattatreyamurthy, & A. R. Sheth. (1981). Isolation & characterization of inhibin from human seminal plasma. Indian J Exp Biol 19(4):307- 313. Thomas, M. G., O. S. Gazal, G. L. Williams, R. L. Stanko, & D. H. Keisler. (1999). Injection of neuropeptide Y into the third cerebroventricle differentially influences pituitary secretion of luteinizing hormone and growth hormone in ovariectomized cows. Domest Anim Endocrinol 16(3):159-169. Thornton, J. E., C. C. Cheung, D. K. Clifton, & R. A. Steiner. (1997). Regulation of hypothalamic proopiomelanocortin mRNA by leptin in ob/ob mice. Endocrinology 138(11):5063-5066. Tian, Q., S. B. Stepaniants, M. Mao, L. Weng, M. C. Feetham, M. J. Doyle, E. C. Yi, H. Dai, V. Thorsson, J. Eng et al. (2004). Integrated genomic and proteomic analyses of gene expression in Mammalian cells. Mol Cell Proteomics 3(10):960-969. Tizioto, P. C., L. L. Coutinho, J. E. Decker, R. D. Schnabel, K. O. Rosa, P. S. Oliveira, M. M. Souza, G. B. Mourao, R. R. Tullio, A. S. Chaves et al. (2015). Global liver gene expression differences in Nelore steers with divergent residual feed intake phenotypes. BMC Genomics 16(1):242. Tizioto, P. C., L. L. Coutinho, P. S. N. Oliveira, A. S. M. Cesar, W. J. S. Diniz, A. O. Lima, M. I. Rocha, J. E. Decker, R. D. Schnabel, G. B. Mourão et al. (2016). Gene expression differences in Longissimus muscle of Nelore steers genetically divergent for residual feed intake. Scientific Reports 6:39493. Tomikawa, J., Y. Uenoyama, M. Ozawa, T. Fukanuma, K. Takase, T. Goto, H. Abe, N. Ieda, S. Minabe, C. Deura et al. (2012). Epigenetic regulation of Kiss1 gene expression mediating estrogen-positive feedback action in the mouse brain. Proceedings of the National Academy of Sciences of the United States of America 109(20):E1294-E1301. Tong, L., M. Wang, Y. Zhou, Y. Li, L. Chen, F. Xu, K. Li, & J. Xiao. (2018). MiR-505-3p, a New Repressor of Puberty Onset in Female Mice. bioRxiv

202

Topaloglu, A. K., F. Reimann, M. Guclu, A. S. Yalin, L. D. Kotan, K. M. Porter, A. Serin, N. O. Mungan, J. R. Cook, M. N. Ozbek et al. (2008). TAC3 and TACR3 mutations in familial hypogonadotropic hypogonadism reveal a key role for Neurokinin B in the central control of reproduction. Nature Genetics 41:354. Topaloglu, A. K., F. Reimann, M. Guclu, A. S. Yalin, L. D. Kotan, K. M. Porter, A. Serin, N. O. Mungan, J. R. Cook, M. N. Ozbek et al. (2009). TAC3 and TACR3 mutations in familial hypogonadotropic hypogonadism reveal a key role for Neurokinin B in the central control of reproduction. Nat Genet 41(3):354-358. Topaloglu, A. K., J. A. Tello, L. D. Kotan, M. N. Ozbek, M. B. Yilmaz, S. Erdogan, F. Gurbuz, F. Temiz, R. P. Millar, & B. Yuksel. (2012). Inactivating KISS1 mutation and hypogonadotropic hypogonadism. N Engl J Med 366(7):629-635. Trama, J., Q. J. Lu, R. G. Hawley, & S. N. Ho. (2000). The NFAT-related protein NFATL1 (TonEBP/NFAT5) is induced upon T cell activation in a calcineurin-dependent manner. Journal of Immunology 165(9):4884-4894. Tremellen, K. (2008). Oxidative stress and male infertility--a clinical perspective. Hum Reprod Update 14(3):243-258. Troussard, A. A., N. M. Mawji, C. Ong, A. Mui, R. St.-Arnaud, & S. Dedhar. (2003). Conditional Knock-out of Integrin-linked Kinase Demonstrates an Essential Role in Protein Kinase B/Akt Activation. Journal of Biological Chemistry 278(25):22374- 22378. True, C., S. Verma, K. L. Grove, & M. S. Smith. (2013). Cocaine- and amphetamine-regulated transcript is a potent stimulator of GnRH and kisspeptin cells and may contribute to negative energy balance-induced reproductive inhibition in females. Endocrinology 154(8):2821-2832. Tsurutani, Y., M. Fujimoto, M. Takemoto, H. Irisuna, M. Koshizaka, S. Onishi, T. Ishikawa, M. Mezawa, P. He, & S. Honjo. (2011). The roles of transforming growth factor-β and Smad3 signaling in adipocyte differentiation and obesity. Biochemical and biophysical research communications 407(1):68-73. Tsutsui, K., G. E. Bentley, G. Bedecarrats, T. Osugi, T. Ubuka, & L. J. Kriegsfeld. (2010). Gonadotropin-inhibitory hormone (GnIH) and its control of central and peripheral reproductive function. Front Neuroendocrinol 31(3):284-295. Urieli-Shoval, S., P. Cohen, S. Eisenberg, & Y. Matzner. (1998). Widespread expression of serum amyloid A in histologically normal human tissues. Predominant localization to the epithelium. J Histochem Cytochem 46(12):1377-1384. Urist, M. R. (1965). Bone: Formation by Autoinduction. Science 150(3698):893-899. Urrutia, R. (2003). KRAB-containing zinc-finger repressor proteins. Genome Biol 4(10):231. Ushizawa, K., C. B. Herath, K. Kaneyama, S. Shiojima, A. Hirasawa, T. Takahashi, K. Imai, K. Ochiai, T. Tokunaga, Y. Tsunoda et al. (2004). cDNA microarray analysis of bovine embryo gene expression profiles during the pre-implantation period. Reproductive biology and endocrinology : RB&E 2:77-77. Utsunomiya, Y. T., A. S. do Carmo, R. Carvalheiro, H. H. Neves, M. C. Matos, L. B. Zavarez, A. M. Pérez O’Brien, J. Sölkner, J. C. McEwan, J. B. Cole et al. (2013). Genome-wide association study for birth weight in Nellore cattle points to previously described orthologous genes affecting human and bovine height. BMC Genetics 14(1):52. Vaiciunas, A., L. L. Coutinho, F. V. Meirelles, A. V. Pires, & L. F. P. Silva. (2008). Leptin and hypothalamic gene expression in early- and late-maturing Bos indicus Nellore heifers. Genetics and Molecular Biology 31:657-664. Van Beest, F. M., A. Mysterud, L. E. Loe, & J. M. Milner. (2010). Forage quantity, quality and depletion as scale‐dependent mechanisms driving habitat selection of a large browsing herbivore. Journal of Animal Ecology 79(4):910-922.

203

Vaquerizas, J. M., S. K. Kummerfeld, S. A. Teichmann, & N. M. Luscombe. (2009). A census of human transcription factors: function, expression and evolution. Nat Rev Genet 10(4):252-263. Vargas, C. A., M. A. Elzo, C. C. Chase, Jr., P. J. Chenoweth, & T. A. Olson. (1998). Estimation of genetic parameters for scrotal circumference, age at puberty in heifers, and hip height in Brahman cattle. J Anim Sci 76(10):2536-2541. Vasiliou, V., A. Pappa, & T. Estey. (2004). Role of human aldehyde dehydrogenases in endobiotic and xenobiotic metabolism. Drug Metab Rev 36(2):279-299. Vidova, V., & Z. Spacil. (2017). A review on mass spectrometry-based quantitative proteomics: Targeted and data independent acquisition. Analytica Chimica Acta 964:7- 23. Vigneault, C., C. Gravel, M. Vallee, S. McGraw, & M. A. Sirard. (2009). Unveiling the bovine embryo transcriptome during the maternal-to-embryonic transition. Reproduction 137(2):245-257. Vizcaino, J. A., A. Csordas, N. del-Toro, J. A. Dianes, J. Griss, I. Lavidas, G. Mayer, Y. Perez- Riverol, F. Reisinger, T. Ternent et al. (2016). 2016 update of the PRIDE database and its related tools. Nucleic Acids Res 44(D1):D447-456. Vogel, C., & E. M. Marcotte. (2012). Insights into the regulation of protein abundance from proteomic and transcriptomic analyses. Nat Rev Genet 13(4):227-232. Vogel, M. J., L. Guelen, E. de Wit, D. Peric-Hupkes, M. Loden, W. Talhout, M. Feenstra, B. Abbas, A. K. Classen, & B. van Steensel. (2006). Human heterochromatin proteins form large domains containing KRAB-ZNF genes. Genome Res 16(12):1493-1504. Wagner, D. O., C. Sieber, R. Bhushan, J. H. Borgermann, D. Graf, & P. Knaus. (2010). BMPs: from bone to body morphogenetic proteins. Sci Signal 3(107):mr1. Wahab, F., F. Ullah, Y. M. Chan, S. B. Seminara, & M. Shahab. (2011). Decrease in hypothalamic Kiss1 and Kiss1r expression: a potential mechanism for fasting-induced suppression of the HPG axis in the adult male rhesus monkey (Macaca mulatta). Horm Metab Res 43(2):81-85. Wahli, W., & L. Michalik. (2012). PPARs at the crossroads of lipid signaling and inflammation. Trends in Endocrinology & Metabolism 23(7):351-363. Walther, T. C., & M. Mann. (2010). Mass spectrometry–based proteomics in cell biology. The Journal of Cell Biology 190(4):491. Wang, Y., N. Ghaffari, C. D. Johnson, U. M. Braga-Neto, H. Wang, R. Chen, & H. Zhou. (2011). Evaluation of the coverage and depth of transcriptome by RNA-Seq in chickens. BMC Bioinformatics 12(10):S5. Wang, Y. H., N. I. Bower, A. Reverter, S. H. Tan, N. De Jager, R. Wang, S. M. McWilliam, L. M. Cafe, P. L. Greenwood, & S. A. Lehnert. (2009a). Gene expression patterns during intramuscular fat development in cattle. J Anim Sci 87(1):119-130. Wang, Z., M. Gerstein, & M. Snyder. (2009b). RNA-Seq: a revolutionary tool for transcriptomics. Nat Rev Genet 10(1):57-63. Wang, Z., J. Hou, L. Lu, Z. Qi, J. Sun, W. Gao, J. Meng, Y. Wang, H. Sun, H. Gu et al. (2013). Small Ribosomal Protein Subunit S7 Suppresses Ovarian Tumorigenesis through Regulation of the PI3K/AKT and MAPK Pathways. PLoS ONE 8(11):e79117. Washburn, M. P., A. Koller, G. Oshiro, R. R. Ulaszek, D. Plouffe, C. Deciu, E. Winzeler, & J. R. Yates, 3rd. (2003). Protein pathway and complex clustering of correlated mRNA and protein expression analyses in Saccharomyces cerevisiae. Proc Natl Acad Sci U S A 100(6):3107-3112. Wathes, D. C., D. R. Abayasekara, & R. J. Aitken. (2007). Polyunsaturated fatty acids in male and female reproduction. Biol Reprod 77(2):190-201.

204

Webb, R., P. C. Garnsworthy, J. G. Gong, & D. G. Armstrong. (2004). Control of follicular growth: local interactions and nutritional influences. J Anim Sci 82 E-Suppl:E63-74. Webb, R., R. G. Gosden, E. E. Telfer, & R. M. Moor. (1999). Factors affecting folliculogenesis in ruminants J Anim Sci 68(2):433-443. Weber, K. L., B. T. Welly, A. L. Van Eenennaam, A. E. Young, L. R. Porto-Neto, A. Reverter, & G. Rincon. (2016). Identification of Gene Networks for Residual Feed Intake in Angus Cattle Using Genomic Prediction and RNA-seq. PLOS ONE 11(3):e0152274. Werther, G. A., A. Hogg, B. J. Oldfield, M. J. McKinley, R. Figdor, & F. A. Mendelsohn. (1989). Localization and Characterization of Insulin-Like Growth Factor-I Receptors in Rat Brain and Pituitary Gland Using in vitro Autoradiography and Computerized Densitometry* A Distinct Distribution from Insulin Receptors. J Neuroendocrinol 1(5):369-377. Whitley, N. C., C. R. Barb, R. V. Utley, J. M. Popwell, R. R. Kraeling, & G. B. Rampacek. (1995). Influence of stage of the estrous cycle on insulin-like growth factor-I modulation of luteinizing hormone secretion in the gilt. Biol Reprod 53(6):1359-1364. Whitlock, B. K., J. A. Daniel, R. R. Wilborn, S. P. Rodning, H. S. Maxwell, B. P. Steele, & J. L. Sartin. (2008). Interaction of estrogen and progesterone on kisspeptin-10-stimulated luteinizing hormone and growth hormone in ovariectomized cows. Neuroendocrinology 88(3):212-215. Wickramasinghe, S., G. Rincon, A. Islas-Trejo, & J. F. Medrano. (2012). Transcriptional profiling of bovine milk using RNA sequencing. BMC Genomics 13:45. Wilkinson, M., R. Brown, S. A. Imran, & E. Ur. (2007). Adipokine gene expression in brain and pituitary gland. Neuroendocrinology 86(3):191-209. Willert, J., M. Epping, J. R. Pollack, P. O. Brown, & R. Nusse. (2002). A transcriptional response to Wnt protein in human embryonic carcinoma cells. BMC Dev Biol 2:8. Williams, G. L., M. Amstalden, M. R. Garcia, R. L. Stanko, S. E. Nizielski, C. D. Morrison, & D. H. Keisler. (2002). Leptin and its role in the central regulation of reproduction in cattle. Domest Anim Endocrinol 23(1-2):339-349. Williams, L. M., B. A. Lago, A. G. McArthur, A. R. Raphenya, N. Pray, N. Saleem, S. Salas, K. Paulson, R. S. Mangar, Y. Liu et al. (2016). The transcription factor, Nuclear factor, erythroid 2 (Nfe2), is a regulator of the oxidative stress response during Danio rerio development. Aquat Toxicol 180:141-154. Wilson, M. E. (1995). IGF-I administration advances the decrease in hypersensitivity to oestradiol negative feedback inhibition of serum LH in adolescent female rhesus monkeys. Journal of Endocrinology 145(1):121-130. Wiltbank, J. N., C. W. Kasson, & J. E. Ingalls. (1969). Puberty in Crossbred and Straightbred Beef Heifers on two Levels of Feed. Journal of Animal Science 29(4):602-605. Wisse, B. E., & M. W. Schwartz. (2003). The skinny on neurotrophins. Nat Neurosci 6(7):655- 656. Wolfenson, D., Z. Roth, & R. Meidan. (2000). Impaired reproduction in heat-stressed cattle: basic and applied aspects. Anim Reprod Sci 60-61:535-547. Wright, W. S., K. A. Longo, V. W. Dolinsky, I. Gerin, S. Kang, C. N. Bennett, S. H. Chiang, T. C. Prestwich, C. Gress, C. F. Burant et al. (2007). Wnt10b inhibits obesity in ob/ob and agouti mice. Diabetes 56(2):295-303. Wu, A. R., N. F. Neff, T. Kalisky, P. Dalerba, B. Treutlein, M. E. Rothenberg, F. M. Mburu, G. L. Mantalas, S. Sim, & M. F. Clarke. (2014). Quantitative assessment of single-cell RNA-sequencing methods. Nature methods 11(1):41. Wu, M. Y., & C. S. Hill. (2009). TGF-β Superfamily Signaling in Embryonic Development and Homeostasis. Developmental Cell 16(3):329-343.

205

Wu, Q., B. B. Whiddon, & R. D. Palmiter. (2012). Ablation of neurons expressing agouti- related protein, but not melanin concentrating hormone, in leptin-deficient mice restores metabolic functions and fertility. Proc Natl Acad Sci U S A 109(8):3155-3160. Wurster, A. L., P. Precht, K. G. Becker, W. H. Wood, Y. Zhang, Z. Wang, & M. J. Pazin. (2012). IL-10 transcription is negatively regulated by BAF180, a component of the SWI/SNF chromatin remodeling enzyme. BMC Immunology 13(1):1-19. Wycherley, G., M. T. Kane, & A. C. Hynes. (2005). Oxidative phosphorylation and the tricarboxylic acid cycle are essential for normal development of mouse ovarian follicles. Hum Reprod 20(10):2757-2763. Xie, H., H. M. Hoffmann, J. D. Meadows, S. L. Mayo, C. Trang, S. S. Leming, C. Maruggi, S. W. Davis, R. Larder, & P. L. Mellon. (2015). Homeodomain Proteins SIX3 and SIX6 Regulate Gonadotrope-specific Genes During Pituitary Development. Mol Endocrinol 29(6):842-855. Xie, W., G. M. Samoriski, J. P. McLaughlin, V. A. Romoser, A. Smrcka, P. M. Hinkle, J. M. Bidlack, R. A. Gross, H. Jiang, & D. Wu. (1999). Genetic alteration of phospholipase C beta3 expression modulates behavioral and cellular responses to mu opioids. Proc Natl Acad Sci U S A 96(18):10385-10390. Xin, X., X. Luan, J. Xiao, D. Wei, J. Wang, D. Lu, & S. Yang. (2005). Association study of four activity SNPs of CYP3A4 with the precocious puberty in Chinese girls. Neuroscience Letters 381(3):284-288. Xu, B., E. H. Goulding, K. Zang, D. Cepoi, R. D. Cone, K. R. Jones, L. H. Tecott, & L. F. Reichardt. (2003). Brain-derived neurotrophic factor regulates energy balance downstream of melanocortin-4 receptor. Nat Neurosci 6(7):736-742. Xu, C., & Z. Wang. (2008). Comparative proteomic analysis of livers from ketotic cows. Vet Res Commun 32(3):263-273. Xu, C., Z. Wang, G. Liu, X. Li, G. Xie, C. Xia, & H. Y. Zhang. (2008). Metabolic characteristic of the liver of dairy cows during ketosis based on comparative proteomics. Asian-Aust. J. Anim. Sci 21(7):1003-1010. Xu, J., M. A. Kirigiti, K. L. Grove, & M. S. Smith. (2009). Regulation of food intake and gonadotropin-releasing hormone/luteinizing hormone during lactation: role of insulin and leptin. Endocrinology 150(9):4231-4240. Xu, J. X., D. Xia, R. Q. Zhao, Q. F. Xu, X. H. Wei, & J. Chen. (2004). [The developmental changes of GHR and IGF-1R gene expressions in porcine hypothalamus and pituitary]. Yi Chuan Xue Bao 31(5):495-501. Xu, X., S. C. Rongali, J. P. Miles, K. D. Lee, & M. Lee. (2006). pat-4/ILK and unc-112/Mig- 2 are required for gonad function in Caenorhabditis elegans. Experimental Cell Research 312(9):1475-1483. Xu, Y., U. M. Bailey, & B. L. Schulz. (2015). Automated measurement of site-specific N- glycosylation occupancy with SWATH-MS. Proteomics 15(13):2177-2186. Xu, Z.-P., A. Dutra, C. M. Stellrecht, C. Wu, J. Piatigorsky, & G. F. Saunders. (2002). Functional and Structural Characterization of the Human Gene BHLHB5, Encoding a Basic Helix-Loop-Helix Transcription Factor. Genomics 80(3):311-318. Yambayamba, E. S., M. A. Price, & G. R. Foxcroft. (1996). Hormonal status, metabolic changes, and resting metabolic rate in beef heifers undergoing compensatory growth. J Anim Sci 74(1):57-69. Yang, H., X.-l. Xu, H.-m. Ma, & J. Jiang. (2016a). Integrative analysis of transcriptomics and proteomics of skeletal muscles of the Chinese indigenous Shaziling pig compared with the Yorkshire breed. BMC genetics 17(1):80.

206

Yang, J., L. Chen, X. Zhang, Y. Zhou, D. Zhang, M. Huo, & Y. Guan. (2008). PPARs and Female Reproduction: Evidence from Genetically Manipulated Mice. PPAR Research 2008:8. Yang, L., K. Tang, Y. Qi, H. Ye, W. Chen, Y. Zhang, & Z. Cao. (2012a). Potential metabolic mechanism of girls' central precocious puberty: a network analysis on urine metabonomics data. BMC Systems Biology 6(3):S19. Yang, X., J. Coulombe-Huntington, S. Kang, G. M. Sheynkman, T. Hao, A. Richardson, S. Sun, F. Yang, Y. A. Shen, R. R. Murray et al. (2016b). Widespread Expansion of Protein Interaction Capabilities by Alternative Splicing. Cell 164(4):805-817. Yang, Y., J. Wang, D. Bu, S. Li, T. Yuan, L. Zhou, J. Yang, & P. Sun. (2012b). Comparative proteomics analysis of plasma proteins during the transition period in dairy cows with or without subclinical mastitis after calving. Czech J Anim Sci 57(10):481-489. Yao, K., Y. L. Yin, W. Chu, Z. Liu, D. Deng, T. Li, R. Huang, J. Zhang, B. Tan, W. Wang et al. (2008). Dietary arginine supplementation increases mTOR signaling activity in skeletal muscle of neonatal pigs. J Nutr 138(5):867-872. Yasin, M., A. C. Dalkin, D. J. Haisenleder, J. R. Kerrigan, & J. C. Marshall. (1995). Gonadotropin-releasing hormone (GnRH) pulse pattern regulates GnRH receptor gene expression: augmentation by estradiol. Endocrinology 136(4):1559-1564. Yates, J. R., 3rd, S. Speicher, P. R. Griffin, & T. Hunkapiller. (1993). Peptide mass maps: a highly informative approach to protein identification. Anal Biochem 214(2):397-408. Ye, R., & P. E. Scherer. (2013). Adiponectin, driver or passenger on the road to insulin sensitivity? Mol Metab 2(3):133-141. Yelich, J. V., R. P. Wettemann, H. G. Dolezal, K. S. Lusby, D. K. Bishop, & L. J. Spicer. (1995). Effects of growth rate on carcass composition and lipid partitioning at puberty and growth hormone, insulin-like growth factor I, insulin, and metabolites before puberty in beef heifers. J Anim Sci 73(8):2390-2405. Yelich, J. V., R. P. Wettemann, T. T. Marston, & L. J. Spicer. (1996). Luteinizing hormone, growth hormone, insulin-like growth factor-I, insulin and metabolites before puberty in heifers fed to gain at two rates. Domest Anim Endocrinol 13(4):325-338. Yen, C.-F., H.-S. Wang, C.-L. Lee, & S.-K. Liao. (2014). Roles of integrin-linked kinase in cell signaling and its perspectives as a therapeutic target. Gynecology and Minimally Invasive Therapy 3(3):67-72. Yen, H. C., Q. Xu, D. M. Chou, Z. Zhao, & S. J. Elledge. (2008). Global protein stability profiling in mammalian cells. Science 322(5903):918-923. Yilmaz, A., M. E. Davis, & R. C. Simmen. (2006). Analysis of female reproductive traits in Angus beef cattle divergently selected for blood serum insulin-like growth factor I concentration. Theriogenology 65(6):1180-1190. Yonekura, S., T. Senoo, Y. Kobayashi, T. Yonezawa, K. Katoh, & Y. Obara. (2003). Effects of acetate and butyrate on the expression of leptin and short-form leptin receptor in bovine and rat anterior pituitary cells. Gen Comp Endocrinol 133(2):165-172. Yoo, S. W., T. Bolbot, A. Koulova, R. Sneeringer, K. Humm, Y. Dagon, & A. Usheva. (2013). Complement factors are secreted in human follicular fluid by granulosa cells and are possible oocyte maturation factors. J Obstet Gynaecol Res 39(2):522-527. Zacchi, L. F., & B. L. Schulz. (2016). SWATH-MS Glycoproteomics Reveals Consequences of Defects in the Glycosylation Machinery. Mol Cell Proteomics 15(7):2435-2447. Zachut, M. (2015). Defining the Adipose Tissue Proteome of Dairy Cows to Reveal Biomarkers Related to Peripartum Insulin Resistance and Metabolic Status. J Proteome Res 14(7):2863-2871.

207

Zachut, M., P. Sood, Y. Levin, & U. Moallem. (2016a). Proteomic analysis of preovulatory follicular fluid reveals differentially abundant proteins in less fertile dairy cows. Journal of Proteomics 139:122-129. Zachut, M., P. Sood, L. Livshitz, G. Kra, Y. Levin, & U. Moallem. (2016b). Proteome dataset of pre-ovulatory follicular fluids from less fertile dairy cows. Data in Brief 7:1515- 1518. Zakaria, M. M., K.-H. Jeong, C. Lacza, & U. B. Kaiser. (2002). Pituitary homeobox 1 activates the rat FSHβ (rFSHβ) gene through both direct and indirect interactions with the rFSHβ gene promoter. Molecular Endocrinology 16(8):1840-1852. Zaki, M. E., K. S. Amr, & M. Abdel-Hamid. (2013). APOA2 Polymorphism in Relation to Obesity and Lipid Metabolism. Cholesterol 2013:289481. Zamah, A. M., M. E. Hassis, M. E. Albertolle, & K. E. Williams. (2015). Proteomic analysis of human follicular fluid from fertile women. Clin Proteomics 12(1):5. Zamorano, P. L., V. B. Mahesh, L. M. De Sevilla, L. P. Chorich, G. K. Bhat, & D. W. Brann. (1997). Expression and localization of the leptin receptor in endocrine and neuroendocrine tissues of the rat. Neuroendocrinology 65(3):223-228. Zhang, Q., H. G. Lee, J. A. Han, E. B. Kim, S. K. Kang, J. Yin, M. Baik, Y. Shen, S. H. Kim, K. S. Seo et al. (2010). Differentially expressed proteins during fat accumulation in bovine skeletal muscle. Meat Sci 86(3):814-820. Zhang, Q. C., D. Petrey, J. I. Garzon, L. Deng, & B. Honig. (2013). PrePPI: a structure- informed database of protein-protein interactions. Nucleic Acids Res 41(Database issue):D828-833. Zhao, C., H. Zhang, W. C. Wong, X. Sem, H. Han, S. M. Ong, Y. C. Tan, W. H. Yeap, C. S. Gan, K. Q. Ng et al. (2009). Identification of novel functional differences in monocyte subsets using proteomic and transcriptomic methods. J Proteome Res 8(8):4028-4038. Zhao, H., M. Huang, Q. Chen, Q. Wang, & Y. Pan. (2012). Comparative gene expression analysis in mouse models for identifying critical pathways in mammary gland development. Breast Cancer Research and Treatment 132(3):969-977. Zhao, J., F. Yang, J. Feng, Y. Wang, B. Lachenbruch, J. Wang, & X. Wan. (2017). Genome- Wide Constitutively Expressed Gene Analysis and New Reference Gene Selection Based on Transcriptome Data: A Case Study from Poplar/Canker Disease Interaction. Frontiers in Plant Science 8(1876) Zhao, Y., X. Yan, X. Li, Y. Zheng, S. Li, & X. Chang. (2016). PGK1, a glucose metabolism enzyme, may play an important role in rheumatoid arthritis. Inflammation Research 65(10):815-825. Zhao, Y. M., U. Basu, M. V. Dodson, J. A. Basarb, & L. L. Guan. (2010). Proteome differences associated with fat accumulation in bovine subcutaneous adipose tissues. Proteome Sci 8:14. Zhen, S., M. Zakaria, A. Wolfe, & S. Radovick. (1997). Regulation of gonadotropin-releasing hormone (GnRH) gene expression by insulin-like growth factor I in a cultured GnRH- expressing neuronal cell line. Mol Endocrinol 11(8):1145-1155. Zhu, H., S. Shah, N. Shyh-Chang, G. Shinoda, W. S. Einhorn, S. R. Viswanathan, A. Takeuchi, C. Grasemann, J. L. Rinn, M. F. Lopez et al. (2010). Lin28a transgenic mice manifest size and puberty phenotypes identified in human genetic association studies. Nat Genet 42(7):626-630. Zieba, D. A., M. Amstalden, S. Morton, M. N. Maciel, D. H. Keisler, & G. L. Williams. (2004). Regulatory roles of leptin at the hypothalamic-hypophyseal axis before and after sexual maturation in cattle. Biol Reprod 71(3):804-812. Zoeller, R. T., P. H. Seeburg, & W. S. Young, 3rd. (1988). In situ hybridization histochemistry for messenger ribonucleic acid (mRNA) encoding gonadotropin-releasing hormone

208

(GnRH): effect of estrogen on cellular levels of GnRH mRNA in female rat brain. Endocrinology 122(6):2570-2577. Zulu, V. C., T. Nakao, & Y. Sawamukai. (2002a). Insulin-like growth factor-I as a possible hormonal mediator of nutritional regulation of reproduction in cattle. J Vet Med Sci 64(8):657-665. Zulu, V. C., Y. Sawamukai, K. Nakada, K. Kida, & M. Moriyoshi. (2002b). Relationship among insulin-like growth factor-I, blood metabolites and postpartum ovarian function in dairy cows. J Vet Med Sci 64(10):879-885.

209

APPENDICES

210

APPENDIX I. Polymorphisms and genes associated with puberty in heifers

Manuscript published in Theriogenology, July 2016

https://doi.org/10.1016/j.theriogenology.2016.04.046

Abstract

Puberty onset is a multifactorial process influenced by genetic determinants and environmental conditions, especially nutritional status. Genes, genetic variations, and regulatory networks compose the molecular basis of achieving puberty. In this article, we reviewed the discovery of multiple polymorphisms and genes associated with heifer puberty phenotypes and discuss the opportunities to use this evolving knowledge of genetic determinants for breeding early pubertal Bos indicus–influenced cattle. The discovery of polymorphisms and genes was mainly achieved through candidate gene studies, quantitative trait loci analyses, genome-wide association studies, and recently, global gene expression studies (transcriptome). These studies are recapitulated and summarized in the current review.

Keywords: Bovine, Genetic markers, Genome-wide studies, Puberty, PENK

211

APPENDIX II. SNP detection using RNA-sequences of candidate genes associated with puberty in cattle

Manuscript published in Genetics and molecular research, March 2017

http://dx.doi.org/10.4238/gmr16019522

Abstract

Fertility traits, such as heifer pregnancy, are economically important in cattle production systems, and are therefore, used in genetic selection programs. The aim of this study was to identify single nucleotide polymorphisms (SNPs) using RNA-sequencing (RNA-Seq) data from ovary, uterus, endometrium, pituitary gland, hypothalamus, liver, longissimus dorsi muscle, and adipose tissue in 62 candidate genes associated with heifer puberty in cattle. RNA- Seq reads were assembled to the bovine reference genome (UMD 3.1.1) and analysed in five cattle breeds; Brangus, Brahman, Nellore, Angus, and Holstein. Two approaches used the Brangus data for SNP discovery 1) pooling all samples, and 2) within each individual sample. These approaches revealed 1157 SNPs. These were compared with those identified in the pooled samples of the other breeds. Overall, 172 SNPs within 13 genes (CPNE5, FAM19A4, FOXN4, KLF1, LOC777593, MGC157266, NEBL, NRXN3, PEPT-1, PPP3CA, SCG5, TSG101, and TSHR) were concordant in the five breeds. Using Ensembl's Variant Effector Predictor, we determined that 12% of SNPs were in exons (71% synonymous, 29% nonsynonymous), 1% were in untranslated regions (UTRs), 86% were in introns, and 1% were in intergenic regions. Since these SNPs were discovered in RNA, the variants were predicted to be within exons or UTRs. Overall, 160 novel transcripts in 42 candidate genes and five novel genes overlapping five candidate genes were observed. In conclusion, 1157 SNPs were identified in 62 candidate genes associated with puberty in Brangus cattle, of which, 172 were concordant in the five cattle breeds. Novel transcripts and genes were also identified.

212

APPENDIX III. Pre and post-puberty expression of genes and proteins in the uterus of Bos indicus heifers: the luteal phase effect post-puberty

Manuscript published in Animal Genetics, September 2018

https://doi: 10.1111/age.12721

Abstract

Progesterone signalling and uterine function are crucial in terms of pregnancy establishment. To investigate how the uterine tissue and its secretion changes in relation to puberty, we sampled tissue and uterine fluid from 6 pre- and 6 post-pubertal Brahman heifers. Post-pubertal heifers were sampled in the luteal phase. Gene expression of the uterine tissue was investigated with RNA-Sequencing, while the uterine fluid was used for protein profiling with mass spectrometry. A total of 4,034 genes were differentially expressed (DE) at a nominal P-value of 0.05 and 26 genes were significantly DE after Bonferroni correction (P < 3.1x10-6). We also identified 79 proteins that were DE (P < 1x10-5) in the uterine fluid, out of 230 proteins. When we compared proteomics and transcriptome results, 4 DE proteins were identified as DE genes: OVGP1, GRP, CAP1 and HBA. Except for CAP1, the other 3 had lower expression post-puberty. The function of these 4 genes hypothetically related to preparation of the uterus for a potential pregnancy is discussed in the context of puberty. All DE genes and proteins were also used in pathway and ontology enrichment analyses to investigate overall function. The DE genes were enriched for terms related to ribosomal activity. Transcription factors that were deemed key regulators of DE genes were also reported. Transcription factors ZNF567, ZNF775, RELA, PIAS2, LHX4, SOX2, MEF2C, ZNF354C, HMG20A, TCF7L2, ZNF420, HIC1, GTF3A and 2 novel genes had the highest regulatory impact factor scores. This data can help to understand how puberty influences uterine function.

213

APPENDIX IV. A pipeline for the analysis of multi-omics data

Manuscript published in Proceeding of 22nd Conference of the Association for the Advancement of Animal Breeding and Genetics, July 2017

http://agbu.une.edu.au/AAABG 2017/66Nguyen22289.pdf

Abstract

We describe an analytical pipeline to exploit the results from RNA sequencing (RNA- Seq) experiments combining a series of processes from data normalization to network inference. The pipeline makes use of numerical approaches aimed at identifying key regulators via the regulatory impact factor (Reverter et al. 2010) metrics. It also employs the partial correlation and an information theory (Reverter and Chan 2008) for the identification of significant edges in the construction of gene co-expression networks. Key nodes in the network include differentially expressed genes, transcription factors, tissue specific genes as well as genes harboring SNPs found to be associated with the phenotype(s) of interest. The pipeline has already been successfully employed in two beef cattle studies, dealing with the onset of puberty and feed efficiency. In the present paper, we describe a pipeline to analyse RNA-Seq data, focus on relevant genes, generate gene co-expression networks and identify emerging clusters within the network to provide new insight about the subject matter under scrutiny.

214

APPENDIX V. Identification of pleiotropic key-regulatory genes related to economically relevant traits in beef cattle through systems biology approach

Manuscript published in Proceeding of the World Congress on Genetics Applied to Livestock Production, February 2018

http://www.wcgalp.org/proceedings/2018/identification-pleiotropic-key-regulatory-genes- related-economically-relevant

Abstract

In this investigation, a multi-breed and multi-omics approach was applied to study the pleiotropic effect from a list of key regulatory genes obtained from three independent studies evaluating fertility traits in beef cattle. A pleiotropic map for 32 traits (including growth, feed intake, carcass, meat quality, and reproduction) and a data-mining analysis using the CattleQTLdb were performed to identify candidate genes. This approach allowed the identification of genes with pleiotropic effects related to crucial biological process for fertility, production and health traits such as IYD, RBM20 and PLA2G4E. These genes will be further investigated to better understand the biological processes related with these complex traits. These genes could assist in improving selection decisions and reducing the frequency of undesirable genotypes.

Keywords: pleiotropic markers; production and reproduction traits, systems biology, multi- omics, beef cattle.

215

APPENDIX VI. ANIMAL ETHICS APPROVAL CERTIFICATE

216

UQ Research and Innovation Director, Research Management Office Nicole Thompson

ANIMAL ETHICS APPROVAL CERTIFICATE 22-Nov-2012

Activity Details Chief Investigator: Professor Stephen Moore, Queensland Alliance for Agriculture and Food Innovation Title: Gene Expression Studies for improving fertility of Australian Brahman cattle AEC Approval Number: QAAFI/279/12 Previous AEC Number: Approval Duration: 01-Aug-2012 to 01-Aug-2014 Funding Body: Group: Production and Companion Animal Other Staff/Students: Marina Fortes, Damien Smith, Sheree Boisen, Robert Johnson, Andrew Kelly, Christopher Shooter, Laercio Porto-Neto, Ashley Ostrofski, Beverley Hutchinson, Milou Dekkers, Michael McGowan, Gry Boe-Hansen, Sophia Edwards, Emily Piper, Lisa Kidd, Timothy Rosignol, Joao Paulo do Rego, Brett Teale

Location(s): Gatton Bldg 360 - CAAS

Summary Subspecies Strain Class Gender Source Approved Remaining Cattle Bos Indicus Adults Female Institutional 20 20 Breeding Colony

Permit(s):

Proviso(s):

Approval Details

Description Amount Balance

Cattle (Bos Indicus, Female, Adults, Institutional Breeding Colony) 18 Jul 2012 Initial Approval 20 20

UQ Research and Innovation Cumbrae-Stewart Building T +61 7 3365 2925 (Enquiries) E [email protected] The University of Queensland Research Road T +61 7 3365 2713 (Manager) W www.uq.edu.au/research/rid/ Brisbane Qld 4072 Australia F +61 7 3365 4455

Page 1 of 2 Please note the animal numbers supplied on this certificate are the total allocated for the approval duration

Please use this Approval Number: 1. When ordering animals from Animal Breeding Houses 2. For labelling of all animal cages or holding areas. In addition please include on the label, Chief Investigator's name and contact phone number. 3. When you need to communicate with this office about the project.

It is a condition of this approval that all project animal details be made available to Animal House OIC. (UAEC Ruling 14/12/2001)

The Chief Investigator takes responsibility for ensuring all legislative, regulatory and compliance objectives are satified for this project. The Chief and Alternative Investigators are jointly responsible for ensuring the project has, at all times, a nominated Chief Investigator. In the event that a regulatory obligation must be fulfulled and the nominated Chief Investigator is unavailable to satisfy that obligation, the obligation will by default fall to the Alternative Investigator.

This certificate supercedes all preceeding certificates for this project (i.e. those certificates dated before 22-Nov-2012)

Page 2 of 2