|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 EVOLUTION AND CELLULAR BIOLOGY OF THE UTERINE SERPINS By MARIA B PADUA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF P HILOSOPHY UNIVERSITY OF FLORIDA 2009
2 M aria B. P adua
3 I dedicate this work t o my mother Carmen, uncle Adn, nanny Zenaida and siblings Carolina and Fernando
4 ACKNOWLEDGMENTS I would like to thank my advisor Dr Peter J. Hansen for his guidance, support and challenge through my PhD program. I am grateful to him for encouraging me as a scientist during all these years. Also, I would like to thank my committee members Dr Nasser Chegini, Dr David Jul ian, Dr Phyllip LuValle, Dr Karen Moore for their contributions and suggestions to improve my research projects and academic training. I am also grateful to Dr William Thatcher for his continuous support and participation. I also thank my lab ma tes, Luciano and Aline Bonilla, Barbara Loureiro, Lilian Oliveira, Justin Fear Silvia Carambula and Jim Moss and my old lab mates, Dr Jeremy Block, Dr Dean Jousan, Moiss Franco, Dr Luiz Augusto de Castro and Amber Brad for their great help through man y different ways. Likewise, I want to thank my international friends Dr Yaser Al Katanani, Dr Fabiola Paula Lopes, Dr Zvi Roth, Dr Paolete Soto, and Dr Olga Ocon for their friendship and additional contributions. Thanks are also extended to all perso nnel in the Department of Animal Sciences, especially to Rick Wenzel for making the cell culture room a more functional place to work. Likewise, I want to thank the International Student Center staff, especially to Debra Anderson and Maud Fraser for their kindness and assistance to all international students. I am also grateful to Neal Bensen and Steve McClellan from the Flow Cytometry Core Laboratory, Dr Gigi Ostrow from the Gene Expression Core Laboratory, Dr Savita Shanker and Xiao Hui Zhou from the DNA Sequencing Core Laboratory and to the University of Florida Di agnostic Referral Laboratories of the Interdisciplinary Center for Biotechnology Research Cancer Genetics Research Complex at the University of Florida for their assistance to perform many o f the experiments presented in this dissertation.
5 Gratitude is also extended to the following University of Florida people for providing access to laboratory equipment: Dr Owen Rae and Shelley Lanhart from the Department of Large Animal Clinical Sciences Dr Pushpa Kalra from the Dep arment of Physiology and Functional Genomics, Dr Lori Warren, Jan Kivilpeto, Dr Lokenga Badinga and Dr Alan Ealy from the Department of Animal Sciences, University of Florida. Likewise, I am grateful to the following p eople for their invaluable assistance in providing tissue samples for some of the experiments presented in this dissertation: Dr Dan C. Sharp, Dr Luciano Silva and Dr. Claudia Klein from the Animal Sciences Department, University of Florida, Dr Ellis Gr einer from the college of Veterinary Medicine, University of Florida and Mirian Y Caas from the Laboratorio de Embriologia and Endocrinologia Molecular, Decanato de Agronomia, Universidad Centroccidental Lisandro Alvarado, Venezuela, Dr Douglas C Antcz ak and Christina Costa from the Baker Institute for Animal Health, College of Veterinary Medicine, Cornell University, Dr. John Verstegen from the College of Veterinary Medicine, University of Florida D r Fernanda Agreste from the Surgery Department, Facu ldade de Medicina Veterinria e Zootecnia, Universidade de So Paulo Brazil and Dr Stephen Shores and Debi Gibson from the Animal Shores Hospital in Gainesville. I am also grateful t o Dean Leed for his care and support. Special thanks go to Dr Andrs K owalski, Mnica Pra do Cooper, Jos Cristobal Nieto, Marlia and Jos Trujillo Armando Luiz and Claudia Garca for their sincere friendship for many years Also, I am grateful to my very good friends in Gainesville Domenicchella Dean, Cristina Caldari Tor res and Milerky Perdomo for their friendship and support in many different ways I really appreciate it.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .................... 4 LIST OF TABLES ................................ ................................ ................................ ................................ 9 LIST OF FIGURES ................................ ................................ ................................ ............................ 10 LIST OF ABBREVIATIONS ................................ ................................ ................................ ............ 12 ABSTRACT ................................ ................................ ................................ ................................ ........ 14 CHAPTER 1 LITERATURE REVIEW ................................ ................................ ................................ ........... 16 Introduction ................................ ................................ ................................ ................................ 16 Serine Proteina se Inhibitor Superfamily ................................ ................................ .................... 17 Mechanism of Action ................................ ................................ ................................ .......... 17 Non Inhibitory Serpins ................................ ................................ ................................ ........ 20 Functions of Inhibitory Serpins Distinct from Proteinase Inhibition ............................... 21 Serpinopathies ................................ ................................ ................................ ...................... 22 Receptor Mediated Internal ization of the Serpin Proteinase Complex ............................ 23 Evolution of Serpins and Their Functions ................................ ................................ ......... 26 Phylogeny of the Serpin Superfamily ................................ ................................ ................. 31 Positive Selection in Serpins ................................ ................................ ............................... 33 Novel Serpins and their Functions ................................ ................................ ...................... 35 Uterine Serpins ................................ ................................ ................................ ............................ 37 Ovine Uterine Serpin ................................ ................................ ................................ ........... 39 Endometrial secretion. ................................ ................................ ................................ 40 Biological function. ................................ ................................ ................................ ...... 41 Porcine Uterine Serpin ................................ ................................ ................................ ........ 44 Endometrial secretion. ................................ ................................ ................................ 44 Biological function. ................................ ................................ ................................ ...... 45 Bovine Uterine Serpin ................................ ................................ ................................ ......... 46 Endometrial secretion. ................................ ................................ ................................ 47 Biological function. ................................ ................................ ................................ ...... 47 Caprine Uterine Serpin ................................ ................................ ................................ ........ 48 Evolution and Phylogeny of Uterine Serpins ................................ ................................ ..... 48 Synopsis and Objectives ................................ ................................ ................................ ...... 50 2 MOLECULAR PHYLOGENY OF UTERINE SERPINS A ND ITS RELATIONSHIP TO EVOLUTION OF PLAC ENTATION ................................ ................................ ................ 51 Introduction ................................ ................................ ................................ ................................ 51 Materials and Methods ................................ ................................ ................................ ................ 53
7 Data Base Queries to Identify Uterine Serpin Genes ................................ ........................ 53 Reverse Transcription Polymerase Chain Reaction (RT PCR) ................................ ........ 53 Identification of cDNA for Equine Uterine S erpin Gene ................................ .................. 54 Sequencing of Amplicons ................................ ................................ ................................ ... 56 Detection of Equine Uterine Serpin by Western Blotting ................................ ................. 57 Amino Acid Sequence Alignments and Analysis of Phylogenetic Tree .......................... 57 Results ................................ ................................ ................................ ................................ .......... 59 Identification of Coding Sequences of Known and New Uterine Serpins Using Blastn ................................ ................................ ................................ ................................ 59 Uterine Serpin Gene Organization in the Bovine and Canine ................................ .......... 62 Characteristics of CanUS and Expression in Tissues ................................ ........................ 62 Uterine Expression and Amino Acid Sequence of EqUS ................................ ................. 62 Endometrial Secre tion of EqUS ................................ ................................ .......................... 66 Lack of Expression of the Uterine Serpin Gene in the Pregnant Cat ............................... 67 Amino Acid Sequence Conservation ................................ ................................ .................. 67 Identification of the Putative Hinge Region and P1 ........................ 70 Phylogenetic Analysis of Uterine Serpins ................................ ................................ .......... 72 Positive Selection of the Uterine Serpin Gene ................................ ................................ ... 72 Discussion ................................ ................................ ................................ ................................ .... 78 3 COMPARISON OF TH E NATIVE AND RECOMBINANT FORMS OF OVINE UTERINE SERPIN FOR INHIBITION OF CELL PROLIFERATION ................................ 85 Introduction ................................ ................................ ................................ ................................ 85 Materials and Met hods ................................ ................................ ................................ ................ 86 Materials ................................ ................................ ................................ ............................... 86 Collection of Uterine Fluid and Purification of Native OvUS ................................ ......... 87 [ 3 H]thymidine Incorporation by D17 and PC 3 Cells ................................ ....................... 87 Induction of Apoptosis in D17 and PC 3 Cells ................................ ................................ 88 Purific ation of His Tagged rOvUS from Conditioned Medium ................................ ....... 89 Galactosidase ................................ ................................ 90 Proliferation of P388D1 and PC 3 Cells ................................ ................................ ............ 90 Statistical Analysis ................................ ................................ ................................ ............... 91 Results ................................ ................................ ................................ ................................ .......... 92 Inhibition of Proliferation and Induction of Apoptosis in D17 and PC 3 Cells .............. 92 Antiproliferative Actions on P388D1 and PC 3 Cell Lines: Comparison of the Native and Recombinant Forms of OvUS ................................ ................................ ...... 92 Discussion ................................ ................................ ................................ ................................ .... 95 4 REGULATION OF DNA SY NTHESIS AND THE CELL CYCLE IN HUMAN PROSTATE CANCER CELL S AND LYMPHOCYTES BY OVINE UTERINE SERPIN ................................ ................................ ................................ ................................ ........ 97 Introduction ................................ ................................ ................................ ................................ 97 Materials and Methods ................................ ................................ ................................ ................ 99 Materials ................................ ................................ ................................ ............................... 99 Purification of rOvUS ................................ ................................ ................................ .......... 99
8 PC 3 Cell Culture ................................ ................................ ................................ .............. 100 [ 3 H]thymidine Incorporation by PC 3 Cells ................................ ................................ .... 101 Cell Proliferation Based on ATP Co ntent ................................ ................................ ........ 101 Cytotoxicity Assay ................................ ................................ ................................ ............. 102 TUNEL Labeling ................................ ................................ ................................ ............... 102 Secretion of IL 8 ................................ ................................ ................................ ................ 103 Cell Cycle Analysis ................................ ................................ ................................ ........... 104 Statistical Analysis ................................ ................................ ................................ ............. 105 Resul ts and Discussion ................................ ................................ ................................ ............. 105 Proliferation of PC 3 cells ................................ ................................ ................................ 1 05 Lactate Rehydrogenase Release ................................ ................................ ........................ 107 DNA Fragmentation (Apoptosis) ................................ ................................ ..................... 107 Interleukin 8 Secretion ................................ ................................ ................................ ...... 107 Cell Cycle Dynamics ................................ ................................ ................................ ......... 112 5 CHANGES IN EXPRESSION OF CELL CYCLE RELATED GENES IN PC 3 PROSTATE CANCER CELLS CAUSED BY OVINE UTERINE SERPIN ...................... 118 Introduction ................................ ................................ ................................ ............................... 118 Materials and Methods ................................ ................................ ................................ .............. 119 Materials ................................ ................................ ................................ ............................. 119 Purification of rOvUS ................................ ................................ ................................ ........ 120 PC 3 Cell Culture ................................ ................................ ................................ .............. 120 Proliferation Assay ................................ ................................ ................................ ............ 121 Cell Culture for RNA Extract ion ................................ ................................ ...................... 121 RNA Extraction ................................ ................................ ................................ ................. 121 cDNA Synthesis and Real Time PCR Array ................................ ................................ ... 122 Statistical Analysis ................................ ................................ ................................ ............. 122 Results ................................ ................................ ................................ ................................ ........ 124 Inhibition of PC 3 Cell Proliferation by OvUS ................................ ............................... 124 Cell cycle Related Gene Expression Profile at 12 h after Treatment with rOvUS ....... 124 Cell cycle Related Gene Expression Profile at 24 h after Treatment with rOvUS ....... 127 Discussion ................................ ................................ ................................ ................................ .. 127 6 GENERAL DISCUSSION ................................ ................................ ................................ ....... 134 LIST OF REFERENCES ................................ ................................ ................................ ................. 141 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ........... 156
9 LIST OF TABLES Table page 2 1 Primers used in RT PCR proce dure to obtain the full length coding sequence of the equine uterine serpin gene. ................................ ................................ ................................ .... 55 2 2 Exon and intron sizes for the uterine serpin gene of the cow and dog. .............................. 63 2 3 estimates and maximum log likelihood of models for positive selection within the protein coding sequence of uterine serpins ................................ ................................ ........... 74 2 4 Test of significance for models for positive selection within the protein coding sequence of uterine serpins. ................................ ................................ ................................ ... 75 5 1 Cell cycle related genes s creened using the RT 2 Profiler TM PCR Array. ......................... 123 5 2 Regulation of cell cycle related genes of PC 3 cells after 12 h of treatment with 200 g/ml recombinant ovine uterine serpin. ................................ ................................ ............ 126 5 3 Down regulation of human cell cycle related genes of PC 3 cells after 24 h of treatment with 200 g/ml recombinant ovine uterine serpin. ................................ ........... 128
10 LIST OF FIGURES Figure page 1 1 antitrypsin. A) Native stressed conformation of the protein. ................................ ................................ ................................ .............................. 18 1 2 Neighbor joining tree of 110 serpin sequences. ................................ ................................ ... 34 2 1 Identification of an incorrectly annotated dog corticosteroid binding globulin (CBG) as an uterine serpin (US). ................................ ................................ ................................ ....... 61 2 2 Representative electrophoretogram of amp licons obtained by reverse transcriptase polymerase chain reaction (RT PCR) of RNA fron canine tissue and canine uterine serpin primers. ................................ ................................ ................................ ........................ 64 2 3 Expression and secretion of equine uterine serpi n ................................ ............................... 65 2 4 Amino acid sequence alignment of the uterine serpins using the ClustalW algorithm. .... 68 2 5 Identification of the P1 site and hinge region (underlined) in uterine serpins. ........... 71 2 6 Phylogenetic tree of the uterine serpin proteins with the ovine uterine serpin (OvUS) as out group. ................................ ................................ ................................ ........................... 73 2 7 Selecton output generated for the uterine serpin group of proteins. ................................ ... 76 2 8 Amino acid sequence alignment of ovine uterine serpin (NP_001009304.1) and h antitrypsin (NP_000286.3) using the ClustalW algorithm ................................ 77 2 9 Phylogenetic tree of placentation in mammals (adapted from Vogel 2005) to illustrate the existence of uterine serpin genes relative to type of placentation. ................ 79 3 1 Effect of OvUS on induction of apoptosis in D17 and PC 3 cells as determined by TUNEL labeling. ................................ ................................ ................................ .................... 93 3 2 Inhibition of [ 3 H]thymidine incorporation of P388D1 cells and PC 3 cells by native (n) and recombinant (r) OvUS. ................................ ................................ .............................. 94 4 1 Inhibition of [ 3 H]thymidine incorporation of PC 3 cells by recombinant ovine uterine serpin (rOvUS) ................................ ................................ ................................ ..................... 106 4 2 Inhibition of proliferation of PC 3 cells by recombinant ovine uterine serpin (rOvUS) as determined by ATP content/well. ................................ ................................ ................... 108 4 3 Lack of cytotoxic effect of recombinant ovine uterine serpin (rOvUS) on PC 3 cells was measured by the release of lactate dehydrogenase. ................................ .................... 109
11 4 4 Representative photomicrographs of PC 3 cells labeled using the TUNEL procedure the control protein ovalbumin. ................................ ................................ ............................ 110 4 5 Effect of recombinant ovine uterine serpin (rOvUS) on DNA fragmentation (apoptosis) of PC 3 cells. ................................ ................................ ................................ ..... 111 4 6 Effect of recombinant ovine uterine serpin (rOvUS) on interleukin ( IL) 8 concentration in cell culture supernatants of PC 3 cells. ................................ ................... 113 4 7 Cell cycle dynamics of PC 3 cells as affected by recombinant ovine uterine serpin (rOvUS). ................................ ................................ ................................ ............................... 114 4 8 Cell cycle dynamics of lymphocytes as affected by recombinant ovine uterine serpin (rOvUS). ................................ ................................ ................................ ............................... 116 5 1 Inhibition of [ 3 H]thymidine incorporation of PC recombinant ovine uterine serpin (rOvUS) ................................ ................................ ........ 125 5 2 Points in the cell cycle where genes were differentially regulated by ovine uterine serpin at 12 and 24 h are repr esented in panel A and B, respectively. ............................. 130 6 1 Possible pathways by which OvUS could block cell proliferation. ................................ .. 138
12 LIST OF ABBREVIATIONS Bo Bovine Cap Caprine CBG Corticosteroid Binding Globulin CD Cluster of Differentiation CDK Cyclin Dependent Kinase Con A Concanavalin A DPBS G CSF Granulocyte Macrophage Colony Stimulating Factor HSP 47 Heat Shoc k Protein 47 IFN Interferon IL Interleukin LPS Lipopolysaccharide LRP1 Low Density Lipoprotein Receptor Related Protein 1 Maspin Mammary S erine P roteinase I nhibitor MENT Myeloid and Erythroid Nuclear Termination S tage specific P rotein Mya Million Years Ago NK Natural Killer OVA Ovalbumin Ov Ovine PAI Plaminogen Activator Inhibitor PC 3 Human Prostate C ancer Cells 3 PEDF P igment E pithelium D erived F actor
13 PHA Phytohemagglutinin PK Protein Kinase Polyinosinic Po lycytidylic Acid Po Porcine PWM Pokeweed Mitogen RAP Receptor Associated Protein Rb Retinoblastoma protein RCL Reactive Center Loop RT PCR Reverse Transcription Polymerase Chain Reaction rOvUS Recombinant Ovine Uterine Serpin SERPIN Serine Proteinase Inhibitor STZ S treptozotocin TNF Tumor Necrosis Fa ctor TUNEL Terminal Deoxynucleotidyl T ransferase (TdT) and fluorescein isothiocyanate conjugated dUTP nick end labeling US U terine S erpin
14 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EVOLUTION AND CELLULAR BIOLOGY OF THE UTERINE SERPINS By M aria B. P adua May 2009 Chair: Peter J. Hansen Major: Animal Molecular and Cellular Biology Uterin e serpins (US) are a unique group of progesterone induced glycoproteins in a restricted group of mammals that belong to the serine proteinase inhibitor superfamily and which are secreted in large quantities into the uterus during pregnancy. The US gene ha s been identified in species with epitheliochorial placentation of the Ruminantia and Suidae orders of the Laurasiatheria superorder of eutherian mammals. One goal of this dissertation was to examine the evolution of the US gene in mammals. A US gene wa s identified in horses and dogs and found expressed in the uterus during pregnancy. The dog is a species with endotheliochorial placenta, suggesting that the US gene is not restricted to species with epitheliochorial placentation. However, its absence in other mammals, and apparent loss in the cat, suggests that the US gene evolved only within the Laurasiatheria superorder. The US do not appear to be functional proteinase inhibitors and to have species specific functions. The most studied member of the g roup, ovine uterine serpin (OvUS), inhibits proliferation of several cell types including activated lymphocytes, bovine pre implantation embryos and some tumor cell lines. A second goal was to evaluate the mechanism by which OvUS inhibits cell proliferati on. Ovine US blocked cell cycle progression of human prostate
15 cancer (PC 3) cells, causing an accumulation of cells at G_2/M at 12 h and at G_0/G_1 at 24 h after treatment addition. Additionally, OvUS blocked the cell cycle progression in phytohemaggluti nin stimulated lymphocytes increasing the number of cells at the G_0/G _1 stage at 96 h after treatment. B locking of the cell cycle progression by OvUS was caused specifically by the up regulation of cell cycle checkpoint and arrest genes such as CDKN1A (p 21), CDKN2B (p15) and CCNG2 (cyclin G2) and down regulation of genes involved in DNA synthesis and cell cycle regulation and progression. The finding that the US gene is only retained in a limited group of mammals suggests its importance for successful pr e gnancies in these species. It is also possible that the US gene evolved a distinct and specie dependent function from an ortholog serpin gene, rather than the typical anti proteolytic activity conserved in most members of the superfamily.
16 CHAPTER 1 L ITERATURE REVIEW Introduction Pregnancy in mammals has evolved in a process that included the appearance of new genes for novel functions. Many of t hese new genes ar o se by gene duplication to generate paralog s which underwent sequence divergence from the parental genes (Louis 2007). Recently, it was determined specifically in the mouse placenta that most of the genes present during early events of placentation, such as those involved in basic growth and metabolism, are evolutionary ancient genes and their ortholog genes are also found within eukaryotes (Knox & Baker 2008). Similar results were found in the human decidua where ortholog genes are also present in all vertebrates (Knox & Baker 2008 ). In contrast, those genes expressed in the mature placenta are rodent specific genes which suggest that these genes are newly evolved since no orthologs were found in other eukaryot e or vertebrate s (Knox & Baker 2008). Some examples of genes involved in reproductive processes that were formed as a result of gene duplication are the primate chorionic gonadotropins (Maston & Ruvolo 2002) and interferon et al. 1999). The uterine serpins (US) present in the uterus of a limited group of mammals are another example. In sheep, cattle, and pigs an ancestral gene or genes of the serine proteinase inhibitor (serpin) superfamily underwent modification to exhibit high expression in the uterus and under the regulation of progesterone. Unlike most of the members of the serpin superfamily, US probably have not reta ined the anti proteolytic activity that defines this superfamily (Irving et al. 2000). Instead, evidence suggests they have divergent functions as for example, regulation of the immune function in the sheep (Hansen 1998 ) and placental iron transport in th e pig ( Roberts & Bazer 1988). The purpose of this dissertation is to more closely examine the evolution of the US gene
17 in mammals and to determine the mechanisms by which the US produced by the sheep inhibits cell proliferation Serine Proteinase Inhibito r Superfamily The serpins ( ser ine p roteinase in hibitors) are a n important superfamily of proteins presents in at least some members of all phyla. In vertebrates serpins are involved in the regulation of proteolytic pathways including blood coagulation an antitrypsin, antithrombin), complement cascade (complement C1 inhibitor) and tissue remodeling (plasminogen activator inhibitor 1 and 2) (Irving et al. 2000) Although the inhibitory activity of serpins is mostly towards serine proteina ses, serpins can also function as inhibitors of proteinases that belong to other families such as cysteine proteinases and papain like cysteine proteinases (Silverman et al. 2001, Irving et al. 2002 a Hungtinton 2006). Mechanism of A ction Inhibitor y serpins inactivate their target proteinases through a unique suicide substrate like inhibitory mechanism The s helices (A sheets (A C) (Figure 1 1). The key region for inhibitory serpins is the reactive center loop (RCL), which is a flexible structure localized on the top of the serpin and contains a complementary sequence to the active site of the target proteinase. The RCL is usually formed by 20 25 amino acids and the hinge region is enclosed in the P15 P9 portion of it (Irving et al. 2000 Hungtinto n 2006 Whisstock and Bottomley 2006). The hinge region provides mobility to the RCL. There is a consensus pattern present in the sequence for inhibitory serpins. Arginine is conserved at the P17 position, arginine, lysine or glutamic acid at P16, g lycine is usually present at P15 threoni ne or serine is present at P14 position and acid residues with short side chains such as alanine, glycine or serine are abundant in positions P12 to P9 (Irving et al. 2000).
18 Figure 1 1. Structure of the inhibitory serpin antitrypsin. A) Native stressed conformation of the protein. sheets are shown in green, red and yellow respectively. The reactive center loop (RCL) is shown in pink (1 ) with the hinge region (2 ). A lso shown is the breach (3 ), gate (4 ) and shutter (5 ). B) Cleaved relaxed conformation of the protein. The sheet A in 7. Reproduced from Irving et al (2000) with permiss ion of Genome Rese arch ( 2000) and after slight modification. 3 1 4 2 5 6 7 A B
19 The proteinase binds covalently to RCL and cleaves the scissile bond of the serpin at the P1 serpin and the proteinase catalytic serine. Then, the amino terminal portion of the RCL with the attached proteinase moves as much as 70 to the opposite side to form th sheet A, and the compression of the proteinase against the base of the serpin causes a distortion in the structure of the proteinase causing its inactivation (Figure 1 1) (Irving et al. 2000, Silverman et al. 2001 Hungtinton 2006). This conformationa l change in the structure, from the stressed (non cleaved RCL) to the relaxed (cleaved RCL) form, increases the overall stability of the serpin. This characteristic makes serpins highly unusual M ost proteins a re in a thermodynamically favorable conforma tion in their native state. In contrast, serpins in their native state (i.e. the stress conformation ) require lower temperatures for denaturation (Silverman et al. 2001 Im et al. 2002 Hungtinton 2006 Whisstock & Bottomley 2006). There are some other im portant regions in the structure of the serpins besides the hinge sheet sheet A occurs (Figure 1 1) (Irving et a l. sheet A. Together with the breach, the shutter helps open the sheets to allow the insertion of the hinge region of the RCL (Irving et al. 2000). Finally, the gate formed by strands s3C and s4C is region where the RCL sheet A without cleavage (Irving et al. 2000). Some serpins are relatively inactive proteinase inhibitors, but they become fully activated as in hibitors by the binding of cofactors such as heparin and vitronectin (Jin et al. 1997, Schvartz et al. 1999 Zhou et al. 2003). This is the case of antithrombin, heparin cofactor II, plasminogen
20 activator inhibitor 1, protease nexin I and protein C inhibi tor, where the RCL of these serpins is sheet A, making the RCL and the P1 residue less accessible for the proteinase. Once the cofactor interacts with the serpin, there is a conformational change, where the RCL is expelled and the P1 residue becomes accessible for the target proteinase (Law et al. 2006, Whisstock & Bottomley 2006). Likewise, heparin can enhance the inhibitory function of already active inhibitory serpins. A set of experimen ts conducted by Gupta & Gowda antitrypsin and thereby increase the inhibitory potential of this serpin. Non Inhibitory S erpins There is a group of serpins that do not inhibit proteinases. Examples of this growing group includes ovalbumin (OV A), the chicken egg storage protein (Benarafa & Remold 2005), mammary serine protease inhibitor (maspin), which inhibits angiogenesis and tumor growth and increases the sensitivity of cancer cells t o undergo apoptosis (Zhang 2000, Sheng 2006), th e hormone transport proteins corticosteroid and thyroxine binding globulin (Pemberton et al. 1988), angiotensinogen, involved in blood pressure regulation, water and salt homeostasis (Morgan et al. 1996 Stanley et al. 2006), the chaperone heat shock prote in (HSP) 47 (Nagata 1998, Sauk et al. 2005) and the multifunctional pigment epithelium derived factor (PEDF), which promotes differentiation of neurons and retinal photoreceptors, and neuronal degeneration while inhibiting angiogenic processes in the retin a, antagonizing the angiogenic effects of vascular endothelial growth factor, platelet derived growth factor and interleukin (IL) 8, and promoting apoptosis in endothelial cells ( Irving et al 2000, Silverman et al. 2001, Tombran Tink & Barnstable 2003).
21 F unctions of Inhibitory S erpins Distinct from Proteinase Inhibition Serpins are versatile proteins and can function in processes through mechanisms not antitrypsin, which i s produced by hepatocytes, neutrophils, macrophages, monocytes, intestinal epithelial cells, astrocytes and some other cells to inhibit the enzymatic activity of different prote in ases such as neutrophil elastase, cathepsin G, proteinase 3, chymotrypsin, tr ypsin thrombin (Travis & Salvesen 1983 antitrypsin is increased by the pro inflammatory molecules IL 1, IL 6, tumor necrosis factor (TNF) (Congote 2007). Thus, it is assumed that the physiological role of this serpin is to prev ent host tissue from the proteolytic damage caused by the activity of these proteases. antitrypsin has functions that are unrelated to its anti proteolytic activity. Breit et al antitrypsin inhibited in a dose dependent manner the proliferative response of human peripheral blood lymphocytes induced by the antitrypsin did not completely inhibit these proliferative responses. Moreo ver, this serpin did not cause any inhibitory activity against lymphocytes activated by the T and B cell mitogen pokeweed (PWM) (Breit et al. antitrypsin competed with diferric transferrin for the bind ing of the transferrin receptors inhibiting the proliferation of human skin fibroblasts (Graziadei et al. 1998). antitrypsin on cytokines have been also pursued. Alpha1 antitrypsin, but not the secretory leukocyte protease inhibitor, in hibited in a dose dependent manner the release of the pro inflammatory cytokines TNF 1, IL 8 and IL from LPS stimulated human monocytes (Janciauskiene et al. 2004). Moreover, in the same antitrypsin increas ed IL 10 release from human monocytes stimulated with LPS.
22 antitrypsin on the pro inflammatory cytokines were not blocked when neutralizing antibodies against IL 10 were included in the experiments (Janciauskiene et a l. 2004). antitrypsin in some human diseases such as type I diabetes has been also studied. Alpha1 antitrypsin prevented the apoptotic effects of TNF streptozotocin (STZ a diabetes inducer ) on the MIN6 mouse insuli noma cell line through caspase 3 activity inhibition (Zhang et al. antitrypsin to the non obese diabetic (NOD) mouse prevented the development of type 1 diabetes and reduced insulin autoantibodies (Song et al. 2004 ). Moreover, Lewis et al. antitrypsin prolonged islet graft survival i n C57BL/6 mice treated with STZ Similar results were obtained from experiments using C57BL/6 mice also treated with STZ, where the ant itrypsin reduced the levels of glucose in the blood, number of apoptotic cells and rate of diabetes (Zhang et al. 2007). Serpinopathies The importance of the serine proteinase inhibitors is also highlighted by the loss of function, deficiency and disea ses (also known as serpinopathy) caused by mutations, leading to a serpin existing in a non functional state. Serpin polymerization is a non functional state which is achieved by the binding of the sheet A, or in some cases to sheet C of other serpin, forming a long inactive polymer. This phenomenon causes the accumulation of polymerized serpins as misfolded protein, therefore decreasing the amount of functional serpins available (Kaiserman et al. 2006 Law et al. 2006). antitrypsin polymerization could lead to emphysema and liver cirrhosis and the polymerization of neuroserpin could cause familial encephalopathy (FENIB) (Kaiserman et al. 2006, Law et al. 2006). The other non functional state is latency
23 whe sheet A without being cleaved by the target proteinase causing the inactivation of the serpin (Kaiserman et al. 2006). Serpin mutations also alter the functionality and stability of the protein. Mutations can induce polym 1 antitrypsin Z mutant), cancer (loss of function of maspin) change in serpin specificity ( antitrypsin Pittsburg mutation, where an amino acid substitution at the P1 site of this serpin generates a potent inhibitor of thrombin and other coagulation enzymes, caus ing death by bleeding disorders) and also loss of cofactor binding site to disrupt the interaction between heparin and antithrombin III (Law et al. 2006, Kaiserman et al. 2006). In addition, gain o f function serpinopathy can occur when genetic mutations lead to overexpression of a serpin. The excess in the synthesis of megsin, a serpin produced by human mesangial cells, causes accumulation and polymerization of the protein in the endoplasmic reticu lum of the kidney glomerular epithelial cells (Inagi et al. 2005, Miyata et al. 2005). This renal serpinopathy is characterized by m esangial matrix expansion stress of the endoplasmic reticulum and augmented IgA deposition which leads to renal failure ( Miyata et al. 2005, Kaiserman et al. 2006 ). T he overexpression of the mutant megsin, which lacks anti proteolytic activity and flexibility at the hinge region loop, does lead to polymerization, but not to the generation of a serpinopathy suggesting the i mportance of the overall structure and function of megsin for the development of the disease (Inagi et al. 2005 Miyata et al. 2005 ). Receptor Mediated I nter nalization of the Serpin Proteinase C omplex Once the target proteinase is bound to the serpin, a stable complex is formed by the inactive proteinase and the cleaved serpin. This serpin proteinase complex is cleared rapidly from the circulation by an internalization pathway mediated by hepatic receptors that leads to lysosomal degradation. Perlmutter et al. (1990) showed a specific cell surface receptor in human
24 monocytes and hepatoma (HepG2) cells, which was referred as the serpin enzyme complex receptor (SEC receptor). The SEC receptor recognized and subsequently internalized the following serpin p rot antitrypsin antichymotrypsin cathepsin G, antithrombin III thrombin and also complement C1 inhibitor C1s but with less affinity (Perlmutter et al. 1990, Poller et al. 1995). Later, experimental data revealed that most of t he serpin enzyme complexes are also internalized by the low density lipoprotein (LDL) family of endocytic receptors which includes the very low density lipoprotein (VLDLR), the low density lipoprotein receptor related protein 1 (LRP1), which was originally known as LRP or occasionally as CD91, the glycoprotein 330 (gp330) also known as megalin or LRP2, and the low density lipoprotein (LDLR) receptors (Kounnas et al. 1996, Skeldal et al. 2006, Lillis et al. 2008). Since the SEC receptor was never sequenced and the LRP1 is abundantly found in the liver, this receptor is the most likely candidate for the internalization of the serpin proteinase complexes. Kounnas et al. (1996) demonstrated that addition of either LRP 1 antibodies the antagonist chaperone rec eptor associated protein (RAP) or chloroquine (an inhibitor of lysosomal degradation) to the HepG2 cells inhibited the internalization and d antitrypsin trypsin, antithrombin III thrombin and heparin cofactor II thrombin complexes. Moreover, these serpin proteinase complexes were not internalized by mouse fibroblasts lacking LRP1 receptors (Kounnas et al. 1996). In addition native or cleaved forms of the serpins alone were not recognized as ligands and subsequently internalized by the LRP1 receptor, suggesting the specificity of the receptor for serpin proteinase complexes only (Kounnas et al. 1996). Moreover, in vivo experi ments using the rat as a model showed that the antagonist RAP delayed the internalization of radiolabeled antithrombinIII thrombin complexes from plasma circulation
25 (Kounnas et al. 1996). Similar experiments demonstrated that other serpin proteinase compl exes such as complement C1 inhibitor antitrypsin neutrophil elastase, protease nexin 1 thrombin, protease nexin 1 factor XIa, neuroserpin tissue type plasminogen activator and protease nexin 1 factor VII activating protease are internalized by the endocytic LRP1 receptor (Storm et al. 1997, Poller et al. 1995 Crisp et al. 2000, Knauer et al. 2000, Makarova et al. 2003, Muhl et al. 2007). The role of h eparin in the binding of the serpin proteinase complexes to the LRP1 receptor and subsequent i nternalization has been also studied, in particular the binding of protease nexin 1 to heparin after the serpin proteinase complex has formed. Experiments performed by Knauer et al. (1997) tested the role of heparin and LRP1 in the internalization of a va riant form of the protease nexin 1 (PN1 K7E) which does not bind heparin. They showed that the complex PN1 K7E thrombin could form but, internalization of the complex and further degradation was greatly reduced in normal mouse embryonic fibroblasts and ne arly abolished in the LRP1 deficient mouse embryonic fibroblasts, when compared with the experimental controls. Moreover, in the absence of endocytosis, the addition of soluble heparin inhibited the binding of the serpin proteinase complex to the surface of human and mouse embryonic fibroblast cells (Knauer et al. 1997). The authors concluded that cell surface heparins mediated the binding of protease nexin 1 thrombin complex to the LRP1 and further internalization. Similar experiments were performed by Crisp et al. (2000) to study the role of heparins, in the internalization of the protease nexin 1 urokinase plasminogen activator (uPA) complex through the LRP1. These authors showed that the mechanisms of cell surface binding and internalization of this serpin proteinase complex are heparin independent. These results indicate that the
26 interaction of the serpin proteinase complex with cell surface receptors in the endocytic pathway is determined by the specific serpin or proteinase involved. There is als o a specific receptor, t he urokinase plasminogen activator receptor (uPAR) known as well as CD87, that plays a role in the internalization of the uPA and tissue type plasminogen activator (tPA) complexes formed with either plasminogen activator inhibitor ( PAI) 1 or PAI 2 (Olson et al. 1992, Wind et al. 2002). Even though active uPA binds with high affinity to uPAR, its internalization is much slower than the uPA PAI 1 complex (Wind et al. 2002). Moreover, the internalization of uPA PAI 1 bound to the uPAR receptor and the later recirculation of the receptor requires the interaction of the macroglobulin receptor/low density lipoprotein receptor surface glycoprotein (Binder et al. 2007). There is experimental evidence showing that the internalization of the uPA PAI 1 complex can be also mediated through the related receptor sorting protein related receptor (sorLA), LRP2, LRP1 and VLDLR, all members of the LDLR family of endocytosis receptors (Wind et al. 2002, Skeldal et al. 2006). Likewise, it was recently demonstrated by Chroucher et al. (2006) that the internalization of the uPA PAI 2 complex is also achieved by the LRP1 receptor. Evolution of S erpins and T heir Functions It was previously believed that serpin genes were restricted to higher eukaryotes. However, recent studies indicate that s erpins have an ancient origin, at least 1000 million years old, and they are not only limited to higher eukaryotes organisms, but instead are present in at least some species of all phyla on Earth. Serpin genes have been found in some complete prokaryote g enomes including Archaea and Bacteria (Irving et al. 2002 b ) In the Bacteria genera, serpin genes are present in cyanobacteria, actinobacteria and formicutes, but not in any proteobacteria family with complete
27 genomes (Irving et al. 2002 b Roberts et al. 2004 Ivanov et al. 2006). The conservation of residues in functional key regions and secondary structural elements of these serpins suggests their ability to inactivate proteinases (Irving et al. 2002 b Roberts et al. 2004). Ivanov et al. (2006) demonst rated that the serpin found in the actinobateria Bifidobacterium longun inhibited human neutrophil elastase and porcine pancreatic elastase. Kang et al. (2006) showed that there are three serpin genes in the formicute Clostridium thermocellum (Clotm serpi n 1, 2 and 3) The Clotm serpins 1 and 2 are close related, with the hinge region of the RCL very conserved with inhibitory serpins whereas Clotm serpin 3 is a diverg ent serpin In the same study, it was demonstrated that serpin 1 did not inhibit trypsin chymotrypsin or papain, but it had inhibitory activity towards subtisilin, a bacterial serine proteinase. Multiple serpin genes were found in two species of the Archaea and Bacteria domains where the variability of the RCL suggested that these genes we re generated by duplication and later diversification (Roberts et al. 2004) Despite the cases described above, serpin sequences are not broadly present in the prokaryotic kingdom. No serpin sequences were identified in 13 Archaea and 56 Bacteria genera with complete genome sequences, which suggested that in prokaryotes the serpin gene was probably acquired by lateral transfer or lost when it was not vital (Irving et al. 2002 b Roberts et al. 2004, Ivanov et al. 2006). In viruses, serpin genes have been only identified in the poxviridae family and they can be localized in extracellular and intracellular compartments (Brooks et al. 1995, Lucas & McFadden 2004). Three active serpin genes were found in the Orthopoxviruses genus. The viral serpin 1 (SP I 1) is a 40 kDa protein that blocked apoptosis in rabbitpox virus infected cells (Brooks et al. 1995). The viral serpin 2 (SPI 2) or CrmA inhibited efficiently caspase 1, preventing the production of IL et al. 1995, Callus & Vaux 2007). Also, it was found that
28 SPI 2 inhibited caspases 3 and 8 (Brooks et al. 1995, Callus & Vaux, 2007). Finally, the viral serpin 3 (SPI 3) is involved with regulation of virus host cell fusion (Brooks et a l. 1995). Although the amino acid sequence of SPI 1 is almost 50% identical to SPI 2, the predicted RCL of these two genes differs significantly (Brooks et al. 1995). Irving et al. (2000) suggested the possibility of gene duplication from a single gene a s the origin for SPI 1 and 2, but separately from SPI 3. Another viral serpin gene, (SERP I) was found in the myxoma virus of the Leporipoxvirus genus. SERP I is a 55 kDa glycosylated secreted protein which binds and inhibits human host plasmin, uPA and tPA. The knockout SPI 1 myxoma virus demonstrated its importance on reducing the immune response of host monocytes and macrophages (Lucas & McFadden 2004). S erpin genes are also present in several green algae and plants. In the plant kingdom, serpin g enes are present in moss and conifers ( Roberts & Hejgaard 2008 ). Moreover, serpins have been found in the phloem sap of the cucurbits pumpkin and cucumber and in very high quantities in the seeds of barley, wheat, rye, oats and apple (Roberts & Hejgaard 2 008). There is very little information about the genomic organization of most of the members of the plant kingdom. There is only one serpin gene present in the unicellular green algae Chlamydomonas reinhardt i i 7 serpin genes in Arabidopsis and 14 in ric e (Roberts & Hejgaard 2008). Although the first plant serpin genes identified were inhibitory, recent findings in maize and rice have shown the expression of close related non inhibitory serpin genes (Roberts & Hejgaard 2008). There are no studies on the evolution of serpin genes in green algae and plants however, Roberts & Hejgaard (2008) have postulated the possibility of a lateral gene transfer from bacteria to algal ancestors and inherited from the most advanced green algae when they evolved to land plants.
29 The function of serpins in green algae and plants remains unclear. Most of the known proteinases are found in plants. Typical animal caspases are absent, but cysteine proteinases are involved in the programmed cell death (Ro berts & Hejgaard 200 8). Green algae and plants also lack genes homologous to those encoding trypsin, elastase, thrombin, granzyme B. One chymotrypsin like proteinase has been discovered in plants, and in vitro studies have shown that serpins from wheat, barley, rye, oat, pu mpkin, apple and Arabidopsis are able to inhibit serine proteinases of the chymotryp sin group (Roberts & Hejgaard 2008). Silverman et al. (2001) postulated that phloem sap serpins could be involved in the host mechanism of defense, since the phloem serpin 1 found in Cucurbita maxima decreased the ability of the piercing sucking aphids, Myzus persicae to survive and reproduce on pumpkins (Yoo et al. 2000). Likewise, it has been proposed that the function of serpins found in high concentrations in plant seed s protects storage proteins from degradation caused by proteolytic enzymes of fungi o r insects (Roberts & Hejgaard 2008). Serpin sequences have been not found in complete genomes of Saccharomyces cerevisiae Schizosaccharomyces pombe (fission yeast) (Roberts et al. 2004) However, a recent study conducted by Steenbakkers et al. (2008) revealed that the presence of one serpin gene in the fungal kingdom. Analysis of the amino acid sequence of this so called s to have all the features of a prototypical functional serpin (Steenbakkers et al. 2008). Moreover, the serpin gene is present in the unicellular eukaryote Entamoeba histolytica which lacks of an N terminus signal peptide even though it is secreted by ac tivated trophozoites (Riahi et al. 2004). The interaction between this serpin and trypsin is typical of non inhibitory serpins where the proteinase is still active while the serpin is cleaved in
30 its C terminus (Riahi et al. 2004). In the same study, it w as shown that the serpin had non inhibitory activity towards chymotrypsin, elastase and cathepsin G. Additionally, serpins are present in early metazoans. Cole et al. (2004) cloned the full length cDNA of a serpin gene called jellypin from the cnidari an jellyfish Cy anea capillata Analysis of the amino acid sequence showed that this serpin lacks of N terminal signal sequence as is characteristic of intracellular serpins. In the same study, it was determined that jellypin contained a functional RCL an d inhibited human chymotrypsin, elastase, and capthesin G, however, under unphysiological conditions (Cole et al. 2004). Since the closest relative of jellypin in the serpin superfamily are plant serpins, the authors suggest that this metazoan serpin aros e approxi mately 1000 million years ago (M ya), around the time of divergence between plants and animals (Cole et al. 2004). In nematodes, eight serpin genes have been found in Caenorhabditis elegan s and the same numbers of serpin genes are present in parasi tic species such as Ascaris suum Brugia malayi Onchocerca volvulus Trichostrong y lus vitrinus and Shistosoma mansoni (Zang & Maizels 2001) The role for these serpins has been related to the evasion of the s immune response and limiting tissue dama ge in response to inflammation since BmSPN 2 inhibits cathepsin G and elastase, both produced b y neutrophils (Zang & Maizels 2001). In these species there are also small residues forming five disulphide bonds where two of these are located at each side of the RCL, and no structural relation with the well known members of the superfamily (Zang & Maizels 2001). In the dog hookworm A. caninum smapins inhibits serine protea ses that are involved in & Maizels 2001).
31 The percent of identity found between the C. elegans serpins CeSPN 3 and CeSPN 4 is 87% while the identity between CeSPN 7 and CeSPN 8 is 78% (Zang & Maizels 2001). T he authors concluded that these serpin genes in C. elegans arose as a local duplication event (Zang & Maizels 2001). Moreover, it was also suggested that a possible intron gain and loss is the model for the serpin gene evolution in nematodes (Zang & Maize ls 2001). Phylogeny of the S erpin S uperfamily In humans, serpins are classified in nine groups A I where the largest group are the extracellular clade A or SERPINA with 13 members found in chromosomes 1, 14 and X and the intracellular clade B or SERPINB with 13 members on chromosomes 18 and 6 (van Gent et al. 2003, Law et al. 2006, Izuhara et al. 2008). The groups SERPINC and SERPIND are represented by antithrombin and heparin cofactor II respectively (van Gent et al. 2003). The serpins PAI 1 and glial 2 antiplasmin and PEDF in SERPINF, and the complement C1 inhibitor in the SERPING group (van Gent et al. 2003 Kaiserman et al. 2006). The SERPINH gro up is represented by the HSP 47, the myoepithel ium neuroserpin are in the SERPINI group (van Gent et al. 2003 Kaiserman et al. 2006). The phylogenetic analysis conducted by Irving et al. (2000) divided the serpin superfamily into 16 clades or classes (A P). The largest clades in the classification are Clade A and B and they are formed by extracellular or antitrypsin like serpins and intracellular or ov like serpins respectively. This last group of serpins is suggested to be ancestors t o the majority of ext racellular serpins. The other clades are formed by serpins that have been found in insects, nematodes, plants and virus; however, there are ten highly diverged orphan serpins which failed to group within any of the clades; these proteins are also known as orphan serpins (Irving et al. 2000 Silverman et al. 2001). T he viral serpins SPI 1 and 2 were clustered together in clade n
32 but SPI 3 was classified alone in clade o (Irving et al. 2000) In addition, the myxoma serpin SERP I was placed in clade e cl ustered together with plasminogen activator inhibitor 1 and protease nexin 1 (Irving et al 2000). A study performed only with the nematode serpins grouped all the C. elegans serpins together in one cluster whereas serpins present in other species were sep arated in different branches as an indication of divergence (Zang & Maizels 2001). These findings were similar to those found by Irving et al. (2000) where the C. elegans serpins were clustered together in clade l w hereas the Schistosoma serpins are clust ered in clade m and all the other nematode serpins are consider orphans. Cole et al. (2004) conducted also a phylogenetic study to determine the relationship of the metazoan serpin jellypin with other members of the serpin family. The results classif ied jellypin as an orphan, but its closest relatives were the plant serpins (Cole et al. 2004), which are clustered in clade p (Irving et al. 2000) More complex phylogenetic analyses were performed by Atchley et al. (2001) (Figure 1 2) and Ragg et al. (2 001) where they took into consideration amino acid sequences, exon intron structures and diagnostic of amino acid sites. The results from the study performed by Atchley et al. (2001) were similar to those presented by Irving et al. (2000) where serpins we re clustered within ten clades or groups with two largest groups, the antitrypsin and ovalbumin clades, and several orphan serpins. Since the study by Ragg et al. (2001) was performed using only antitrypsin and ovalbumin groups being the largest All three studies classified antithrombin III and HSP 47 in separate clades or groups (Irving et al. 2000, Ragg et al. 2001, Atchley et al. 2001). However, the intron exon data from Atchley et al. (2001) and Ragg et al. (2001) showed that neuroserpins,
33 PAI 1 and glia derived nexin are grouped together within a clade suggesting a single evolutionary lineage, whereas Irving et al. (2000) clustered neuroserpins into a separate group from the other two serpins. Likewise, the intron exon data clustered toge ther complement C1 antiplasmin (Atchley et al. 2001, Ragg 2001) whereas Irving et al. (2000) placed the protease C1 inhibitor as a separate clade. Positive Selection in Serpins Positive or Darwinian selection refers to selective p ressure acting on proteins that can provide evolutionary advantage to an organism The most used method to detect positi vely synonymous substitutions (dN), where a nucle otide substitution will lead to an amino acid replacement of the encoded protein, with the synonymous substitutions (dS) where the substitution of a nucleotide will not change the amino acid leaving the encoded protein unchanged (Wong et al. 2004). When t evidence of neutral or purifying selection respectively (Wong et al. 2004). There is enough ev idence suggesting that serpins, like a few other protein families, have undergone positive or Darwinian selection. Positive selection has occurred at the RCL region of the protein, in particular at the P1 site proximal to the scissile bond of the RCL whic h changes the selectivity for th e target proteinase (Brown 1987, Hill & Hastie 1987, Goodwin et al. 1996, Zang & Maizels 2001, Barbour et al. 2002). Experimental data from Barbour et al. (2002) showed that antitrypsin isoforms from two different species of mouse inhibited different serine proteinases. Moreover, in the same study it was determined that this diversity in the functionality of these isoforms was caused in particular by the variations in the RCL area
34 Figure 1 2. Neighbor joining tree of 110 serpin sequences. The clusters of proteins show close correspondence to the groups of proteins described by the analyses of exon intron structure Atchely et al. (20 01) by permission of the Oxford Journals, Oxford University Press.
35 (Barbour et al. 2002). In contrast, van Gent et al. (2003) did not find any positive selection in a study that compared some serpin groups between human and rodents, human and artiodactyla and rat versus mouse. The p ositive selection of the serpin genes has been reported in different species in which it is suggested the advantage of this process during the evolution of the serpins in order to neutralize a broader range of proteinases For example, it has been po stulated that caspase inhibitor genes in viruses have undergone positive selection as a defense mechanism against the et al. 1995, Callus & Vaux 2007). Likewise, in nematodes, the rapid diversification of the RCL has allowed novel changes in these serpins and smapins in order & Maizels 2001). There is also circumstantial evidence suggesting that the evolution of the RCL of serpins present in plant seeds is b ecause of positive Darwinian selection to inhibit exogenous proteinases from insects and pathogens (Roberts & Hejgaard 2008). Nove l Serpins and their F unctions As previously mentioned, the mechanisms of actions and functions of some serpins has been exte nsively studied and well documented. However, recent discoveries have found new serpins with novel characteristics and interesting biological functions. One example is the novel nuclear serpin, myeloid and erythroid nuclear termination stage specific pro tein (MENT) which has a n a large insertion between the C and D helices. The M loop is also known as the CD loop and contains a nuclear localization signal and the AT hook motif which is required for chromatin and DNA binding (Silverman et al. 2 001 Irving et al. 2002 a ). This intracellular serpin is the major non histone chromatin protein localized in the heterochromatin and it functions by inducing higher order chromatin condensation by connecting separate nucleosomes in vitro (Silverman et al. 2001 Irving et al. 2002 a ). In addition, MENT is one of those cross
36 class serpins that inhibits cysteine proteinases and papain like cysteine proteinases such as cathepsins K, L and V (Irving et al. 2002 a ). Experimental data showed that MENT is involved in cell cycle progression and therefore in cell proliferation by inhibiting the enzymatic activity of the nuclear cysteine proteinase SPase, a cathepsin L like proteinase, whose function has been related to the degradation of the phosphorylated form of th e retinoblastoma (Rb) protein, a known regulator of the cell cycle (Irving et al. 2002 a ). Plasminogen activator inhibitor type 2 is another interesting serpin with novel function. The bi topological PAI 2 is predominantly translated as an intracellular 47 kDa non glycosylated protein, but it can be also a 60 kDa glycosalyted secreted protein as a combination of an inefficient nuclear localization signal and the lack of a conventional hydrophobic amino terminal sequence (Medcalf & Stasinopoulos 200 5, Crouch er et al. 2008). Like the serpin MENT, the CD loop is present in PAI 2, which binds non covalently to several proteins including the Rb protein (Medcalf & Stasinopoulos 2005). Although the physiological role of the intracellular PAI 2 remains unknown, it has been suggested that PAI 2 could be involved in the cell cycle regulation since this serpin protected Rb from degradation by a distinct anti proteolytic mechanism ( Medcalf & Stasinopoulos 2005, Croucher et al. 2008). A therapeutic role of PAI 2 in th e reduction of tumor growth and metastasis of different types of cancer has been extensively demonstrated using in vitro systems and rodent tumor models where uPA was inhibited completely and a significant decrease in the extra cellular matrix degradation achieved (Croucher et al. 2008). High levels of expression of PAI 2 in breast cancer tumors are associated with reduced metastasis, tumor size and prolonged survival whereas opposite effects are expected when this serpin is expressed at low levels (Crouch er et al. 2008).
37 More recently, another novel serpin was discovered in the visceral adipose tissue of the obese type 2 diabetes rat model. This new serpin called visceral adipose tissue derived serpin (vaspin) was found to be up regulated in the obese type 2 diabetes rat model and down regulated in the non obese diabetes resistant rat model (Hida et al. 2005). The vaspin transcript was up regulated in the subdermal white adipose tissue by the administrations of insulin and thiazolidinedion es ( TZD ) a compound that enhances insulin sensitivity (Hida et al. 2005). The authors suggested from the experimental data that vaspin inactivates unknown target proteases derived from fat or other tissue that inhibited the effects of insulin (Hida et al 2005). However, it was demonstrated in the same study that vaspin lacks anti proteolytic activity towards common serine proteinases such as trypsin, factor Xa, elastase, urokinase, collagenase and dipeptidyl peptidase (Hida et al. 2005). The vaspin gen e is also expressed in the white adipose tissue of mouse and human (Hida et al. 2005). The US are also novel with respect to serpins by virtue of their limited distribution among mammals, since n o human homologue of the US gene has been reported (van Gent et al. 2003). The primary site of expression for the US is in the endometrium and regulation of their expression by progesterone. A detailed description of their characteristics, properties and plausible functions for this group of serpins are described below. Uterine Serpins Uterine serpins were first described as progesterone induced proteins in the uterine fluid of the pregnant sheep, pig and cow (Moffatt et al. 1989, Baumbach et al. 1986, Leslie et al. 1990). The protein in sheep, discovered f irst, was called uterine milk protein (UTMP) because it was the major protein in uterine fluid (called uterine milk by Aristotle). The two US present in the pig were originally known as uteroferrin associated protein or uteroferrin associated basic protei ns ( UfAP / UABP ) because they form non covalently heterodimers with uteroferrin (Baumbach et al.
38 1986, Roberts & Bazer 1988). The connection between UTMP and UfAP was first made when the genes were sequenced. Ing & Roberts (1989) found greatest homology be tween the cDNA antitrypsin and classified UTMP as a member of the serine proteinase inhibitor (serpin) superfamily. Shortly thereafter, Malathy et al. (1990) found that both pig proteins were serpins and also highly similar to ovine UTMP. Mathialagan & Hansen (1996) identified the bovine UTMP as a serpin closely related to sheep UTMP and pig UfAPs and proposed that old designations be eliminated and proteins termed as uterine serpins (US) preceded by the name of the specie f rom where it was identified such as (Bo) for bovine, (Ov) for ovine and (Po) for porcine. Since this paper, additional US were identified in the goat (Tekin et al. 2005 a ) and a sequence for water buffalo ( Bubalus bubalis ) has been submitted to Genbank (Ac cession number: DQ 661648.1 ) Uterine serpins have been only characterized in species with epitheliochorial placenta. All US proteins that have been described to date possess a 25 amino acid signal peptide sequence and are secreted into the uterine lumen ( Mathialagan & Hansen 1996, Tekin et al. 2005 a ). Mathialagan & Hansen (1996) determined that most of the amino acids present in OvUS, BoUS and PoUS 1 and 2 necessary for the serpin backbone structure are conserved at the same pos antitrypsin. However, a nalysis of the hinge regions of all US indicates that these serpins are not conserved with inhibitory serpins and are probably not functional proteinase inhibitors (Irving et al. 2000 Tekin et al 2005 a ). I n the hinge region of US, the only amino acid that is conserved with the inhibitory serpins is alanine at the P12 position whereas the other positions are not (Tekin et al. 2005 a ). In addition, the amino acid at P1 nserved among species where valine is usually found with the exception of OvUS which has an alanine at this position (Tekin et al. 2005 a ).
39 The OvUS was screened for anti proteinase activity against several proteinases without identification of a target p roteinase ( Ing & Robets 1989). Both BoUS and OvUS are capable of inhibiting pepsin at very high concentrations ( Mathialagan & Hansen 1996, Peltier et al. 2000 a ), but this activity occurs even when the protein is proteolytically cleaved and may be distinct from the classical serpin inhibitory mechanism involving a functional RCL ( Peltier et al. 2000 a ). Uterine serpins have a conserved KVP sequence at the P4 P2 site of the RCL (Tekin et al. 2005 a Mathialagan & Hansen 1996). There is also a unique 39 amin o acid insertion in the BoUS, which incorporates two KAKEVPAVVKVPM repeats within the putative P1 one KEVPVVVKVP sequence right after that P1 Mathialagan & Hansen 1996, Tekin et al. 2005 a ). The KEVPVVVKVP motif is also found further upstream as a conserved sequence among all known US ( Mathialagan & Hansen 1996) The real function for all these repeats remains unknown. Mathialagan & Hansen (1996) suggested that US may inhibit some members of the aspartic proteinase family, because o f the resemblance of the motif to the VVVK present in pep sinogens However Peltier et al. (2000 a ) showed that the addition of a synthetic peptide corresponding to the putative P7 P15 region of OvUS did not inhibit pepsin activity Ovine Uterine S erpin This is the most studied protein of the US group. Ovine US is a secreted protein which is present in the uterine fluid as a pair of basic glycoproteins with molecular weights of 55,000 and 57,000, derived from a single 54,000 precursor (Moffatt et al. 198 7 Hansen et al. 1987 a ). This basic glycoprotein possess an isoelectric point of 9.2 and two N linked glycosylation sites, which suggests that the two major forms of the protein may differ in the number of carbohydrate chains (Hansen et al. 1987 a Ing & R obets 1989). The a mino acid co mposition was found to be rich in lysine, leucine and threonine, but low in tyrosine and devoid of tryptophan ( Hansen et al. 1987 a ). Analysis of the carbohydrate content of OvUS showed that it consists of 2.8% neutral
40 sugars 2.5% amino sugars, and 0.3% sialic acid (Hansen et al. 1987 a ). Additionally, OvUS possess a mannose 6 phoshate the so called lysosomal recognition marker ( Hansen et al. 1987 a ). Endometrial secretion. The major regulator of OvUS secretion is the hormon e progesterone (Moffatt et al. 1987, Ing et al. 1989, Leslie & Hansen 1991, Padua et al. 2005). Experiments performed using the ovariectomized ewe as a model showed that after 6 days of treatment with progesterone, OvUS can be detected in endometrial epit helial cells (Ing et al. 1989). Moreover, progesterone therapy for 14 60 days induced a large increase in the secretion of the protein (Ing et al. 1989, Leslie & Hansen 1991 Padua et al. 2005). Ovine US can be detected in uterine secretions of ovariecto mized ewes treated for 30 days with a combination of progesterone plus estrone, however, the protein could not be detected when they were treated with estrone only (Moffatt et al. 1987). In addition, levels of OvUS mRNA increased in the glandular epitheli um of ovariectomized ewes receiving uterine infusion of ovine placental lactogen and/or ovine growth hormone combined with IFN as compared with the levels in ovariectomized ewes infused with control protei ns (Spencer et al. 1 999, Noel et al. 2003). It was also shown that the co administration of estradiol with progesterone where estradiol up regulates the expression of progesterone receptors in the endometrium, decreased OvUS mRNA in the glandular epithel ium ( Spencer et al. 1999 ). During the estrous cycle, OvUS mRNA can be detected in uterine endometrium around days 13 16 (Stewart et al. 2000). In the pregnant sheep, the transcript for OvUS can be first detected at days 13 15 (Ing et al. 1989, Stewart et al. 2000). Then, a 3 fold increase of steady state levels of OvUS mRNA in the glandular epithelium occurs between days 20 and 60, another 3 fold between days 60 and 80, and a decline at day 120 of pregnancy (Stewart et al. 2000).
41 Between days 20 and 50 of pregnancy, the expression of OvUS mRNA is lower in the deep glandular epithelium than in the upper glandular epithelium of the uterine stratum spongiosum. There was however, no difference in mRNA expression between the upper and lower glandular epithel ium of the stratum spongiosum between days 50 and 60 of gestation (Stewart et al. 2000). High levels of expression of the transcript for OvUS can be detected in all glandular epithelium in the stratum spongiosum between days 60 and 120 of pregnancy (Stewa rt et al. 2000). At day 1 of postpartum, OvUS mRNA was only detected in the stratum spongiosum of the glandular epithelium, but during postpartum days 7 and 28, the transcript for OvUS was not longer detected (Gray et al. 2003). It appears that OvUS is initially produced only by the uterine glands and then its expression spreads to the luminal epithelium Although OvUS protein was found only in glandular epithelium at day 60 of pregnancy, it was immunolocalized in the luminal and glandular epithelium of the intercaruncular endometrium by days 120 to 140 of pregnancy (Moffatt et al. 1987, Stephenson et al. 1989 a ). Ovine US has been also detected in amniotic and allantoic fluids (Newton et al. 1989, McFarlane et al. 1999), presumably because of transplace ntal transport. Biological function. The biological function of OvUS is still not known with certainty although evidence suggests it functions primarily to inhibit cell proliferation. Experiments performed by Ing & Roberts (1989) tested t he anti protease activity of OvUS against trypsin, chymotrypsin, plasmin, thrombin, elastase and plasminogen activator without significant inhibition. Likewise, OvUS had non inhibitory activity towards recombinant human cathepsin D, porcine cathepsin D, recombinant cathe psin E, cysteine proteinase cathepsin B ( Mathialagan & Hansen 1996) or dipeptidyl proteinase IV (DPPIV) ( Liu & Hansen 1995).
42 Mathialagan & Hansen (1996) demonstrated that freshly purified OvUS inhibited pepsin A and C (chymosin) activity at both pH 2.0 and 4.5, but an excess of 35 and 8 fold molar of OvUS was required for a 50% inhibition of both aspartic proteinases, respectively. Moreover, the complex formed by OvUS and pepsin could not be detected electrophoretically in the presence of Sodium Dodecyl S ulfate (SDS) ( Mathialagan & Hansen 1996) whereas, complexes formed by inhibitory serpins and their partner proteinases (at 1:1 ratio) are so stable that they can be resolved using the same approach (Potempa et al. 1994 ). Furthermore, the binding of OvUS t o pepsin seems to be electrostatic since it occurred at pH 8 and low salt concentration where OvUS has positive charge and pepsin has a negative charge, but not at pH 10.25 where both proteins are negatively charged ( Peltier et al. 2000 a ). Some studi es have been conducted to determine the possible role of OvUS as a binding or carrier protein. Ovine US can bind pregnancy associated glycoproteins ( Mathialagan & Hansen 1996 ), which are inactive aspartic proteinases secreted in large amounts by the ungul ate placenta (Xie et al. 1991 Green et al. 1998). Ovine US can also bind to activin, a member of the transforming growth factor et al. 1999). However, the binding affinity of activin for OvUS is much lower than activin affinity towards fo l listatin, suggesting that unlike fo l listatin, OvUS is not a candidate to neutralize the biological activity of activin. In addition, OvUS can bind IgM and IgA ( Hansen & Newton 1988). However, the binding of OvUS to immunoglobulins was inhibited by the presence of high salt concentratio ns, indicating the ionic nature of the binding ( Hansen & Newton 1988). A plausible function for OvUS is the inhibition of immune cell proliferation during pregnancy to provide protection for the allogeneic ally distinct conceptus (Hansen 1998). Purified O vUS inhibited lymphocyte proliferation induced by mitogens such as the antigen
43 Candida albicans Con A, PHA and in the mixed lymphocyte reactions (Segerson et al. 1984, Hansen et al. 1987 b Stephenson et al. 1989 b Skopets & Hansen 1993, Skopets et al. 199 5). In addition, OvUS reduced the antibody titer in ewes immunized against the T cell dependent antigen OVA (Skopets et al. 1995). However, antitrypsin ( Breit et al. 1983), OvUS did not cause any inhibitory activity against lymphocytes activated by the T and B cell mitogen PWM ( Skopets & Hansen 1993). Ovine US also inhibits NK cell activity. In vivo experiments using the pregnant mi ce as a model (C), which is a 50 base pair compound consisting of double stranded RNA that activates NK cells, and reduced basal splenocyte NK cell activity ( Liu & Hansen 1993 ). Additional experiments conducte d by Liu & Hansen (1993) demonstrated that OvUS inhibited NK like activity in sheep lymphocytes and mouse splenocytes against K562 and YAC 1 target cells. Moreover, the lytic activity of NK like cells in sheep peripheral blood lymphocytes and endometrial epithelium against D 17 cells infected with bovine herpes virus 1 was inhibited by OvUS ( Tekin & Hansen 2002 ). Some experiments have been conducted to elucidate the mechanism by which OvUS inhibits cell proliferation. The protein does not act by inducin g cytotoxicity because OvUS did not cause cytotoxic effects on lymphocytes ( Stephenson et al. 1989 b Skopets & Hansen 1993). Binding of OvUS to lymphocytes is specific, dose dependent and saturable (Liu et al. 1999). However, it is not known if OvUS bind s to a specific cell surface receptor or competes with other molecules for a receptor binding site. Experiments with Rp 8 Cl cAMPS, a selective inhibitor of cAMP dependent type I protein kinase (PK) A, indicated that effects of OvUS on proliferation of PH A stimulated lymphocytes are not dependent on this kinase ( Tekin et al. 2005 b ). A study performed by Peltier et al. (2000 b ) showed that OvUS reduced proliferation of
44 phorbol myristol acetate (PMA) activated lymphocytes, suggesting that OvUS inhibits some PKC mediated events. In the same study, the protein blocked IL 2 induced proliferation and reduced expression of CD25 (IL 2 mRNA caused by Con A (Peltier et al. 2000 b ). Skopets & Hansen (19 93) demonstrated that the antiproliferative effect of OvUS on lymphocytes was not blocked by addition of neutralizing antibody to transforming growth factor immunosuppr essive growth factor. Porcine U terine S erpin The pig is the only species shown to contain two US genes (PoUS 1 and PoUS 2) with predicted molecular weights of 45,123 and 46,009 respectively ( Mathialagan & Hansen 1996). The PoUS 1 has four potential N gl ycosylation sites, whereas PoUS 2 presents only three, lacking the one at ( Asn 107 ) (Malathy et al. 1990, Mathialagan & Hansen 1996). Analysis of purified PoUS showed three different peptides. There are two larger forms of the protein with molecular weigh ts of 50 and 46 kDa and identical NH 2 amino terminus, but different in the number or complexity of glycosylated chains (Baumbach et al. 1989, Murray et al. 1989). There is also a small glycosylated polypeptide of 40 kDa, distinct from the other two in its NH 2 amino terminus (Baumbach et al. 1986 Murray et al. 1989) that may represent a proteolytic cleavage product. These three polypeptides seemed to derive from a single 45 kDa precursor and like OvUS, all of them posses the mannose 6 phosphate lysosome r ecognition signal (Murray et al. 1989). Endometrial secretion. The porcine protein is present in uterine flushings of pregnant, pseudo pregnant and progesterone treated pigs (Baumbach et al. 1986). In addition, it can be detected in allantoic fluid of co nceptuses at mid pregnancy (Baumbach et al. 1986, Murray et al. 1989). In
45 endometrial tissue, the protein was immunolocalized only in the glandular epithelium of pregnant pigs at days 45 and 60 (Murray et al. 1989). During pregnancy, PoUS mRNA expressio n is low at days 13 and 30, but the expression of the transcript increases until the last month of gestation, which in the pig lasts 115 days (Malathy et al. 1990). Moreover, high level of expression of the PoUS transcript was also detected in endometrial tissues from day 68 pseudo pregnant gilts (Malathy et al. 1990). In contrast, PoUS mRNA expression was no detected in either endometrium or liver of non pregnant or ovariectomized pigs (Malathy et al. 1990). Biological function. There is little evidence that PoUS 1 or PoUS 2 are inhibitory serpins. Purified PoUS did not inhibit proteinases such as trypsin and chymotrypsi n (Malathy et al. 1990). Mathialagan & Hansen (1996) tested the ability of PoUS to inhibit porcine pepsin using uterine flushes from l ong term progesterone treated pigs. The inhibitory activity of the protein was only present in some batches, had very low stability, and the activity was lost completely when the batch was kept at 4 o C for more than one day. The main role of the PoUS has been related to the iron transport from the mother to the pig conceptuses by its association with the uteroferrin protein. Porcine uteroferrin is a purple colored alkaline phosphatase whose secretion by the glandular epithelium is also stimulated by prog esterone (Chen et al. 1975, Buhi et al. 1982 Baumbach et al. 1986, Murray et al. 1989). Uteroferrin is transported by the areolae of the placenta into the allantoic fluid (Chen et al. 1975, Ducsay et al. 1982, Renegar et al. 1982). Uteroferrin carries t wo bound iron atoms per polypeptide chain (Baumbach et al. 1986, Roberts & Bazer 1988, Nuttleman & Roberts 1990). Data from experiments performed by Buhi et al. (1982) in which [ 59 Fe] labeled uteroferrin was injected into selected allantoic sacs of days 5 8 and 67 pregnant gilts showed that the [ 59 Fe] was
46 released from uteroferrin and quickly associated with the protein transferrin, also present in the allantoic fluid, and then from transferrin to the fetus. In addition, it was demonstrated that uteroferri n was more susceptible to proteolysis after iron depletion (Buhi et al. 1982 Nuttleman & Roberts 1990). Porcine US binds non covalently to uteroferrin. This pink color heterodimer has a molecular weight of 80 kDa and is very stable for long periods of time in the presence of oxygen whereas purple uteroferrin is not (Baumbach et al. 1989; Roberts & Bazer 1988). Also, the heterodimer conformation (Baumbach et al. 198 9). Experimentally, the heterodimer can be dissociated by low pH, antibodies raised against either uteroferrin or all three PoUS peptide forms, or by dialysis at low salt concentration (Baumbach et al. 1989). Vallet (1995) suggested that another pos sible role for PoUS during pregnancy is to prevent embryonic losses caused by uteroferrin ascorbic acid induced lipid per oxidati on which is harmful for smaller or less develop conceptuses. Experiments performed on microsomal membranes from reproductive ti ssues showed that the uteroferrin PoUS heterodimer in the presence of ascorbic acid was less able to provoke lipid peroxidation damage than uteroferrin itself (Vallet 1995). Bovine Uterine Serpin The bovine uterine serpin (BoUS) is also a basic protein a nd several molecular weight forms can be detected in uterine secretions from pregnant and progesterone treated cows by western blotting, with two major bands at 57,000 and 38,000 (Leslie et al. 1990, Leslie & Hansen 1991). Unlike the porcine and ovine US, the BoUS has only one potential site for glycosylation (Asn 243 ), which is conserved with OvUS and PoUS 1 and 2 ( Mathialagan & Hansen 1996).
47 Endometrial secretion. Secretion of BoUS into the uterus is dependent upon progesterone. Using western blotting, Leslie & Hansen (1991) demonstrated induction of BoUS in uterine secretions of ovariectomized cows treated with progesterone for 10 to 30 days. The bovine US is secreted by the endometrial epithelium of the cow. The protein was immunolocalized in the gl andular epithelium of endometrial tissues from day 135 of pregnancy and ovariectomized cows treated for 30 days with progesterone (Leslie et al. 1990, Leslie & Hansen 1991). However, no immunostaining was detected in either the glandular or luminal epithe lium of endometrial tissues from cows on days 17, 19, 21 of pregnancy or days 17 and 19 of the estrous cycle (Leslie et al. 1990). Unlike OvUS, the BoUS mRNA is expressed during estrus, with particular highly expression in the cranial area of the uterine horns (Bauersachas et al. 2005). During estrous, t he mRNA for BoUS was highly expressed in the superficial uterine glands and more weakly expressed in the deep glandular epithelium and some dispersed epithelial cells of the endometrium (Bauersachas et al 2005). In addition, the BoUS transcript can be detected in the caruncule, ovary and cotyledon (Khatib et al. 2007). Biological function. The biological function of BoUS remains unknown. Mathialagan & Hansen (1996) showed that activity of porcine pep sin was inhibited by crude uterine flushing obtained from progesterone treated cows. The importance of BoUS for physiologic function is highlighted by recent experiments by Khatib et al. (2007) indicating that a single nucleotide polymorphism (A/G) at pos ition 1269 of the BoUS gene wa s associated with a significant increase in productive life in cattle populations, where the G allele expression at that position has been linked to the productive life. Productive life is in large part determined by culling rate and
48 culling is largely for problems with reproductive health or infectious diseases. Thus, BoUS is likely to play a role in one or more processes controlling reproduction or immune function. Caprine Uterine Serpin This US was recently found discove red. Caprine (Cap) US was immunolocalized in the endometrial glandular epithelium at day 25 of pregnancy, but it was not detected at day 5 of the estrous cycle (Tekin et al. 2005 a ). Western blot of uterine fluid from a pregnant goat identified a major ba nd with a molecular weight of approximately 57,000 and some other low molecular weight bands (Tekin et al. 2005 a ). Like the ovine and porcine, the CapUS had two predicted N glycosylation sites at positions 222 and 268, where the position 268 is conserved with the other known US and the 222 position is also conserved in almost all the other US, with the exception of the BoUS (Tekin et al. 2005 a ). No study has been performed to identify the possible biological role of this US. Evolution and Phylogeny of Uterine S erpins Krem & Cera (2003) performed a study looking for possible evolutionary markers for antitrypsin numbering) of the shutter region, where the serine 56 is related to the TCN AGY codon usage dichotomy that is associated to the protostome deuterostome division. The authors determined that the OvUS, BoUS, PoUS 1 and 2 posses a conserved serine (encoded by TCN codons) at position 53 which is also preserved in all the sepins used fo r the study. At the position 56 they found that the PoUS 1 and 2 also posses a conserved serine 56 (AGY codon) also found in almost all the serpins present in chordates (Krem & Cera 2003). However, OvUS and BoUS were among some other few serpins where di stinct residue was found at position 56, where instead of serine, alanine was present at this position (Krem & Cera 2003).
49 Uterine serpins have been classified as a highly antitrypsin clade (Irving et al. 2000). Accordi ng to the results obtained from another phylogenetic study, OvUS and PoUS 1 and 2 were grouped into a large clade containing serpins with different antitrypsin, antichymotrypsin, angiotensinogen where genes co ding for these serpins contain three introns at homologous areas in the conserved part of the coding region (Atchley et al. 2001). A similar study performed only on vertebrate serpins by Ragg et al. (2001) inferred that OvUS and PoUS 1 and 2 could classif y within the antitrypsin group. However, the authors stated the need for further verification for these results due to the lack of genomic organization for these three US. In contrast, Irving et al. (2000) created phylogenetic trees with preexisting alignments of the Pfam database. In this analysis, the US and the angiotensin like serpins were grouped as separate clades rather than being included antitrypsin group. A phylogenetic study by Peltier et al. (2000 c ) also resulted in US being considered as a separate clade in the serpin superfamily. The identity of the predicted amino acid sequence of OvUS with the CapUS is 96%, 82% with the BoUS and 55 and 56% with the PoUS 1 and 2 respectively ( Mathialagan & Hansen 1996, Tekin et al. 2005 a ). Peltier et al (2000 c ) estimated that OvUS and BoUS diverged from the PoUS 1 and 2 about 60 Mya and the PoUS 1 and 2 diverged from each other around 5 Mya Computational analysis of US motifs suggested that OvUS and BoUS contain similar amounts of casein kinase 2 and cAMP phosphorylation sites. In contrast, neither PoUS 1 nor 2 contain cAMP phosphorylation ( Peltier et al. 2000 c ). The same study also showed that these three species have similar amounts of phosphorylation sites for tyrosine kinase s as well as for N myr istoylation. However, the PoUS 1 and 2 contain more phosphorylation sites for casein kinase 2 and PK C and N linked glycosylation sites than OvUS and BoUS.
50 Synopsis and Objectives Among the members of the serpin superfamily, the US group is a very intrigu ing group of proteins. Based on this literature, there is evidence showing that US are hormonal regulated proteins and highly expressed into the uterus during pregnancy. The US group lacks apparently of anti protei nase activity and they plausibly diverge in biological function with respect to the other members of the superfamily. Also, the limited group of species which posses the US gene suggests the possibility of the evolution of the US gene within mammals for specific functions. The specific object ives that addressed in this dissertation are 1) To determine the presence of the US gene in species with non epitheliochorial placenta and also to examine the evolution of the US gene in mammals. 2) To establish whether the US produced by the sheep (OvUS) inh ibits cell proliferation by a) causing cell death such as apoptosis or necrosis b) altering cell cycle dynamics to cause cell cycle arrest. 3) To determine which cell cycle related genes are regulated by OvUS
51 CHAPTER 2 MOLECULA R PHYLOGENY OF UTERINE SERPINS A N D ITS RELATIONSHIP T O EVOLUTION OF PLACENT ATION Introduction Evolution of placentation in mammals has been dependent upon new uses of existing genes as well as appearance of new genes arising from gene duplication and selection for se quence divergence. Knox & Baker (2008) have shown that genes expressed preferentially by the placenta and decidua early in development tend to be ancient genes while the genes expressed preferentially by the mature placenta have been formed recently. Exa mples of such trophoblast genes are the interferon et al. 1999) and the primate chorionic gonadotropins (Maston & Ruvolo 2002). Another group of uterine genes formed by gene duplication are the uterine serpins (US) (al so called uterine milk proteins). These proteins are members of the ser ine p roteinase in hibitor (serpin) superfamily and have been identified as secretory products of the endometrial epithelium of the pregnant sheep, goat, cow and pig (Moffatt et al. 1987 Leslie et al. 1990, Malathy et al. 1990, Tekin et al. 2005 a ). Uterine serpins have been classified as either a separate clade of the serpin superfamily (Peltier et al. 2000 c ) or as a highly diverge group of the antitrypsin clade ( Irving et al. 2000). Recently, they have been designated as SERPINA14. The prototypical serpins are competitive inhibitors of serine and cysteine proteinases. The proteinase binds covalently to reactive center loop (RCL), which is l ocalized on the top of the serpin and contains a complementary sequence to the active site of the target proteinase. The scissile bond at the P1 sheet A (I rving et al. 2000, Hungtinton 2006, Whisstock & Bottomley, 2006). The inactivation of the proteinase is accompanied by an irreversible
52 conformational change of the serpin structure that increases its overall stability (Silverman et al. 2001, Hungtinton 20 06, Whisstock & Bottomley 2006). A f ew serpins such as the US have evolved biological functions that do not involve p roteinase inhibitory activity. Other examples are the intracellular serpin mammary serine protease inhibitor (maspin) (Sheng 2006), corti costeroid a nd thyroxine binding globulins (Pemberton et al. 1998), pigment epithelium derived factor (PEDF) (Tombran Tink & Barnstable 2003) and the heat shock protein 47 (Sauk et al. 2005). Inhibitory serpins are usually recognized by a consensus sequenc e in the hinge region which is localized within the RCL of the serpin (Irving et al. 2000) but the putative hinge region of US is not conserved with inhibitory serpins (Irving et al. 2000 Tekin et al. 2005 a ) The ovine (Ov) and bovine (Bo) US exhibit eit her no or very weak inhibitory activity towards a range of serine and aspartic proteinases (Ing & Roberts 1989, Liu & Hansen 1995, Mathialagan & Hansen 1996, Peltier et al. 2000 a ). All of the species shown to possess US genes are in the Ruminantia and Su idae orders of the Laurasiatheria superorder of eutherian mammals and all have epitheliochorial type of placentation characterized by limited invasiveness and maintenance of epithelial layers in endometrium and trophoblast. The epitheliochorial placenta h as evolved from a more primitive hemotropic placenta three separate times in evolution in a subgroup of Laurasiatheria consisting of cetartiodactyls, pigs, horses, and spiny anteaters, in a species of mole, and in lemurs (Vogel 2005). The evolutionary p ressure for development of an epitheliochorial placenta has been attributed to increase d efficiency of nutrient transport and gas exchange (Leiser et al. 1997), greater maternal control of nutrient distribution to the conceptus (Mess & Carter 2007) and an altered immunological relationship between mother and conceptus (Moffett & Loke 2006).
53 The objectives of this study were to identify novel US genes in other species within and outside of the Laurasiatheria superorder to determine whether the presence of the US gene is restricted to species with epitheliochorial placentation and to evaluate whether US gene has been subject to positive selection. Materials and Methods Data Base Queries to Identify Uterine Serpin G enes The nucleotide and protein sequences for OvUS were used as query sequences to perform a blastn search in the nucleotide collection (nr/nt) database of the National Center for Biotechnology Information ( NCBI ) website to identify known and novel uterine serpin genes. The Genbank accession numb er for the OvUS nucleotide sequence was NM_001009304.1 and the accession number for the protein sequence was NP_001009304.1. Subsequently, a genomic blast (blastn) of completed genomic database sequences in the NCBI was performed to identify sequences tha t have similarities with the OvUS query sequence. Species examined were human ( Homo sapiens ), rhesus macaque ( Macaca mulata ), chimpanzee ( Pan troglodytes ), dog ( Canis lupus familiaris ), cow ( Bos taurus ) mouse ( Mus musculus ) rat ( Rattus norvegicus ), opossu m ( Monodelphis domestica ) and duck billed platypus ( Ornithorhynchus anatinus ), as well as other vertebrates like chicken ( Gallus gallus ), zebra fish ( Danio rerio ) and puffer fish ( Tetraodon nigroviridis ). Additionally the unfinished w hole shotgun genomic sequences for t he horse ( Equus cabalus previously unfinished) and cat ( Felis catus ) were included in this search Reverse Transcription Polymerase Chain Reaction (RT PCR) Uterine endometrial tissue s were collected from pregnant mares at day s 21 and 59 o f gestation, a pregnant bitch at day 60 of gestation, a pregnant queen at day 60 of gestation as well as from the intercaruncular area from ovariectomized ewes treated with 100 mg/ml of progesterone for 60 days (Padua et al. 2005). D og liver tissue and o varies were also tested.
5 4 Total RNA was isolated after homogenization with the TRI reagent (Sigma Aldrich, St Louis, RNA purity and concentration were determined spectrophotometrically. The RT PCR was pe rformed using the SuperScript One Step RT PCR kit with Platinum Taq (Invitrogen, Carlsbad, CA). cDNA for OvUS, equine (Eq), Canine (Can) and Feline (Fe) US were amplified from total RNA using primer sets for sheep (forward ACA GAT GCT TTA CAG CCG GTC AGA; reverse TGA ACT TAA CAA CCA CCG GGA CCT), horse (forward GCT GCA GAA ATG TCC CAC AGG AAA; reverse AGA GGA AAT CCC TGT GCT TCA GGT) dog (forward ACC CAG TCT CGT CAT GGG AAG TTT; reverse TCA CGT CAT ACA TCG CCT GTG TGT) and cat (forward TAC GAG ATC CAC AAC GCG CAC TAA ; reverse AAG TCA GTC ATC TGG GCC TTC ACA ). To verify that PCR products were amplified from RNA only, the SuperScript reverse transcriptase /Platinum Taq mix was omitted from control reactions and an equivalent concentration of Taq DNA polymeras e (Invitrogen) was added cDNA synthesis and pre denaturation reactions were 1 cycle of 50 o C for 30 min and 94 o C for 2 min respectively. PCR amplification reactions were performed in 40 cycles of 94 o C for 15 sec for denaturation, 50 o C for 30 sec for ann ealing and 72 o C for 1 min extension. A final extension cycle was performed at 72 o C for 5 min Identification of cDNA for Equine U terine Serpin G ene The complete sequence of the EqUS gene was obtained by primer walking and the rapid amplification of cDNA he oligo dT
55 Table 2 1. Primers used in RT PCR procedure to obtain the full length coding sequence of the equine uterine serpin gene. Amplicons are designated based on the G enbank Accession Number of the sequences obtained from the RT PCR. The full length was obtained by overlapping seven sequences as shown in the diagram. Amplicon EU 810388 Fw GCTGCAATGTCCCACAGGAAA; Rv AGAGGAAATCCCTGTGCTTCAGGT EU 810389 Fw CTCTGGATTGCTGCAGAAATG; Rv CTACTCAGCTATGGGGTTGAA EU 810390 Fw CTCTGGATTGCTGCAGAAATG; Rv CTACTCAGCTATGGGGTTGAA EU 810391 Fw TCGACCTCCAAAGAATCAGAGCGT; Rv AGGACACGTTTCCAGTGTAAGGCA EU 810392 Fw AAACGTGTCCTTAGTCCTCGTGCT; Rv TTGAAGACTTGGCCCAC AAAGAGC EU 810393 Rv /Phos/TGTCCCTAAAGGAGA Fw TGTTCGAGGCTCTGTCAGTTGAGT; Rv TGTGCTTCAGGTGCCTCTGTCTAT Fw GCTGCAGAAATGTCCCACAGGAAA; Rv ACTCAACTGACAGAGCCTCGAACA EU 810394 Fw AAACGTGTCCTTAGTCCTCGTGCT
56 shown in T able 2 1. Amplicons were electrophoresed on 3% (w/v) agarose gel containing 1 g/ml ethidium bromide in Tris acetate EDTA buffer (40 mM Tris acetate, 2 mM EDTA pH 8.5), visualized on a ultraviolet transluminator and photographed with a Canon G 7 (Canon Inc, Japan) power shot digital camera using the Digi Doc IT TM imaging system (UVP, LLC, Upland, CA). Amplicons were purified from the corresponding agarose gel bands using the QIAquick DNA concentration was determined spectrophotometrically and purity was assessed by electrophoresis as previously described. Sequencing of A mplicons DNA amplicons were sequenced in both directions at the University of Florida DNA Sequencing Core Laborat ory using the ABI Prism 3130 genetic analyzer (Applied Biosciences, Foster City, CA). Briefly, ABI prism BigDye Terminators v.1.1 cycle sequencing reactions were assembled in a total volume of 20 l containing 500 ng DNA, 10 pmols primer, 4 l of BigDyer terminator, and 5% (v/v) dimethyl sulfoxide. Sequencing reactions were performed on an ABI GeneAmp 9700 thermal cycler (Applied Biosystems) with an initial denaturation at 95 o C for 10 min and followed by 45 cycles of denaturation at 95 o C for 30 s ec annea ling at 55 o C for 20 s ec and extension at 60 o C for 4 min. The excess of dye labeled terminators were removed by using MultoScr een 96 well filtration system (Millipore, Bedford, MA). The purified extension products were dried in SpeedVac (ThermoSavant, H olbrook, NY) and then suspended in Hi di formamide. Sequencing reactions were performed using POP 7 sieving matrix on 50 cm capillaries in the ABI Prism 3130 Genetic Analyzer and were analyzed by ABI Sequencing Analysis software v. 5.2 and KB Basecaller. A nalyses of the deduced amino acid sequence were conducted using the program Scan Prosite at the Ex PASy Molecular Biology Server ( http://ca.expasy.org ) The predicted signal
57 peptide cleavage sites were obtained from SignalP 3.0 ( Bendtsen et al. 2004) and predicted glycosylation sites were obtained from NetNGlyc 1.0 both from the Ex PASy Molecular Biology Server Detection of Equine Uterine S erpin by Western B lotting Ovariectomized pony mares were treated daily with pr ogesterone (150 mg/ml), estradiol benzoate (10 mg/ml) progesterone plus estradiol benzoate or vehicle as a control. After 28 days, uterine fluids were collected by flushing the uterus with 0.9% (w/v) NaCl (Hansen et al. 1985). Samples were concentrated using C entricon plus 20 devices (Millipore Corporatio n, Billerica, MA). Aliquots of 20 l of concentrated mare samples were separated under reducing conditions using one dimensional discontinuous sodium dodecyl sulfate polyacrylamide gel electrophoresis a t 4 15% (w/v) polyacrylamide, Tris HCl gels (Bio Rad, Hercules, CA). Aliquants of 0.5 g of protein from uterine fluid of an ovariectomized ewe treated with 100 mg/ml of progesterone for 60 days and from the control serpin ovalbumin (OVA) were used as pos itive and negative control, respectively. Western blotting was performed as described elsewhere (Padua et al. 2005) with minor modifications. Briefly, the monoclonal anti OvUS (HL 218; Leslie et al. 1990) and IgG1 (MOPC21) (Sigma Aldrich) (1:32,000 dilut ion) were used as a primary and control antibodies, respectively. The second antibody used was ECL TM anti mouse IgG peroxidase linked; 1:25,000 dilution (Amersham GE Healthcare Bio Sciences Corp, Piscataway, NJ) and the ECL plus western blotting detection kit (Amersham) was used as a detection system Amino Acid S equence A lignments and A nalysis of P hylogenetic T ree ClustalW (Chenna et al. 2003) was used to obtain the amino acid alignments which were then prepare d for publication using the Boxshade multiple alignments designer program version 3.21 ( http://www.ch.embnet.org ).
58 The phylogenetic tree that includes corticosteroid binding globulin (CBG) was constructed using the Neighbor Joining (NJ) method (Saitou & Nei 1987) of the Molecular Evolutionary Genetics Analysis (MEGA4) software, version 4.0 ( Tamura et al. 2007 ). The NJ tree was based on the distance calculation under the Jones Taylor Thornton (JTT) matrix substituti on model (Jones et al. 1992) after removing position containing gaps (complete deletion option). The NJ method (Saitou & Nei 1987) was also used to construct the phylogenetic tree for d JTT distance matrix (Jones et al. 1992). The plotted tree was obtained by using the drawgram software of PHYLIP (Phylogeny Inference Package) version 3.5c (Felsenstein 1989, 1993). The gamma shape parameter and amino acid frequencies were estimated fro m the data using the Tree Puzzle software by maximum likelihood analysis for amino acids (Strimmer & von Haeseler 1996, Schmidt et al. 2002) and the reliability of the trees was estimated by the bootstrap test with 1,000 repetitions (Felsenstein 1985). An alysis of Ratio of Non S ynonymous and S ynonymous S ubstitutions The ratio of non was used to determine whether the pressure of selection induces purifying or positive selection at specific areas of the sequence ( Yang & Nielsen 2002 ) or lower than 1 are an indication of positive (Darwinian) or purifying selection, respectively. The codon based substitution model in the CODEML program of the phyloge netic analysis by maximum likelihood (PAML, version 4.0b) software package (Yang 2007) was used to identify the effect of selective pressures on the uterine serpin genes. Aligned sequences were tested using different models of variable dN/dS ratios among sites which includes M0 (null model), M1 (nearly neutral, where
59 et al. 2000, Bielawski & Yang 2003, Wong et al. 2004). The likelihood ratio test (LRT) was obtained b y twice the difference in log likelihood (2 L) from two different models and the significance of the test was estimated by using the 2 distribution with degrees of freedom calculated from the estimated parameters (Anisimova et al. 2001, Yang & Nielsen 200 2). The models M1 and M7 were tested against models M2 and M8, respectively. Additionally, the Selecton server (Server for the identification of site specific positive and purifying selection, version 2.4) (Stern et al. 2007) was used to identify an empirical Bayesian method Aligned coding sequences in FASTA format for uterine serpins were tested with the M8 model, s s cores were also tested by using LRT test which compares the positive selection model (M8) and s set to 1); allowing only for purifying and neutral selection, respectively. Results Identification of Coding S equences of K nown and New Uterine S erpins U sing B lastn Uterine serpin sequences have been previously described for goat (Capra hircus) cow ( Bos t auru s ) and pig ( Sus scrofa ) Results from the blastn search in the nucleotide collection database of the NCBI identified an additional uterine serpin coding sequence that has been submitted to Genbank for water buffalo ( Bubalus bubalis Genbank: DQ 661648.1 ) This is a species with epitheliochorial placenta that is closely related to the cow. The blast search also identifi ed a gene in the dog ( Canis lupus familiaris Genbank: XM 850115.1) with similarity to the query sequence, OvUS. The dog is a member of Laurasiatheria but, unlike other species with uterine serpin genes, the dog has an endotheriochorial placenta. Current ly, XM_850115.1 is annotated as one of the two CBG genes in the dog. Phylogenetic analysis shows, however, that while the
60 other predicted dog CBG (Genbank: XP 547960.2 ) is clustered together with CBG proteins present in other species such as human, chimp anzee, squirrel and rhesus monkey, pig, sheep, mouse and hamster, XP 855208.1 clusters with the US (Figure 2 1). A genomic blastn search of complete genomic sequences identified uterine serpin genes for the cow and dog. No US genes were identified in the order Anthropoidea, represented by human, rhesus monkey and chimpanzee, or in Rodentia, represented by the mouse and rat. Likewise, the US gene was no identified either in the opossum, a species with a primitive epitheliochorial placentation within the M arsupialia order, in the duck billed platypus, an egg laying mammal that belongs to the Monotremata order, or in any of the other vertebrates with complete genome sequences (chicken, puffer and zebra fish). A genomic blast was also performed using OvUS as query for the w hole shotgun genomic sequence of the horse, which was incomplete at the time of analysis. There were four stretches of nucleotides of 664, 232, 72 and 86 nucleotides in length ( Genbank: gbAAWR02019930 ) that matched with the coding sequence of OvUS (68,75, 86 and 79% identity respectively). These sequences were all localized on chromosome 24 SERPINA11. A genomic blast search of the w hole shotgun genomic sequence of the cat using OvUS as the query identifie d one stretch of 198 nucleotides in length ( Genbank: gbA CBE01215383.1) that matched with the coding sequence of OvUS with 77% of identity. However, using CanUS as the query sequence identified four stretches of nucleotides of 525, 273, 146 and 199 nucleot ides in length (same accession number ) that matched the coding sequence of CanUS with 82 82, 80 and 89 % identity respectively.
61 Figure 2 1. Identification of an incorrectly annotated dog corticosteroid binding globulin (CBG) as an ute rine serpin (US). The sequence incorrectly annotated as CBG [Genbank : XP_855208.1 ] clusters with the US while another dog CBG [Genbank: XP 547960.2 ] clusters with CBG in other species The Neighbor Joining tree was constructed using the Molecular Evoluti onary Genetics Analysis (MEGA4) software The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown next to the branches. The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the Jones Taylor Thornton (JTT) matrix based method and are in the units of the number of amino acid substitutions per site. A ll positions containing gaps and missing data were eliminated from the dataset (Complete deletion option).
62 Uterine S erpin Gene Organization in the Bovine and C anine The BoUS gene is localized in chromosome 21 for the cow and is organized in 6 exons, with two small untranslated regions (UTR) contained in exons 1 and 6 respectively. The CanUS is localized in chromosome 8 and also contains 6 exons but with no UTR present. The BoUS gene clusters with SERPINA1, SERPINA11 and SERPINA12. The CanUS gene cluster s with SERPINA1, SERPINA9, SERPINA11, and SERPINA12. The re are variations in the length of exons and introns between BoUS and CanUS gene as indicated in T able 2 2 However, the size of exon 3 is identical in both species. Characteristics of Can US and E xpression in T issues The gene for CanUS (Genbank: XM_850115 ) is 1 416 bp in length that encod es a polypeptide of 471 amino acids wi th a theoretical isoelectric point of 9.43 and predicted molecular mass of 54 25 kDa. The CanUS protein sequence lacks a sign al peptide suggesting the intracellular localization of the protein. Using XM_850115 primers were designed for RT PCR. Amplicons were obtained from endometrium of a bitch at 60 days of gestation, but there was no expression in the liver ( Figure 2 2 ). In addition, transcripts for CanUS were identified in the ovary by RT PCR (data not shown). Uterine E xpression and Amino Acid S equence of EqUS Using primers designed from the EqUS sequences identified in the blast search, expression of EqUS in endometruim from a pregnant mare at Day 21 (data not shown) and Day 59 of gestation was identified (Figure 2 3A). T he complete sequence of EqUS was obtained by performing RT gestation. In this ma nner, seven overlapping sequences were obtained (Genbank: EU810388, EU810389, EU810390, EU810391, EU810392, EU810393, EU810394 ) with a total length of The
63 Table 2 2. Exon and intron sizes for the uterine serpin gene of the cow and dog. Cow exon1 (UTR) exon2 intron2 exon3 intron3 exon4 intron4 exon5 exon6 (UTR) Total bp 16 637 3204 268 1808 148 2205 288 65 8639 Dog exon1 intron1 exon2 intron2 exon3 intron3 exo n4 intron4 exon5 Total bp 210 1863 148 1580 268 876 666 5034 124 10768
64 Figure 2 2. Representative electrophoretogram of amplicons obtained by reverse transcriptase polymerase chain reaction (RT PCR) of RNA fron cani ne tissue and c an ine uterine serpin primers RNA was isolated from dog liver or endometruim from a pregnant bitch at Day 60 of gestation Control reactions excluded reverse transcriptase (w/o RT ) Arrows on the left represent size of standards while the arrow on the right show s the expected amplicon s ize
65 Figure 2 3. Expression and secretion of e q uine uterine serpin Panel A represents an electrophoretogram of amplicons obtained by reverse transcriptase polymearse chain reaction (RT PCR ) of either RNA from endometruim of an ovariectomized ewe treated with progesterone for 60 days (P4 treated ewe) or RNA from endometruim of a pregnant mare at Day 59 of gestation. Control reactions excluded reverse transcriptase (w/o RT). Arrows show a s ize of the standard and expected size of the amplicons for the sheep and mare respectively. Panel B represents results of a western blotting experiment using antibody to OvUS and either uterine fluid from an ovariectomized ewe treated with progesterone fo r 60 days, uterine flushings from ovariectomized mares treated for xx days with either vehicle progesterone (P 4 ), estr adiol benzoate (E), or the combination (P 4+E), or endometrial tissue from a pregnant mare at 59 days of gestation. As negative control, the serpin ovalbumin (OVA) was also subjected to western blotting.
66 sequence was submitted to NCBI GenBank as TPA (Third Party Annotation, GenBank accession number: BK 006618 ). The complete coding sequence of the EqUS was 1 263 bp which is 100% identical t o genomic sequence. T he deduced amino acid sequence is shown in Figure 2 4 The gene encod es a polypeptide of 421 amino acids wi th a theoretical isoelectric point of 9.49 and predicted molecular mass of 48 79 kDa. Analysis of the EqUS protein sequence sh owed a predicted signal peptide, most likely between positions (Cys 25 ) and (Glu 26 ) and two potential glycosylated sites at (Asn 212 ) and (Asn 246 ) Endometrial S ecretion of EqUS Uterine flushings from ovariectomized mares treated with steroids were examined by western blotting using an antibody to OvUS to determine if EqUS was secreted into uterine lumen and, if so, whether as for other uterine serpins secretion was under control of progesterone. The western blot is shown in Figure 2 3B Immunoreactive pro tein bands were obtained from uterine flushes from ovariectomized mares treated with estr adiol benzoate progesterone or progesterone and estrogen. There was an absence of immunoreactive protein in flushes of ovariectomized mares treated with vehicle. Th e molecular weight of the immunoreactive EqUS was similar to OvUS identified by western blotting of uterine fluid from a progesterone treated ewe (55,000 57,000). The immunoreactive EqUS is larger than the predicted molecular mass of 48 79 kDa suggest ing the protein is glycosylat ed Multiple immunoreactive bands were also seen for supernatant from lysed endometrial tissue of a mare at Day 59 of gestation. The variety of lower molecular weight bands probably represent proteolytic products as has been des cribed for OvUS (Moffatt et al. 1987, Leslie et al. 1990, Peltier et al 2000 a Leslie & Hansen 1991). No immunoreactive bands were identified for the control serpin OVA (Figure 3B) or when IgG replaced the primary antibody (data not shown).
67 Lack of E xp ression of the Uterine S erpin G ene in the P regnant C at RT PCR experiments with primers designed for a putative FeUS sequence identify by blastn against CanUS (525 bp in length) failed to yield an amplicon from cDNA obtained from pregnant endometruim of a pregnant queen at Day 60 of gestation (data not shown). Moreover, stop codons were identified after translation of the deduced amino acid sequences for at least two of the four stretches of nucleotides identified by blastn. Depending upon the frame used for codon identification, the 525 bp sequence had 1 10 stop codons, the 272 bp sequence had 1 6 stop codons, the 199 bp sequence had from 0 3 stop codons, and the 146 bp sequence had from 0 4 stop codons. Simila r l y, the 198 bp sequence identified in cat by genomic blast search using OvUS had 1 4 stop codons. Amino Acid Sequence C onservation The ClustalW amino acid sequence alignments of uterine serpins are shown in Figure 2 4 There is a 9 amino acid insert present in OvUS and CapUS (KINLKHLLP) that is no t present in uterine serpins from other species. Likewise, the water buffalo US (BuUS) contains a 13 amino acid insert (MNAKEVPVVVKVP) that is highly conserved with 39 amino acid insert for BoUS. There is also a 3 amino acid insert in the same relative p osition for the PoUS 1, PoUS 2, CanUS and EqUS. This insert is conserved between the PoUS 1 and 2, but the insert for CanUS and EqUS is not conserved with any other species. The CanUS sequence is unique among uterine serpins in that there are two start codons upstream from the start codons for other uterine serpins (Figure 2 4) so that the CanUS polypeptide could be as much as 48 amino acid longer than other serpin sequences and not contain a signal peptide. There is a KEVPVVVK motif located near the putative P1 the BoUS that has been postulated to be a pepstatin like domain (Mathialagan & Hansen 1996).
68 Figure 2 4. Amino acid sequence alignment of the uterine serpins using the ClustalW algorithm. S ites of conservation are shown in shade d columns using the Boxshade software. Black shade d columns show identical amino acids and grey shade d columns represent similar amino acids. Distinctive amino acid motifs limited to a subset of uterine serpins ar e marked with red rectangles. Start codons are shown by red arrows.
69 Figure 2 4 Continued
70 This domain is completely conserved in OvUS, CapUS, PoUS 1 and PoUS 2, is nearly completely conserved in BuUS (only one amino acid substitutio n), but is not conserved in CanUS and EqUS. Identification of the Putative Hinge R egion and P1 ite of the RC L ClustalW alignment was used to align all US with inhibitory serpins antichymotrypsin (Genbank: NP_001076 and NP_998952.1), human plasminogen activator inhibitor (Genbank: NP_000593), plasmin inhibitor (Genbank: NP_777095), human leukocyte e lastase inhibitor (Genbank: P30740), human placental thrombin inhibitor (Genbank: P35237) and human squamous cell carcinoma antigen 1 (Genbank: NP_536722) ( Figure 2 5). Note that while most of the US have a valine at the same position. The exception is for EqUS and CanUS which have an isoleucine and arginine at this position, respectively. The hinge region of the RCL ( underlined in Figure 2 5 ) is highly conserved among inhibitory serp ins. The consensus sequence of RCL involves an arginine at P17, arginine, lysine or glutamic acid at P16, glycine at P15, serine or threonine at P14 and for the hinge region an alanine, glycine or serine at P12 P9 (Irving et al 2000). For the US, the co nserved arginine at P17 is present in almost all US with the exception of CanUS which has a lysine at this position For the P16, US posses s an aspartic acid or histidine with the exception of the CanUS (glutamic acid) and EqUS (lysine) Moreover, the co nserved glycine at P15 is present in only PoUS 1, PoUS 2, EqUS and CanUS. Uterine serpins do not contain threonine or serine at P14 nor alanine, glycine or serine at positions P12 to P10 (with the exception of the PoUS 1 and 2 which have an alanine at P1 0). However, US have a conserved serine or alanine at position P9. Taken
71 Figure 2 5. Identification of the P1 Represented alignment of a portion of the amino acid sequences of some inhibitory members of the serpin superfamily and the uterine serpin (blue square). Inferred amino acid sequences of uterine serpins were aligned with inferred amino acid sequences of selected inhibitory serpins u sing the ClustalW software. Shown is a portion of the polypeptides near the P1 ites of conservation are shown in shade d column s using the Boxshade software. Black shade d columns show identical amino acids and grey sha de d columns represent similar amino acids.
72 together, resul ts indicate that the RCL and the hinge region required for inhibitory activity has not been conserved in uterine serpins. Phylogenetic Analysis of Uterine S erpins The phylogenetic tree constructed using the NJ method is presented in Figure 2 6. The sheep and goat proteins as wells as the water buffalo and cow proteins were clustered together, and the horse and dog were clustered together in a sister group and the pig proteins were closer to the horse and dog than to the ruminants. Positive S election of t he U terine S erpin G ene Aligned sequences of uterine serpins were tested for evidence of positive selection using different models of variable dN/dS ratios among sites (Table 2 3). The LRT for positive selection was significant when comparing M2 and M8 ag ainst their null model (Table 2 4). The Bayes empirical Bayes method used in PAML to calculate the posterior probabilities at each codon site (Yang et al. 2005) indicate A121 G343 and K380 as positive selected sites for M2 and I35 A121, G343, K365 V378 F373 and K380 as positive selected sites for M8. These results partially agreed with those obtained with the Selecton program where six amino acid residues ( Figure 2 7 ) were identified as sites with potential positive selection ( I35, A121, D156, G343, K 365 and K380 ) and thirty nine amino acid residues were likely to be under purifying selection ( M1, H3, M6, S22, I23, V38, A58, F62, E68, K72, I75, F76, A86, D167, I178, V182, T195, T201, K226, E227, F229, T235, V237, V239, M241, K244, M248, S251, E254, N26 8, D277, K324, E349, E357, K388, F393, E398, V411 and N413 ). To identify whether codons undergoing positive selection were associated with regions antitrypsin to sheets of the protein (Figure 2 8). A total of six amino acid residues were identified as being subject to positive selection by at least two of the three
73 Figure 2 6. Phylogenetic tree of the uterine serpin protei ns with the ovine uterine serpin (OvUS) as out group. The Jones Taylor Thornton matrix was used for distance calculation with gamma corrected distances and tree was inferred by the Neighbor joining method The t ree reliability was tested with 1000 bootst rap replicates (not shown).
74 Table 2 3. Ratios estimates and maximum log likelihood of models for positive selection within the protein coding sequence of uterine serpins a The numbers of free parameters are represented by (p) assigned to each class of Model dN/dS Parameter estimate(s) a L M0 (one ratio) 0.8815 = 0.8815 6270.25 M1 (Nearly Neutral) 0.7506 p 0 = 0.29151, p 1 0 1 = 1 6233.37 M2 (Positive selection) 1.0717 p 0 = 0.25346, p 1 = 0.61790, p 2 0 1 2 = 3.20732 6219.15 M7 (Beta) 0.7799 p= 0.02945 q= 0.00709 6 235.54 1.0860 p= 0.49271, q= 0.19375, p 0 = 0.83214, p 1 s = 2.91144 6219.83
75 Table 2 4. Test of significance for models for positive selection w ithin the protein coding sequence of uterine serpins a Models 2 L 2 value df P values Positive selected sites (BEB) b M1 versus M2 2( 6219.15+6233.37) 28.44 2 <0.001 A121**, G343*, K380* M7 versus M8 2( 6219.83+6235.54) 31.42 2 <0.001 I35*, A121**, G343 *, K365* V378*, F373*, K380** a The l ikelihood ratio test statistics were calculated as two times the difference in the log likelihood between models (2 L). S ignificance wa s tested by using the 2 test with degrees of freedom calculated from the paramet er estimates. b Sites of positive selection were estimated by calculating the posterior probability that each codon was from the site class of positive selection under models M2 and M8 by the Bayes empirical Bayes method (* P >0.95; ** P >0.99).
76 Figure 2 7. Selecton output generated for the uterine serpin group of proteins. Shown is the sequence for PoUS 1. Color intensity indicates likelihood of purifying (dark purple) and positive selection (dark yellow) at each amino aci d residues. Six amino acid residues were estimated to undergo positive selection (I35, A121, D156, G343, K365 and K380) while thirty nine were identified to be under purifying selection. Amino acid residues identified as under positive selection by the B ayes empirical Bayes method for models M2 and M8 of the PAML program are shown by an asterisk (*) and diamond ( ), respectively. The putative hinge region of the reactive center loop is underlined and the putative scissile bond at P1 by an arrow.
77 Figure 2 8. Amino acid sequence alignment of ovine uterine serpin (NP_001009304.1) and antitrypsin (NP_000286.3) using the ClustalW algorithm Helices are shown as green cylinders and pleated sheets are shown as yellow arrows. The location and designation of these regions are based on Silverman et al. (2004). Amino acid residues identified as being subject to positive selection by at least two of three statistical methods are indicated by arrowheads.
78 s tatistical methods. Of the six positive selected sites, two are in colied regions, one is in the putative Helix D, one is in the putative Helix E, one is in the reactive center loop and one is the 2 8 ). Discussion Serpin genes are of ancient origin and are found in bacteria, fungus, nematodes, archaea and virus (Irving et al. 2000, 2002 b Steenbakkers et al. 2008). The fact that uterine serpins (SERPINA14) are apparently limited to species in the La urasiatheria supero rder means that they either aro se recently, after Laurasiatheria diverged from Euarchontoglires (~87 Mya) (Murphy et al 2007), or they have been retained during evolution in this superorder only. Their uterine expression and the loss o f amino acid motifs required for proteinase inhibitory activity in the P1 for maintenance of pregnancy. Based on experiments with sheep and pigs, uterine ser pins may function to inhibit cell proliferation (Padua & Hansen 2008) or interact with other uterine proteins (Baumbach et al 1986, Hansen & Newton 1988, McFarlane et al 1999). The uterine serpins are thus an example of a new gene arising from gene dupl ication and selection for sequence divergence (Louis 2007) that plays a particular role in pregnancy. While it was not possible to examine complete genomic sequences in each order of mammals, the available evidence points strongly to restriction of t he uterine serpin to Laurasiatheria only. The distribution of known uterine serpins as well as orders where blast search of complete genomic sequences fail to identify a uterine serpin is shown in Figure 2 9. All the identified uterine serpins are in Rum inantia, Suidae, Perissodactyla or Carnivora orders of Laurasiatheria. No gene that significantly matched OvUS was found in within the orders
79 Figure 2 9. Phylogenetic tree of placentation in mammals (adapted from Vogel 2005) to il lustrate the existence of uterine serpin genes relative to type of placentation. Shown are the 4 m ajor s uperorders of eutherian mammals (Laurasiatheria, Euarchontoglires, Xenarthra and Afrotheria) with the order Marsupiala as the out group of the tree. B lue branches represent orders with epitheliochorial placenta and orange branches represent orders with either endotheliochorial or hemochorial type of placentation. The yellow branches represent unresolved situations in the phylogeny. The symbol repres ents those orders within the Laurasiatheria super or der where the uterine serpin gene has been identified. The symbol represents those orders where the uterine serpin gene was not identified after blast search of complete genomic sequences
80 Anthropoidea and Rodentia of the superorder Euarchontoglires, Marsupialia or in non mammalian species examined. Given the fact that all uterine serpins identified before this study were species exhibiting epitheliochorial placentation (Figure 2 9), it was hypothesi zed that uterine serpin genes serve sole role uniquely required for species with this type of placentation. While this hypothesis may be correct it is not true that the gene is only present in species with epitheliochorial placentation. Of the two new ut erine serpins identified in the present study, one is for species with epitheliochorial placentation (horse), but one is for species with endotheliochorial placentation (dog). Phylogenetic analysis showed that the genes in the dog and horse are closely re lated to each other. In fact, CanUS and EqUS are closer to each other than to PoUS 1 and PoUS 2. For example, the KEVPVVVK in other uterine serpins including pig is lost in both horse and dog. In some mammalian phylogenies, the horse is more closely rel ated to the pig than carnivores (Vogel 2005) whereas in others, the horse is closer to carnivores than to pigs (Nishihara et al. 2006) or the horse is equidistant from carnivores and pigs (Murphy et al 2007). Even though the dog has a uterine serpin gene there was no evidence for functional uterine serpin gene in the closely related cat. Results must be tentative because a complete genomic sequence of cat is not yet available. However, the fact that nucleotide sequences identified in cat as being simil ar to uterine serpins contained stop codons suggests that a functional uterine serpin gene has been lost in this species Additionally, the dog uterine serpin has undergone significant change from other serpins in that it has acquired two putative start codons upstream from the start codons in other species. If in fact this start codon is correct, the CanUS gene encodes for an intracellular protein without a signal peptide while all other species with uterine serpin encode for secretory proteins (Moffatt et al. 1987, Leslie et al. 1990, Malathy et al. 1990, Tekin et al.
81 2005 a ). The newly discovered EqUS was also shown to be secreted into the uterine lumen as revealed by western blotting of uterine flushings. Taken together, it is possible that the uterin e serpin gene in species with epitheliochorial placentation has been required as a secretory protein while the gene in species with endotheliochorial placentation has either being changed to become an intracellular protein (dog) or a pseudogene (cat). If this idea is correct, uterine serpins play some important role in the uterine lumen of species with epitheliochorial placentation that is not required for species with either endotheliochorial placentae. This hypothesis is also supported by the absence of a homologous uterine serpin gene in human and rodents, species with hemochorial placentation. The function of uterine serpins is probably not to inhibit serine or cysteine proteinases. Ovine and BoUS exhibit no inhibitory activity towards a range o f proteinases (Ing & Roberts 1989, Liu & Hansen 1995, Mathialagan & Hansen 1996, Peltier et al 2000 a ). In addition, there is evidence, as shown here, that the P1 from inhibitory serpins. The RCL co ntains a complementary sequence to the active site of the target proteinase and the hinge region of RCL is highly conserved among inhibitory serpins (Irving et al 2000). In uterine serpins, however, the amino acids of the consensus sequence is absent at P15 (except for PoUS 1, PoUS 2, EqUS and CanUS), P14 and P12 to P10 (except for PoUS 1 and 2 which have an alanine at P10). Furthermore, an amino acid in the RCL and at the 1 site is not con served with inhibitory serpins. Among the serpin superfamily, the RCL is important for selectivity of the serpin for the target proteinase and has undergone an accelerated rate of evolution (Brown 1987, Hill & Hastie 1987, Goodwin et al. 1996, Zang & Maizels 2001, Barbour et al. 2002).
82 The identification of the P1 V instead of P V) that was identified in the original paper by Ing & Robets (1989) and used as the basis for assigning its location in other papers (Mathialagan & Hansen 1996, Peltier et al 2000 c Tekin et al. 2005 a ). The accessibility to more complete sequences from inhibitory serpins whose P1 s in addition to improvement in alignment programs results in a more reliable estimate of the P1 uterine serpins. Unique functions of uterine serpins in species with epitheliochorial placentae must be linked to the unique characteristics of placentation in this species. Unlike other type of placentae in eutherian mammals, where maternal blood either spills directly onto trophoblast (hemochorial placenta) or where trophoblast invasion and erosion of endometrial tissue leaves the trophoblast in contact with maternal endothelium, the epitheliochorial placenta is characterized by limited invasiveness of trophoblast and an intact endometrial epithelium that sometimes forms a synctium with trophoblast cells. The efficiency of placental transport is improved in species with epitheliochorial placentae because of countercurrent exchange of nutrients and gases between fetal and maternal blood vessels (Leiser et al 1997). It has also been speculated that the presence of an intact maternal endometrial epithelium may reduce the recognition of trophoblast antigens by maternal immune system (Moffet & Loke 2006) although available evidence indicates that the maternal immune response to the trophoblast is similar in species with epitheliochorial placentatio n to that of other species (Oliveira & Hansen 2008). Ovine uterine serpin has been shown to inhibit proliferation of lymphocytes and a variety of cancer cells (Padua & Hansen 2008) and it is possible that one or more uterine serpins function to inhibit cel l proliferation during pregnancy. Inhibition of lymphocyte proliferation
83 has been interpreted as signifying role for uterine serpins in prevention of rejection of conceptus tissue by the maternal immune system (Hansen 1998). In addition, the limited inva siveness of trophoblast tissue in species with epitheliochorial placentation could be achieved, at least in part, by inhibition of trophoblast proliferation. Through inhibition of the cell cycle (Padua & Hansen 2008), it may be also that uterine serpins p articipate in the process of trophoblast binucleate cell formation, wh ich occurs in sheep goat, cattle and horse (Hoffman & Wooding 1993). Another characteristic of uterine serpins is the prop ensity to bind other proteins. Ovine uterine serpin has been f ound to bind activin (McFarlane et al. 1999 ), IgA and IgM (Hansen & Newton 1988), and the so called pregnancy associated glycoproteins (Mathialagan & Hansen 1996) that are inactive members of the aspartic proteinase family (Xie et al. 1991, Green et al. 19 98). The pig uterine serpins bind to an iron binding protein secreted by the uterus called uteroferrin and stabilize the iron bound to uteroferrin (Baumbach et al 1989). In some cases, this binding can be ascribed to the highly basic isoelectric point f or uterine serpins. Binding to IgA and pepsin can be reduced by high salt concentrations (Hansen & Newton 1988, Peltier et al 2000 a ). However, a physiological role for protein binding remains a possibility. Binding to uteroferrin may enhance iron trans port to the fetus since this protein delivers its iron to fetal fluids (Buhi et al 1982). The function of pregnancy associated glycoproteins is not known but the conservation of the KEVPVVVK motif located near the P1 like domain (Mathialagan & Hansen 1996) may mean that binding to pregnancy associated glycoproteins is an important function. Progesterone is the hormone that induces the synthesis and secretion of uterine serpins into the uterine lumen of sheep, pig, c ow and goat (Ing et al 1989, Leslie et al 1990, Malathy et al 1990, Tekin et al 2005 a ). It is likely that the equine protein is also stimulated by progesterone.
84 In addition, in the cow (Khatib et al 2007) as well as in dog, the uterine serpin gene i s expressed in the ovary. The presence of uterine serpin in other reproductive tissues distinct to the uterus raises the possibility that uterine serpins may have different functions according to the cell type and stage of differentiation. It is possible that uterine serpin functions in the ovary to regulate follicular growth by binding to activin. Activin improves follicular growth and granulosa cell proliferation (Knight & Glister 2006). OvUS binds to this member of the TGF (McFarlane et al. 1999). In summary, t he uterine serpin gene is present only in a restricted group of species within the Laurasiatheria superorder of eutherian mammals and likely evolved under positive selection which suggests diversifying functionality of these proteins from the proteinase inhibitory activity of most members of the serpin superfamily. This evidence also suggests the uterine serpins gene is an example of gene duplication followed by selection. The finding that the uterine serpin has been retained as a se cretory protein in species with epitheliochorial placentation within the Laurasiatheria superorder also suggests that the protein has an important role during pregnancy in these species while in species with endotheliochorial placenta the gene has undergon e changes suggesting that it is playing either a new role in those species (dog) or no required (cat) during pregnancy.
85 CHAPTER 3 COMPARISON OF THE NA TIVE AND RECOMBINANT FORMS OF OVINE UTERI NE SERPIN FOR INHIBITIO N OF CELL PROLIFERAT ION Introduction The uterine serpins (US) are a group of glycosylated proteins secreted into the uterus of the sheep, cow, goat and pig during mid and late pregnancy (Moffatt et al. 1987 Leslie et al. 1990 Baumbach et al. 1986 Tekin et al. 2005 a ). These proteins are membe rs of the superfamily of serine proteinase inhibitors (serpins) ( Ing & Robets 1989, Mathialagan & Hansen 1996). Most of the members of the serpin superfamily are inhibitors of serine or cysteine proteinases, which are inactivated by an irreversible suicid e substrate like mechanism (Silverman et al. 2001). However, there are some members of this superfamily that have roles distinct from the inhibition of proteinases. Some examples of non inhibitory serpins include the hormone transport proteins corticoste roid and thyroxine binding globulin (Pemberton et al. 1988), the chaperone heat shock protein 47 (Nagata 1998) and angiotensinogen which is involved in the regulation of blood pressure (Morgan et al. 1996). Another serpin that seems to have a divergent f unction is ovine uterine serpin (OvUS). This US has been linked to the protection of the allogeneic conceptus through inhibiti on of immune cell proliferation during pregnancy (Hansen 1998). Several experiments have demonstrated that OvUS inhibited lympho cyte proliferation induced by mitogens such as c on canavalin A (Con A), the antigen Candida albicans phytohemagglutinin and in the mixed lymphocyte reactions (Segerson et al. 1984 Hansen et al. 1987 b Stephenson et al. 1989 b Skopets & Hansen 1993 Skopet s et al. 1995). Ovine US also inhibits natural killer (NK) cell activity ( Liu & Hansen 1993). Moreover, OvUS inhibited the lytic activity of NK like cells in sheep peripheral blood lymphocytes and e ndometrial epithelium against D 17 cells infected with bo vine herpes virus 1 (Tekin & Hansen 2002).
86 However, OvUS did not cause any inhibitory activity against lymphocytes activated by the T and B cell mitogen PWM ( Skopets & Hansen 1993). Ovine US had no effect on the reduction of skin fold thickness caused by Mycobacterium tuberculosis in sheep (Skopets et al. 1995) and T + cells induced by Con A ( Peltier et al. 2000 b ), suggesting the apparent selectivity of the protein to inhibit some immune cell types. Little is k nown about the antiproliferative actions of OvUS in other non immune cell types. The objectives pursued on this study are to test whether OvUS inhibited proliferation of non immune cells, specifically tumor cells, and to determine whether OvUS inhibits ce ll proliferation by apoptosis. Finally, the antiproliferative potency of a recombinant form of OvUS (rOvUS) was compared with the native form of the protein. Materials and Methods Materials The Essential Medium (MEM), ed Eagle Medium Nutrient Mixture F 12 H am (DMEM Medium (DMEM), streptomycin were purchased fr om Sigma Aldrich (St Louis, MO) The fetal bovine serum was from Inte rgen (Purchase, NY ) and the heat inactivated horse serum was from Hyclone (Logan, UT). The D 17 ( canine primary osteo genic sarcoma), PC 3 (human prostate cancer) and P388D1 (mouse lymphoma) cell lines were purchased from ATCC (Rockville, MD). The c entric on ultrafiltration devices were from Amicon (Beverly, MA ) or Millipore ( Bedford, MA ) Carboxymethyl Sepharose, Sephacryl S 200 and His Trap columns were obtained from Amersham Biosciences (Piscataway, NJ) and [ 3 H]thymidine (6.7 Ci/mmol) was from ICN (Irvi ne, CA) The RQ1 RNase free DNase was from Promega (Madison, WI), in situ cell death detection kit [terminal deoxynucleotidyl transferasemediated dUTP nick end labeling (TUNEL)] was from Roche (Indianapolis, IN). The
87 ProLong Antifade kit was purchased fr om Molecular Probes (Eugene, OR), RNase A was from Qiagen (Valencia, CA ) and the Precast Tris HCl gradient gels were obtained from Bio Rad ( Richmond, CA) Other reagents were from either Sigma Aldrich or Fisher. Collection of Uterine Fluid and Purificatio n of N ative OvUS Ewes of Rambouillet genotype (n=4) were made unilaterally pregnant as described elsewhere (Bazer et al. 1979) by ligating one uterine horn. Ewes were slaughtered at day 140 of pregnancy by captive bolt stunning and exsanguination. Crude uterine fluid was collected from the ligated uterine horn after slaughter by aspiration. Uterine fluid was clarified by centrifugation at 3600 x g and stored at 20 o C until purification of native ovine uterine serpin (nOvUS) by a combination of cation ex change chromatography with carboxymethyl Sepharose and gel filtration chromatography with Sephacryl S 200 as previously described ( Liu & Hansen 1993) The purity of the protein was assessed by sodium dodecyl sulfate, polyacrylamide gel electrophoresis (SD S PAGE) using 4 15% polyacrylamide Precast Tris HCl gradient gels and staining by Coomassie Blue G 250 After purification, nOvUS was buffer exchanged with DPBS and concentra ted using Centricon PL 20 ultrafiltration devices. Protein concentration was det ermined using the Bradford assay with bovine serum albumin as standard (Bradford 1976). [ 3 H ]thymidine Incorporation by D17 and PC 3 C ells The D17 cells were cultured continuously in complete medium [MEM supplemented with 10% (v/v) heat inactivated fetal bo vine serum, 200 U/ml penicillin and 2 mg/ml streptomycin]. The PC 3 cells were grown in DMEM F12 medium supplemented with 10% (v/v) heat inactivated fetal bovine serum, 200 U/ml penicillin and 2 mg/ml streptomycin. Both cell lines were cultured continuou sly at 37 o C in a humidified 5% (v/v) CO 2 At confluence cells were trypsinized, centrifuged for 5 min at 110 x g and resuspended in fresh complete medium. Cell
88 viability was assessed by trypan blue exclusion and cell concentration adjusted to 1 x 10 5 ce lls/ml For the [ 3 4 cells were cultured in 96 well plates at 37 o C and 5% (v/v) CO 2 with all treatments in triplicate. After 24 h in culture, treatments consisting of 1 mg/ml of nOvUS or ovallbumin (OVA, as negative control) were added. For control wells, an equivalent volume of DPBS was added instead of proteins. 3 H]thymidine in 10 harvested onto glass fiber filters using a cell harvester device at 24 h after thymidine addition. Filters were counted for radioactivity using scintillation spectrometry. Induction of A poptosis in D 17 and PC 3 C ells To determine apoptosis, 1 x 10 4 mg/ml nOvUS or OVA For the vehicle treatment, an equivalent volume of DPBS was added instead of OvUS or OVA The final volume in all wells was adjusted with c cultured at 37 o C and 5% (v/v) CO 2 for 24 h and then harvested for the detection of apoptotic cells by terminal deoxynucleotidyl transferase mediated dUTP nick end labeling (TUNEL) All treatments were performed in trip licate and the experiment was performed on three (D17) or one (PC 3) separate occasions. Cells were processed for TUNEL staining as follows. Cells were washed by centrifugation with 0.1 M sodium phosphate, pH 7.4 with 0.9% (w/v) NaCl (PBS) and containin g 1 mg/ml polyvinyl pyrrolidone (PBS/PVP), and th in 0.2 M sodium phosphate, pH 7.4 with 0.9% (w/v) NaCl for 1 h at room temperature. Cells was transferred to a pol y l lysine coated slide and allowed to dry for at least 24 h at room
89 temperature. For TUNEL staining, slides were washed twice in PBS/PVP (2 min each) and then incubated with permeabilization solution [0.1% (v/v) Triton X 100 containing 0.1 % (w/v) sodium citrate in PBS] for 1 h at room temperature. Positive controls were incubated with RQ1 RNase free DNase (50 U/ml) at 37 o C for 1 h. Slides were washed in PBS/PVP and then conjugated dUTP and the enzyme terminal deoxynucleotidyl transferase as prepared by the manufacturer) for 1 h at 37 o C in t he dark. Negative controls were incubated in the absence of the enzyme. After washing, slides were incubated with RNase A then with 12.5 4 times in PBS/PVP to remove excess propidium iodide and coverslips mounted with mounting medium containing Prolong Antifade as recommended by the manufacturer. Slides were observed using a Zeiss Axioplan 2 fluorescence microscope with dual filter (Carl Z eiss, Inc., Gttingen, Germany ). Images were acquired using AxioVision software and a high resolution black and white AxioCam MRm di gital camera (Carl Zeiss, Inc., Thorwood, NY). Percent apoptotic cells were determined by counting the total number of TUN EL labeled nuclei in at 10 different sites on the slide. Purific ation of His Tagged rOvUS from Conditioned M edium The human embryonic kidney (HEK) 293F (Gibco Invitrogen, Carlsbad, CA) cells transfected with a plasmid construct containing the gene for OvUS (Tekin et al. 2006) was cultured continuously in selective medium [FreeStyle TM 293 expression medium containing 700 o C in a humidified 8% (v/v) CO 2 incubator according to the Conditioned medium from 293 F r OvUS cells was clarified by centrifugation at 4000 x g for 15 min and t he supernatant was removed and concentrated using Centricon PL 80 or 20 concentration devices (molecular weight exclusion limit = 30000).
90 Purification of the His tagged rOvUS was achieved using HisTrap columns. The binding buffer was 20 mM sodium phospha te buffer, pH 8.0 15 mM imidazole 0.3 M NaCl, and the elution buffer was 20 mM sodium phosphate buffer, pH 8.0, 0.3 M NaCl and with 300 or 500 mM imidazole. The concentrated medium was diluted with at least an equal volume of binding buffer and loaded onto the His Trap column s. The protein recovered after elution was buffer exchanged with DPBS and concentrated using Centricon PL 20 filters. P urity of rOvUs was assessed by SDS PAGE using 4 15% polyacrylamide Precast Tris HCl gradient gels and c oncentra tion was determined using the Bradford assay ( Bradford 1976 ) Expression and P Galactosidase The pcDNA3.1/V5 His TOPO/lacZ galactosidase ( Gibco Invitrogen, Grand Island, NY ) was used to express a His gala ctosidase as a control protein. Freestyle TM 293 F cells were transfec ted as described some where else (Tekin et al. 2006). Transfected cells were then selected as described for the rOvUS. After harvest ing, the cell pellet was collected by centrifugation at 100 x g for 10 min at room temperature. The pellet was resuspended in 20 mM sodium phosphate, pH 8.0 containing 0.3 M NaCl and 35 mM ecommendations (Sigma Aldrich). Cells were lysed by two freeze thaw cycles. The supernatant was collected by c entrifugation and the recombinant His tagged protein purified as previously described for rOvUS. Proliferation of P388D1 and PC 3 Cells The P388D1 cells were cultured continuously in DMEM supplemented with 10% (v/v) heat inactivated horse serum, 200 U/ml penicillin 3 cells
91 were grown in DMEM F12 medium (Gibco Invitrogen) supplemented with 10% (v/v) heat The proliferation assay was performed in 96 wel l culture plates at 37 o C and 5% (v/v) CO 2 with all treatments in triplicate. For P388D1 cells, 2 x 10 3 were cultured for 6 h before the addition of various concentrations (0, 125, 250, 500 and 1000 and OVA (negative control). For PC 3 cells, 1 x 10 4 rOvUS or OVA. For control wells, an equivalent volume of DPBS was added instead of proteins. T 3 glass fiber filters using a cell harvester device at 24 h after thymidine addition Filters were counted for radioactivity using scintillation spectrometry. Experiments were performed on 3 and 5 separate occasions for P388D1 and PC 3 cells respectively. Another experiment with PC 3 cells was performed as described above except with re galatcosidase at Statistical Analysis Data on [ 3 H]thymidine incorporation were analyzed by least squares analysis of variance using the General Linear Models Procedure of SAS (SAS for Windows, Release 8.02; SAS Inst., Inc., Cary, NC). Replicate was considered as a random effect and other main effects were considered fixed. Error terms were determined based on calculation of expected mean squares. In some analyses, the p diff mean separation test of SAS was performed to determine treatments that differed from untreated cells.
92 Results Inhibition of Proliferation and Induction of Apoptosis in D17 and PC 3 C ells Ovine uterine serpin inhibited proliferation (P<0.05) of both D1 7 and PC 3 cells (Figure 3 1A). The proportion of cells that were TUNEL positive was very low (1.1% for control D17 cells and 0.5% for control PC 3 cells) and was slightly increased by OvUS (2.2% for D17 cells treated with OvUS and 3.3% for PC 3 cells tre ated with OvUS; Figure 3 1B). Nonetheless, the number of apoptotic cells was very low for all treatments (see Figures 3 1C for examples of TUNEL staining). Antiproliferative Actions on P388D1 and PC 3 Cell Lines: Comparison of the Native and Recombinant Forms of OvUS Both nOvUS and rOvUS inhibited proliferation of P388D1 cells as measured by incorporation of [ 3 H]thymidine into DNA (Fig. 3 2). All concentrations of rOvUS tested were n contrast, there was reduction in proliferation at any concentration was greater for rOvUS than for OvUS. The control protein, ovalbumin, did not inhibit proliferatio n. Similar results were obtained for PC 3 cells (Figure 3 2) except that, in this case, the only significant inhibition was for rOvUS. All concentrations tested, including a concentration as low 3 H]thy midine incorporation (P<0.001) While not significant, the 3 H]thymidine incorporation. In galactosidase on proliferation of PC 3 cells. Incorporation of [ 3 H]thymidine in cells treated with 50, 100 and 200 galactosidase.
93 Figure 3 1. Effect of OvUS on induction of apoptosis in D17 and PC 3 cells as determined by TUNEL l abeling. Data in panel A illustrate that addition of 1 mg/ml OvUS reduced incorporation of [ 3 H]thymidine into D17 cells and PC 3 cells. Results for D17 cells are least square means + SEM and are based on three separate assays and with two separate batche s of OvUS. Results for PC 3 cells are least square means + SEM and are based on one assay with two separate batches of OvUS. Asterisks represent means that differ from the control value at P<0.05. Data in panel B represent the percent of cells tha t wer e TUNEL positive after 24 h culture. Data are from a representative assay. Panel C shows representative patterns of TUNEL labeling for D17 cells (top row) and PC 3 cells (bottom row). Note that yellow cells are considered apoptotic.
94 Figure 3 2. Inhibition of [ 3 H]thymidine incorporation of P388D1 cells and PC 3 cells by native (n) and recombinant (r) OvUS. Cells were cultured with various concentrations of nOvUS (open circles), rOvUS (filled circles and solid li ne) or ovalbumin (OVA) (filled circles and dashed line). Data represents least square means SEM. Means
95 Discussion Results from the present study showed that the antiprol iferative effect of OvUS is not restricted to immune cells, specifically the mitogen stimulated lymphocytes. Ovine US can inhibit the proliferation of three different tumor cell lines (D17, PC 3 and P388D1). I t is important to mention that there are diff erences between the signaling pathways of these tumor cells and the mitogen activated lymphocytes. Usually, tumor cells present mutations on genes that are related to cell growth control. In contrast, lymphocytes are in a resting state and once activated by mitogens, the expression and production the IL 2 receptor is crucial for proliferation and activation of these immune cells ( Ellery & Nicholls 2002 ) Thus, it is possible that the signaling pathways affected by OvUS to inhibit cell proliferation of tu mor cells and lymphocytes are different. It was previously shown that OvUS blocked IL 2 induced proliferation and reduced expression of CD25 (IL 2 mRNA caused by the mitogen Con A (Peltier e t al. 2000 c ). There are few serpins that inhibit cell proliferation. The mammary serine proteinase inhibitor (maspin) is one example where the function of this serpin has been related to the suppression of tumor growth by the induction of apoptosis of ca ncer cells and the reduction of angiogenesis (Sheng 2006). Another example is the pigment epithelium derived factor (PEDF) which promotes the expression of FasL, activating signal transduction pathways, in particular caspases 8 and 3 leading to endothel ial cell death (Tombran Tink & Barnstable 2003). Unlike maspin and PEDF, which inhibit proliferation by inducing apoptosis, OvUS had only a slight effect on DNA fragmentation (apoptosis) of D17 and PC 3 cells. There are two other serpins that inhibit cell proliferation, but by a caspase independent mechanisms. One is the intracellular serpin, myeloid and erythroid nuclear termination stage specific protein (MENT) which is involved in cell cycle progression and therefore in cell
96 proliferation. This serpi n inhibits the activity of the nuclear cysteine proteinase SPase, whose function has been related to the degradation of the phosphorylated form of the retinoblastoma (Rb) protein, a known regulator of the cell cycle (Irving et al. 2002 a ). The other serpin is the intracellular plasminogen activator inhibitor type 2 (PAI 2). This protein could be involved in the regulation of the cell cycle since it protected Rb from degradation by a n independent anti proteolytic mechanism ( Medcalf & Stasinopoulos 2005, Cro ucher et al. 2008). However, the effects of OvUS on cell cycle progression as a mechanism to inhibit cell proliferation remain to be determined.
97 CHAPTER 4 REGULATION OF DNA SY NTHESIS AND THE CELL CYCLE IN HUMAN PROST ATE CANCER CELLS AND LYMPHOCYTES BY O VINE UTERINE SERPIN Introduction Ser ine p roteinase in hibitors (serpins) inactivate their target proteinases through a suicide substrate like inhibitory mechanism The proteinase binds covalently to the reactive center loop (RCL) of the serpin and cleaves the scissile bond at the P1 sheet A and a distortion in the structure of the proteinase that results in its inactivation (Irving et al. 2000 Van Gent et al. 2003 Law et al. 2006). Not all serpins, however, exert proteinase inhibitory activity. Some examples are corticosteroid and thyroxine binding globulins, which function as hormone transport proteins (Pemberton et al. 1988), the chaperone heat shock protein 47 ( Sauk e t al. 2005), mammary serine protease inhibitor (Maspin), which increases the sensitivity of cancer cells to undergo apoptosis (Sheng 2006), and pigment epithelium derived factor (PEDF), which has neurotrophic, neuroprotective, antiangiogenic, and proapopto ticactions (Fernandez Garcia et al. 2007). Another class of serpins without apparent proteinase activity is the uterine serpins. These proteins, which are produced by the endometrial epithelium of the pregnant cow, sow, sheep, and goat (Moffatt et al. 19 87, Ing & Roberts 1989, Malathy et al. 1990, Leslie et al. 1990, Mathialagan & Hansen 1996 Tekin et al. 2005 a ), have been classified as either a separate clade of the serpin superfamily ( Peltier et al. 2000 c ) or as a highly diverge group of the antitrypsin clade (Irving et al. 2000). The best characterized protein of this unique group of serpins is ovine uterine serpin (OvUS). This basic glycoprotein is a weak inhibitor of aspartic proteinases (pepsin A and C) ( Mathialagan & Hansen 1996 Pel tier et al. 2000 a ), but it does not inhibit a broad range of serine proteinases (Ing & Roberts 1989, Liu & Hansen 1995). Additionally, amino acids in the hinge region of inhibitory serpins are not conserved in uterine serpins and
98 OvUS behaves different in the presence of guanidine HCl than for inhibitory serpins (Peltier et al. 2000 a ). The biological function of OvUS during pregnancy may be to inhibit immune cell proliferation during pregnancy and provide protection for the allogeneically distinct concept us (Hansen 1998). Ovine US decreases proliferation of lymphocytes stimulated with concanavalin A, phytohemagglutinin (PHA), Candida albicans and the mixed lymphocyte reaction (Segerson et al. 1984 Hansen et al. 1987 b Skopets & Hansen 1993 Skopets et a l. 1995 Peltier et al. 2000 b ). In addition, OvUS decreases natural killer cell cytotoxic activity, abortion induced by Liu & Hansen 1993) and the production of antibody in sheep immunized with ovalbumin (Skopets et al 1995). Th e antiproliferative actions of OvUS are not limited to lymphocytes. Ovine US decreases development of the bovine embryos and proliferation of mouse lymphoma, canine primary osteogenic sarcoma and human prostate cancer cell lines (Tekin et al. 2005 b 2006) The mechanism by which OvUS inhibits proliferation of cells is unknown. The protein could block activation of cell proliferation, inhibit the cell cycle at other points or induce apoptosis or other forms of cell death. For the PC 3 prostate cancer li ne, inhibition of cell proliferation by OvUS might involve reduction in interleukin 8 (IL 8) secretion because of the importance of autosecretion of this cytokine for cell androgen independent proliferation (Araki et al. 2007). The goal of the present stu dy was to evaluate the mechanism by which OvUS inhibits cell proliferation. Using PC 3 cells as a model system, it was tested whether inhibition of DNA synthesis involves cytotoxic action of OvUS, induction of apoptosis or disruption of the IL 8 autocrine loop. I t was also tested whether OvUS blocks specific steps in the cell cycle for PC 3 cells and lymphocytes.
99 Materials and Methods Materials The human prostate cancer (PC 3) cell line was purchased from ATCC (Rockville, MD), the FreeStyle TM 293 express F 12 (DMEM F12) and 0.25% Trypsin EDTA were obtained from Gibco Invitrogen (Carlsbad, CA), the RQ1 RNase free DNase and the CellTiter Glo Luminescent Cell Viability Assay kit were obtained from Promega, (Madison, WI), the DHL TM Cell Cytotoxicity Assay kit was from Anaspec (San Jose, CA) and the ELISA MAX TM Set Deluxe kit for human IL 8 was obtained from BioLegend (San Diego, CA). The in situ cell death detection kit [terminal deoxynucleotidyl t ransferase mediated dUTP nick end labeling (TUNEL)] was purchased from Roche (Indianapolis, IN), the DNase free RNase A was obtained from Qiagen (Valencia, CA), Precast Tris HCl gradient Ready gels were from BioRad (Richmond, CA) and [ 3 H]thymidine (6.7 Ci /mmol) was from ICN (Irvine, CA). The Prolong Antifade kit was purchased from Molecular Probes (Eugene, OR), Geneticin was from Research products international (Mount Prospect, IL), Centricon filter devices were from Millipore Corporation (Bedford, TX), niquel Sepharose chromatography medium (high performance) was from Amersham Biosciences (Piscataway, NJ), fetal bovine and horse serum from Atlanta Biologicals (Norcross, GA). Other reagents were obtained from either Fisher (Pittsburg, PA) or Sigma Aldrich (St. Louis, MO). Purification of rOvUS The His tagged rOvUS was purified from conditioned medium of FreeStyle TM human embryonic kidney (HEK) 293F cells (Gibco Invitrogen, Carlsbad, CA) transfected with a plasmid construct containing the gene for OvUS. Details of the cell line are provided elsewhere (Tekin et al. 2006). Cells were cultured continuously in selective medium [FreeStyle TM 293 Geneticin ] at 37 o C in a humidified 8% (v/v) CO 2
100 rOvUS was diluted 1:1 (v/v) in binding buffer [20 mM sodium phosphate buffer, 35 mM imidazole, 0.3 M NaCl, pH 8.0] and loaded into a nickel Sepharose column that was pre equilibrated with binding buffer. The His tagged rOvUS was eluted with 20 mM phosphate buffer, 500 mM imidazole, 0.3 M NaCl, pH 8.0, concentrated and buffe r exchanged into using Centricon plus 20 concentration devices. Purity of the rOvUS was assessed by sodium dodecyl sulfate polyacrylamide gel electrophoresis using precast 4 15% polyacrylamide Tris HCl gradient gels. The protein concentration was determined by Bradford assay (Bradford 1976) using bovine serum albumin as standard. For each experiment, rOvUS and the control protein, OVA, were added to culture wells of PC cells or lymphocytes dissolved in DPBS. The vehicle control included addition of DPBS at the same volume as for rOvUS and OVA. The actual volume of protein or vehicle added varied between experiments but was generally 14 25 so that the volume of DPBS was the same in all wells. PC 3 C ell C ulture The PC Nutrient Mixture F 12 (DMEM F12) supplemented wit h 10 % (v/v) heat inactivated fetal bovine serum, 200 U/ml penicillin and 2 mg/ml streptomycin at 37 o C in a humidified 5% (v/v) CO 2 incubator. For the IL 8 experiment only, the medium was modified to reduce the fetal bovine serum concentration to 4% (v/v) For all the experiments, cells were cultured in 75 cm 2 flasks until they reached 50 70% of confluence. Cells were then trypsinized, centrifuged at 110 x g for 5 min and resuspended in fresh complete medium. Cell viability was assessed by trypan blue exclusion and cell concentration was adjusted according to the requirements of each experiment.
101 [ 3 H]thymidine Incorporation by PC 3 C ells PC 5 cell/ml in a 96 well plate. Afterward s, various concentrations of rOvUS (0, 0.5, 1, 2, 4, 8, 16, 32, 64, 125 and [ 3 H]thymidine in 10 edium were added. Cells were harvested 24 h after [ 3 H] thymidine addition onto fiber glass filters using a cell harvester (Brandel, Gaithersburg, MD). Filters were counted for radioactivity using scintillation spectrometry (Beckman Coulter Inc., Fullerton CA). Each concentration of protein was tested in triplicate and the experiment was performed in six different replicates using a different batch of rOvUS for each replicate. Cell Proliferation Based on ATP C ontent 3 cells (1 x 10 5 cell s /ml ) were cultured for 24 h in a dark wall clear bottom 96 well plate. Then, treatments consisting of vehicle (DPBS) or three different lbumin; OVA) and were prepared to determine background. At 48 h after addition of treatments, ATP content per well was determined using the CellTiter Glo Lu minescent Cell Viability Assay kit according to Glo reagent were added to each well, contents of the plate were mixed on a shaker for 2 min and then incubated at room temperature for 10 mi n. Chemiluminescence was quantified using a multi detection micr oplate reader (FLX 800, BioTek, Winooski VT). All treatments were performed in triplicates and the assay was performed on three different occasions using a different batch of rOvUS for each replicate.
102 Cytotoxicity A ssay The assay was based on the release of lactate dehydrogenase into culture medium following loss of cell membrane integrity accompanying cell death (Decker and Lohmann Matthes, 1988) Procedures for cell culture and treatmen ts were similar to those described for the ATP assay. At 48 h after addition of the treatments, release of lactate dehydrogenase into the medium was determined using the DHL TM Briefly, the solution or DPBS. To facilitate cell lysing, the plate was placed on a shaker for 2 min. A total of After 10 min at room intensity was measured using a multi detection microplate reader (FLX 800) with excitation and emission wavelengths of 530 560 nm and 590 nm, respectively. Percent cytotoxicity was calculated by dividing 100 x fluorescence from the unlysed cells by fluorescence of the lysed cells. For each assay, each treatment was performed in six wells. The assay was replicated five different times using a different batch of rOvUS for each replicate. TUNEL L abeling 3 cells were cultured overnight in chamber slides at a final concentration of 1 x 10 4 cells/ml. Then, treatments consisting of vehicle ( DPBS) 50, 100 or 200 added to produce a final volume of 300 PBS/PVP [100 mM sodium phosphate pH 7.4, 0.9% (w/v) NaCl, 1 mg/ml polyvinyl pyrrolidone] and fixed with 4% (w/v) paraformaldehyde for 1 h at r oom temperature. Cells then were washed in PBS/PVP and stored at 4 o C for the TUNEL (terminal deoxynucleotidyl transferase and fluorescein isothiocyanate conjugated dUTP nick end labeling) procedure.
103 For TUNEL labeling, fixed cells were incubated for 1 h a t room temperature with permeabilization solution [PBS, pH 7.4, 0.1 (v/v) Triton X 100, 0.1% (w/v) sodium citrate). containing terminal deoxynucleotidyl transferase and fluorescein isothiocyanate conjugated dUTP, for 1 hour at 37 o C. Positive controls wer e preincubated with RQ1 RNase free DNase (50 U/ml) and negative controls were incubated without transferase. Slides were washed with ure. Slides were washed with PBS/PVP and Prolong Antifade was used to mount coverslips. Samples were observed using a Zeiss Axioplan 2 fluorescence microscope with dual filter (Carl Zeiss, Inc., Gttingen, Germany). Percent of cells with DNA fragmentati on was determined by counting the total number of nuclei and total number of TUNEL labeled nuclei at 10 different sites on the slide. The experiment was performed using three different batches of rOvUS. Secretion of IL 8 PC 3 cells (100 well plate overnight at a final concentration of 1 x 10 5 cells/ml. Treatments were then added including vehicle (DPBS, similar volume as for rOvUS and OVA treatments), and three different concentrations of rOvUS and OVA medium. At 48 h after addition of treatments, cell culture supernatants were collected, centrifuged and stored at 20 o C until ELISA for IL 8. Treatments were performed i n triplicate for each assay; the experiment was repeated on three different occasions using three different batches of the recombinant protein. For the measurement of IL 8, the ELISA MAX TM Set Deluxe kit for human IL 8 was used according to the manufacture conditioned medium.
104 Cell Cycle A nalysis PC 5 OVA were added after addition of treatments, cells were collected by trypsinization and washed with DPBS. Cells were fixed overnight in 70% (v/v) ethanol at 4 o C, washed with DPBS and resuspende d with 500 100, 0.05 mg/ml DNase free RNase A, 5 propidium iodide]. Cells were then analyzed by flow cytometry using a FACSort flow cytometer (Becton Dickinson, Franklin Lakes, NJ) and the r ed fluorescence of single events was recorded at wavelengths of 488 nm (excitation) and 600 nm (emission) Data were gated using pulse width and pulse area to exclude doublets, and the percent of cells present in each phase of the cell cycle was calculate d using ModFITLT V3.1 software (Verity Software, Topsham, ME). The experiment was performed on three occasions with five different batches of rOvUS. For the sheep lymphocyte experiment, mononuclear cells were purified by density gradient centrifugation from the buffy coat of heparinized peripheral blood collected by jugular venipuncture from non pregnant Ram bouillet ewes (Tekin & Hansen 2002). After removing red blood cells by incubation with red cell lysis buffer (0.01 M Tris HCl pH 7.5 containing 8.3 g/L of ammonium chloride), cell viability was assessed by trypan blue exclusion, and concentration adjusted to 4 x 10 6 cells/ml. Cells were then suspended in a culture medium consisting of Tissue Culture Medium 199 containing 5% (v/v) horse serum, 200 U/ ml penicillin, 0.2 mg/ml streptomycin, 2 mM glutamine and 10 5 mercaptoethanol and aliquots of cells cultured in 4 well plates and with treatments of
105 72 and 96 h in culture at 37 o C in a humidified 5% (v/v) CO 2 incubator, lymphocytes were collected and washed with DPBS. Thereafter, lymphocytes were fixed and treated as described above. The experiment was performed s eparately for lymphocytes from four different sheep. Three different batches of rOvUS were tested for each she ep Statistical A nalysis Data were analyzed by least squares means analysis of variance using the General Linear Models Procedures of SAS (SAS Sy stem for Windows, Version 9.0; SAS Institute, Cary, NC, USA). Error terms were determined based on calculation of expected mean squares with replicate considered random and other main effects considered fixed. For the cytotoxicity and IL 8 data, orthogon al polynomial contrasts were used to determine the linear and quadratic effects of rOvUS and OVA. In other analysis, the pdiff mean separation test of SAS was used to distinguish the difference of various levels of a treatment. Results and Discussion P rol iferation of PC 3 c ells The antiproliferative effects of rOvUS on proliferation of PC 3 cells were evaluated by two different assays. In the first experiment, it was shown that rOvUS caused a concentration dependent decrease in incorporation of [ 3 H]thymid ine into DNA (P<0.001) with the minimum 1). The antiproliferative actions of OvUS using [ 3 H]thy midine uptake as the measure of proliferation has been demonstrated previously for PC 3 cells and other cell typ es (Segerson et al. 1984 Hansen et al. 1987 b Skopets & Hansen 1993 Skopets et al. 1995 Peltier et al. 2000 b Tekin et al. 2005 b 2006). To confirm this effect of rOvUS reflected an inhibition in cell proliferation and not a disruption in [ 3 H]thymidine uptake by the cells, antiproliferative effects were also evaluated by an assay in which the relative total
106 Figure 4 1. Inhibition of [ 3 H]thymidine incorporation of PC 3 cells by recombinant ovine uterine serpin (rOvUS) The inset graph is provided to clarify the effects of rOvUS g/ml). Data represent least squares means SEM. Values that differ from untreated cells are indicated by asterisks (***P<0.001).
107 number of cells per well was estimated by the ATP content per well. Treatment with rOvUS 2). In contrast, the control serpin, ovalbumin, did not cause effect in the ATP content per well. The finding tha t rOvUS reduced ATP content per well confirms that the effects of OvUS to reduce [ 3 H]thymidine incorporation reflect a reduction in cell proliferation rather than interference with [ 3 H]thymidine transport into the cell. Lactate Rehydrogenase R elease Poss ible cytotoxic effects of rOvUS on PC 3 cells were evaluated by measurements of lactate dehydrogenase release into culture medium (Figure 4 3). None of the concentrations of rOvUS or OVA tested caused an increase in the percent of lysed cells during cultu re. Thus, rOvUS does not inhibit proliferation through induction of cell death. DNA F ragmentation ( A poptosis) The TUNEL procedure was used to test whether rOvUS decreased cell proliferation by induction of DNA fragmentation characteristic of apoptosis an d other forms of cell death. Representative images of TUNEL labeled cells are shown in Figure 4 4 and the average percent of cells that were TUNEL positive is shown in Figure 4 5. Treatment of PC 3 with either rOvUS or the control protein OVA did not inc rease the percent of cells that were TUNEL positive at either 24 or 48 h after treatment; the percentage of cells that were TUNEL positive was low for all groups (< 5.7 %). The fact that rOvUS did not induce apoptosis makes the action of this serpin disti nct from that of two other serpins that inhibit cell proliferation. Both maspin (Sheng 2006) and PEDF (Fernandez Garcia et al. 2007) are proapoptotic serpins. Interleukin 8 Secretion Interleukin 8 accumulation in the medium was measured because of the autocrine effect of this chemokine on prostate cell proliferation (Araki et al. 2007). In addition, at least one class of
108 Figure 4 2. Inhibition of proliferation of PC 3 cells by recombinant ovine uterine serpin (rOvUS) as determine d by ATP content/well. Ovalbumin (OVA) was used as a negative control. Data represent least squares means SEM. Means that differ from untreated ce <0.01;***P<0.001)
109 Figure 4 3. Lack of cyto toxic effect of recombinant ovine uterine serpin (rOvUS) on PC 3 cells was measured by the release of lactate dehydrogenase. Ovalbumin (OVA) was used as a control protein. Data represent least squares means SEM
110 Figure 4 4. Repr esentative photomicrographs of PC 3 cells labeled using the TUNEL procedure the control protein ovalbumin. Cells in panel D were treated with DNAse as a positive control.
111 Figure 4 5. Effect of recombinant ovine uterine serpin (rOvUS) on DNA fragmentation (apoptosis) of PC 3 cells. Results show the percent of TUNEL positive cells at 24 (A) and 48 (B) h after addition of the treatments. Data represen t least squares means SEM. Ovalbumin (OVA, 200 g/ml) was used as control protein.
112 molecule that inhibits PC 3 cell proliferation, soy isoflavones, also reduces IL 8 secretion (Handayani et al. 2006). As shown in Figure 4 6, there was, however, no effect of rOvUS on accumulation of IL 8 into cond itioned cultured medium. Thus, rOvUS does not block PC 3 cell proliferation through inhibition of IL 8 secretion. Cell Cycle D ynamics Dynamics through the different phases of the cell cycle were affected by the treatment of PC 3 cells with rOvUS. Repre sentative DNA histograms after treatment with vehicle or 200 Figures 4 7A and 4 7B while least squares means SEM for for results at 12 and 24 h after treatment are shown in Figures 4 7C and 4 7D, respectively. At 12 h after addition of tr eatments, rOvUS decreased (P<0.01) the percent of cells in S phase and increased the percent of cells in the G 2 /M phase (Figure 4 7C). There was no effect of rOvUS on the percent of cells in G 0 /G 1 0 /G 1 (P<0.001), decreased the percent of cells in S phase (P<0.01), and did not affect the percent of cells in G 2 /M phase (Figure 4 7D). Control of the cell cy cle dynamics by rOvUS was also evaluated in a second cell type the peripheral blood lymphocyte. Representative DNA histograms for PHA treated lymphocytes are shown for control cells and cells treated with 200 8A and 4 8B, respec tively while least squares means SEM are shown in Figures 4 8C and 4 8D. At both 72 (Figure 4 8C) and 96 h (Figure 4 8D) after addition of PHA, rOvUS increased (P<0.001) the proportion of lymphocytes in the G 0 /G 1 phase and decreased (P<0.05) the proport ion of cells in the S phase. In contrast, there was no effect of the control protein (OVA) on the distribution of cells in the cell cycle. These results indicate that OvUS block cell proliferation through cell cycle arrest in both PC 3 and lymphocytes. The differences in specific stages at which the cell cycle was blocked between PC 3 cells and
113 Figure 4 6. Effect of recombinant ovine uterine serpin (rOvUS) on interleukin (IL) 8 concentration in cell culture supernatants of PC 3 cel ls. Ovalbumin (OVA) was used as control serpin. Data represent least squares means SEM.
114 Figure 4 7. Cell cycle dynamics of PC 3 cells as affected by recombinant ovine uterine serpin (rOvUS). Controls included vehicle (co ntrol) and ovalbumin (OVA). Representative DNA histograms for analysis at 12 h after treatment with vehicle or 200 g/ml rOvUS are shown i n panels A and B, respectively. The least squares means for results of three separate assays are shown in panels C an d D for analysis at 12 h (C) and 24 h (D) after treatment. Bars with d P<0.10, others at P<0.05 or less).
115 lymphocytes is likely to be caused by differences in activation and regulation pathways for these two cell types Unlike PC 3 cells, lymphocytes are arrested at G 0 until proliferation is induced by addition of PHA. Inhibitions at points in the cell cycle other than G 0 /G 1 are less likely to be seen since few cells progress to later stages of the cell cycle. In add ition, it is possible that genetic mutations in PC 3 cells compromise some potential regulatory mechanisms. In particular, unlike lymphocytes, PC 3 cells lack functional p53 (Isaacs et al. 1991) which causes cell cycle arrest at G1/S by inducing p21 cip1 t hat inhibits cyclin dependent kinases (Harper et al. 1993, Duli et al. 1994). The mechanism by which OvUS inhibits cell cycle dynamics is not understood. One serpin has been identified which can affect cell cycle regulatory proteins. This serpin, myel oid and erythroid nuclear termination stage specific protein (MENT), is a nuclear protein that inhibits cell proliferation through interactions with a nuclear protein with papain like cysteine proteinase activity (Irving et al. 2002 a ) Inhibition of the p roteinase prevents degradation of the cell cycle protein Rb although antiproliferative effects may depend more on other actions of MENT to mediate euchromatin condensation in an Rb independent manner (Irving et al. 2002 a ) In any case, OvUS, is apparently without proteinase inhibitory activity and is an extracellular protein that is unlikely to achieve a nuclear localization. The antipepsin activity of OvUS is probably not biologically significant. Ovine US is a very weak inhibitor of pepsin C [a 35 and 8 fold molar excess of OvUS was required to inhibit pepsin A and C ( Mathialagan & Hansen 1996)] and the binding of OvUS to pepsin is electrostatic ( Peltier et al. 2000 a ). Moreover, pepsin shows an acidic pH optimum and is unlikely to be involved in cell proliferation under the conditions utilized.
116 Figure 4 8. Cell cycle dynamics of lymphocytes as affected by recombinant ovine uterine serpin (rOvUS). Controls included vehicle (control) phytohemagglutinin (PHA) and ovalbumin (OV A). Representative DNA histograms for analysis at 7 2 h after treatment with PHA or 200 g/ml rOvUS are shown i n panels A and B, respectively. The least squares means for results of four separate assays are shown in panels C and D for analysis at 7 2 h (C) and 96 h (D) after treatment. Bars with d ifferent superscripts differ ( P<0.05 or less)
117 A major point in the cell cycle regulated by OvUS is transition from G 0 /G 1 to S phase: rOvUS decreased the proportion of cells in S phase in all experiments and increased the proportion of cells in G 0 /G 1 at 24 h after treatment for PC 3 cells and at both times examined for lymphocytes. Ovine uterine serpin can bind to cell membranes (Liu et al. 1999) and, perhaps, OvUS inhibits proliferation by activation of signal transduction systems that inhibit transition from G 0 /G 1 to S phase or prevents pro pro liferative molecules in culture medium from binding their receptors. Experiments with Rp 8 Cl cAMPS, a selective inhibitor of cAMP dependent type I protein kinase A, indicated that effects of OvUS on proliferation of PHA stimulated lymphocytes are not dep endent on this kinase (Tekin et al. 2005 b ). Studies to determine activation of other anti proliferative signal transduction systems by OvUS are warranted. Taken together, the present study indicates that the mechanism by which OvUS inhibits prolifera tion of PC 3 cells and lymphocytes involves cell cycle arrest and not, at least for PC 3 cells, apoptosis, cytotoxicity or inhibition of IL 8 secretion.
118 CHAPTER 5 CHANGES IN EXPRESSIO N OF CELL CYCLE RELATED GENES IN PC 3 PROSTATE CANCER CELLS CAUSED BY OVIN E UTERINE SERPIN Introduction Uterine serpins (US), also known as uterine milk proteins (UTMP), are member s of the serine proteinase inhibitor (serpin) superfamily ( Ing & Robets 1989 Mathialagan & Hansen 1996) and are designated as SER PINA14. These progesterone induced glycoproteins are secreted in large quantities into the uterus of a restricted group of mammals during pregnancy (Malathy et al. 1990, Leslie et al. 1990, Tekin et al. 2005 a ). The best studied of the US is the protein f ound in the sheep. Ovine uterine serpin ( OvUS ), which is the most abundant protein in uterine secretions of the pregnant sheep (Moffatt et al. 1897 Hansen et al. 1987 a ), is an example of a serpin that has gained a new function while apparently losing pro teinase inhibitory activity characteristic of serpins. O ther examples include the heat shock protein 47 (Nagata 1998), corticosteroid and thyroxine binding globulin (Pemberton et al. 1998) and angiotensinogen (Morgan et al. 1996). Inhibitory serpins ina ctivate their target proteinases by an irreversible suicide substrate like mechanism after the proteinase binds to the reactive center loop (RCL) (Silverman et al. 2001). Usually, inhibitory serpins are recognized by a consensus sequence in the hinge regi on which is localized within the RCL of the serpin (Irving et al. 2000) but the hinge region of OvUS is not conserved with inhibitory serpins (Tekin et al. 2005 a Irving et al. 2000). Ovine US does not inhibit cathepsins B, D and E ( Mathialagan & Hansen 1 996), dipeptidyl proteinase IV ( Liu & Hansen 1995), trypsin, chymotrypsin, plasmin, thrombin, elastase and plasminogen activator ( Ing & Robets 1989). While OvUS inhibits the aspartic proteinases pepsin A and C, this inhibition is atypical for serpins sinc e an excess of 35 and 8 fold molar of OvUS was required for a 50% inhibition of pepsin A and C, respectively ( Mathialagan & Hansen 1996).
119 T he role of OvUS during pregnancy has been linked to the protection of the allogeneic conceptus against the maternal immune system (Hansen 1998). It exerts this role in large part by inhibiting the proliferation of activated lymphocytes (Hansen et al. 1987 b Peltier et al. 2000 b ) The antiproliferative effect of OvUS is also exerted on some other cell types including m ouse lymphoma (P388D1), canine primary osteogenic sarcoma (D 17) and human prostate cancer (PC 3) cell lines (Tekin et al. 2005 a 2006). The mechanism by which OvUS inhibits cell proliferation is poorly understood. Ovine US did not cause cytotoxic or a poptotic effects on lymphocytes or PC 3 cells (Tekin et al. 2005 b Skopets & Hansen 1993, Padua & Hansen 2008). It was recently determined that OvUS blocks cell cycle progression in mitogen stimulated lymphocytes and increases the number of cells at the G 0 /G 1 stage at 96 h after addition of the protein ( Padua & Hansen 2008). Ovine US also blocks the progression of the cell cycle of PC 3 cells in a manner that leads to an accumulation of cells at G 2 /M at 12 h after addition of the protein and at G 0 /G 1 at 2 4 h after treatment ( Padua & Hansen 2008). The objective of the present study was to understand the mechanism by which OvUS blocks cell cycle progression in PC 3 cells by determining cell cycle related genes whose expression is altered by OvUS. Materials and Methods Materials The FreeStyle TM Mixture F 12 (DMEM F12) and 0.25% Trypsin EDTA were purchased from Gibco Invitrogen (Carlsbad, CA). The G418 disulfate ( geneticin) was purchased from R esearch products international (Mount Prospect, IL), nickel Sepharose chromatography medium (high performance) from Amersham Biosciences (Piscataway, NJ), Precast Tris HCl gradient Ready gels were obtained from BioRad (Richmond, CA) and Centricon filter devices were from
120 Millipore Corporation (Bedford, TX). The human prostate cancer (PC 3) cell line was from ATCC (Rockville, MD), [ 3 H]thymidine (6.7 Ci/mmol) was from ICN (Irvine, CA) and fetal bovine and horse sera from Atlanta Biologicals (Norcross, GA). Other reagents were obtained from either Fisher (Pittsburg, PA) or Sigma Aldrich (St. Louis, MO). Purification of rOvUS Human embryonic kidney (HEK) 293F (Gibco Invitrogen, Carlsbad, CA) cells transfected with a plasmid construct containing the gene for OvUS (Tekin et al. 2006) were cultured continuously in selective medium [FreeStyle TM g eneticin] at 37 o C in a humidified incubator containing a gas environment of 8% (v/v) CO 2 in air The rOvUS was purified by using immobilized metal ion (nickel) exchange chromatography as described by Padua & Hansen 2008. Briefly, rOvUS was eluted with 20 mM phosphate buffer, 500 mM imidazole, 0.3 M NaCl, pH 8.0, concentrated and buffer phosphate buffered saline (DPBS) using Centricon plus 20 concentration devices. Sodium dodecyl sulfate polyacrylamide gel electrophoresis under reducing conditions using 4 15% polyacrylamide Tris HCl gradient gels and Coomassie Blue were used to assess the purity of the rOvUS. After filter concentration of the protein was determined by Bradford assay (Bradford 1976) using bovine serum albumin as standard. PC 3 Cell C ulture The PC 3 cell line wa s cultured into 75 cm 2 flasks continuously in complete medium 12 (DMEM F12) supplemented with 10 % (v/v) heat inactivated fetal bovine serum, 200 U/ml penicillin and 2 mg/ml streptomycin] at 37 o C in a hu midified incubator with a gas environment 5% (v/v) CO 2 in air. Cells were then trypsinized after reaching 50 70% of confluence, centrifuged at 110 x g for 5 min and
121 resuspended in fresh medium. The via bility of the cells was assessed by trypan blue exclu sion and cell concentration was adjusted according to each experiment. Proliferation A ssay PC 3 cells were cultured in of 1 x 10 4 cells/ml After 24 h in culture, cells were treated with either rOvUS, ovalbumin (OVA, a control serpin dissolved in DPBS) or vehicle ( DPBS ). The vehicle was added to control wells at an equivalent volume ( 6 ) as for the rOvUS and OVA Additional culture medium was added to all wells to bring the final vo lume to 200 l. The final concentration of rOvUS and Ci [ 3 H]thymidine was added to the wells. Cells were collected on fiber glass filters by using a cell har vester (Brandel, Gaithersburg, MD) 24 h after [ 3 H]thymidine addition. Radioactivity on the filters was counted by scintillation spectrometry (Beckman Coulter Inc., Fullerton, CA). The experiment was replicated on five different occasions using a total of four different batches of rOvUS. For each replicate, each treatment was tested in triplicate. Cell Culture for RNA Extraction PC 3 cells were cultured in 4 well plates at a final concentration of 4 x 10 5 cells/ml in 100 After 24 h of culture treatments and complete medium were added to achieve a final concentration of 200 or an equivalent volume of DPBS as experimental control in medium was removed from the plates and cell lysed for total RNA cell extraction as described below. The experiment was replicated four times, with a different batch of rOvUS for each replicate. RNA Extraction Total RNA was extracted using the RNeasy plus micro kit (Qiagen Inc, Valencia, CA) 3 cells were lysed in wells for 5 min by
122 microcentrifuge tu bes, vortexed for 1 min and placed into gDNA eliminator columns to remove genomic DNA. After mixing with 70% (v/v) ethanol, samples were transferred RNeasy MinElute spin columns, washed and total RNA eluted with RNase free water. RNA concentration and qu ality was determined by the Agilent 2100 BioAnalyzer (Agilent Technologies, Santa Clara, CA) at the Gene Expression Core Laboratory of the Interdisciplinary Center of Biotechnology Research, University of Florida. High quality RNA was used for the RT PCR cDNA Synthesis and Real Time PCR A rray The cDNA for each RNA sample was obtained by using the Super Array RT 2 First Strand nstructions. Briefly, after genomic DNA elimination, the reverse transcription reaction was performed at 42 o C for 15 min and then heated at 95 o C for 5 min to inactivate the enzyme. The cDNA was mixed with the RT 2 SYBR green/ROX q PCR master mix (SABiosci ences Corporation) and 25 l aliquots were loaded into each well of the RT 2 Profiler PCR Array (SA Biosciences Corporation, catalog number PAHS 020A). The PCR array was designed to study the profile of 84 human cell cycle related genes (Table 5 1). The P CR array experiments were performed on an ABI 7300 instrument (Applied Biosystems, Foster City, CA). The conditions for amplification were as follows: 1 cycle of 10 min at 95 o C followed by 40 cycles of 15 sec at 95 o C and 1 min at 60 o C. Statistical Analy sis The General Linear Models procedure of SAS (SAS System for Windows, Version 9.0; SAS Institute, Cary, NC, USA) was used to analyze the data from the proliferation experiments by the least square means analysis of variance. All main effects were consid ered fixed and the
123 Table 5 1 Cell cycle related genes screened using the RT 2 Profiler TM PCR Array. 1 House keeping genes. 2 Human Genomic DNA contamination control. 3 Reverse Transcription control. 4 Positiv e PCR control. ABL1 ANAPC2 ANAPC4 DIRAS3 ATM ATR BAX BCCIP BCL2 BIRC5 BRCA1 BRCA2 CCNB1 CCNB2 CCNC CCND1 CCND2 CCNE1 CCNF CCNG1 CCNG2 CCNH CCNT1 CCNT2 CDC16 CDC2 CDC20 CDC34 CDK2 CDK4 CDK5R1 CDK5RAP1 CDK6 CDK7 CDK8 CDKN1A CDKN1B CDKN2A CDKN2B CDKN3 CHEK1 CHEK2 CKS1B CKS2 CUL1 CUL2 CUL3 DDX11 DNM2 E2F4 GADD45A GTF2H1 GTSE1 HERC5 HUS1 KNTC1 KPNA2 MAD2L1 MAD2L2 MCM2 MCM3 MCM4 MCM5 MKI67 MNAT1 MRE11A NBN PCNA RAD1 RAD17 R AD51 RAD9A RB1 RBBP8 RBL1 RBL2 RPA3 SERTAD1 SKP2 SUMO1 TFDP1 TFDP2 TP53 UBE1 B2M 1 HPRT 1 RPL13A 1 GADPH 1 ACTB 1 HGDC 2 RTC 3 RTC 3 RTC 3 PPC 4 PPC 4 PPC 4
124 model included effects of treatments and batch of rOvUS. Differences between levels of a treatment were determined by the pdiff mean separation test of SAS. 35 cycles were considered as non detectable and assigned a value of 35. The average of four house keeping genes [Beta 2 microglobulin (B2M), Hypoxanthine phosphoribosy ltransferase 1 (HPRT1), Glyceraldehyde 3 phosphate dehydrogenase actin (ACTB)] was used Ct of the control groups. The fold change for each gene was calculated by 2 and the statistical analysis to determine differences between treatments was performed using the RT 2 Profiler PCR Array Data Analysis web based software (SABiosciences Corp o ration ). Results Inhibition of PC 3 C ell P roliferation by OvUS The inhibitory effect of OvUS on the proliferation of PC 3 cells is shown in Figure 5 1. Incorporation of [ 3 H]thymidine into the DNA of PC 3 cells was reduced (p<0.05) by rOvUS. In contrast, the control serpin OVA did not affect [ 3 H]thymidine incorporation. Cell C ycle Related Gene E xpressi on Profile at 12 h after T reatment with rOvUS The mRNA expression of 17 genes was significantly altered by rOvUS (Table 5 2). Three genes involved in cell c ycle checkpoint and arrest were up regulated. These genes were CDKN1A (p21 cip 1 ), CCNG2 (cyclin G2), and CDKN2B (p15 ink ). In addition, the mRNA for 14 genes was decreased by rOvUS. Among these are 3 genes (MCM3, MCM5 and PCNA), whose gene products are re quired at the S phase of the cell cycle for DNA synthesis and replication. Others are genes involved in the regulation and progression at the M phase [CDC2, CKS2, CCNH (cyclin H), BIRC5 (survivin), MAD2L1, MAD2L2] and at the G 1 phase (CDK4, CUL1,
125 Fig ure 5 1. Inhibition of [ 3 H]thymidine incorporation of PC recombinant ovine uterine serpin (rOvUS) Ovalbumin (OVA) was used as control serpin and Data repr esent least squares means SEM. Bars with different letters differ (p<0.05).
126 Table 5 2 Regulation of cell cycle related genes of PC 3 cells after 12 h of treatment with 200 g/ml recombinant ovine uterine serpin. a CDK2 associated dual specificity phosphatase b Mitotin ( S. cerevisiae ) c Cell division cycle 46 ( S. cerevisiae ) Gene Symbol Description F old Change 95%CI P value Up regulated Control Treat CCNG2 Cyclin G2 7.545 0.29 6.134 0.31 2.6585 (1.12, 4.20) < 0.05 CDKN1A CDK inhibitor 1A ( p21, Cip1) 4.078 0.05 2.727 0.24 2.5502 (1.70, 3.40) < 0.01 CDKN2B CDK inhibitor 2B p15 8. 25 0.36 7.037 0.32 2.3184 (0.81, 3.83) < 0.05 Down regulated BIRC5 Survivin 6.885 0.11 7.514 0.16 0.6465 (0.47, 0.82) < 0.05 CCNH Cyclin H 3.783 0.03 4.042 0.06 0.8354 (0.76, 0.91) < 0.01 CDC2 Cell d ivision cycle 2 1.45 0.07 1.844 0.06 0.7608 (0.67, 0.85) < 0.01 CDK4 Cyclin dependent kinase 4 3.45 0.06 3.942 0.08 0.7111 (0.61, 0.81) < 0.01 CDKN3 Cyclin dependent kinase inhibitor 3 a 1.265 0.05 1.519 0.07 0.8384 (0.74, 0.94) < 0.05 CKS2 CDC28 protein kinase regulatory s ubunit 2 1.13 0.07 1.437 0.06 0.8084 (0.71, 0.91) < 0.05 CUL1 Cullin 1 4.375 0.07 4.629 0.06 0.8384 (0.74, 0.94) < 0.05 MAD2L1 MAD mitotic deficient like 1 3.06 0.06 3.622 0.17 0.6774 (0.51, 0.85) < 0.05 MAD2L2 MAD mitotic deficient like 2 4.593 0.06 4.944 0.09 0.7836 (0.66 0.90) < 0.05 MCM3 Minichromosome maintenance deficient 3 b 2.563 0.17 3.632 0.25 0.4765 (0.28, 0.67) < 0.05 MCM5 Minichromosome maintenance deficient 5 c 5.965 0.16 6.604 0.05 0.642 (0.50, 0.79) < 0.05 PC NA Proliferating Cell Nuclear Antigen 2.225 0.20 2.909 0.08 0.6223 (0.44, 0.80) < 0.05 RAD1 RAD1 homolog ( S. pombe ) 7.41 0.06 7.719 0.07 0.807 (0.71, 0.91) < 0.05 RBBP8 Retinoblastoma Binding Protein 8 4.233 0.15 4.927 0.15 0.618 (0.44, 0.8 0) < 0.05
127 CDKN3). The last two genes that were down regulated were DNA damage checkpoint and repair genes RBBP8 (retinoblastoma binding p rotein 8) and RAD1. Cell C ycle Related Gene Expression Profile at 24 h after T reatment with rOvUS Treatment of PC 3 cells with rOvUS for 24 h caused down regulation of 16 genes (Table 5 3). Some of them (BIRC5, CDK4, CDKN3, CKS2, MAD2L2 and RAD1) were also down regulated at 12 h. The others are genes related to the regulation and progression through the M (CCNB1, CDK5RAP1, CDC20 and E2F4) and G 1 ( CDK4 and TFDP2 ) Likewise, r OvUS down regulated t hree DNA damage checkpoint related genes ( RAD17 KPNA2 an d BRCA1 ) whose activation blocks cell cycle progression at all stages of the cell cycle. The expression of BCCIP (BRCA2 and CDKN1A interactin g protein) a gene involved in DNA repair, spindle formation and cytokinesis was also down regulated by rOvUS. In addition, rOvUS caused the down regulation of M KI67. The gene product of MKI67 (k i 67) is a marker of cell p roliferation. Discussion It was previously shown that OvUS inhibited cell proliferation of PC 3 cells by disrupting cell cycle progression ( Padua & Hansen 2008). The results presented in this study corroborate those findings and provide an overview of the changes in gene expression that are associated with alterations in the cell cycle. In particular, the inhibition of the cell cycle progression is initially associated with increased expression of genes that block cell cycle and decreased expression of genes needed for progression through G 1 S and M phases. After more prolonged treatment, the inhibition of expression of genes required for cell cyc le progression is extended to a wider range of genes. In an earlier study, OvUS caused accumulation of PC 3 cells at the G 2 /M phase at 12 h
128 Table 5 3. Down regulation of human cell cycle related genes of PC 3 cells after 24 h of treatment with 200 g/ml recombinant ovine uterine serpin. Gene Symbol Description Fold change 95%CI P value Down regulated Control Treat BCCIP BRCA and CDKN1A interacting protein 2.63 0.16 3.17 0.10 0.6881 (0.51, 0.87) < 0.05 BIRC5 Survivin 6.21 0.09 7.32 0.29 0.4643 (0.27, 0.65) < 0.05 BRCA1 Breast cancer 1, early onset 6.41 0.16 7.29 0.30 0.5445 (0.29, 0.80) < 0.05 CCNB1 Cyclin B1 1.10 0.06 2.00 0.30 0.537 (0.31, 0.76) < 0.01 CDC20 Cell Division Cycle 20 homolog 0.53 0.20 1.55 0.17 0 .4916 (0.32, 0.67) < 0.01 CDK4 Cyclin dependent kinase 4 2.88 0.11 3.41 0.12 0.6941 (0.54, 0.85) < 0.05 CDK5RAP1 CDK5 regulatory subunit associated protein 1 9.07 0.19 9.65 0.13 0.6704 (0.46, 0.88) < 0.05 CDKN3 Cyclin dependent kinase inhibitor 3 a 1.02 0.01 1.68 0.24 0.6354 (0.43, 0.84) < 0.05 CKS2 CDC28 protein kinase regulatory subunit 2 0.91 0.08 1.64 0.20 0.6042 (0.43, 0.78) < 0.05 E2F4 E2F transcription factor 7.14 0.14 7.62 0.07 0.7148 (0.57 0.86) < 0.05 KPNA2 1.97 0.23 2.89 0.25 0.5269 (0.29, 0.77) < 0.05 MAD2L2 MAD2 mitotic arrest deficient like 2 3.99 0.08 4.69 0.04 0.6148 (0.54, 0.69) < 0.001 MKI67 Antigen identified by monoclonal antibody Ki 67 3.35 0.24 4.48 0 .34 0.4579 (0.20, 0.72) < 0.05 RAD1 RAD1 homolog ( S. pombe ) 6.96 0.06 7.33 0.13 0.7741 (0.62, 0.92) < 0.05 RAD17 RAD17 homolog ( S. pombe ) 6.38 0.06 6.56 0.02 0.8831 (0.80, 0.96) < 0.05 TFDP2 Transcription factor Dp 2 7.21 0.09 7.49 0.04 0.82 11 (0.71, 0.93) < 0.05 a CDK2 associated dual specific phosphatase.
129 after treatment ( Padua & Hansen 2008). It is clear from examination of Figure 5 2A that different phases of the cell cycle are disrupted by OvUS at 12 h. Ovine US caused an increase in mRNA expression of CDKN1A (p21 cip1 ), CDKN2B (p15 ink ) and CCNG2 (cyclin G2), all of which are involved in cell cycle checkpoint and arrest. It is very likely that the upregulation of expression of these genes is the proximal cause for the inhi bition of cell cycle progression. CDKN1A for example, belongs to the cyclin dependent kinase inhibitor (CDKI) family and inhibits cell cycle progression by inhibiting CDK2 and CDK4 and by blocking DNA replication and repair by binding to PCNA (Harper et a l. 1993, Waga et al. 1994, Li et al. 1994, Cayrol et al. 1998) This inhibitor of cell cycle progression causes arrest at G 1 S and G 2 phases (Harper et al. 1993, Krug et al. 2002). CDKN2B also belongs to the CDKI family and binds to CDK4 and CDK6 to pre vent their association with cyclin D, thereby blocking the cell cycle at G 1 (Krug et al. 2002). Finally, CCNG2 (cyclin G2) a non typical cyclin wh ose expression is independent of p53 (Bates et al. 1996) blocks cell cycle progression at the G 1 /S phase by a ssociation with the active protein phosphatase 2A (Bennin et al. 2002). The CCNG2 gene is also expressed at the late S and G 2 phases (Le et al. 2007). At 12 h, OvUS caused down regulation of genes involved in DNA replication (PCNA, MCM3 and MCM5) at the S phase and those implicated in the regulation and progression of the cell cycle at M (CDC2, CCNH, MAD2L1, MAD2L2) and G 1 (CDK4, CUL1, CDKN3) phases (Figure 5 2A). As an example, CCNH (cyclin H) is the regulatory subunit of the cdk activating kinase (CAK ) and is distinct from mitotic cyclins because is expressed constantly through the cell cycle. The function of cyclin H is related to the phospholylation of different cyclin dependent kinases (CDKs) and components of the transcriptional machinery (Kaldis 1999). At
130 Figure 5 2. Points in the cell cycle where genes were differentially regulated by ovine uterine serpin at 12 and 24 h are represented in panel A and B, respectively. Up regulated genes are in green and down regulated genes in red. Genes that block cell cycle progression are underlined. Genes that were regulated at 12 and 24 h are shown with an asterisk (*) Genes involved in DNA repair are in the center of the cycle to represent that many of these are involved in s everal stages of the cell cycle.
131 24 h, downregulation of cell cycle related genes became more widespread (Figure 5 2B). Some genes whose mRNA were decreased at 12 h remained so at 24 h ( BIRC5, CDK4, CDKN3, CKS2, MAD2L2 and RAD1 ). Additional genes were a lso decreased at 24 h (CCNB1, CDK5RAP1, CDC20, E2F4 TFDP2 RAD17 KPNA2 and BRCA1, BCCIP and MKI67 ) Down regulation of genes such as CCNB1, BIRC5, CDK4, CDC20 and TFDP2 would impede progression through the G 1 S and G 2 /M phases (Figure 5 2B). For examp le, CDC20 is the activator of the anaphase promoting complex/cyclosome (APC/C) required for the metaphase anaphase progression during mitosis (Baker et al. 2007). It was inhibited by 50% by OvUS. The down regulation of MKI67 at 24 h suggests that a propor tion of PC 3 cells entered the resting state (G 0 ) since the gene product for MKI67 (ki 67) is a cell marker linked to proliferation and is present in all stages of the cell cycle with the exception of the G 0 stage (Gerdes et al. 1984). This is consistent with the global decrease in gene expression observed at 24 h and also with earlier observations where OvUS caused an increase in the proportion of PC 3 cells at G 0 /G 1 phase at 24 h after treatment ( Padua & Hansen 2008). Upregulation of expression of C DKN1A (p21 cip ), CDKN2B (p15 ink ) and CCNG2 (cyclin G2) at 12 h is likely to be a cause for the down regulation of genes at 12 and 24 h. It has been shown that high levels of expression of CDKN1A (p21 cip ) down regulates the expression of BIRC5 (survivin) in other cells (Lhr & Mritz 2003 Xiong et al. 2008). Also, it is possible that the down regulation of expression of some genes is the cause for the inhibition of transcription of other genes. As an example, the down regulation of CKS2 causes a reduction in the transcription of CCNB1 (cyclin B1) and CDC2 (CDK1) (Martinsson Ahlzn et al. 2008). The activity of CDK1 is required for the regulation of some DNA repair pathways (Aylon et al. 2004, Branzei & Foiani 2008).
132 T he lack of functional p53 on PC 3 cel ls (Isaacs et al. 1991) probably affects feedback loops in response to OvUS. This protein is an effector molecule in the DNA repair pathway and lack of p53 abolishes the DNA checkpoints response and apoptosis (Sancar et al. 2004, Gatz & Wiesmller 2006). Moreover, c ells with disrupted p53 can overcome controls at the G 2 /M checkpoint and fail to maintain a sustained arrest at this stage (Bunz et al. 1998). A ll genes studied that are related to DNA damage checkpoints or repair (RBBP8, RAD1, RAD17, BRCA1, KP NA2 PCNA and BCCIP) were down regulated either at 12 or 24 h after OvUS treatment. The products of these genes are induced in response to incomplete DNA replication or damage at specific or several points of the cell cycle (Waga et al. 1994, Sancar et al 2004, Sartori et al. 2007 Teng et al. 2003 Lu et al. 2005). Thus, OvUS disrupts the transcription of some genes involved in nucleotide excision, mismatch, homologous recombination repair and translesion synthesis pathways in addition to the sensor (RA D1 and 17) and mediator molecules (BRCA1) of the DNA damage checkpoint. The failure of OvUS to induce apoptosis in PC 3 cells (Tekin et al. 2005 b Padua & Hansen 2008) despite these changes in gene expression, could reflect the lack of p53 or an as yet un described anti apoptotic action of OvUS. Ovine US is one of the few serpins identified that alters cell cycle dynamics. The other is the intracellular protein MENT which also has a very basic isoelectric point (9 v ersus 5 6.5 for other serpins) (Silverman et al. 2001) MENT inhibits the enzymatic activity of the nuclear cysteine proteinase SPase, a cathepsin L like proteinase involved in the degradation of the phosphorylated form of the retinoblastoma (Rb) protein, a known regulator of the cell cycle (Irv ing et al. 2002 a ). Another intracellular serpin, PAI 2, can protect Rb from degradation by an independent anti proteinase mecha nism (Croucher et al. 2008) Unlike MENT and PAI 2, OvUS
133 is an extracellular protein that can bind to ce ll membranes (Liu et al 1999). Nonetheless, it is possible that OvUS inhibits cell proliferation by being internalized. Other extracellular serpins can become internalized antitrypsin, which enters and resides in the cytoplasm of a mouse insulinoma cell line and protects against apoptosis through inhibition of caspase 3 activation (Zhang et al. 2007). Alternatively, OvUS blocks cel l proliferation through blockage or induction of signal transduction systems. In lymphocyte s OvUS inhibits proliferation of phorbol myristol acetate stimulated lymphocytes suggesting that the protein blocks downstream actions of the protein kinase ( PK ) C pathway (Peltier et al. 2000 b ). Like OvUS, transforming growth factor (TGF) causes cell cycle arrest by upregulation of CDKN2B (p15 ink ), CDKN1A (p21 cip1 ) and CCNG2 (cyclin G2) (Horne et al. 1997, Gartel & d Tyner 2002). Interferon (IFN) CDKN1A (p21 cip1 ) expression, independent of p 53 by the S TAT 1 pathway (Gartel & Tyner 2002). Both, TGF p53 independent mechanism, where p21 protects cells against p53 independent apoptosis which is induced by these signals (Gartel & Tyner 2 002) Signal transduction pathways affected by OvUS have not been determined, but OvUS may regulate components of signaling pathways shared with TGF or IFN In summary OvUS inhibits proliferation of PC 3 cells through disruption of the cell cycle dynamics. Disruption involves increased expression of cell cycle checkpoint and arrest genes CDKN1A (p21 cip1 ), CDKN2B (p15 ink ) and CCNG2 (cyclin G2) and down regulation of genes involved in cell cycle progression.
134 CHAPTER 6 GENERAL DISCUSSION T he serpin gene s are an ancient origin gene family, being identified in bacteria, fungus, nematodes, archaea and virus (Irving et al. 2000, 2002 b Steenbakkers et al. 2008). E vidence presented in this dissertation strongly indicate s that one of the subcl ades in this superfamily, the uterine serpins, evolved only within a restricted group of mammals the Laurasiatheria super order of eutherian mammals. The uterine serpins are thus an example of a new gene arising from gene duplication and selection for seq uence divergence (L ouis 2007) that plays a particular role in pregnancy in a group of mammals with unique attributes for gestation (an epitheliochorial placenta) The distribution of known uterine serpins as well as orders where a uteri ne serpin was not id entified is shown in Figure 2 9 All the identified uterine serpins are in Ruminatia (sheep, goat, water buffalo and cow), Suidae (pig), Pe risodactyla (horse) or Carn ivora (dog) orders of Laurasiatheria. An important question is whether uterine serpin ge nes are in the orders within the Laurasiatheria superorder (Cetacea, Hippopotamidae, Pholidota, Chiroptera, Erinaceomorpha, Scalopus and Talpa). Most of these orders have epitheliochorial placenta although the Chiroptera, Erinaceomorpha and Talpa orders h ave either endotheliochorial or hemochorial type of placentation. Experiments in Chapter 2 strongly suggested that the uterine serpin gene is modified (dog) or become a pseudogene (cat) in carnivore species with endotheliochorial placenta. Examination of the situation in species of the remaining orders of the Laurasiatheria superorder with either endotheliochorial or hemochorial placenta would allow determination of whether similar changes in uterine serpin genes are occurring within these orders. If the gene is either lost or modified in species of the Chiroptera, Erinaceomorpha and
135 Talpa orders, the idea that uterine serpins play a specific role required for pregnancy in species with epitheliochorial type of placentation would be strengthened. As indica ted in Figure 2 9 there is an additional order of animals having epitheliochorial placenta that is not within Laurasiatheria, the lemurs and lorises of the order Lemuriformes within the Euarchontoglires superorder. Identification of the uterine serpin ge ne in this species would indicate that the genes predate the diversion of Laurasiatheria from Euarchontoglires superorder and that epitheliochorial placentation is absolutely dependent upon the uterine serpin genes. Peltier et al ( 2000 c ) estimated that u terine serpin genes diverged from CBG before the divergence of mammals so much a result remains possible. It is more likely, however, that uterine serpins will not be found in Lemuriformes since, even if the uterine serpin gene is older than the common an cestor of Laurasiatheria and Euarchontoglires the gene would be likely be lost in the ancestors of lemurs and lorises that did not have epitheliochorial placenta. The results from the positive (Darwinian) selection study reported in Chapter 2 strongly su ggest that there is pressure on the evolution of the uterine serpin gene to affect its functionality. Positive selection has occurred at the RCL region of the inhibitory serpins in particular at the P1 site proximal to the scissile bond of the RCL which changes the selectivity for the target proteinase (Brown 1987, Hill & Hastie 1987, Goodwin et al. 1996 Zang & Maizels 2001 Barbour et al. 2002 ). T he uterine serpins apparently lack anti proteinase activity (Ing & Roberts 1989, Liu & Hansen 1995, Mathia lagan & Hansen 1996, Peltier et al. 2000 a ; see also Chapter 2 for lack of conservation at residues important for anti proteinase activity ) There is also evidence that the uterine serpin gene has evolved different functions within mammals with the gene. I t is possible that the uterine serpin gene in species with epitheliochorial placentation has been retained as a secretory protein while the gene in species with endotheliochorial
136 placentation has either been changed to become an intracellular protein (dog ) or to be lost (cat). If this idea is correct, uterine serpins play an important role in the uterine lumen of species with epitheliochorial placentation that is not required for species with endotheliochorial placentae. Even within species with epitheli ochorial placenta, there might be different functions. The PoUS 1 and 2, for example, regulate iron stability in uteroferrin and may be involved in iron transport to the fetus (Baumbach et al 1989, Roberts & Bazer 1988) while the OvUS can inhibit prolif eration of lymphocytes and certain other cell types ( Tekin et al 2005 b Padua & Hansen 2008; also see Chapters 3, 4 and 5). Inhibition of lymphocyte proliferation has been interpreted as signifying a role for uterine serpins in prevention of rejection o f conceptus tissue by the maternal immune system ( Hansen 1998 ) The sheep protein OvUS for example, has been shown to inhibit proliferation of mitogen stimulated lymphocytes, NK cell activity a nd a variety of cancer cells (Segerson et al. 1984, Hansen et al. 1987 b Stephenson et al. 1989 b Skopets & Hansen 1993, Skopets et al. 1995 Padua & Hansen 2008). In addition, the limited invasiveness of trophoblast tissue in species with epitheliochorial placentatation could be achieved, at least in part, by inhib ition of trophoblast proliferation by uterine serpin It may also be that uterine serpins participate in the process of trophoblast binucleate cell formation, a phenomenon characterized by nuclear endoreduplication without proper cytokinesis, wh ich occurs in sheep goat, cattle and horse (Hoffman & Wooding 1993) Ovine uterine serpin acts to inhibit cell proliferation by inducing changes in gene expression of proteins involved in the cell cycle and, most notably, the up regulation of cell cycle inhibitor s such as CDKN1A (p21 cip ), CDKN2B (p15 ink ) and CCNG2 (cyclin G2) (Chapter 5). These cell cycle inhibitors have been related to differentiation of stromal cells into decidual
137 cells formation of polyploidy stromal cells as well as terminal differentiation and apoptosis of luminal and stromal cells during decidualization (Tan et al 2002, Yao et al 2003, Yue et al 2005, Li et al 2008) Ovine US is one of the few serpins identified that inhibits cell proliferation by blocking cell cycle progression (Chapte r 4) In PC 3 cells, OvUS decreased the percent of cells in S phase and increased the percent of cells in the G 2 /M phase at 12 h after treatment addition. At 24 h, OvUS increased the percent of cells in G 0 /G 1 and decreased the percent of cells in S phase In the mitogen (PHA) stimulated lymphocyte, a t both 72 and 96 h after stimulation, OvUS increased the proportion of lymphocytes in the G 0 /G 1 phase and decreased the proportion of cells in the S phase (Chapter 4) There are two other serpins that can al ter cell cycle progression. One is the intracellular protein MENT which also inhibits cell proliferation by blocking cell cycle progression, inhibit ing the enzymatic activity of the nuclear cysteine proteinase SPase, a cathepsin L like proteinase involved in the degradation of the phosphoryla ted form of the Rb protein, a kn own regulator of the cell cycle (Irving et al. 2002 a ). The other is the intrace llular serpin, PAI 2 that can protect Rb from degradation by an independent anti proteinase mechanism (Cro ucher et al. 2008). The signal transduction pathway activated or down regulated by OvUS remains to be determined. Unlike MENT and PAI 2, OvUS is an extracellular protein that can bind to cell membranes (Liu et al. 1999). Ovine US could either regulate s ignal transduction pathways by binding to a specific membrane receptor or by becoming internalized into the cell and interacting with intracellular binding partners (Figure 6 1 ) It is possible that OvUS inhibits cell proliferation by being internalized t hrough endocytosis. Other extracellular serpins can become antitrypsin, which enters and resides in
138 Figure 6 1 Possible pathways by which OvUS could block cell proliferati on. A) OvUS pathway that blocks cell cycle progression. B) OvUS could block the binding site of a ligand needed for cell cycle progression so that the pathway for activation is n ot stimulated. C) OvUS could become internalized and block cell cycle progression after internalization by binding too and disrupting intracellular binding partners that are the effectors of the cell cycle arrest. Yellow circles represent the phosphorylat ed status of the proteins in the signal transduction pathway.
139 the cytoplasm of a mouse insulinoma cell line and protects against apoptosis through inhibition of caspase 3 activation (Zhang et al 2007). Alternatively, OvUS could bind to a specific cell s urface receptor or compete with other molecules for a receptor binding site (Figure 6 1 ), like antitrypsin which competes with diferric transferrin for the trasnferrin receptors inhibiting the proliferation of human skin fibroblasts (Graziadei et al. 1988). In lymphocytes, OvUS inhibits proliferation of phorbol myristol acetate stimulate d lymphocytes, suggesting that the protein blocks downstream actions of the PKC p athway (Peltier et al. 2000 b ). OvUS could regulate components of signaling pathways share d with TGF Like Ov US, TGF causes cell cycle arrest by upregul ation of CDKN2B (p15 ink ), CDKN1A (p21 cip1 ) and CCNG2 (cyclin G2) (Horne et al. 1997, Gartel & Tyner 2002). Interferon (p21 cip1 ) expression by the S TAT 1 pathway (Gartel & Tyner 2002). Thus, it is possible that uterine serpins are in volved i n the process of tropho blast binucleate cell formation and/or uterine luminal and stromal cell differentiation. In the dog ( Chapter 2) as well as in the cow (Khatib et al 2007), the uterine serpin gene was recently identified to be expressed in th e ovary. The presence of uterine serpin in other reproductive tissues distinct to the uterus raises the possibility that perhaps uterine serpins may have different functions according to the cell type and stage of differentiation. One possibility is that uterine serpin functions in the ovary to regulate follicular growth by binding to activin. Activin improves follicular growth and granulosa cell proliferation (Knight & Glister 2006). OvUS can also this member of the TGF (McFarlane et al. 1999) However, the binding affinity of activin for OvUS is much lower than acti vin affinity towards fo l listatin. I t is also possible that OvUS is participating in the differentiation of granulose cells into luteal cells after ovulation. In the ovary, granul osa cells luteinize under the influence of
140 luteinizing hormone The expression of cell cycle inhibitors has been involved in the inhibition of proliferati on of differentiating granulosa cells in mice, specifically the CDK inhibitors p21 Cip1/Waf1 and p27 Ki p1 which are up regulated throughout this period (Burns et al. 2001, Jirawatnotai et al. 2003). Likewise, p15 i nk4b another CDK inhibitor has been linked with luteinization and granulosa cell differentiation (Burns et al. 2001). Results presented in this dissertation showed that OvUS blocked cell cycle progression and up regulat ed the expression of the CDK inhibitors CDKN1A (p21 cip1 ) and CDKN2B (p15 ink ) and also caused the down regulation of MKI67 (Chapter 5) The gene product of MKI67 (Ki 67) is a marke r of cell proliferation and is present in all stages of the cell cycle with the exception of the G 0 stage (Gerdes et al. 1984 ) In summary, t he uterine serpin gene is present only in a restricted group of species within the Laurasiatheria superorder of eut herian mammals and likely evolved under positive selection which suggests diversifying functionality of the protein within these species. The finding that the uterine serpin has been retained as a secretory protein in most species with epitheliochorial pl acentation within the Laurasiatheria superorder also suggests that the protein has an important role during pregnancy in species with this kind of placenta. The finding that OvUS inhibits cell proliferation by blocking cell cycle progression, specifically by the upregualtion of cell cycle inhibitors suggests other possible functions for uterine serpins in reproductive tissue, in particular trophoblast binucleate cell formation, uterine cell differentiation and granulosa cell differentiation in the ovary.
141 LIST OF REFERENCES Anis imova M, Bielawski JP & Yang Z 2001 Accuracy and power of likelihood ratio test to detect positive selection at amino acid sites Mol Biol Evol 18 1585 1592. Araki S, Omori Y, Lyn D, Singh RK, Meinba ch DM, Sandman Y, Lokeshwar, VB & Lokeshwar BL 2007 Interleukin 8 is a molecular determinant on androgen independence and progression in prostate cancer cells. Cancer Res 67 6854 6862. Atchley WR, Lokot T, Wollenberg K, Dress A & Ragg H 2001 Phylogenetic analyses of amin o acid variation in the serpin proteins. Mol Biol Evol 18 1502 1511. Aylon Y, Liefshitz B & Kupiec M 2004 The CDK regulates repair of double strand breaks by homologous recombination during the cell cycle. EMBO J 23 4868 4875. B aker DJ, Dawlaty MM, Gal ardy P & van Deursen JM 2007 Mitotic regulation of the anaphase promoting complex. Cell Mol Life Sci 64 589 600. Bates S, Rowan S & Vousden KH 1996 Characterization of human cyclin G1 and G2: DNA damage inducible genes. Oncogene 13 1103 1109. Bauersach s S, Ulbrich SE, Gross K, Schmidt SEM, Meyer HHD, Einspanier R, Wenigerkind H, V ermehren M, Blum H, Sinowatz F & Wolf E 2005 Gene expression profiling of bovine endometrium during the oestrus cycle: detection of molecular pathways involved in functional c h anges. J Mol Endocrinol 34 889 908. Barbour KW, Goodwin RL, Gu illonneau F, Wang Y, Baumann H & Berger FG 2002 proteinase inhibitor family: role of hypervariable reactive center loop. Mol Biol Evol 19 718 727. Baumbach GA, Ketcham CM, Richardson DE Bazer FW & Roberts RM 1986 Isolation and characteriza tion of a high molecular weight stable pink form of uteroferrin from uterine secretions and allantoic fluid of pigs. J Biol Chem 261 12869 12878. Baumbach GA, Sand ers PTK, Bazer FW & Roberts RM 1984 Uteroferrin has N asparagine linked high mannose type o ligosaccharides that contain mannose 6 phosphate. Proc Natl Acad Sci USA 81 2985 2989. Bazer FW, Roberts RM, Basha SM, Zavy MT, Caton D & Barron DH 1979 Method for obtaining ovine uterine secretions from unilaterally pregnant ewes. J Anim Sci 49 1522 15 27. Benarafa C & Remold E 2005 The ovalbumin serpins revisited: perspective from the chicken genome of clade B serpin evolution in vertebrates. Proc Natl Acad Sci USA 102 11367 11372.
142 Bendtsen JD, Nielsen H, von Heijne G & Brunak S 2004 Impro ved prediction of signal peptides: SignalP 3.0. J Mol Biol 340 783 795. Bennin DA, Don ASA, Brake T, McKenzie L, Rosenba um H, Ortiz L, DePaoli Roach AA & Horne MC 2002 Cyclin G2 associates with protein phosphatase 2A catalytic and s in active complexes and induces nuclear aberrations and a G1/S phase cell cycle arrest. J Biol Chem 277 27449 27467. Bielawski JP & Yang Z 200 3 Maximum likelihood methods for detecting adaptive evolution after gene duplication. J Struct Funct Genomics 3 201 212. Binder BR, Mihaly J & Prager GW 2007 uPAR uPA PAI 1 interactions and signaling: a Thromb Haemost 97 336 342. Bradford MM 1976 A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein dye binding. Anal Biochem 78 : 248 254. Branzei D & Foiani M 2008 Regulation of DNA repair throughout the cell cycle. Nat Rev Mol Cell Biol 9 297 308. Breit SN, Luckhurst E & Penny R 1983 antitrypsin o n the proliferative response of human peripheral blood lymphocytes. J Immunol 130 681 686. Brooks MA, Ali AN, Turner PC & Moyer RW 1995 A rabbitpox virus serpin gene controls host range by inhibiting apoptosis in restrictive cells. J Virol 69 7688 7698. Brown AL 1987 Positively darwinian molecules? Nature 326 12 13. Buhi WC, Ducsay CA, Bazer FW & Roberts RM 1982 Iron transfer between the purple phosphatase uteroferrin and transferring and its possible role in iron metabolism of the fetal pig. J Biol Chem 257 1712 1723. Bunz F, Dutriaux A, Lengauer C, Waldman T, Zhou S, B rown JP, Sedivy JM, Kinzler KW & Vogelstein B 1998 Requirement for p53 and p21 to sustain G 2 arrest after DNA damage. Science 282 1497 1501. Burns KH, Yan C, Kumar TR & Matzuk MM 20 01 Analysis of ovarian gene expression in follicle Endocrinology 142 2742 2751 Cayrol C, Knibieh ler M & Ducommun B 1998 p21 binding to PCNA causes G1 and G2 cell cycle arrest in p53 deficient cells. Oncogene 16 311 3 20 Callus BA & Vaux DL 2007 Caspase inhibitors: viral, cellular and chemical. Cell Death Differ 14 73 78.
143 Chen TT, Bazer FW, Geb hardt BM & Roberts RM 1975 Uterine secretion in mammals: synthesis and placental transport of a purple acid phosphatase in p igs. Biol Reprod 13 304 313. Chenna R, Sugawara H, Koike T, Lopez R, Gibson TJ, Higgins DG & Thompson JD 2003 Multiple sequence alignment with the Clustal series of programs. Nucleic Acids Res 31 3497 3500. Cole EB, Miller D, Romero D, Greenberg RM, Br mme D, atalpete S, Pak SC, Mills DR, Silverman GA & Luke CJ 2004 Identification and activity of a lower eukaryotic serine proteinase inhibitor (serpin) from Cyanea capillata : analysis of a jellyfish s erpin, jellypin. Biochem 43 11750 11759. Congote LF 2007 Serpin A1 and CD91 as host instruments against HIV 1 infection: are extracellular antiviral peptides acting as intracellular messengers Virus Res 125 119 34. Crisp RJ, Knauer DJ & Knauer MF 2000 Roles of the heparin and low density lipid receptor related protein binding sites of protease nexin 1 (PN1) in urokinase PN1 complex catabolism. J Biol Chem 275 19628 19637. Cr oucher D, Saunders DN, Lobov S & Ranson M 2008 Revisiting the biological roles of PAI2 (SERPINB2) in cancer. Nature Rev Cancer 8 535 545. Croucher D, Saunders DN & Ranson M 2006 The urokinase/PAI 2 complex. J Biol Chem 281 10206 10213. Ducsay CA Buhi WC, Bazer FW & Roberts RM 1982 Role of uteroferrin in iron transport and macromolecular uptake by allantoic epithelium of the porcine conceptus. Biol Reprod 26 729 743. Ellery JM & Nicholls PJ 2002 Alternate signaling pathways from the interleukin 2 receptor. Cytokine Growth Factor Rev 13 27 40. Felsenstein J 1985 Confidence limits on phylogenies: A n approach using the boot strap. Evolution 39 783 791. Felsenstein J 1993 PHYLIP (Phylogeny Inference Package) version 3.5c. Distributed by the author. Department of Genetics, University of Washington Seattle Felsenstein, J 1989 PHYLIP -Phylogeny Inference Package (Version 3.2). Cladistics 5 164 166. F ernandez Garcia NI, Volpert OV & Jimenez B 2007 Pigment epithelium derived factor as a multifunctional antitumor factor. J Mol Med 85 15 22.
144 Gartel AL & Tyner AL 2002 The r ole of the cyclin dependent kinase i nhibitor p21 in a poptosis Mol Cancer Ther 1 639 649. Gatz SA & Wiesmller L 2006 p53 in recombination and repair. Cell Death Differ 13 1003 1016. Gerdes J, Lemke H, Baisch H, Wacker HH, Schwab U & Stein H 1984 C ell cycle analysis of a cell proliferation associated hu man nuclear antigen defined by the monoclonal antibody Ki 67 J Immunol 133 1710 1715. Goodwin RL, Baumann H & Berger FG 1996 Patterns proteinase inhibitors in mammals. Mol Biol Evol 13 346 358. Gray CA, Stewart MD, Johnson GA & Spencer TE 2003 Postpartum uterine involution in sheep in endometrial gene expression. Reproduction 125 185 198. Graziad ei I, Khler CM, Wiedermann CJ & Vogel W 1998 The acute antitrypsin inhibits trasferrin receptor bindin g and proliferation of human skin fibroblasts. Biochim Biophys Acta 1401 1 7 0 1 76. Green JA, Xie S & Roberts RM 1998 Pepsin related molecules secreted by trophoblast. Rev Reprod 3 62 69. Gupta VK & Gowda LR 2008 Alpha 1 proteinase inhibitor is a heparin binding serpin; molecular interactions with Lys rich cluster of helix F domain. Biochimie 90 749 761. Handayani R, Rice L, Cui Y, Medrano TA, Samedi VG, Baker HV, Szabo NJ & Shiverick KT 2006 Soy isoflavones alter expression of genes associated with can cer progression, including interleukin 8, in androgen independent PC 3 human prostate cancer cells. J Nutr 136 75 82. Hansen PJ 1998 Regulation of uterine immune function by progesterone lessons from the sheep. J Reprod Immunol 40 63 79. Hansen PJ & Ne wton GR 1988 Binding of immunoglobulins to the major progesterone induced proteins secreted by the sheep uterus. Arch Biochem Biophys 260 208 217. Hansen PJ, Ing NH, Moffatt RJ, Baumb ach GA, Saunders PTK, Bazer FW & Roberts RM 1987 a Biochemical charact erization and biosynthesis of the uterine milk proteins of the pregnant sheep uterus. Biol Reprod 36 405 418. Hansen PJ, Segerson EC & Bazer FW 1987 b Characterization of immunosuppressive substances in the basic protein fraction of uterine secretions of pregnant ewes. Biol Reprod 36 393 403.
145 Harper JW, Adami GR, Wei N, Keyomarsi K & Elledge SJ 1993 The p21 Cdk interacting protein Cip1 is a potent inhibitor of G1 cyclin dependent kinases. Cell 75 805 816. Hida K, Wada J, Eguchi J, Zhang H, Baba M, Sei da A, Hashimoto I, Okada T, Yasuhara A, Nakatsuka A, Shikata K, Hourai S, Futami J, Watanabe E, Matsuki Y, Hiramatsu R, Akagi S, Makino H & Kanwar YS 2005 Visceral adipose tissue derived serine protease inhibitor: a unique insulin sensitizing adipocytoki ne in obesity. Proc Natl Acad Sci USA 102 10610 10615. Hill RE & Hastie ND 1987 Accelerated evolution in the reactive centre regions of serine protease inhibitors. Nature 326 96 99. Hoffman LH & Wooding FB 1993 Giant and binucleate trophoblast cells o f mammals J Exp Zool. 266 559 5 77. Horne MC, Donaldson KL, Goolsby GL Tran D, Mulheisen M, Hell WJ & Wahl AF 1997 Cyclin G2 is up regulated during growth inhibition and B cell antigen receptor mediated cell cycle arrest. J Biol Chem 272 12650 12661. Hungtinton JA 2006 Shape shifting serpins advantages of a mobile mechanism. Trends Biochem Sci 31 427 435. Im H, Woo MS, Hwang KY & Yu MH 2002 Interactions causing the kinetic trap in serpin protein folding. J Biol Chem 48 46347 46354. Inagi R, Nang aku M, Useda N, Shimizu A, Onogi H, Izuhara Y, Nakazato K, Ueda Y, Oishi H, Takahashi S, Yamamoto M, Susuki D, Kurokaw a K, van Ypersele de Strihou C & Miyata T 2005 Novel serpinopathy in rat kidney and pancreas induced by overexpression of Megsin. J Am S oc Nephrol 16 1339 1349. Ing NH & Roberts RM 1989 The major progesterone induced proteins secreted into the sheep uterus are members of the serpin superfamily of protease inhibitors. J Biol Chem 264 3372 3379. Ing NH, Fr ancis H, McDonnell JJ, Amann J F & Roberts RM 1989 Progesterone induction of the uterine milk proteins: major secretory proteins of sheep endometrium. Biol Reprod 41 643 654. Irving JA, Pike RN, Lesk AM & Whisstock JC 2000 Phylogeny of the serpin superfamily: Implications of patterns of amino acid conservation for structure and function. Genome Res 10 1845 1864. Irving JA, Shushanov SS, Pike RN, Popova EY, Bromme D, Coetzer THT, Bottomley SP, Boulynk o IA, Grigoryev SA & Whisstock JC 2002 a Inhibitory activity of a heterochromatin asso ciated serpin (MENT) against papain like cysteine proteinases affects chromatin structure and blocks cell proliferation. J Biol Chem 277 13192 13201.
146 Irving JA, Steenbakkers PJM, Les k AM, Op den Camp HJM, Pike RN & Whisstock JC 2002 b Serpins in Prokaryot es Mol Biol Evol 19 1881 1890. Isaacs WB, Carter BS & Erwing CM 1991 Wild type p53 suppresses growth oh human prostate cancer cells containing mutant p53 alleles Cancer Res 51 4716 4720. Ivanov D, Emonet C, Foata F, Affolter M, Delley M, Fisseha M, Blu m Sperisen S, Kochhar S & Arigoni F 2006 A serpin from the gut bacterium Bifidobacterium longum inhibits eukaryotic elastase like serine proteases. J Biol Chem 281 17246 17252. Izuhara K, Ohta S, Kanaji S, Shiraishi H & Arima K 2008 Recent progress in und erstanding the diversity of human ov serpin/clade B serpin family. Cell Mol Life Sci 65 2541 2553. Janciauskiene S, Larsson S, Larsson P, Virtala R, Jansson L & Stevens T 2004 Inhibition of lipopolysaccharide mediated human monocyte activation, in vitro, antitrypsin. Biochem Biophys Res Commun 321 592 600. Jin L, Abrahams JP, Skinner R, Petitou M, Pike RN & Carrell RW 1997 The anticoagulant activation of antithrombin by heparin. Proc Natl Acad Sci USA 94 14683 14688. Jirawatnotai S, Moons DS, S tocco CO, Franks R, Hales DB, Gibori G & Kiyokawa K 2003 The cyclin dependent kinase inhibitors p27 kip1 and p21 cip1 cooperate to restrict proliferative life span in differentiating ovarian cells. J Biol Chem 2 7 8 17021 17027. Jones DT, Taylor WR & Thornto n JM 1992 The rapid generation of mutation data m atrices from protein sequences. Comput Appl Biosci 8 275 82. Kaiserman Dion, Whisstock JC & Bird PI 2006 Mechanisms of serpin dysfunction in disease. Expert Rev Mol Med 8 1 19. Kaldis P 1999 The cdk acti vating kinase (CAK): from yeast to mammals. Cell Mol Life Sci 55 284 296. Kang S, Barak Y, Lamed R, Bayer EA & Morrison Mark 2006 The functional repertoire of prokaryote cellulosomes includes the serpin superfamily of serine proteinase inhibitors. Mol M icrobiol 60 1344 1354. K hatib H, Schutzkus V, Chang YM & Rosa GJ 2006 Pattern of expression of the uterine milk protein gene and its association with productive life in cattle. J Dairy Sci 90 2427 2433. Knauer MF, Kridel SJ, Hawley SB & Knauer DJ 1997 The efficient catabolism of thrombin protease nexin 1 complexes is a synergistic mechanism that requires both the LDL receptor related protein and cell surface heparins. J Biol Chem 272 29039 29045.
147 Knauer DJ, Majumdar D, Fong PC & Knauer MF 2000 SERPIN regulation of factor XIa J Biol Chem 275 37340 37346. Knight PG & Glister C 2006 TGF Reproduction 132 191 206 Knox K & Baker JC 2008 Genomic evolution of the placenta using co option and duplication and divergence Genome Res 18 695 705. Koun nas MZ, Church FC, Argraves WS & Strick land DK 1996 Cellular internalization and degradation of antithrombin III thrombin, heparin cofactor II antitrypsin trypsin complexes is mediated by the low density lipoprotein receptor related protein. J Biol Chem 271 6523 6529. Krem MM & Di Cera E 2003 Conserved Ser residues, the shutter region, and speciation in serpin evolution. J Biol Chem 278 37810 37814. Krug U, Ganser A & Koeffler HP 1996 Tumor suppressor genes in normal and malignant hematopoiesis Oncogene 13 1103 1109. Law R HP, Zhang Q, McGowan S, Buckle AM, Silverman GA, Wong W, Rosado CJ, L angendorf CG, Pike RN, Bird PI & Whisstock JC 2006 An overview of the serpin superfamily. Genome Biol 7 216. Le XF, Arachchige Don AS, Mao W, Horne MC & Bast RC Jr 2007 Roles of human e pidermal growth factor receptor 2, c jun NH2 terminal kinase, phosphoinositide 3 kinase, and p70 S6 kinase pathways in regulation of cyclin G2 expression in human breast cancer cells Mol Cancer Ther 6 2847 2857. Leiser R, Krebs C, Ebert B & Dantzer V 19 97 Placental vascular corrosion cast studies: a comparison between ruminants and humans Microsc Res Tech 38 76 87. Leslie MV & Hansen PJ 1991 Progesterone regulated secretion of the serpin like proteins of the ovine and bovine uterus. Steroids 56 589 5 97. Leslie MV, Hansen PJ & Newton GR 1990 Uterine secretions of the cow contain proteins that are immunochemically related to the major progesterone induced proteins of the sheep uterus. Domest Anim Endocrinol 7 517 526. Lewis EC, Shapiro L, Bowers OJ & Dinarello CA 2005 antitrypsin monotherapy prolongs islet allograft survival in mice. Proc Natl Acad Sci USA 102 12153 121 58. Li F, Devi S Bao L, Mao J & Gibori G 2008 Involvment of cyclin D3, CDKN1A (p21), and BIRC5 (survivin) in interleukin 11 stimulation of deciduali zation in mice. Biol Reprod 78 1 27 1 33.
148 Li R, Waga S, Hannon GJ, Beach D & Stillman B 1994 Differential effects by the p21 CDK inhibitor on PCNA dependent replication and repair. Nature 371 534 537. Lillis AP, van Duyn LB, Murphy Ullrich & Strickland DK 2008 LDL Receptor related protein 1: unique tissue specific functions revealed by selective gene knockout studies. Physiol Rev 88 887 918. Liu WJ & Hansen PJ 1993 Effect of the progesterone induced serpin like proteins of the sheep endometrium on natu ral killer cell activity in sheep and mice. Biol Reprod 49 1008 1014. Liu WJ & Hansen PJ 1995 Progesterone induced secretion of dipeptidyl peptidase IV (cell differentiation antigen 26) by the uterine endometrium of the ewe and cow that costimulates ly mphocyte p roliferation. Endocrinology 136 779 787. Liu WJ, Peltier MR & Hansen PJ 1999 Binding of ovine uterine serpin to lymphocytes. Am J Reprod Immunol 41 428 432. Louis EJ 2007 Making the most of redundancy. Nature 449 673 674. Lu H, Guo X, Meng X Liu J, Allen C, Wray J, Nickoloff JA & Shen Z 2005 The B RCA2 interacting protein BCCIP f unctions in RAD51 and BRCA2 f ocus formation and h omologous recombinational r epa ir Mol Cell Biol 25 1949 19 57. Lucas A & McFadden 2004 Secreted immunomodulatory vira l proteins as novel biotherapeutics. J Immunol 173 4765 4774. Lhr K, Mritz C, Contente A & Dobbelstein M 2003 p21/CDKN1A mediates negative regulation of transcription by p53. J Biol Chem 278 32507 32516. Makarova A, Mikhailenko I, Bugge TH, List K, Lawrence DA & Strickland DK 2003 The low density lipoprotein receptor related protein modulates protease activity in the brain by mediating the cellular internalization of both neuroserpin and neuroserpin tissue type plasminogen activator complexes J Bio l Chem 278 50250 50258. Ma lathy PV, Imakawa K, Simmen RC & Roberts RM 1990 Molecular cloning of the uteroferrin associated protein, a major progesterone induced serpin secreted by the porcine uterus, and expression of its mRNA during pregnancy. Mol Endoc rinol 4 428 440. Martinsson Ahlzn H, Liberal V, Gr nen felder B, Chaves SR, Spruck CH & Reed SI 2008 Cyclin dependent kinase associated proteins cks1 and cks2 are essential during early embryogenesis and for cell cycle progression in somatic cells. Mol C ell Biol 28 5698 5709. Maston GA & Ruvolo M 2002 Chorionic gonadotropin has a recent origin within primates and an evolutionary history of selection. Mol Biol Evol 19 320 335.
149 Mathialagan N & Hansen TR 1996 Pepsin inhibitory activity of uterine serpins. Proc Natl Acad Sci USA 93 13653 13658. illips DJ, Jenkin G, Hearn MTW & Kretser DM 1999 Uterine milk protein, a novel activin binding protein, is present in ovine allantoic fluid Endocrinol ogy 140 4745 4752. Me dcalf RL & Stasinopoulos SJ 2005 The undecided serpin. The ins and outs of plaminogen activator type 2. FEBS J 272 4858 4867. Mess A & Carter AM 2007 Evolution of the placenta during the early radiation of placental mammals. Comp Biochem Physiol A Mol Integr Physiol 148 769 779. Miyata T, Inagi R, Sugiyama S & Usuda N 2005 Serpinopathy and endoplasmic reticulum stress. Med Mol Morphol 38 73 78. Moffatt J Bazer FW, Hansen PJ, Chun PW & Roberts RM 1987 Purification, secretion and immunocytochemical l ocalization of the uterine milk proteins, major progesterone induced proteins in uterine secretions of th e sheep. Biol Reprod 36 419 430. Moffett A & Loke C 2006 Immunology of placentation in eutherian mammals. Nat Rev Immunol 6 584 594. Morgan L, Brou ghton Pipkin F & Kalsheker N 1996 Angiotensinogen: molecular biology, biochemistry and physiology. Int J Biochem Cell Biol 28 1211 1222. Muhl L, Nykjaer A, Wygr ecka M, Monard D, Preissner KT & Kanse SM 2007 Inhibition of PDGF BB by factor VII activating protease (FSAP) is neutralized by protease nexin 1, and the FSAP inhibitor complexes are internalized via LRP. Biochem J 404 191 196. Murphy WJ, Pringle TH, Cri der TA, Springer MS & Miller W 2007 Using genomic data to unravel the root of mammal phylogen y. Genome Res 17 413 421. Nagata K 1998 Expression and function of heat shock protein 47: a collagen specific molecular chaperone in the endoplasmic reticulum. Matrix Biol 16 379 386. Newton GR, Hansen PJ, Baze r FW, Leslie MV, Stephenson DC & Low BG 1 989 Presence of major progesterone induced proteins of the sheep endometrium in fetal fluids Biol Reprod 40 417 424. Nishihara H, Hasewaga M & Okada N 2006 Pegasoferae, an unexpected mammalian clade revealed by tracking ancient retroposon insertions. P roc Natl Acad Sci USA 103 9929 9934. Noel S, Herman A, Johnson GA, Gray CA, S tewart MD, Bazer FW, Gertler A & Spencer TE 2003 Ovine placental lactogen specifically binds to endometrial glands to the ovine uterus. Biol Reprod 68 772 780.
150 Nuttleman PR & R oberts RM 1990 Transfer of iron from uteroferrin (purple acid phosphatase) to transferring related to acid phosphatase activity J Biol Chem 265 12192 12199. Oliveira L & Hansen PJ 2008 Deviations in populations of peripheral blood mononuclear cells and endometrial macrophages in the cow during pregnancy. Reproduction 136 481 490. Olson D, Pllnen J, Hyer Hansen G, Rnne E, Sakaguchi K, Wun TC, Appella E, Dan K & Blasi F 1992 Int e r nalization of the urokinase plasminogen activator inhibitor type 1 comp lex is mediated by the urokinase receptor J Biol Chem 267 9129 9133. Padua MB & Hansen PJ 2008 Regulation of DNA synthesis and the cell cycle in human prostate cancer cells and lymphocytes by ovine uterine serpin BMC Cell Biol 9 5. Padua MB, Tekin S, Spencer TE & Han sen PJ 2005 Actions of progesterone on uterine immunosuppression and endometrial gland development in the uterine gland knockout (UGKO) ewe. Mol Reprod Dev 71 347 357 Peltier MR, Grant TR & Hansen PJ 2000 a Distinct physical and struc tural properties of the ovine uterine serpin. Biochim Biophys Acta 1479 37 51. Peltier MR, Liu WJ & Hansen PJ 2000 b Regulation of lymphocyte proliferation by uterine serpin: interleukin 2 mRNA production, CD25 expression and responsiveness to interleukin 2. Proc Soc Exp Biol Med 223 75 81. Peltier MR, Ra ley LC, Liberles DA, Benner SA & Hansen PJ 2000 c Evolutionary history of the uterine serpins. J Exp Zool 288 165 174. Pemberton PA Stein PE, Pepys MB, Potter JM & Carrell RW 1988 Hormone binding gl obulins undergo serpin conformational change in inflammation. Nature 336 257 258. Poller W, Willnow TE, Hilpert J & Herz J 1995 antitrypsin antichymotrypsin cathepsin G complexes by the low density lipoprotein receptor related protein. J Biol Chem 270 2841 2845. Potempa J, Korzus E & Travis J 1994 The serpin super family of proteina se inhibitors: structure, function and regulation. J Biol Chem 269 15957 15960. Ragg H, Lokot T, Kamp PB, Atchley WR & Dress A 2001 Vertebrate serpins: construction of a conflict free phylogeny by combining exon intron and diagnostic site analyses. Mol Biol Evol 18 577 584. Renegar RH, Bazer FW & Roberts RM 1982 Placental transport and distribution of uteroferrin in the fetal pig. Biol Reprod 27 1247 1260.
151 Riahi Y, Siman Tov R & Ankri S 2004 Molecular cloning, expresin and characterization of a serin e proteinase inhibitor from Entamoeba histolytica Mol Biochem Parasitol 133 153 162. Roberts RM & Bazer FW 1988 The functions of uterine secretions. J Reprod Fert 82 875 892. Roberts RM Ealy AD, Alexenko AP, Han CS & Ezashi T 1999 Trophoblast interf erons. Placenta 20 256 264. Roberts TH, Hejgaard J, Saunders NFW, Cavicchioli R & Curmi PMG 2004 Serpins in unicellular Eukarya Archaea and Bacteria : sequence analysis a nd evolution. J Mol Evol 59 437 447. Saitou N & Nei M 1987 The neighbor joining m ethod : a new method for rec onstructing phylogenetic trees. Mol Biol Evol 4 406 425 Sancar A, L indsey Boltz LA, Unsal Ka maz K & Linn S 2004 Molecular mechanisms of mammalian DNA repair and the DNA damage checkpoints. Annu Rev Biochem 73 39 85. Sartori AA, Lukas C, Coates J, Mistrik M, Fu S, Bartek J, Baer R, Lukas J & Jackson SP 2007 Human CtIP promotes DNA end resection Nature 450 509 5 14. Sauk JJ, Nikitakis N & Siavash H 2005 Hsp47 a novel collagen binding serpin chaperone, autoantigen and ther apeutic target. Front Biosci 10 107 118. Seg erson EC, Moffatt RJ, Bazer FW & Roberts MR 1984 Suppression of phytohemagglutinin stimulated lymphocyte blastogenesis by ovine uterine milk protein. Biol Reprod 30 1175 1186. Sch midt HA, Strimmer K, Vingron M & von Haeseler A 2002 TREE PUZZLE: maximum likelihood phylogenetic analysis using quartets and parallel c omputing. Bioinformatics 18 502 504. Schvartz I, Seger D & Shaltiel S 1999 Molecules in focus Vitronectin. Int J Biochem Cell Biol 31 539 544. Sh eng S 2006 A role of novel serpin maspin in tumor progression: the divergence revealed through efforts t o converge. J Cell Physiol 209 631 635. Silverman GA, Bird PI, Carrell RW, Church FC, Coughlin PB, Gettins PGW, Irving JA, Lomas DA, Luke CJ, Moyer R W, Pemberton PA, Remold nnell E, Salvesen GS, Travis J & Whisstock JC 2001 The serpins are an expanding superfamily of structurally similar but functional diverse proteins. J Biol Chem 276 33293 33296. Silverman GA, Whisstock JC, Askew DJ, Pak SC, L u ke CJ, Cataltepe S, Irving JA & Bird PI 2004 Human clade B serpins (ov serpins) belong to a cohort of evolutionarily dispersed
152 intracellular proteinase inhibitor clades that protect cells from promiscuous proteolysis. Cell Mol Life Sci 61 301 325. Skeldal S, Larsen JV, Pedersen KE, Petersen HH, Egelund R, Christ ensen A, Jensen JK, Gliemann J & Andreasen PA 2006 Binding areas of urokinase type plasminogen activator plasminogen activator inhibitor 1 complex for endocytosis receptors of the low density lipopr otein receptor family, determined by site directed mutagenesis. FEBS J 273 5143 5159. Skopets B & Hansen PJ 1993 Identification of the predominant proteins in uterine fluids of unilaterally pregnant ewes that inhibit lymphocyte proliferation. Biol Repro d 49 997 1007 Skopets B, Liu WJ & Hansen PJ 1995 Effects of endometrial serpin like proteins on immune responses in sheep. Am J Reprod Immunol 33 86 93. Song S, Goudy K, Campbell Thompson M, Wasserfall C, Scott Jorgensen M, Wang J, Tang Q, Crawford JM, Ellis TM, Atkinson MA & Flotte TR 2004 Recombinant adeno associated virus mediated alpha 1 antitrypsin gene therapy prevents type I diabetes in NOD mice. Gene Ther 11 181 186. Spencer TE, Gray A, Johnson GA, Taylor KM Gertler A, Gootwine E, Ott TL & Baz er FW 1999 Effects of recombinant ovine interferon tau, placental lactogen, and growth hormone on the o vine uterus. Biol Reprod 61 1409 1418. Stanley P, Serpell LC & Stein PE 2006 Polymerization of human angiotensinogen: insights into its structural mec hanism and functional significance. Biochem J 400 169 178. Steenbakkers PJM, Irving JA, Harhangi HR, Swinkels WJC, Akhmanova A, Dijkerman R, Jetten MSM, Van Der Drift C, Whisstock JC & Op den Camp HJM 2008 A serpin in the cellulosome of the anaerobic fun gus Piromyces sp strain E2. Mycol Res 112 999 1006. Stephenson DC, Leslie MV, Low BG, Newton GR, Hansen PJ & Bazer FW 1989 a Secretion of the major progesterone induced proteins of the sheep uterus by caruncular and intercaruncular endometrium of the pregnant ewe from days 20 140 of gestation. Domest Anim Endocrinol 6 349 362. Stephenson DC, Hansen PJ, Newton GR, Bazer FW & Low BG 1989 b Inhibition of lymphocyte proliferation by uterin e fluid from the pregnant ewe. Biol Reprod 41 1063 1075. Stern A, Doron Faigenboim A, Erez E, Martz E Bacharach E & Pupko T 2007 Selecton 2007: advanced models for detecting positive and purifying selection using a Bayesian inference approa ch. Nucleic Acids Res 35 W506 W511.
153 Stewart MD, Johnson GA, Gray CA, Burghar dt RC, Schuler LA, Joyce MM, Bazer FW & Spencer TE 2000 Prolactin receptor and uterine milk protein expression in the ovine endometrium during the estrous cycle and pregnancy. Biol Reprod 62 1779 1789. Storm D, Herz J, Trinder P & Loos M 1997 C1 inhibito r C 1 s complexes are internalize and degraded by the low density lipoprotein receptor related protein. J Biol Chem 272 31043 31050. Strimmer K & von Haeseler, A 1996 Quartet puzzling: A quartet maximum likelihood method for reconstructing tree t opologies. Mol Biol Evol 13 964 969. Tamura K, Dudley J, Nei M & Kumar S 2007 MEGA4: Molecular Evolutionary Genetics Analys is (MEGA) software version 4.0. Mol Biol Evol 24 1596 1599. Tan J, Raja S, Davis MK, Tawfik O, Dey SK & Das SK 2002 Evidence for coordinate d interaction of cyclin D3 with p21 and cdk6 in directing the development of uterine stromal cells decidualization and polyploidy during implantation. Mech Dev 111 99 113. & Hansen PJ 2002 Natural killer like cells in the sheep: Functional char acterization and regulation by pregnancy associated pro teins. Exp Biol Med (Maywood ) 227 803 811. Brad AM & Hansen PJ 2005 b Antiproliferative actions of ovine uterine serpin. Am J Reprod Immunol 53 136 143. ML & Hansen PJ 2006 Expression and properties of recombinant ovine uterine serpin. Exp Biol Med (Maywood) 231 1313 1322. GR & Hansen PJ 2005 a Identification and cloning of caprine uterine serpi n. Mol Reprod Dev 70 262 270. Teng SC, Wu KJ, Tseng SF, Wong CW & Kao L 2003 Importin KPNA2, NBS1, DNA repair and tumorigenesis. J Mol Hist 37 293 299. Travis J & Salvesen GS 1983 Human plasma proteinase inhibitors. Annu Rev Biochem 52 655 709. Tombr an Tink J & Barnstable CJ 2003 PEDF: a multifaceted neurotrophic factor. Nat Rev Neurosci 4 628 6 36 Vallet JL 1995 Uteroferrin induces lipid peroxidation in endometrial and conceptus microsomal mambranes and is inhibited by apotransferrin, retinol bin ding protein, and the uteroferrin associated proteins. Biol Reprod 53 1436 1445. Van Gent D, Sharp P, Morgan K & Kalsheker N 2003 Serpins: structure, function and molecular evolution. Int J Biochem Cell Biol 35 1536 1547.
154 Vogel P 2005 The current mol ecular phylogeny of eutherian mammals challenge s previous interpretations of placental evolution Placenta 26 5 91 596 Waga S. Hannon GJ, Beach D & Stillman B 1994 The p21 inhibitor of cyclin dependent kinases controls DNA replication by interaction with PCNA. Nature 369 574 578. Whisstock JC & Bottomley SP 2006 Molecular gymnastics: serpin structure folding and misfolding. Current Opinion Struct Biol 16 761 768. Wind T, Hansen M, Jensen JK & Andreasen PA 2002 The molecular basis for anti proteolytic and non proteolytic functions of plasminogen activator inhibitor type 1: roles of the reactive center loop, the shutter region, the flexible joint region and the small serpin fragment. Biol Chem 2002 383 21 36. Wong WSW, Yang Z, Goldman N & Nielsen R 20 04 Accuracy and power of statistical methods for detecting adaptive evolution in protein coding sequences and for identifying positively sele cted sites. Genetics 168 1041 1051. Xie S, Low BG, Nagel RJ, Kramer KK, Anthony RV, Zoli AP, Becker s JF & Roberts RM 1991 Identification of the major pregnancy specific antigens of cattle and sheep as inactive members of the asp artic proteinase family. Proc Natl Acad Sci USA 88 10247 10251. Xiong J, Hu L, Li Y, Dou L, Cai P, Tang Z & Wang L 2008 Effect on survivin regulation of transcription level by p21 waf1 overexpression in HepG2 hepatocellular carcinoma cells. J Huazhong Univ Sci Technol Med Sci 28 308 313. Yang Z 2007 PAML 4: Phylogenetic analysis by maximum likelihood. Mol Biol Evol 24 1586 1591. Yang Z & Nielsen R 2002 Codon substitution models for detecting molecular adaptation at individual sites along specific lineages. Mol Biol Evol 19 908 917. Yang Z, Nielsen R, Goldman N & Pedersen MK 2000 Codon substitution models for heterogeneous selection pres sure at amino acids sites. Genetics 155 431 449. Yang Z, Wong WSW & Nielsen R 2005 Bayes empirical Bayes inference of amino acid sites under positive selection. Mol Biol Evol 22 2472 2479. Yao MWM, Lim H, Schust DJ, Choe SE, Farago A Ding Y, Michaud S Church GM & Maas RL 2003 Gene expression profiling reveals progesterone mediated cell cycle and immunoregulatory roles of Hoxa 10 in the preimplantantion uterus. Mol Endocrinol 17 610 627 Yoo BC, Aoki K, Xiang Y, Campbell LR, Hull RJ, Xoconostle Czare s B, Monzer J, Lee JY, Ullman DE & Lucas WJ 2000 Characterization of cucurbita maxima phloem serpin 1 (CmPS 1). A developmentally regulated elastase inhibitor. J Biol Chem 275 35122 35128.
155 Yue L, Daikoku T, Hou X, Li M, Wang H, Nojima H, Dey SK & Das SK 2005 Cyclin G1 and cyclin G2 are expressed in the periimplantation mouse uterus in a cell specific and progesterone dependent manner: evidence for aberrant regulation with hoxa 10 deficiency. Endocrinology 146 2424 2433. Zang X & Maizels RM 2001 Serine proteinase inhibitors from nematodes and the arm race between host and patho gens. Trends Biochem Sci 26 191 197. Zhang B, Lu Y, Campbell Thompson M, Spec er T, Wasserfall C, Atkinson M & Song S cells from apoptosis. Diabet es 56 1316 1323. Zhang M 2000 Maspin is an angiogenesis inhibitor. Nat Med 6 196 199. Zhou A, Huntington JA, Pannu NS Carrell RW & Read RJ 2003 How vitronectin binds PAI 1 to modulate fibrinolysis and cell migration. Nat Struct Biol 10 541 544.
156 BIOG RAPHICAL SKETCH Maria Beatriz Padua was born in 1968 in Caracas, Venezuela. She graduated from El Carmelo High School in the same city in 1985 and enrolled the next year in the School of Agronomy at the University Centroccidental Lisandro Alvarado in Barq uisimeto, Lara State, Venezuela where she received her Bachelor of Science degree in Agricultural Engineering in 1994. In 1995, she worked as sales representative for the veterinary health supply firm Grupo Catalina C.A. in Caracas, Venezuela. From 1996 to 2000, she was a pharmacist assistant in Celbefar C.A. Farmacia El Roble in the same city. She participated in an English language program at the University of Florida in 2001 and she enrolled in the Animal Molecular and Cellular Biology Graduate Progra m at the University of Florida under the supervision of Dr Peter J. Hansen in April, 2002. She received her Master of Science degree in 2004 and her thesis was elected as the best thesis for the Department of Animal Sciences. She is currently a Doctor o f Philosophy candidate. Upon completion of her degree, she will pursue a postdoctoral position.