PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 6/01 c:/ClRC/DateDue.p65-p. 15 GLUCOCORTICOID RECEPTOR GENE MUTATIONS ASSOCIATE WITH GLUCOCORTICOlD-RESPONSIVE LEUKOCYTE AND PRODUCTION TRAITS IN CATTLE By Jennifer Brigitte Jacob A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements - for the degree of DOCTOR OF PHILOSOPHY Department of Animal Science 2003 ABSTRACT GLUCOCORTICOID RECEPTOR GENE MUTATIONS ASSOCIATE WITH GLUCOCORTICOID-RESPONSIVE LEUKOCYTE AND PRODUCTION TRAITS IN CATTLE BY Jennifer Brigitte Jacob Glucocorticoids hormones were named for their effects on glucose metab- olism, but are also known to effect many other physiological processes, includ- ing embryonic development, blood pressure, lactation, behavior, and immunity. These hormones effect physiology through regulation of glucocorticoid-respon- sive gene expression. Glucocorticoids bind their cellular receptors, glucocorti- coid receptor (GR), which are ligand activated transcription factors. Following hormone binding, activated GR bind to regulatory sequences in glucocorticoid- responsive genes and regulate transcription via their first transactivation domains (called T1). Studies of human GR have identified naturally occurring mutations in DNA encoding the ligand-binding (Karl et al., 1993; Mendonca et al., 2002), DNA- binding domain (Koper et al., 1997; Ruiz et al., 2001), and first transactivation domains (Karl et al., 1996a; de Lange et al., 1997), which change GR numbers and function. In addition to natural mutations identified in human GR 1:1- encoding DNA, recent work has shown that bulls challenged with glucocorti- coids display significant additive genetic variation in glucocorticoid-responsive leukocyte traits (Tempelman et al., 2002; Burton et al, in review). This implies that variation in glucocorticoid-responsive leukocyte traits, mediated through bovine GR, are inherited. Therefore, the hypothesis of this study was that bovine glucocorticoid receptor genes harbor naturally occurring mutations in DNA encoding the bovine 11 region, and that these mutations are biologically relevant. Following PCR amplification of GR T1-encoding exon 2 (GR2) from bovine genomic DNA and subsequent sequencing, multiple single nucleotide polymorphisms (SNPs) were identified in and across six dairy and beef cattle breeds. Several of these mutations change predicted amino acid sequences, possibly resulting in a change to GR2 structure and (or) function. In order to identify whether these amino acid-changing SNPs were biologically relevant, SNPs were associated with glucocorticoid-sensitive leukocyte and milk production traits. Seventeen SNPs were found to be significantly associated (P s 0.05) with leukocyte and milk production traits, and several were associated with more than one trait. ‘ These findings imply that SNPs found within bovine GR2 loci are partially responsible for the way in which a bull’s leukocytes and metabolism respond to glucocorticoid challenge. Furthermore, three-dimensional homology modeling of two divergent Holstein alleles revealed changes to B-pleated sheet and loop structures resulting from identified amino acid-changing SNPs in bovine GR2 T1-encoding DNA. This study has therefore shown that glucocorticoid receptor DNA which encodes the first transactivation domain is polymorphic in cattle, and that these mutations may enable some animals to display a stress-resistant phenotype. Copyright by JENNIFER BRIGITTE JACOB 2003 To my Father, who gave me the love of nature and my Mother, who gave me the love of learning ACKNOWLEDGMENTS To begin with, I would like to acknowledge the Michigan Agricultural Experiment Station (MAES), without whose support this work could not have been done. Thank you for believing in me and for giving me this opportunity. I would also like to acknowledge Drs. Melvin Pagan and Douglas Maskek (as well as the Maskek farm) for providing Brahman and Brown Swiss DNAs (respectively) for identification of polymorphism within these breeds, and the American Jersey Cattlemen’s Association (AJCC) and NorthStar Select Sires (Lansing, MI) for providing Jersey semen for study of GR2 polymorphism with- in this breed. I would like to acknowledge Dr. Ted Ferris for his assistance with Jersey bull selection, and Dr. Patty Weber for GR2 primer design and also for contributing the GR bovine tissue distribution data, thank you. Additionally, I cannot forget my help; Tracy Davis, Nicole Weber, and Julie Grelewicz, for assisting with DNA isolations. Of course, I need to acknowledge and thank my committee, who has guid- ed me through some interesting times! Thank you very much for your time, ears, and wisdom; Dr. Coussens, Dr. Doumit, Dr. Erskine, and Dr. Tempelman. This study could not have been done without the support of my family and friends, who pushed, pulled, and cajoled me through this doctorate. The list has to start with my Major Professor and Mentor, Dr. Jeanne Burton. When I look back to where we started and where we are ending.....wow, it has been a TRIP! Thank you for your support and occasional well-placed kicks in the rear. Mara....l miss our chats over and around tip-boxes, you were the perfect sen- ior grad! Patty, you have taught me so much and have given me a life-long appreciation for GLOVES (smalls, please!). Ah, the lmmunogenetics vi Laboratory, an ever-changing world of graduate students, undergraduates, and post-docs, and never a dull moment. Sally, thanks for the midnight calv- ings....l didn’t know until then that I could tail-bleed a cow upside-down and stuffed into a corner. Dr. Anantachi Chaiyotwitayakun, pope goon mai na ka, and Khob kun ka. My buddies; Valencia, Rebecca, Chris. The world (and South Lansing) will never quite be the same. Have you been ‘Kissed’ today? We solved many a problem and laughed ourselves silly over coffee, ah bliss! Dr. Peter Saama, thank you for your patience and your critique, I like the way you think. Marquita, thank you, for everything. Always take The Long Way Home...l won’t forget what has brought me here. The best is always last. My family. My parents did a wonderful job of starting me toward this goal. They gave me a sense of wonder, of undying curiosity, the love of learning, and a love for animals and nature. They also gave me the sense of responsibility, toward myself and others, that has, at times, been what has kept me moving. My brothers have been sounding boards, who willingly commiserate with me and are willing to make anyone “sorry” that I don’t like. Thank you so much for simply being there! Jan. What can I say? Thank you for making me laugh and making my world so much brighter. Where my parents gave me the start toward this goal, Mark, you have pulled me through to the end. Half of this degree is yours, truly. Thank you for loving me and for doing all of those things that are too numerous to record! vii TABLE OF CONTENTS CHAPTER 1. INTRODUCTION ............................ 1 CHAPTER 2. HYPOTHESIS AND OBJECTIVES ................. 7 CHAPTER 3. LITERATURE REVIEW ........................ 8 A. The Glucocorticoid Receptor: Tissue Distribution, Structure, and Function ............................................... 8 i. Tissue Distribution of the Glucocorticoid Receptor ............... 9 ii. Domain Structure of the Glucocorticoid Receptor ................ 12 iii. Regulation of Transcription by the Glucocorticoid Receptor ....... 29 B. The Glucocorticoid Receptor Gene and mRNA ................... 44 C. Human Glucocorticoid Receptor Genes Harbor Multiple Biologically Relevant Mutations ....................................... 48 D. Mutations in the T1-Encoding Region of Human GR Genes ......... 57 CHAPTER 4. IDENTIFICATION OF MUTATIONS IN THE T1-ENCODING REGION OF BOVINE GR GENES ....................... 67 A. Introduction ............................................. 67 B. Materials and Methods ..................................... 69 i. Genomic DNA from Six Cattle Breeds ......................... 69 a. Isolation of Genomic DNA from blood ........................ 69 b. Isolation of Genomic DNA from semen ....................... 70 ii. PCR Amplification of GR2 Loci ............................. 71 iii. PCR-Restriction Fragment Length Polymorphism (PCR-RFLP) Screening of GR2 Amplicons .............................. 72 iv. Molecular Cloning of GR2 Genes ........................... 73 v. DNA Sequence analysis of Cloned GR2 Inserts ................ 75 vi. Analysis of GR2 Mutations at the DNA and Predicted Amino Acid Sequence Levels ...................................... 76 vii. Visualization of GR2 Mutations Using Three-Dimensional Protein Modeling ............................................. 76 C. Results ................................................ 77 i. PCR Amplification of Bovine GR2 and PCR-RFLP Analysis ......... 77 ii. DNA Sequence Analysis of Bovine GR2 Genes ................. 89 iii. Amino Acid Sequences Predicted from GR2 DNA Sequences ...... 98 vi. Three-Dimensional Homology Modeling of Mutant 131 Regions . . . .103 D. Discussion ............................................. 111 E. Conclusions ............................................ 113 viii CHAPTER 5. MUTATIONS IN EXON 2 OF BOVINE GLUCOCORTICOID RECEPTOR GENES ASSOCIATE WITH LEUKOCYTE SENSITIVITY TO GLUCOCORTICOIDS .............................. 1 14 A. Introduction ............................................ 114 B Methods ............................................... 11 7 i. Animals .............................................. 117 ii. Isolation of Genomic DNA ................................ 117 iii. PCR Amplification of GR2 ................................ 118 iv. Molecular Cloning of GR2 ................................ 119 v. GR2 DNA Sequence Analysis ............................. 119 vi. Polymorphism Identification .............................. 120 vii. Defining GR2 Polymorphism at the Amino Acid Sequence Level . . .120 viii. Study of GR2 Associations with Glucocorticoid-Responsive Leukocyte and Milk Production Traits ...................... 121 C. Results ............................................... 124 i. Cloning and Sequencing of Bovine T1-Encoding GR2 ............ 124 ii. Identification of GR2 Single Nucleotide Polymorphisms .......... 125 iii. Identification of Changes in Predicted Amino Acid Sequences ..... 126 iv. Associations of Identified GR2 SNPs with Glucocorticoid- Responsive Traits ..................................... 136 D. Discussion ............................................. 147 CHAPTER 6. ASSOCIATIONS BETWEEN MILK PRODUCTION TRAITS AND GLUCOCORTICOID RECEPTOR GENE POLYMORPHISM IN AMERICAN JERSEY BULLS .................................. 150 A. Introduction ............................................ 150 B. Materials and Methods .................................... 152 i. Identification of Jersey Bulls with High Versus Low Milk SCS PTA . .152 ii. Isolation of Genomic DNA from Semen, PCR Amplification of GR2, and Molecular Cloning and DNA Sequencing of GR2 ......... 154 iii. GR2 Polymorphism Identification at the DNA and Predicted Amino Acid Sequence Levels ................................. 154 iv. Estimation of Association Between GR2 SNPs and Milk Production PTAs .............................................. 155 C. Results ............................................... 155 i. The GR2 Locus is Polymorphic in American Jersey Bulls ......... 155 ii. GR2 Mutations Translate into Amino Acid Substitutions in Predicted Protein Sequence ..................................... 156 iii. GR2 SNPs Associate with Milk Production PTAs But Not Milk SCS PTA ........................................... 157 D. Discussion ............................................. 169 CHAPTER 7. GENERAL DISCUSSION AND CONCLUSION ....... 174 CHAPTER 8. RECOMMENDATIONS FOR FUTURE WORK ........ 190 APPENDICES ...................................... 193 Appendix A. Activation of the Hypothalamic-Pituitary-Adrenal, or Stress, Axis ................................................. 195 Appendix B. Reference of Amino Acids .......................... 196 Appendix C. Coding for Animals Used in the Multi-Breed Study of Chapter 4 ............................................. 197 Appendix D. Tertiary Structure Predictions for the Remaining Three Holsteins Cows Studied in Chapter 4 ......................... 199 Appendix E. Pedigree Information for Holstein Bulls Studied in Chapter 5 ............................................. 200 Appendix E. Pedigree Information for Holstein Bulls Studied in Chapter 5 ............................................. 201 Appendix F. Identification Coding and Pedigree Information for Jersey Bulls Studied in Chapter 6 ................................. 202 BIBLIOGRAPHY .................................... 203 LIST OF TABLES Table 1. Naturally-Occurring Mutations in the Ligand Binding and DNA Binding Domains of Human GR .............................. 52 Table 2. Naturally-Occurring Mutations in the First Transactivation Domain of Human GR ........................................... 56 Table 3. Site-Directed Mutagenesis of GR’s First Transactivation Domain Reveals Key Amino Acid Residues Required for Transcriptional Regulation by the Receptor Molecule ......................... 61 Table 4. Selected Immune Traits and Their Heritability Estimates In Holstein Bulls Challenged with Exogenous Glucocorticoid .......... 66 Table 5. Table of GR2 SNPs Identified Within Each of the Cattle Breeds . .96 Table 6. Summary Table Depicting the Leukocyte and Milk Production Traits Studied .......................................... 123 Table 7. Summary of Significant Associations Between GR2 SNPs and Leukocyte and Milk Production Traits ........................ 144 Table 8. Summary of Significant Associations Between GR2 Amino Acid Changes and Leukocyte and Milk Production Traits .............. 146 Table 9. Rankings of Jersey Bulls by Somatic Cell Score (SCS) Predicted Transmitting Ability (PTA) ................................. 153 Table 10. Summary of GR2 SNP Positions Commonly Held Between Holstein and Jersey Bull Populations ........................ 173 Table 11. Summary of SNPs Identified Across All Sequenced GR2 Loci .180 Table 12. Summary of SNPs Found in GR2 Loci from More Than One Cattle Population ........................................ 182 Table 13. Summary of Amino Acid Changes Identified in the GR2- Enooding 1:1 Region. ..................................... 185 xi Table 14. Summary of Amino Acid Changes Found in GR 11 Regions of More Than One Cattle Population ........................... 186 Table 15. Summary of Significant Amino Acid Changes in Human and Bovine T1 Regions and the Resulting Change in hGR Activity ...... 189 xii TABLE OF FIGURES Figure 1. Tissue Expression of Bovine Glucocorticoid Receptor (a) and B-actin (b) mRNA ........................................ 11 Figure 2. Functional Domains of Three Documented lsoforms of the Human Glucocorticoid Receptor (hGR) ....................... 13 Figure 3. Tertiary Protein Structures of hGRa Functional Domains Predicted from X-ray Crystallography ......................... 18 Figure 4. General Transcriptional Regulatory Mechanism of GR T1 ..... 24 Figure 5. Model of Direct Action of Hormone-Bound GRor on Glucocorticoid-Responsive Gene Expression in a Glucocorticoid- Sensitive Target Cell. .................................... 31 Figure 5. Model of Direct Action of Hormone-Bound GRa on Glucocorticoid- Responsive Gene Expression in a Target Cell .................. 32 Figure 6. Models of Six Mechanisms by Which Hormone-Activated GRor Can Regulate Glucocorticoid-Responsive Gene Expression ........ 34 Figure 7. Structure of the Human GR Gene, mRNAs, and Proteins ..... 46 Figure 8a. Design of Primers to PCR Amplify Bovine GR2 ............ 77 Figure 8b. Bovine GR2 Loci Amplified by PCR .................... 79 Figure 8c. DNA Sequence of the Bovine GR2 Locus ................ 81 Figure 8d. Predicted Protein Sequence of the Bovine GR T1 Region . . . .82 Figure 9a. Taql PCR-RFLPs Between DNA Pools of Five Cattle Breeds .85 Figure 9b. Haelll PCR-RFLPs for Holstein Cattle DNA Pools .......... 86 Figure 9c. Mspl PCR-RFLPs in DNA from Holstein Pools and Individual Animals within a Pool ..................................... 88 xiii Figure 10. Identification of Heterozygosity in GR2 via Analysis of Pooled and Individual Cloned GR2 Alleles ........................... 90 Figure 11a. Summary of Mutations in the GR2 Locus Across Six Cattle Breeds ................................................ 92 Figure 11b. Summary of SNPs in GR2 Loci Across Six Cattle Breeds . . .94 Figure 12a. Summary of Changes in the Predicted Amino Acid Sequences of GR2-Encoding T1 Regions Across Six Cattle Breeds . .100 Figure 12b. Summary of Changes in the Predicted Amino Acids of GR2- Encoding T1 Regions Across Six Cattle Breeds ................. 102 Figure 13. X-Ray Crystallized Tertiary Protein Structure for Pectate Lyase ................................................ 104 Figure 14a. Secondary Structural Prediction of the Human GR T1 Region ................................................ 106 Figure 14b Secondary Structural Prediction of the Bovine GR T1 Region ................................................ 108 Figure 15. Predicted Changes to the Tertiary Structure of Mutant T1 Regions ............................................... 11 0 Figure 16. Holstein Bull GR2 Amplicons ......................... 124 Figure 17a. Summary of Single Nucleotide Polymorphisms (SNPs) Observed in the GR2 Locus of 40 Holstein Bulls ................ 128 Figure 17b. Summary of GR2 SNP Frequencies in a Population of 40 Holstein Bulls ........................................... 131 Figure 18a. Summary of Polymorphism Observed in the Predicted Amino Acid Sequences for the T1 Region of GR in a Population of 40 Holstein Bulls ........................................... 133 Figure 18b. Summary of the Frequencies of Changes in the Predicted Amino Acids of GR T1 Regions in 40 Holstein Bulls .............. 135 xiv Figure 19. Associations Between Individual GR2 SNPs and a Variety of Glucocorticoid-Responsive Leukocyte Traits and Milk Production PTAs ................................................. 138 Figure 19b. % B Cells ....................................... 139 Figure 19a. "/0 CD4 Cells .................................... 139 Figure 19d. CD18 .......................................... 140 Figure 190. MHC II ......................................... 140 Figure 19f. Milk PTA ........................................ 141 Figure 19e. % C08 Cells .................................... 141 Figure 199. Milk Protein PTA .................................. 142 Figure 19h. Milk SCS PTA ................................... 142 Figure 20a. Summary of GR2 Polymorphism Observed in 20 Jersey Bulls ................................................. 159 Figure 20b. Summary of GR2 SNP Frequencies in a Population of 20 Jersey Bulls ............................................ 161 Figure 21a. Summary of Polymorphism Observed in Predicted Amino Acid Sequence Encoded by the GR2 Locus .................... 163 Figure 21b. Summary of Polymorphism Observed in the Predicted Amino Acid Sequences for GR2 Loci .............................. 165 Figure 22b. Milk Protein Weld PTA ............................. 167 Figure 22a. Milk Weld PTA ................................... 167 Figure 220. Milk SCS PTA .................................... 168 Figure 23. Summary of the Occurrence of SNPs Across all GR2 Loci Sequenced ............................................ 184 Figure 24. Summary of the Frequencies of Amino Acid Changes Across All GR2 Loci Sequenced .................................. 188 Images in this dissertation are presented in color XV LIST OF ABBREVIATIONS ACTH - adrenocorticotrophic hor- mone Al - artificial insemination AIPL-USDA - Animal Improvement Program Laboratory-United States Department of Agriculture AJCC - American Jersey Cattlemen’s Club ANOVA - analysis of variance AP-1 - activator protein-1 AR - androgen receptor ATP - adenosine triphosphate BLUP - best linear unbiased predic- tion bp - base pair CIEBP - CAAT/enhancer binding pro- tein cAMP - adenosine 3’,5’-cyclic monophosphate (cyclic AMP) CBP - p300/CREB binding protein CD11a - cluster of differentiation 11a CD18 - cluster of differentiation 18 CD40 - cluster of differentiation 40 CDGZL - cluster of differentiation 62L cDNA - copy deoxyribonucleic acid CIITA - major histocompatibility com- plex class II transactivator CRE - CAMP response element bind- ing protein (CREB) response ele- ment CREB - CAMP response element binding protein CRH - cortiootrophin releasing hor- mone dATP - deoxynucleotide adenosine triphosphate dCTP - deoxynucleotide cytosine triphosphate DBD - DNA binding domain of the glucocorticoid receptor dGTP - deoxynucleotide guanine triphosphate DNA - deoxyribonucleic acid DRB3 - bovine major histocompatibili- ty complex class II locus DTT - di-thiothreotal dTTP - deoxynucleotide tyrosine triphosphate E-box - response element for myc/max transcription factors EDTA - ethylenediaminetetraacetic acid ER - estrogen receptor Fcy- Fc receptor gamma GAPDH - glyceraldehyde-S-phos- phate dehydrogenase GCG - Genetics Computer Group GM-CSF - granulocyte-macrophage colony-stimulating factor GnRH - gonadotrophin releasing hor- mone GR - glucocorticoid receptor GRa - glucocorticoid receptor alpha isoform GRB - glucocorticoid receptor beta isoform GRP - glucocorticoid receptor isoform rho GR2 - glucocorticoid receptor exon 2 CBS - glucocorticoid response ele- ment GRIP1 - glucocorticoid receptor inter- acting protein 1 hGR - human glucocorticoid receptor HPA - hypothalamic pituitary adrenal axis hsp - heat shock protein lCAM-1 - intercellular adhesion mole- cule 1 lFN-ry- interferon gamma IL-1 - interleukin 1 lL-1a - interleukin 1 alpha IL-1j3 - interleukin 1 beta ’ IL-2 - interleukin 2 lL-4 - interleukin 4 IL-5 - interleukin 5 IL-6 - interleukin 6 IL-8 - interleukin 8 IPTG - isopropylthiogalactoside Kb - kilobase kDa - kilodalton LBD - ligand binding domain of gluco- corticoid receptor mGR - membrane glucocorticoid receptor MHC II - major histocompatability complex class II MR - mineralcorticoid receptor mRNA - messenger ribonucleic acid NAAB - National Association of Animal Breeders NF-1 - nuclear factor-1 xvii NF-AT - nuclear factor-AT NFKB - nuclear factor kappa B nGREs negative GRE NLS - nuclear localization signal PAGE - polyacrylamide gel elec- trophoresis PBS - phosphate buffered saline PCR - polymerase chain reaction PEPCK - phophoenolpyruvate car- boxykinase PIC - preinitiation complex PKC - protein kinase C POMC - proopiomelanocortin PR - progesterone receptor PTA - predicted transmitting ability PVN - paraventricular nuclei RANTES - regulated upon activation, normal T cell expressed and secreted RAR - retinoic acid receptor RFLP - restriction fragment length polymorphism rGR - rat glucocorticoid receptor RNA - ribonucleic acid RT - room temperature RXR - vitamin D receptor SAS - Statistical Analysis Software SCM - synergy control motif SCS - somatic cell score 808 - dodecyl sodium sulfate SNP - single nucleotide polymor- phism 1:1 - first transactivation domain 1:2 - second transactivation domain TBE - tris boric acid EDTA TBP - TATA binding protein TEMED - N,N,N’,N’-tetramethyI-1 ,2- ethanediamine TGF-B - transforming growth factor beta TFIIA - transcription factor poly- merase ll subunit A TFIIB - transcription factor poly- merase II subunit B TFIID - transcription factor poly- merase I| subunit D TFIIE - transcription factor poly- merase lI subunit E TFIIF - transcription factor poly- merase II subunit F TFIIH - transcription factor poly- merase II subunit H TFIIJ - transcription factor poly- merase ll subunit J xviii TH1 - type 1 T helper cell (CD4+) TI-I2 - type 2 T helper cell (CD8+) TNF-a - tumor necrosis factor alpha TR - thyroid hormone receptor UTR - untranslated region UV - ultraviolet WT - wild type xix CHAPTER 1. INTRODUCTION “T here is no need to be a doctor or a scientist to wonder why the human body is capable of resisting so many harmful agents in the course of everyday life...disease does not strike everyone indifferently. For some individuals who go down at the attack, there are others who have immunity to a greater or lesser extent”. (Metchnikov, 1908). The innate ability of an animal’s immune system to mount inflammatory responses and fight infection is vital to the maintenance of health and survival. For this reason, the study of genetic diversity between livestock animals that vary in their ability to fight infection is critical. In addition, the ability to predict an animal’s inherent sensitivity to harsh or stressful environmental and physio- logical stimuli, which could be provided by knowledge of an animals’ genetic makeup, will be key to improved management of animal health and production on farms (Freeman and Lindberg, 1993). Most cells and organ systems, including the immune system, respond to harsh or stressful stimuli by regulat- ing the expression of various genes. Examples of stimuli that are predicted or shown to affect gene expression include extremes in heat and cold, pain and fear, infection by bacteria, viruses, yeast, and fungi, and a variety of common husbandry practices (Filion et al., 1984; Mitchell et al., 1988; Nanda et al., 1990; Minton, 1994; Wohlt et al., 1994; Burton et al., 1995; Grandin, 1997; Kehrli et al., 1999; Preisler et al., 2000a, Weber et al., 2001; Madsen et al., 2002). In response to these stimuli, the paraventricular nuclei (PVN) of the hypothalamus release corticotrophin releasing hormone (CRH), which acti- vates the hypothalamic-pituitary-adrenal (HPA) axis and the release of adreno- corticotrophic hormone (ACTH) into the blood (See Appendix A). ACTH, in turn, targets the adrenals to synthesize and release glucocorticoids. In response to glucocorticoids, responsive cells may either up-regulate or down- regulate expression of genes that encode such proteins as transcription fac- tors, horrnone receptors, heat shock proteins, cytokines, proteins involved in energy metabolism, antigen presentation, and cell-to-cell communication. Such changes in gene expression allow specific adaptation of cell phenotypes such that the whole animal can cope with the harsh or stressful situation. Regulation of gene expression can occur at several levels, one of which is transcription. Transcriptional regulation of gene expression is mediated through multiple and often complex interactions between a variety of transcrip- tion factor complexes and DNA sequences (motifs) contained in the regulatory regions of target genes (Schule et al., 1988; Tjian and Maniatis, 1994; Scheinman et al., 1995; Wolberger, 1998; Brogan et al., 1999). One transcrip- tion factor important in mediating phenotypic adaptation during stress belongs to a large superfamily of ligand-activated transcription regulators and is called the glucocorticoid receptor (GR). GR mediates the actions of glucocorticoid stress hormones in animals. Therefore, GR is one transcriptional regulator at the forefront of research aimed at studying gene expression and phenotypic changes in cells of animals exposed to stress-induced increases in blood glu- cocortiooids. The superfamily of molecules to which GR belongs represents the largest known family of ligand-activated transcription factors in eukaryotes (Evans, 1988; Tsai and O’Malley, 1994). In addition to GR, other members of this superfamily include the steroid hormone receptors for estrogens (ER), andro- gens (AR), progesterone (PR), mineralcorticoids (MR), and thyroid hormones (TR). Additional members considered part of this superfamily are the retinoic acid receptor (RAR), the vitamin D receptor (RXR), and a set of proteins whose genes have DNA sequence similarity to steroid hormone receptors, called orphan receptors because their ligands are not well characterized. However, the main members of this family of ligand-activated transcription reg- ulators are GR, ER, MR, AR, PR, and TR, the human and rodent genes of which have all been cloned, sequenced, and their exon-intron boundaries established (Danielsen et al., 1986; Encio and Detera-Wadleigh, 1991). Known promoter regions for the TR, ER, AR, PR, and GR genes resemble pro- moters of genes that are constitutively expressed in most cells, called house- keeping genes (Arriza et al., 1987; Huckaby et al., 1987; Baarends et al., 1990; Sakurai et al., 1992). Examples of such genes are B-actin, glyceralde- hyde-3-phosphate dehydrogenase (GAPDH), and cathepsin D. Like the steroid receptor superfamily genes, promoters for these housekeeping genes are characterized by the presence of a GC rich “island”, absence of a TATA box, and the presence of multiple transcription initiation sites (Tsai and O’Malley, 1994; Smale, 1997; Segal et al., 1999). This similarity of promoters suggests that, like housekeeping genes, the steroid hormone receptor genes may be continuously expressed in appropriate cells because the receptors they encode are so crucial to animal health, metabolism, reproduction, and survival. Another characteristic of the steroid receptor superfamily of genes is their size and structure. Most of the genes are greater than 60 kilobases (Kb) in length, with a range of 38 Kb for PR to 140 Kb for ER. These genes also possess multiple exons, with intervening introns, where individual exons typi- cally encode specific functional domains in the complex globular receptor mol- ecule (Tsai and O’Malley, 1994). The human GR gene is 80 Kb with 9 exons and 9 introns (Encio and Detera-Wadleigh, 1991), and preliminary copy DNA (cDNA) sequence from our laboratory suggests the same GR gene structure exists in cattle (Weber and Burton; unpublished data). Therefore, GR structure and function is predicted to be similar in cattle and humans. Use of the synthetic glucocorticoid, dexamethasone, in human patients suffering from inflammatory diseases and Cushing's-like symptoms has revealed individuals who exhibit variable sensitivity to glucocorticoid therapy (Huizenga et al., 1998). Importantly, dexamethasone has extremely high specificity and affinity for GR (Roth and Kaeberle, 1982; Miller et al., 1994), suggesting that variable sensitivity to this hormone might be due to variation in GR stmcture and (or) function. This possibility has led to an extensive search for GR gene polymorphism in human patients, in an effort to identify specific DNA mutations that might explain phenotypic variation in glucocorticoid sensi- tivity (Lamberts et al., 1986; de Lange et al., 1997). Such studies have identi- fied numerous GR gene polymorphisms, some of which appear to impact cel- lular gene expression and disease status in patients harboring one or more of the mutations (Lamberts et al., 1992; Ashrof and Thompson, 1993; Karl et al., 1993; Malchoff et al., 1993; Karl et al., 1996b; Koper et al., 1997; Huizenga et al., 1998; Mendonca et al., 2002). At the molecular level, most of the pheno- type-altering GR mutations discovered to date affect the number of cellular GR expressed, glucocorticoid binding affinity of GR, or degree of interaction between GR and DNA motifs in regulatory regions of glucocorticoid sensitive genes (Koper et al., 1997; Huizenga et al., 1998). All of these human GR gene polymorphisms have been shown to have dominant negative effects in cells of affected patients. However, more subtle glucocorticoid sensitivities most probably exist and may be caused by polymorphism in other important domain-encoding exons of GR genes. This has been shown recently in humans, where clusters of polymorphisms in exon 2 that encodes the main transactivation domain of GR (called tau-1, or T1) may produce modest but detectable effects on target gene expression (de Lange et al., 1997; Koper et al., 1997). As this domain is critical to transcriptional regulation of glucocorti- coid-sensitive genes by GR, molecular genetic variation in its coding DNA might help explain phenotypic variation in important traits such as the stress response, immunity, and health in humans and other animals. This is particu- larly relevant to livestock production because modern husbandry practices that lead to unavoidable stress, such as parturition or handling, can also be associ- ated with heightened disease-susceptibility in some animals (Smith et al., 1973; Guidry et al., 1976; Wohlt et al., 1994). Evidence for the existence of genetic variation in response to stress and experimental glucocorticoid challenge is available in the diary cattle literature, and suggests that a search for bovine GR gene polymorphism is warranted. For example, Detilleux et. al. (1994) found moderate to high heritabilities (h2 ranged from 0.11 to 0.61) in a wide variety of immune response traits meas- ured in vitro on leukocytes from parturient dairy cows. Importantly, these ani- mals undergo an extreme stress response around calving, which results in high concentrations of blood glucocorticoids, progesterone, and estradiol (Smith et al., 1973), and is associated with changes in immune cell gene expression and phenotype (Lee and Kehrli, 1998; Preisler et al., 2000a,b; Burton et al., 2001; Weber et al., 2001; Madsen et al., 2002; Burton and Erskine, 2003). Similarly, experimental dexamethasone challenges in pedi- greed Holstein bulls revealed modest to significant genetic variation in altered protein expression in neutrophils (Tempelman et al., 2002; Kelm et al., in review) and mononuclear leukocytes (Abdel-Azim et al., in review; Burton et al., in review; Kelm et al., in review), indicating genetic influences in an ani- mal’s immune system sensitivity to glucocorticoids. This suggested to us that, like human GR, bovine GR genes may be polymorphic in regions that encode important functional domains of the receptor protein. This led to our hypothesis that bovine GR genes harbor biologically significant mutations that affect gene expression. The overall goals of this Ph.D. study were to characterize DNA mutations in the T1 -encoding exon 2 (GR2 locus) of bovine GR genes and to estimate the degree of association between these mutations and key leukocyte traits that are sensitive to glucocorticoids and may influence health and pro- ductivity in dairy cattle. CHAPTER 2. HYPOTHESIS AND OBJECTIVES Mutations in the first transactivation domain (T1) of glucocorticoid receptors (GR) may exist and affect GR function and thus the ability of animals to cope with stressful environments. Several biologically relevant mutations have been identified in exon 2 of human GR genes, which encodes T1 and its flanking protein regions. The hypothesis of the current study is that the T1-encod- ing GR2 locus of bovine GR genes is polymorphic and that this polymor- phism Is biologically relevant. The overall goals of this study were to identi- fy mutations in the bovine GR2 locus and determine whether these mutations associate statistically with heritable traits of mammary health and leukocyte sensitivity to glucocorticoids in dairy cattle. Four specific objectives were developed to test the hypothesis. These were: Specific Objective i) To determine if bovine glucocorticoid receptor genes are polymorphic in the GR2 locus Specific Objective ii) To characterize identified GR2 polymorphism at the DNA and amino acid sequence levels, and at the predicted tertiary protein structure level. Specific Objective iii) To determine the frequency of identified GR2 polymor- phisms within defined Holstein and Jersey bull popula- tions. Specific Objective iv) To determine if GR2 polymorphisms associate with leukocyte sensitivity to glucocorticoids and (or) milk pro- duction traits in defined Holstein and Jersey populations. CHAPTER 3. LITERATURE REVIEW A. THE GLUCOCORTICOID RECEPTOR: TISSUE DISTRIBUTION, STRUCTURE, AND FUNCTION. Glucocorticoid receptors are present in most cells predominantly as cytoplasmic receptors (GR), but they can also exist as plasma membrane (mGR) receptors in some cell types. The expression Of mGR has, to date, been correlated with the induction of apoptosis via initiation Of the Jnk signal transduction pathway, particularly in lymphoma cells (Gametchu, 1987; Sackey et al., 1997; Chen et al., 1999; GametChu and Watson, 2002). Additionally, human monocytes and B lymphocytes express mGR in their membranes, albeit at low levels relative to lymphoma cells (Buttgereit and Scheffold, 2002). Membrane GR has a size of approximately 97 kDa and has been found to be inactive at a lower size (Gametchu and Watson, 2002). Membrane GR binds hormone and triggers activation of the Jnk pathway and cell apoptosis (Gametchu et al., 1991). Cytoplasmic GR are 94 kDa in size, are the main cellular receptors for glucocorticoid hormones, and the primary mediators Of normal steroid actions in target cells. The end result Of all glucocorticoid action in normal cells is reg- ulation of glucocorticoid-responsive gene expression for the purpose of cellular adaptation to stress. In order to regulate target gene expression, glucocorti- coid hormones bind to and activate GR. Binding Of glucocorticoids turns GR into a ligand-activated transcription factor capable of altering gene expression in target cells through a variety of transcriptional regulatory routes, or the stabi- lization or destabilization Of specific messenger RNA (mRNA) transcripts. Because GR is so critical to survival, the domain structure and function of this complex protein have been extensively studied. In fact, human GR (hGR) was the first mammalian transcription factor to be studied any great detail. i. Tissue Distribution of the Glucocorticoid Receptor: As the molecules that mediate the activities of glucocorticoid hormones in target cells, GR are vital to the adaptation of glucocorticoid-responsive tis- sues tO stress. In fact, GR are absolutely required for tissue development and ultimate survival (Cole et al., 1995). Generation Of GR knock-out mice showed that although embryos were able to develop into full term fetuses, they died shortly following birth from severe abnormalities including atelectasis Of the lung, hyperplasia of the adrenal cortex, and impaired expression of gluco- neogenic enzymes in the liver (Reichardt et al., 1998). AS implied by their requirement for survival, GR have been found to be present in nearly all cells Of the body (Ballard, et. al., 1974; Reul, et. al., 1989). Initially, message for GR was not identified in jejunum, adipose, bladder, seminal vesicle, uterus, prostate tissue (Ballard et al., 1974), or the bovine lens (Jobling and Augusteyn, 2001). However, in 1994 Pedersen et al. identified the presence of hGR in both omental and subcutaneous adipose tissue, and more recently GR has been found in rat embryonic gonadal, gastrointestinal, and urogenital tis- sue (Kitraki et al., 1997). Although these Observations appear to contradict previous findings, it is possible that GR is expressed in certain tissues exclu- sively during embryonic development and is subsequently undetectable in adult tissue. In fact, the concentration of GR within target tissues varies depending on stage Of development, age, cell cycle stage, and endocrine sta- tus Of the animal (Burnstein et al., 1994 ; Yudt and Cidlowski, 2002). Our labo- ratory has shown variable but near ubiquitous tissue expression of GR mRNA in adult cattle, with abundance variations dependent on the tissue (Figure 1). Interestingly, isolated populations of bovine neutrophils and mononuclear leukocytes cells express relatively abundant levels Of GR mRNA and proteins (Preisler et al., 2000a,b; Weber et al., 2001; Figure 1) and thus are predicted .3639 E .._m «o scram umoom .._m E :mEEQEoR HSow .._m Co 333 undooom .._m Co 5665 ”mm? .__on new counm ”mam? .._m 6 :0:an 86330320 2 952mm 9m .moSoox:o_ Emacsnm £5, 32%: 3333.3. 229:2th .333: 333 ES @5265 52903. So E0: 3:02, 2.255 3.23593 93m: £5 E :2,ch mocmncsnm «.sz m0 .ocoeocmck .Gmmr .._m 6 398E439 m__oo CE 3 mo .6 363536 2 35.9 2896 m_ 36230336 2 33:63.03 .50 63mm: :03 .8 _o>o_ commoaxo czomé 96:33:00 9: w_ 3V 6ch c. .333: 350 $9: 5 2993.3 :5 m_ m_ 363536 E bm> Sn nowmmaxm 232505: 2.3: mam 5905 new .4sz $9 5339 20230336 .3626 388m 350 new .9239 .mcmEsz c. .om Co 3.32ng 33:. .9 23m."— 10 30: 38>. u or a mp 3:3 3:55 u 2 a 2 85.. =6 u up a S 35.. 2.9.3:. 0:3 u 3 d m 3:3 «.2353: 303 n on 0:3 3:30:08 303 n 2 3:3 0.03.... up a up 3:3 32m .323 .I. a a 5 3:3 m::. u c a m 3:3 33: 33.3 n v d a 3:3 323 n N a F 3:3 5.3.8 3... is, it 5. a.“ I :31. on 2. wwhwwwmwvaFNFFFowmwhum—yawF "mac“; 3. In: ._. . I. .m . ON 2. 2. t. 2. m_. E. ‘ i... i C . i. I-‘ § bu...“ ., hr MFNFFPomehomVnNF "3:3 .3 4sz 3V :=3._n 318.233... u.3.t33=.0 o:.>om So 3.30..an 33.... .F 959". 11 to be extremely sensitive to glucocorticoid hormones (Burton et al., 1995; Burton and Kehrli, 1995, 1996). Developmental changes in cellular GR expression, combined with its nearly ubiquitous distribution in cattle and other species, demonstrates the importance and far-reaching effects of this regulator of gene expression during stress (Bamberger et al., 1996). Ii. Domain Structure of the Glucocorticoid Receptor. In 1985, full length copy DNA (cDNA) for hGR was cloned from human lymphoid and fibroblast cells and the cDNA sequenced (Hollenberg et al., 1985). From this work it was determined that hGR had two splice variants, a 90-94 kDa GRoz isoform (dependent on the state of glycosylation) of 777 amino acid residues in length (Figure 2a) and a GRB isoform of 742 amino acids (Hollenberg et al., 1985; Figure 2b). A decade later, an additional iso- form, GRP, was isolated from lymphocytes of patients displaying glucocorticoid resistance and was found to be 697 amino acid residues in length (Moalli and Rosen, 1994; Figure 2c). Therefore, the main structural difference across the 3 GR lsoforms appears to be the length of each molecule, with important impli- cations for function (Figure 2). Hereafter, the GR acronym will refer to the GRa isoform. Initially, hGR was thought to be composed of three main functional domains, a region that bound steroid, one that interacted with DNA, and anoth- er that interacted with nuclear proteins (Dellweg et al., 1982). In order to study the function(s) of these domains, hGR was randomly fragmented into a num- ber of small pieces by proteolytic cleavage and each piece studied for various activities (Vedeckis et al., 1983). In that work, certain fragments retained some but not all the characteristics (DNA-binding, ligand binding, and transactivation) of the full receptor. Importantly, this work verified for the first time that hGR contains several autonomous domain regions with ligand-binding, DNA bind- 12 (a) (b) LBD DBD Figure 2. Functional Domains of Three Documented lsoforms of the Human Glucocorticoid Receptor (hGR). hGR has been identified as having three different splice isoforrns; or, B, and P. The hGRa (a) has a full ligand binding domain (LBD) and DNA bind- ing domain (DBD) as well as two trans- activation domains (11 and I2) and is thus considered the “normal” GR as it binds glucocorticoid and profoundly affects the expression of glucocorti- coid-responsive genes in target cells.The hGRB (b) has a truncated LBD and is thus unable to bind gluco- corticoids. One paradigm put forward in the literature is that hGRB provides a fast negative feedback loop for hGRa, turning off glucocorticoid signaling by competing with hGRa for protein bind- ing and (or) glucocorticoid response element (GRE) binding in promoters of hormone-responsive genes. Little is known about hGRP (c) although it is thought to lack both a LBD and 1:2. It retains the ability to bind DNA through its DBD and increased expression of this isoform by cells appears to be associated with resistance to gluco- corticoids (Moalli and Rosen, 1994). 13 ing, and transactivation activities (Vedeckis et al., 1983). In 1985, Hollenberg et al. refined the original three-domain model of hGR to include an additional immunogenic domain. This domain was located in the amino-terminus of the protein and was able to bind and cross-react to both monoclonal (Govindan and Sekeris, 1978; Okret et al., 1981; Gametchu and Harrison, 1984; Harmon et al., 1984) and polyclonal antibodies (Govindan, 1979; Eisen, 1980; Okret et al., 1981) within a species. In addition, locations of the DNA-binding region (in the middle of hGR) and the glucocorticoid binding site (near the carboxy-termi- nus ) were identified during those studies. However, it was not until 1986 that Giguere et al. (1986) identified the precise boundaries of what are now estab- lished as the four main functional domains of hGR (Figure 2a). This work employed the technique of random mutational analysis to determine the pre- cise exons in the hGR cDNA that encoded the domains involved in DNA-bind- ing (1 domain; DBD), glucocorticoid-binding (1 domain; LBD), and regulation of glucocorticoid-responsive gene transcription (2 domains; 1:1 and 12). By rea- son of their degree of effects on transcription of target gene expression in vitro, the two transcription regulation domains of hGR were called the first transacti- vation domain (tau-1, or T1, having the greatest effect on gene transcription), and the second transactivation domain (tau-2, or 12, more subtle effects on gene regulation), respectively. In addition to the four major functional domains, other regions of hGR were soon identified that afforded the molecule other important functional character- istics, such as nuclear translocation, phosphorylation, and protein-protein inter- actions (Muller and Renkawitz, 1991). Shortly following the elucidation of these hGR domains and subdomains, the murine GR gene was cloned, sequenced, and similar domains for ligand-binding, DNA-binding, and transac- tivation defined (Danielsen et al., 1986). Elucidation of boundaries between 14 the functional domains allowed further biochemical and functional analyses of each domain and their contributions to the actions of glucocorticoids on hor- mone-responsive genes. One of the first GR domains to be studied in such depth was the ligand- binding domain (LBD). Because this domain binds glucocorticoid hormone, it is primarily responsible for initiating translation of ligand message into observed biological effects in target cells. The LBD of hGR is 16 kDa in size and spans amino acid residues 537 through 777 (Simons et al., 1987; Carlstedt-Duke et al., 1988; Simons et al., 1989). In effect, LBD makes up the entire carboxy-terminal end of the hGR molecule. In particular, residues M622, C656, and C754 bind steroid and are located within hydrophobic segments of the LBD (Carlstedt-Duke et al., 1988). This would indicate that there is a hydrophobic pocket within the three-dimensional hGR, which binds hormone and activates the receptor to regulate target gene expression. Recent X-ray crystallographic structures of this domain depicts an a-helical “sandwich” struc- ture of LBD, where the hormone is buried in the middle (Figure 3a; Huang and Simons, 1999 and references therein). In addition to the hormone-binding function of LBD, this domain appears to repress in vitro transactivation when not ligated with hormone (Muller and Renkawitz, 1991). Using mutagenesis, GR LBD truncation mutants were created as well as LBD mutants with inserted amino acid residues. GR LBD mutants containing amino acid residues 1 through 532 were not active, while GR containing residues 1 through 582 retained 40% of “normal” or wild-type (WT) activity. By fusing amino acid residues 533 through 777 back into the mutants displaying reduced activity, two subdomains of the LBD were identified that possess inhibitory effects on GR activity (Hollenberg et al., 1989). These subdomains lie between amino acid residues 530 and 582 and residues 697 to 777 (Hollenberg et al., 1989). 15 Therefore, the unligated LBD acts as a ligand-dependent “switch” that keeps the remainder of the GR molecule in an inactive state. While the LBD of the carboxy-tenninus of GR is critical for initiation of hormone activity in cells, stud- ies have also shown that loss of the amino-terminus of GR leads to only weak transactivation of target genes in the presence of glucocorticoids (Encio and Detera-Wadleigh, 1991). It was thus concluded from these studies that initial receptor activation via ligand binding in the LBD is not enough to effect full reg- ulation of target gene expression, and subsequently led to further studies into the DNA-binding and transactivation domains of GR. The DNA-binding domain (DBD) is responsible for linking GR with specific DNA targets in regulatory regions of glucocorticoid-responsive genes. The general properties of such DBD were first characterized for the Xenopus tran- scription factor lllA, which was observed to possess peptide loops that form “fingers” stabilized by centralized zinc ions (Brown et al., 1985; Miller et al., 1985). Since then, this general zinc finger motif has been identified in numer- ous DNA-binding proteins, including the large family of steroid hormone recep- tors. The DBD of GR is comprised of 93 amino acids, spanning residues 440 to 510 in full-length GR (Luisi et al., 1991; van Tilborg et al., 1995). GR’s DBD contains two subdomains, each with a peptide loop finger containing four cys- teines bound to a central zinc ion (Luisi et al., 1991; Figure 3c). The amino- terminal or “first” finger contains aromatic and hydrophobic residues that form a B-sheet structure. This acts as a core to stabilize the relative orientation of both zinc fingers within the DBD and allows interaction of the first finger with a second GR molecule following homodimerization (Luisi et al., 1991). Due to its stable orientation, the first zinc finger is considered ‘specific’ in its binding to regulatory DNA in glucocorticoid-responsive genes. In contrast, the carboxy- terminal or “second” zinc finger forms an oc-helix structure followed by a B- 16 Figure 3. Tertiary Protein Structures of hGRa Functional Domains Predicted from X-ray Crystallography. Two functional domains of GR have been x-ray crystallized; (a) the ligand bind- ing domain (LBD; PDB Accession #1QKU; Eiler et al., 2001) and (c) the DNA binding domain (DBD; PDB Accession #1 RGD; van Tilborg et al., 1995). The ”:2 domain forms part of the LBD and is shown in (a) to lie just above the groove of LBD that forms the hormone binding pocket or “sandwich”. The LBD and 1:2 appear to be mostly a-helical (cylinders) in structure. The DBD (c) is comprised of two zinc finger regions, which are positioned in the GR molecule for effective interaction with the major groove of the DNA double helix as well as with glucocorticoid response elements in the regulatory DNA of target genes. Though 1:1 (b) is critical for transactivation or transrepression of gluco- corticoid-responsive genes by GR, this domain has proven elusive in x-ray crystallography experiments due to its lack of tertiary structure at neutral pH. Therefore, the crystal structure of GR’s T1 domain is still unknown. 17 Figure 3. Tertiary Protein Structures of hGRa Functional Domains Predicted from X-ray Crystallography (a) (b) (c) 18 sheet, leading to a finger with a more flexible orientation. The role of this sec- ond zinc finger appears to be to make contact with DNA-associating proteins (Luisi et al., 1991). In addition, this is the finger that binds to DNA with particu- lar affinity for sequence contained in the major groove of the double helix. Interestingly, the second fingers of GR homodimers bind to successive major grooves while orienting around, but not binding to, the minor groove of the double helix (Luisi et al., 1991; Tsai and O’Malley, 1994). This may allow GR to orient the double helix in such a way that the first fingers of GR’s DBD can then access promoters of target genes, and (or) stabilize dimerized GRs on the double helix. The second zinc finger is considered to have a more ‘non- specific’ DNA binding role than the first zinc finger due to its ability to bind to multiple DNA sequences and their associated proteins. The two zinc fingers of the DBD thus act in concert to modify the DNA double helix for subsequent binding of activated and dimerized GR to specific DNA motifs in regulatory regions of glucocorticoid-responsive genes. Therefore, DBD domain activity provides another critical step in the cascade of events leading to glucocorti- coid-induced regulation of target genes by ligand-activated GR. Lying within and flanking the LBD and DBD are additional regions, or sub- domains, of GR that have important functions in the activity of the receptor molecule. One of these subdomains lies directly downstream of the DBD and is composed of 15 amino acids (residues 510 through 525; Dahlman-Wright, et al., 1991), with a core of 5 amino acids called the dimerization or D-loop. The D-loop is responsible for homodimerization of ligand-bound GR (Luisi et al, 1991). Specific amino acids that flank the D-loop and lie within the second zinc finger of DBD also contribute to efficient dimerization of ligand-bound GR. In fact, mutation of just one of these amino acids (eg. A458T) completely abol- ishes GR dimerization (Newton, 2001 ). Dimerization of ligand-activated GR is 19 important because it contributes to the specificity and affinity of GR interac- tions with DNA (Dahlman-Wright et al., 1995). Another important subdomain of GR is a stretch of basic amino acids located directly downstream of the DBD and its D-loop region. This region car- ries one of two nuclear localization signals present in the GR molecule (Muller and Renkawitz, 1991; Sackey et al., 1996). The first nuclear localization signal is located in the amino-terminus of GR (Muller and Renkawitz, 1991). In rat GR, this DBD-associated nuclear localization subdomain is 28 amino acids long and helps mediate nuclear localization of GR molecules that become dimerized upon ligand binding (Sackey et al., 1996). The importance of this subdomain was elucidated when a single point mutation (R484H) was shown to reduce the number of nuclear GR by 60%, although whole cell expression of GR was the same as for cells with wild-type GR (Sackey et al., 1996). This indicated that fewer activated mutant GR were able to translocate into the nucleus and regulate expression of glucocorticoid-responsive genes. GR dimerization is thus critical not only for proper interactions with the DNA double helix but also for translocation of the activated receptor complex from the cytosol through the nuclear membrane. This is accomplished when the GR dimers interact with important nuclear import proteins, such as importin or, which enables GR to be transported across the nuclear envelope via special- ized nuclear pores (Prufer and Barsony, 2002). Without dimerization and nuclear localization, GR is unable to mediate the effects of glucocorticoids on expression of target genes in otherwise sensitive cells (Maden and DeFranco, 1993; Sackey et al., 1996). The GR molecule has a third subdomain that is exposed for protein-protein interaction when the receptor is not bound by hormone. This subdomain binds a variety of heat shock proteins (hsp), including hsp90, hsp70, and hsp56 20 (Schreiber, 1991; Walsh et al., 1992; Czar et al., 1994). While little is known about this subdomain, it appears to form part of the LBD and may mask the nuclear localization site when GR is associated with these hsp complexes and not bound by hormone (Czar et al., 1994). The hsp bound to this subdomain are thought to act as GR chaperones (Caamano et al., 2001), trafficking newly synthesized GR to the cytoplasm of the cell and keeping cytoplasmic GR in a state that promotes high affinity for ligand-binding by LBD (Schreiber, 1991; Walsh et al., 1992). Upon binding of hormone to LBD, the hsp subdomain pre- sumably undergoes a conformational change as the entire hsp complex is released (Czar et al., 1994). Release of the hsp complex would then expose the nuclear localization subdomain, since it is after such release that hormone- activated GR dimerize and translocate into the nucleus (Sackey et al., 1996). Therefore, the main LBD and DBD domains as well as the dimerization, nuclear localization , and hsp subdomains of GR are critical for glucocorticoid binding and subsequent activity of the hormone bound receptor in the nucleus. However, none of these domains and subdomains actually cause changed expression of honnone-responsive genes. This is the job of additional transac- tivation domains. Initiation or repression of messenger RNA synthesis is a primary point of control in the regulation of gene expression. Transactivation domains present in transcription factors, including GR, control mRNA synthesis via interactions with other transcription co-factors and proteins of the basal transcriptional com- plex (Ma and Ptashne, 1987; Sadowski et al., 1988; Mitchell and Tjian, 1989; Jiang et al., 1994). The first transactivation domains to be defined in eukaryot- ic transcription factors were for the Gal4 (Ma and Ptashne, 1987; Kakidani and Ptashne, 1988; Sadowski et al., 1988) and GCN4 (Hope and Struhl, 1986) transcription factors in yeast. These transactivation domains were found to be 21 comprised of a high percentage of glutamine, proline, and acidic amino acid residues (Mitchell and Tjian, 1989). Other glutamine-rich transactivation domains have been identified in transcriptional coactivators including oct-1, AP-1, and Drosophila homeotic proteins (Bohmann et al., 1987; Williams et al., 1987; Sturrn et al., 1988; K0 et al., 1988), although their specific functions have not been well characterized. Proline-rich transactivation domains are thought to function by disrupting a-helices in proteins that interact with them (Williams et al., 1987; K0 et al., 1988; Stmhl et al., 1988; Norman et al., 1988). Transactivation domains that are rich in acidic amino acids have been identi- fied in Gal 4 (Ma and Ptashne, 1987) as well as hGR (Hollenberg and Evans, 1988) and may facilitate transcription initiation by interacting in a relatively non- specific manner with other transcriptional regulators. One characteristic shared by acidic transactivation domains is their ability to remain highly unstructured at the tertiary protein level when placed at physiological (neutral) pH (Sigler, 1988; Dahlman-Wright et al., 1995). It has been theorized that these acidic transactivation domains remain unstructured until they contact other regulatory proteins, at which time they develop the tertiary protein struc- ture necessary for regulation of transcription initiation and repression (Sigler, 1988). The 1:1 of hGR is contained within a major immunogenic region in the amino-terminus of the molecule (Hollenberg et al. 1985; Weinberger et al., 1985; Hollenberg et al., 1986). This region was considered immunogenic due to its ability to cross-react with species-specific monoclonal (Govindan and Sekeris, 1978; Okret et al., 1981; Gametchu and Harrison, 1984; Harmon et al., 1984) and polyclonal antibodies (Govindan, 1979; Eisen, 1980; Okret et al., 1981), implying that it is variable. lmmunogenic domains are present in all members of the superfamily of steroid hormone receptors (Evans 1988). 22 —— While there is variation in the size of this domain across family members, each contains amino acid sequences that confer strong in vitro transactivation activi- ty (Giguere et al., 1986; Hollenberg et al., 1987; Godowski et al., 1988). In fact, the transactivation activity from 11 contained within these immunogenic domains occurs independent of hormone or DNA binding of a full length recep- tor molecule (Danielsen et al., 1987; Godowski et al., 1987; Hollenberg et al., 1987; Reichardt et al., 1998; Wright et al., 1991; Dahlman-Wright et al., 1995; Scheinman et al.,1995). Also, when the T1 domain is relocated in GR or other members of this steroid receptor family, it still exhibits strong transactivation activity (Dahlman-Wright, et al., 1994). Therefore, GR’s 151 domain is consid- ered to have constitutive transactivational capacity and, as such, is predomi- nantly responsible for the regulation of glucocorticoid-responsive genes by hor- mone-activated nuclear GRs (Hollenberg and Evans, 1988). The 1:1 of hGR spans amino acid residues 77 through 262 (Wright et al., 1991). The T1 domain region has a “core” region consisting of 41 amino acids, spanning residues 187 through 227. Although this core region is absolutely necessary for transactivation from 1:1, alone it only displays 60-70% of the full 1:1 activation of glucocorticoid-responsive genes (Dahlman-Wright et al., 1994). Thus, the protein sequences flanking the 171 core region are also important determinants of transactivation. Indeed, a 58 amino acid stretch (residues 187 through 244) of 11 in hGR has been identified that retains almost all of the transactivation activity of the core (Wright et al., 1991). However, the remain- ing transactivation activity from the immunogenic region of GR must come from flanking amino acid sequences, as mutations in these flanking regions have been shown to both positively and negatively affect transactivation from GR’s 1:1 (Wright et al., 1991). 23 GR I1 works by binding to either the transcription preinitiation complex (PIC) or a multitude of other transcriptional co-factors, and by that means, either activates or represses the expression of glucocorticoid-responsive genes (Figure 4). Human GR I1 contains clusters of acidic residues, which classi- fies it as an acidic transactivator (Mitchell and Tjian, 1989). Multiple phospho- rylation sites on conserved serine residues contribute further to the overall acidity of I1. This is important in light of the fact that a strong positive correla- tion exists between the acidity of such domains and their transactivation activi- ty (Danielsen et al., 1987). Despite this observation, it is also likely that certain hydrophobic amino acids in I1 domains are critical for full transactivation activ- Figure 4. General Transcriptional Regulatory Mechanism of GR I1. Figure 4. General Transcriptional Regulatory Mechanism of GR I1 . Glucocorticoid-responsive gene expression is regulated via GR I1 (arrows), through protein-protein interactions with the basal transcriptional machinery (a) as well as interaction with other transcriptional co-factors (b, c, and (1), such as AP-1 (e). Thus I1 of GR are intimately involved in the regulation of target gene expression. 24 ity as these are highly conserved across such proteins (Cress and Treizenberg, 1991; Reiger et al., 1993). Thus, it appears that a combination of domain charge and tertiary protein structure following protein-protein interac- tions dictate the transactivation potential of I1. A prediction of hGR I1 tertiary structure was recently reported using circu- lar dichroism and nuclear magnetic resonance spectroscopy (Dahlman-Wright et al., 1995). Unlike other domains of GR, I1 is largely unstructured at neutral pH (6.9-7.7; Sigler et al., 1988). However, I1 forms a-helices in the presence of trifluoroethanol, an or-helical stabilizer (Dahlman-Wright et al., 1995 and ref- erences within). No evidence of B-sheet formation was observed even in the presence of the B-sheet stabilizer, SDS (Dahlman-Wright et al., 1995). This work indicates that hGR I1 is a conformationally flexible domain in neutral physiological solutions (such as the cytoplasm of a cell) but acquires a more rigid, possibly oc-helical structure upon change in pH. Such flexibility would suggest that I1 is able to interact with multiple different proteins, readily adapt- ing its “binding groove” to accommodate the wide variety of proteins it is required to interact with for regulation of gene expression. In support of this proposition, segments of I1 have been predicted to contain oc-helices, two that reside in the core I1 region (Dahlman-Wright et al., 1995; Dahlman-Wright and McEwan, 1996). The first oc-helix spans amino acids 189 to 201, the second residues 216 through 227, and the third residues 234 through 240, with the first and second helices making up the 41 amino acid I1 core (Dahlman-Wright et al., 1995; Dahlman-Wright and McEwan, 1996). Mutation of proline residues at the edges of these three helices disrupt helical structure and cause dramatic reductions in transactivation activity of I1 (Dahlman-Wright and McEwan, 1996). This suggests that proline residues contributing to I1 oc- helices are absolutely required for transactivation. It is thus not surprising that 25 these same proline residues are highly conserved in the I1 domains of GR across mammalian species (Dahlman-Wright and McEwan, 1996). To date, the precise tertiary protein structure of I1 remains unknown. Despite this, the research reported above is conclusive: I1 is the primary domain of GR that is responsible for transcriptional regulation of glucocorticoid-responsive genes. An additional region of GR critical to the regulation of target gene tran- scription was identified in the mid-1980’s by Giguere et al. (1986). Mutagenesis of hGR revealed a region within the amino-terminus of the LBD (between amino acids 526 and 556) that retained wild-type steroid binding affinity but displayed reduced in vitro transcriptional activity (Giguere et al., 1986; Milhon et al., 1997). Further study of this region revealed that it retained transactivation activity independent of its position within hGR (Hollenberg and Evans, 1988). These findings confirmed the presence of a second weaker transactivation domain contained within the main LBD of GR; as such the domain was called the second transactivation domain, or I2. The GR I2 is composed of two structurally and functionally divergent subdomains. The amino-terminal subdomain spans amino acids 532 to 547 and is predicted to have an or-helical structure (Milhon et al., 1997). Importantly, one face of the I2 a—helix makes contact with a-helices in the main part of the LBD, forming a sandwich structure into which a glucocorticoid hormone molecule fits (Figure 3a; Milhon et al., 1997). In effect, the oc-helix of I2 functions to stabilize the LBD of GR following hormone binding. Another face of the I2 or-helix is exposed to the surface of GR and is thus accessible for interacting with other proteins (Milhon et al., 1997). Therefore, the a—helices of I2 may effect hor- mone-dependent transactivation by their stabilizing effects on the LBD and through contact with other transcription factors and co—factors. The carboxy- terminal subdomain of I2 spans residues 548 and 561 and is predicted to form 26 p exposed to the outside of the GR molecule and thus is fully accessible teractions with other proteins (Milhon et al., 1997). It is thought that this area is the primary transactivation region of I2. Indeed, a reduction in t gene expression of up to 50% was observed following mutagenesis lues L550A and SS61A) of this hGR I2 region (Milhon et al., 1997). afore, the full GR I2 domain acts both as a stabilizer of the hormone- d LBD as well as a regulator of target gene expression. 1 addition to its position within the GR molecule, I1 differs from I2 in sev- egards. Importantly, I1 can regulate transcription in in vitro gene expres- systems independent of hormone binding while I2 requires hormone bind- I display full transactivation activity (Schmitt and Stunnenberg, 1993). .errnore, I2 has a well-defined tertiary protein structure whereas I1 ins unstructured until it encounters other regulatory proteins. Therefore, ay interact with a smaller selection of similar proteins whereas I1 is pre- d to have the capacity to interact with a large variety of diverse proteins. may explain the relatively strong transactivation activity of I1 compared to Because of these differences, the two transactivation domains of GR most co-activate receptor regulation of glucocorticoid-responsive gene expres- regardless of cell type involved. 'he domain structure of GR described thus far has been for the full-length form of this receptor. As previously mentioned, however, other isoforms 2 also exist, namely GRB (Hollenberg et al., 1985) and GRP (Moalli and n, 1994; Krett 1995; de Lange 2001; Figure 2). Full GR regulation of glu- rticoid-responsive genes probably involve all three isoforms. GRB is >logous with GRoz between amino acids 1 and 727 (Hollenberger et al., ) but differs in its carboxy-terminus (Encio and Detera-Wadleigh., 1991). difference occurs because of an alternative splice site in exon QB of the 27 GR gene that splices out a large region of the LBD-encoding DNA during tran- scription (Encio and Detera-Wadleigh, 1991; Oakley et al., 1996). Therefore, GRB has a truncated LBD and is unable to bind glucocorticoids. GRB is also devoid of an intact I2 region (Encio and Detera-Wadleigh, 1991). Together with the truncated LBD, GRB is incapable of ligand-dependent transactivation. Although GRB possesses all other domains and subdomains found in GRa, it has been found to be transcriptionally inactive (Yudt and Cidlowski, 2002). Despite the ability of GRB to bind hsp molecules, they are located primarily in the nucleus as monomers in the absence of ligand, and form dimers with hor- mone-activated GRoc (Oakley et al., 1996; Oakley et al., 1997). One paradigm put forth in the biomedical literature is that GRB acts as an important modifier of GRoc activity on target gene expression. As such, GRB blocks ligand- dependent transactivation of the target gene by competing with hormone-acti- vated GRa for dimerization and DNA binding (Bamberger et al., 1995; Oakley et al., 1996; Oakley et al., 1997; Brogan et al., 1999; Oakley et al., 1999). In this way, GRB may act in a short-loop fashion to curtail a cell’s responsiveness to an excessive or prolonged glucocorticoid stimulus through competition with hormone-activated GRa. Similarly, GRP is associated with glucocorticoid resistance, although specific regulatory mechanisms of this isoform are not known. In summary, GR is present in cells in at least 3 known isoforms, all bearing the important DBD, hsp, and I1 domains, but each containing varying amounts of LBD and I2 domains. Only GRoc can bind glucocorticoid hormones to regu- late target gene expression in response to these hormones. In this case, the I1 domain appears to possess most of the transactivation activity of the mole- cule with I2 contributing some activity. However, GRB and GRP may have important modifying functions on the activity of GRoc, possibly protecting cells 28 from “hyper-responsiveness” during extreme or chronic exposure to high glu- cocorticoid concentrations. While the DBD is important for gene regulation in systems that require DNA-binding by GR, the I1 domain acts independently of hormone binding and thus initiates transactivation (or transrepression) in all glucocorticoid-responsive gene systems. The importance to gene regulation of I1 will be highlighted further in the subsequent section on GR function. iii. Regulation of Transcription by the Glucocorticoid Receptor: In the absence of ligand, GR is found mainly in the cytoplasm of glucocor- ticoid-sensitive cells. Inactive GR is bound to multiple accessory proteins including the hsp complex. At least four heat shock protein molecules are known to bind to inactive GR, 2 molecules of hsp90 and one molecule each of hsp70 and hsp56 (Figure 5b). These hsps act as chaperones that facilitate the transport of newly synthesized GR from the Golgi apparatus to the cyto- plasm following translation of the receptor protein. Hsp90 molecules chaper- one protein folding; hsp70 has protein ‘unfoldase’ activity and binds to other proteins during translation, while hsp56 plays a role in both GR folding and trafficking (reviewed in Czar et al., 1994). Binding of hsp70 is a prerequisite for subsequent binding of two hsp90 molecules, after which hsp56 binds to hsp90 (Czar et al., 1994). The main function of this hsp heterocomplex is to keep GR inactive in the cytoplasm and in a high affinity state for hormone binding in the LBD (Bamberger et al., 1996). Because glucocorticoid hormones are Iipophilic, they readily diffuse across the lipid bi-layer making up the plasma membranes of cells (Figure 5a). Once in the cytoplasm, the hormone can bind with high affinity to the GR-hsp com- plex. This ligand binding induces a conformational change in GR that causes dissociation of all hsp (Figure 5c; Bamberger et al., 1996). Though poorly understood, the ligand-induced conformational change in GR is believed to be 29 the most important consequence of ligand binding, exposing the dimerization sequence, nuclear localization signal (NLS), and perhaps the I2 domain of the receptor (Tsai and O’Malley, 1994). Shortly following hsp dissociation, GR becomes hyperphosphorylated on conserved serine residues, which appears to activate the receptor even further (Bamberger et al., 1996). Two ligand-acti- vated GR dimerize, either in the cytoplasm just prior to nuclear localization (Bamberger et al., 1996), or just following translocation. The activated GR are transported across the nuclear membrane through nucleoporins, most probably through binding of nuclear transport protein importin a (or B) to the GR NLS (Figure 5d; Prufer and Barsony, 2002). The relevance of the process from lig- and binding to nuclear translocation is that it allows GR to become appropri- ately positioned in the nucleus of the cell for subsequent transcriptional regula- tion of hormone-responsive genes (Figure 5e). However, GR’s target DNA is compressed and packaged tightly around histone and non-histone proteins into a nucleosome. In order to gain access to appropriate DNA motifs in regulatory regions of target genes, GR must remodel the nucleosomes. To accomplish this, translocated GR interacts with co-factors including the SWl/SNF and Ada- Gcn5 complexes (Henriksson et al., 1997; Robyr and Wolfe, 1998; McEwan, 2000). These complexes use energy from ATP hydrolysis to alter histone-DNA interactions, such that GR’s DBD can gain access to appropriate docking motifs (see below) in the target DNA (McEwan 2000 and references within; Figure 5e). The zinc fingers of the DBD then bind to the DNA’s double helix, keeping the underlying DNA exposed and allowing regulation of gene expres- sion by GR (Luisi et al., 1991; Tsai and O’Malley, 1994; Figure 5f, 9, h, and i). Transcriptional regulation by the ligand-activated GR is complex, occurring via several direct and indirect mechanisms, and involves physical interaction of GR with several other transcriptional regulatory proteins. In general, these mechanisms can be divided into two categories; regulation by direct GR-DNA 30 39:03:: 3.3 w:. 9.03:0 2363...: 3:3 363.6 m:. .o A. 6:... 5 3:353 :665 263333 6::v 3. 3:353 > .00 36> -_.om-m:o::o: 3:6 :65 23.693: 3.6. 6cm 8. 3: 33m 3.0 66068820. 3 0:65: :80 .3. 2936 m:. .0. 3.5.6 :9: .o 9.3 m :_ 63 63:8 3:. :_ 20.33: 658:: :3: 0. 3:03:30 3 .3 300 o. 6:20: .85 36.0.: xoo:m .3: .3636... mm :65 0:6:6 6. A3 2.3 .306. .o 33:63.: 3.3.: 3:. 390 3.868820 2:52.: 3.00 303... o>=.m:om..u.oo.toooo:_0 a :. 363.53 336 m>.w:on3m.-u.oo.toooo:_0 :o 806 65.50.3950... .0 :o=o< «3.5 .o .30.... .m 050.". 31 commoaxm. c.0290 :_ 00:0:0 :o_0m0axm. :_0§n. :_ 00:20 a I. 0 :. 5.000..an 0:00 3. =00 02:20.0 06280820 Ecofiooooza .__oo .090... . $5.... .0 :o 8:6 Uczomuococto: *0 co=o< «ooh—D *0 .0605. .m 0.59.". I '1- ~11: 32 binding, and regulation that is independent of GR binding to DNA. Direct regu- lation of gene expression involves the binding of GR homodimers to specific DNA motifs within regulatory regions of glucocorticoid-responsive genes (Figure 5e). These motifs are called glucocorticoid response elements (GRE) and occur in several forms, depending on the target gene. The simplest mechanism of action is binding of GR to a consensus GRE (Newton, 2000; Reichardt et al., 2000; Figure 6a) containing the sequence, GGTACAnnnTG'l‘l'CT (where the ‘n’ represents an arbitrary base; Scheidereit et al., 1983). These GREs act to tether GR’s transactivation domains close to promoters of target genes, allowing for GR regulation of gene expression (Beato, 1989; Beato, 1991; Miner and Yamamoto, 1992 and references within). Most known cases of this simple GR-GRE binding lead to up-regulation, or transactivation, of target gene expression. Examples of genes regulated through simple GR-GRE binding include growth hormone (Slater et al., 1985), tyrosine hydroxylase, (the rate limiting enzyme in catecholamine release; Hagerty et al., 2001), the inflammatory regulators lipocortin l and calpactin binding protein (Newton 2000), BZ-adrenoreceptors, secretory leukocyte pro- tease inhibitor, the decoy lL-1 type II receptor (Newton 2000), myelin protein 22 (Desamaud et al., 2000), and sgk1 (a regulator of epithelial sodium chan- nels; Itani et al., 2002). Additionally, GR-GRE binding transactivates the CD40 ligand gene in both TH2 and B lymphocytes (Barnes 2001; Jabara et al., 2001). GR binding to simple GREs can also repress transcription of many immune and inflammatory genes. In this context, GR-GRE binding inhibits expression of the acute phase protein a1-acid glycoprotein gene (Nishio et al., 1993), multiple cytokine genes (IL-1a, lL-1B, lL-4, lL-8, and lFN-y, reviewed in Almawi et al., 1996), adhesion molecule genes (CD18; Agura et al., 1992; and CD11a, Pitzalis et al., 2002), the macrophage Fcy receptor gene (Sivo et al., 33 Figure 6. Models of Six Mechanisms by Which Hormone-Activated GRa Can Regulate Glucocorticoid-Responsive Gene Expression. (3) (b) Pu | I I (c) ATAAITAAT Box Figure 6. Models of Six Mechanisms by Which Hormone-Activated GRa Can Regulate GI “w... . r -- Gene Expression. ln panel (a), hormone-activated GR binds to a consensus DNA sequence motif, called the glucocorticoid response element (GRE), in the promoter region of glucocorticoid-responsive genes. In doing so, the 131 domain of GR is well posi- tioned to recmit proteins of the basal transcription machinery, causing transacti- vation of the affected gene. In panel (b), a less well defined glucocorticoid response element, called nGRE, binds hormone-activated GR, but in this case the receptor transrepresses the hormone-responsive gene. A variety of nGRE motifs have been found in promoters, introns, and untranslated regions (UTRs) of glucocorticoid-responsive genes and may act by blocking movement of RNA 34 Figure 6. (Cont’d). (d) (e) polymerase dun'ng transcription of the affected gene. In panel (c), competitive nGREs are GRE sequence motifs closely placed to response elements for other transcription factors (as shown for NF-KB) resulting in the inability of that transcription factor to bind the DNA once GR has bound. In this model of gene regulation, GR blocks normally active gene expression. In panels (d) and (e), composite GREs are GREs that overlap the response element for another tran- scription factor [shown are NF-KB in (d) and AP-1 in (e)], leading either to inhibi- tion (d) or activation (e) of gene transcription through protein-protein interac- tions. In panel (f), a tethering GRE is shown in which a transcription factor (AP- 1) already bound to its response element recruits hormone-activated GR and thus blocks binding of other transcriptional co-factors to AP-1. This ultimately represses normal activation of the down-stream gene. 35 1993), and the TGF-B inhibitor gene TGF-BI (Parrelli et al., 1998), and MHC II genes (Chen et al., 2000). Therefore, this GR-GRE mechanism of action that is considered simple and straightforward causes complex phenotypic changes in cells by virtue of both activation and repression of glucocorticoid-responsive gene expression. Some genes contain poorly defined GREs whose expression, when bound by GR, is repressed. These GREs have been termed “negative GREs” or nGREs (Newton, 2000; Figure 6b). Because nGREs have no consensus sequence, their existence is rather controversial. In fact, some investigators believe that nGRE are actually GRE half-sites (Sakai, et al., 1988; Drouin, et al., 1989; Subramaniam, et al., 1997; Zhang et al., 1997) that have lower affini- ty for GR dimers than full GREs (Malkoski and Dorin, 1999), or that bind with monomers of GR, possibly GRB (Subramaniam et al., 1997). One of the best characterized nGREs is in the promoter of the pro-opiomelanocortin gene (POMC; Drouin et al., 1989a,b,c; Nakai et al., 1991). This nGRE has been identified as CGTCCA in one strand of DNA and TGACC in the complementary DNA strand (Drouin et al., 1989a). Other genes identified as harboring nGREs in either their promoter or other regulatory regions include prolactin (Sakai et al., 1988; Subramaniam et al., 1997, 1998 ), and osteocalcin (Newton 2000). To date, only one cytokine gene has been identified as containing a nGRE, the IL-1B gene (Zhang et al., 1997). In another mechanism of action, GR can bind to a GRE or nGRE, and compete with another transactivating factor for response element binding (Figure 6c), resulting in inhibition of gene transcrip- tion. In fact, several cytokine genes are regulated through this mechanism; IL- 2, IL-6, and TNF-cr gene expression is inhibited by competition for binding between GR and nuclear factor-KB (NF-KB; Almawi and Melemedjian, 2002). Additionally, GR binding to a GRE in the ER gene, inhibits normal binding of 36 the CIEBP transcription factor in ER’s promoter, thereby repressing transcrip- tion of ER (Lethimonier et al., 2002). Thus, predominantly negative effects on transcription occur through GR-nGRE binding, with increased transcriptional inhibition from steric interference of other transcription factors by bound GR. In addition to simple GREs and putative nGREs, composite GRE motifs also exist in promoters of some glucocorticoid-responsive genes. Composite GREs are GRE sites that overlap the binding site(s) of other transcription fac- tors (Newton, 2000), including activator protein-1 (AP-1; Diamond et al.,1990; Miner and Yamamoto, 1992), NF-KB (Scheinman et al., 1995; de Bosscher et al., 1997; Liden et al., 2000), cAMP response element (CRE; Scott et al., 1998), the binding site of the myc/max family (E-box; Tseng et al., 2001), and oct-1 (Chandran et al., 1996). GR binding to composite GREs can either block binding by other transcription factors, effectively inhibiting transactivation of the target gene by these other activators (Figure 6d), or enhance transcription by the co-activator sharing the composite GRE (Figure Be). In the case of AP-1, this transcription factor can be made up of c-fos homodimers, c-jun homod- imers, or c-fos/c-jun heterodimers (Herrlich 2001), determined by the ratio of c- 103 to c-jun within the cell (Diamond et al., 1990). GR binding to a composite GRE in the presence of c-fos homodimers caused no change in proliferin gene expression, whereas the presence of c-jun homodimers resulted in activation of transcription and c-fos/c-jun heterodimers resulted in repression of proliferin gene transcription in the presence of GR (Diamond et al., 1990). Examples of genes harboring composite GREs, which are transrepressed by GR, include phophoenolpyruvate carboxykinase (PEPCK) through a composite GRE/CRE site (lmai et al., 1993; Scott et al., 1998), gonadotrophin-releasing hormone (GnRH) through a GRE/oct-1 binding element (Chandran et al., 1996), the B1- adrenergic receptor through GRE/E-box binding (Tseng et al., 2001), and the 37 CRH (Malkoski and Dorin, 1999), osteocalcin (Morrison and Eisman, 1993), and proliferin genes (Pearce and Yamamoto, 1993; Miner and Yamamoto, 1992; Diamond et al., 1990) through composite GRE/AP-1 binding sites. Binding of GR to GREs, nGRES, or composite GREs in disrupted chromatin allows other regulators and transcription factors to access GR’s T1 and 1:2 domains, thus initiating transcriptional regulation of glucocorticoid-responsive target genes (McEwan, 2000). Many glucocorticoid-responsive genes do not contain GRE, nGRE, or composite GRE motifs, suggesting that glucocorticoid regulation of such genes is not dependent on GR-DNA binding. Instead, these glucocorticoid-respon- sive genes are regulated through interactions between activated GR and other transcription factors that ultimately contact DNA. These DNA elements that use GR-protein interactions to regulate gene expression are termed “tethering” GRE or nGREs (Newton 2000). Tethering elements are the response ele- ments for transcription factors, and GR binds with the transcription factor already bound to its response element (Figure 6f; Jonat et al., 1990; Schule et al., 1990; Yang Yen et al., 1990; Konig et al., 1992). Transcription factors iden- tified capable of tethering GR include AP-1 (Schule et al., 1990; Pfahl 1993; Heck et al., 1994; Herrlich 2001), NF-KB (Baeuerle and Henkel, 1994; Ray and Prefontaine, 1994; Scheinman et al., 1995; De Bosscher et al., 1997; Wissink et al., 1997; McKay and Cidlowski, 1998), GR|P1 (Hong et al., 1999), nuclear factor AT (NF-AT; Refojo et al., 2001), and p300/CREB-binding protein (CBP; Kino et al., 2002b; Refojo et al., 2001; Kamei et al., 1996). Most of the genes identified to be regulated through GR tethering involve the AP-1 or NF-KB tran- scription factors. NF-KB is a p50/p65 heterodimer (Kawakami et al.,1988; Baeuerle and Baltimore, 1989) and, when activated, is involved in a number of immunological responses such as toxic shock, acute inflammatory responses, 38 graft-versus-host reactions, and the acute phase response (Baeuerle and Henkel, 1994). GR interacts with the p65 subunit of NF-KB (Ray and Prefontaine, 1994; Wissink et al., 1997) and prevents NF-IcB binding to DNA and the resultant activation of gene transcription (Scheinman et al., 1995; de Bosscher et al., 1997; Liden et al., 2000) and, by this means, regulates tran- scription of glucocorticoid-responsive genes. Interaction between GR and AP- 1 generally results in decreased in transcription of genes normally induced by AP-1. Examples of genes that are regulated through GR binding to AP-1 are protein kinase C (PKC; Maroder et al., 1993), collagenase (Liu et al., 1995), collagenase I (Jonat et al., 1990), TGF-B (Periyasamy and Sanchez, 2002; AyanlarBatuman, et al., 1991), the proapoptotic gene Bax (Amsterdam et al., 2002), and TH1-mediating cytokines IL-2, IL-2 receptor, and IFN4y (Refojo et al., 2001). Likewise, GR interaction with NF-IcB generally results in inhibition of transcription. Genes affected via GR-NF-KB interaction include the adhe- sion molecules E-selectin (Ray et al., 1997) and lCAM-1 (Liden et al., 2000; Caldenhoven et al., 1995), the chemokine RANTES (Wingett et al., 1996), and cytokines lL-6 (Vanden Berghe et al., 1999) and GM-CSF (Refojo et al., 2001). A less studied transcription factor believed to interact with GR is NF-AT, which results in suppression of GM-CSF (Smith et al., 2001), lL-2 and IL-4 in T cells, TNF—a in B cells, and lL-4 and IL-5 in mast cells (Refojo et al., 2001). Additionally, interaction between GR and CAMP-response element-binding pro- tein transcription factor, or CREB, stimulates uteroglobin gene expression (involved in differentiation of lung epithelium; Cato et al., 1984; Nord et al., 2000), and possibly the MHC Class II trans-activator gene (CIITA; van der Stoep et al., 2002). The CD18 gene harbors not only binding elements for AP- 1, but also oct-1 and CREB (Rosmarin et al., 1992), any or all of which could interact with GR to suppress gene expression. Therefore, activated GR regu- 39 lates glucocorticoid-responsive gene expression through a variety of mecha- nisms in target immune and other cells, underscoring the potent and broad- reaching effects of glucocorticoids. Despite the specific mode of transcriptional regulation by GR, the T1 domain is almost always involved. As early as 1982 it was suggested that GR regulation of transcription via ”:1 may be through direct interaction of this domain with transcription initiators like RNA Polymerase II (Dellweg et al., 1982). Initiation of transcription by RNA Polymerase ll requires an assembly of general factors (the basal transcriptional machinery) and polymerase at or near the transcription start site. The basal transcription factors include TFIIA, TFIIB, TFIID, TFIIE, TFIIF, TFIIH, and TFIIJ and comprise a pre-initiation complex (PIC) (Tjian and Maniatis, 1994). TFIID is in itself a multiprotein complex including the TATA-binding protein (TBP) and tightly associated factors (Ford et al., 1997). Binding of TBP to the TATA box of target gene DNA is required for subsequent formation of PIC and gene activation. In addition to PIC, numer- ous transcriptional co-factors have been identified and work by modifying the activity of basal transcriptional machinery to form a more stable complex (McEwan et al., 1993; Tjian and Maniatis, 1994). In 1991, Wright et al. observed interaction of GR’s 1:1 with the DNA of a basal yeast promoter. In this study, it was noted that co-factors were unnecessary for transcriptional regulation by 121, and that the core 11 region was identified as being responsi- ble for 11’s interaction with the basal transcriptional machinery (Wright et al., 1991). However, Cordingley et al. (1987) showed that chromatin remodeling by GR exposes a nuclear factor-1 (NF-1) binding site, suggesting that 11 transactivates via cofactor proteins. Models of direct PIC interaction or cofac- tor interactions with 11 are not mutually exclusive in light of GR’s ability to reg- ulate transcription with or without DNA binding. These studies, and the fact 40 that 1:1 is unstructured prior to protein binding, makes it probable that 1:1 regu- lates transcription by binding basal transcriptional machinery as well as tran- scriptional co-factors. Studies performed by McEwan (et al., 1994) and Dahlman-Wn’ght (et al., 1994) found that although T1 bound basal transcription factors, it did not directly bind TBP. Direct interaction of 1:1 with basal tran- scriptional machinery, however, is not fully explained by 171 binding kinetics (McEwan et al., 1995). The 1:1 displayed maximum transactivation function between 20 and 40 minutes, while the PIC is fully assembled within 25 min- utes. This suggests that 1:1 interacts with not only the PIC but also with other transcriptional co-factors and that transcriptional regulation by 1:1 is more com- plex than previously thought (McEwan et al., 1995). Indeed, it was shown that 11 does interact with the SWI/SNF (Dahlman-Wright et al., 1994) and GRIP (Hong et al., 1996; Eggert et al., 1995) cofactor proteins to regulate transcrip- tion. In 1997, Ford et al. revisited the question of a 11 -TBP interaction and showed that 1:1 does indeed interact with the TFIID complex and at least partly with TBP itself. Interaction and recruitment of TFIID by 1:1 is yet another mechanism for regulation of glucocorticoid-responsive genes by hormone-acti- vated GR. Regardless of whether or not GR binds DNA to effect regulation of target gene expression, T1 is involved through its interactions with either the preinitiation complex and (or) transcriptional co-factor such as AP-1 or NF-K’B (McEwan, 2000). Therefore, 1:1 affords GR its main power to effect transcrip- tional regulation of glucocorticoid-responsive genes. The GR has an additional mechanism whereby it can regulate gene expression, without directly influencing transcription of the gene. In this capac- ity, GR does not necessarily use T1. Ligand activated GR has been shown to bind directly to the 3’ ends of certain mRNA transcripts, ultimately decreasing the stability of the message and leading to inhibition of protein expression. 41 This mechanism of GR regulation is best characterized for its own mRNA (Okret et al., 1986; Kalinyak et al., 1987; Rosewicz et al., 1988; Bumstein et al., 1990), where hormone-activated GR effectively limits GR gene expression by destabilizating its mRNA (Dong et al., 1988; Bumstein et al., 1991; Bumstein et al., 1994). Other proteins affected through this mechanism include cytokines IL-1or, IL-18, and IL-6 (Amano et al., 1993). Alternatively, mRNA for the pro-survival protein Bcl-x is stabilized by GR (Amsterdam et al., 2002). Therefore, cells may be able to modify target gene responsiveness to glucocorticoids through ligand-activated GR destabilization of mRNAs in addi- tion to the other GR regulatory mechanisms. A final series of gene regulation events in the face of glucocorticoid chal- lenge are provided by the other isoforms of GR, namely GRB and GRP (Bamberger et al., 1996; Oakley et al., 1996; de Lange et al., 2001). Early studies of the non-hormone binding GRB isoform suggested that it inhibited activity of the GRor isoform (Bamberger, et. al., 1996), implicating an auto-reg- ulatory function for GRB in a cell’s response to glucocorticoids. However, there are no clear mechanisms for this inhibition. GRB has been found to be bound to heat shock protein 90, can bind to activated GRa (de Castro et al., 1996; Oakley et al., 1999), and has an inhibitory effect on the activity of GRor in vitro (Oakley et al., 1996; Brogan et al., 1999; Oakley et al., 1999). Also, GRB is expressed at significantly higher levels in macrophages, eosinophils, and poly- morphonuclear cells from patients with glucocorticoid-insensitive asthma (Christodoulopoulos et al., 2000; Gagliardo et al., 2000; Webster et al., 2001) and arterial hypotension (Bamberger et al., 1997) than in cells from normal patients, suggesting that the GRB inhibition of GRa function is in part responsi- ble for these dysfunctions. However, GRB has not been found to affect tran- srepression by GR): of glucocorticoid-sensitive genes on GRE, AP-1, or NF-KB 42 sites, showing that the only known function of GRB is competitive inhibition of GRa transactivation. In direct opposition to the observed actions of GRB, GRP is theorized to up-regulate the activity of GRor. Recently, patients that were diagnosed with hematological malignancies and examined for the presence and prevalence of GR isoforms in bone marrow cells, were shown to express significantly lower GRB than GRor or GRP (de Lange et al., 2001). Reduced expression of GRB with normal expression of GR: as well as GRP suggested that the regulatory function of GRor is supported by GRP. Obviously, more research of this iso- form is warranted. However, if GRP does activate GRoc, it may be indirectly involved in GR regulation of glucocorticoid-responsive genes. In summary, GR is the primary mediator of the effects of glucocorticoids on target cells and tissues. It is present in nearly every cell in the body in either the cytoplasm (GR) and (or) membrane (mGR), and is alternatively spliced into several isoforms (GRa, GRB, and GRP). GRoc is the main and most abundant GR isoform, primarily responsible for transcriptional activities of GR on gluco- corticoid-responsive target genes. Scant literature indicates that the GRB and GR-P isoforms may act to modify the activities of GRa on target gene expres- sion. GR is composed of four main functional domains and several regulatory subdomains. The LBD with it’s 1:2 subdomain is responsible for forming a sta- ble hormone “sandwich”, while other GR subdomains are responsible for hsp binding, receptor dimerization, nuclear localization, and chaperone receptor folding and movement to and from the cell’s cytoplasm. The DBD uses its zinc fingers to interact with the major groove and regulatory DNA of target genes once the hormone-activated receptor translocates into the nucleus. Lastly, the 1:1 transactivation domain plays a critical and direct role in gene regulation by recruiting both PIC and other important regulatory proteins to the promoters of 43 glucocorticoid-responsive genes. GR uses numerous, complex, direct and indirect mechanisms to effect activation and (or) repression of target genes in glucocorticoid-responsive target cells. Target genes include those involved in glucose and energy metabolism, cellular differentiation, inflammation and immunity, and reproduction and lactation. The 1:1 domain is significantly involved in the regulation of these genes. Therefore, 1:1 is a good candidate domain for study of GR regulation of target gene expression and cellular phe- notypic changes in response to elevated circulating glucocorticoids. B. THE GLUCOCORTICOID RECEPTOR GENE AND MRNA Animal to animal variation in sensitivity to glucocorticoids can be explained in part by mutations in GR genes. These mutations, or polymorphisms, can result in altered receptor protein molecules or numbers, resulting in variation in glucocorticoid sensitivity of target cells. Following cloning and sequencing of the full-length hGR cDNA, investigators were able to determine that the cDNA sequence is approximately 9500 base pairs (bp) in length, and the genomic gene 80 kb in length (Hollenberg et al., 1985; Encio and Detera-Wadleigh, 1991). Identification of exonfintron boundaries was the next step in elucidating the structure of the hGR gene. GR was shown to be composed of 9 major exons separated by introns (Figure 7a; Encio and Detera-Wadleigh, 1991). The first exon of the GR gene is comprised of the first 182 bp, makes up part of the GR gene promoter, and is transcribed into a primary mRNA transcript but not into the secondary transcript, which will be ultimately translated into protein (Zhong et al, 1990; Encio and Detera-Wadleigh, 1991). Sequencing of the full GR promoter revealed up to eighteen binding sites for the activator pro- tein SP-1, four transcription initiation sites, two nGREs, as well as AP-1 and NF-IcB binding sites (Zhong et al., 1990; Govindan et al., 1991; Yudt and 44 Cidlowski, 2002). However, the GR promoter does not harbor either a TATA- or a CAAT-box (Zhong et al., 1990; Govindan et al., 1991; Nobukuni et al., 1995). Promoters that have multiple SP-1 binding sites, contain multiple initia- tion sites, and lack a TATA box are typical of housekeeping genes, examples including certain growth factors, oncogenes, cytoskeletal elements, and tran- scription factors (Azizkhan et al., 1993), all of which are required for basic cell structure and (or) function. Accordingly, housekeeping genes typically display ubiquitous and constitutive expression. Since GR is expressed in almost every cell within the body and mediates the activity of glucocorticoid hormones (see Section 3A), it is not surprising that the GR promoter fits a profile consistent with that of housekeeping genes. Indeed, the promoter elements of the GR gene indicate that GR proteins are expressed in a continuous manner in gluco— corticoid-sensitive cells, consistent with their critical roles in cellular develop- ment and adaptation during stress (Bamberger, et al., 1996). The GR genes of both mouse and man have been identified to harbor multiple distinct promot- ers. Mouse GR contains three promoters (exons 1A, 1B, and 1C; Strahle et al., 1992), while human GR has been found to contain five untranslated exon 1 sequences (Breslin et al., 2001). These multiple promoters have been theo- rized to give rise to cell type-specific regulation of GR (Breslin et al., 1998), including regulation of mGR (Chen et al., 1999a,b; Yudt and Cidlowski, 2002). Exons 2 through 9 of GR are transcribed and translated into the functional domains and subdomains of the receptor protein. The first intron of GR (intron A) is a 4 kb stretch of DNA lying 3’ of exon 1 (Encio and Detera-Wadleigh, 1991) and is followed by the 1197 bp exon 2. Exon 2 contains the transcrip- tional start site and therefore is the first exon of the GR gene that is tran- scribed. Exon 2 also encodes the entire immunogenic domain, including the II core and its upstream and downstream flanking regions (Giguere, et. al., 45 Figure 7. Structure of the Human GR Gene, mRNAs, and Proteins. (a)GenomIc DNA . 3 147<-—base Transcription and Alternative Splicing (b) Messenger RNA Messenger RNA Transactivation Transactivation Transactivation Transactivation Domaln Domain Domaln Domain DBD Figure 7. Structure of the Human GR Gene, mRNAs, and Proteins. The hGR gene (a) is composed of 9 exons with interspersing introns. Sizes of the exons (in base pairs) are listed above each one. Exon 1 encompasses the promoter of the GR gene and is transcribed into a primary but not a sec- ondary mRNA transcript. Exons 2 through 9 are transcribed into mRNA (b), which exists primarily as one of two splice variants, GRa mRNA (encoded by exons 2 through 90L) or GRB (encoded by exons 2 through 93). These mRNAs are translated into two vastly different GR proteins, the GRor isoform that binds glucocorticoid and the GRB isoform that does not (c). The main functional domains are the first transactivation domain (I1) encoded by exon 2, the zinc (Zn) finger containing DNA binding domain (DBD) encoded by exons 3 and 4, the second transactivation domain (I2) encoded by the 5’ end of exon 5, and the ligand binding domain (LBD) encoded by the 3’ end of exon 5, and by exons 6, 7, 8, and Qa. 46 1986; Hollenberg and Evans, 1986; Encio and Detera-Wadleigh, 1991). As previously described, I1 is the domain primarily responsible for transactivation and transrepression of cellular genes by GR in response to glucocorticoids. Following exon 2 is a 30 kb stretch of DNA called intron B. Next are exons 3 and 4 composed of 167 bp and 117 bp, respectively. A 400 bp intron C lies between exons 3 and 4. Together , exons 3 and 4 encode the DBD and its two zinc fingers, responsible for GR’s ability to interact with target DNA. Between exons 4 and 5 is a large intron D, composed of a 16 kb stretch of DNA. Exon 5 is 279 bp long and encodes part of the LBD, dimerization and nuclear translocation domains, and all of I2. Therefore, the sequence of exon 5 is partly responsible for hormone binding, receptor dimerization, receptor translocation, and transactivation of glucocorticoid-responsive genes (T sai and O’Malley, 1994). Introns E (2 kb), F (4.6 kb), G (13.5 kb), and H (800 bp) are interspersed between exons 5 and 6, 6 and 7, 7 and 8, and 8 and 9, respec- tively. Exon 6 is 145 bp long and together with the 3’ end of exon 5, all of exons 7 (131 bp long), and 8 (158 bp long), and the 5’ end of exon 9 encodes the full LBD of GRa (Encio and Detera-Wadleigh, 1991). In total, exon 9 is comprised of almost 4000 bp and contains an alternative splice site as well as a long (3.8 kb) 3’ untranslated region (UTR). The splice site enables exon 9 to give rise to the two predominant GR isoforms, a and B (Hollenberg et al., 1985). GRP remains uncharacterized but may be the result of a second splice site in or somewhere immediately 5’ of exon 6. Upon GR gene transcription, exon 1 and introns A through H are spliced out, exons 2 through 9(1, 2 through 98, or 2 through ~ 6 are ligated together (Figure 7b), and the various GR isoforms are translated and expressed as GR proteins in the cell (Figure 7c). Exon-specific encoding of GR’s individual 47 functional domains and subdomains has facilitated studies of GR structure, function (described in Section 3A above) and dysfunction (see Section F below). C. HUMAN GLUCOCORTICOID RECEPTOR GENEs HARBOR MULTIPLE BIOLOGICALLY RELEVANT MUTATIONS In light of the fact that GR mediates the effects of glucocorticoid hormones in target cells, receptor mutations have long been suspected as being respon- sible for multiple glucocorticoid-related diseases in human patients. These dis- eases include extreme sensitivity or resistance to glucocorticoids, and depend- ence on glucocorticoid therapy for survival. An early study on glucocorticoid- resistance compared GR function in glucocorticoid “resistant “ guinea pigs ver- sus glucocorticoid “sensitive” mice (Kraft et al., 1979). Results of this study revealed that guinea pig GR had a 20—fold decrease in glucocorticoid binding affinity as compared to the hormone binding affinity of mouse GR (Kraft et al., 1979). When the guinea pig GR gene was finally cloned, sequenced, and compared with human, monkey, rat, and mouse GR gene sequences, over 24 different changes were observed in the region Of DNA that encodes the LBD (Keightley and Fuller, 1994). These differences translated into a dramatically altered LBD in the guinea pig-(Keightley and Fuller, 1995), and are undoubt- edly responsible for the characteristic low glucocorticoid affinity Of guinea pig GR compared with GR from the more sensitive species (Keightley and Fuller, 1994; 1995). Primates (Marmoset) have also been shown to harbor multiple and significant changes in the functional domains of their GR (Brandon et al., 1991). 48 Glucocorticoid resistance has since been studied in humans with identifi- able glucocorticoid-resistance phenotypes, including hirsutism, pattern bald- ness, and sexual abnormalities (e.g., irregular menstruation, isosexual precoci- ty, pseudoherrnaphroditism; Lamberts et al., 1986; Mendonca et al., 2002). As shown in Table 1, abnormalities of GR from these patients included reduced affinity for glucocorticoids (Chrousos et al., 1982; Tomita et al., 1986; Nawata et al., 1987; Lamberts et al., 1992; Kamada et al., 1994), dramatic reduction in the number of expressed cytoplasmic GR (lida et al., 1985; Lamberts et al., 1992; Kamada et al., 1994), and altered thermolability of GR (Bronnegard et al., 1986). These abnormal properties of GR acted to significantly reduce glu- cocorticoid uptake by target tissues, resulting in dramatically elevated blood glucocorticoid concentrations, glucocorticoid-resistance, and disease pheno- types. By studying families displaying biochemical and/or physical glucocorti- coid resistant phenotypes it was found that such resistance was inherited (Tomita et al., 1986; Linder and Thompson, 1989; Hurley et al., 1991; Lamberts et al., 1992). This finding launched numerous studies aimed at the identification of causative mutations in the GR genes of these patients. One of the first suspect domains of GR was the LBD, as many glucocorticoid-resist- ance abnormalities related to reduced hormone binding affinity by GR. In fact, cloning and sequencing Of murine GR genes had revealed that one abnormali- ty in glucocorticoid binding with GR was the result of a single glutamine to glycine substitution at amino acid residue 546 (Danielsen et al.,1986). Later, Linder and Thompson (1989) analyzed the DNA sequence encoding the LBD of hGR from a patient with known resistance to glucocorticoids, and identified the additional Of a Bglll restriction endonuclease cut site compared with corre- sponding DNA from healthy patients. Using computer analysis, the LBD- encoding DNA was analyzed and found to contain four bp substitutions that 49 explained the added Bglll site and resulted in altered LBD function in the patient (Linder and Thompson, 1989). Since that study, additional mutations have been found in the LBD-encoding sequence of hGR (Table 1). One such DNA mutation was an A2054T point mutation that resulted in an important V641 D amino acid substitution (see Appendix B for a table of amino acid codes) leading to co-dominant inheritance of glucocorticoid resistance (Hurley et al., 1991). Another mutation of the LBD of hGR is an A2317G nucleotide change that results in an I729V amino acid substitution and a marked decrease in GR’s affinity for glucocorticoids (Malchoff et al., 1993). In other studies, a silent A/C change at asparagine residue 766 was not associated with any glucocorticoid resistance phenotypes (Koper et al., 1997), while a mutation causing an l559N amino acid change totally abolished ligand binding (Karl et al., 1996b) and a L753F residue change led to increased ligand-GR dissociation and glucocorticoid-resistance (Ashrof and Thompson, 1993; Palmer et al., 1991). Mendonca et al., (2002) have identified a V571A amino acid change within GR’s LBD that results in a 10-50-fold decrease in GR trans- activation and pseudoherrnaphroditism, while a G67QS mutation in exon 8 results in a 50% reduction in GR transactivation (Ruiz et al., 2001) and an I747M mutation results in decreased affinity of GR for hormone (Kino et al., 2002a; Vottero et al., 2002). In addition to these mutations, Karl et al. (1993) identified a 4 bp deletion at the intron-exon boundary of exon 6 in GR genes from patients with overt glucocorticoid resistance. This deletion removes a donor splice site in the affected allele, resulting in an unstable transcript and a 50% reduction in cytoplasmic GR content (Karl et al., 1993). A frameshift insertion of an ‘A’ at hp 2439 in exon 9, meanwhile, results in decreased num- bers of GR and is associated with Lupus Nephritis (Jiang et al., 2001). Taken together, these studies show that the LBD-encoding sequences Of human GR 50 .000:0:0 000:. .:0E:006 .0:. 08:00.0. 0:. 6:0 00:30:03 00 0:..600. 0:. .60.00..0 0:90 6:0 0:.0E06 6:265. 00 0:. 0.0 950:0 00_< .0:0..0.:E 0:00 IO 60.0.. 0:. .0 €2.05. 5 0:0...000 6.00 0560 6:0 06:00.03: 0:. 0.0065 0.:0. 0.:. .0 0:83.00 6:0:-..0. 0: .r 00:00 00 ._0:. :_ 8060.00.20: 5:0: .0:. 60.._.:06. :00: 00.0 0>0: 00.0000 0:.: .0 >.0.._0> < .0.:0..0a 60.00..0 :. 00:0.0.00.-6_00.:0000:_0 6:0 00:30:08 00 .0:0..0::.0>6 :..>> 60.060000 0.0 08:06.: 000:. .0 >:0.>. 006006.: 5.0000. 0:. .0 0:658 052.: .000. <20 6:0 .00... 0:65: 6:00.. 0:. 0:600:0 0:0x0 :. 0:0..0.:E 030.053: .090: 0. 950:0 :00: 0>0: 00:00 00 :0E:I .00 00:51.0 050.600 06.650 <20 6:0 06.650 6:00... 0:. :. 00060.22 05.530.265.02 ._. 0.00... 51 0000 .00000. 000.0.0:0...>.000 .00.:0:00.0.o.:0 u ..<0: 00 20:20:. .o 09.. 2.; u 3.0 0....:000 0200.. :...s 000060000 .00 .0 room ...0 .0 9.0.... 0.00:5: U0000..00U ”00.0.3.0 £500.00.“— 0 coxw\0m. 00:000.... (mavw 60. .20 .0 .008. .820 000:0 oz 0 8.0.00. 200.2 0 o. F 000. .:o...:Eo:P 0:0 .920... :00. .._m .0 0:20: 820.800.: 00-28.5880 00000.05 00.. .005 50002. 080 .20 8.02.0020: 00 0:0 .0 o.0:o> 0080 .20 .0 0:2 00.00.3880 5. 2.5.... 00000.00: :00 .26: 00.00: 000. .2... .0 29.202 0.2.00 :0 Ease :_ 8:030. 20.-.. 0 800.00.. .00; 5.000 :00. .20 .0 .082 E020 000:0 oz 0 80.0.00. 0&0: e o. 0 0.200.000.0006 .0. 3.02.0 c0260.. .00. .20 .0 20:31 2:82:00 ”2.2.00 :20 0...: :. oz 00. >50: 0000.. 50:8: 8:00: :o: 000. .._0 .0 :00. 000.002: 0.09:0: 00 0. 00.6060. $00 0 09.0. 0.6 00:00 0.06 0: v .80 .._m .0 NS: 0.2.8 000.0202 0:5 .0 $00 0 80:00.. 00.00 200000 0000 .20 .0 8:00:02 E0_._:o.::0::0:880n_ 0 800.00.. 53> 030: :03. .20 .0 :8. 0202:8820 0:.: o. 030:: 0 800.00.. 2009 <82. 60. .20 .0 .89. 8:0.0_00.-_8_:8 5.2. 820.800... 02 :00 00.0.0 .000 .._m .0 ~50 0:00:02: 02 e 80.000: :5... 2.0030 room .._0 .0 000.000.”. 002.00.002.00 .00....00 600.0 “00000.0... 00.0590 0.5 2.55... 000. .._m .0 00:: :0 .5 053.928.5880 8.0.2 :0. 82.00:. a: 4 20 :00 $005.05. .00. .20 .0 80:30 :_ :0 .o 0.:...0 080:0: 00000.00: 00.. 2:222 2:222 000. ”.00. 0.5. 0:: .0203. 00.: 3:50. :_ 8:20.00. 28.8885 00.. 2:222 0.0.5.2 0000.0.00 002.000.... 00 00xm.l<0.0E0016.0< 00.E< 06:00.032 .00 0005... .0 00.0800 00.650 <20 600 00.650 6000.: 00. 0. 00060.32 00...:000.>__0.5.0z .F 0.00... 52 genes are polymorphic, often resulting in amino acid changes that cause reduced glucocorticoid binding and (or) decreased expression of cellular GR in affected patients. Another GR domain of interest from glucocorticoid-resistant patients is the DBD. Although DNA binding by GR is not essential for regulation of some glu- cocorticoid-responsive genes (T jian and Maniatis, 1994; Tsai and O’Malley, 1994; Scheinman et al., 1995), it is necessary for GR regulation of many other genes. Mutations in this region not only influence GR’s ability to interact with DNA, but also affect transactivation from 12. Site-directed mutagenesis stud- ies have identified numerous amino acid residues that, when mutated, abro- gate GR binding to DNA and resultant transactivation (Oro, et al., 1988; Schena et al., 1989). In 1993, Zandi et al. characterized induced GR mutants (H451 N/S459G) that actually caused increased DNA binding capacity. Another interesting mutation found in the rat GR DBD was a lysine to alanine switch at residue 461 (Starr et al., 1996). This change resulted in a GR 132 variant that replaced normal AP-1 dependent transrepression with 1:2 transactivation (Meyer et al., 1997). Based on these studies, patients with glucocorticoid resistance were also tested for the presence of natural mutations in the DBD- encoding region of the GR gene (Table 1). In one study, a G-to-C mutation at hp 1549 was identified but did not associate with glucocorticoid resistance in the patients tested (Koper et al., 1997). More recently, a G to A mutation resulting in a R477H amino acid change was identified in GR’s DBD, but also was not associated with glucocorticoid-resistant phenotypes (Ruiz et al., 2001). So, to date, the bulk of mutations affecting the biological function of GR’s DBD have resulted from mutagenesis studies and are not naturally occurring. Whereas naturally occurring mutations in the LBD often abrogate the activity of GR, mutagenesis of the DBD has displayed a variety of effects that include all 53 possible variations in transactivation and transrepression activity of 1:2. Recently, mutations have also been identified in the promoter region of hGR genes. One such mutation alters the recognition site for the Bcll restric- tion endonuclease in the promoter region of the GR gene (Panarelli et al., 1998; Table 2). The precise location of this mutation is not yet known, but the mutation is statistically associated with clinical glucocorticoid resistance in the forms of abdominal obesity, insulin resistance, and elevated blood pressure (Rosmond et al., 2000). Another mutation was identified within the hGR gene promoter using the restriction enzyme, Tth111l (Rosmond et al., 2001). The subjects tested in this study had significantly increased circulating glucocorti- coid concentrations but no other overt physical, endocrine, or metabolic disor- ders (Rosmond et al., 2001). Although the consequences of the presence of such mutations in the promoter regions of GR genes have not been fully eluci- dated, it is possible that they may determine the density of cytoplasmic GRs. In summary, it is clear that numerous biologically relevant mutations exist in the promoter, DBD-encoding, and LBD-encoding regions of human GR genes. These molecular genetic variations lead to changes in the number, structure, and (or) function of mutant GRs, some of which lead to discemable health and metabolic disorders in humans. While it is quite possible that such mutations could also affect cellular gene expression during glucocorticoid chal- lenge, the patients harboring these forms of mutant GR appear to compensate for the glucocorticoid resistance by increasing the secretion of glucocorticoids from the adrenal cortex. It is this elevation in basal levels of blood glucocorti- coids that alert doctors to the presence of a glucocorticoid-resistance disease; however, effects of mutations in the GR gene promoter, LBD and DBD on gene expression per se are not clear. Because GR regulation of gene expres- sion is primarily through the T1 domain, this domain would be an interesting 54 .00m:0:0 000:. .:0E0000 .00. 000:0.0.0. 0:. 0:0 .000>.0:0:0 m0 0:. .500. 0... 00.0000 0:90 0:0 50:50 .0:0..0:0. mo 0... 0.0 :2.0..0 00_< 0:0..05... 00.0.. 0:. .0 0:0...000 0.00 0:.E0 0:0 02.0080: 0:. 0.8.0:. 0.0.0. 0.... .0 0:80.00 0:2-..0. 0:: 0:00... :0E0: .0 002.0:000 .00 :. 000:0:0 .:00...:m.0 5.2. 00.0.00000 :000 .0: 0>0: 0:00.058 000... .0 .00.: ..0>02.0... 0.0.0.0 mo 0... .0 .5. 50800 :0..0>..000:0.. .0... 0:. 0000000 .0... :0x0 0:. 00:00 00 0.06.0 .0 0 :0x0 :. 00....c00. :000 0>0: 000.005.: 0:...0000 2.0.0.0: _0.0>00 .00 00:5: .0 50500 00.83.0005... .2... 0.... 0. 00030.32 m:...:0oo 3.0.302 .N 0.00.. 55 .000 .._0 .0 000000 0.0000 00.00.. 0...... 00.0.0000< 0 00.0.... 00002 <000._0 0300000000 .000 .._0 .0 0000.000 0.000020 000 0.0000 0...... 00.0.0000< 0 00.0.... 00002 <000.0 mwmoxm 0000 .._0 .0 0:02.023. 0000.000 000.00 0...... 000000000 02 0 00.0.... 00002 <000.0 02020000020 0. .0.... 000. .._0 .0 000000... .0000 00000.00. 0...... 00.0.8000 00 .0.... 0 00.0.... 00002 .8005 000. .._0 .0 00v. 00.000000 00000.00 02 0 00.0.... 00002 <000.0 0000. .._0 .0 00v. 0.0.0.0 00 00.0002. 0 00x0... 00.0005 000.. 0000 _._0 .0 000“. .00__0 00.000000 02 0 00x0... 0.000.. 5.20 gm: 00.000000 00000.00 02 0 00x0... 20000 <0000 0_0_ 00.000000 00000.00 02 0 00x0... .0.... 00000 .00...0 .0 00.00.030.000 000 0.000 00.000000 03000.00 02 0 00x0... .00“. 00.00 000...0 .0 .003. .000. ...0 .0 0000.. 00 00.000000 0.00.00.00 02 0 003.... v.00... 4.0000 000. .._0 .0 .008. .0000 .._0 .0 000”. .00__0 ”00000000 00000.00 02 0 00.0.... 0000 4.0000 mm .0 0.00.00: 08. ...0 .0 50.0000 00..0> .00..0...0000000> 0...... 00.0.0000< 0 09m... 000:. 0..0 :8 000. .._0 .0 0:00.05 .0. 00.0.0000 00000.00. 0...... 00.0.0000< 0 coxm: 00..:_ 0..0 __0m 000.02. 00000 000. .._0 .0 .0>00>> 0. 0.E0:._000..00>0 0...... 00.0.0000< 0 000m... 0000. 0..0 ..0m .00. .._0 .0 000005 .0000..0.>. 0. 0000000 0.0.00.6 0 00.0.... 0.0.00.2 0.00.0.2 0000.20: 00.000000 00 09.90.0050 ”-0” :0- 0_o< 00.02 00000.82 . 4 56 region to study biologically relevant mutations that cause changes in gene expression and thus phenotypic changes during stress responses of humans and other species. Studies outlined in the subsequent sections and in Table 2 indicate that the 11 domain of hGR is indeed polymorphic. D. MUTATIONS IN THE 1:1 -ENCODING REGION OF HUMAN GR GENES Transactivation domains enhance one or more rate-limiting step(s) in the assembly of the preinitiation complex by interacting directly or indirectly with transcription factors or co-factors. The first transactivation domain of GR is composed of 185 amino acids (between residues 77 and 262 of hGR cDNA) with a core 41 amino acid region necessary for transactivation or transrepres- sion activity (Dahlman-Wright et al., 1994). GR 1:1 interacts with members of the basal transcriptional machinery, as well as numerous transcriptional co-fac- tor (see Section Diii, above). For this reason the importance of GR’s T1 is its involvement in every aspect of target gene regulation by glucocorticoid hor- mones. Naturally occurring and site-directed mutants of human GR 11 have been identified that display varying effects on transcriptional regulation by GR. Naturally occurring GR I1 mutants that have been identified all stem from point mutations harbored in exon 2 of human GR genes (Karl et al., 1993; Koper et al., 1997; Huizenga et al., 1998; etc. Table 2). Two such mutations with a small separation between them were identified in the N-terminal region of 11 (Koper et al., 1997). The first G to A change at bp 198 was silent (E22E) while the second G to A mutation at bp 200 changed an arginine residue to a lysine (R23K; Koper et al., 1997). Other mutations have been identified that changed amino acid sequence but led to no discernable changes in phenotype (Table 2). An additional mutation was identified as a G to A change at hp 1220 57 in exon 2 of hGR genes, resulting in an asparagine to serine change at amino acid residue 363 (N363S; Karl et al., 1993). None of these mutations were found to be associated with phenotypes characteristic of glucocorticoid-resist- ance (Koper et al., 1997). However, the N3638 mutation was further investi- gated for possible association with cellular sensitivity to glucocorticoids (Huizenga et al., 1998). Changes in insulin and cortisol levels following gluco- corticoid treatment suggested that patients harboring the N3638 mutation tended toward increased glucocorticoid sensitivity, but values for these vari- ables were not significantly different from normal patients with “wild type” T1. Additionally, there was no direct association between the mutant GR 1'1 domains and degree of transactivation of a reporter gene in vitro (Huizenga et al., 1998). More recently, several studies have analyzed the N3638 mutation for possible associations with adrenal androgen excess (Kahsar—Miller et al., 2000), glucocorticoid-sensitivity (Rosmond et al., 2001), and central obesity as a factor of coronary disease (Dobson et al., 2001). Only central obesity in a group of Dutch men was associated in any way with this mutant form of ”c1 (Dobson et al., 2001). This work on the N3638 GR mutant suggests that 1:1 may be able to tolerate multiple mutations without dramatically altering its transactivation activity. Certainly, the role of GR as the key regulator of gluco- corticoid-responsive gene expression would suggest that 1'1 be necessarily tol- erant to mutations. However, site directed mutagenesis studies do not support this finding. In marked contrast to the naturally occurring mutations so far identified in hGR 1:1, site directed mutagenesis of 11 has identified important residues that effect significant changes in the transactivation activity of this domain (Table 3). Most of these studies have focused on the 11 core region. In the first of such studies, mutagenesis of four proline residues, or combina- tion of residues (at amino acid positions 194, 197, 220, 223), reduced the 58 transactivation potential of T1 between 50 and 80% compared to wild type “:1 (Dahlman-Wright and McEwan, 1996; Table 3). These residues were selected for study because they flank the putative oz-helices of 1:1, structures which may be partially responsible for 1:1 interactions with other proteins. As expected, mutagenesis affected the stability and function of the 131 core. In a subsequent study, Almlof et al. (1997) mutated nearly every amino acid in the 131 core in order to identify residues critical for “:1 function. The core was divided into four regions for site-directed mutagenesis; helices 1, 2, 3, and a loop region. Mutations that resulted in reductions of helix 1 activity were found primarily at three hydrophobic amino acid residues; F191, I193, and D196 (see Appendix B for classification of amino acid residues). Replacement of these hydropho- bic residues with either alanine or an acidic residue resulted in up to 80% reduction in the transactivation of a reporter gene. Mutation of the two residues, I193 and D196, displayed increases in transactivation of up to 151% (Almlof et al., 1997). Within the putative loop region of T1, the amino acid residue W213 displayed the most dramatic reduction of 70%, but substitutions with other hydrophobic residues (W213F and W213Y) restored transcriptional ability. Single substitutions within helix 2 affected transcription to a mild degree, with multiple substitutions of both L224 and L225 showing the most significant down-regulation of transcription (64-72%). One cysteine residue (C223G,R) also showed rather significant reductions in transactivation upon mutagenesis, causing speculation that this residue may be involved in forma- tion of disulfide bridges. Double mutations in helix 3 displayed the greatest reductions in transcriptional activity (60-89%). The exception to this finding was amino acid L236, which displayed 83% reduction in activity when it was the only mutated amino acid in helix 3 (Almlof et al., 1997). The effect of this single mutation is interesting in light of the fact that GR 11 has been shown to 59 60:. 6.68 3.8.6 9: w. .209" doom .998... new 63.4-3306. new 59 :6 6 62.7339... .2 E. :9 6...... .59 ...m .o 66.2. 6cm camp 695%.). 6:... Em..>>-:wE_:mn. :. B. 8.8:: :o 688 9m 98:5: 9669 66m o:_E<+ .mcmEocmca 89.: 6988386 .9: 88999 9: 9m: 9:28 68: E9. 2... 53:8 m0 C.>>. 9.5 6.3., 6 8868.8 8 8.286.: mm 698:... 6 53:8 8.62.889. 9: 8:86... 9:28 966:: 9.... .8266... 9: 5. 69856 8.3.8.7, 9:92. 9: 9m: 5:28 :6: 9.... .688... 8:93;. 638 9: 6cm 8: -mSE .9 698.9 6.8 oEEm 9: 88.88 5:28 6cm:-:m. 9.... 53:8 3. c. 8890:. 68586 c. 69.89 2209 :6... .968 8.8.6 9: 2 95:92: 2:80:66... 53:8 3. 26. c. 69.89 :26 :9: :o 699:... 8...; 8...; .03 68 .m& .03 88.8. .o 8:896 9: 53, 53:8 E. 6:86 26.6988 2 69.89 995 .8869 6.8 866 2: -:_:E 9 96:93:. .28. 9: 898 m:.~..mEE:m 53:8 8.62688: mo 9 8088 68586 .938 c. 69.89 .98 E. 9.: c. 9.622.. .mo 6 869 2.: 59:8 8.62688: :9: 9: 55.; 88.8. 6 8:92: 69096-96 6.38.0.2 .0582". 9.: .3 8.6.38: 6:25.882... .2 8.68: 336.8: 6.04. o:_E< >8. 985: 5.82. 8.62888... 8...". PEG :0 988855. 8.8.5.35 .m 0.8... 60 9m. 888:. 6.2-NF 5.0m :99. a mmva. 9:. 6890:. 6.97m 5.0m mm 5v. 88 69mm: 68 62.7.8306. 888:. 6.2.: H.20m mmev. 0.:. $8 m M. N x.6... :va>> w SN: 0.:. $3 m a N x=2. :va>> w 6er 0.:. $0: N x.6... .SNN: w XmFNm. 0.:. $8 m x8... :va>> 0.:. $3 N vso... ._oNN: 3m: :6 6 28.4-3395 $0NP N x.6... XmFNm 9m. $.VN m x.6... DOVNZ w :mmNo 0.:. $8 m x..o... DOVNz w 2:3: 0.:. $5 m x.6... >mmN.. 0.:. $8 m x.6... 6qu w >mmN: 0.:. $2 m x.6... >mmN4 w .68: 0.:. $mN N x.6... >mNN4 w >VNN.. 0.:. $vt N x.6... :mNN._ 9m. $mmN N x.6... : FNNm. 9m. 09: :8. >25 9:. $8 :8. :95 0.:. $mN F x.6... >8 5 0.:. $me F x.6... :mm: .59 :6 :o 6.62 $mN w x.6... n. 5:: .o :52: 0.:. $9 N a. F x.72.. :mNNz :ONN. :35 :35 0.:. $8 N v2.2.. :mNNz w :ONN. 89 .82,on :8 39.58.80 $0N F x.6... :33 w :83. :0 ts 2 8.2.200 8 6223.2 8:9lu:oF 9.0” Stage: a :o com—588.3%. r... .o :2mow. #63335. Eo< OEE< 6.8222 .288: «5 .3 c2538: 6:22.022... .0.. 82:8: 8:28: 20< o:.E< >3. 285: 22:00 8.62588... 8...: 9:0 :0 28885.2 88.5.26 .m 28... 61 tolerate single mutations much better than multiple mutations (lniguez-Lluhi, et al., 1997). In fact, it has been theorized that “C1 has pleomorphic activation “surfaces”, which interact directly with transcriptional co-factors to regulate transcription. The Almlof study above indicated that outside of helix 1, single point mutations in 1:1 are not as important to transcriptional activity as multiple mutations. The “surface” theory combined with the unstructured nature of 11 at neutrophil pH would explain why it is able to tolerate single mutations much better than multiple mutations. Moreover, helix 1 of the 1:1 core is most likely located at the surface of the receptor protein and may be accessible for inter- actions with target proteins (McEwan et al., 1994). If true, mutations in helix 1 would be expected to have a significant impact on If function. One amino acid within the putative loop structure, W213, appears to be critical for this interaction surface and would explain the dramatic effect when mutated (Almlof et al., 1997). In another mutagenesis study, the full If region of rat GR (rGR) was used for mutagenesis work (lniguez-Lluhi et al., 1997). In rGR, the 11 region is located between amino acid residues 108 and 317 and, like hGR, is comprised of a large percentage of acidic residues (Tasset et al., 1990). In the lniguez- Lluhi et al (1997) study, rGFt 131 domains harboring multiple mutations (rather than point mutations) were generated and then fused to the DBD and LBD of rGFi and expressed in mammalian cells along with a luciferase reporter gene. No specific cluster of mutations in the T1 region regulating luciferase gene activity were generated, but an inverse relationship was observed between overall transcriptional activity of mutant 1:13 and the number of amino acids changed (lniguez-Lluhi et al., 1997). The more residues mutated in rGR 1:1, the greater the reduction in luciferase gene transcription. Several strongly affected mutants had an altered tryptophan (W234) residue in common (Table 62 3). Interestingly, mutation of a tryptophan residue (W213) in a homologous region of hGR was also found to significantly affect transcription of the luciferase reporter gene (Almlof et al., 1997). Both studies would indicate that this tryptophan residue is of particular importance to transcriptional regulation mediated by GR 1:1. This work also indicates that individual amino acid muta- tions within 11 may have little effect on transcription regulation of target genes by GR, whereas clusters of two or three critical mutations may cause dramatic down-regulation of transcriptional activity. This finding argues in favor of GR ”:1 possessing surfaces that function to regulate gene activation or repression, where multiple rather than one or two residues makes up a contact area for target co-factors. This notion of contact surfaces in T1 would also be in line with the known flexibility of this domain (described in Section Dill), and implies that ”[1 selectively exposes or uses different regulatory surfaces depending on the context of the specific gene promoter affected and the cellular proteins involved in GR’s regulation of the gene. If true, it is easy to envision that sin- gle mutations in T1 are better tolerated than clusters of mutations, which may modify an entire protein-binding surface of the domain. In a subsequent study of rGR, lniguez-Lluhi and Pearce (2000) identified a specific motif within the rGR T1 region which, when mutated, dramatically increased the transactivation of 1:1. This motif, determined as (IN)KXE, where either an isoleucine or valine is followed by a lysine, any amino acid residue, and a glutamic acid residue, appears twice within a 25 amino acid stretch and is highly conserved across the hormone receptor superfamily (lniguez-Lluhi and Pearce, 2000). Mutation of the first lysine (to glutamic acid) resulted in a 6-fold increase in rGR 1:1 activity (Table 3), while the mutation of only the second conserved lysine (with- in 25 amino acids) resulted in a 5-fold increase in 11 activity and mutation of both lysines resulted in a 12-fold increase in rGR ”c1 transactivation of a 63 reporter gene (Table 3). Due to the apparent synergistic increase in rGR activity, they called this motif [(lN)KXE-X12-(IN)KXE; where X represents any amino acid residue] the synergy control motif (SCM) and hypothesized that this motif is responsible for controlling GR from run-away transactivation of target genes. Additional studies are required to investigate the possibility that natu- rally occurring mutation clusters exist in the 1:1 region, including the SCM, of GR. This possibility is supported by the fact that 1:1 lies within the larger and variable immunogenic domain of the GR molecule (Okret et al., 1982; Gametchu and Harrison, 1984; Harmon et al., 1984; Hollenberg et al., 1985). Transcriptional regulation of glucocorticoid-responsive genes by GR is of particular interest in the bovine species due to the deleterious effects of stress- related increases in blood glucocorticoid concentrations on animal health and productivity (Filion et al., 1984; Mitchell et al., 1988; Nanda et al., 1990; Minton et al., 1994; Burton et al., 1995; Grandin, 1997; Kehrli et al., 1999; Preisler et al., 2000a,b; Weber et al., 2001; Burton et al., 2001). The glucocorticoid surge at parturition has been documented to induce immunosuppression, which is heritable (Detilleux et al., 1994). Additionally, bulls challenged with exogenous glucocorticoids display significant additive genetic variation, and modest to high heritability estimates, in a variety of immune traits (Table 4) including sev- eral neutrophil functions, lymphocyte blastogenesis traits, and the percentages of circulating T cell subsets (Tempelman et al., 2002; Abdel-Azim et al., in review; Burton et al., in review; Kelm et al., in review). Identification of biologi- cally-relevant mutations within 11 of bovine GR would not only allow for a bet- ter understanding of variation in the regulation of animal sensitivity to stress in high producing cattle, but would also provide important information to aid future biomedical studies on stress-related disease susceptibility. This was the goal of the current dissertation research. 64 .._m: 0.09:0: 0:. 00.:0E3000 :03; 08:90.0: 0:. 0:0 00.08.60 3:50. :0: .0 00:0: 0:. 950.. 05:28 9.2050: 0:. 5.3 6030:? 0..0:00 EmozEQm 0:30.36 0..0: 0::EE. 0.0.. 5:200 u:m:-:0. 0:: 00.08.50 3.09:0... EmocEQw o. .0009: .020: 9.0: 000:. .0 .60.). 099.20 0.00.:000036 0:0: -mmoxm 9.30:0. 0.:.: 50.0.0... .0 2.0: 0:255 030:0:5: :_ 0020000 :00: mm: :0_.m:m> 0..0:0m 02.600 .cmo:_:m_w .0.00.t0000:_0 0:0:0m0xm 5.? 0092.20 0.:.: 50.0.0: :_ 0208:»: 3:55:01 :.0...._. 0:0 3.0:... 0:358. 00.00.0m é 0.00... 65 :0_.0..0.5:.00 0.00.:000020 :0_.0..0_5:.00 0505000020 0. 02.0.0. 500.... :0 500:0000 0. 0.0:...00 5:50:01. 0. 02.0.0. .00 :0 500:0000 0. 0.0:...00 5.50501. 2.050. 5 ...0 .0 5.0V. .5050. 5 .._0 .0 5.0V. 2050. 5 ...0 .0 :53. 2.050. 5 .._0 .0 5.0V. .5050. 5 ...0 .0 :500. 2.050. 5 ...0 .0 5.0V. 0000 ...0 .0 :0E_00:.0... 2.050. 5 .._0 .0 :50v. .5050. 5 .._0 .0 5.0V. 2050. 5 ...0 .0 :0000 2.050. 5 .._0 .0 :0000 2.050. 5 ...0 .0 :0000 2.050. 5 .._0 .0 :0000 2.050. 5 .._0 .0 :0.50 2.050. 5 ...0 .0 00.50 00050 .000 0. 0..0 000000008005 .0 0.005800% __:00.502 .000 0. .00 3.2.00 00020500000.x0.000_0>:. 5:00.502 K00 0. 3.0 0:050 00.08.0000 .0 :0..0.0:00 550.502 00:05:05 .0 :0_. .000 0. 0..0 .0008 0:0 :.0:000.0:. 02.00.08 ....00.502 ES 0. .00 05.... 58.50.. 500880-555... .000 0.00.0 05.00.50: .0 500.05. :.00:00 00.0 0. 0..0 05:00.50: 05.05050 :0 050 .0 05000.03 < 5.0>0:00:00 :52. :0_. .000 0. 0..0 055.0 0530.5. 0.00:000005 05005055.. .000 0. 00.0 00050200.: 0.509.052 0000:05000 $00.0 0. .00 0..00 00.25055 :0 .. 01.2 .0 5.00053 000.0 0. .000 0..00 50.05805 :0 _ 01.2 .0 :0.000.0xm. _. 01.2 000.0 0.00.0 05000.90 0..00 .00.0=:0:0:. .0 000500.00 5.0000. :00 .. 0.. 0... 0000.0 0. 000.0 05000.0x0 0..00 ... 05.05050 .0 000500.00 0500.0... 5.0000000 000 0:. 0 50.0 0. 00.0 000.90 .05. 0..00 h 05.05050 .0 000500.00 05005:. 5.0000000 .50 0... 0.00.0 0.00.0 000.90 .0... 0..00 .. 05.05050 .0 000500.00 2.050. 5 .._0 .0 5.0V. .000 0. 0..0 0....005000 05.05050 .0 000500.00 2.050. 5 ...0 .0 :0000 0000.0 0. 00.0 0..00 0 0:..05050 .0 000500.00 00:20.01 0.05:0m 3:30.001 :0..0..0> 0..0:00 0:50.005 50.... 0:05.... .0_00...0000:_0 03509.0 5.; 0005.350 0.50 50.0.2. .5 00.08.00 5.33:0: 52: 0:0 0..0... 0:35.... 00.00.00 .0 200.. 66 CHAPTER 4. IDENTIFICATION OF MUTATIONS IN THE 11- ENCODING REGION OF BOVINE GR GENES A. INTRODUCTION Glucocorticoids have far-reaching effects in animals that include the sup- pression of inflammation (Almawi et al., 1996; Barnes 1998; Newton 2000), induction of gluconeogenesis (Cole et al., 1995; Hanson and Reshef, 1997), modulation of blood pressure (Whitworth et al., 2001; Mitchell and Webb, 2002), and promotion of cellular differentiation and development of various organs and tissues (Cole et al., 1993; Brewer et al., 2002), and regulation of behavior (Takahasi, 1996; Korte 2001). The end result of a glucocorticoid challenge is regulation of hormone-responsive gene expression in target cells (Bamberger et al., 1996). The receptor for glucocorticoids, GR, is composed of four main functional domains. These are the ligand binding (LBD), DNA binding (DBD), first transactivation (T1), and second transactivation (1:2) domains (Hollenberg et al., 1985; Giguere et al., 1986). In particular, T1 plays a critical and direct role in regulation of gene expression by recmiting basal transcriptional machinery and other important regulatory proteins to the pro- moters of glucocorticoid-responsive genes (Ford et al., 1997; Hittelman et al., 1999; McEwan 2000). GR uses numerous direct (DNA-binding) and indirect (protein-protein interaction) mechanisms to effect activation or repression of target genes in glucocorticoid-responsive cells. Thus, the 11 domain plays a central role in 68’s regulation of gene expression and is a good candidate for polymorphism studies. As with other members of the steroid receptor superfamily, GR’s individual functional domains are transcribed from specific exons. GR’s 1:1 is encoded by the second exon, called the GR2 locus. In humans, 1:1-encoding GR2 has 67 been shown to be polymorphic (de Lange et al., 1997; Koper et al., 1997; Huizenga et al., 1998; Dobson et al., 2001; see Table 3). Additionally, specific mutations in the GR2 locus of human and rat have been studied for transcrip- tional regulatory effects using mutagenesis. These studies have demonstrated that multiple mutations in a small region of GR2 have a greater effect on GR transactivation activity than any single mutation (Almlof et al., 1997), raising the possibility that there are functional surfaces in 11 that are disrupted by mutagenesis of two or more amino acid residues (lniguez-Lluhi et al., 1997). Amino acid changing mutations within GR2 could have far-reaching effects on glucocorticoid-responsive gene expression. In fact, additive genetic variation has been identified in bovine leukocyte traits known to be sensitive to gluco- corticoids (T empelman et al., 2002; Abdel-Azim et al., in review; Burton et al., in review, Kelm et al., in review; Table 4), indicating that the manner in which leukocytes respond to glucocorticoids is inherited in cattle. These variable sensitivities to glucocorticoid hormones are likely due to polymorphism in numerous pathway genes, including GR itself. If true, the study of polymor- phism in 1:1-encoding GR2 would be logical. The hypothesis of this study was that the bovine GR2 locus is polymor- phic. The objectives were to isolate genomic DNA from six cattle breeds, amplify the GR2 locus from genomic DNA, identify the presence of mutations within GR2, characterize corresponding amino acid changes, and predict if the identified mutations may effect the tertiary protein structure of the 1:1 region of bovine GR. 68 B. MATERIALS AND METHODS I. Genomic DNA from Six Cattle Breeds. Whole blood was collected from 50 Angus, 4O Polled Hereford, and 140 Holstein cattle from the Michigan State University Beef and Dairy Research and Teaching farms (Appendix C). Whole blood was also collected from 62 Brown Swiss cattle (Mashek Farm, Iowa), semen samples for 6 Brahman were generously donated by Dr. Melvin Pagan (Puerto Rico), and DNAs were obtained from semen of 10 Jersey bulls (generously donated by NorthStar Select-Sire, Lansing, MI; Appendix C). Genomic DNA was isolated from either leukocytes or sperm from these samples. 3. Isolation of Genomic DNA from blood. Whole blood was drawn into 6 ml vacutainers containing the anti-coagulant acid-citrate dextrose. Blood was transferred to sterile 50 ml conical tubes (Fisher Scientific, Hanover Park, IL) and erythrocytes lysed by addition of a 0.5 x volume of cold sterile hypotonic lysis solution (10.56 mM Na2HP04, 2.67 mM NaH2P04, at pH 7.3) and gentle inversion for 1.5 minutes. Then a 2 x volume of sterile ice—cold hypertonic solution (10.56 mM Na2HPO4, 2.67 mM NaH2PO4, 0.43 M NaCl, at pH 7.3) was added to restore isotonicity. The mix- ture was then centrifuged at 1100 x g at room temperature (RT) for 5 minutes and the supernatant discarded. The remaining leukocyte pellet was washed with 6 ml of sterile phosphate buffered saline (PBS; 0.13 M NaCI, 2.68 mM KCI, 4 mM Na2HPO4, 1.76 mM KH2PO4, at pH 7.4), followed by centrifuga- tion at 1100 x g at RT for 5 minutes and one wash with 3 ml of sterile PBS. After centrifugation (as previously described), the supernatant was discarded and the leukocyte pellet flash frozen at —80°C for 5 minutes to lyse the cells. Cells were thawed on ice for 15 minutes and refrozen at —80°C for an addition- al 5 minutes to ensure cell lysis. Following thaw, 25 pl of 10% dodecyl sodium 69 sulfate (SDS), 5 pl of 20 mg/ml Proteinase K (Roche, Indianapolis, IN), and 400 pl of proteinase K buffer (50 mM KCI, 1.5 mM MgCI2, 10 mM Tris-HCl, 0.5% Tween 20) were added to each sample to finish cell lysis and degrade proteins, and incubated at 55°C for 2 hours. The Proteinase K enzyme was then heat inactivated at 94°C for 10 minutes followed by incubation on ice for 1.5 hours. Next, 1.5 ml of sterile PBS was added, the samples vortexed, and centrifuged at 2500 x g for 10 minutes at RT. The supematants containing genomic DNA were then removed into sterile 15 ml polypropylene tubes (Fisher Scientific, Hanover Park, IL) and DNA precipitated by the addition of 300 pl of 3 M Na acetate (pH 5.2) and 4 ml of ice-cold 100% ethanol. Precipitated DNA was transferred to sterile 1.5 ml microfuge tubes and micro- centrifuged at 15,000 x g for 10 minutes at 4°C. Supematants were then decanted and the DNA pellet washed by the addition of 1 ml of ice-cold 70% ethanol. The ethanol was decanted and the DNA pellet dried resuspended in milIi-Q water to 100 ng/pl until use in the polymerase chain reaction (PCR). b. Isolation of Genomic DNA from semen. Semen straws were stored at —80°C until use for isolation of genomic DNA. At that time, semen was thawed and dispensed into a series of sterile 50 ml polypropylene conical tubes (Fisher; Hanover Park, IL), after which 350 pl of 30% STE [0.1M NaCl, 0.05M Tris base, 1 mM ethylenediaminetetraacetic acid (EDTA)], 55 pl of 20% SDS, 25 pl of 4.3% 1M di-thiothreotal (D'l'l‘; Sigma, St. Louis, MO), and 150 pl of 26% Proteinase K (20 mg/ml; Roche, Indianapolis, IN) were added. The mixture was vortexed, allowed to incubate at 60°C for 1 hour, and then twice the volume of Tris base-EDTA was added and mixed thoroughly. Twice the volume of phenol and chloroform (1 :1 ratio) was then added and mixed gently at RT for 10 minutes. The solution was cen- trifuged at 2,000 x g for 10 minutes at RT and the aqueous layers containing 70 genomic DNA transferred to new sterile tubes. Equal volumes of a solution containing chloroformzisoamyl alcohol (24:1 ratio) were then added to the aqueous layer and mixed for 10 minutes at RT. Tubes were centrifuged for 10 minutes at 2,000 x g at RT and the aqueous layers again removed and placed into sterile 15 ml conical tubes. DNA was precipitated by the addition of 1/10"1 the volume of 3 M sodium acetate (pH 5.2) and 2 x volumes of ice-cold 100% ethanol and the tubes rotated gently until DNA precipitates appeared. The DNA was removed into sterile 1.5 ml microfuge tubes, washed once with ice- cold 70% ethanol, suspended to 200 ng/pl in sterile miIIi-Q water, and stored at 4°C until use in PCR. Ii. PCR Amplification of GR2 Loci. GR2 loci were amplified from genomic DNA samples using PCR. For effi- ciency of GR2 polymorphism screening, a series of 200 ng/pl aliquots of DNA for 8 animals per aliquot from 5 of the 6 breeds (discluding Brahman) were pooled and the pooled samples PCR-amplified. Primers for the PCR reaction were designed based on DNA sequence for hGR exon 2 and were: Forward primer 5’-GGGACTGTATATGGGAGAGAC-3’ Reverse primer 5’-GGAC'ITTGAACTTCTCTGCTCGATC-3’ These primers encompassed the 1:1 core-encoding domain and most of its 5’ and 3’ flanking DNA in hGR exon 2, with an anticipated PCR amplicon size of 903 bp. Bovine GR2 was then amplified using the pooled genomic DNA samples previously described. To perform PCR, 200 ng of genomic DNA/pooled sample was combined with a PCR master mix containing 10 mM of deoxynucleotide triphosphates (dATP, dCTP, dCTP, dTT P; Invitrogen, Carlsbad, CA), 60 mM of Tris-$04 (pH 9.1), 18 mM of (NH4)ZSO4, and1.2 mM M92804. Then, 12.5 mM each of the forward and reverse primers and 2 pl of ELONGASE Enzyme mix [20 mM Tris-HCI at pH 8.0, 0.1 mM EDTA, 1 71 mM DTT, stabilizers, 50% (v/v) glycerol; Invitrogen, Carlsbad, CA] were added to the reaction mixtures for total PCR reaction volumes per sample of 50 pl. Ingredients were combined in sterile polypropylene PCR tubes and heated at 94°C for 2 minutes to totally denature the genomic DNA, followed by 25 cycles of 94°C for 1 minute (denaturing), 56°C for 2 minutes (annealing), and 72°C for 1 minute (elongation), using a Robocycler PCR machine (Stratagene, La Jolla, CA). A final 72°C elongation step was applied for 10 minutes. Aliquots (~5 pl) of the resulting PCR amplicons were then checked for size and single bands on 1.2% agarose checking gels with ethidium bromide staining and the remain- ing sample stored in milli-Q water at 4°C until GR2 polymorphism screening was performed. iii. PCR-Restriction Fragment Length Polymorphism (PCR-RFLP) Screening of 632 Amplicons. The PCR-RFLP technique was used along with 11 restriction endonucleas- es [Alul, Bfal, BstUl, Dpnll, Haelll, Hhal, Msel, Mspl, Nlalll, Rsal, Taqal (all from New England Biolabs, Beverly, MA)] for preliminary screening of GR2 polymorphism. In this protocol, 2.5 pl of PCR amplicons/sample were com- bined with 1 U of the respective restriction endonucleases and appropriate buffers that were provided with the enzymes. Reactions were allowed to incu- bate at 37°C for 12 hours and were then resolved on 2.5/4% stacked, nonde- naturing polyacrylamide gels. For polyacrylamide gel electrophoresis (PAGE), a 4% gel solution [80% milIi-Q water, 10% 10 x tris-boric acid-EDTA (T BE), 10% of a 40% (19:1) acrylamidezacrylamide bis solution; Bio-Rad, Hercules, CA] was mixed and placed under vacuum for 20 minutes (degassing), during which time the Mini-Protein ll electrophoresis chamber (Bio-Rad, Hercules, CA) was assembled. Following degassing, 0.10 x of 10% ammonium persul- fate and 0.015 x of N,N,N’,N’-Tetramethyl-1,2-Ethanediamine (TEMED; Bio- 72 Rad; Hercules, CA) were added to the gel solution, which was rapidly pipetted into the Mini-Protein ll gel rigs. The unpolymerized gels were covered with water and allowed to polymerize for 40 minutes at RT, after which time the water was removed and a second, 2.5% stacking gel [84% milli-Q water, 10% 10 x TBE, 6% of a 40% (19:1) acrylamidezacrylamide bis solution; Bio-Rad, Hercules, CA] was added. Gels were allowed to polymerize for an additional 40 minutes at RT. They were then covered with 1 x TBE and subjected to electrophoresis at 4°C. The digested GR2 amplicons were then combined with 4 pl of sample loading buffer and pipetted into individual lanes of the gels. The Amplisize molecular mass ruler (Bio-Rad, Hercules, CA) was added to one lane of each gel. Gels were electrophoresed at 70 volts for 4 hours at 4°C and PCR-RFLP banding patterns visualized using ethidium bromide staining and photographed on a UV light source (GelDoc 2000; Bio-Rad, Hercules, CA). For partial confirmation of differences in pooled samples revealed by PCR- RFLP, DNAs from individual animals making up the DNA pools were PCR amplified in the GR2 locus and digested for PCR-RFLP analysis as described above. The individual samples were amplified using the protocol outlined in Section ii above, digested using the restriction endonuclease(s) that revealed clear differences in the pooled DNA samples across or within breed, and ana- lyzed by electrophoresis and ethidium bromide staining for differences in band- ing patterns (as described above). Iv. Molecular Cloning of 632 Genes. To elucidate the precise mutations contributing to GR2 polymorphism, GR2 amplicons from five animals of each breed (including Brahman) were cloned and the DNA inserts sequenced. To do this, individual GR2 amplicons were purified using the Wizard Plus Minipreps DNA Purification System (Promega, Madison, WI) and suspended in milli-Q water to a final concentration of 200 73 ng/pl. These GR2 amplicons were then combined with 2 x of T4 DNA ligase buffer, 50 ng of pGEM-T Easy vector, and 3 U of T4 DNA ligase (all from Promega, Madison, WI) for molecular cloning of individual GR2 alleles. Ligation reactions were incubated overnight at 4°C and aliquots (2 pl) com- bined with 50 pl of 1 x 108 cfu/mg high efficiency competent JM109 Escherichia coli cells, per the manufacturers protocol (Promega, Madison, WI). Bacterial suspensions were incubated at 4°C for 20 minutes, heat shocked at 42°C for 45 seconds, and cold shocked (4°C) for 2 minutes. Room tempera- ture SOC media (2.0 g tryptone, 0.5 g yeast extract, 1ml 1M NaCl, 0.25 ml 1M KCI, 0.5 ml 1M MgCl2-6H20, 0.5 ml 1M MgSO4-7H20, 2 ml 1M glucose at pH 7.0) was added for a total volume of 1 ml. Bacteria were then incubated at 37°C for 60 minutes and 50 pl of the cell suspension spread on Luria agar plates containing 100 mg/ml of ampicillin, 50 mg/ml of X-gal dissolved in 5- bromo-4-chIoro-3-indolyl-b-D-galactoside, and 0.1M of isopropylthiogalactoside (IPTG, Sigma, St. Louis, MO). Plates were incubated at 37°C for 12-16 hours and a blue-white colony screening conducted. Where possible, nine white colonies per animal were individually picked and grown in 5 ml of Luria broth for an additional 12-16 hours. Where 9 colonies could not be identified, as many white colonies as were available were picked and grown as described. Aliquots of each colony were used to verify the presence of plasmid containing the inserted GR2 DNA (GR2 insert). To do this, one microliter of cells for each clone was combined with the EcoRI restriction endonuclease and its buffer (Invitrogen, Carlsbad, CA) and incubated at 37°C for 1 hour to release the cloned GR2 inserts, which were visualized on 1.2% agarose checking gels using ethidium bromide staining. 74 v. DNA Sequence analysis of Cloned GR2 Inserts. DNA sequencing of cloned GR2 alleles for each animal was performed on 3 sets of isolated plasmids, two that were purified from two separate clones per animal (and thus representing individual GR2 alleles) and one from the pool of 9 plasmids per animal (thus representing both GR2 alleles for elucida- tion of heterozygotes). The DNA sequencing reactions were prepared using 200 ng of plasmid DNA and the ABI PRISM Dye Pn'mer Cycle Sequencing Ready Reaction Kits (with AmpliTaq DNA polymerase; Applied Biosystem, Foster City, CA), for 15 cycles of 96°C for 10 seconds, 55°C for 5 seconds, 70°C for 60 seconds. This was followed by 15 cycles at 96°C for 10 seconds, and 70°C for 60 seconds. Sequencing reactions were concentrated by the addition of 0.1 x of 3M Na-acetate (pH 5.2) and 2 x of RT 95% ethanol, incu- bated at 4°C for 15 minutes, centrifuged at 4°C and 15,000 x g for 20 minutes, rinsed with 1 x of RT 70% ethanol, and the labeled DNA dried in a speed-vacu- um (Savant Speed Vac Plus, ThermoSavant, Holbrook, NY). The sequencing reactions were then electrophoresed on 4.75% denaturing polyacrylamide gels (40% 19:1 acrylamidezacrylamide bis, sterile water, urea) with the addition of 10 x TBE, 0.10 x of 10% ammonium persulfate, and 0.015 x of TEMED (Bio-Rad; Hercules, CA). Gels were allowed to polymerize for 2 hours, at which point 4 pl of sequencing loading buffer [1 :5 ratio of blue dex- tran and EDTA in sterile water with AmberLite resins (Roche, Indianapolis, IN) in forrnamide] was added to the samples. Samples were vortexed and heat- denatured at 90°C for 3 minutes. Sample mixtures were immediately chilled on ice and loaded into individual lanes of the sequencing gel. All cloned GR2 inserts in the individual plasmids and pooled plasmids were sequenced in both the forward and reverse directions, using forward and reverse M13 and M13 primers (Applied Biosystem, Foster City, CA). Fonrvard and reverse sequenc- 75 ing typically yielded 500 to 550 bp of usable GR2 sequence, with 100 to 150 bp of overlapping sequence for later alignment of full GR2s. vi. Analysis of 632 Mutations at the DNA and Predicted Amino Acid Sequence Levels. DNA sequences were analyzed using the Genetics Computer Group (GCG; Madison, WI) software. Following reformatting (‘chopup’ and ‘reformat’ commands) of the forward and reverse sequences, they were joined using the ‘assemble’ command. Multiple alignment files were generated using the ‘pile- up’ and ‘pretty’ commands for identification of polymorphism between alleles and a consensus sequence, which was the sequence of nucleotides that occurred with the greatest frequency across the population of test animals. Multiple sequence alignment files for DNA and predicted amino acid sequences were imported into an Excel spreadsheet (Microsoft Office v. x, Redmond, WA) for generation of graphs depicting position and frequency of identified mutations. vii. Visualization of 632 Mutations Using Three-Dimensional Protein Modeling. Following translation of DNA sequences into predicted amino acid sequences, the most divergent predicted protein sequences from the Holstein breed were submitted to the FOLD-DOE server at UCLA (http://fold.doe- mbi.ucla.edu). Predictions of a—helix, B—pleated sheet, loop, and turn struc- tures were thus obtained. The FOLD-DOE server was then used in conjunc- tion with Insight lI modeling software (Molecular Simulations Incorporated; San Diego, CA) to homology model the GR protein structures for divergent alleles from each breed. A crystallized protein structure from Chrysanthemum (pec- tate Iyase) with close secondary protein structural homology to GR2 was iden- tified by the FOLD-DOE server for homology modeling. This protein has been 76 identified by other researchers as having a secondary protein stmcture closely related to that of GR2 (Iniguez—Lluhi, personal communication). The amino acids of pectate Iyase were changed to reflect the residue sequence of GR2, for the selected alleles, using Insight ll software for visualization as a three- dimensional protein model. The GR2 models were then compared for structur- al differences resulting from GR2 amino acid changing polymorphism. C. RESULTS Images in this dissertation are presented In color i. P63 Amplification of Bovine 632 and PCR-3FLP Analysis. Due to the fact that the PCR primers used to amplify bovine GR2 DNA were designed from human sequence (Figure 8a), there was no guarantee that bovine GR2 could be amplified. However, the primers produced single bands in all animals across the breeds (Figure 8b) and, upon DNA sequence analysis, these were shown to be 92% homologous to the corresponding DNA sequence of hGR2 (Weber and Burton, unpublished; Figure 8c). However, PCR amplicons were 915 bp in length instead of the expected 903 bp, due to Figure 8a. Design of Primers to PCR Amplify Bovine 632. Exons 1 2 9a um 9r um {5 1 . _ .4 Eli: ' ‘ : l_ H t 1- im. ‘ .".Z.‘. ‘503 b;— Flgure 8a. Design of Primers to PCR Amplify Bovine 632. DNA sequence for exon 2 of the human GR gene (Hollenberg et al., 1985; Giguere et al., 1986) was used to design primers (see methods) for PCR amplification of the bovine GR2 locus. The primers used for this purpose were based on human GR2 sequence and were predicted to amplify a 903 bp fragment of bovine GR2 that included the 1:1-enoodng domain. 77 Figure 8b. PCR Amplification of Bovine GRZ Genes. Primers designed from human GR2 were used to amplify bovine GR2. In this figure, agarose checking gels show a series of single bands from the PCR amplification of bovine GR2 using the primers shown in Figure 8a and genomic DNA from representative DNA pools from 8 animals of six cattle breeds as template. Lanes M are the Amplisize DNA sizing marker; lanes N are negative controls (no DNA template added during PCR). In panel (I) the checking gel shows representative GR2 bands for the Angus (lanes 1 and 2), and Hereford (lanes 3 and 4) breeds, while in panel (ii), the checking gel shows representative GR2 bands for the Holstein (lanes 5, 6, and 7), Brown Swiss (lanes 8 and 9), and Jersey (lanes 10 and 11) breeds. The check gel in panel (iii) shows representative GR2 bands for the Brahman breed (lanes 12 and 13). These gels show that the primers designed from human GR2 were able to amplify bands of the correct approximate size from bovine genomic DNA. 78 Flgure 8b. Bovine 632 Loci Amplified by PCR. (i) Lanes: 915 bp ('7) 2 Kb 1.5 Kb 1 Kb 0.7 Kb 0.5 Kb 0.3 Kb 0.1 Kb 915 bp 915 bp 79 .oocoacmm NEG 5:5: 2 muoonEoc £on mm; 853% <20 «CO 338 $5 .__mLo>O .8833 Nam 58:: 60: umfiuoa an mom or: :9: 659 an m5 9299: 9m 3a 0.59". c_ c2623 mcoozaEm «am 9.38 on... 605303 <20 NEG cmEsc c_ 952 6c Axon .35 m 3 “6229295 #5 .6 Sam. mcioocoéoo F on. 55? 6.95. $8526. 208 8:59: ou_soo_o:c-mw m mango»: NEG 9.38 85 $505 E9593 85303 mi... .296. .938 Eon c_ .565 m_ 8533 SEE Amoromwooms c2388 xcmmcoov :95: new Am£>o£ EoEoI 95 comics woocoaaow <20 NEG Lo cowtmano ucm EoEcm=m 950% 9:0: 2:... 6:00.. «m0 o:_>om 05 co cocoaoow on Goad: mcfl>on c065: mcfi>on c055: mcfl>on c085: wcfl>on 002:0 ocfl>on cmesn mcfi>on G065: mcfl>on c0853 wcw>on .300.— «mu o£>om 05 *0 cocoa—vow <29 .8 0.59m 81 Figure 8d. Predicted Protein Sequence of the Bovine G3 11 Region. bovine human bovine human bovine human bovine human bovine human bovine human 1 60 GLYMCETETK VMGNDLGFPQ QGQISLSSGE TDFRLLEESI ANLNRSTSVP ENPKNSASTA -------------------------------- LK------ -------—-- ----S----- 61 120 VSAAPTEKEF PKTHSDVSSE QQNLKGQKGS NGGNMKLYTT DQSTFDIflrk RLQDLEFSSG ---------------------- H----T-T ----V----- ------— --------- 121 180 SPSKETSESP WSSDLLIDEN CLLSPLAGED DPFLLEGSSN EDCKPLVLPD TKPKIKDNGD --G---N--- -R ------------------- S ----- N -------- I ------------- 181 240 LILPSPSSVP LPQVKTEKED FIELCTPGVI KQEKLGPVYC QASFSGANII GNKMSAISVH -V-S---N-T -------------------------- T ------- P --------------- 241 300 GVSTSGGQMY HYDMNTASLS QQQDQKPIFN VIPPIPVGSE NWNRCQGSGD DNLTSLGTLN 304 FSGR _p__ Figure 8d. Predicted Protein Sequence of the Bovine GR “:1 Region. Comparison of predicted amino acid sequences encoded by GR2 between one Holstein (bovine) and human (translated from GenBank accession #BC015610.1). Dashes in the human sequence represent amino acids that are conserved between bovine and human 632. As was predicted from GR2 DNA sequence in Figure 8c, amino acid sequence alignment shows that the bovine sequence has a 4 amino acid insertion (bold lowercase letters) within the core of 1:1 (highlighted by a gray box) that was not present in human 11. Overall, the bovine amino acid sequence of bovine 1:1 was 92.6% homologous to the human 11 region. 82 the insertion of 12 nucleotides near the 5’ end of the 1:1 core-encoding DNA that are not found in hGR2 (Figure 8c). Following successful amplification of GR2 from pooled DNAs, amplicons were subjected to digestion with a battery of 11 restriction endonucleases. Digestion products were then electrophoresed and visualized by ethidium bro- mide staining for changes in banding patterns resulting from the mutation of various restriction endonuclease recognition sites. Differences in banding pat- terns were observed following digestion with three enzymes; Taql, Haelll, and Mspl (Figure 9a, b, c). Digestion with Taql resulted in banding patterns that tended to reveal differences between, rather than within, breeds of cattle (Figure 9a), while banding pattern differences following digestion with Haelll were exclusive to the Holstein breed (Figure 9b). In addition, Mspl digestions revealed banding patterns in Holsteins with varying degrees of band intensity (Figure 9c). Differences in the intensity of ethidium bromide staining in DNA pools indicate variable amounts of DNA resident within affected bands. As PCR amplicons from 8 animals were represented in each PCR-RFLP sample, variation in staining intensity between pools implied differences in the banding patterns of individual DNAs making up those samples. Changes in PCR-RFLP banding patterns were observed using Taql, Haelll, and Mspl, implying possi- ble polymorphism in bovine GR2 genes. In order to verify differences observed between pooled samples, individual DNAs from a selected pool (H3) were analyzed using PCR-RFLP for identification of differences between 632 in individual animal’s DNAs. Individual DNAs from Holsteins digested with the Mspl restriction endonuclease demonstrated several different banding patterns (Figure 9c), which were represented in the pooled sample from these DNAs (Figure 90, lane 3). These findings imply changes to restriction endonuclease 83 Figure 9a. Taql PCR-RFLPs Between DNA Pools of Five Cattle Breeds. Following PCR amplification of bovine GR2 from genomic DNAs pooled within a breed, amplicons were subjected to digestion with restriction endonucleases and electrophoresed on polyacrylamide gels. Possible differences in banding patterns (PCR-RFLPs) were visualized by ethidium bromide staining. Digestion of GR2 amplicons with the restriction endonuclease Taql revealed differences in banding patterns between breeds thought of as dairy versus beef. Lanes M are the DNA sizing ladder, lanes U are 632 amplicons from one Holstein cow that were left uncut, and lanes C are digested (cut) GR2 amplicons from the same cow. In Panel (i) the gel picture depicts banding patterns from Holstein (lanes 1, 2, and 3), and Brown Swiss (lanes 4, 5, and 6) pools, while in panel (ii) the gel picture depicts banding patterns for Angus (lanes 7 and 8) and Hereford (lanes 9 and 10) pools, as well as Jersey pools (lanes 11 and 12). This figure shows differences between the banding pat- terns for dairy (Holstein, Brown Swiss, and Jersey) versus beef(Angus and Hereford) breeds. 84 Figure 9a. Taql PCR-RFLPs Between DNA Pools of Five Cattle Breeds. Holstein Brown Swiss MUC123456 915 bp 500 bp 275 bp 125 bp 50 bp (”I Angus Hereford Jersey MUC789101112 915 bp 500 bp 275 bp 125 bp 50 bp 85 Figure 9b. Haelll PCR-RFLPs for Holstein Cattle DNA Pools. (0 M 1000 bp 700 bp 400 bp 300 bp 250 bp 200 bp f2: 2: 100 bp Figure 9b. Haelll PCR-RFLPs for Holstein Cattle DNA Pools. Digestion of GR2 amplicons using the restriction endonuclease Haelll revealed differences within the Holstein breeds. Lane M is the DNA sizing lad- der, lane U is a GR2 amplicon from one cow that was left uncut, lane C is the digested (cut) G32 amplicon from the same cow. This gel picture shows banding patterns from Holstein pool 4 (lane 1), pool 7 (lane 2), pool 8 (lane 3), pool 10 (lane 4), pool 12 (lane 5), and pool 14 (lane 6). As shown, there appears to be differences in banding patterns between Holstein DNA pools fol- lowing digestion with Haelll. 86 Figure 9c. Mspl PCR-RFLPs in DNA from Holstein Pools and Individual Animals within a Pool. Digestion of GR2 amplicons using the restriction endonuclease Mspl revealed polymorphism in the Holstein breed. Lanes M are the DNA sizing ladder, lanes U are uncut GR2 amplicons from one cow, and lanes C are digested (cut) GR2 amplicons from the same cow. In panel (l) the gel shows differ- ences in banding patterns between Holstein pool 1 (lane 1), pool 2 (lane 2), pool 3 (lane 3), pool 4 (lane 4), and pool 5 (lane 5). Holstein pool 3 (boxed lane 3) was particularly interesting due to differences in banding patterns and variation in band intensities. GR2 was amplified from the eight individual ani- mals making up Holstein pool 3 and subjected to PCR-RFLP, as shown in pane (ii). In panel (ii) the gel shows differences between Holsteins H2538 (lane 6), H3245 (lane 7), H3093 (lane 8), H3145 (lane 9), H3052 (lane 10), H2887 (lane 11), H2783 (lane 12), and H2622 (lane 13), the 8 animals consti- tuting pool 3. Differences between banding patterns from individual Holstein GR2 amplicons digested with Mspl may partially explain variation in Holstein pool 3 band intensities. 87 Figure 90. Mspl PC3—3FLPs in DNA from Holstein Pools and lndlvidual Animals within a Pool. (l) Pools of Holstein DNAs M uc c 1 2 E 4 5 r, .211- 1000bp'5j 900bp‘ 915bp 700bp 500bp <— 440bp 400bp“ 300bp <— 275bp <— 200bp 10° DP [‘4— 100bp (ll) Individual Holstein DNAs Used in Pool 3 MUCCIEIG 7 e 910111213 1500 bp 1000 bp 900bp 915bp 700 bp 500 bp 440bp 400bp 30° ”P 275 bp 200 bp 200 bp 100 bp 100 bp recognition sites in Holsteins and the other breeds and therefore the presence of 6R2 polymorphism. ii. DNA Sequence Analysis of Bovine 632 Genes. In order to elucidate the nature of possible 6R2 polymorphism observed by PCR-RFLP, cloned 6R2 inserts from 5 animals of each of the six cattle breeds were DNA sequenced and the sequences aligned for identification of single nucleotide polymorphisms (SNPs). Three GR2 sequences were gener- ated for each animal. Two individual clones as well as an aliquot of all clones (pools) were sequenced in both the forward and reverse directions for identifi- cation of individuals as homozygous versus heterozygous at each SNP posi- tion (Figure 10). This approach identified numerous SNPs in bovine 6R2 genes (Figure 11), which appeared to be much more complex in nature than observed previously using PCR-RFLP. Highlighted in bold red letters of Figure 11a are SNPs which fall within double-pass sequence while SNPs falling to the left and right of the red letters were generated from single-pass sequence. SNPs tended to cluster in the 5’ and 3’ sequences flanking the 1:1 core-encoding DNA of GR2 (Figure 11a), and several of these SNP were present as multiple allelic forms across the breeds (Figure 11b). Of the ani- mals studied, 19 were heterozygous and 4 were homozygous. Interestingly, several SNPs were found exclusively in each breed (Table 5); SNPs 6346A and T81 SC in Angus, A520T and A617T in Hereford, A5456 and 0578A in Brahman, C776 and A786C in Brown Swiss, C4416, T4540, A480T, A492T, A5290, and CB20T in Holstein, and T147C, A4726, A5106, A5316, and T5656 in Jersey. Therefore, DNA sequence analysis of cloned GR2 alleles revealed that the bovine GR2 locus is polymorphic with several breed-specific SNPs, although no particular pattern was attributable to a specific breed. 89 Figure 10. Identification of Heterozygosity In 632 via Analysis of Pooled and Individual Cloned 632 Alleles. TCACTTNGGGCAGT CACTTAGGGCAG CACTTGGGGCA (a) AIG peak 03) Insert 1 shows (c) Insert 2 shows as an N an A peak shows a 6 peak Pool of 9 Clones Insert 2 Figure 10. Identification of Heterozygosity in 632 via Analysis of Pooled and Individual Cloned GR2 Alleles. To identify whether an animal was heterozygous or homozygous at each nucleotide, two cloned 6R2 inserts and a pool of all cloned inserts per animal were DNA sequenced. Heterozygosity at a nucleotide position would be observed as an ambiguous base pair or ‘N’ in the sequence. In this example, the ‘N’ (in panel a) is an overlay of ‘6’ and ‘A’ peaks. GR2 insert 1 (b) con- tains an ‘A’ at the same nucleotide position, while 632 insert 2 (c) contains a ‘6’. This would confirm that the animal is indeed heterozygous at this nucleotide position, and thus that a SNP exists in this position. 90 Figure 11a. Summary of Mutations in 632 Loci Across Six Cattle Breeds. Following sequencing, alignments of 6R2 alleles for 30 animals revealed 47 unique SNP locations in the 915 bp bovine 6R2. In this figure, the vertical axis lists the animals’ identification numbers (see Appendix C for coding; A1, A2, A3 etc.) along with the allele (1 or 2) for a particular row of DNA sequence. The consensus GR2 DNA sequence (Cons) generated by 606 is shown at the top and bottom of the figure. The horizontal axis at the bottom of the figure shows the nucleotide positions in the 915 bp 6R2 fragments where SNPs were identified. Letters intermpting vertical lines are the actual changed nucleotides (relative to consensus). Four SNP locations (346, 377, 417, 429) fell within the T1 core-encoding region of GR2 (underlined in the consensus sequence). Highlighted in bold red letters are those SNPs which fell within double-pass sequence, while the remaining SNPs on either side were generated from single-pass sequence. 91 Figure 11a. Summary of Mutations in the 632 Locus Across Six Cattle Breeds. Con- c'rccccc1rccc r G 66 GG 666 ] cccc cc 3 G) ()0 0 0—0 H no—di 666 x; Get; éé cc 0) H U1 l NHNI—‘NI—‘Nl—‘NI—‘NHNHNI—‘NH coon ee.t one Cl: (D to I Nl—‘HI—‘Nl—‘NH A \ll Individual (1 or 2) 632 Alleles for 30 Cattle of Six Breeds III to r 0 Z) o——no——§»——nnnn G C G G T L TA 2 * * G TA A A T A w L H ‘J:C': nc I I IL I I lccc i0 COB. ‘TC ' ‘ ' ‘ ' ‘C A 'I ' W 5124711111133444444444SS55555555555566677778888 2167224579471245577891222346666789913602881236 387713677914924020089150135862873035363002 Nucleotide Position in the 915 bp 632 Inserts 'r'T Key: A = Angus, Br = Brahman, Bs = Brown Swiss, H = Holstein, He = Hereford, J = Jersey, Cons = Consensus sequence 92 .0000 E. 05 0000000 .05 <20 .0 30280000: .0 000 .m 002m E00050. 000 0050.0: 0... .0 Eu: 0:0 :0_ 05 0. 0.00 0__0m . 0.: 000003000 00:0 :0 .0 00060000 05 00 000000. 0.0 000000 00.00 0 .0 5000 c. 0.0EE0 m E0: .00_ Nam 00 05 05 0. 0020000 002w 20:0 .059. 0.5 c. 0000.6 0.300 5m 000.0< .004 «m6 0. 002m .0 E0EEam .0.... 0.50.“. 93 Flgure11b. Summary of SNPs in 632 Loci Across Six Cattle Breeds. salallv pasta emeo ssoaov sts :0 abetueoied T O In 94 Table 5. Table of 632 SNPs Identified Within Each of the Cattle Breeds. This table details each of the 47 SNP locations identified across the cattle breeds and elucidates which SNPs were harbored by which breed. The left- hand column lists the nucleotide position of the SNP and the remaining 6 columns indicate which breed harbored the SNP. Several SNPs appear exclu- sively in each breed (in bold). 95 Table 5. Table of 632 SNPs* Identified Within Each of the Cattle Breeds. Position of Brown SNP Angus Hereford Brahman Swiss Holstein Jersey 5 056 C56 12 T12A T12A 21 C216 0216 C216 C216 46 C466 C466 0466 77 C776 123 C1236 01236 01236 128 01286 01286 01286 147 T1470 157 C1576 C1576 01576 171 01716 01716 01716 193 C1936 01936 01936 C1936 01936 346 G346A 377 A3776 A3776 417 A4176 A4176 A4176 A4176 429 T4296 T4296 T4296 T4296 441 04416 454 T4540 459 T4590 T4590 472 A4726 474 6474A G474A 480 A480T 492 A492T 510 A5106 520 A520T 528 A5286 A5286 529 A5290 531 A5316 545 A5456 560 A56OT A560T 561 65610 65610,A 65610 65610 65610 65610,A 563 A5636 A5636 565 T5656 578 0578A 586 A5860 A5860 592 A592T A5926 598 A5980 A5980 617 A617T 633 A6330 A6330 A6330 A6330 A6330 A6330 660 66600 66600 66600 66600 66600 66600 703 67030 67030 67030 67030 67030 67030 725 67250 67250 67250 67250 67250 67250 783 A783T A783T *Differences are coded as the consensus nucleotide followed by the nucleotide position and the changed nucleotide, or SNP. 96 Table 5. (Cont’d) Position of Brown SNP Angus Hereford Brahman Swiss Holstein Jersey 786 A7860 813 T8130 820 0820T 830 68306 68306 68306 68306 68306 68306 862 68620 68620 68620 68620 68620 68620 *Differences are coded as the consensus nucleotide followed by the nucleotide position and the changed nucleotide, or SNP. 97 iii. Amino Acid Sequences Predicted from 632 DNA Sequences. To determine whether identified 6R2 SNPs alter 1:1 region predicted pro- tein sequences, DNA sequences were translated into predicted amino acid sequences. As expected from the large number of GR2 SNPs identified, multi- ple amino acid substitutions were detected in the predicted 1:1 region peptides. It is known that some amino acid substitutions are more disruptive to protein secondary and tertiary structure than others. Historically, these changes were categorized as conservative or non-conservative amino acid changes, such as T73 which represents a neutral to basic (respectively) change, or F153S which represents a nonpolar to polar change (Figure 12a, see Appendix B for residue classifications). Changes at amino acid positions 195 and 211 (high- lighted by an astrerix in Figure 12b) are located within a synergy control motif of GR2 which enhances transcription 6 to 12 fold when mutated (lniguez-Uuhi and Pearce, 2000). Changes indicated in bold red letters in Figure 12a are those generated by double-pass DNA sequence, while changes on either side are generated from single-pass DNA sequence. Several predicted amino acid changes were located within the 11 core (residues 101 to 145; underlined in Figure 12a and identified by hatched bars in Figure 12b), however, the major- ity of amino acid changes clustered in the amino-terminal and carboxy-terminal regions (Figure 12a) of the predicted protein sequences (Figure 12b). Therefore, it is possible that amino acid changes resulting from SNPs in GR2 genes could cause changes to the structure of the 1:1 region and possibly GR’s ability to regulate glucocorticoid-responsive gene expression. 98 Figure 12a. Summary of Changes in the Predicted Amino Acid Sequences of 632-Encoding 1:1 Regions Across Six Cattle Breeds. Thirty-seven amino acid residue locations were predicted to contain substitu- tions encoded by SNPs in the 6R2 loci shown in Figure 11. In this figure, the vertical axis lists the animals’ identification numbers (eg. A1, A2, A3, etc.) along with the allele (1 or 2) shown for a particular row of amino acid sequence. The consensus sequence (Cons) identified by 606 is shown at the top and bottom of the panel. The horizontal axis at the bottom of the panel shows the amino acid residue position within the predicted peptide sequence where amino acid substitutions occurred due to 6R2 SNPs. Letters interrupting vertical lines are the actual changed residues (relative to consen- sus). Three substitutions occurred within the 151 core (126, 139, and 143, underlined in the consensus sequence), predicted to lie between residues 101 and 146. Bold red letters indicate amino acid changes which fall within dou- ble-pass sequence, while changes on the left and right are changes generat- ed from single-pass sequence. 99 Figure 12a. Summary of Changes in the Predicted Amino Acid Sequences of 632-Encoding 1'1 Regions Across Six Cattle Breeds. Cons LNTDLALVNAfilfiAPSNKDKDKSSQKBEwKCHV VV Al-l v RE GV K. I] CPSZCL L A2-1 A2-2 A3-l I A4-l A4-2 i? A Brl- Brl- Br2- Br2— Br3- Br3- Br4- Br4- BrS- BrS- le- B31- B32- B32- 833- B33- BS4- BS4- % .5 'J L J rfi rm K0 "heir. I: L NHNI—‘NI—‘NHNI—‘NI—‘NI—‘NHNH 5) v 1 J C) 2% Az——-3»ao-3: ill Individual (1 or 2) 632 Alleles for 30 Cattle of Six Breeds :1: :1: 0 H 1: L NHHHNHNH N” G Full (V QQ rep I I H ——KK%KB&——6&——K N TSIL Ir SIL Cons 1111mm. {IMWJZSSQM'JKCHVQGHN 2471244455111111111111111111222222222 5613927234455667778889999012346677 693738040672783579610421217 Amino Acid Residue Position in the 1:1 Region Key: A = Angus, Br = Brahman, Bs = Brown Swiss, H = Holstein, He = Hereford, J = Jersey, Cons = Consensus sequence 100 00050. 000.5 .00 0 0000000000 0000000 000 .008 .00000 0.050 0 0.0...; 0.. 00.0.5 0000. 0.0 5.000 00 0 00.00.00... 0000000 0.00 E 000 000.00.. 000.00. .00_E.0..->x00.00 000 00.0.0 00.. 0. 0000000 0000.00. .00.. 000 00. 000 0. 0.00 0:00 0.00 E. 000 0.0...... 0030. 0000000 .5000 0.3 000 .00. .03 0000.000 030.00. .0 0.00 00000... .0000000 0.0..0 000010 0000.000 020.00. 000 0.5000 0.0-x 05 000 000000 00000000 0.0..0 .0 000000.00 00. 0.5000 0.x0.> 00 .. .2 050.... 0. 0020 0.00 0.0.. 00000.0 0000200030 030.00. 0.00 00.0.0 0..0..0 0.5000 0.30.. 0.0... 0000.5 0.000 .00 00000< 000.00... B. 00.0000m-~m0 *0 00.0< 00.E< 030.0000 0.... 0. 0000000 00 E0EEam .00? 050.". 101 'ILLZA V IL3A d3930 1 I930 1373A “783W 80330 * .L I I3)I 89031 C166 I3 GL6 I3 * N96 I)l >186 IO 9/888 IS S187 A38I| SLLIG N9L I)! IE)OLICI W179 I)I l09IN N89IS * 889 H BLVI SLSV )l39N V617A A8171 E) IVV A931 39 IO HLL )lI’W Flgure 12b. Summary of Changes in the Predicted Amino Acids of 632-Encoding I1 Regions Across Six Cattle Breeds , , , . . A3? 3 3 «‘9: 8 R 8 I”- 2 "’ 36MN 99318 emeo ssomv sefiueqo ,to abetueoeied 102 L13 IL3 393 I93 393 9‘83 033 I I3 903 66 I L6 I 96 I 86 I 88 I L8 I 38 I U. I 91 I OL I 179 I 08 I 89 I 89 I L? I \;_;\\\§j ‘ :0.“:n:.>-.;~~;>..<§c\ \ r~c ‘ > :x; 968 13 \ \ “I V93 H. 93 I L9 39 6? I? 83 SI Amino Acid Position in 1:1 Regions vi. Three-Dimensional Homology Modeling of Mutant 1'1 Regions Three-dimensional homology protein modeling is a powerful method for visualizing the effect of amino acid changing mutations on tertiary protein structure (Kuhn et al., 1990; Stemberg, 1996). Modeling entails identifying changes in key protein structures, such as a—helices and B—pleated sheets, and is based on the availability of crystal structures for highly homologous (or the same) molecules (Rice and Eisenberg, 1997). Since 63’s “:1 region has not been crystallized, a protein having a secondary structure closely related to that of the bovine 1:1 region was identified and subsequently used in the modeling process. For this work, the crystallized protein was pectate Iyase (Figure 13). Secondary structural predictions for both human and bovine 11 regions were determined. As predicted by secondary structural analysis (FOLD-DOE server; Rice and Eisenberg, 1997), the 1:1 regions of both human and bovine GR are predicted to possess fourteen B-pleated sheets and numerous loops, as well as four oc-helices, one of which is predicted to lie within the 1:1 core (highlighted by gray boxes in Figures 14a and b). The four additional amino acid residues observed within bovine 11 core (Figure 14a) results in an elon- gated or-helix when compared to the human 1:1 core (Figure 14b). In bovine 11, several SNPs previously shown to change amino acid sequences (Figure 12a) are also predicted to change secondary protein stmcture, shown in parentheses above the amino acid sequence in Figure 14b. These might be expected to alter tertiary protein structure (as viewed by homology modeling). Within the Holstein breed, homology modeling did predict some structural changes to B-sheets and loops in 1:1 regions of two animals with divergent 11 predicted amino acid sequences (Figure 15). In portions of the model, alter- ations in orientation and length of B—sheets and loops can be observed. Four areas containing notable changes in predicted structure are highlighted by cir- 103 Figure 13. X-Ray Crystallized Tertiary Protein Structure for Pectate Lyase. Figure 13. X-Ray Crystallized Tertiary Protein Structure for Pectate Lyase. The protein pectate Iyase (NDB Accession #1AIR; Lietzke et al., 1996) was identified as the crystallized protein with closest homology to bovine 632- encoding 11 based on predicted secondary protein structure (confidence value of 2.7). Therefore, this protein was used in the subsequent homology modeling of mutant bovine 11 regions. Pectate Iyase has been identified to contain 4 or-helices and 17 B-sheets. Loops and turns are found throughout the tertiary structure. 104 .x00 .00 0 0 0200.000 0. 0.00 B 00... 00.3. 000 0000. 030.00.30 000 0.0000 00.00.00 3 000.008 0 0.0.000 0. 00.0000 0. 0.0 000.30 .0 00.00. E. 00 .. 0202000000 000.. 000 00000.. 00 00.3. 000 0000. 000 .03000 00 0.0000 00.00.00 .0.000..>0 00 02.000 0.0 002.008 .0 000..0.00.0 5.6.00 00..0.00.0 .0.3.03..0 00000000 00. 0...... 2.0. 0000 .0 00. 00. 000.00 00.0.. 0. 00003000 0.00 00.80 00. 0.30.. 0.0. 0. .0.3.03..0 0.0.0.0 00000000 .0 02.0000 .0. 00......030 00.5 00.00. E. 00.>00 00. .0 .00. 0. 00000000000 00003000 0.00 00.0.0 .000 000.00. E 00.>00 000 000.30 0003.00 0.3.02.0 0.0.0.0 00.2000 0. 00000000 0000000 0. .00.0 0. 00.00... B. «.0 000.3... 0.... .0 02.0.00... .0.3.03..m 30000000 .03 0.30.“. 105 if!!! 0000000020000000020200000H00H020HumanooomqmdazznwmwH000009000>0H00H02 com 000 OHHZflUmhmdomw>m0AMMOHHAUQBUAMHNQNMQBM>OmAm>mmmmmAHAQUZQMHMQSHQmfl>AmSUQMZ.mmU wNN HmH qummflnwofldmeflUZMQHQQQmMBQMNmBMHmmmwmmhmflnNAHQhHmOQHHVAMEZUUZmUMOUMAZOOMmm> «mH bb QmflfifimhNMMHN4¢m>¢HmmBmNZH24HmNMQQMhQBMUWmAmHOUOOthANZUS>MHMBMUSHAU mh H 00.00: 3. m0 00.03... 00. .0 00.8.00... 3.3.03.5 30000000.. .03. 0.30.“. 106 .000..00-0 000. 0.00. 0000. 000 0.0000-.. .00.—.0 0. 000. 000 .72.. .0. 00003000 0.00 00.0.0 00. 0>000 0.000.00.00 0. 02600 0.0 9-0.... 00.0000. 0.0..0 000000 00. 0003.00 00003000 0.00 00.0.0 00.0.00.0 0. 0000000 .03 0.30.0. 000.30 .0. 00.0.00.0 .00. 000. 0030.00. 0.00 00.0.0 0 >0 .0000. 0. x._00-0 00. 0030000 .080 0.0 0 >0 00.00..00.0. 0.00 E. 00. 0.0...... 0.. 0. 00.0.00.0 0.0 0.0000-.. N 000 0.00.0 000 00.3. 000 0000. 030.0050 000 0.0000-.. 0. 002.00 -0 v 0.0.000 0. 00.0.00.0 00.0 0. 00 00.>00 .0 00.00. .9 00. .C .00 000.30 0.... ..>_0>..00000.. 000.. 000 00000.. 00 00.3. 000 0000. 000 .03000 00 0.0000 00.00.00 0000.08 00 0.5000 0.0 000..00-0 .0 000..0.00.0 00003000 0.00 00.0.0 00. 2.0.00 .3200 00..0.00.0 .0.3.02.0 00000000 00. 0..; .50. 0000 .0 00.00.000.00 00.0.. 0. .00.. 000 0: 0050.0 0. 7.... 0.0..0 0.0... 00003000 0.00 00.0.0 00. 0.30.. 0.0. 0. 0.3.02.0 0.0.0.0 00000000 .0 00..0.00.0 .0. 00...... -030 0.0.5 000003000 0000. .5 0.0206000 .0. 000003000 0.00 00.0.0 00.0.00.0 0. 0000000 .0 00..00....000. 00.32.00 00.000 0.. .00 00.>00 00. .0 00..0.00.0 .0.3.03.5 E0000005 .03 0.30.0 107 gift-SI»! ”MEGAWHAZQQUm¢OUMZmZNmUQmHmmHAZmHAMOQOOOMAmdfizznfimMHOUUWBHQUm>mHdeMZUHHZ won A>v A>V AS. mNN <0mhm¢0mfi>mUAMMOHHAUQBOJMHNQMMQBM>0Adm>mmmmmqHAQUZQMHMASBQQA>QQSUQMZ.mmUmqqm wNN AU. AM. AMva AZ. HmH mmamwdqmquUZHQHAAQmmkmmNmBWHmmmOmmmNQQNAMMMSHOthdnhfiwqMZZUUZmUNOUMAZOONmm> NmH bb QmmBMmhHMHHmddm>4HmdmZMmZMA>mHmNZA24HwNMAflMhQBMUWWQmHOUOOAMUQNZUSbMBMBMUZHAU m0 H .0200...— B. .06 00.>00 00. .0 00..0.00.0 3.0.02.0 30000000 .03 050.0 108 00.00. :- 0.m.0 .0 0.3.02.0 >.0...0. 00. .0..0 >00. 0020 00.0 00.00000 0.00 00.0.0 00.00 .00. 00:00.. 0.30.. 0.0... ..v 00.0. 0.0000-.. .00.0 .0 00..03.0.0 000 .0 00.0. 0000. .0 00..0.00 -..0 00. 0. 0000000 00 ..0.5 00 .8 000 P 00.0. 0.00000 .0 00.000. 00. 0. 0000000 00..0.000 .0 0030.0. .. 000.0 .30. 0.0 >0..0.50.00 .00.). .00.3.02.0 0.5. 00. 0005.00 .00.0...0 00 0. 00.0.00.0 0.0.5 0000. 000 0.0000-.. .0.0>00 .00.0..0 0.05 002.008 00 00300..< 0.5000 000 .0980 00.2.0 >0 00.00..00.0 0.3.02.0 0.0.0.0 >.0...0. 0. 0000000 00.0.00.0 0...5 0.30.. 0.0. 0. 0.5000 0.0 00.0..0 00. .0 0.0000. 00.0.0000 .0.300.00 000.00. E 00.>00 .0. 00.0.00.0 0000. 0...5 000>. 0.0.000 .0 0030.00. 0.00 00.0.0 00. 00.00.00. >0 00.0000. 0.0.5 ..00. 0.30.0 3.01 000 7:... 000003000 0.00 00.0.0 00.0.00.0 ..00. 0. 0.00020 .0220 0.0.0.0... 0.5..- 000.00. E- 00 00.>00 .0000. >00.00.00 0. 0003 00.5 000>. 0.0.000 000.00: E. .0235. .o 0.3.03.5 >.0...0._. 00. 0. 0000000 00.0.00.0 .0. 0.30.0 109 .000. .0... E. .00.:5. .0 0.3.03.5 0...5-r 0.... 0. mo . c000 00.0.00.0. .0. .0. 230.0 110 cles and boxes (Figure 15). Such changes in protein structure may be expected to influence the ability of bovine GR molecules to bind transcriptional coactivators and regulate transcription of glucocorticoid-responsive genes (Vottero et al., 2002). D. DISCUSSION The ability to PCR amplify the tl-encoding region of bovine GR genes (GR2 locus) allowed for the investigation of naturally-occurring polymorphism in populations of animals. Polymorphism was tentatively identified in GR2 using PCR-RFLP. However, since restriction endonucleases are insensitive to SNPs falling outside of their specific recognition sequences, a more rigorous investigation of GR2 polymorphism was warranted. Therefore, differences in PCR-RFLP banding patterns were used to generate preliminary data for a wider polymorphism search via DNA sequence analysis. in total, DNA from 28 individual animals was sequenced. Sequencing revealed 47 unique nucleotide positions harboring SNPs, 37 of which were predicted to change amino acid sequences. Of interest are 2 changes located within a synergy control motif (amino acid positions 195 and 211), and which enhance transcription by GR from 6 to 12 fold when mutated (lniguez-Lluhi and Pearce, 2000). in addition, several SNPs were found to be exclusive to specific cattle breeds (bolded in Table 5), which may be partially responsible for breed-specific responses to stress. The complexity of mutations identified in the GR2 loci of 6 cattle breeds by DNA sequence analysis rivals that of the bovine MHC II (DRBS) locus (van Eijk et al., 1992). The role of the highly variable DRBS locus is to give rise to MHC II antigen binding grooves capable of binding and presenting numerous anti- genic peptides to CD4 T cells that ultimately help trigger adaptive immune responses (Tonegawa, 1988). Similarly, the role of GR as a broad-spectrum 111 transcription factor requires that 1:1 interact with numerous transcriptional co- regulators and proteins of the basal transcription machinery. Variability of the 11 regions of GR that directly interact with transcriptional regulators may cause subtle differences in individuals and breeds in the expression of glucocorticoid- responsive genes, and thus of “stress phenotypes”. The GR2 SNPs identified in this study, especially those SNPs that changed amino acids leading to alter- ations in the tertiary protein structure of 11, may contribute to such variability in gene expression during stress. It has been documented that changes in 2 or 3 amino acids within a stretch of 11’s protein sequence may have a greater impact on the overall function of this transactivation domain than single amino acid changes (Almlof et al., 1997). Iniguez-Lluhi et al (1997) postulated that sequential amino acid residues form active surfaces within and surrounding the 11 core, surfaces that presumably provide contact sites for the multitude of transcriptional co-regula- tors that GR interact with. If true, the multiple mutations identified in this study could be more likely to disrupt such surfaces and affect 11’s transactivation function than any single mutation. Indeed, the GR2 SNPs were predicted to change several amino acids in the bovine 1:1 region which, upon homology modeling did appear to influence B-sheet and loop structures that could be per- ceived as surfaces. However, the tertiary protein modeling approach did not show such gross structural changes in most of the alleles modeled (see Appendix D for examples), suggesting that other changes to the 11 region (ie. charge) may have occurred instead. That said, the usefulness of homology modeling in this study was limited by the lack of a crystallized structure for GR’s 112 1:1. This is a particularly difficult problem since the 11 region of GR is unstmc- tured until it interacts with other transcriptional proteins (Sigler, 1988; Yamamoto et al., 1998). Thus, even if a crystal structure for 171 were available, it might represent only one of many possible structures for this key region of GR. Accordingly, other methods of assessing the possible biological significance of GR2 polymorphism in cattle are required. In the subsequent chapters this is done by estimating statistical associations between GR2 SNPs and heritable leukocyte and mammary health traits that are sensitive to GO. E. Concwsuous The bovine GR2 locus, which encodes the 1:1 region of the GR protein, is polymorphic in 6 breeds of cattle. Sequencing of GR2 genes revealed changes at 47 unique nucleotide positions, several of which occurred across breeds. Additionally, several SNPs were found to be breed-specific, including 1 found exclusively within beef breeds (Angus and Hereford), which will allow for exploration of QTL in a breed-specific fashion. Multiple SNPs identified in GR2 were observed to change 1:1 amino acid sequences and may therefore affect 1:1 ’5 charge, hydrophobicity or hydrophilicity, hydrogen bonding, or terti- ary protein structure. Protein modeling did not reveal significant changes in tertiary protein structure, however, this approach precluded for modeling of charge and bonding. Given our findings, it would be useful to develop a rapid PCR-based genotyping assay for future research in determining association between GR2 polymorphism and traits of biologic or economic importance to the cattle industry. 113 CHAPTER 5. MUTATIONS IN EXON 2 OF BOVINE GLUCOCOR- TICOID RECEPTOR GENES ASSOCIATE WITH LEUKOCYTE SENSITIVITY TO GLUCOCORTICOIDS A. lmnooucrlou Suppression of the immune system following adrenal release of glucocorti- coids has been extensively studied. The steroids effect profound changes in both the innate and adaptive branches of the bovine immune system (Kehrli et al., 1999). Changes include repressed expression of important trafficking pro- teins on neutrophils, antigen presentation molecules on monocytes and B cells, reduced numbers of circulating 78 T cells, and changed proportions of circulating B cells, CD4 T cells, and C08 T cells (Ashwell et al. 2000; Burton and Erskine, 2003). Of particular concern in cattle and other livestock species are glucocorti- coid-induced immune cell changes that lead to a reduced ability of leukocytes to migrate and recognize invading pathogens in infected tissue, thereby leav- ing animals vulnerable to infection and clinical disease. Natural immunosup- pression in dairy cows is striking following parturition (Detilleux et al., 1995; Burton and Erskine, 2003), during which time glucocorticoid concentrations are 7 to 10 times higher than basal concentrations (Preisler et al., 2000a; Weber et al., 2001). Parturient immunosuppression can be mimicked by treating cattle with the synthetic glucocorticoid, dexamethasone, which also causes reduced expression of CDBZL and CD18 adhesion molecules in neutrophils (Burton et al., 1995; Burton and Kehrli, 1995; Weber et al., 2001), reduced synthesis of numerous proinflammatory cytokines (Almawi et al., 1996; Nonnecke et al., 1997) and tissue receptors for these molecules (Cronstein et al., 1992; Yu et al., 1997), reduced numbers of circulating 76 T cells (Burton and Kehrli, 1996; Anderson et al., 1999), reduced expression of MHC II molecules on antigen 114 presenting cells (Burton and Kehrli, 1996), and reduced lymphocyte respon- siveness to mitogens and antigens in vitro (Oldham and Howard, 1992; Nonnecke et al., 1997). Milk production traits have also been shown to be responsive to glucocorti- coids, and are observed as reductions in milk yield following stressful hus- bandry events (Bruckmaier et al., 1993) and a reduction in production of milk proteins B-casein, a-lactalbumin, and whey associated protein (Reichardt et al., 2001; Houdebine et al., 1985). Milk somatic cell score (SCS), an indicator trait of mammary gland health, is a measurement of the number of somatic cells in milk, the majority of which in infected glands are blood-derived neu- trophils (Paape et al., 2000; Riollet et al., 2000). Neutrophil adhesion mole- cules are profoundly down-regulated following glucocorticoid challenge (Burton et al., 1995) presenting the possibility of transiently reduced migration of phagocytic neutrophils into infected mammary tissue and increased mastitis susceptibility (Burton and Kehrli, 1995). Therefore, the sensitivity of leukocytes and mammary cells to glucocorticoids partly determines health and productivity of dairy cows. The cellular receptor that mediates glucocorticoid effects is called the glu- cocorticoid receptor, or GR. It is expressed in bovine neutrophils (Preisler et al., 2000a; Weber et al., 2001) and mononuclear leukocytes (Preisler et al., 2000b) as well as in cells of most other tissues (Bamberger et al., 1996). GR is a ligand-activated transcription factor that regulates expression of glucocorti- coid-responsive genes (Bamberger et al., 1996). Upon glucocorticoid binding, cytosolic GR becomes activated to shed its accessory molecules, dimerizes with another hormone-activated GR, and translocates into the cell’s nucleus (Tsai and O’Malley, 1994). Once in the nucleus, GR interacts in various direct and indirect ways with regulatory DNA in and around glucocorticoid-sensitive 115 genes to regulate their transcription (Tsai and O’Malley 1994; Bamberger et al., 1996; McEwan 2000). GR is suspected to directly regulate gene expression of adhesion molecules such as CD62L on neutrophils (Weber et al., 2001), multi- ple cytokine genes (Gessani et al., 1988; Almawi et al., 1996; Brack et al., 1997; Nonnecke et al., 1997), cytokine receptors on multiple cell types (Almawi et al., 1996; Yu et al., 1997; Galon et al., 2002), and antigen presentation pro- teins (Burton and Kehrli, 1996; Jabara et al., 2001; Galon et al., 2002). Genetic variation in some of these glucocorticoid-sensitive traits has been observed in parturient dairy cows (Detilleux et al., 1995a,b) and dairy bulls used as breeding stock following exogenous treatment with dexamethasone (Tempelman et al., 2002; J.L. Burton in review; Kelm et al., in review). In fact, dexamethasone-induced immunosuppression in these bulls has recently been shown to be predictive of health status in their daughters, especially of mam- mary gland health indicated by milk SCS (G.A. Abdel-Azim in review). Therefore, genetic variation in infectious disease susceptibility in cattle may be mediated in part by genetic variation in the sensitivity of leukocytes to gluco- corticoids (Tempelman et al., 2002). The goal of the current study was to determine if GR genes of Holstein dairy cattle contain mutations that could partly explain genetic variation in leukocyte sensitivity to glucocorticoid challenge and relevant milk production traits. We focused our polymorphism search in GR gene exon 2 (GR2 locus) because this locus contains DNA sequence that encodes the 1:1 region of GR proteins (Hollenberg et al., 1985; Giguere et al., 1996). The 1:1 domain directly interacts with proteins of the basal transcription machinery and other transcrip- tion regulatory proteins (Ford et al., 1997; Hittelman et al., 1999) and is thus key to a cell’s sensitivity to glucocorticoids. The same Holstein bulls already determined to exhibit genetic variation in neutrophil and mononuclear leuko- 116 cyte sensitivities to glucocorticoid challenge in vivo (Tempelman et al., 2002; J.L. Burton in preparation) was the test population selected for this study. Our main objectives were to: (1) obtain DNA sequences for GR2 genes of 40 previ- ously phenotyped Holstein bulls; (2) identify GR2 polymorphism; (3) determine if GR2 polymorphism translated into amino acid sequence polymorphism; and (4) estimate statistical associations between GR2 polymorphisms and heritable traits of leukocyte sensitivity to glucocorticoids and milk production genetic merit of the test bulls. 8 METHODS i. Animals DNA from 40 Holstein bulls with known leukocyte phenotypes in response to glucocorticoid (dexamethasone) challenge was available for this study (see Appendix E). The leukocyte traits included in this study are listed in Table 6 and were: neutrophil expression of CD18; mononuclear leukocyte expression of MHC II; percentages of blood CD4 and CDB T cells; and the percentage of blood 8 cells. All traits were measured flow cytometrically before, during, and after dexamethasone was administered to the bulls (as described in Tempelman et al., 2002) using antibodies and immunostaining protocols reported by Burton and Kehrli (1995; 1996). ii. Isolation of Genomic DNA Holstein bull genomic DNA was obtained from ACD-anticoagulated blood samples donated for this study by Genex, Cooperative Resources lntemational (formerly 21st Century Genetics, Shawano, Wl). Extraction of genomic DNA from leukocyte pellets was performed by adaptation of a method described in Lahiri et al. (1992). In brief, non-DNA cellular components were precipitated and removed by addition of saturated sodium chloride and resulting DNA pre- 117 cipitated by addition of 100% ethanol. To lyse leukocytes for this DNA extrac- tion, approximately equal volumes of blood (5 mL) and solution A (10 mM Tris- HCI at pH 7.6, 10 mM KCI, 10 mM MgCI2, 2 mM EDTA in milli-Q water) were added to 15 ml conical tubes, followed by addition of 100 mL of 2.0% Triton X- 100 (Invitrogen, Carlsbad, CA) and mixed by inversion. Cellular debris and nuclei were separated by centrifugation at 1000 x g for 10 minutes at room temperature and the supematants removed. The resulting nuclei pellets were suspended in the original volume of solution A, followed by a repeat of the cen- trifugation and supernatant removal steps. For lysis of nuclear pellets, 0.8 pl of solution 2 (10 mM Tris-HCI at pH 7.6, 10 mM KCI; 10 mM MgCI2, 400 mM NaCl, 2 mM EDTA in milli-Q water) and 50 pl of 10% SDS were added and vortexed to suspend the pellet. This was followed by incubation at 55°C for 10 minutes, addition of 300 pl of 6 M NaCl, and mixing to precipitate proteins. Contents were transferred to 1.5 ml microcentrifuge tubes and proteins removed by centrifugation at 12,000 x g for 5 minutes. Supematants contain- ing genomic DNA were transferred to fresh sterile tubes and DNA strands pre- cipitated by the addition of 2 volumes of 100% ethanol at room temperature. Precipitated DNA strands were removed with a glass pipette and transferred into tubes containing 1 ml of ice-cold 70% ethanol. Tubes were centrifuged at 12000 x g for 5 minutes at 4°C. DNA pellets were dried and suspended in 0.5 ml of 10 mM Tris-HCl, 1 mM EDTA at pH 8.0 and incubated at 65°C for 15 minutes. Genomic DNA was stored at 4°C until use in PCR to amplify the GR2 loci. lll. PCR Amplification of 632 The GR2 locus was amplified from bovine genomic DNA using PCR as described in Chapter 4, Section Bil. Briefly, primers were designed from human GR2 gene sequence to include the 1:1 core-encoding DNA and most of its 5’ and 3’ flanking DNA. GR2 was PCR amplified using genomic DNA from 118 the 40 Holstein bulls and amplicons visualized on 1.2% agarose checking gels with ethidium bromide staining. PCR products were stored in milli-Q water at 4°C until molecular cloning. iv. Molecular Cloning of GR2 PCR amplicons of GR2 were purified using the Wizard Plus Minipreps DNA Purification System (Promega, Madison, WI) and suspended in milli-Q water to a final concentration of 200 ng/pl. These GR2 fragments were then ligated and cloned as described in Chapter 4, Section Biv. Aliquots of each selected clone were used in PCR to verify the presence of cloned GR2 inserts. To do this, 1 pl of the cells for each clone were pipetted into wells of a sterile 96-well plate (Fisher Scientific, Hanover Park, IL), which were incubated at 94°C for 10 minutes to lyse the cells. Then, 48 pl of PCR master mix com- prised of 10 mM dNTPs, 5 pl PCR mix [200 mM Tris-HCI (pH 8.4), 500 mM KCI], 25 mM MgCl+2, 12.5 mM each of forward primer (5’-AGTCCTTG- GAGCTCAGA-3’) and reverse primer (5’-G'l‘|'TATI'GTCAGGCAGC-3’), 36.5 pl MQ water, and 1 pl of Taq Polymerase (Invitrogen, Carlsbad, CA) were added to each well and the plate was subjected to 94°C for 10 minutes to denature the genomic DNA, and 25 cycles of 94°C for 30 seconds of denaturing, 53°C for 30 seconds of annealing, and 72°C for 1 minute of elongation in a MJ100 PCR machine (MJ Research, Reno, NV), with a final 72°C elongation for 10 minutes. PCR reactions were visualized for single bands on 1.0% agarose checking gels following electrophoresis using ethidium bromide staining to con- firm the presence of the GR2 inserts. Clones that were positive for GR2 inserts in the PCR were next used for DNA sequencing. v. 632 DNA Sequence Analysis Two of the picked clones per bull were selected for GR2 DNA sequencing. Additionally, aliquots of all picked clones were pooled for each bull and 119 sequenced for identification of heterozygotes. Sequencing reactions were pre- pared for sequencing by aliquotting 250 pl of appropriated bacterial cells into wells of 96-well plates (Fisher Scientific). The plates were shipped on dry-ice to Genome Therapeutics (Waltham, MA) for DNA sequencing of the cloned GR2 inserts in both the forward and reverse directions using M13 forward and M13 reverse primers to prime the reactions. Each sequencing reaction resulted in approximately 600 bp of usable GR2 sequence in both the forward and reverse directions from the cloned 915 bp GR2 inserts, resulting in 300 bp of overlapping sequence. Electropherograms for pooled clones were analyzed for heterozygosity by the identification of peaks representing individual nucleotides that were overlaid (as shown in Figure 10 from Chapter 4). Pooled and individual clone sequences were aligned and used for comparison and subsequent polymorphism identification. vi. Polymorphism Identification GR2 sequences were analyzed using GCG (Madison, WI) software. Following reformatting (‘chopup’ and ‘reformat’ commands) of the forward and reverse sequences, they were joined using the ‘assemble’ command. Multiple alignment files were generated using the ‘pileup’ and ‘pretty’ commands for generation of a consensus GR2 sequence (based on most frequently occurring nucleotide at each nucleotide position within the population of test bulls). Polymorphism was demonstrated when a nucleotide at a particular position in GR2 was different from consensus. The multiple sequence alignment files were then imported into an Excel spreadsheet (Microsoft Office v. x, Redmond, WA) for generation of graphs depicting position and frequency of identified SNPs vll. Defining 632 Polymorphism at the Amino Acid Sequence Level DNA sequences for each allele of each bull were translated into predicted 120 amino acid sequences using GCG’s “translate” command and multiple align- ment sequences generated using the ‘pileup’ and ‘pretty’ commands for gener- ation of a consensus sequence and identification of amino acid changes as described above for DNA sequences. Once again, multiple sequence align- ment files of predicted amino acid sequences were imported into an Excel spreadsheet and graphs generated as a way of visualizing the frequency and position of amino acid changes relative to consensus. viii. Study of 682 Associations with Glucocorticoid-Responsive Leukocyte and Milk Production Traits Several leukocyte traits exhibited significant additive genetic variation dur- ing recovery from dexamethasone challenge in vivo (Tempelman et al., 2002; Burton et al., in review), and these were used for analyses of GR2 SNP-trait associations. Heritability estimates for these recovery traits were modest to high, ranging from 0.06 to 0.56 (Table 6). The actual traits used for the asso- ciation analyses were neutrophil expression of 0018, mononuclear leukocyte expression of MHC II, percentages of CD4 and C08 T cells in blood, and the percentage of blood B lymphocytes. in addition, bulls’ genetic merit (as PTA) for milk yield, milk protein yield, and milk SCS was obtained from the NFL- USDA web site (http://www.aipl.arsusda.gov/). SAS was used to determine the best linear unbiased prediction (BLUP) of linear recovery for each of these traits. GR2 SNP or amino acid-trait associations were conducted in SAS using one-way analysis of variation (ANOVA). Results were plotted as ‘SNP position’ against the ‘P-value’ from the F-statistic, with a line of significance added across the plots where P = 0.05. The SNPs falling on or below this line were considered significantly associated with traits. 121 Table 6. Summary Table Depicting the Leukocyte and Milk Production Traits Studied, Including the Acronyms Used for Each Trait and Their Physiological Relevance. Leukocyte and milk production traits are listed in the left-hand column. The leukocyte traits were chosen because they exhibited significant genetic varia- tion in recovery from dexamethasone challenge in this group of 40 test bulls (Tempelman et al., 2002; Burton et al., in preparation). Acronyms used in this study for each trait are listed in the middle column along with an arrow indicat- ing leukocyte traits which were down-regulated following glucocorticoid admin- istration. The physiological relevance and heritability estimates of each stud- ied trait are briefly summarized in the right-hand columns of the table. 122 Table 6. Summary Table Depicting the Leukocyte and Milk Production Traits Studied. Leukocyte or Milk Acronym Physiological Estimated Production Trait Relevance Heritability Percentage of circu- by, CD4 Helper T cells; induction 0.23 to 0391a Iating T cells that express the CD4 accessory molecule Percentage of circu- i°/o CDB Iating T cells that express the CD8 accessory molecule Percentage of circu- We 8 cells Iating B cells Expression of MHC II iMHC II on mononuclear cells Expression of CD18 lCD18 on circulating neu- trophils Predicted transmitting Milk PTA ability for milk yield Predicted transmitting Milk ability for milk protein Protein yield PTA Predicted transmitting Milk scs abili for milk somat- ic ce I score of inflammatory, cell- mediated, and humoral immune responses fol- lowing recognition of for- eign antigen presented in the context of MHC II molecules on antigen presenting cells Cytotoxic T cells; kill altered host cells pre- senting foreign antigen in the context of MHC I molecules Antibody production upon activation by anti- gen and CD4 T cells Presentation of antigen to CD4 T cells Adhesion molecules used by blood neu- trophils for migration into infected tissue. Estimation of genetic merit for milk yield in daughters Estimation of genetic merit for milk protein yield in daughters Estimation of genetic merit for milk somatic depending on day relative to glucocorticoid administration 0.06 to 0.391a depending on day of gluco- corticoid administration ~O.151a 0.21 to 0.561a depending on day of gluco- corticoid administration 0.10 to 0.542b depending on day relative to glucocorticoid administration 0.133c 0.09C 0.17d cell counts in daughters bTempelman et al. (2002) dCranford and Pearson (2001) aBurton et al. (in review) 0De Groot et al. (2002) 123 C. RESULTS Images in this dissertation are presented in color i. Cloning and Sequencing of Bovine fl-Encoding 632. A population of 40 Holstein bulls, exhibiting significant additive genetic variation in several leukocyte traits in response to glucocorticoids (Tempelman et al., 2002; Burton et al., in preparation), were used to identify polymorphism in the GR2 locus. The full-length bovine GR cDNA has yet to be sequenced, therefore human GR cDNA sequence was used to design primers that would be expected to amplify a 903 nucleotide amplicon of bovine GR2 using PCR. These primers were designed to include DNA sequence encoding the 11 core as well as flanking DNA regions (Figure 8a). Figure 16 shows that the primers successfully amplified GR2 from bovine genomic DNA (six representa- tive bulls shown). Subsequent sequence analysis of the GR2 amplicon from a randomly selected Holstein cow, followed by BLASTn analysis, showed that bovine GR2 contained 12 nucleotides in the “c1 core-encoding region that are not present in human GR2 (GenBank accession #BCO1-5610.1; Figure 8c). Figure 16. Holstein Bull GR2 Amplicons. M N 1 2 3 4 5 6 1000 bp In“ a... ‘ ” m... m 2......«915 bp 700 hp at: ,, ~ 500 bp uni-ii ‘ Figure 16. Holstein Bull GR2 Amplicons. A 1.2% agarose checking gel Showing a series of single bands following PCR amplification of bovine GR2 from six representative Holstein bulls. The PCR products were 915 bp in length. Lane M is the Amplisize DNA Sizing marker; lane N is a negative control (no DNA template added during PCR); lanes 1-6 are single 915 bp PCR products amplified from genomic DNA templates of six Holstein bulls. 124 Thus, our PCR fragment was 915 bp in length rather than the 903 bp we expected based on human sequence. Overall, bovine GR2 DNA sequence was highly homologous (92%) to corresponding human GR2 sequence (Figure Be). When GR2 DNA sequence from the Holstein cow was translated into predicted amino acids, the peptide was 92.6% homologous to human pre- dicted amino acid sequence (Figure 8d). Thus we have demonstrated that the bovine GR2 locus is relatively homologous to human GR2, and translates into a homologous predicted amino acid sequence. ll. Identification of 632 Single Nucleotide Polymorphisms. In order to elucidate individual alleles for each bull, PCR-amplified GR2 DNA amplicons were cloned into the pGEM T-Easy vector and propagated in JM109 Escherichia coli competent cells. Several individual clones per bull, each containing a single allele, were selected and pooled for DNA sequence analyses (as described in Chapter 4). The Genetics Computer Group soft- ware was then used to generate sequence files for each of 76 successfully sequenced alleles (4 bulls had unacceptable sequence for one allele each), which were then compared for identification of SNPS. We observed differ- ences from the GCG-generated consensus sequence at 52 individual nucleotide positions within GR2 (Figure 17a), many of which had frequencies of 1.3% or 2.6% across the entire population (e.g., see nucleotide positions 24, 62, 77, etc. in Figure 17b). However, several SNPS at nucleotides 28, 289, 571, and 617 Showed higher frequencies in the bull population, with a discem- able “SNP cluster" lying between nucleotides 780 and 801 (Figure 17b). In Figure 17b, the hatched bars at positions 346 and 415 represent SNPS found within the t1 core-encoding region of GR2. All other solid bars to the left and right of these hatched bars represent SNPS found in the 5’ and 3’ (respective- ly) flanking DNA of GR exon 2 that surrounds the t1 core-encoding region. 125 Seventeen SNPS occurred in the 5’ flanking area and 33 SNPS occurred in the 3’ flanking DNA. ili. Identification of Changes in Predicted Amino Acid Sequences. GR2 alleles were further analyzed by GCG for potential influences of SNPS on predicted amino acid sequences. This analysis revealed that 29 amino acid residue positions were affected by SNPS in the GR2 locus (Figure 18a). These predicted peptide sequences showed mostly single amino acid changes (Figure 18b), including 11 that were considered conservative and 19 that were considered non-conservative. One amino acid position (residue 234) contained both conservative and non-conservative changes (M234V, M234T), depending on the bull (Figure 18a and 18b). Additionally, there were four amino acid changes involving proline substitutions (L35P, Q193P, P207l, P267H; Figure 18b), which could cause steric hindrance in the peptide regions of GR that flank 11. Of interest is a change at amino acid position 195, which falls within a synergy control motif of GR and which has been found to enhance transcription by 9 fold when mutated (lniguez-Lluhi and Pearce, 2000). It was interesting to note that neither of the two SNP locations occur- ring in the actual t1 core-encoding region (nucleotide positions 346 and 415 in Figure 17) translated into amino acid changes in the 11 core itself, which lies between residues 101 and 146 in the predicted amino acid sequences. In total, 29 unique 11 region amino acid sequences were identified in this Holstein bull population. Therefore, many SNPs in bovine GR2 are predicted to change amino acid sequence, which could result in changes to the protein structure and charge, possibly affecting the transcription regulatory functions of GR. The 171 core itself did not contain any predicted amino acid residue changes in this group of Holstein bulls, suggesting that this critical functional domain of bovine GR is relatively stable and intolerant to change. 126 Figure 17a. Summary of Single Nucleotide Polymorphisms (SNPS) Observed in the GR2 Locus of 40 Holstein Bulls. Fifty-two SNP locations were identified in the 915 bp bovine GR2 PCR ampli- cons that were cloned for DNA sequencing of individual alleles. In the figure, the vertical axis lists the bulls’ identification numbers, from least to greatest, (eg. 4536, 4591, 4628, etc.) along with the allele (1 or 2) for a particular row of DNA sequence. The consensus GR2 DNA sequence (Cons) generated by GCG is shown at the top and bottom of the figure. The horizontal axis at the bottom of the figure shows the nucleotide positions within the original 915-bp GR2 fragments where SNPS were identified. Letters interrupting vertical lines are the actual changed nucleotides (relative to consensus). Two SNP loca- tions (346, 415) fell within the t1 core-encoding region of GR2 and are under- lined in the consensus sequences. Highlighted in bold red letters are those SNPS which fell within double-pass sequence, while the remaining SNPS on either side were generated from Single-pass sequence. 127 Figure 17a. Summary of Single Nucleotide Polymorphisms (SNPS) Observed in the GR2 Locus of 40 Holstein Bulls. 1((( 'll ‘ &"T I “10101 (‘ll'x '11 (‘(~ rt" ‘ l. “i (3 It It (3 It hlblbbbbhfihkfihbhkbfibhfihkbfin \tqqqqqqqqqmmmmmmmmmmmmmmmo biblbih \le mm WV 5) n “I N r A u; mwmmfihmwpHqquqqmmNNHH4qmmwwmeHHmmqqmmmmeHHquqmmHHmmmmHHm I NHNHNHMHNHNHNHNHNHNHNHNHNHNHNHHNHNHNHNHNHNHNHNHNHNHNHNHNHNHH 'l J ‘I J L G *_1AA AA(APAAAPAP'QATCCALAL1'lr an 23444455555555666667'777778888888888 02334445 1841499346777891121700355890112244585 538905746 79652261521896737155123490911:0818115 Nucleotide Position in the 915 bp GR2 Inserts _ fl OO ‘ 1KK‘ I l (A: l .679 1111 7 n ##fibfifibkfibifibfihbkbbhhifib O E emu—m {I'll—3 U" 128 Figure 17a. (cont’d) bibiblb mxmooooooooo pppbwwmmmoooo I I I qmmHqummmmmw I NHHNHNHNHNHHNHNH b \0 lb lb ‘0 \O I I an mm mm TH Cone AAF“.JgfiATTGAGGGthATAACACAAACACTCA ,CAP TTA- in“. - ' EML'IIMCACAMCAL . L ' L I I I I I L I 2267911111111122234444555555556666677777778888888888 4827502334445504841499346777891124700355890112244589 53890574607965226152189673715512349091130818115 Nucleotide Position in the 915 bp GR2 Inserts 129 ..o\oo.m 2 some >0 -5309. 8:9: c. c9580 Pom ucm .omw K5 .Em .mmm .mm 20:80. ..m mmzw .8526... .o\oo.m cam «on; .0 36:26 -9. 26. c. 8.580 mdzw cozzcog .0 3.8.5:. on“ 85 93% 059. «E... .562 9685.28 3 m5 mcicmz <20 39:889. .m Em .m as g 28. mmzm 250% 92 86.2 9.. .0 29. new :2 2. 9 22 28 9: .mmo .0 of. .0925 mo_m__m __m .0 83588 9: mm commoaxm 2.3 Sofie... 910 .8. mac an m5 on“ c. anmeo mdzm o__o__m .o 8.9.039". .m..:m £220... 9. .o cote-anon a :. 3.2.032“. nzm «m6 no 22555 .2: 0.59... 130 Figure 17b. Summary of GR2 SNP Frequencies in a Population of 40 Holstein Bulls. : \w: &x\\\\\\\\\\\\\\\\\\\ < .\‘- \ New ‘ ‘ \ ~\\\~'\'\\\\\\\\\\\\\\\\\\\\\’ >.\\ v;\\ -. 'T 6 ID V M N 1- O salaliv "n9 ugetS|oH SSOJOV Sst 40 afietueoied 131 .oocmacom $3-996 E9. “.9995... 995 9.6 .9...o co mdzm 9.2929 9.. 2...; .8533 $9.. .9928 $5.2. =9 no.2; mmzw mmoc. 9w 99.9 v9 v.9. 2. cozozcmf .999. m< .03 new For 82.69 $95.9. m.._ o. “9069.. 99> 20...; .98 E .958 9.. 55.2, .9580 32.3.9.3 Bum 9...»... oz .Amzmcmmcoo 0. 9.6.9. 332.09 3359.0 .958 9.. 9m 99... .mo.tm> 9.32.9... 99..."... .mmzw «mm 0. on... .9580 32.2.9.3 Bum 8.6... 99.2, 85:03 9.39. no.0.u9a E. ..3. 9.. 55.2. 9.2.69. 9669 Bow 056m 9.. 26...... 29.9. 9.. .o Eozon 9.. .m 8.8 .9899. 9.... .929. some .0 Eozon ucm no. 9.. .m 5.6% m. 000 .3 nozzcoc. .980. 8539. 9659.8 9.... .8533 Eon 02:8 .0 2,9 8.3.th m .5. E265 Am .0 5 99... 9.. 5.2. 9.2m 69¢ .mmwv .mev .ommv do. $9890 0. .98. E9. .99.:5: co..mo....cmu. .232 9.. .m._ 99.9. 59. :. mmxm .mo.to> 9.. .939. m.... c. .0. 9:9“. c. .2505 .8. «CO 9.. :. wdzm >9 caboose 22.3.9.3 59:8 0. c9069.. 99... 9.2.9.0. 9.2.0.9 Bow o:.Em o:.c->.cm>>._. .m..:m 5990... ow .0 5.3.2.0“. o c. :0 .o .339“. B. 9.. .0. «09.958 .u.o< o:.E< 020.35 9... :. notomno Em._.n._oE>_om .o EaEan .aww 05m."— 132 C .9 m 0 .C a .E HNHNHNHNHNHNHNHNHNHNHNHNHNHHNHNHNHNHHN c llllIlIlIlIIllllllIlIllllIIIllIlIlIlIl O NHHNNQWFmebbfll-tuiulu VVI:I\ VI. Vu H . - [\HH‘OKOPHH a I: gNNmmmmmmvvmmmmmmbbbr 'mlzzlzflfl 3 O a 6vv¢vvvvvvwvvvvvvwwvvvwvvvv¢v¢¢w¢¢vvv¢dU I-II-‘INHNr—Ir-h-INr-Ir—II-INHHHNHNHNHNHNHNHNHHNHNHNHN lllllllllll llllllll Illlil IIIIIIII Illll IDHHmmmmr-h—ILOLDDFmeHNNmkotokobbLnLnr-IHDCDGDMMKOUDFl\ I fimmmmmmeoxoxokob[\mmmmmmmmmmmmxbmb[\bbbmmmmmm fl OmanKDUDKOlOWKOQKOLOKOKOWF[\FFFDFFDFFFFFFPFFFFFI‘F O UV‘fi'fi'fi'flfi'fi'Vfifi‘fi'fi‘fl'Vflfi‘fi'Vfi'vfi'Vflfi‘fl‘Vfld‘V‘V‘fi‘fi‘V‘VV'Q'V‘V‘U s"“8 uISISIOH 07 10! salauv 389 (Z 10 L) IBnPMIPUI Fl ure 18a. Summary of Polymorphism Observed in the Predicted Amino Acid Sequences for the 1:1 Region of G: in a Population of 40 Holstein Bulls. 133 .2033. 33 Ba .93. com .3033. 38 303.3. 33 396.39 .3 69.38 9.2.3.333 6.33 oc.E3 322333 263.69. .33E 9.. .3... 32.9.3 9:9. 3...... ....3E .9603 28.9.3 3 2....2. 32:3. 33 39.9.3 2:. 33....232 mm. .2239. 233 oc.E3 3533 .9933 9.... ..9. ..333 3633 .320 93 .2632 96.39 ..333 .3 9.2.2333 6.33 SE... .3:.o< 3399.3 22.3 333... .o 32:39. 96.39 9.. 32.2.3 23* 9.. 6:3 63.33.36 3833...... 3.2.3 .3 333.3933 9.. 32.9.3 3.3..» 9.... 63.9.3 33.2.3 ..3 .3 3339.399. 33 6333993 6:3 2 9:9". c. 3n.zm mam E9. 63.2693 3:25.333 96.39 233 oc.E3 2.2.3.3 33.9.3339. 32.9.3 9:9. 3.2... .355 2.3.3.0... 32 E 3:93: 5. me .o 36.o< o:.E< 63.3.63... 9.. :. 3399.0 .0 33.39.69“. 35 .o EuEEam .33 359". 134 ugatsloH SSOJOV sefiueuo ;o afietuaoJad O Q .5 WGZJ. voz § 97888 we: g. GSBZN sea I 00933 on: I F {3 1.97. 5 E A0988 09?. ‘6 98980 as: i OLSZH L98 3 NLVSZW m g :IQZZS 9:: g IOLZd 1.0: g V9031 90?. .§ 066L3 an E * N960! S“ t )iSGLO set i lL8LL m '3 A88“ am g: 9am m 2 avg»: m 2 N96)i 96 3 H880 as '3 saw as G g (MB is g VGVA 617 ‘3 sun. 1.7 5 19178 99 "g Ava ' W a d98'l ‘39 E 983 a 3 T 1 T T 1 . fl 0) h 0 In 1- n N v- o a saleuv line 2 3 U, i: Holstein Bulls. 135 Amino Acid Position in Predicted t1 Region Protein Sequence iv. Associations of Identified GRZ SNPs with Glucocorticoid- Responsive Traits. In order to estimate potential associations between GR2 polymorphism and inherited glucocorticoid-sensitive leukocyte traits or milk production poten- tial in the 40 test bulls, the MIXED procedure of Statistical Analysis System (SAS; Cary, NC) was used to periorrn one-way analysis of variance (ANOVA), which was used to statistically associate individual SNPs with the battery of traits shown in Table 6. These included 6 glucocorticoid-sensitive leukocyte traits [percentages of circulating T lymphocytes expressing CD4 (% CD4) and 008 (% CDB); percentage of circulating B lymphocytes (% B cells); neutrophil expression of the adhesion molecule CD18 (CD18 expression); and mononu- clear leukocyte expression of MHC II molecules (MHC II expression)]. Each of these leukocyte traits responded to glucocorticoid challenge in the test bulls and possessed significant additive genetic variation in the recovery phase fol- lowing termination of the steroid treatments (Tempelman et al., 2002; Burton et al., in review; Table 6). Therefore, the recovery phase of these leukocyte traits were used in the analyses of SNP-phenotype associations. Genetic merit (pre- dicted transmitting abilities, or PTA) for the milk production traits were obtained for each bull from a web site housed at the Animal Improvement Programs Laboratory of the United States Department of Agriculture (AlPL-USDA, http://www.aipl.arsusda.govl). The PTAs studied included milk yield, milk pro- tein yield, and milk SCS (Table 6). These traits of genetic merit (PTAs) were weighted against their reliability estimates (the number of daughter records which were used to generate the milk production PTAs) before SNP-trait asso- ciations were estimated. Results of the association analyses are shown in the panels of Figure 19, which presents a series of trait plots where the x-axes are the nucleotide posi- 136 tions of identified GR2 SNPs and the y-axes are the P-values from F-tests per- formed during the one-way ANOVAs. A total of 13 unique SNP positions were identified as associating significantly with the listed traits at P S 0.05 (Table 7). In Figures 19a, b, and c, SNPs at positions 24, 77, 95, 105, 145, 200, 617, 754, and 813 significantly (P s 0.05) associated with °/o CD4 (Figure 19a), % B cells (Figure 19b), and MHC II (Figure 19c). Additionally, several SNPs were found to associate significantly with multiple traits. These SNPs are highlight- ed by solid black data points in the panels of Figure 19. For example three SNPs at nucleotides 621, 780, and 820 associated significantly (P .<_ 0.05) with MHC II (Figure19c) and CD18 (Figure 19d), and a SNP at position 139 sig- nificantly associated with both CD18 MFI (Figure 19d) and % CDB (Figure 19e). Summarizing across Figures 19a through 19c, SNPs at nucleotide posi- tions 139, 621, 780, and 820 in GR2 may be important to GR function because each influenced the phenotype of at least two glucocorticoid-sensitive leuko- cyte traits. It was noteworthy that neither of the two SNPs occurring within the I1-encoding DNA, at positions 346 and 415, were associated with any ana- lyzed traits. This may have been due to the fact that these SNPs did not change amino acids (Figures 17 and 18). SNPs at GR2 nucleotide positions 621, 780, and 820 were also shown to associate (P s 0.05) with a bull’s predicted transmitting ability for milk yield (Figure 191‘), but not milk protein or milk somatic cell score PTAs. Interestingly, these SNPs also influenced MHC II expression (Figure 19c) and 0018 expression (Figure 19d; Table 7), two traits that are critical for early inflammatory responses that clear intramammary infections in dairy cows. Following translation of GR2 DNA sequences into amino acid changes, these changes were analyzed to determine whether changes to the GR protein would also associate with traits of immune dysfunction. Changes at amino 137 Figure 19. Associations Between Individual GRZ SNPs and a Variety of Glucocorticoid-Responsive Leukocyte Traits and Milk Production PTAs. In each panel, P-values, on a log10 scale, from one-way ANOVAs of SNP-trait analyses are given on the y-axes, with a horizontal black line representing sig- nificance at P s 0.05. The x-axes in each panel show nucleotide positions within the GR2 locus where SNPs were identified. SNPs that associated with more than one trait in panels (a) through (f) are indicated with solid black cir- cles (.), SNPs that associated with only the trait in a given panel are indi- cated by hatched circles (@), and SNPs that were not associated with any measured trait are indicated by open circles (O ). (a) % CD4 T cells; (b) % B cells; (c) mononuclear leukocyte expression of MHC II; (d) neutrophil CD18 expression; (e) % CDB T cells; (f) milk yield PTA, (9) milk protein yield PTA, and (h) milk SCS PTA. SNPs at nucleotide positions 139, 621, 780, and 820 all associated significantly (P s 0.05) with multiple leukocyte traits (listed above) as well as with milk yield (nucleotides 621, 780, 820). 138 Figure 19a. % CD4 Cells 1.0-3 T (ET 0.50 « 0.” . 0.10 4 0.05 P - Value G 0.02 . 0.01 I I III e I I I III I ‘ I III I I 1 Trr I' rfi' 7" I I I ITI I I 1' III I I I III I I T I'T 383§§§ §§§§§33§§§ III—i 6!" Position of Single Nucleotide Polymorphisms Figure 19b. % B Cells 1.0 1 1 4 00m ‘ 1 0.20« 0.10 1 0.05‘ 0.02‘ P - Value 0.01 « is °s§é§s§§§§$§§§§§§ s Position of Single Nucleotide Polymorphisms @ = Associates with only this trait . = Associates with 2+ traits O = No association with any traits 139 Figure 19c. MHC II P - Value 1.01 m, A; 1 0.20‘ 0.10 1 0.05 ‘ 1 0.02 . 0.01 1 1 mmfit 9 0.002 1 0.001 é II'II'I °88 d Figure 19d. CD18 P - Value III I IéI I ‘I r III I I I I I III I r I III I rrIT' I ' I IéI I l I I'I I -§§§§§§§§§§§.§§§ Position of Single Nucleotide Polymorphisms 141: 010‘ ° 0 4 1 0.02 . 0.01 ‘ 1 d 0.005 1 4 l 0002‘ . 0.001 ‘ I I'I r I Frr r j—rT‘IT r 1 1 .1. "TIVUTTII‘ 1*1 #11 . r" 'r' ' 1 TI‘ TTII' i ' 'rr ' I °3§§§§§§§§§§§§§§§§§ Position of Single Nucleotide Polymorphisms @ = Associates with only this trait . .-_ Associates with 2+ traits Q = No association with any traits 140 Figure 19a. % 008 Cells 1.0 1 0.50 1 i 0.20 . 0.10 1 0.05 ‘ 0.02 . P - Value 0001 1 1 1 1 0.005 . 4 0.002 ‘ 0.001- r III I I—rrTI I I1I I III I l I III I I °3§§§§§§§§§§§§§§§§§ Position of Single Nucleotide Polymorphisms Figure 191. Milk PTA 1.0 1 1 0.501 4 0.20' 0.10 - P - Value 0.05 0.02 . 0 CC 0.01 I I rII I I I III I I I II I I I °8§§§§§§§§§§§§§§§§§ Position of Single Nucleotide Polymorphisms @ = Associates with only this trait . = Associates with 2+ traits O = No association with any traits 141 Figure19g. Milk Protein PTA 1.00 '3 . . . 0.90 1: 0 o 0.00 Phi 0.70 0.60 ‘: o o 0.50 ‘ 0.40 ‘. 0.30 ' 020 0.10 P - Value rII'II ITII'II'I lll‘lIll‘Illl'Tj] jjll'II'Il'Il' =00'0'000'0'000000'0 0000 Position of Single Nucleotide Polymorphisms Figure 19h. Milk scs PTA 1.00 -, 0.90 9 Q 0.00 0.70 0.60 g 0.50 0 . 0.40 g 0.00 3 \L J) I 0 g 0 c 0.20 O . . 0.10 0‘ TT I l =0 '000 0"0 .. ”0'00 0'00 0. 00 Position of Single Nucleotide Polymorphisms O LLlllAllllll P - Value AA. _ - O - @ = Associates with only this trait . = Associates with 2+ traits O = No association with any traits 142 3::3 m:.:.3:.9 3:. :3:. 3:33 .. 300 .o 3339:3933 3:9: 33: 3:33 .. .33 0.. 5.2. 35.3.8333 2.332.333 3:23 33.23.83 3:3 $3338.95 .333 w n. .3 3.0.. 52. 33.3.8333 2.333 -...:m.3 3.32. 332w :33..:3n. .>.3>..33333. .3. 3:3 .3. 3.3..33.33:3 >3 33.33.3:. 93 $333... w n: .:33...:m.3 2:9: 932. .3 .63.: w 0.. .:33...:m.3 932. .35 35.3.8333. ...3.. .3... .33:: 33.3.. 3.3 ..3.. 3 :..2. 33.3.33333 33 o. 3::3. 33:3:333 NEG 3:3:33:33 3... E9. 333:3:3 33.83.32 .33. 3:. 33933 33.3.. 3.3 333>.3:3 833.8333 .3. 333: 3:9. 3:. 3:3 332m 35.393: 3:3...333 33:33.33: NEG 3:. 3.3.333 :E:_33 3:3:-..3. 3: .r 3:3... 32m 3:. 3: 333... .3:. 33:33.8: 333:3:3 3:. 3:3 83333 32823:: 3:. >3 332.3..3. 3.3.. 33..33_3:: 3:3:33:33 3:. 33 33333 3.3 332m NEG .3..3.... :o..3:3o.3 v.52 3:3 333953.. 3:3 332m «:6 c333.3m 3:0..3.oo33< .:33...:m.m .o E3EE:m K 3.33... 143 3333.3 w m .3 232.333 .33.: 3 33.3 m n. .3 .333_._c3.3 3 30333< 30333... 30333< 333 303.3... 3.3 3.330 3333.0 3453.0 33. 333 35.35 33. 3....330 3....330 3....330 .33 30h .3< h .3 333 3.3330 003 333.... 3... 3033 .< 3033 .< 33. 3333.. 33. 33.330 33 05.350 NB 3.0%? VN «F... 3:30 3:30 5.33936 5.33.35 3:30 323 3.3; 5.5. ._. 300 .3 m 33 .. 0...5. 3. no ._. 30 .3 ".0 29:30: 3:3... 5.5332: 5.: 3:3 3:39.33 3:3 3:23 3:3 383.33 22.3.3333... 2.32.223 .3 33.5.53 .. 3.33.. 144 .333 w n. .3 3.3.. 5.2. 33.3.8333 3.332.333 3.32. 3333333 3.33 33.93 32”. ...3.. .33. .33:: 33.3.. 3.3 ..3.. 3 3:2. 33.3.33333 33 3. 3:3. 33:3:333 3.33 33.93 33.0 3:3:33:33 3:. :.3.. 3333330 .33. 33. 333.33 33.3.. 3.3 333233 :3..3.33333 .3. 333: 3:9. 33. 3:3 333:3:3 3:333 -.33 3:3...333 3.33 3:....3 3E0 3:. 3.3.333 :E:_33 3:3:-..3_ 33... .933... 320 33.0 3:. >3 33333:3 332. .33. 33:3:3 3.33 3:.E3 3:. 3:3 :3...333 3.33 3:.:.3 33. >3 3332.3. 9.3.. 3.33 3:.E3 3:3:33:33 33.33 33333 3.3 393.33.33.23: .3..3.._. :3..3:33.3 3...... 3:3 33.39334 3:3 333:3:0 3.3< 3:.E< 3:0 3333.33 3:3..3.3333< .:33...:3.m .3 E3EE:w .3 3.33... 145 >3333 >3333 >833 333 «.3333 <333h <333k <333P 333 ..333 33 333.. 33 .33 3 SE 3:30 33.33353 33.33355 3:30 39.30 3.3< 35:3 3.3; 3...-2 ._. 333 .x. __ 3...: 3. 33 ._. 333 .3 “.3 29:39. 33.3.... 3333332.... 3:.: 3:3 333933.. 3:3 33333.5 3.3< 35:3 «:0 3333.8 3:3..3.3333< 23355.3 .3 E38533 .3 3.33... 146 acid positions 8, 35, 46, 206, and 260 all associated significantly (P s 0.05) with MHC II expression, °/o CDB, % CD4, CD18 expression, and milk yield PTA within this group of Holstein bulls (Table 8), indicating that changes to the GR protein may indeed effect immune responses to stressful stimuli. Therefore, GR2 SNPs identified in this study associate significantly with both inherited glucocorticoid-sensitive leukocyte traits and milk yield PTA, suggesting that these may be potential candidate molecular genetic markers for health and productivity in dairy cows, particularly as 5 of these SNPs translated into amino acid changes which also associated significantly with altered immune function. D. DISCUSsz A population of 40 Holstein bulls with documented additive genetic varia- tion in leukocyte traits known to be responsive to glucocorticoids were used for GR gene polymorphism identification. A PCR-amplified region of the GR2 locus of 915 bp in length was found to be 92% and 93% homologous to the corresponding GR2 DNA and protein sequences in humans. We also found that the GR2 locus in Holsteins contained SNPs in 52 locations, more than half of which translated into conservative and (or) non-conservative changes in pre- dicted amino acid sequences in and around the 11 core of bovine GR. The amino acid change at position 195 is interesting since it falls within a GR2 syn- ergy control motif, in vitro mutation of which resulted in a 9 fold enhancement of transcription by human GR (lniguez-Lluhi and Pearce, 2000). Association of GR2 SNPs with traits of dexamethasone-induced leukocyte changes highlight- ed key mutations (at nucleotide positions 139, 621, 780, 820) that related to glucocorticoid sensitivity of a bull’s leukocytes. interestingly, the same muta- tions at nucleotide positions 621, 780, 820 associated significantly with a bull’s PTAs for milk yield. Milk protein and.milk SCS PTAs only tended toward an 147 association (PS 0.20) with any GR2 SNP. Therefore, we have not only found that the GR2 locus is polymorphic in Holsteins, but also that some of these polymorphisms are predicted to change protein sequence and to associate with heritable glucocorticoid-responsive leukocyte traits and genetic merit for milk production. Although this is the first study of polymorphism in bovine GR genes, bio- medical literature is abundant in which protein-coding regions of human GR genes have been studied for mutations suspected to be associated with dis- ease phenotypes such as primary glucocorticoid resistance (Malchoff et al., 1993), Cushing’s syndrome (Karl et al., 1996a,b), increased sensitivity to glu- cocorticoids (Huizenga et al., 1998), high blood pressure (Koper et al., 1997), and obesity (Dobson et al., 2001; Tables 1 and 2). Some work has even iden- tified point mutations in the human GR2 locus (reviewed in de Lange et al., 1997), but these had no apparent associations with the specific diseases or conditions under study. In contrast, site-directed mutagenesis of the human GR2 locus has revealed key amino acids that are critical for GR’s transcription regulation of target gene expression (Almlof et al., 1997; Table 3). In that work it was noted that clusters of 2 or 3 mutations had significantly more impact on subsequent GR regulation of glucocorticoid-responsive gene expression than any single mutation had on its own. Additionally, such mutagenesis studies in the human GR2 locus revealed what appeared to be different “surfaces” of the 1:1 domain that either activated or repressed transcription of glucocorticoid- sensitive genes (lniguez-Lluhi et al., 1997). It was also noted that several amino acid substitutions within these surfaces were required to significantly alter the transcription regulatory functions of GR’s 11 domain (lniguez-Lluhi et al., 1997). In the current study we have identified several clusters of naturally occurring mutations in bovine GR2 that change predicted amino acid 148 sequences and also related to multiple leukocyte and milk production pheno- types with known sensitivity to glucocorticoids. Together with findings from site-directed mutagenesis of human GR2 (Almlof et al., 1997; lniguez-Lluhi et al., 1997), it is possible that the SNP clusters we have identified in bovine GR2 genes changes one or more “activation surfaces” of bovine GR proteins, thus explaining associations with the traits we observed. However, analyses utiliz- ing the association of multiple SNPs with leukocyte traits need to be performed in order to begin determining whether GR 121 surfaces are disrupted by GR2 amino acid changing SNPs. If true, these SNPs may be useful in identifying bulls who may transmit beneficial “stress resistance” phenotypes to their daughters, which may in turn benefit health and milk production in these cows. 149 CHAPTER 6. ASSOCIATIONS BETWEEN MILK PRODUCTION TRAITS AND GLUCOCORTICOID RECEPTOR GENE POLYMORPHISM IN AMERICAN JERSEY BULLS A. INTRODUCTION Inflammation of the mammary gland, or mastitis, affects milk-producing mammals and is responsible for significant monetary losses for the US dairy industry (DeGraves and Fetrow, 1993). Within the bovine mammary gland the primary immune cells that fight invading pathogens are blood-derived neu- trophils (Paape et al., 2000; Kehrli and Harp, 2001), which cause the charac- teristic increase in milk somatic cell counts during mastitis. Milk somatic cell counts are averaged over a cow’s lactations and put on a Iog10 scale to deter- mine her somatic cell score (SCS; Schutz and Powell, 2003), which is used by veterinarians and producers as an indicator of mastitis for management deci- sions related to improved mammary gland health. One such decision which is utilized currently is genetic selection, where bulls with PTAs for low milk SCS can be selected to produce daughters with decreased milk somatic cells. However, milk SCS may not be the optimal trait for genetic selection against mastitis because its heritability is moderate (0.10 to 0.14; Detilleux 2002) and it is influenced by many environmental factors in addition to the presence of intramammary infection. As a result, researchers have studied a variety of immune traits as possible indicators of mastitis susceptibility and resistance. One such group of traits are leukocyte sensitivities to glucocorticoid hormones because these steroids achieve high concentration in blood around parturition and other stress periods when cows are the most susceptible to mastitis (Burton and Erskine, 2003). Previous studies have demonstrated that when cattle experience signifi- 150 cantly increased levels of circulating glucocorticoids, either during parturition (Preisler et al., 2000a) or exogenous administration of the hormone (Burton et al., 1995), their blood neutrophils display marked decreases in expression of the adhesion molecules CDGZL and 0018 (Lee and Kehrli, 1998; Weber et al., 2001) and reduced respiratory burst activity (Detilleux et al., 1994), rendering the leukocytes dysfunctional in their migration and bactericidal activities and leaving the mammary gland vulnerable to colonization by opportunistic bacteria (Burton and Kehrli, 1995; Shuster et al., 1996). Glucocorticoids have also been shown to affect various phenotypes of lymphocytes and monocytes/macrophages involved in inflammatory and adaptive immune responses (Burton and Kehrli, 1996; Nonnecke et al., 1997; Yu et al., 1998), all of which may contribute to mastitis susceptibility in honnone-treated and par- turient dairy cows (Burton and Erskine, 2003). The effects of glucocorticoids on leukocytes are mediated by homologous cytosolic receptors called glucocorticoid receptors or GR. GR is a ligand-acti- vated transcriptional regulator of glucocorticoid-sensitive genes (Bamberger 1996; Tsai and O’Malley, 1994). To affect gene expression, hormone activated GR form homodimers that readily translocate into the nucleus. Once in the nucleus, GR interacts with other transcription regulatory proteins through its TI domain to affect expression of glucocorticoid-responsive genes through both direct and indirect mechanisms (Bamberger et al., 1996; Ford et al., 1997). Recently, our laboratory has shown that bovine leukocyte sensitivities to glucocorticoid challenge possess additive genetic variation (Tempelman et al., 2002; Burton et al., in review). These findings implicated GR as possibly con- tributing to genetic variation in leukocyte dysfunctions that correlate with masti- tis susceptibility in parturient and hormone-treated cows. We substantiated this possibility by showing that the t1-encoding region of bovine GR genes 151 (called the GR2 locus) is polymorphic in six cattle breeds (Chapter 4) and in Holstein bulls (Chapter 5), and that certain key amino acid-changing GR2 mutations associated statistically with a bull’s neutrophil and lymphocyte sensi- tivities to glucocorticoid challenge and his PTA for milk yield (Jacob et al., in preparation; see Flgure 19 in Chapter 5). Therefore, GR2 polymorphism may explain some of the genetic variation in mammary gland immunity against mastitis, including milk SCS. In the current study we sought to determine whether the GR2 locus is also polymorphic in a population of Jersey bulls selected for high versus low milk SCS PTA, and if certain GR2 mutations associate significantly with a bull’s PTA for milk SCS, milk yield, and (or) milk protein yield. We identified a group of 20 Jersey bulls with milk SCS PTAs 2 2 standard deviations above (n = 11) or below (n = 9) the mean SCS for all of their contemporaries and used genomic DNA from sperm of these animals to confirm that the GR2 locus is polymorphic and that one key mutation associated with the milk production traits studied. 8. MATERIALS AND METHODS i. Identification of Jersey Bulls with High Versus Low Milk SCS PTA Twenty Jersey bulls were identified for this study based on milk SCS PTA. As shown in Table 9, 11 of the bulls had milk SCS PTA 2 2 standard deviations from the mean SCS PTA (3.27) of their contemporaries and were labeled HSCS1, HSCSZ, HSCS11. The remaining 9 bulls had milk SCS PTAs S 2 standard deviations from the contemporary mean (LSCSl, LSCSZ, ..., LSCSQ). The various sibling relationships between the 20 test bulls are also indicated in Table 9 and were determined from records of the bulls’ sires and maternal grandsires resident at the National Association of Animal Breeders 152 Table 9. Rankings of Jersey Bulls by Somatic Cell Score (SCS) Predicted Transmitting Ability (PTA). Bull Identification Milk SCS PTA Milk Protein PTA Milk Yield PTA HSCS1II: 3.69 39 1061 3:: 9. 9:: H 3 ' ' ' H 30541. § 3.62 24 1387 ”8835*“ 3'23; 3.3 339 HS 36 ' H8087: § 3.58 19 1050 HSCSBT 3.57 27 1490 H SC 39 3.55 22 1289 “383”” 3'23 3 333 H 11 ° Contemporary Mean SCS PTA = 3.27 LSCSQO 3.28 26 807 LSCS8 3.26 23 734 LSCS7 3.25 1 4 303 LSCS6 3.24 30 1 122 LSCSST A 3.18 27 234 LSCS4 3.1 6 21 328 LSCS3‘I’ A 3.1 6 5 840 LSCS2 3.16 21 1145 LSCSI o 3.04 24 731 “Denotes animals which are in half-sibling group 1; tDenotes animals which are in 3/4 sib group 1; ODenotes animals which are in 3/4 sib group 2; ADenotes animals in 3/4 sib group 3; §Denotes animals in 3/4 sib group 4. Table 9. Rankings of Jersey Bulls by Somatic Cell Score (SCS) Predicted Transmitting Ability (PTA). The left-hand column depicts the Jersey bulls ranked by SCS PTA and coded as high SCS (HSCS) or low SCS (LSCS), from highest to lowest scores. The traits studied are listed across the top and each bull’s PTA for that trait listed in the table. Bulls related to each other as half-sib or three-quarter sib groups are indicated by symbols. 153 web site (NAAB, http://www.naab-css.org; see Appendix F for sire and mater- nal grandsire information) and provided by the American Jersey Cattlemen’s Club (AJCC, Reynoldburg, OH). Straws of semen for each bull were located in the field and purchased by AJCC for this study, and the straws shipped in liquid nitrogen to our laboratory at Michigan State University for isolation of genomic DNA. ii. Isolation of Genomic DNA from Semen, PCR Amplification of GR2, and Molecular Cloning and DNA Sequencing of GR2 Genomic DNA was isolated from the donated semen samples as described in Chapter 4,Section Bla. GR2 fragments were amplified from genomic DNA using PCR as described in Chapter 4, Section Bil, and cloned into pGEM-T Easy vectors for subsequent DNA sequencing, as described in Chapter 4, Section Biv and v. Iii. GR2 Polymorphism Identification at the DNA and Predlcted Amino Acid Sequence Levels GR2 allelic sequences were analyzed for mutations using GCG (Madison, WI) software as described in Chapter 4, Section Bvi. Briefly, ‘chopup’, ‘refor- mat’, and ‘assemble’ commands of GCG were used to align and join the for- ward and reverse sequences of GR2 alleles. The ‘map’ and “translate’ com- mands of GCG were then used to determine the open reading frame and obtain predicted amino acid sequences from each GR2 DNA sequence. SNPs were compared across half-sibling groups to ensure Mendelian inheritance was not violated. Consensus DNA and amino acid sequences and multiple sequence files were generated using the ‘pileup’ and ‘pretty’ commands of GCG, and the sequence files for all GR2 alleles imported into an Excel spread- sheet (Microsoft Office v. x, Redmond, WA) for generation of plots and figures used for data presentation. This approach allowed ready visualization and 154 summary of SNPs and resulting amino acid sequence changes in the bovine GR2 gene. iv. Estimation of Association Between GR2 SNPs and Milk Production PTAs Original PTA for bulls’ milk SCS (Sept. 1999), and recent (Sept. 2002) PTAs for milk yield and milk protein yield were obtained from the Animal Improvement Programs Laboratory of the United States Department of Agriculture (AlPL-USDA; www.aipl.arsusda.gov). Associations of milk yield PTA, milk protein yield PTA, and milk SCS PTA with individual GR2 SNPs were determined by using the MIXED procedure of SAS to conduct one-way ANOVAs. Results were graphed as ‘SNP position’ against the ‘P-value’ of the F-statistic with a line of significance added at P = 0.05. SNP positions falling on or below this line were considered significantly associated with the trait. C. RESULTS Images in this dissertation are presented in color l. The GR2 Locus is Polymorphic in American Jersey Bulls DNA sequence analysis of GR2 from 20 American Jersey bulls revealed mutations (SNPs) at a total of 38 nucleotide positions within the 915 bp stretch of amplified GR2 (Figure 20a). When compared across half-sibling groups, SNPs did not violate Mendelian inheritance. The frequency of most SNPs was very low (one or two occurrences in the population of 20 bulls), but the SNPs at nucleotide positions 319, 429, 562, 660, and 703 occurred more often with allelic frequencies ranging from 10.5% to 18% (Figure 20b). The SNP at posi- tion 561 was very polymorphic (52% frequency) in this population of Jersey bulls. It should be noted that two alleles from two separate bulls produced poor quality DNA sequences, so these sequences were discarded. Thus, the 155 SNP frequencies reported here are based on 38 instead of 40 alleles for the population of 20 bulls. Figure 20b also shows that there were SNPs at nucleotide positions 561 and 562, with multiple combinations of possible nucleic acid changes apparent in this area of GR2. While most of the 38 SNP locations occurred in DNA sequence encoding regions of GR that flank the actual 1:1 core, 4 SNPs at nucleotides 319, 378, 422, and 429 lying within the t1 core-encoding DNA also harbored SNPs (underlined in the consensus sequence in Figure 20a and highlighted by hatched bars in Figure 20b). Therefore, the GR2 locus that encodes the 1:1 domain of bovine GR are polymorphic in Jersey bulls with high versus low milk SCS PTAs. ii. 682 Mutations Translate into Amino Acid Substitutions In Predicted Protein Sequence Twenty-three of the 38 GR2 nucleotide positions harboring SNPs were shown to change amino acids in predicted protein sequences (Figures 21a and b). Amino acid changes were both conservative (stayed within an amino acid class) and non-conservative (changed the class of amino acid from hydrophobic to hydrophilic, charged to uncharged, etc. See Appendix B for residue classification). 0f the 23 predicted amino acid changes, 10 were con- servative and 14 non-conservative, with residue 187 containing both conserva- tive and non-conservative amino acid changes (Figure 21 a). Due to the high degree of DNA sequence polymorphism at multiple SNP locations (Figure 20), many amino acid residue changes indicated more than 2 possible residue sub- stitutions (Figure 21 b). in particular, amino acid residue 187, which is translat- ed from the codon at nucleotides 561, 562, and 563, was highly polymorphic across the 20 Jersey bulls. Within the 11 core itself, four residue changes occurred (D1060, T126R, C1416, L143R; Figure 21; see Appendix B for 156 one-letter residue coding), corresponding to the nucleotide substitutions at residues 319, 378, 422, and 429 (Figure 20). Four amino acid substitutions occurred in the amino-terminal region flanking the 11 core but the majority of protein sequence polymorphisms occurred in the carboxy-terminal region of the 1:1 domain (Figure 21). The C141G (in 1:1), 02208, and O262P (carboxy- terminai flanking region) non-conservative amino acid substitutions are note- worthy because they might be expected to have effects on steric hindrance and (or) protein folding properties of GR’s 1:1 domain. Additionally, changes to amino acids at positions 195 and 211 fall within a synergy control motif, which has been found to enhance transcription by human GR by 9 to 12 told when mutated in vitro (lniguez-Liuhi and Pearce, 2000). iii. GR2 SNPs Associate with Milk Production PTAs But Not Milk SCS PTA Results of the one-way ANOVAs associating GR2 SNPs with mammary health and milk production PTAs are shown in Figure 22, which presents a series of plots where the X-axes indicate the nucleotide positions within GR2 where SNPs occurred and the Y-axes indicate the level of significance of asso- ciations between individual SNPs and the corresponding PTAs. One SNP, at position 598, was identified in this study as being significantly associated with milk production PTAs at P s 0.05. SNPs at nucleotide position 598 were also significantly associated with milk protein yield PTA (Figure 22b). While there were no SNP-milk SCS PTA associations detected that were significant at P S 0.05, several SNPs at positions 378, 454, 529, 582, 587, 786, 799, 862, 877, and 894 tended toward an association with milk SCS PTA at P S 0.15 (Figure 220). None of these SNPs were tending towards significance with milk protein or milk yield PTAs. These results suggest that SNPs at position 598 may 157 Figure 20a. Summary of GR2 Polymorphism Observed in 20 Jersey Bulls. Thirty-eight SNP locations were identified in the 915-bp region of GR2 ampli- fied by PCR and then cloned for sequencing of individual alleles. In this figure, the vertical axis lists bull identification numbers (ie. HSCS11, HSCS10, HSCSQ, etc.) along with the allele (1 or 2) shown for a particular row of DNA sequence. The consensus sequence identified by GCG (Cons) is shown at the top and bottom of the figure. The horizontal axis at the bottom of the figure shows the nucleotide positions within the original 915 bp fragments where SNPs were identified. Letters interrupting vertical lines are the actual changed nucleotides relative to consensus. Four SNP positions (319, 378, 422, 429) are located within the t1 core-encoding DNA of GR2 and are underlined in the consensus sequences. Highlighted in bold red letters are those SNPs which fell within double-pass sequence, while the remaining SNPs on either side were generated from single-pass sequence. 158 Figure 20a. Summary of GR2 Polymorphism Observed in 20 Jersey Bulls. Cons CIA'J'CC'C‘C‘FC‘ 'I‘GAM'I‘ HSCSl- HSCSl- HSCSZ- HSCSZ- HSCS3- HSCS3- HSCS4— HSCS4- HSCSS- HSCSS- HSCSG- HSCSG- HSCS7- HSCS7- HSCS8- HSCS8- HSCS9- i C C C C C T T ”.l‘ HSCS9-2 C? ‘ HSCSlO- I C C E C C C C G G E (DC) HNHNHNHNHNHNHNHNH A l Ont-3‘. 1 C C C HSCSlO-Z HSCSll-l HSCSll-Z LSCS9-1 LSCS8-l L LSCS8- LSCS7- LSCS7- LSCS6- LSCSG- -LSCS5- LSCSS- LSCS4- LSCS4- LSCS3- LSCS3- LSCSZ- LSCS2- Cons GI LSCSl- LLUJ. Posits 14i2411112 21602570 48713 I A U1 1‘ 'J Individual (1 or 2) GR2 Alleles for 20 Jersey Bulls HNHNHNHNHNHNHN C} C 0 N3 '5 r U10 U1 8qu ml: mvi 01" m0 ”’9 l—‘ \l [\J N 00 U" W om \D 00 N \O pU’l (DUI om HON NON Nucleotide Position in the 915 bp GR2 Inserts Key: HSCS = high somatic cell score bulls LSCS = low somatic cell score bulls Cons = consensus sequence 159 .8533 33.296 E0: “398950 99$ 2% 555 :0 mdzm 05:3an 9: of; 605303 mmmaefizoo £55, =2 :22; mmzm mon 9m 932 no. Eon 5 0959.59: 8:3 .6 cozflzaoa £5 E 223:? 25.32th mm; 5m 8:68 am mozofloac 9.0 .309 2 $de $6533: 5:9: 3m umtaooo mom new .08 Now .mmv dz.” 30:68 8 665m .Ao\oo.mv 2:833: 26. >625? E 09580 mmzm 60:. 223 .420 9.685-900 E on. Co 362883: .m new .m 0:32 mmzw 26% 2mg cocoa; or: .6 Em: new :2 05 9 93 2.0m . on... 622.6 mo_o__m __m Co «3352mm 8 nmmmmaxo vofiam cormsaoa :3 >38... 9: 5 80. «$0 an m5 9: E nozomno quw Co 36:33: £505 939.. $5. .m_.=m >333 ow .o cozsaod a :. «flop—0:62“. dzm «:6 .o EaEan .93 2:9“. 160 zzv Flgure 20b. Summary of GR2 SNP Frequencies in a Population of 20 Jersey Bulls. salauv "n9 Raster ssomv '5st to afietuemed O In 161 1768 £L8 398 088 88!. 98L SZL SOL [.99 099 Inserts 969 N l69 a: £89 0 999 Q :99 ass :2 l99 or 099 c 095 '- BZS o 909 E 009 on - O #97 0. 88? SW so: 3 m a 1.91 2 eat trOl l! at Figure 21 a. Summary of Polymorphism Observed in Predicted Amino Acid Sequence Encoded by the GR2 Locus. Twenty-three amino acid residue locations were predicted to contain substitu- tions encoded by the GR2 SNPs shown in Figure 20. In this figure, the verti- cal axis lists the bull’s identification (ie. HSCS11, HSCSlO, HSCSQ, etc.) along with the allele (1 or 2) shown for the particular row of amino acid sequence. Consensus sequences (Cons) are shown at the top and bottom of the figure. The horizontal axis at the bottom of the figure shows the amino acid positions within the full 1:1 region predicted protein sequence where amino acid substitu- tions occurred. Letters interrupting vertical lines are the actual residue changes relative to consensus. Four amino acid positions within the 1:1 core were predicted to be changed and are underlined in the consensus sequences. Highlighted in bold red letters are those SNPs which fell within double-pass sequence, while the remaining SNPs on either side were generat- ed from single-pass sequence. 162 Figure 21 a. Summary of Polymorphism Observed in Predicted Amino Acid Sequence Encoded by the GR2 Locus. Cons MTLKD'J. VIN. HSCS1- HSCSl— HSCS2- HSCS2- HSCSB— HSCSB- HSCS4— HSCS4— HSCSS- HSCSS- HSCS6- HSCS6- HSCS7- HSCS7- HSCSB- HSCSB- HSCS9- HSCS9-2 F HSCSlO— L Q GRL SEKCMV?KVQM '11'11 ”—8 (1)—(Pr l l I l I I I' I HNHNHNHNHNHNHNHNH W21—NIP H \r r \f‘F II [I I—d I. H L HSCSlO— HSCS11— HSCS11- LSCS9-1 LSCS8-1 LSCSB- LSCS7- LSCS7— LSCSG- LSCS6— LSCSS- LSCSS- LSCS4- Lscs4- LSCS3- LSCS3- Qfi R L383? L _ gi R LSCSl- “1 Cone . VLKL éixcm KVQ']. 473611111111111222222222 5802446678889123466789 66137963789104226768 t" ' J NI—‘NH individual (1 or 2) GR2 Alleles for 20 Jersey Bulls :1: tr} H 4. we P—zzwm—memzwwxzmmuw J‘r l-‘Nl-‘Nl—‘Nl—‘Nl—‘Nl—‘Nl—‘N F30) Amino Acid Residue Position in the 1:1 Region 163 .50.: 63:00 3.05.0 m0 0 E 09000. 90 09900 :0 .3 0900 ..0:.. SN 0:0 m9 0:03.000 0.00 0E5...v 30 000:0:0 30.0.3000 0.:. E 0.00.3? 2:9: 002, SF 5.3.000 .0 0.00 0:.E0 0:... 00.30.3000 :3 >099. 0.5 E5? 009 0. $0 80: mEmc9 00.0:0309. 00: 0E0.:90E>.00 .0.0>00 330 .o\om.m 2 0.0.0 .o 8.8832. 39 83:20 3.8 8.50 05 .o 3.5.05 9:. .92 3922 .3 30.3383. 90 .9: i: .09 .02 0:03.000 0.00 8.80 .0 .900 E. 0:. E5...) 003:0:0 00 0. 090.090 0030.09 0.00 0:.E< .300 :000 0>000 :023 90 :03000. 030.09 :000 30 0:032:88 0.00 oEEm .0303. 000:0:0 0.0..0 000:. .0 5.3.000 030.09 0:. 026:0 0.x0-x 0:. 0:0 090900 m00:90t.0 0.0..0 .0 03020900 0:. 030:0 m.x0.> 0:... 00.030 00.0.30 ..0 .0 0039:0900 00 0000990 0:0 0a 0.50.“. E 002m NEG E0: 090.090 0:0..350030 030.09 0.00 oEEm 0..0..0 .0 00.0:0309. 030:0 933: 0...... .0.-3m >099. on .0 5.6.3000 0 c. .00.. «m0 :0. 000:0300m 0.0< 0:.E< 00.0.00.0 05 n. 0020000 89:90:50“. ..o 305E3w .n R 0:30.“. 164 Figure 21b. Summary of Polymorphism Observed in the Predicted Amino Acid Sequences for GRZ Loci. IBGZJ. m “ILLZA 1.1.: 99: 0 N993>i 3 £930 :9: 3 'IszA m g mm m .E 80830 07.3 % *JJLZM uz E 0666 661 E * N96L)l 96L (5 '0 HBQLS 88L 3 E 1.8l '2 P m a O. :iGQL'I as: .E 'iLQLA 1.9l g HSVL‘i \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\ a” ‘5 9mg \\\\ m a HQZLJ. 0 9n 2 OQOLE) \\\\\\\\\\\\\\\\\\\\\\\ 90l- (<9 389M 99 g 3991 E am < 17w {azééséezééééémo sung Rosier oz 1s6uowv spgov oung pafiueug JO afieluaoiad 165 Figure 22. Associations between individual GR2 SNPs and milk produc- tion PTAs for Jersey bulls. In each panel, P-values from one-way analysis of variance of associations between SNPs and PTAs are given on the Y-axis (log10 scale), with a horizon- tal black line representing significance at P s 0.05. The X-axes in each panel show nucleotide positions within GR2 where SNPs were identified. SNPs that associated with more than one trait in panels (a) through (c) are indicated with solid black circles ( .), SNPs that associated with only the trait in the panel are indicated by hatched circles (@), and SNPs that were not significantly associated with any measured trait are indicated by open circles (O). (a) milk yield PTA; (b) milk protein yield PTA; (c) milk SCS PTA. SNPs at position 598 were significantly (P s 0.05) associated with both milk yield (a) and protein yield PTAs (b). GR2 SNPs at nucleotides 378, 454, 529, 582, 587, 786, 799, 862, 877, and 894 tended toward associations (P S 0.15) with milk SCS PTA (c). 166 Figure 22. Associations between individual GR2 SNPs and milk produc- tion PTAs for Jersey bulls. Figure 223. Milk Yield PTA 1.0 0.50 1 J i 1 ' o 1 . . J 0.20‘ 010* C 0 0 i 0.05 P - Value i w 0 0.02 ' 0.01 °3§§§ §§§§§§§§§§§§§ Position of Single Nucleotide Polymorphisms Figure 22b. Milk Protein Yield PTA 1.01 o o o .(o o o 1 0501 o . 0.10 1 d 1 I 0.02 ‘ P - Value 0.01 1 6 1 i 0.005 * 1 1 om‘ I 0.001‘ n- °3§§§§§§§§§§§§§§§§§ Position of Single Nucleotide Polymorphisms @ =Associates with only this trait O = No association I ' I III 1 T I'" II . = Associates with 2+ traits 167 Figure 22. (cont’d) Figure 22c. Milk SCS PTA 1.oo~ , , 0.90 i . 0.80 . 0 0.70 . 0.60 ‘ 0.50 . 0.40 ‘ —K) O 030* ° P - Value 0304 . . 0.15 0.125 0.10 [I III I l I III I l I III I I I I l I l I III I l I III I wssmség'fisé'més Position of Single Nucleotide Polymorphisms @ = Associates with only this trait . = Associates with 2+ traits O = No association 168 account for some of the genetic variation in milk production traits, while differ- ent SNPs at positions 378, 454, 529, 582, 587, 786, 799, 862, and 894 may influence mammary health, but each must be studied in a larger population of animals before final conclusions can be drawn. D. Discussm The study reported here has identified numerous mutations in and sur- rounding the 1:1 core-encoding locus of GR genes (GR2 locus) in a population of 20 active Al Jersey bulls. Several of the identified GR2 mutations were pre- dicted to change amino acid sequence, including 2 changes within GR’s syner- gy control motif; a mutation at SNP position 598 associated significantly with both milk yield and milk protein yield PTAs; and several SNPs at various 3’ positions tended toward associations with milk SCS PTA. These results were not surprising given that expression of at least three milk protein genes [3- casein, a-lactalbumin, and whey associated protein are controlled by glucocor- ticoids, most likely through interactions between GR and certain transcriptional coactivators (Houdebine et al., 1985; Reichardt et al., 2001). Indeed, increased blood glucocorticoid concentrations have been shown to reduce milk production in lactating ruminants (Bruckmaier et al., 1993; Sevi et al., 2001). While results of the current study must be substantiated in a larger population of bulls and their daughters, combined with available literature on glucocorti- coid regulation of milk protein production, they suggest that variation in the DNA sequence of GR2 may be partially responsible for genetic regulation of milk production during stress. Several studies other than this one have looked for associations between milk production traits and polymorphism in bovine candidate genes. In particu- lar, highly polymorphic alleles of the bovine DRB3 locus (encoding the antigen binding groove of MHC II) have been associated with milk SCS, milk yield, milk 169 protein yields, the incidences of milk fever, retained placenta, and mastitis (Sharif et al., 1998a, b; Dietz et al., 1997a), and the competency of the immune system in dairy cattle (Dietz et al., 1997b). Yao et al., (1996) also identified associations between growth hormone gene polymorphism and milk yield, milk fat yield, and milk protein yield in Holstein cattle. Additionally, recent work from our group has shown that several GR2 mutations associate with glu- cocorticoid-sensitive traits in bovine leukocytes, including neutrophil expression of C018 and the percentages of circulating T cells expressing CD4 and CDB accessory molecules (Chapter 5 and Jacob et al., in preparation). Three GR2 mutations also associated (P s 0.05) with milk yield PTA in the population of 40 Holsteins bulls used for that study, but those were at different GR2 nucleotide positions (Figure 191’) than the one position identified in the current study (Figure 22a, b). Interestingly, when SNPs identified in the current study are compared to those present in a population of Holstein bulls (Chapter 5 and Jacob et al., in preparation) several SNPs tending toward an association with milk SCS PTA (P S 0.20) were found to be held in common (or in close proximity within the GR2 gene) between the two populations of bulls (Table 10). Within the current Jersey population these occurred at nucleotide positions 203, 550, 582, 587, 598, 703, and 894, and within the Holstein population (Chapter 5) the SNPs occurred at positions 200, 545, 586, 597, 702, and 895 (Table 10). Three SNP positions located within 1 bp of each other (586/587, 597/598, and 702/703) translated into the same amino acid substitutions (at residues K195N, D1990, and M234l/T) in both populations, while two additional SNPs translat- ed into neighboring amino acids (L182F in Jersey and |183V in Holstein). SNPs that translated into amino acid substitutions at residues 182 and 183 (I182V and L183F) tended to associate with milk protein yield PTA in both pop- 170 ulations of bulls, while SNPs that translated into a K195N substitution tended to associate with milk SCS PTA in both populations, and SNPs that translated into E1990 substitutions tended to associate with milk yield PTA in the Holstein population (P s 0.20) and associated significantly (P s 0.05) with milk yield PTA in the Jersey population (Table 10). These results raise the suggestion that common GR2 SNPs may be identified that are conserved across dairy breeds which may have an impact on milk production traits following a stress challenge. Given the known effects of glucocorticoids on lactation, additional studies using larger populations of animals are warranted to substantiate this possibility. In conclusion, this study reports the identification of numerous naturally occurring mutations within the GR2 locus of a small population of active Jersey bulls, chosen for study based upon their divergence in milk SCS PTA. Because SNPs at one of these GR2 nucleotide positions associated signifi- cantly with genetic merit for milk production traits of key economic importance to the dairy industry in this small population of Jersey bulls, GR2 polymorphism may prove useful for future studies on glucocorticoid regulation of mammary gland health and productivity. Ultimately, certain GR2 polymorphisms may be proven to be useful molecular genetic markers for genetic improvement of mastitis resistance and milk production in high producing, intensively reared dairy cattle. 171 Table 10. Summary of GR2 SNP Positions Commonly Held Between Holstein and Jersey Bull Populations. Comparison of SNPs which tended (P S 0.20) toward associations with milk production traits in both Holstein and Jersey bulls identified seven nucleotide positions in GR2 that were either held in common or occurred in close proximi- ty to each other (within 5 bp). These are listed in the left-hand column, fol- lowed by the population the SNPs occurred in. The right-hand column lists the amino acid changes resulting from the SNPs at each position. Three amino acid changes were common between the populations, at residues 195, 199, and 234. SNPs occurring at positions 545 (Holstein) and 550 (Jersey) translat- ed into neighboring amino acids (182 and 183) and were associated with milk protein yield PTA in both populations, while SNPs at position 586 (Holstein) and 587 (Jersey) changed the same amino acid (K195N) and were associated with milk SCS PTA in both populations. Additionally, SNPs at positions 597 (Holstein) and 598 (Jersey) which changed the same amino acid (E1990) were associated with milk yield PTA in both populations. 172 Table 10. Summary of GR2 SNP Positions Commonly Held Between Holstein and Jersey Bull Populations. SNP Trait Amino Acid Position Population Association Change 200 Holstein Protein Yield PTA, Silent C018 Expression 203 Jersey Milk Weld PTA K68E 545 Holstein Milk Protein Yield PTA I183V 550 Jersey Milk Protein Weld PTA, Milk SCS PTA L182F 560 Jersey Milk SCS PTA Silent 562 Holstein Milk Yield PTA, CD18 Expression Silent 586 Jersey No significant associations K195N 587 Jersey Milk SCS PTA Silent 586 Holstein Milk SCS PTA, 0018 Expression K195N 597 Holstein Milk Yield PTA E1990 598 Jersey Milk Weld PTA, Milk Protein Weld PTA E1990 702 Holstein Milk Weld PTA, CD18 Expression M234T 703 Jersey Milk SCS PTA M234V M234I 799 Jersey Milk SCS PTA K266N 801 Holstein Milk Weld PTA, Milk Protein Weld PTA P267H CD18 Expression 894 Jersey Milk SCS PTA T298l 895 Holstein 0018 Expression Silent 173 CHAPTER 7. GENERAL DISCUSSION AND CONCLUSION The response to glucocorticoid challenge is varied among species, individ- uals within a species, and different tissues, cell types, and stages of the cell cycle within individuals (Yudt and Cidlowski, 2002, and references therein). Responsiveness is mediated at the molecular level through GR and the tran- scriptional co-factors with which the receptor associates. Hormone-bound GR interacts either with specific DNA motifs in promoters of target genes and (or) a variety of transcriptional coactivators/suppressors to regulate gene expres- sion and change the phenotype of cells during stress. Following the identifica- tion of variation in sensitivity to glucocorticoids between human, mannoset, and guinea pig GFls (Brandon et al., 1991; Keightley and Fuller, 1994; Keightley and Fuller, 1995), numerous naturally occurring mutations were iden- tified in key domains of hGR which affect their transcriptional functions (Tables 1 and 2, Chapter 3). Many of these hGR gene mutations affected the pheno- type of individuals harboring them. Recently, studies have also shown that there is significant additive genetic variation in the sensitivity of the bovine immune system to glucocorticoid challenge (Tempelman et al., 2002; Abdel- Azim et al., in review; Burton et al., in review; Kelm et al., in review; Table 4, Chapter 3). Therefore, the hypothesis of the current series of studies was that bovine GR genes are polymorphic and that corresponding SNPs associate with traits of immunological and economic relevance to the cattle industry. The goals were to determine whether the bovine GR2 locus harbors polymorphism, to characterize this polymorphism at the DNA and amino acid sequence levels, and to elucidate possible associations between polymorphisms and a variety of immunological and production traits measured in populations of pedigreed cat- tle. 174 This dissertation describes the presence of complex GR2 SNPs in three populations; 6 cattle breeds selected for either milk production (Brown Swiss, Holstein, and Jersey), beef production (Angus and Polled Hereford), or dual milk and beef production (Brahman; Chapter 4); a population of 40 active Al Holstein bulls with known leukocyte sensitivities to glucocorticoid challenge (Chapter 5); and a population of active AI Jersey bulls divergent (2 2 standard deviations from their contemporaries) in milk SCS PTA (Chapter 6). Many of the identified SNPs translated into changes in predicted amino acid sequences. Three-dimensional homology modeling of divergent Holstein alle- les predicted only modest changes in tertiary structure of 11 regions encoded by mutant GR2s (Chapter 4), however this approach does not reveal changes in hydrophobicity or hydrophilicity, charge, or bonding. Holstein bulls with doc- umented genetic variation in leukocyte sensitivity to glucocorticoids harbored numerous SNPs within their GR2 loci (Chapter 5). In these glucocorticoid challenged bulls, several SNPs associated significantly (P 5 0.05) with changes in neutrophil CD18 expression, MHC II expression on mononuclear cells, and percentages of circulating CD4 and CDB T cells. Additionally, sever- al SNPs associated significantly with the bulls’ PTAs for milk yield (Chapter 5). Associations between GR2 SNPs, neutrophil 0018 expression, and milk yield PTA in the Holstein bulls suggested that GR2 polymorphism may be linked to mammary health and productivity in this breed of cattle. In the dairy industry, mammary health is monitored using milk SCS as an indicator trait, which is a reflection of the ability of neutrophils to migrate (via 0018 adhesions) into infected quarters. Following identification of complex polymorphism in a popu- lation of 20 Jersey bulls of divergent milk SCS PTAs, some GR2 SNPs associ- ated significantly with milk yield and protein yield PTA (P g 0.05), but no SNPs associated significantly at P s 0.05 with milk SCS PTA (Chapter 6). However, 175 this may have been due to the small number of animals studied and the ten- dency (P S 0.20) for one key SNP to associate with milk SCS PTA suggests that this w0ik should be repeated in a larger number of animals. In order to elucidate the extent of GR2 polymorphism, all sequenced GR2s across populations were analyzed for SNP location and frequency. GR2 loci from a total of 89 animals have been sequenced, resulting in 167 usable alle- les. All totaled, these alleles harbored 103 unique SNP positions (Figure 23 and Table 11), 43 of which are found in more than one population [Divergent Breeds, Jersey, and (or) Holstein; Table 12]. In addition, several SNPs were found exclusively in each breed (Table 5), and may be markers for breed-spe- cific variation in glucocorticoid-responsiveness. Additionally, several SNP posi- tions occurred in relatively high frequency (2 3% ) across alleles, with nucleotide position 561 being particularly variable (30%; Figure 23). The majority of SNPs were located in 5’ and 3’ regions flanking the 1:1 core-encod- ing DNA, with 8 SNPs falling within the 1:1 core itself (highlighted by white bars in Figure 23). These changes may result in alteration of secondary and (or) tertiary protein structure, bonding, or charge, and therefore be partially respon- sible for variation in a cell’s responsive to glucocorticoids. Ultimately, alterations in transcriptional regulation by GR’s 11 domain are a result of changes to the protein structure or charge, therefore predicted amino acid changes resulting from GR2 SNPs were analyzed across populations. Across all animals studied, a total of 62 unique amino acid changes were pre- dicted to occur within the 11 region of GR (Figure 24 and Table 13), 22 of which were found in more than one cattle population (Table 14). Twelve residue changes had relatively high frequencies (2 3%) across animals (Figure 24). Most predicted protein changes were located in the amino— and carboxy- terrninal regions surrounding the 1:1 core, however, 5 residue changes 176 occurred within core 1:1 (highlighted by hatched bars in Figure 24). Such changes in predicted amino acid sequences would most likely result in alter- ation of tertiary protein structure or charge, and lead to changes in 1:1 transac- tivation activity. Examples of residue substitutions which could possibly alter protein structure or charge are L35P found in Holsteins, which may result in steric hindrance due to a bulky proline residue; S46L also in Holsteins, which could be expected to change the polarity of the residue and perhaps orienta- tion (Table 14; see Appendix B for classification of amino acid residues); N52K found in Brown Swiss and Angus cattle (Figure 12a) or K195N which is found in Jersey and Holstein (Table 14), both of which would be expected to alter the charge of the residue; $188R found in Jersey, which could change both charge and polarity (Table 14 and Figure 12a); and C2208 in all of the six cattle breeds, which has the possibility of dismpting disulfide bridges, and 6225F in Holsteins which would change the polarity of the amino acid (Table 14). Indeed, changes to the bovine 1:1 region at amino acid positions 143 (L143R) and 158 (S158N) in hGR have been found to influence GR’s ability to transcribe a reporter gene from 28-48% and 24-39% relative to wild type GR, respectively (Table 15; Almlof et al., 1997). Additionally, changes at amino acid position 158 (F1538 in bovine) in humans reduces the ability of hGR to bind the TATA binding protein (84% reduction) as well as the CREB binding protein (28% reduction), thus reducing transcription by mutant GR relative to wild type GR (Almlof et al., 1999). In 2000, lniguez-Lluhi and Pearce identified a region in the human GR 11 which appears to suppress transcription, termed the synergy control motif. This motif is highly conserved across the members of the hormone receptor superfamily as (lN)KXE where IN indicates the pres- ence of either an isoleucine or valine at that position and X is any amino acid. Disruption of this motif by mutation of these conserved amino acid residues 177 results in a 3 to 12 fold increase in GR transcriptional activity relative to wild type GR (lniguez-Lluhi and Pearce, 2000). Interestingly, bovine 11 regions contain mutations at 2 amino acid residues within the synergy control motif (Table 15). Mutation of residue 298 (195 in bovine GR) results in a 6 fold increase in hGR transcriptional activity in vitro, while mutation of residue 315 (211 in bovine GR) resulted in a 12 fold increase in hGR activity relative to wild type (lniguez-Lluhi and Pearce, 2000). Therefore, not only is the bovine GR2 locus polymorphic, but many of the identified SNPs change predicted amino acid charge and polarity and presumably the ability of GR to regulate target gene transcription. it has been suggested that GR 11 regions regulate transcription of gluco- corticoid-responsive genes by using active surfaces to interact with other tran- scriptional co-regulators (lniguez-Lluhi et al., 1997). Therefore, mutations har- bored by bovine GR2 loci that changed predicted amino acids may alter the folding of GR’s 11 region, or its overall charge, and ultimately the phenotype of glucocorticoid-responsive cells. Why would variation in 11 region-encoding GR2 be present in animal pop- ulations? Changes to GR 11 regions resulting from polymorphism in GR2 could result in changes to the folding and (or) charge of the 11 region. Such changes in the 11 domain of GR could affect how it interacts with other tran- scriptional co-regulators (11-protein interactions), leading to alterations in GR’s regulation of glucocorticoid-responsive gene expression. Formation of unsta- ble 151-protein interactions as a result of identified mutations could lead to reductions in the transactivation or transrepression of gene expression by GR. Alternatively, more stable 11 -protein interactions could increase GR’s transacti- vation or transrepression of target gene expression. Such changes in GR’s ability to regulate transcription could cause changes in the mRNA levels of glu- 178 cocorticoid-responsive genes, and might alter the amount of protein expressed by the cell. Protein changes could result in changes in the cell’s phenotype, which in turn, can change the phenotype of an organ or tissue, and ultimately the entire animal. Changes in an animal’s responsiveness to glucocorticoids, and therefore to stressful stimuli, could influence the evolutionary selection of animals with superior genetics. Prior to the domestication of cattle, increases in blood glucocorticoid concentrations in response to perceived stress (Appendix A) enabled them to react to predatory danger through increased blood glucose concentrations and heart and respiration rates, thus allowing for an enhanced fight or flight response. Animals who were able to evade preda- tion and remain healthy would be more likely to pass on their genetics. Therefore, any genetic changes to GR, including its 11 domain, which resulted in appropriate adaptation of gene expression required for health and survival would be preserved. Today, presence of polymorphism in the 11-encoding GR2 locus could reflect these evolutionary struggles for survival and still underlie subtle variations in glucocorticoid (stress)-responsive phenotypes of the immune, mammary, reproductive, and circulatory systems that partially determine an animal’s genetic merit. Clearly, more studies using larger animal numbers and precise measurements of glucocorticoid-sensitive traits of eco- nomic importance need to be performed to substantiate the findings of the cur- rent study and determine if the GR2 locus is a critical factor in predicting ani- mal performance in today’s husbandry environments. In addition, studies are required to define precisely which GR2 SNPs or SNP clusters are critical in regulating expression of glucocorticoid-responsive genes with known involve- ment in animal health, reproduction, and production. 179 Table 11. Summary of SNPs Identified Across All Sequenced GR2 Loci. SNP Position & SNP Position 8: SNP Position & SNP Position 8: Population Population Population Population 1 Jer 203 Jer 531 Breeds, Hol 701 Hol 4 Jer, Breeds 247 Hol 545 Breeds, Hol 702 Hol 5 Hol 289 Hol 550 Jer 703 Breeds, Jer 12 Breeds, Jer 319 Jer 560 Breeds, Jer 725 Breeds, Jer 21 Breeds, Jer 346 Breeds, H0! 561 Breeds, Jer 733 Hol 24 Hol 377 Breeds 562 Hol, Jer 754 Hol 28 Hol 378 Jer 563 Breeds 759 Hol 46 Jer 415 H0! 565 Breeds 780 Hol 62 Hol 417 Breeds 571 Hol 783 Breeds 77 Breeds, Hol 422 Jer 578 Breeds, Hol 786 Breeds, Jer 95 Hol 429 Breeds, Jer 579 Hol 799 Hol, Jer 104 Jer 441 Breeds 582 Jer 801 Hol 105 Hol 442 Hol 586 All 811 Hol 123 Breeds, Hol 454 Breeds, Jer 587 Jer 813 Breeds, Hol 128 Breeds, Jer 459 Breeds 591 Jer 820 Breeds, Hol 138 Hol 472 Breeds 592 Breeds 828 Hol 139 Hol 474 Breeds 597 Hol 830 Breeds, Jer 140 Hol 480 Breeds 598 Breeds, Jer 841 Hol 145 Hol 492 Breeds, Hol 613 Hol 848 Hol 147 Breeds, Hol 496 Hol 617 Breeds, Hol 851 Hol 154 Hol 500 Jer 621 Hol 862 Breeds, Jer 156 Hol 505 Jer 633 Breeds, Jer 877 Jer 157 Breeds, Jer 510 Breeds 645 HoI 881 Hol 171 Breeds, Jer 520 Breeds 660 Breeds, Jer 894 Jer 193 Breeds 528 Breeds 667 Jer 895 Hol 200 Hol 529 Breeds, Jer 675 HoI Key: Breeds = All six cattle breeds, Jer = Jersey bull population, Hol = Holstein bull population. Table 11. Summary of SNPs Identified Across All Sequenced GR2 Loci. Following analysis of the presence of SNPs across sequenced GR2 loci, 103 unique nucleotide SNP locations were identified in the 915 bp sequences of GR2. SNPs are listed in the columns followed by the population in which they are identified. SNP locations that fall within the 11 core-encoding DNA are ital- icized and boldfaced. 180 Table 12. Summary of SNPs Found in GR2 Loci from More Than One Cattle Population. Twenty-nine SNP positions were identified in GR2 loci of more than one cattle population. The first column lists the position of the identified SNP, with the exact change found across the breeds (Breeds; Chapter 4) following it. Similarly, the change at each SNP position found within the Jersey bull (Jersey Bulls; Chapter 6) and Holstein Bull (Holstein Bulls; Chapter 5) pop- ulations indicated in the next columns. The right column lists the amino acid position encoded by the SNP(s) found within the cattle populations studied. SNPs that fell in the 11 core-encoding DNA are highlighted by a gray box. 181 Table 12. Summary of SNPs Found in 0R2 Loci from More Than One Cattle Population. Position Breeds Jersey Holstein Amino Acid of SNP Bulls Bulls Position 12 T12A (K) T120 (l) M4K, l 21 C21 G 021 G T7R 46 0460 C460 G15E 77 0770 (V) C77T (F) L26V, F 104 0104T L35F 105 T1050 L35F 123 01230 (G) 0123T (V) A41G, V 1 28 01280 01 280 L43V 147 T1470 T1470 V49A 156 A1560 N528 157 01570 01570 N52K 171 01710 01710 A570 346 G346A 0346A Silent 377 A3770 T126A 378 03780 T126R 429 T4290 T4290 L143R 454 T4540 T4540 Silent 492 A492T (M) A4920 (R) K164M, K164R 528 A5280 K176R 529 A5290 (N) A5290 (K) K176N 531 A5310 A5310 01770 545 A5450 A545G I182V 560 A560T A5600, T (187) 561 0561A, C 05610, T 8187T,Y,N 562 C562T, G, A 05620 (T187l) T187S,l,R,L,F,P 578 0578A C578A Q193K 586 A5860 A5860 A5860 K195N 591 A591 G E1970 592 A5920(E), T(D) E1970, Silent 597 A597C E1990 598 A5980 A5980 E1990 617 A61Tl' A6170 T2068 633 A6330 A6330 K211T 660 06600 06600 C2208 702 T7020 M234T 703 07030 07030 M234l 725 07250 07250 V242L 786 A7860 A7860 Q262P 799 A7990 (N) A7990 (K) K266N, Silent 813 T8130 T8130 V271A 820 0820T A8200 Silent 830 08300 08300 V277L 862 08620 08620 Silent .300 :0...000 00:00.035 0000 o. 000.0 E9. 050:9 00.0:0309. 00: 0020 .050 .0.0>00 .000 w >0:0309. 0 5.? 09.3000 0020 .0 3:0..0E 0:... 0:0.09 055:0: <20 .0 0:0 .0 0:. 5 0:30. 0020 030:0 0.0: 9...; 0:. .0 E0: 0:0 ..0. 0:. 0. 0..0: 0..00 . 0:. 0:0 0020 .0 :0...000 00:00.03: 0:. 030:0 0.0-x 0: 0. 00.030 00.0..0 ..0 .0 0000.:0900 00 0000990 8’ 030.. 5 550:0. 00.0..0 ~00 000:03000 :0 00:00 0020000 0020 .0 00.0:0309. 026:0 930.. 0.: 0. 000003000 .00.. «:0 :0 000.04. 0020 .o 0000:3000 0:. .0 30.5030 .00 0.30.0 183 Figure 23. Summary of the Occurrence of SNPs Across all 0R2 Loci Sequenced. 768 198 W8 8Z8 8 I8 108 98!. 08L VSL SZL ILL 991 [.91 0171 831 VOI- 9? 73 Z I- O a O In 0 In 1'9 N 1- 1- seieuv nv ssoiov SdNS to afietueoied 01 184 Table 13. Summary of Amino Acid Changes Identified in the GR2- Encoding 11 Region. Amino Acid Position Amino Acid Position Amino Acid Position & Population & Population 8: Population L2V Breeds E 1390 Breeds E1990 Breeds, Jer M4K Breeds C1410 Jer E1990 Hol M4T Jer L143}? Breeds, Jer T2068 Breeds T7R Breeds, Jer A147R Breeds T206A Hol E80 Hol F1538 Breeds P207l Hol D15E Breeds 8158N Breeds K211T Breeds, Jer L26V Breeds N160| Breeds 02208 Breeds, Jer L35F Jer K164M Breeds 8225F Hol L35P Hol K164R Hol M234l Breeds, Jer A410 Breeds V167L Jer M234T,V Hol A41V Hol L169F Jer V242L Breeds, Jer L43V Breeds 01700 Breeds H2510 Hol S46L Hol K176R,N Breeds 02530 Hol T478 Hol D1770 Breeds, Hol 8260Y Hol V49A Breeds, Hol I182V Breeds, Hol 0261 L Breeds E510 Hol L183F Jer 0262P Breeds, Jer N52K Breeds S187T,Y,N Breeds K266N Jer N528 Hol T187S,N,K,L,F,H,R Jer P267H Hol A57G Breeds T187l Hol V271A Breeds K68E Jer 8188R,G Breeds V277L Breeds, Jer 082H Hol S188R Jer E2800 Hol K96N Hol Q193K Breeds N2830 Hol D1060 Jer Q193K,P Hol R2840 Hol T126A Breeds K195N All T294A Hol T1261? Jer E1970 Breeds T298l Jer Key: Breeds = All six cattle breeds, Jer = Jersey bull population, Hol = Holstein bull population. Table 13. Summary of Amino Acid Changes Identified in the 0R2- Encodlng 11 Region Seventy-five unique amino acid residue changes were identified across the 304 residue bovine 11 region of the 89 animals studied. Residue changes that fell within the 11 core are italicized and boldfaced. 185 Table 14. Summary of Amino Acid Changes Found in GR 11 Regions of More Than One Cattle Population. Position Breeds Jersey Holstein SNP of AA Bulls Bulls Position(s) 4 M4K M4T 12 7 17R T7R 21 35 L35F L35P 104, 105 41 A410 A41V 123 49 V49A V49A 147 52 N52K N528 156, 157 126 T126A T1261? 377, 378 143 L143R, P L143R 429 164 K164M K164R 492 177 D1770 01770 531 182 I182V I182V 545 187 8187Y,T,N T187S,N,R, T187l 560, 561, 562 K,L,F 193 0193K 0193K 578 195 K195N K195N K195N 586 199 E1990 E1990 E1990 597,598 206 T2068 T206A 617 211 K211T K211T 633 220 02208 02208 660 234 M234l M234l M234T 702, 703 242 V242L V242L 725 262 0262P O262P 786 277 V277L V277L 830 Table 14. Summary of Amino Acid Changes Found in GB 11 Regions of More Than One Cattle Population. Twenty-two unique amino acid residue positions were found in GR2-encoded 11 regions of more than one cattle population. The first column lists the posi- tion of amino acid changes encoded by identified SNPs, with the exact change found across the breeds (Breeds; Chapter 4) following it. Similarly, the change at each amino acid position found within the Jersey bull (Jersey Bulls; Chapter 6) and Holstein Bull (Holstein Bulls; Chapter 5) populations indicated in the next columns. The right column lists the SNP position(s) found within the cattle populations studied. Residue changes within the 11 core are highlighted by a gray box. 186 ..00. 030.00.. 0000 0. :00 0:0 .000 .03 0030.00.. 0. E9. 050:9 00.0:0309. 00: 0:0...000 030.00. .90 ->00 ..0>026: .000 w 6:0309. 0 :..2, 09.3000 000:0:0 .0 5.0..0E 0: 0. .3. 0:2:0: 000:03000 50.0.0 30260000.. 0.0:.E.0.->x00.00 0:0 0:.E0 0:. 5 000:0:0 0.00.05 .:0.. 0:0 ..0. 0:. 0. 0.0: 0..00 0.32. .900 E. 0:. 5 000:0:0 00.0.00.0 00.00.05 03 0:0 .3. .00. .00. .00. 0030.00. .0 0.0: 00:0.0... 000:0:0 0.0..0 000:. .0 :0...000 030.09 0:. 026:0 0.x0-x 0:. 0:0 00.00.00 000:90...0 0.0..0 .0 000600.00 0:. 026:0 0.x0.> 0: 0. 00.020 00.0..0 ..0 .0 0000 -.:00.00 00 0000900 0.. 0.00... 5 0020 00.0 E9. 00.0.00.0 0:0..3...0:30 030.00. 0.00 0:.E0 0..0..0 .0 00.0:03090 000:03000 .00.. «mm. =< 000.0< 000:0:0 0.0< 0:.E< .o 00.9.0300... 0:. .o E0EE30 éu 0.30.0 187 0:0.00. E. 0:. :. :o...0o0 0.0< o:.E< 37..703—67363372637.Z70Irlrlrlrlrlrlrlrlrlrlrlrlrlrlrlrlrlrlrlrlrillr nggggsve-CZIrooeeg gall-£99 ssvvvezosgsgvvv veal» LLLQZIroeerVSOIrI-gstsc 6701-906]. aslctggggL-CIPGI. Iron—gsslg seieuv "v ssomv sefiueua to afietuemed In N O ('9 mm 000:03000 .00.. «00 ..< 00o.0< 000:0:0 0.0< 05:2 .0 00.000300.“— 0:. .o E85030 .00 0.30.0 188 Table 15. Summary of Significant Amino Acid Changes in Human and Bovine 11 Regions and the Resulting Change in hGR Activity. Position of Corresponding Transactivation of a Reference hGR Amino Amino Acid Change Reporter Gene as for hGR Acid Change in Bovine GR2 Compared to WT hGR L225V L143R 28-48% Almlof et aL,1997 F235L or V F153NS 84% decrease in bind- Almlof et al., ing affinity for TBP and 1998 28% decrease in CBP binding N2400 S158N 24-39% Almlof et al., 1997 K298E K195N 6-fold increase Iniguez-Lluhi and Pearce, 2000 K315E K21 1 T 12-fold increase Iniguez-Lluhi and Pearce, 2000 Table 15. Summary of Significant Amino Acid Changes in Human and Bovine 11 Regions and Resulting Changes in hGR Activity. Amino acid residue changes identified within hGR 11 regions or the synergy control motif are listed in the left-hand column with their bovine counterparts listed in the next column (11 core are indicated by bold blue). The functional change observed in an in vitro reporter gene system or protein binding affinity is reported in the third column and the reference is listed in the right-hand col- umn. Changes to hGR residues 225 and 240 result in decreased GR tran- scription while changes to residue 235 reduce the ability of GR to bind the TATA binding protein (T BP) as well as the CREB binding protein (CBP). Residues 298 and 315 (italicized) are located within the synergy control motif and enhance hGR transcription when mutated. 189 CHAPTER 8. RECOMMENDATIONS FOR FUTURE WORK In order to determine true biological relevance of the GR2 mutations identi- fied through the current series of studies, several additional studies would need to be conducted both in vivo and in vitro. First of all, associations of GR2 SNPs with other heritable glucocorticoid-responsive leukocyte traits, such as lymphocyte blastogenesis, random migration of neutrophils, and neutrophil oxidative metabolism and generation of superoxide anions, etc. (described in Kelm et al., in review), should be conducted. Additionally, modification of the current one-way ANOVA protocol to test multiple SNPs (instead of individual SNPs) for associations with the glucocorticoid-responsive traits studied may reveal significant clusters of SNPs that form response surfaces that may be worthy of further study. In addition, association of individual and multiple amino acid changes with glucocorticoid-responsive immune traits could high- light important protein residues in 11 regions. Furthermore, development of an efficient PCR-based genotyping method to facilitate future studies in larger populations is essential. One such study could be the genotyping of signifi- cantly larger populations of bulls and their daughters, from the Jersey or Holstein breeds, for analysis of milk production and peripartum immune traits and determination of associations between key 0R2 SNPs and these traits. On a larger population of animals, truly significant (PS 0.001) GR2 SNPs may be revealed and allow for effective GR2 genotyping and (or) haplotyping for genetic selection purposes. To further identify biological relevance of GR2 SNPs and also to continue linking glucocorticoid-sensitive immune traits to mammary health, animals which have been genotyped or haplotyped for 0R2 SNPs could be challenged with glucocorticoids (dexamethasone) and immune traits collected and ana- 190 lyzed, particularly those which have been shown to possess genetic variation (Tempelman et al., 2002; Burton et al., in review; Kelm et al., in review) and also to be significantly associated with GR2 mutations in the current study (ie. % CD4 and 008 T cells, CD18 expression by neutrophils, MHC II expression by mononuclear cells, etc.). Not only would this type of study allow further analysis of glucocorticoid-responsive immune functions, but would allow for SNP-trait associations better poised to reveal truly significant SNPs. Additionally, microarrays could be run on GR2 “glucocorticoid-sensitive” and “glucocorticoid-resistant” genotyped animals for the identification of target genes which might be differentially effected in expression by GR2 mutations. The studies outlined above are mainly designed to screen for GR2 muta- tions and to determine whether they are truly biologically relevant to immune function, mammary health, and milk production of the whole animal. A concur- rent study could be conducted to determine the effect and importance of indi- vidual amino acid changing SNPs on GR function in vitro. This study would make use of a highly defined gene expression system to study 0R2 haplo- types that associate with transactivation of a specific reporter gene, thus identi- fying key 0R2 alleles that may be important to stress susceptibility. Making use of the yeast transcription factor Gal4, which is not found in other higher mammalian cells, various 0R2 mutants can be fused to the Gal4 DNA-binding domains and used to transactivate expression of a reporter gene such as the green fluorescent protein (GFP) or luciferase gene. Following identification of ‘functionally relevant” GR2 haplotypes in vitro, relevant SNPs could be identi- fied that were responsible for changes to 11 transactivation/transrepression function. These may make the best candidates for future studies of GR2 poly- morphism in vivo. Finally, screening and investigation of the remaining exons 191 of bovine 0R gene may reveal significant polymorphisms which are associated with sensitivity or resistance to glucocorticoids. Ultimately, these suggested studies would identify functionally relevant GR2 SNPs, both in vitro and in vivo, which could be used in the future for management decisions or genetic selection of cattle whose immune systems will be less sensitive to stressful husbandry practices, leading to healthier and more productive progeny. 192 APPENDICES 193 APPENDIX A. ACTIVATION OF THE HYPOTHALAMlC-PITUITARY-ADRENAL, on STRESS, Axrs. Following exposure of an animal to stimuli perceived as stressful (ie. extremes in heat and cold, pain and fear, infection, common husbandry practices), par- aventricular nuclei (PVN) within the hypothalamus release corticotrophin releasing hormone (CRH) into the portal blood (a). CRH binds to CRH recep- tors (CRH-R1) in the anterior lobe (AL) of the pituitary (b), resulting in the release of adrenocorticotropin releasing hormone (ACTH) into the circulation. ACTH, in turn, targets the adrenal glands to synthesize and release glucocorti- coids (c). Genetic variation in disease susceptibility and production potential is observed following glucocorticoid challenge, pathogen exposure and the sub- sequent immune response (d), and an animals’ genotype will determine gene expression following a stress challenge. As the receptor for glucocorticoids, molecular genetic variation in GR and resulting glucocorticoid-responsive gene expression is an intriguing possibility to explain genetic variation In the health and productivity of husbandry-stressed cattle. 194 APPENDIX A. ACTNATION or THE HYPOTHALAMIC-PITUITARY-ADRENAL, on STRESS, Axrs. (a) \. ! I STRESS (b) ACTH (Q my... W) Animal's Genotype Genn ation SusceD lllty and Immune ” Productlgi Potential K Pathogen Response Exposure 195 APPENDIX B. REFERENCE or AMINO Acros. Listed are amino acid residue names and their three letter and one letter codes. One letter codes are used throughout this dissertation. In the right- hand column is listed the classification (or class) of the amino acid residue. Amino Acid Full Three-Letter Single-Letter Residue Name Code Code Classification Alanine Ala A Neutral, Nonpolar Cysteine Cys 0 Basic, Polar Aspartic Acid Asp D Acidic Glutamic Acid Glu E Acidic Phenylalanine Phe F Neutral, Nonpolar Glycine Gly 0 Neutral, Nonpolar Histidine His H Acidic Isoleucine lie I Neutral, Nonpolar Lysine Lys K Basic Leucine Leu L Neutral, Nonpolar Methionine Met M Neutral, Nonpolar Asparagine Asn N Neutral Proline Pro P Neutral, Nonpolar Glutamine Gln 0 Neutral, Polar Arginine Arg R Basic, Polar Serine Ser 8 Neutral, Polar Threonine Thr T Neutral, Polar Valine Val V Neutral, Nonpolar Tryptophan Trp W Neutral, Nonpolar Tyrosine Tyr Y Basic, Polar 196 APPENDIX C. CODING FOR ANIMALS USED iN THE MULTl-BREED STUDY or CHAPTER 4. Codes for animals making up the Breeds Populations are listed on the left-hand side, along with their breed. Barn identifiers and the source of the genomic DNA are listed on the right. Code Breed Barn lD Source of DNA A1 Angus a3d28 MSU Beef Teaching and Research A2 Angus a9603 MSU Beef Teaching and Research A3 Angus a9609 MSU Beef Teaching and Research A4 Angus a9707 MSU Beef Teaching and Research Br1 Brahman br21 Donated By Dr. Pagan, Puerto Rico Br2 Brahman br22 Donated By Dr. Pagan, Puerto Rico Br3 Brahman bi23 Donated By Dr. Pagan, Puerto Rico Br4 Brahman br26 Donated By Dr. Pagan, Puerto Rico Br5 Brahman br27 Donated By Dr. Pagan, Puerto Rico Bs1 Brown Swiss bs31 Donated by Dr. Mashek, Iowa B32 Brown Swiss bs38 Donated by Dr. Mashek, Iowa BS3 Brown Swiss bsS1 Donated by Dr. Mashek, Iowa Bs4 Brown Swiss bs60 Donated by Dr. Mashek, Iowa H1 Holstein h1175 MSU Dairy Teaching and Research H2 Holstein h2941 MSU Dairy Teaching and Research H3 Holstein h3159 MSU Dairy Teaching and Research H4 Holstein h3223 MSU Dairy Teaching and Research H5 Holstein h3229 MSU Dairy Teaching and Research He1 Polled Hereford he608 MSU Beef Teaching and Research He2 Polled Hereford he40e MSU Beef Teaching and Research He3 Polled Hereford he409 MSU Beef Teaching and Research He4 Polled Hereford he58c MSU Beef Teaching and Research He5 Polled Hereford hes35 MSU Beef Teaching and Research J1 Jersey j211 NorthStar Select Sire, Michigan J2 Jersey j311 NorthStar Select Sire, Michigan J3 Jersey j344 NorthStar Select Sire, Michigan J4 Jersey j355 NorthStar Select Sire, Michigan J5 Jersey j357 NorthStar Select Sire, Michigan 197 .32 950...: m_m__m TFI Eofioz DE E0: EDEED >=cmoEc9m on 9 “39286 Bo: 9m; 9569 E oc_cooco-mm0 $9: is .8325 So amp 0.59... c: 9.. 2.66.01 .2 “39285 30:30 3.202% 05 95:3 sm .8303 .e .2920 E nontowmu 339509 3.3 9.: 05m: “3280.: 30.080: 995 A3 9.. EB any v: A3 NI Emfior .2 wm_m__m «mo .Q £53210 2. mum-Dam m300 mz_m._.w._0_._ mum—IF Gz.z_<2mm 5:. Con. mZOFoamm—l wmaoamhm >¢<_._.mm._. .0 xazmna< 198 ‘ { InfinHU k in “a PM“! «in I. P2 ;» 4 Ag 2: E ..v mmhnazo 2. Oman...” mBOU mzfihmJOI mum—C. 02.2.(Emm NIP COL MZOFoammn— NEH-Foam...“ >m