|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 CHARACTERIZATION OF PAQR PROTEINS USING Saccharomyces cerevisiae: THE HUMAN MEMBRANE PROGESTIN RECEPTORS By JESSICA L. SMITH A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008
2 2008 Jessica L. Smith
3 To my family. Thank you for your support.
4 ACKNOWLEDGMENTS I would like to thank m y fam ily, especially my parents John and Jean, and all of my brothers and sisters (Jennifer, Jackie, Jeff, Jill, Janessa, and Josh) and their families for their love and support. I thank God for getting me through a lot of rough times in the past few years. I would also like to thank my advisor, Tom Lyons, for allowing me to work in his lab, for his support, for the projects, and for allowing me to begin developing as a sc ientist. I also thank Dr. Fanucci, Dr. Horenstein, Dr. Gulig, and Dr. de Crecy for their advice, support, and for serving on my committee. I would like to thank a ll of my friends and my fellow la b mates, Lisa, Nancy, Stephanie, Brian, Julie, Ibon, Liz, Lidia, A ndrew, Kim, Marilee, and Matt, fo r their support. I am thankful to Nancy Villa for her technical advice, friendship, and encouragement. I thank Brian for the initial work with the Class II PAQR s that motivated me to continue this aspect of the project. I thank Ibon for his motivation and the collaborative work we did to help each other progress. I thank Lisa for training me and for helpful sugge stions. I thank my undergraduate, Kim, for her help. I also thank my dog, Monkey, for listeni ng to my practice talks and for keeping me company. .
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES.........................................................................................................................9 ABSTRACT...................................................................................................................................12 CHAP TER 1 INTRODUCTION..................................................................................................................14 The Progestin and AdipoQ Receptor (PAQR) Family........................................................... 14 Structural Features of the PAQRs...........................................................................................15 History of the PAQRs: From Prokaryotes to Eukaryotes ....................................................... 16 PAQRs in Prokaryotes.....................................................................................................16 PAQRs in Eukaryotes...................................................................................................... 17 Saccharomyces cerevisiae as a Model Organism for PAQR Characterizations..................... 19 Summation..............................................................................................................................22 2 CLASS II PAQRS: MEMBRANE PROGESTIN RECEPTORS..........................................30 Introduction................................................................................................................... ..........30 Classic Steroid Effects.....................................................................................................30 Nonclassic Steroid Effects............................................................................................... 30 Membrane Progestin R eceptors (mPRs)......................................................................... 31 Class II PAQRs: Candidate Membrane Progestin Receptors (mPRs)..................... 33 Class II PAQRs: Novel Members of th e G-protein Coupled Recepto r (GPCR) Superfamily?.........................................................................................................36 Results and Discussion......................................................................................................... ..39 Overexpression of Some PAQRs Causes Ligand-Independen t Repression of FET3 .....40 Overexpression of Some PAQRs Causes Ligand-Dependent Repression of FET3 ........41 Reduced PAQR Expression Alleviates FET3-lacZ Repression by Constitu tively Active PAQRs..............................................................................................................41 Detection of PAQR Expression in Yeast......................................................................... 42 Dose-Dependent Repression of FET3-lacZ by Certain Steroids .....................................43 PAQR-Mediated Repression of FET3 Does Not Absolutely Require G-proteins ..........44 PAQR-Mediated Repression of FET3 Requires th e Presence of the PKA Yeast Homologue...................................................................................................................47 Identification of Novel A gonists of Class II PA QRs...................................................... 47 Materials and Methods...........................................................................................................50 Plasmids...........................................................................................................................50 Yeast Strains and Assays.................................................................................................51 Expression of Hemaglutanin (HA)-T agged Class II PAQRs.......................................... 52
6 Total Membrane Protein Extraction and Detection by W estern Blot.............................. 53 Summation..............................................................................................................................54 3 CLASS II PAQRS: LOCALIZATION AND TOPOLOGY.................................................. 68 Introduction................................................................................................................... ..........68 Class II PAQR Localization?..........................................................................................68 Class II PAQR Topology?............................................................................................... 68 Determining Protein Localiza tion and Topology in Yeast .............................................. 69 Green Fluorescent Protein........................................................................................69 Immunofluorescence................................................................................................70 The 3xHA-Suc2-His4C Dual Topology Reporter (DTR)........................................70 Results and Discussion......................................................................................................... ..72 Localization of Class II PAQRs with GFP...................................................................... 73 Protein Localization of Class II PA QRs by Immunofluorescence.................................. 74 Topology of Class II PAQRs by Immunof luorescence and the Dual Topology Reporter........................................................................................................................75 Materials and Methods...........................................................................................................78 Plasmids...........................................................................................................................78 Yeast Strains and Assays.................................................................................................78 Expression of GFP-tagged Class II PAQRs.................................................................... 79 Total Membrane Protein Extraction and Detection by W estern Blot.............................. 79 Immunofluorescence of S. cerevisiae Expressing the Class II PAQRs. .......................... 80 Fluorescence Microscopy Experiments........................................................................... 81 Histidinol Growth Assays................................................................................................81 Glycosylation Analysis....................................................................................................82 Summation..............................................................................................................................83 4 HEMOLYSIN III PROTEINS: BACTERIAL HOMOLOGUES OF PAQR PROTEINS .... 92 Introduction to Bacterial PAQRs............................................................................................ 92 Roles in Virulence?.........................................................................................................92 Other Possible Functions.................................................................................................93 Mechanisms of Action?................................................................................................... 94 Results and Discussion......................................................................................................... ..95 Overexpression of some bacter ial PAQRs causes repression o f FET3 ...........................95 Expression of HlyIIIs in E. coli .......................................................................................96 Materials and Methods...........................................................................................................98 Cloning of the Bacterial Hemolysin III Genes................................................................ 98 Expression of the Hemolysin III Proteins in S. cerevis iae ............................................100 galactosidase Reporter Assays ..................................................................................100 Expression of the HlyIII Homologues in Escher icia coli .............................................100 Summation............................................................................................................................102
7 5 FUTURE DIRECTIONS FOR CLASS II PAQR CHARACTERIZATIONS ..................... 107 Introduction................................................................................................................... ........107 Identification of Novel Interactions Between PAQRs and Other Proteins ...................107 Possible Intrinsic Enzymatic Activity of PAQR Proteins ............................................. 111 Expression of Class II PAQRs in Sf9 Insect Cells........................................................ 112 Ligand Binding Experiments......................................................................................... 113 Results and Discussion......................................................................................................... 113 Expression of Class II PAQRs in Insect Cells............................................................... 114 Ligand Binding Experiments......................................................................................... 115 Materials and Methods.........................................................................................................117 Plasmids.........................................................................................................................117 Expression of PAQR5 and PAQR6 in Sf9 Insect Cells................................................ 117 Ligand Binding Assays for P AQRs Expressed in Yeast ............................................... 118 Summation............................................................................................................................120 LIST OF REFERENCES.............................................................................................................124 BIOGRAPHICAL SKETCH.......................................................................................................133
8 LIST OF TABLES Table page 2-1 Primers used for cloning PAQRs into pGREG536 (listed from 5 3) ........................... 67 2-2 Codons in PAQR sequences that are rarely used in S. cere visiae .....................................67 3-1 Primers used for cloning PAQRs into pGREG575, pGREG600, and pJK90 (listed from 5 3).......................................................................................................................91 4-1 Primers used for cloning HlyIIIs (listed from 5 3) .....................................................106 5-1 Primers used for cloning PAQRs into pIEx-4 (listed from 5 3) ................................. 123
9 LIST OF FIGURES Figure page 1-1 The conserved motifs of the PAQR family of proteins..................................................... 23 1-2 Phylogenic tree of PAQR proteins..................................................................................... 24 1-3 Topology models for the PAQR proteins..........................................................................25 1-4 The yeast pheromone signal transduction pathway........................................................... 26 1-5 Yeast promoter-reporter assay for nuclear steroid receptor (nSR) analysis...................... 27 1-6 Yeast promoter-reporter assays of PAQR activation.........................................................29 2-1 Mechanisms of progestin steroids......................................................................................56 2-2 Overexpression of some PAQRs caus e repression of FET3-lacZ .....................................57 2-3 Repression of FET3-lacZ is ligand-dependent for som e PAQRs......................................57 2-4 Constitutive repression of FET3-lacZ can be alleviated and becom e ligand-dependent under low expression induction conditions........................................................................ 58 2-5 Expression of PAQRs in yeast can be detected by Western blot....................................... 59 2-6 Ligand-dependent repression of FET3-lacZ by Class II PAQRs occurs in a dosedependent m anner and is steroid specific..........................................................................61 2-7 Class II PAQR signaling model......................................................................................... 62 2-8 Repression of FET3-lacZ does not require yeast G -proteins. .........................................63 2-9 Expression of truncated hum an Class II PAQRs cause li gand-dependent repression of FET3-lacZ ......................................................................................................................64 2-10 Yeast cells that express constitutively activ e yeast G proteins do not demonstrate repression of FET3-lacZ ....................................................................................................64 2-11 Repression of FET3-lacZ requires yeast Tpk2p, a subunit for the yeast homologue of hum an PKA...................................................................................................................... ..65 2-12 Synthetic compounds tested fo r activation of Class II PAQRs .........................................65 2-13 Dose response of PAQR7 to RU-486................................................................................66 3-1 The yeast dual-topology reporter (DTR)........................................................................... 84
10 3-2 Overexpression of N-terminally GFP-tagged PAQRs causes repression of FET3lacZ ...................................................................................................................................85 3-3 Human Class II PAQRs do not fluoresce when tagged with GFP at the N-term inus........ 85 3-4 Western blot of GFP-tagged PAQRs expressed in yeast ...................................................86 3-5 Overexpression of C-terminally GFP-tagged PAQRs causes repression of FET3lacZ ...................................................................................................................................86 3-6 Western blot of GFP-tagged PAQRs expressed in yeast ...................................................87 3-7 Immunofluorescence of GFP-tagged and HA-tagged PAQRs..........................................88 3-8 Immunofluorescence of soluble G FP in unperm eabilized spheroplasts............................ 89 3-9 Overexpression of C-terminally DT R-tagged P AQRs causes repression of FET3lacZ ....................................................................................................................................89 3-10 Histidinol growth assays of DTR-tagged PAQRs ............................................................. 90 3-11 Endo H digestion of DTR-tagged PAQR6 and PAQR7....................................................90 3-12 Endo H digestion of RNase B............................................................................................90 4-1 Overexpression of some bacterial PAQR hom ologues causes repression of FET3lacZ ..................................................................................................................................103 4-2 SDS-PAGE of HlyIII-1 expre ssion s from the pKM260 vector....................................... 104 4-3 SDS-PAGE of HlyIII-2 expre ssion s using the pKM260 vector...................................... 104 4-4 SDS-PAGE of soluble cytoplasmic samp les and insoluble cytoplasm ic samples for HlyIII-2 in pKM260.........................................................................................................105 5-1 Western blot of PAQR 5 and PAQR 6 expressed in Sf9 insect cells. ............................. 121 5-2 Binding of 3H-PG in a single-point binding assay w ith total membranes isolated from yeast cells expressing PAQR5 or PAQR6.......................................................................122 5-3 Specific binding of 3H-PG in a single-point binding assay with total membranes isolated from yeast cells expressing PAQR6................................................................... 123
11 LIST OF ABBREVIATIONS 17,21-diOH PG: 17,21-dihydroxyprogesterone DTR Dual-topology reporter (3xHA-Suc2-His4C) FET3 gene encoding Fet3p Fet3p Ferrous transport protein GPCR G-protein coupled receptor HlyIII Hemolysin III 17-OH PG: 17 -hydroxyprogesterone 21-OH PG: 21-hydroxyprogesterone IZH Implicated in Zinc Homeostasis gene Izhp Implicated in Zinc Homeostasis protein lacZ gene encoding the -galactosidase enzyme mPR membrane progestin receptor PAQR Progestin and adipoQ Receptor PG Progesterone TS Testosterone 7TM Seven transmembrane
12 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CHARACTERIZATION OF PAQR PROTEINS USING Saccharomyces cerevisiae: THE HUMAN MEMBRANE PROGESTIN RECEPTORS By Jessica L. Smith May 2008 Chair: Thomas Lyons Major: Chemistry Some studies indicate that a few members of the Progestin and AdipoQ Receptor (PAQR) family (PAQR5, PAQR7, and PAQR8) act as memb rane progestin receptors (mPRs) and signal in a G-protein dependent manner, but these studies have been dis puted. To clarify the role of human PAQRs in progestin signaling, the yeast Saccharomyces cerevisiae was used as a simple model organism. When PAQRs are expressed at low leve ls, repression of the yeast promoter FET3 is observed in a FET3-lacZ promoter-reporter assay only when certain progestins are present. To explore the molecular mechanism by which the FET3 promoter is affected, a genetic mutational analyses was used. The effect of PAQRs on FET3-lacZ is not dependent on the presence of human or yeast G-proteins, but all human PA QRs and their yeast homologues share a common intracellular signaling mechanism in yeast involving tpk2, the yeast homologue of human PKA. In addition, evidence is demonstrated to support that human PAQR6 and PAQR9 act as progestin receptors, which confirms functional predictions based on bioinformatic sequence analyses. The localization and topology of some PAQR proteins is also di sputed, so studies of these characteristics were also pursued in yeast. Many problems were encountered and few conclusions could be drawn.
13 Initial characterizations of PA QR homologues from bacteria are also described. Although attempts were made to express these proteins in E. coli there was little success. However, the bacterial proteins were also expressed in S. cerevisiae for the FET3-lacZ assay and it was demonstrated that some of the prot eins tested can cause repression of FET3
14 CHAPTER 1 INTRODUCTION The Progestin and AdipoQ Receptor (PAQR) Family The hum an progestin and adipoQ receptor (PAQR) family of proteins consists of eleven proteins predicted to have least seven tran smembrane (7TM) domains (Tang et al., 2005). Homologues of PAQRs can be found in simple and complex eukaryotes as well as in prokaryotes (Lyons et al., 2004; Fernandes et al., 2005; Thomas et al., 2007). In simple eukaryotes, such as the yeast Saccharomyces cerevisiae the PAQRs are involved in metal and lipid metabolism (Lyons et al., 2004; Karpichev et al., 2002). In more complex eukaryotes, such as humans, the PAQRs act as receptors for adiponectin, which st imulates glucose uptake and fatty acid oxidation (Kadowaki and Yamauchi, 2005). The human PAQR s may also act as r eceptors for steroid hormones (Thomas et al., 2007). In prokaryotes, such as Bacillus cereus, B. anthracis, Bacteroides fragilis and Vibrio vulnificus the PAQR homologues, calle d hemolysin III proteins, are putatively involved in causi ng lysis of red blood cells (Bai da and Kuzmin, 1995; Klichko et al., 2003; Robertson et al., 2006; Chen et al., 2004). Besides understanding the normal physiological role s of the PAQRs, it is also important to understand their roles in pathophysiological cond itions. Thus far, the human PAQRs are known to be important in diseases such as diabetes and obesity (Kadowaki and Yamauchi, 2005) or to have potential importance in breast cancer development (Dressing and Thomas, 2007). In addition, the human PAQRs may be activated by endocrine-disrupt ing chemicals (EDCs), as the homologues in fish are (Tokumoto et al., 2007). This is especial ly interesting because EDCs are known to cause many health problems, including developmental abnormalities, reproductive problems and cancers (Newbold et al., 2006).
15 Structural Features of the PAQRs All PAQRs contain a motif, na m ed UPF0073, which is unique to the PAQR family (Tang et al., 2005). Each PAQR has the following motifs: ExxNxxH just before transmembrane domain 1 (TM1), SxxxH at the end of TM2, and HxxxH in TM7 (Figure 1-1). Other than these conserved motifs, there is little homology between the members, but further sub-classifications can be made based on sequence homologies such that the PAQRs can be divided into three classes (Fernandes et al., 2005): Class I (PAQR s 1-4), Class II (PAQRs 5-9), and Class III (PAQRs 10 and 11) (Figure 1-2). The PAQRs are expected to have similar orientations in the membrane. Evidence suggests that the Class I and Class III PAQRs have se ven transmembrane (7TM) domains with an intracellular N-terminus and extr acellular C-terminus (Yamauchi et al., 2003; Kim et al., 2006; Daley et al., 2005). For the human Class I PAQR s, immunofluorescence was used to determine that the N-terminus was inside, while the C-te rminus was outside the cell (Yamauchi et al., 2003). Based on experiments using the yeast du al-topology reporter (DTR) Suc2-His4C (which is described in more detail in Chapter 3) and sequence predictions, it was determined that the yeast Class I proteins (IZHs) had an extracellula r C-terminus (Kim et al., 2006). For the Class III proteins, C-terminal tagging with the alkaline phosphatase and green fluorescent protein was used to determine that the C-terminus of the E. coli homologue (called YqfA) was located in the periplasmic space rather than in the cytoplasm (Daley et al., 2005). This putative topology results in the clustering of the conserved motifs on the intracellular side of the membrane (Figure 1-3A), suggesting that they may be important for a conserved signaling mechanism (Tang et al., 2005). In contrast, predictions suggest that the Class II PAQRs have either 7TM domains like the other PAQRs (Zhu et al., 2003a; Thomas et al., 2007) or eight transmembrane domains (Lyons et
16 al. 2004; Fernandes et al., 2005) (F igure 1-3C). Some topological analyses ev en suggest that the Class II PAQRs have an orientation in the membrane that is opposite to th at of the Class I and Class III PAQRs (Thomas et al., 2007). Accordi ng to immunocytochemical analyses of cells expressing the human PAQR7, N-terminally dir ected antibodies were able to bind without permeabilization of cells, but C-terminally direct ed antibodies were not, suggesting that PAQR7 has an extracellular N-terminus and an intracellular C-terminus (T homas et al., 2007). With this topology, the Class II PAQRs would not have all of the conserved PAQR family motifs on the same side of the cell as the Class I and Class III PAQRs (Figure 1-3B), which could mean that the Class II PAQRs signal via a different mechanism. In fact, some experimental evidence suggests that the Class II PAQRs function as G-pr otein coupled receptors (GPCRs), which have a topology with the N-terminus outside the cell and the C-terminus inside the cell (Thomas et al., 2007; Karteris et al., 2006). History of the PAQRs: From Prokaryotes to Eukaryotes Findings from studies conducted on m embers of th e different classes of PAQRs reveal that these proteins vary widely in function. The PA QR family has a common eubacterial origin, as homologues to the Class III proteins have been characterized as eubacterial hemolysin III proteins (Tang et al., 2005). PAQRs in Prokaryotes The first PAQR to be studied was the Bacillu s cereus hemolysin III (HlyIII) protein (Baida and Kuzmin, 1995). Homologues of HlyIII from B. cereus (Baida and Kuzmin, 1996), Vibrio vulnificus (Chen et al., 2004), Bacillus anthracis (Klichko et al., 2003), and Bacteroides fragilis (Robertson et al., 2006) have all been purported to have hemolytic activity towards human red blood cells (RBC) by forming pores in RBC membra nes. However, the hemolytic activity of HlyIIIs has only been characterized with whole bacterial cells or cell extracts, rather than
17 purified HlyIIIs (Baida and Kuzmin, 1996; Chen et al., 2004; Klichko et al., 2003; Robertson et al., 2006). The possibility remains that the 7T M HlyIIIs are not hemolytic themselves, but instead act as membrane proteins that initia te a signaling cascade to activate or increase expression of a different protein that has hemo lytic activity. More on these proteins will be discussed in Chapter 4. PAQRs in Eukaryotes Mem bers of the eukaryotic P AQRs were first described when a study identified the gene for PAQR11 (also named monocyte-to-macrophage differentiation fact or 1, or MMD1) in mRNA differential display studi es (Rehli et al., 1995). The studi es were aimed at identifying genes with enhanced expression in mature macr ophages versus monocytes (Rehli et al., 1995). Much later, PAQR10, the closest human homologue of PAQR11, (also called macrophage/microglia activation factor) was identif ied in differential display studies of genes involved in microglia activation af ter brain trauma (Bra uer et al., 2004). Both studies predicted that these proteins have seve n transmembrane domains (Rehli et al., 1995; Brau er et al., 2004), but this has not been confirmed. The human Class III proteins have been proposed to act as ion channels or receptors (Rehli et al., 1995), alt hough ligands for these proteins have not been identified. In 2002, the first eukaryotic Class I P AQR was described in studies of a Saccharomyces cerevisiae protein called YOL002c (Che rry et al., 1998; Karpiche v et al., 2002). YOL002c was originally proposed to have a role in lipid a nd phosphate metabolism (Karpichev et al., 2002). Later, YOL002c as well as thre e closely related homologues in S. cerevisiae were shown to be involved in metal metabolism (Lyons et al., 2004 ). These proteins were named Izh1p, Izh2p, Izh3p, and Izh4p (Izh, or I mplicated in Z inc H omeostasis). In additi on, Izh2p acts as a receptor
18 for the plant defensin Osmotin, which is structur ally similar to human adiponectin (Narasimhan et al., 2005). The human Class I proteins (PAQRs 1-4) were first described in studies that identified them as receptors for adiponectin (also know n as adipoQ) (Yamauchi et al., 2003). These proteins, called AdipoR1 (PAQR 1) and AdipoR2 (PAQR2), act as receptors for adiponectin, a protein which induces fatty acid oxidation and glucose uptake (Yamauchi et al., 2003). PAQR1 and PAQR2 are plasma membrane localized, they ar e predicted to have 7TM domains, they have a topology with an intra cellular N-terminus and extracellular C-terminus (Yamauchi et al., 2003), and they have been shown to signal via th e scaffolding protein A PPL1 (adaptor protein containing pleckstrin homology domain, phosphotyrosine binding (PTB) domain and leucine zipper motif) (Mao et al., 2006). PAQR3 was recen tly reported to be a go lgi-localized protein that sequesters the kinase Raf-1 to this organe lle, resulting in extra cellular signal regulated kinase (ERK) pathway inhibition (Feng et al., 2007). PAQR3 was thus renamed Raf Kinase Trapping to Golgi (RKTG) (Fe ng et al., 2007). PAQR4 has not been characterized yet. Members of the Class II PAQRs (PAQRs 5-9) we re first described when studies identified genes that encode membrane proteins that putatively mediate the nongenomic effects of progestins (Zhu et al., 2003a; Zhu et al., 2003b). Originally identified in fish, the Class II PAQRs were shown to play a role in oocyte me iotic maturation induced by progestins (Zhu et al., 2003a; Zhu et al., 2003b). Because they have been proposed to act as membrane progestin receptors (mPRs), PAQR5, PAQR7, and PAQR8 are also known as mPR mPR and mPR (Zhu et al., 2003a; Zhu et al., 2003b). Both fish and human Class II PAQRs have been shown to cause decreases in cAMP levels that could be inhibited by the pertussis toxin, which ADP-ribosylates and inactivates G-proteins
19 (Zhu et al., 2003a; Thomas et al., 2007). This s uggests that the Class II PAQRs function as Gprotein coupled receptors (GPCRs) to affect cAMP levels (Thoma s et al., 2007; Karteris et al., 2006). Experiments also demonstrated that PAQR7 and inhibitory G-proteins coimmunoprecipitate, suggesting that these proteins directly interact (Thomas et al., 2007). In contrast, results from one study show that human Class II PAQRs localize intracellularly to the endoplasmic reticulum, do not bind progesterone and do not activate GPCR-related signaling cascades (Krietsch et al., 2006). Other studie s have shown that PAQR7 localizes to an intracellular tubuloretic ular network (Fernandes et al., 2005) and that PAQR8 is associated with lysosomes (Suzuki et al., 2001). Most of the studies on the Class II P AQRs have focused on PAQR5, PAQR7, and/or PAQR8. Further studies of th e Class II PAQRs, including PAQR6 and PAQR9, are needed to clarify the functions of this group of proteins. Saccharomyces cerevisiae as a Model Organ ism for PAQR Characterizations The yeast S. cerevisiae is a unicellular eukaryotic organism that is useful as a simple model system for the study of biomolecules and biologi cal processes of more complex organisms. Many of the genes encoded by the yeast genome have homologues in humans and many of the mechanisms involved in different cell processes, such as signal tran sduction, are conserved (Sturgeon et al., 2006). S. cerevisiae can be used for functional expression of eukaryotic proteins, including those from humans. S. cerevisiae serves as a good model organism to study human proteins because, unlike mammalian cells, yeast cells are easily to handle in molecular biological procedures and cellular components needed for hu man protein function are often c onserved in yeast (Mentesana et al., 2002). In addition, simple phenotypic scre ens of mutant receptors can be easily performed to identify amino acid residues involved in the signaling mechanisms (Ladds et al., 2005).
20 Thus far, yeast has proven to be an effec tive tool for the simplified study of many human proteins. For example, promoter-reporter assays have been developed for the study of G-protein coupled receptors (GPCRs) (Mentesana et al., 2002) and nuclear steroi ds receptors (McEwan, 2001). The successful use of yeast for characteri zations of GPCRs and steroid receptors makes this system particularly attractive for the char acterization of the Class II PAQRs, which are debated to be GPCRs and are putative receptors for progestin steroids (Krietch et al., 2006; Thomas et al., 2007). For the characterization of GPCRs, the si gnal transduction pathway of the yeast -factor mating pheromone receptor (Ste2p) is often used (Mentesana et al., 2002). Ste2 is a GPCR which couples to a heterotrimeric G-protein consisting of Gpa1p (G ), Ste4p (G ), and Ste18p (G ) (Mentesana et al., 2002). U pon ligand activation of Ste2, G is activated, the G dimer is released and activates a mitogen activated protein kinase (MAPK) cascade to activate the transcription factor Ste12p and ultimately lead to increased transcription of the FUS1 gene (Figure 1-4) (Mentesana et al., 2002). When the promoter of FUS1 is fused to a reporter gene, such as lacZ activation of the -pheromone mating pathway can be detected by measuring increases in transcription levels of FUS1 by a simple colorimetric -galactosidase assay (Mentesana et al., 2002). Because the yeast G-protein alpha subunit Gp a1p is highly homologous to the mammalian inhibitory G-protein Gi Dowell and Brown, 2002), it is able to couple to many human GPCRs (Ladds et al., 2005). Successful coupling of Gpa1p to human GPCRs allows for the characterization of various human GP CRs with a simple colorimetric FUS1-lacZ promoterreporter assay. Importantly, this simple assay has been used to screen for novel synthetic
21 agonists and antagonists (Pausch, 1997) and for simplified struct ure-function studies (Beukers and IJzerman, 2005) of GPCRs. For the characterization of nuclear steroid recepto rs (nSRs), yeast is an especially attractive organism to use because yeast cells do not have endogenous steroid receptors that could interfere with analyses of individual heterologous recep tors (McEwan, 2001). A promoter with a steroid response element (SRE), which is a DNA sequence recognized by a nSR, in combination with a reporter gene, such as lacZ, is used (Figure 1-5) (McEwan, 2001). With this system, the expressed nuclear receptor binds directly to the SRE to activ ate transcripti on, the amount of which can be measured in a colorimetric assay (McEwan, 2001). Receptors for steroids such as progesterone, estrogen and glucoc orticoids have all been charact erized with promoter-reporter systems in yeast and these types of studies have b een useful for structure-function studies as well as identifying signaling path way components (McEwan, 2001). In addition to human GPCRs a nd nuclear receptors, other PAQRs have been successfully characterized with promoter-repor ter assays in yeast as well (Figure 1-6) (Kupchak et al., 2007). Signal transduction pathways ac tivated by the Class I PAQRs from yeast (IZH proteins) and humans (PAQR1 and PAQR2) have been characterized with the FET3-lacZ promoter-reporter. Fet3p is involved in high-affinity iron uptake (A skwith et al., 1994). Typically, under low iron conditions, FET3 transcription is upregulated by Aft 1p (Yamaguchi-Iwai et al., 1995); however, when the yeast or human Class I PAQRs are overexpressed, FET3 transcription is repressed in a manner which is not dependent on Aft1p (Kupchak et al., 2007). The mechanism by which yeast or human Class I PAQRs repress FET3-lacZ was explored and it was found that this effect is dependent on the presence of cAMP-dep endent kinase (called protein kinase A, PKA) and AM P-dependent protein kinase (AMPK) (Kupchak et al., 2007).
22 Interestingly, the involvement of AMPK in PAQR1 and PAQR2 signaling has previously been reported (Yamauchi et al., 2003). These findings sugge st that at least all Class I PAQRs have an intracellular signaling mechanism that activates signal transduction pathways likely to be conserved from yeast to humans (Kupchak et al., 2007). Summation Thus far, the Class II PA QRs have been proposed to act as GPCRs and to act as receptors for progestin steroids (Thomas et al., 2007); how ever, both of these ideas have been disputed (Tang et al., 2005; Krietsch et al., 2006). Becau se yeast has been successfully used to study many human proteins, including GP CRs, nuclear steroid receptors, and some PAQRs, we have chosen to use this organism to characterize the human Class II PAQRs. Our results confirm that the human Class II PAQRs mediate cell signa ling in response to progestin steroids. Interestingly, we show that th e cell signaling in yeast is not dependent on the presence of human or yeast G-proteins, but is de pendent on PKA similarly to th e Class I PAQRs. We also attempted to localize these proteins in yeast and to determine their topology in the cell membrane. In addition, we show that overexpression of Class III PAQRs from bacteria causes cell signaling in yeast.
23 Figure 1-1. The conserved motifs of the PAQR family of proteins. The lo cation of the motifs are indicated by letters. Motif A (ExxNxxH) is located just before TM1, motif B (SxxxH) is located at the end of TM2, and motif C (HxxxH) is located just before TM7.
24 Figure 1-2. Phylogenic tree of PAQR proteins. The three classes of PAQRs are shown: Class I (PAQRs 1-4), Class II (PAQRs 5-9), and Class III (PAQRs 10-11). The tree is rooted with the distantly related Class III PAQRs. The branch lengths represent the distance between sequences with 0.1 substitutions per site according to the scale bar. Numbers at the nodes refer to the probability that a particular grouping is made per 1000 trees drawn. This figure was generated by Tom Lyons with the ClustalX program.
25 Figure 1-3. Topology models for the PAQR protei ns. In panel A, the Class I and Class III PAQRs have the N-terminus located inside the cell, while the C-terminus is located outside the cell. This topology would result in the presence of the conserved motifs (represented by the encircled letters A, B, or C) on the intracellular side of the membrane. Motif A is ExxNxxH, motif B is SxxxH, and motif C is HxxxH. In panel B, the Class II PAQRs have been suggested to have a topology similar to the GPCRs, with an extracellular N-terminus and an intracellular C-terminus While all three coserved motifs are depicted here as bei ng on the outside surface, Zhu et al. (2003a) have proposed an alternativ e topology in which the firs t two conserved motifs would be outside and the last one would be inside. In panel C, the Class II PAQRs have also been suggested to have eight transmembrane domains.
26 Figure 1-4. The yeast pheromone signal transduc tion pathway. Upon activation of Ste2p by the pheromone -factor, G is activated, the G dimer is released and activates a mitogen activated protein kinase (MAPK) cascade that leads to activation of the transcription factor Ste12p and ultima tely increased transcription of the FUS1 gene. [This figure was modified from Mentes ana PE, Dosil M and Konopka JB (2002) Functional assays for mammalian G-prot ein-coupled receptors in yeast. Methods Enzymol 344:92-111 with permission from Elsevier (Figure 2, page 96).]
27 Figure 1-5. Yeast promoter-reporter assay for nuclear steroid receptor (nSR) analysis. Upon diffusion of the steroid into the cell, the st eroid binds the nSR to activate it. The nSR recognizes and binds to a DNA sequence (ste roid response element, SRE). This causes transcription of lacZ which can be measured via the -galactosidase assay.
28 A B
29 C Figure 1-6. Yeast promoter-repor ter assays of PAQR activat ion. Yeast cells are doubly transformed with the pFET3-lacZ and GAL1 -PAQR vectors. The cells are grown in LIM containing galactose as an inducer of PAQR expression. In panel A, when no PAQRs are present, the FET3 promoter is activated causing high levels of lacZ transcription, and thus high levels of -galactosidase as determined by a colorimetric assay of ONPG hydrolysis. In panel B, when a PAQR is overexpressed but not activated, PAQR inititated signal transduction does not occur and the FET3 promoter is activated as described above. In pa nel C, when a PAQR is overexpressed and activated, either constitutively or via a lig and, signal transduction occurs, resulting in repression of FET3-lacZ This results in decreased -galactosidase activity, as determined by decreased hydrolysis of ON PG measured in a colorimetric assay.
30 CHAPTER 2 CLASS II PAQRS: MEMBRANE PROGESTIN RECEPTORS Introduction Steroid hormones are known to affe ct cell physiology by at leas t two processes. The most studied physiological effects are known as the cla ssic steroid effects, while the other effects are called the nonclassic steroid effects. Classic Steroid Effects Upon diffusi on of the steroid across the plasma membrane, the steroid binds to a nuclear steroid receptor (nSR) (Chen and Farese, 1999). The activated receptor translocates to the nucleus where it binds steroid response elem ents (SRE) in the DNA sequence and initiates transcription of target genes (Fi gure 2-1) (Falkenstein et al., 2000). The steroid-dependent effect on the genome is sensitive to inhibitors of transc ription and translation (F alkenstein et al., 2000). The alterations in gene transcription occur with in hours to days, and so the classical steroid effects are considered to be nonrapid (Farach-Carson and Da vis, 2003; Losel and Wehling, 2003). Nonclassic Steroid Effects In addition to the classic effects of steroids at the genom ic level, it has long been recognized that steroids can also have rapid, nongenomic effects (F alkenstein et al., 2000), which do not absolutely require alterations in gene tr anscription (Losel and Wehling, 2003; Thomas et al., 2002). These effects are referred to as noncla ssical steroid effects (Losel and Wehling, 2003; Thomas et al., 2002) and are not blocked by transc ription inhibitors (Fal kenstein et al., 2000; Losel and Wehling, 2003). Instead of involving soluble nuclear receptors rapid steroid effects are thought to be initiated at the plasma membrane by membra ne-bound steroid receptor s (Falkenstein and
31 Wehling, 2000). Activation of the nonclassical steroid receptors alters production of second messengers, such as cAMP, free intracellular cal cium, and phosphoinositides, and causes protein kinase cascade activation (Falkenstein et al., 2000). While th e nonclassical effects do not absolutely require transcripti on, the downstream eff ects of altered signaling pathways could ultimately affect transcription levels of some ge nes (Falkenstein et al., 2000; Losel and Wehling, 2003). The mechanism by which the classical genomic st eroid effects occur is attributed to the nuclear steroid receptors; howev er, many points of contention surr ound the issue of nonclassical steroid effects. Although it is widely accepted that steroid hormones can have rapid, nongenomic effects, the existence of novel, memb rane-bound steroid receptors is contended for several reasons. First, some st udies suggest that the classic nu clear steroid receptors can be membrane localized and may mediate at leas t some of the nonclass ical steroid effects (Falkenstein et al., 2000). Also, reproducib ility problems exist for studies that have demonstrated certain proteins to have characteristics of membra ne steroid receptors (Wehling et al., 2007). More details on these problems will be di scussed later in this chapter. Clearly, the mechanisms of nonclassical steroid effects are not established, so further studies of candidate membrane steroid receptors are needed. Membrane Progestin Receptors (mPRs) The first steroid for which rapid effects were observed was progesterone (Falkenstein et al., 2000). P rogestins, particularly progesterone, are important steroids th at are involved in many different biological processes, including oocyte maturation, maintenance of pregnancy, stimulation and inhibition of cell proliferati on, prevention of bone mass loss, spermiogenesis, modulation of sexual behavior, a nd alterations in immune system response (Graham and Clarke, 1997; Oettel and Mukhopadhyay, 2004; Dosiou et al., 2008 ). Importantly, progestins have also
32 been used as therapeutics for many reproductive disorders, such as endometrial cancers, because they limit the growth and pro liferation of such cancerous cells (Zhou et al., 2007). Progestins and antiprogestins have also been used to treat abnormal bleeding in the uterus, prevention of miscarriage, and prevention of prem ature labor (Fernandes et al., 2007). Like other steroids, progestins exert their effects via either classical genomic mechanisms or rapid, nongenomic mechanisms. Classic genomi c progestin effects are mediated by either isoform (A or B) of the nuclear progestin r eceptors (nPR), which bind to progestin response elements (PREs) in the promoters of target genes, to activate gene transcription (Figure 2-1) (Graham and Clarke, 1997). Nonclassical effects of progestins include activation of phosphatidy linositol 3-kinase (PI3K), activation of MAPK, and influx of Ca2+ (Norman et al., 2004). The mechanism by which these nonclassical effects occu r is controversial but is thought to be initiated at the plasma membrane (Falkenstein et al., 2000 ) (Figure 2-1). The existence of novel, nonclassical progestin receptors is debated, but there is much evidence to support their existence. For instance, the rapid action of progesterone on amphibian oocyte maturation can occur in enucleated oocytes, suggesting that oocyte maturation is a nongenomic effect of the steroid (Morrill and Kostellow, 1999). Also, rapid progesterone effects on sperm cells can not be inhibited by the potent nPR antagonist RU-486 (Baldi et al., 1991). Furthermore, the nPRs are not found to be expressed at detectable levels in human sperm (Castilla et al., 1995). Likewise, while progesterone is well known to affect cytokine production in lymphocyt es, the presence of nPRs in lymphocytes is often undetectable (Dosiou et al., 2008). As for other steroids, the existence of me mbrane-bound nongenomic progestin receptors is debated, especially because classical nPRs can be membrane-associated (Martinez et al., 2007)
33 and capable of mediating many of the rapid nongeno mic effects (Evaul et al., 2007). Also, some proteins that were originally proposed to act as membrane-bound progestin receptors (mPRs) could have a completely unrelated function (Weh ling et al., 2007) or were not found to localize to the plasma membrane and to have an unknown function (Krietsh et al ., 2006; Fernandes et al., 2007). One group of candidate mPRs is part of the Progestin and AdipoQ Receptor (PAQR) family (Tang et al., 2005) and have been proposed to initiate some of the nonclassical effects of progestin steroids, such as activating mitogenactivated protein kinases (MAPKs) and inhibiting adenylyl cyclase to cause decr eases in cyclic adenosine mono phosphate (cAMP) levels (Zhu et al., 2003a; Zhu et al., 2003b). Some studies show that the mPRs that are part of the PAQR family, called PAQR5 (or mPR ), PAQR7 (or mPR ), and PAQR8 (or mPR localize to the cell membrane, display membrane binding to pr ogestins, and function as G-protein coupled receptors (GPCRs) to affect signaling pathways in the cell (Zhu et al., 2003a; Zhu et al., 2003b; Hanna et al., 2006; Thomas et al., 2007). Howe ver, results from one study show that these proteins localize intracellularly to the endopl asmic reticulum, do not display membrane binding to progestins, and do not activate signaling cascad es in response to progestins (Krietsch et al., 2006). Class II PAQRs: Candidate Membrane Progestin Receptors (mPRs) Mem bers of the PAQR family of proteins are characterized by having at least seven transmembrane domains (7TMs) and certain cons erved amino acid sequence motifs (as described in Chapter 1). Phylogenetic sequence analyses s how that the human PAQR family of proteins consists of eleven members that can be subdi vided into three classe s based on their sequence homology (Fernandes et al., 2005): Class I (PAQRs 1-4), Class II (PAQRs 5-9), and Class III
34 (PAQRs 10 and 11) (Figure 1-2). Despite some di stinctions between classes, such as sequence homologies, proposed ligands, and physiological functions, it has b een suggested that all of the PAQR proteins function via a conserved intrace llular signaling mechan ism (Tang et al., 2005), but this remains to be determined. Although the PAQRs are similar to GPCRs in that they have 7TM domains and the Class II PAQRs have been proposed to act as GPCRs (Thomas et al., 2007), it has been suggested that none of the P AQRs function via coupling to G-proteins (Tang et al., 2005). Furthermore, it has been suggest ed that the Class II P AQRs have an eighth potential transmembrane domain (Lyons et al., 2004; Fernandes et al., 2005). While the Class II PAQRs from many different species have been partially characterized (Zhu et al., 2003a; Zhu et al., 2003b; Hanna et al., 2006; Karteris et al., 2006; Krietsch et al., 2006; Thomas et al., 2007; Josefsberg Ben-Ye hoshua, 2007), a physiological role for the human homologues is unclear. Studies indicate that the human Cla ss II PAQRs are differentially expressed in pregnancy tissues and may be important in regulating the onset of labor (Karteris et al., 2006; Fernandes et al., 2005). Importantl y, PAQR7 may have a ro le in breast cancer development, as expression is increased in breast tumors versus normal tissue (Dressing and Thomas, 2007). Also, PAQR7 and PAQR8 may be involved in the effects of progesterone on immune system response, as these two proteins we re detected in T-lymphoc ytes and Jurkat cells (Dosiou et al., 2008). Whether the Class II PAQRs function as membrane-bound progestin receptors is debated in the literature. Two recent st udies aimed at characterizing the human Class II PAQRs and used the same cell line, MDA-MB-231 human breas t cancer, which does not express the nPRs (Thomas et al., 2007; Krietsch et al., 2006). Ho wever, these studies f ound conflicting results regarding progestin binding to and involvement of Class II PAQRs in progestin-induced cell
35 signaling events (Thomas et al., 2007; Krietsch et al., 2006). According to Thomas et al. (2007), membranes from MDA-MB-231 cells transfected with PAQR7 bind progesterone 2.5 times more than untransfected cells. Subce llular fractionation from these cells showed that the increased progesterone binding detected in radioligand binding assays was specific for plasma membranes rather than organelle membranes (Thomas et al., 2007). Furthermore, changes in second messenger production (cAMP) in human cells tr ansfected with PAQR7 occurred only when progesterone was added (Thomas et al., 2007). These changes were not observed for the untransfected control (Thomas et al., 2007). Despite the results obtained by Thomas et al. (2007), anot her group did not find evidence to support that the Class II PAQR s are membrane progestin receptors. Krietsch et al. (2006) used the same cell type that Thomas et al (2007) used (MDA-MB-231) as well as human embryonic kidney (HEK) 293 cells. These studies showed that neither exogenously expressed nor endogenously expressed Class II PAQRs were localized to th e plasma membrane (Krietsch et al., 2006). Furthermore, the same group did not find evidence to support that cAMP levels changed in Class II PAQR transfected cells in response to progest erone, nor did crude membranes from these cells bind progesterone more than untransfected cells (Krietsch et al., 2006). One problem with the approach of Krietsch et al. (2006) in trying to demonstrate specific progesterone binding to membranes of cells expressing the Class II PAQRs is that they used crude membrane preparations. (In these radioligand binding assays, the same amount of total protein was used for the vector control and the Class II PAQR samples. Specific binding was defined as the difference between the total binding, when only radiolabeled progesterone was used, and the nonspecific binding, which is bindi ng in the presence of 1000-fold excess cold
36 progesterone). Because each sample had the sa me amount of total prot ein and because it is likely that the Class II PAQRs are not abundant in crude membrane preparations, it is not surprising that there were insignificant differences in specific binding for the vector versus the Class II PAQR samples. In contrast, Thomas et al. (2007) used the membranes in which the PAQRs localized and observed specific binding fo r the PAQR7 samples in binding reactions that were conducted in a manner similar to that which was used by Krietsch et al. (2006). As a result of the conflicts in the literatu re, the involvement of the Class II PAQRs in progestin-induced cell signaling is unclear. Therefore, the use of a simple system for the characterization of the Class II P AQRs could help to clarify their role in progestin-induced cell signaling. In this chapter, the use of the simple eukaryotic organism Saccharomyces cerevisiae to characterize the Class II PAQR s as mediators of progesteroneinduced cell signaling events is described. The justification for the use of this organism was described in Chapter 1. Class II PAQRs: Novel Members of the G-protein Coupled Receptor (GPCR) Superfamily? Thus far, a conserved intracellular signaling m echanism for the PAQR family has not been established. There are certain amino acid se quences that are conserved in each PAQR throughout evolution (Figure 1-1). These motifs include ExxNxxH just before TM1, SxxxH at the end of TM2, and HxxxH in TM7. If these motifs are all located inside th e cell, they could be involved in a conserved intracellular signaling mechanism (Tang et al., 2005). Current experimental evidence in the literature suggests that the PAQRs do not share a common signaling mechanism. For instance, the Class I P AQR signaling mechanism requires the scaffolding protein a daptor protein containing p leckstrin homology domain, p hosphotyrosine-binding domain, and l eucine zipper motif (APPL1) (Mao et al., 2006). The Class II PAQRs ha ve been suggested to
37 act via coupling to inhibitory G-proteins (Gi-proteins) (Thomas et al., 2007). This claim is partly based on cAMP measurements and Gi-protein-PAQR co-immunoprecipitation experiments. First, activated Gi-proteins inhibit adenylate cyclase, which causes decreased cAMP production. For PAQR7-transfected human cel ls treated with progesterone (for human PAQR7) or 17,20 21-trihydroxy-4-pregnen-3-one (for sea trout PAQR7), cAMP levels decreased (Thomas et al., 2007). While Thomas et al. (2007) demonstrated th at this effect could be blocked for sea trout P AQR7 upon addition of pertussi s toxin, an inhibitor of Gi activation (Thomas et al., 2007), similar inhibitory experi ments were not conducted for the human PAQR7. Thus, it is unclear if the decrease in cAMP associated with the human proteins can be similarly inhibited by pertussis toxin. Thomas et al. (2007) also claimed that th e human PAQR7 co-immunoprecipitated with Giproteins, but the experimental evidence was not clear. After immunoprecip itation with an antiGi antibody, Western blots were performed and the membranes were probed with an antiPAQR7 antibody (Thomas et al., 2007). The band for human PAQR7 was quite faint (Thomas et al., 2007) and there was no negative control shown to demonstrate that the same faint band was not present when PAQR7 was not expressed. In addition, Gi-proteins and the PAQR7 proteins both localize to membranes. Im munoprecipitation of membrane proteins likely leads to coimmunoprecipiatation of other memb rane proteins, whether there is direct interaction between the two proteins or not. Although no cell signaling mechanis m has been proposed for the human Class III PAQRs, the bacterial homologues (Hemolysin IIIs) have been suggested to oligomerize and form pores in the membranes of human red blood cells to cause cell lysis (Baida and Kuzmin, 1996).
38 While the conserved PAQR sequence motifs would be expected to be on the same side of the plasma membrane if they are involved in a conserved signaling mechanism (Tang et al., 2005), studies in the literature sugge st that the three classes of PA QRs have different orientations in the cell membrane. Two of the PAQRs (PAQR1, AdipoR1, and PAQR2, AdipoR2), two of the yeast Izh proteins (Izh1p and Izh4p) and one of the bacterial PAQR homologues (YqfA) have been determined to have a topology with an in tracellular N-terminus and an extracellular Cterminus (Figure 1-3A) (Yamauchi et al., 2003; Kim et al., 2006; Daley et al., 2005), which is opposite to that of GPCRs. In contrast, the Clas s II PAQRs were found to have a topology that is similar to that of GPCRs (Figure 1-3B) (T homas et al., 2007). If these topological determinations are correct, then the conserved amino acid sequence motifs of the PAQR family would not be located on the same side of the cell for all PAQRs. Th is does not support the involvement of the conserved motifs in a common intracellular signaling mechanism. Much debate continues regarding Class II PAQR function, signaling mechanism, localization, and topology. To alleviate some of the contr oversies, we used the yeast Saccharomyces cerevisiae as a simple model organism. S. cerevisiae is a good model organism to study human proteins, such as GPCRs, because many of the cellular components needed for their function are conserved in this organism (Mentesana et al., 2002). However, to be functional in yeast, mammalia n GPCRs often require chimeras of yeast/mammalian G-protein subunits (Mentesana et al., 2002). Yeast has al so been useful for studying nuclear steroid receptors because, unlike mammalian cells, yeast lack nuclear steroid receptors that could interfere with analysis of indi vidual heterologous steroid rece ptors (McEwan, 2001). Thus, yeast can simplify the study of the Class II PAQR role in progestin steroid signaling.
39 Some members of the PAQR family of proteins have been partially characterized in yeast (Lyons et al., 2004; Kupchak et al., 2007). Becau se we are interested in characterizing all members of the human PAQR family, we have chos en to use yeast to study the PAQRs that have been proposed to act as mPRs. Our studies of the Class II PAQRs describe the first characterizations of any of the human membrane steroid receptors in yeast. We used a yeast promoter-reporter ( FET 3lacZ ) assay that has been previous ly used to characterize the homologous adiponectin receptors (Kupchak et al., 2007). Our results confirm that the human Class II PAQRs are involved in progestin-mediate d signaling, which has b een disputed (Krietsch et al., 2006, Thomas et al., 2007). We show that the progestin-mediated effect of the Class II PAQRs on FET 3-lacZ response is not dependent on the pres ence of human or yeast G-proteins. Thus, we have evidence that human Class II PAQRs are able to signal in a G-protein independent manner. We show that, like the Cl ass I PAQRs, the human Class II PAQRs require the yeast homologue of human protein kinase A (called Tpk2p) for signaling. In addition, we provide the first experimental evidence that human PAQR6 and PAQR9 mediate cell signaling in response to progestin, which confirms functional predictions based on bioinformatic sequence analyses (Thomas et al., 2007). We also used the FET 3-lacZ assay to perform structure-function studies and identify new agonist s for the Class II PAQRs. Results and Discussion Previous attem pts have been made to establish whether the human Class II PAQRs mediate progestin signaling; however, these studies yi elded conflicting results (Thomas et al., 2007; Krietsch et al., 2006). In addi tion, because the classi cal progesterone receptors seem to be capable of mediating many of the nongenomic effect s that are attributed to membrane progestin receptors (Evaul et al., 2007), the existence of distinct membrane progestin receptors is debated. Thus, classification of the Class II PAQRs as memb rane progestin receptors is still controversial.
40 To clarify the role of human Class II PAQRs in progestin signaling, we chose the simple model organism Saccharomyces cerevisiae and a promoter-reporter assay that has been previously used to characterize the related adip onectin receptors (Kupchak et al., 2007). With this system, we have found that expression of any of the PAQRs in the presence of their proposed ligands causes negative regulation of FET3 gene expression under conditions (low amounts of iron in the media) that no rmally cause induction of this gene. Overexpression of Some PAQRs Causes Ligand-Independent Repress ion of FET3 Under conditions of low iron media (LIM) and full induction of PAQR expression via the GAL1 promoter, some of the human PAQRs cause repression of the yeas t high affinity iron uptake gene FET3 Previously published data demons trated that all f our of the PAQR homologues in yeast (Izh1p, Izh2p, Izh3p, and Izh4p) repressed FET3-lacZ (Kupchak et al., 2007). Control experiments showed that this e ffect was not due to problems with growth, transcription, translation or reporter activity; nor were these effects due to expression of membrane proteins or the re porter construct (Kupchak, 2008). In addition, it was demonstrated that hete rologously expressed human Class I PAQRs (PAQR1 and PAQR2) could repress FET3-lacZ in an adiponectin-dependent manner (Kupchak et al., 2007). Thus, human PAQR homologues can be functionally expresse d in yeast and it was hypothesized that all of the P AQRs, including the Class II PAQR s, could cause repression of FET3-lacZ via a conserved si gnaling mechanism. Here, we tested the effects of 6x-histid ine tagged Class II and Class III PAQR overexpression (via the GAL1 promoter) on FET3-lacZ in LIM and found that some (PAQR5, PAQR8, and PAQR11) but not all of these proteins could cause repression of FET3-lacZ under full protein expression induction (2 % galactose) (Figure 2-2).
41 Overexpression of Some PAQRs Causes Ligand-Dependent Repression of FET3 Under growth and induction conditions sim ilar to those described above, it was dem onstrated that repression of FET3-lacZ occurs in cells overexpressing PAQR6, PAQR7, or PAQR9 when their proposed liga nd, progesterone, was included during growth (Figure 2-3). Control experiments showed that progesterone did not affect FET3 in yeast carrying the empty expression vector indicating the effect of progesterone was not nonspecific. In addition, progesterone did not cause repression of FET3-lacZ in cells overexpressing the Class III PAQR10 protein (Figure 2-3), nor did it cause repression in ce lls overexpressing the Class I PAQR2 protein (Garitaonandia, 2008), demonstra ting that overexpression of other membrane proteins does not cau se repression of FET3-lacZ in the presence of progesterone. Reduced PAQR Expression Alleviates FET3-lacZ Repression by Constitu tively Active PAQRs Interestingly, within each of the classes of human PAQRs, the proteins can be divided into sub-groups based on sequence relatedness (Figur e 1-2). For Class I PAQRs, PAQR1 and PAQR2 are highly similar. For Class II PAQRs, PAQR5 and PAQR6 are highly similar to one another and PAQR7 and PAQR8 are highly similar to one anothe r, while PAQR9 is the most distant member. For Class III PAQRs, PAQR 10 and PAQR11 are the only members, but are highly related. Based on previous results fo r Class I PAQRs (Kupcha k et al., 2007) and the results presented here, it is a pparent that one member of each sub-group causes repression of FET3-lacZ during full protein expression inducti on (PAQR1 (Kupchak et al., 2007), PAQR5, PAQR8, and PAQR11 (Figure 2-2), while the other member of each sub-group does not. When some receptors are overexpressed, they can signal in the absence of their ligands and the level of this constitutive signaling is dependent on the leve l of expression (Tiberi and Caron, 1994; Milligan, 2003). Therefore, it was hypothesized that the level of FET3-lacZ repression
42 was dependent on the level of PAQR expr ession. This was tested by performing FET3-lacZ assays with PAQR5 cultures that had varying concentrations of indu cer added (galactose) (Figure 2-4A). As expected, when less inducer was present, PAQR5-me diated repression of FET3-lacZ was alleviated, but with increasing amounts of inducer, the repression of FET3-lacZ occurred in a dose-dependent manner. Simila r results were obtained with IZH2p and PAQR1 (Kupchak et al., 2007). Interestingly, for all PAQRs tested, when GAL1 -mediated expression was not fully induced (0.05% galactose), even the constitutively activ e Class II PAQRs required the addition of their proposed ligands (Figure 2-4B). Control experiments showed that the distantly related Class III PAQRs (PAQR10 and PAQR11) (Figure 2-4B) and the adiponectin receptors (Class I PAQRs) did not respond to progest erone (Garitaonandia, 2008). Detection of PAQR Expression in Yeast In addition, it was hypothesized that, when expr ession is fully induced (2% galactose), the constitutively activ e Class II sub-group members express at higher levels than the nonconstitutive members. To determine this, semi-quantitative Western blots were performed. Initial attempts to demonstrate expression of 6x-histidine tagged Class II PAQRs were unsuccessful, as PAQR5 was the only protein successfully detected (data no t shown). Therefore, 7x-HA-tagged PAQRs were constructed and expr ession was successfully detected by Western blot at approximately their expected sizes (4750 kDa) (Figure 2-5A). These constructs were also functional (Figure 2-5B). Our results indi cate that there is not a major difference in expression levels, but PAQR5 and PAQR8 do expr ess slightly more than PAQR6 and PAQR7. The frequency of codons present in the P AQR sequences that are rarely used in S. cerevisiae are given in Table 2-2. According to this, PAQR5 has fewer rarely used codons than the other Class II PAQRs, suggesting that this prot ein could express more readily in S. cerevisiae On the other
43 hand, the number of rarely used codons in P AQR6, PAQR7, and PAQR8 is similar. Thus, there is no correlation between the fre quency of rarely used codons in the PAQR sequences and their expression levels in S. cerevisiae. It has been noted that very different levels of constitutive activity can occur for closely related GPCRs, ev en when their expression levels are similar (Milligan, 2003). Because the 6x-histidine tag is significantly smaller (840 Da) than the 7x-HA tag (~7600 Da) and because several attempts to clone P AQR9 into the pGREG536 vector for 7x-HA tagging were unsuccessful, most of the FET3-lacZ characterizations were performed with the 6xhistidine tagged protein. Dose-Dependent Repression of FET3-lacZ by Certain Steroids For the Clas s II PAQRs, FET3-lacZ repression in the presence of progesterone occurred in a dose-dependent manner (Figure 2-6A). These findings suggest that PAQR receptor overexpression and activation modulate a similar signaling pathway in yeast and that some PAQRs have basal signaling in the absence of their activating liga nd. To study the ligand specificity of the Class II PAQRs, we tested se veral compounds that are similar to progesterone, 17 -hydroxyprogesterone, 21 -hydroxyprogesterone, 17,21-dihydroxyprogesterone, and testosterone (Figure 2-6B), with a previously studied Class II PAQR (PAQR5) as well as the uncharacterized Class II PAQR s (PAQR6 and PAQR9) (Figure 2-6 C-F). Thomas et al., conducted competitive binding assays with membranes containing overexpressed PAQR7 and showed that some of the compounds we tested ha d relative binding affinities (RBA, measured as the ability of unlabeled steroid, compared to unlabeled progesterone, to displace 50% of radiolabeled progesterone in competitive bindi ng assays) of 22.4% (testosterone), 19.7% (21hydroxyprogesterone), and less than 1% (17 -hydroxyprogesterone) (17,21-
44 dihydroxyprogesterone was not tested) (Thomas et al., 2007). Interestingly, like progesterone, we found that all compounds, except fo r testosterone, caused repression of FET3-lacZ When PAQRs were expressed, similar results were obtained for PAQR7 in our lab ( Brian Kupchak, personal communication ). Interestingly, while Thomas et al. (2007) detected significant binding of testosterone, but not 17 -hydroxyprogesterone, we obser ved a cellular response to 17 hydroxyprogesterone but not testoster one. It has not been reported whether testosterone causes a Class II PAQR-mediated cellular response in mammalian cells models. Experiments performed by Brian Kupchak in our lab show that the proge sterone dose-response for PAQR5 shifts to the right in the presence of testoste rone (Figure 2-6G), suggesting th at testosterone could be having an antagonistic effect in our system. Whether testosterone is bindi ng to the Class II PAQRs expres sed in our yeast cells is unknown, but will be the subject of future studies. Regardless, the progestin response of yeast cells expressing the Class II PAQR s strongly suggests that these receptors sense and respond to progestins and that they are, indeed, progestin receptors. PAQR-Mediated Repression of FET3 Does No t Absolutely Require G-proteins The human Class II PAQRs are proposed to act as Gi-protein coupled receptors that cause decreased adenylate cyclase activity (Thomas et al., 2007; Dressing and Thomas, 2007) (Figure 2-7). Upon activation of Gi-proteins, GDP bound to the Gi-proteins is converted to GTP, which causes inactivation of adenylate cyclases and a subsequent decrease in cAMP. For cell membranes from PAQR7-transfected human cel ls treated with progesterone, immunoassay experiments demonstrated that cAMP production decreased and this eff ect could be blocked upon addition of pertussis toxin, an inhibitor of Gi activation (Thomas et al., 2007). Coimmunoprecipitation experiments with anti-Gi-protein antibodies demonstrated significant
45 increases in membrane-bound GTP (T homas et al., 2007). Despite th is experimental evidence, it has also been previously suggested that none of the PAQRs are coupled to G-proteins (Tang et al., 2005). Therefore, it was of interest to dete rmine the involvement of G-proteins in Class II PAQR signaling in our yeast system. The Class II PAQR activity observed in the FET3-lacZ assays does not require the presence of human G-proteins, as these proteins are not expresse d in our yeast system; however, yeast have two genes encoding heterotrimeric G-proteins related to human Gi ( GPA1 and GPA2 ) (Slessareva et al., 2006), so it is possible that the PAQRs function as Gpa1por Gpa2plinked GPCRs in yeast. Thus, the Cl ass II PAQR-mediated repression of FET3-lacZ in the absence of the yeast G-proteins was tested. Knock-out mutants for each of the known yeast Gproteins (Gpa1p and Gpa2p) were used in the FET3-lacZ assay (Figure 2-8). These results show that neither G-protein is require d for the response of PAQR5 to progesterone. A double-mutant for the GPA genes could not be obtained because the GPA1 single-knockout is lethal, so knockouts of GPA1 must be in combination with a mutation in STE7 (Henrick Dohlman, personal communication). A triple knockout in GPA1 STE7 and GPA2 would be required, and this was not readily available. Regardless, the GPA1 and GPA2 genes are involved in completely separate signaling pathways and do not have overlapping functions (Slessareva and Dohlman, 2006), so knockout mutants for GPA1 are not expected to be complimented by GPA2 and knockout mutants for GPA2 are not expected to be complimented by GPA1 To further address the possibili ty that the yeast G-proteins are involved in Class II PAQRmediated repression of FET3-lacZ vectors for the overexpression of truncated PAQR5 and PAQR7 were made. The truncat ed proteins include the firs t 289 out of 330 amino acids for PAQR5 and the first 310 out of 346 amino acids for PAQR7 such that the pr oteins lack the final
46 transmembrane domain and C-terminal amino acids. These truncated PAQRs lack a domain that is homologous to a part of the spotted sea trout PAQR7 previously shown to be important for Gprotein activation (Thomas et al ., 2007). Our human PAQR trunca tions expressed similarly to the full-length proteins (Figure 2-9A) and progesterone signaling was unaffected (Figure 2-9B), suggesting that the regions of th e proteins purported to activate G-proteins in sea trout PAQR7 (Thomas et al., 2007) are not needed for sensing pr ogesterone and signaling, at least in yeast. In addition, by characterizing the Class II PAQRs in yeast in the absence of any other human proteins, such as Gi it was demonstrated that the Class II PAQRs are still capable of signaling. Together, these results suggest that the capability of the Cl ass II PAQRs to signal does not absolutely require coupling to human or yeast G-proteins. Finally, to further study the possible involve ment of G-protein activation in causing FET3lacZ repression, we performed FET3-lacZ assays in cells transformed with plasmids that constitutively express the mutants GPA1Q323L or GPA2Q300L. These mutants lack GTPase activity and are constitutively active forms of the yeast G-proteins Gpa1p (Dohlman et al., 1996) and Gpa2p (Harashima and Heitm an, 2002), respectively. While GPA1Q323L was previously shown to cause transcripti on of the pheromone-pathway activated FUS1 promoter even in the absence of pheromone (G uo et al., 2003), an effect of GPA1Q323L on FET3-lacZ in LIM containing galactose was not obse rved, nor was an effect of GPA2Q300L on FET3-lacZ (Figure 2-10). These results in dicate that the repression of FET3-lacZ is unrelated to activation of G-proteins and that the Class II PAQR signaling that leads to FET3-lacZ repression is unlikely to involve activation of G-proteins.
47 PAQR-Mediated Repression of FET3 Requires the Presen ce of the PKA Yeast Homologue Besides involving G-proteins, the signaling pathway of the Class II PAQRs has been proposed to involve a decrease in cAMP that l eads to a decrease in cAMP-dependent kinase (PKA) activation (Figure 2-7) (T homas et al., 2007; Dressing and Thomas, 2007). Interstingly, it was previously shown that the presence of the yeast Tpk2p isoform of human PKA was essential for Class I PAQR signaling in yeast (Kupcha k et al., 2007). In addition, Tpk1p, Tpk2p, and Tpk3p happen to be upstream of several yeast pr oteins that undego ch anges in phosphorylation in the presence of PAQR7 and progesterone (Reg alla, 2007), which also suggests that these proteins are important for the signaling pathwa y of the PAQRs. Thus, we performed the FET3lacZ assay in the knockout mutant for TPK2 As shown in Figure 2-11, TPK2 is also required for the Class II PAQR signaling th at leads to repression of FET3-lacZ This suggests that all of the PAQRs that we have studied share the same signaling pathway. Identification of Novel A gonists of Class II PAQRs Previously, prom oter-reporter assays in yeast ha ve been used as a convenient tool for the identification of novel agonists a nd antagonists of drug targets, such as human GPCRs (Pausch, 1997). Thus, we decided to test repression of FET3-lacZ in cells expressing the Class II PAQRs and in the presence of compounds other than progestins. The synthetic estrogen diethylstilbestrol (DES ) and the antiprogestin RU-486 were chosen (Figure 2-12A). Although DES is a potent estrog en, it was chosen becaus e it showed increased binding to membranes from cells expressing gold fish PAQR7 and it was previously shown to induce oocyte maturation in goldfish, which is a nonc lassical effect of proge stins, suggesting that DES can have agonistic effects on PAQR7 (Tokum oto et al., 2007). In addition, DES is an interesting endocrine-dis rupting compound (EDC) to study because it has been banned from use
48 due to its serious negative effects, such as carcinogenici ty and teratogenicity (Greenwald et al., 1971; Herbst et al., 1971). RU-486 was chosen because, while it is known to be a potent antagonist of the nPRs, it does not show increased binding to membranes fr om cells expressing human PAQR7 (Thomas et al., 2007). An effect of this compound was unexpected. As an initial screen, assays of the repression of FET3-lacZ were performed during reduced expression (0.05% galactose inducer) of PAQR5, PAQR6, PAQR7, and PAQR8 in the presence of 10 M progesterone, 10 M DES or 10 M RU-486. A concentration of 10 M was chosen because this was the highest conc entration of progesterone tested in the dose-response curves (Figure 2-5). Neither DES nor RU-486 caused repression of FET3-lacZ activity in the vector control cells, as compared to the untreated vector control cells (Figure 2-12). While a t-test analysis indicates that the increase in FET3-lacZ activity that is observed for the vector control cells treated with progesterone, DE S, and RU-486 was significant (P < 0.05) in this particular assay, vector control cells did not always have an increase in FET3-lacZ activity in response to progesterone (for example, see Figures 2-4, 2-6, and 2-8). The increase in FET3-lacZ activity was also not observed in the RU-486 dose response curves for the vector control (Figure 2-13). Interestingly, for the Class II PAQR samples, while neither DES nor RU-486 were as effective as progesterone for causing repression of FET3-lacZ both of these compounds did cause repression of FET3-lacZ (Figure 2-12B) for some of the PAQRs. A t-test analysis indicates that repression of FET3-lacZ in the presence of RU -486 was significant ( P < 0.05) for cells expressing PAQR5, PAQR7, and PAQR8, and that the repression of FET3-lacZ in the presence of DES was significant ( P < 0.05) only for cells expressing PAQR7. This is the first demonstration of an effect of DES on any of the human PAQRs.
49 This is also the first demonstration of an effect of RU-486 on the human PAQRs. So, although the Class II PAQRs seem to respond more to progestins (Figure 2-6B), some activation of these proteins in the presence of RU-486 was observed (Figure 2-12B). Because the effect of RU-486 was significant for more of the PAQRs than the effect of DES, and because an effect of RU-486 was unexpected, dose-response analysis of this compound was performed (Figure 2-13). These results were especially interesting becau se we did not expect to see an agonistic effect of RU-486 on the Class II PAQRs for several reasons: 1.) RU-486 is us ed clinically as an antiprogestin (Gass et al., 1998), 2.) the activity of RU-486 is well-known to directly interact with the nPRs to prevent efficient interaction of nPRs with the transcriptional machinery (Beck et al., 1993), and 3.) Thomas et al. showed th at RU-486 competed poorly with progesterone for binding to membranes from cells expressing PAQR7 (Thomas et al., 2007). Despite the primary effect of RU-486 as an an tagonist (Gass et al., 1998), it has been noted that in some cell models, RU-486 can act as an agonist to strengthen the effects of progesterone on inducing gene expression (Rodrig uez et al., 2002). This result has been suggested to indicate that these particular progeste rone effects are nonclassical in nature (Oettel and Mukhopadhyay, 2004). Also, under some conditions, such as stimul ation of cAMP signaling pathways, RU-486 can have a stimulatory effect on tr anscription (Beck et al ., 1993). This effect is thought to result from cross-talk between the signaling pathwa ys of second messengers and steroid receptors (Beck et al., 1993). This study monitored expressi on of an nPR target gene, and it was suggested that stimulation of cAMP signa ling pathways alter phosphorylati on of nPR or transcriptional machinery components (Beck et al., 1993). Such m odifications could cause increased efficiency
50 of nPR-RU-486 interactions with the transcriptional machinery a nd thus, increased transcription (Beck et al., 1993). No effects on nPR-target gene expression we re observed with either cAMP signaling pathway stimulation or RU-486 alone (Beck et al., 1993); however, it is possible that RU-486 could stimulate cAMP signaling pathways to cause other, nonclassical progesterone-like effects on cells. These effects could occur via the Class II PAQRs. Interestingly, the activity of the Class II PAQRs alters cAMP levels (Thomas et al., 2007), which then causes altered cAMP dependent kinase (PKA) activity. We have demonstrated that the repression of FET3-lacZ requires the presence of the yeast Tpk 2p isoform of human PKA (Figure 2-11). Materials and Methods Plasmids. The plasm ids used were pYES260 (Melcher, 2000) and pGREG536 (Jansen et al., 2005). All plasmids allowed for GAL1 promoter-driven expression of PAQR proteins. All pYES260 primer sequences were designed by Julie Russell to amplify the gene from commercially available cDNA and included 30-40 bases fo r homologous recombination into pYES260 (Melcher, 2000). PAQR5, PA QR6, PAQR7, and PAQR8 were sub-cloned from pYES260 constructs into pGREG536. Fo r cloning into pGREG536, primer s were designed to amplify PAQR5, PAQR6, PAQR7, and PAQR8 and includ ed sections of DNA for homologous recombination into pGREG536 (Jansen et al., 20 05) (Table 2-1). For cloning of truncated PAQR5 and PAQR7, the reverse primers were designe d to anneal to the ge nes at the location of the desired truncation. Proteins expressed in pYES260 have an N-terminal 6x-histidine tag and proteins expressed in pGREG536 have an Nterminal 7x-HA tag. The National Center for Biotechnology Information (NCBI) accession number s for the cDNA sequences of the Class II PAQRs are given in Table 2-2. In addition, th e frequencies of codons in the PAQR sequences
51 that are rarely used by S. cerevisiae that are also given in Table 2-2 were determined with the Gene Designer program (V illalobos et al., 2006). Empty pAD4M and pAD4M expressi on vectors carrying the yeast G -proteins GPA1, GPA1Q323L, and GPA2Q300L were obtained from Henrick Dohlma n, University of North Carolina at Chapel Hill. These constructs allow ADH1 promoter-driven expression of GPA1, GPA1Q323L, and GPA2Q300L. Expression of genes in this vector is constitutive. The GPA1Q323L, and GPA2Q300L mutants are constitutively active forms of GPA1 (Dohlman et al., 1996) and GPA2 (Harashima and Heitman, 2002), respectively. Yeast Strains and Assays. All yeast strains used in this study were obt ained from Euroscarf unless otherwise noted. tpk2 and gpa2 mutants were in the BY4742 (Mat ) background and gpa1 ste7 was obtained from Henrick Dohlman (University of No rth Carolina at Chapel Hill) and was in the BY4741 (Mat a) background. The following are the manufacturers for the carbon sources that were used: D-(+)-glucose (170080025, Acros Organics), D-(+)-galactose (150610010, Acros Organics), and D-(+)-raffinos e (19567-1000, Acros Organics). All growth conditions and galactosidase assays were performed as previo usly described (Kupchak et al., 2007). Briefly, cells were grown in Low Iron Me dium (LIM) containing 1 M Fe3+, conditions which normally induce the expression of FET3 (Eide and Guarente 1992). Galactose was used as a carbon source to indu ce full target protein expression (2%) or reduced expression (0.05% gala ctose/1.95% raffinose). Cells were harvested, washed and permeabilized prior to assay for -galactosidase activity ( lacZ ). -galactosidase assays were performed on permeabilized cells as describe d (Guarente, 1983). The substrate was onitrophenyl-galactopyraniside (ONPG, 4 mg /mL, 128820050, Acros Organics). -
52 galactosidase activity is presented as a percenta ge of activity seen in cells expressing empty expression vector treated with vehicle (ethanol). For individual experiments, each data point was done in triplicate and the error bars represent +/1 standard deviation. Experiments were generally performed at least three times and a representative experi ment is shown. All constructs were tested for the ability to repress FET3-lacZ in a progesterone-dependent manner. Experiments to ensure that diffe rent tags did not affect PAQR activity were performed at least twice or experiments performed as initial screens for responsiveness to the ligands diethylstilbestrol and RU-486 were performed once or twice with further validation of responses to RU-486 as shown in the Results and Disucssion. When needed, the following compounds were added to the LIM during overnight growth: progesterone (P8783, Sigma), testosterone (T-1500, Sigma), 21-hydroxyprogesterone (D6875, Sigma), 17-hydroxyprogesterone (H5752, Sigm a), 17,21-dihydroxyprogesterone (R0500, Sigma) adiponectin (RD172029100, Biovendor Laborat ory medicine, Inc.), diethylstilbestrol (D4628, Sigma), and RU-486 (M8046, Sigma). These compounds were all dissolved in ethanol. The untreated control cultures contained the same concentration of ethanol as the steroid-treated cultures. As noted in the captions of the appropriate figures, some -galactosidase assay data were collected in collaboration with Brian Kupchak and Ibon Garitaonandia. Some of these data or similar trends are also presented in K upchak (2008) and Ga ritaonandia (2008). Expression of Hemaglutanin (HA)-Tagged Class II PAQRs. For Class II PAQR localization studies, overnig ht cultures in SD me dia were reinoculated to an optical density (OD) at 600 nm (OD600) equal to approximately 0.2 in LIM containing 2% galactose for full target protein expression at 30oC. After the OD600 reached 1, the cultures were
53 harvested, washed twice with 25 mL of cold water, and stored at -80oC until total membrane protein preparations were made. Total Membrane Protein Extraction a nd Detection by Western Blot. After thawing, the cell pellets were resuspended in 250 L of m embrane isolation buffer [MIB, 0.6 M D-mannitol (M-4125), 20 mM HEPES, pH 7.4, 1 mM EDTA, 1 mM phenylmethanesulfonyl fluoride (P7626, Sigma), Co mplete, EDTA-free protease inhibitor tablet (11836170001, Roche Diagnostics), and pr otease inhibitor cocktail (P8215, Sigma)], similar to that described by Gitan and Eide (2000). Gl ass beads (G8772, Sigma) were added, and the suspension was vortexed at 4oC six times for 1 minute each, with 1 minute on ice in between. The samples were centrifuged at 3000 x g for 10 minutes, the supernatant was removed and centrifuged at 130,000 x g for 90 minutes. The supern atant was saved and gl ycerol was added to a final concentration of 15%. The pellets were resuspended in fresh mannitol buffer supplemented with 15% glycerol (250 L) and aliquots were stored at -80oC. Before SDSPAGE analysis, protein concentration determinat ions were performed with the bicinchoninic acid (BCA) Protein Assay Kit ( 23227, Pierce). To detect expression of HA-tagged proteins, equal amounts of protein (in g) were loaded onto a 10% poly acrylamide gel and electrophoresis was performed according to standard procedures (Sambrook and Russell, 2001). Dual Color Precision Plus ProteinTM Standards (161-0374, Bio-Rad) were used for molecular weight markers. Western blots were performed using nitrocellulose according to standard procedures (Sambrook and Russell, 2001). Transfer buffer was Tris-glycine, pH 8.3, 10% methanol, and 0.1% SDS. The transfer was performed at 80 volts for 60 minutes. Membranes were blocked overnight at 4oC with 1% bovine serum albumin (BSA ) in phosphate buffered saline (BP3994, Fisher Scientific). For HA-tagged PAQRs, the primary antibody (rabbit polyclonal IgG HA-
54 probe (Y-11), SC-805, Santa Cruz Biotechnology, Inc.) was diluted 1:500 in PBS containing 1% BSA. The secondary antibody used was horse radish peroxidase-con jugated goat anti-rabbit IgG-HRP (1:10000, SC-2004, Santa Cr uz Biotechnology, Inc.). The blot was incubated with SuperSignal West Pico Substrat e Working Solution (Pierce) for 5 minutes, exposed to film and developed. The membrane wa s stripped with RestoreTM Western Blot Stripping Buffer (21059, Thermo Scientific) according to th e manufacturers directions and re-probed with an anti-porin primary antibody (diluted to 0.5 g/mL, anti-porin yeast mitoc hondrial mouse IgG monoclonal, A6449, Invitrogen). The secondary antibody us ed was goat anti-mouse IgG (H+L)-HRP (1706516, Bio-Rad) (1:10000). Again, the blot was incubated with SuperSignal West Pico Substrate Working Solution (Pierce) for 5 minutes exposed to film and developed. The ImageJ software program (NIH) was used as described (http://rsb.info.nih.gov/ij/) to obtain values for the integrated densities of the HA-tagged P AQRs or the corresponding porin band for each sample. The density for each porin band was used to normalize the bands for the HA-tagged PAQRs so that comparisons of PAQR expression levels could be made. Summation Here, it was dem onstrated that, like the Cla ss I PAQRs (Kupchak et al., 2007), expression of Class II PAQRs can have a physiologica l effect on yeast (i.e. repression of FET3 in LIM) and this effect can be induced by the presence of pr ogesterone. By screeni ng different steroids, it was demonstrated that 17-hydroxyprogesterone, 21-hydroxyprogesterone, and 17,21dihydroxyprogesterone, but not testosterone, also induce repression of FET3 when Class II PAQRs are expressed. In addition, this heterologous system allowed the study of the role of the Class II PAQRs in sensing and responding to progesterone in the absence of other steroid receptors. Experimental evidence presented here suggests that th e Class II PAQRs do not
55 absolutely require human or yeast G-proteins to respond to progesterone. This system was also used to begin structure/function studies on human Class II PAQRs and to identify potential new agonists for these proteins.
56 Figure 2-1. Mechanisms of progestin steroids. Progestins exert their effects via either classical genomic mechanisms or rapid, nongenomic m echanisms. Classic genomic progestin effects are mediated by the soluble nuclear progestin receptors (nPR), which bind to the promoters of target genes to activat e gene transcripti on. Rapid, nonclassical effects of progestins are t hought to occur by membrane-bou nd receptors, but this is debated.
57 VectorPAQR5PAQR6PAQR7PAQR8PAQR9PAQR10PAQR11 Beta-galactosidase activity % of vector control 0 20 40 60 80 100 120 Figure 2-2. Overexpression of so me PAQRs cause repression of FET3-lacZ Yeast cells doubly transformed with the pFET3-lacZ vector and GAL1 -6x histidine-PAQR vector were grown in LIM containing 1 M Fe3+ and 2% galactose and -galactosidase assays were performed as described in the Ma terials and Methods. PAQR5, PAQR8, and PAQR11 cause constitutive repression of FET3-lacZ The data for this figure were obtained in collaboration with Br ian Kupchak and Ibon Garitaonandia. VectorPAQR6PAQR7PAQR9PAQR10 Beta-galactosidase activity % of untreated vector control 0 20 40 60 80 100 120 untreated progesterone (100 nM) Figure 2-3. Repression of FET3-lacZ is ligand-dependent for some PAQRs. Yeast cells doubly transformed with the pFET3-lacZ vector and GAL1 -6x histidine-PAQR vector were grown in LIM containing 1 M Fe3+ and 2% galactose and -galactosidase assays were performed as described in the Materi al and Methods. Cultu res were untreated, or had progesterone (100 nM) added during the overnight gr owth in LIM. The data for this figure was obtained in collab oration with Brian Kupchak and Ibon Garitaonandia.
58 A 0.0 0.5 1.0 1.5 2.0 Beta-galactosidase activity % of vector control 0 20 40 60 80 100 120 140 160 180 Galactose (%) B VectorPAQR5PAQR6PAQR7PAQR8PAQR9PAQR10PAQR11 Beta-galactosidase activity % of untreated vector control 0 20 40 60 80 100 120 140 Untreated Progesterone (10 M) Figure 2-4. Constitutive repression of FET3-lacZ can be alleviated and become liganddependent under low expression induction cond itions. In panel A, the repression of FET3-lacZ by PAQR5 increases with increa sing amounts of expression inducer (galactose). In panel B, yeast cells doubly transformed with the p FET3-lacZ vector and GAL1 -6x histidine-PAQR vector were grown in LIM containing 1 M Fe3+ and 0.05% galactose and -galactosidase assays were pe rformed as described in the Material and Methods. Cultures were untreated, or had progesterone (10 M) added during the overnight growth in LIM. Th e data for this figure was obtained in collaboration with Brian Kupchak and Ibon Garitaonandia.
59 A 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 PAQR5PAQR6PAQR7PAQR8Ratio of integrated density (HA band/porin band) B vectorPAQR5PAQR6PAQR7PAQR8 Beta-galactosidase activity % of untreated control 0 20 40 60 80 100 120 140 160 Untreated Progesterone (100 nM) Figure 2-5. Expression of PAQRs in yeast can be detected by We stern blot. In panel A, the PAQRs were re-cloned in vector such th at they had N-terminal 7x-HA tags. The arrow indicates the approximate position of the 50 kDa band of the molecular weight marker. After expression of PAQRs wa s detected with an anti-HA antibody (top left), the membrane was stripped and re-p robed with an anti-p orin antibody (bottom left) for normalization (chart right). In panel B, FET3-lacZ repression by the HAtagged PAQRs was confirmed in LIM containing 0.05% galactose and progesterone.
60 A [progesterone], nM 0110100100010000 Beta-galactosidase activity % of untreated vector control 0 20 40 60 80 100 120 vector control PAQR5 PAQR6 PAQR7 PAQR8 PAQR9 B C [steroid], nM 0110100100010000 Beta-galactosidase activity % of untreated vector control 0 20 40 60 80 100 120 140 PG TS 17-OH-PG 21-OH-PG 17,21-diOH-PG D [steroid], nM 0110100100010000 Beta-galactosidase activity % of untreated vector control 0 20 40 60 80 100 120 PG TS 17-OH-PG 21-OH-PG 17,21-diOH-PG E [steroid], nM 0110100100010000 Beta-galactosidase activity % of untreated vector control 0 20 40 60 80 100 120 PG TS 17-OH-PG 21-OH-PG 17,21-diOH-PG
61 F [steroid], nM 0110100100010000 Beta-galactosidase activity % of untreated vector control 0 20 40 60 80 100 120 PG TS 17-OH-PG 21-OH-PG 17,21-diOH-PG G [progesterone], nM 0 1 10 100100010000 Beta-galactosidase activity % of untreated vector control 0 20 40 60 80 100 120 PAQR5 PAQR5 +TS (10 M) Figure 2-6. Ligand-dependent repression of FET3-lacZ by Class II PAQRs occurs in a dosedependent manner and is steroid specific. In panel A, all Class II PAQRs were used in -galactosidase assays and different concen trations of progesterone were included in the growth media. In panel B, differe nt steroids were test ed: progesterone (PG), testosterone (TS), 17 -hydroxyprogesterone (17-OH-PG ), 21-hydroxyprogesterone (21-OH-PG), or 17,21-dihydroxyprogesterone (17,21-diOH-PG) were tested with the vector control (panel C), P AQR5 (panel D), PAQR6 (panel E), and PAQR9 (panel F). In panel G, the progesterone dose-response for PAQR5 shifts to the right in the presence of testosterone (the experi ment in panel G was performed by Brian Kupchak).
62 Figure 2-7.Class II PAQR signaling model. A porti on of the model proposed in the literature (Thomas et al., 2007; Dressing and Thomas, 2007) is presented here. This model shows that the Class II PAQRs are localized to the plasma membrane, are activated by progesterone, and are coupled to inhib itory G-proteins. Upon activation of Giproteins, GDP bound to the Gi-proteins is converted to GTP, which causes inactivation of adenylate cyclases and a subsequent decrease in cAMP, leading to decreased PKA activity. [This figure was m odified from Dressing GE and Thomas P (2007) Identification of membrane progestin receptors in human breast cancer cell lines and biopsies and their potent ial involvement in breast cancer. Steroids 72:111116 with permission from Elsevier (Figure 3, page 114).]
63 A B Figure 2-8. Repression of FET3-lacZ does not require yeast G -proteins. In panel A, the repression of FET3-lacZ by PAQR5 still occurs in a gpa1 mutant ( gpa1) In panel B, the repression of FET3-lacZ by PAQR5 still occurs in a gpa2 mutant ( gpa2) For both panels, the black bars are for untre ated cultures, and the gray bars are for cultures treated with progesterone (100 nM) in LIM containing 0.05% galactose.
64 A B [progesterone], nM 0 110100100010000 Beta-galactosidase activity % of untreated vector control 0 20 40 60 80 100 120 Vector PAQR5 truncated PAQR5 PAQR7 truncated PAQR7 Figure 2-9. Expression of trun cated human Class II PAQRs cause ligand-dependent repression of FET3-lacZ In panel A, Western blots were performed on yeast total membrane preparations on cells expressing full-lengt h PAQR5 (lane 1), truncated PAQR5 (lane 2), full-length PAQR7 (lane 3), and truncated PAQR7 (lane 4). In panel B, yeast cells doubly transformed with the p FET3-lacZ vector and GAL1 -7xHA-PAQR vector were grown in LIM containing 1 M Fe3+ and 0.05% galactose. -galactosidase assays were performed as described in th e Materials and Methods Without the last hydrophobic domain, the truncated PAQRs still cause ligand-dependent repression of FET3-lacZ vectorGpa1Gpa1Q323LGpa2Q300L Beta-galactosidase activity % of vector control 0 20 40 60 80 100 120 140 160 Figure 2-10. Yeast cells that expr ess constitutively active yeast G proteins do not demonstrate repression of FET3-lacZ Wild-type Gpa1p, or constitutively active forms of the yeast G-proteins Gpa1p (Gpa1Q323L) or Gpa2p (Gpa2Q300L) were constitutively expressed by the ADH1 promoter. Yeast cells doubly transformed with the p FET3lacZ vector and ADH1Gpa vector were grown in LIM containing 1 M Fe3+ and 2% galactose and -galactosidase assays were performed as described in the Material and Methods.
65 Figure 2-11. Repression of FET3-lacZ requires yeast Tpk2p, a subun it for the yeast homologue of human PKA. The repression of FET3-lacZ by PAQR5 does not occur in a tpk2 strain. The black bars are for untreated cultures, and the gray bars are for cultures treated with progesterone (100 nM) in LIM containing 2% galactose. A B Figure 2-12. Synthetic compounds tested for acti vation of Class II PAQRs. In panel A, the structures of RU-486 and diet hylstilbestrol are shown. In panel B, the effect of 10 M RU-486 or diethylstilbestrol on the Cl ass II PAQR-mediated repression of FET3lacZ is demonstrated. Yeast cells doubly transformed with the p FET3-lacZ vector and GAL1 -7xHA-PAQR vector were gr own in LIM containing 1 M Fe3+ and 0.05% galactose. -galactosidase assays were performe d as described in the Materials and Methods. *, P < 0.05. vectorPAQR5PAQR6PAQR7PAQR8 Beta-galactosidase activity % of untreated control 0 20 40 60 80 100 120 140 160 untreated progesterone DES RU-486 * * * *
66 [RU-486], uM 0.0 0.1 1.010.050.0100.0 Beta-galactosidase activity % of untreated control 0 20 40 60 80 100 vector PAQR5 PAQR7 Figure 2-13. Dose response of PAQR7 to RU-486. The effect of varying concentrations of RU486 on PAQR7-mediated repression of FET3-lacZ is demonstrated. Yeast cells doubly transformed with the p FET3-lacZ vector and GAL1 -7xHA-PAQR vector were grown in LIM containing 1 M Fe3+ and 0.05% galactose. -galactosidase assays were performed as described in the Ma terials and Methods. These data were collected in collaboration with Brian Kupchak.
67 Table 2-1. Primers used for cloning P AQRs into pGREG536 (listed from 5 3) Primer name Primer sequence Fwd PAQR5 GAATTCGATATCAAGCTTATCGATACCGTCGACAATGCTGAGCCTGAAGCT Rev PAQR5 GCGTGACATAACTAATTACATG ACTCGAGGTCGACTCATGTTTCTTTTTTATGTAAT TCTGG Rev Trunc5 GCGTGACATAACTAATTACATGACTCGAGGTCGACTCAGGGCTTGGAGGTGGCCAG Fwd PAQR6 GAATTCGATATCAAGCTTATCGA TACCGTCGACAATGCTCAGTCTCAAGCTGC Rev PAQR6 GCGTGACATAACTAATTACATGACTCGAGGTCGACTTACTGTTGTTTGGCCTGGG Fwd PAQR7 GAATTCGATATCAAGCTTATCGATACCGTCGACAATGGCCATGGCCCAGAAA Rev PAQR7 GCGTGACATAACTAATTACATG ACTCGAGGTCGACTCACTTGGTCTTCTGATCAA Rev Trunc7 GCGTGACATAACTAATTACATG ACTCGAGGTCGACTCAGTGC GTGTGCAGAGGCTC Fwd PAQR8 GAATTCGATATCAAGCTTATCGATACCGTCGACAATGACGACCGCCATCTTGG Rev PAQR8 GCGTGACATAACTAATTACATGAC TCGAGGTCGACTCAGGAATCTTTCTTGGTCAG Table 2-2. Codons in PAQR sequen ces that are rarely used in S. cerevisiae. Rare codon analysis was conducted with the Gene Designer software (Villalobos et al., 2006). Gene name NCBI Accession Number of codons used with <5% frequency Number of codons used with <10% frequency PAQR5 Q9NXK6 2 17 PAQR6 Q6TCH4 4 30 PAQR7 Q86WK9 4 24 PAQR8 Q8TEZ7 4 25 PAQR9 Q6ZVX9 5 25
68 CHAPTER 3 CLASS II PAQRS: LOCALIZATION AND TOPOLOGY Introduction The Class II PAQRs consist of five prot eins, PAQRs 5-9. PAQR5, PAQR7, and PAQR8 are debated as being G-p rotein coupled receptors for progesterone that localize to the plasma membrane (Thomas et al. 2007, Krietsch et al. 2006) while PAQR6 and PAQR9 have yet to be characterized. In addition to their localization, the topology of the Class II PAQRs is also uncertain. Class II PAQR Localization? Studies which dem onstrate that PAQR5, PA QR7, and PAQR8 are involved in mediating nonclassical effects of progesterone have also sh own that these proteins are plasma membrane localized (Thomas et al., 2007). However, another group, which found no evidence to support that the Class II PAQRs mediate nonclassical e ffects of progesterone, showed that PAQR5, PAQR7, and PAQR8 localize to the endoplasmic reti culum (Krietsch et al., 2006). Other studies have shown that PAQR7 localizes to an intracellular tubuloretic ular network (F ernandes et al., 2005) and that PAQR8 is associated with lyso somes (Suzuki et al., 2001). So the cellular location of the Class II PAQRs is unclear. Class II PAQR Topology? Other points of contention surrounding the Class II PAQRs are their topology and whether they act as G-protein coupled receptors for progesterone. Besides evidence from second messenger and G-protein inhibitor assays, immunofluorescence studies of the human Class II PAQRs in mammalian cells (Thomas et al., 2007 ) suggests that these proteins have an orientation in the membrane that is similar to that of GPCRs, with an extracellular N-terminus and an intracellular C-terminus (Thomas et al., 2007).
69 In contrast to the Class II PAQRs, immunofluorescence stud ies showed that the human Class I PAQRs expressed in mammalian cell systems have the N-terminus located inside the cell and the C-terminus located outside the cell (Yam auchi et al., 2003). A global topology study of yeast membrane proteins (Kim et al., 2006) sugg ested that the yeast Class I PAQRs have the same topology as the human Class I PAQRs. Fi nally, while a topology st udy of the human Class III PAQRs has not been conducted, the Class III homologue from Eschericia coli (called YqfA) was characterized in E. coli and was found to have a Class I PAQR-like topology (Daley et al., 2005). If the topological determinations reported for the PAQR family members thus far are correct, then the three classes of PAQRs do not shar e similar structural orie ntations. Instead, the Class II PAQRs may have a topology that is similar to that of the GPCR supe rfamily of proteins. Because the localization and topology of the Class II PAQRs is unc lear, we were interested in studying each of these characterist ics in the yeast model organism as well and our approaches are described in this chapter. Determining Protein Localization and Topology in Yeast There are nu merous approaches that can be used to localize proteins in cells. The techniques that are often used with yeast and that have been used in the work for this dissertation are described below. Green Fluorescent Protein One approach for localization of proteins in y east is to use green fluorescent protein (GFP) to tag the proteins of interest (Huh, et al., 2003). The GFP prot ein, a 238 am ino acid protein, has a -can fold, with a -sheet surrounding a -helix (Yang et al., 1996). The fluorophore in the wild-type GFP is formed by spontaneous cycliz ation of Ser65 and Gly67 coupled with reduction of the C-C bond of Tyr66 to form a conjugated system (Yang et al., 1996).
70 Many mutants of GFP have been engineered for various applications (Shaner et al., 2005). One of these mutants is the S65T mutant, which fo lds four times more rapidly (2 hours) than the wild-type GFP (8 hours) and has increased fluor escence and photostability (Heim et al., 1995). This mutant has been used successfully when fused to other proteins for protein localization in subcellular locations such as the yeast plasma membrane and endoplasmic reticulum (Huh et al., 2003; Jansen et al., 2005). Both of these locat ions have been proposed for the Class II PAQRs (Thomas et al., 2007; Krietsch et al., 2006; Fernandes et al., 2005). Immunofluorescence Alternatively, immunofl uorescence has been successf ully used to localize proteins in yeast (Pringle et al., 1991) as well as determine their topology in the plasma me mbrane (Severance et al., 2004). For either of these methods, spheropl asts, which are yeast cel ls without their cell walls, must be made to allow access of the antib ody to the cell (Baggett et al., 2003). For localization, conditions which pe rmeabilize the cells are used so that the primary antibody can access the epitope, whether it is located inside or outside the cell (Ba ggett et al., 2003). A fluorophore-labeled secondary antibody is then used for detection of the primary antibody (Baggett et al., 2003). For topology studies, conditions which pe rmeabilize or keep the cells unpermeabilized are used (Severance et al., 2004). So if the epitope is locat ed inside the cell, it will only be detected under cell permeabilization conditions; however if the epitope is located outside the cell, it can be detected whet her the cells are permeabilized or not. The 3xHA-Suc2-His4C Dual Topology Reporter (DTR) Finally, an alternative method that has been widely used for topology studies in yeast is the 125 kDa 3xHA-Suc2-His4C dual-topology reporter (DTR) (Figure 3-1) (Sengstag, 2000). The DTR is useful for proteins which are located in the m embranes of some organelles, such as the ER, or the plasma membrane (Kim et al., 2003) When this method was used in a global
71 topology study of yeast membrane proteins, the yeast Class I PAQRs (the IZHs) were found to have the C-terminus located outside the cytopl asm (Kim et al., 2006). The DTR has also been successfully used to determine the topology of a yeast ubiquitin ligase and its human homologue (Kreft et al., 2006). For topology studies, the DTR is fused to th e C-terminus of membrane proteins to determine whether this part of the protein is cytoplasmic or not (Sengstag, 2000). First, the HIS4C portion of the tag encodes the full-length y east histidinol dehydrogenase, which can only convert the substrate histidinol to histidine if His4Cp is present in the cytoplasm. If His4Cp is located in the ER, histidinol can not be converted to histidine because polar, charged molecules such as histidinol do not pene trate the ER membrane (Deshaies and Schekman, 1987; Kim et al., 2003). The SUC2 portion encodes a part of the enzyme i nvertase that has several acceptor sites for N -linked glycosylation, as does the His4Cp porti on. These sites do not become glycosylated unless they have been exposed to the lumen of the ER during protein folding (Kim et al., 2003). The topology of the protein determined during fo lding and insertion in the ER membrane is maintained during transport to other cell membrane s (Alberts et al., 2002). Thus, if the protein under study has a topology such that the Suc2p-Hi s4Cp tag is located in the cytoplasm during folding in the ER membrane, the tag does not become glycosylated. This topology will be maintained such that the tag remains in the cytopl asm after the protein is sorted to its appropriate subcellular location. To determine the topology of a DTR-tagged protein, two assays are performed. First, growth assays on histidinol-conta ining plates lacking histidine ar e performed in a yeast strain that is his4(STY50) (Sengstag, 2000). The pl ates are typically grown at 30o C from 3-5 days and the presence of growth is determined (Sengstag, 2000). If the yeast cel ls grow on histidinol,
72 it is concluded that the C-terminus is located in the cytoplasm; how ever, errors in topology determinations have been made when this was the only assay performed and intermediate growth levels were observed (Sengstag, 2000). Therefor e, a second assay is performed in which the glycosylation status of the expressed protein is determined with endoglycosidase H (Endo H) digestion (Sengstag, 2000; Kim et al., 2003). Western blots are performed with an anti-HA antibody to detect whether there is a mobility sh ift of the 3xHA-Suc2-His4C tagged protein after Endo H treatment (Sengstag, 2000; Kim et al., 2003). If the protein has evidence of glycosylation, it is concluded that the C-terminus has translocated to the ER lumen (Kim et al. 2003). Usually, the localization assignment of the C-terminus is made only if the results of the two assays are in agreement (i.e. if there is growth on histidinol, then there should be no glycosylation of the protein and the C-terminus is assigne d to a cytoplasmic location. Alternatively, if there is no gr owth on histidinol, then the pr otein should be glycosylated, indicating that the C-terminus of the protein was located in the ER lumen during secretion and is considered to be extracytoplasmic) (Kim et al., 2003; Kim et al., 2006). This chapter describes attempts to localize the Class II PAQRs with GFP fluorescence and immunofluorescence. It also desc ribes attempts to determine the topology of the proteins by immunofluorescence and th e DTR 3xHA-Suc2-His4C. Results and Discussion Previously, experim ental evidence has been published that shows plasma membrane localization for the Class II PAQRs and that thes e proteins function as receptors for progestin steroids (Thomas et al., 2007). However, it has also been pub lished that the Class II PAQRs are localized to the endoplasmic retic ulum and are not progestin recep tors (Krietsch et al., 2006). Many attempts have been made to localize the Class II PAQRs in mammalian cells, with varying results (Thomas et al., 2007; Krietsch et al., 2006; Fernandes et al., 2005; Suzuki et al., 2001). In
73 addition, although data suggest that the Class II PAQRs have a topology similar to that of GPCRs (Thomas et al., 2007), other members of the PAQR family have the opposite topology (Yamauchi et al., 2003; Kim et al. 2006, Daley et al., 2005). So the cellular location and topology of the Class II PAQRs is unclear. Because our yeast system has demonstrated functional expression of the PAQR proteins, we wanted to determine the localization and topology of the Class II PAQRs in th is system. The results of the attempts to do so are described below. Localization of Class II PAQRs with GFP GFP has been used successfully to determ ine th e locations of many yeast proteins (Huh et al., 2003). To try to determine where the Class II PAQRs were expressed in the cell, PAQR5 and PAQR7 were tagged at the N-terminus with the S65T mutant vari ant of GFP using the pGREG575 vector (Jansen et al., 2005). The constructs were tested in FET3-lacZ assays and were able to cause repression of FET3-lacZ in low iron media (Figure 3-2). Despite this, neither of the GFP constructs showed fluorescence as observed with an epifluorescence microscope, while expression of soluble GFP alone showed much fluorescence localized to the cytoplasm (Figure 3-3). Some of the Cl ass I PAQRs cloned into the same vector (pGREG575) fluoresced and seemed to localize around the periphery of the cell (Garitaonandia, 2008). Western blots were performed using total membrane preparati ons from yeast cells expressing GFP-PAQR5 and GFP-PAQR7 and showed that neither of the c onstructs could be dete cted by Western blot (Figure 3-4). C-terminal GFP-tagged constructs for P AQR5, PAQR6, PAQR7, and PAQR8 were made and the cells expressing these constructs were observed with a fluorescence microscope. FET3lacZ assays were performed and the C-terminal ly GFP-tagged Class II PAQRs were able to respond to progesterone (Figure 35); however, fluorescence could not be observed for any of the
74 constructs (data not shown). Western blots were performed using total membrane preparations from yeast cells expressing the C-terminally GFP-tagged PAQRs and bands for PAQR5-GFP and PAQR6-GFP were observed (Figure 3-6), wh ile bands for PAQR7-GFP (Figure 3-6) and PAQR8-GFP were not detected (F igure 3-4). It should be note d that the bands for PAQR5-GFP and PAQR6-GFP were located at around 50 kDa, approximately 15 kDa lower than expected. Whether this disparity is due to a lack of a fu ll-length GFP tag is unknown, but if it is, this could be a reason for a lack of signal during fluorescen ce microscopy. An alternative explanation for a lack of fluorescence is that the GFP tag could be misfolded but not degraded. According to Wooding and Pelham (1998), GFP may not fold properly in the yeast ER lumen. The cause for the lack of detection of e ither the N-terminally GFP-tagged PAQR5 and PAQR7 or the C-terminally GFP-tagged PAQR7 and PAQR8 by fluorescence microscopy or Western blot with chemiluminescence is unknown and there are several possible reasons. For instance, these proteins may not be expressing in sufficient quantities for observation. The addition of a GFP tag to a protei n could cause the protein to misf old and lead to degradation of both the protein and the GFP tag (Pedelacq et al., 2006). Also, it is possible that the GFP tag is cleaved for some of the constructs, but not all. Regardless, all of the c onstructs were able to express in sufficient quantity to allow for progesterone-dependent repression of FET3-lacZ This may be because PAQR-mediated repression of FET3-lacZ does not require much expression of the PAQR (see Figure 2-4A). Protein Localization of Class II PAQRs by Immunofluorescence Because exp ression of PAQR5-GFP and PAQR 6-GFP was detectable by Western blot, localization studies using imm unofluorescence were continued w ith these constructs. In addition, HA-PAQR5 and HA-PAQR6 were also used for localization studies using immunofluorescence.
75 Although numerous attempts were made, only a few spheroplasts appeared in the brightfield light images that were also visible in the fluorescence images. Examples of images for these few spheroplasts are presented here. Fo r negative control cells stained with the antiGFP primary antibody (Figure 3-7A) or the an ti-HA primary antibody (Figure 3-7E), there appears to be minimal fluorescence, indicating that the conditions used were sufficient to reduce nonspecific binding of primary and secondary an tibodies. The fluorescence that was observed for the PAQR5-GFP, PAQR6-GFP, HA-PAQR 5 and HA-PAQR6 appears to outline the perimeter of the cell (Figure 3-7 C, D, F, G), suggesting a possible plasma membrane localization for these proteins. For soluble G FP, the fluorescence appears to be distributed throughout the whole cell (Figure 3-7B). Topology of Class II PAQRs by Immunofluorescence and the Dual Topology Reporter To determ ine the topology of the Class II PAQRs, immunofluorescence was performed with and without permeabilization of yeast sphe roplasts for the N-term inally HA-tagged PAQR5 and PAQR6 and the C-terminally GFP-tagged P AQR5 and PAQR6. In similar topology studies of the PAQRs expressed in mammalian cells, immunofluorescence studies showed that the human Class I PAQRs expressed in mammalian cell systems have the N-terminus located inside the cell and the C-terminus loca ted outside the cell (Yamauchi et al., 2003). In contrast, immunofluorescence studies in human cells transf ected with human PAQR7 expression plasmids demonstrated that the N-terminus could be clearly detected without permeabilization of the cells, while detection of the C-terminus required ce ll permeabilization (Thomas et al., 2007). These results suggest that the human Class II PAQRs have an orientation in the membrane that is the opposite to that of the Class I PAQRs, with an extracellular N-terminus and an intracellular Cterminus. Therefore, an immunofluorescence approach to topology determination was used for the Class II PAQRs expressed in yeast.
76 Unfortunately, the negative control for the unpermeabilized conditions (cytoplasmic GFP) was detected during every attempt (Figure 3-8). This problem o ccurs because many of the steps involved in yeast spheroplast generation and manipulation cause unwanted permeabiliztion of the cells (Dr. Colin Macdiarmid, University of Missouri, personal communication). So, although similar experiments were conducted for PAQR5 and PAQR6 (data not sh own), no conclusions could be made regarding the loca tion of their Nor C-termini. An alternative method for topology determinatio n in yeast was used. To determine PAQR C-terminal locations, PAQR6 and PAQR7 were tagged with 3xHA-Suc2-His4C (Figure 3-1). These fusions were tested in FET3 lacZ assays to ensure that the proteins retained their ability to repress the FET3 promoter (Figure 3-9). When histidi nol growth assays were conducted with Cterminally 3xHA-Suc2-His4C tagged human PAQRs, growth at 30oC after 3 days was not apparent. However, when plates were incubated longer (5-8 days), STY50 cells expressing tagged PAQRs had intermediate levels of growth (Figure 3-10) when compared to STY50 cells that were expressing the positive control Ost4 -3xHA-Suc2-His4C or the negative controls 7xHA-PAQR6 and 7xHA-PAQR7. (The negative controls show slight growth and this is likely due to the presence of residual histidine from th e media used prior to the histidinol growth experiments). Although this result suggests that the C-terminus of the PAQRs is located in the cytoplasm, intermediate growth phenotypes on hi stidinol are not conclusive (Sengstag, 2000). Therefore, the glycosylation status of the Suc2 portion of the protein was determined (Figure 311). A shift in PAQR-3xHA-Suc2-His4C migr ation upon Endo H treatment was not observed. Prior to Endo H treatment, immunoprecipita tion was performed, and the immunoprecipitated proteins were released from the agarose beads wi th SDS sample buffer containing 2% SDS. To make sure that the Endo H was still active in th is buffer, a reaction was performed using RNase
77 B as a positive control. A shift in the migration of the undigested verses the digested RNase B demonstrated that the conditions used for the Endo H reaction (SDS sample buffer with 2% SDS) did not inhibit the activity of Endo H (Figure 3-12). The positive results for the histidinol growth assay and the lack of glycosylation indicate that the C-termini of PAQR6 and PAQR7 are loca ted in the cytoplasm of the cell. However, because histidinol growth assays are normally carried out for 3-5 days, and only intermediate levels of growth were observed for the DTR-tagged PAQR6 and PAQR7 after 7-8 days, it is possible that the assignment of the C-termin i of PAQR6 and PAQR7 to the cytoplasm is incorrect. Because they express with higher yi elds in yeast (Figure 2-5A), DTR-tagged PAQR5 and PAQR8, might have stronge r growth phenotypes on histidi nol, which would be a better indicator of a cytoplasmi c location for the C-termini of the Class II PAQRs. Furthermore, as was previously done with another human membrane protein (Kreft et al., 2006), a more detailed analysis of topology using the DTR tag in comb ination with various PAQR truncations could also lead to better conclusions about the topology of these proteins. If the Class II PAQRs have only seven tran smembrane domains, and the C-termini are located inside the cell, then th e Class II PAQRs have a topology that is similar to GPCRs, but highly divergent from the other PAQRs, as has b een previously reported (Thomas et al., 2007). A few examples of closely related membrane proteins that have opposite topologies have been reported (Sf et al., 1999; Rapp et al., 2006). Topology is larg ely determined by the number of lysines and arginines in the loop regions of membrane proteins, with a bias towards having more positively charged residues (K+R bias) in the cytoplasm (von Heijne, 1989). In fact, single-charge point mutations can alter the topology of membra ne proteins that have a low K+R bias, demonstrating that even highly relate d proteins can have oppos ite topologies (Rapp et
78 al., 2006). So, even though the PAQRs are rela ted to each other, there are quite a few differences in their sequences, and it is possi ble that the different classes of PAQRs have different topologies. Materials and Methods Plasmids. Hum an PAQR5, PAQR6, PAQR7, a nd PAQR8 were sub-cloned from expression plasmids previously constructed by other lab members (Lisa Regalla and Julie Russell). Primer sequences were designed to amplify the gene and incl uded bases for homologous recombination into pGREG575 for N-terminal GFP-tagging of PA QRs (Jansen et al., 2005), pGREG600 for Cterminal GFP-tagging of PAQRs (Jansen et al ., 2005), or pJK90 for C-terminal 3xHA-Suc2His4C tagging of PAQRs (Kim et al., 2003) (Table 3-1). All plasmids except for pJK90 allowed for GAL1 promoter-driven expression of PAQR pr oteins. The pJK90 plasmid allowed for TPI promoter-driven expression of proteins (Kim et al., 2003). Cloning was perfor med by gap repair. Yeast Strains and Assays. All yeast strains used in this study were obt ained from Euroscarf unless otherwise noted. All growth conditions and -galactosidase assays were perf ormed as previously described (Kupchak et al., 2007). Briefly, cells were grown in Low Iron Medium (LIM) (Eide and Guarente 1992) supplemented with 1 M Fe3+ to induce the expression of FET3 Galactose was used as a carbon source to indu ce full target protein expression (2%) or reduced expression (0.05% galactose/1.95% raffinose). Cells were harvested, washed and permeabilized prior to assay for -galactosidase activity ( lacZ ). -galactosidase assays were performed on permeabilized cells as described (Guarente, 1983). -galactosidase activity is presented as a percentage of activity seen in cells expressing empty e xpression vector treated
79 with vehicle (ethanol). For individual experiment s, each data point was done in triplicate and the error bars represent +/1 standa rd deviation. Experiments were performed at least three times and a representative experiment is shown. All constructs, including the GFP-tagged PAQRs, were tested for the ability to repress FET3-lacZ in a progesterone-dependent manner. Expression of GFP-tagged Class II PAQRs. For Class II PAQR localization studies, overnig h t cultures in SD me dia were reinoculated to an OD600 equal to approximately 0.2 into LIM cont aining 2% galactose for full target protein expression at 30oC. After the OD600 reached 1, the cultures (150 mL) were harvested for total membrane protein preparations, washed twice with 25 mL of cold water, stored at -80oC before total membrane extraction (see below). Alternativ ely, the cultures (5 mL) were used directly in fluorescence microscopy experiments (5 mL). Total Membrane Protein Extraction a nd Detection by Western Blot. Total m embrane protein samples were prepar ed as described in Chapter 2. For GFPtagged PAQRs, the primary antibody (anti-G FP rabbit IgG fraction, A11122, Invitrogen) was diluted 1:1000 in PBS containing 1% BSA. Th e secondary antibody (goat anti-rabbit IgG-HRP (SC-2004, Santa Cruz Biotechnology, Inc.) was dilu ted 1:10000. The blot was incubated with SuperSignal West Pico Substrat e Working Solution (Pierce) for 5 minutes, exposed to film and developed. The membrane was stripped with RestoreTM Western Blot Stripping Buffer (21059, Thermo Scientific) according to th e manufacturers directions and re-probed with an anti-yeast mitochondrial porin primary antibody (diluted to 0.5 g/mL, anti-porin yeast mitochondrial mouse IgG1, monoclonal antibody, A6449, Invitrogen). The secondary antibody was diluted 1:10000 (goat anti-mouse IgG (H+L)-HRP,170-6516, Bi o-Rad). Again, the blot was incubated with SuperSignal West Pico Substrate Worki ng Solution (Pierce) for 5 minutes, exposed to film and developed.
80 Immunofluorescence of S. cerevisiae Expressing the Class II PAQRs. To determ ine the localization of overexpresse d Class II PAQRs by immunofluorescence, a procedure obtained from Dr. Colin Macdiarmid (University of Miss ouri) was modified. Overnight cultures in SD media were reinoculated to an OD600 equal to approximately 0.2 in 25 mL of LIM containing 2% ga lactose and incubated at 30oC. After the OD600 was 0.8-1, the cells were harvested at 1000 x g and washed twice with PBS. The cells were resuspended in 9 mL of PBS and 1 mL of 37% formaldehyde (BP531, Fish er Scientific) was added. The cells were incubated at 30oC for 2 hours with agitation (230 RPM). The cells were washed twice with 10 mL of cold PBS and resuspended in 2 mL of cold PBS containing 10 mM dithiothreitol. The OD600 was determined and a solution of Zymolyas e 20T was added at a ratio of 2 units per OD600 to remove the cell wall. The cells were incubated for 2 hours at 30oC with gentle agitation (100 RPM). The spheroplasts were collected by centrif ugation (200 x g), washed twice with 10 mL of PBS, and resuspended in either 5 mL of cold methanol for permeabilization or 5 mL of PBS for unpermeabilized spheroplasts. The permeabilized spheroplasts were stored at -20oC for 1 hour, while the unpermeabilized spheroplasts were stored at 4oC for 1 hour. The spheroplasts were washed twice with 5 mL of PBS and resuspended in 100 L of PBS. An aliquot (50 L) of spheroplasts was applied to the well of a slide pre-treated with poly-L-lysine. The slides were incubated for 1 hour at 4oC to allow cells to adhere and the slides were washed twice with PBS to remove unbound cells. The slides were blocked for 1 hour at 4oC in PBS containing BSA (1%). For permeabilized spheroplasts, the blocking solution also contained 0.1% Tween-20. The primary antibody was diluted (1:250, anti-G FP rabbit IgG fraction, A11122, Invitrogen or 1:50, rabbit polyclonal IgG HA-pr obe (Y-11), SC-805, Santa Cruz Biotechnology Inc.) in the appropriate blocking buffer and incu bated with the spheroplasts for 1 hour at room temperature.
81 The slides were washed 5 times for 5 minutes each with PBS either without or with 0.5% Tween20 for unpermeabilized or permeabilized spheroplasts respectively. The procedure was repeated with the fluorescently labele d secondary antibody (Alexa Fl uor 555 goat anti-rabbit IgG, A31629) diluted 1:200 in the approp riate blocking buffer. After the last wash step, 1 drop of Mowiol solution was added to the slide and a co verslip was placed over the well. Slides were allowed to harden overnight at 4oC before viewing. Fluorescence Microscopy Experiments. For Class II PAQR localization studies, overnig h t cultures in SD me dia were reinoculated into LIM containing 2% galactose for full target protein expression at 30oC. For GFP fluorescence, after 18-24 hours of growth, a small al iquot of cells was placed on a slide coated in poly-L-lysine (7799,Lab Scientific, Inc.) and cove rslip was placed on top of the cells. A Zeiss Axiovert s100 microscope with a 63X objective was used (courtesy of Dr. R.J. Cousins, University of Florida). The same microscope was used for visualization of immunostained spheroplasts or an Olympus FV500-IX81 confocal microscope with a 100X objective was used (courtesy of Dr. W. Tan, University of Florida). Histidinol Growth Assays. Experim ents were conducted essentially as described (Kim et al., 2003). STY50 transformants containing each of the 3xHA-Suc2-H is4C-tagged proteins were streaked onto SD containing histidine (0.6 mM) or histidinol (6 mM) plates. Alte rnatively, cell pellets from liquid cultures in SD containing his tidine (0.6 mM) were washed with sterile water. The optical density at 600 nm (OD600) was adjusted to 0.5, 0.1, and 0.01. Aliquots of 5 uL of each dilution were spotted onto SD-Ura contai ning histidine (0.6 mM) or hi stidinol (6 mM) plates. All histidinol assays were incubated at 30C and monitored for 3-8 days.
82 Glycosylation Analysis. Experim ents were conducted essentially as described (Kim et al., 2003). Briefly, STY50 transformants were grown to OD600 0.8 to 1 in 250 ml of SD-Ura. After cells were harvested, the pellets were washed with dH2O and stored at -80C. Frozen cells were resuspended in SDS sample buffer (1 mL per 50 mL of cells, 50 mM Tris-H Cl, pH 6.8, 5% glycer ol, 2% SDS, 50 mM dithiothreitol, 5 mM EDTA, 1 mM phenylmethylsulf onyl fluoride, protease inhibitor mixture (Roche Diagnostics, Inc. ), 0.01% bromphenol blue), incubated at 60C for 10 to 15 min and centrifuged for 10 min at 13,000 rpm in an Eppendorf mi crofuge. Soluble fractions were transferred to new tubes. The solution was adju sted to contain 80 mM potassium acetate, pH 5.6. Endo H (0.01 Units, product #11088726001, Roche Diagnostics, Inc.) was added. Samples were incubated at 37C for 2 hours. Nega tive control samples were treated and incubated similarly but without Endo H. RNase B (P7817S, New England BioLabs, Inc.) was used as a positive control for the Endo H reaction and was separated on a 10% polyacrylamide gel that was subsequently stained with Coomassie Blue dye. Solublized proteins from the 3xHA-Suc2-His4C samples were separated on 6.5% SDS-poly acrylamide gels, transferred ont o nitrocellulose membranes, and the membranes were probed with HRP-conjugated anti-HA antibody (Santa Cruz Biotechnology, Inc.). Dual Color Precision Plus ProteinTM Standards (161-0374, Bio-Rad) were used for molecular weight markers. Alternativel y, frozen cell pellets were lysed with glass beads in MIB buffer as described above for total membra ne protein isolation. After the samples were centrifuged at 3000 x g for 10 minutes, the supe rnatant was removed and immunoprecipitation was performed with agarose beads conj ugated to anti-HA mouse monoclonal IgG2a (HA-probe (F-7): sc7392 AC, Santa Cruz Biotechnology) at 4oC with constant shaking (800 rpm) overnight. The samples were washed three times with 1X phosphate-buffered saline (PBS) (Fisher Scientific) and resuspended in SDS sample bu ffer (50 mM Tris-HCl, pH 6.8, 80 mM potassium
83 acetate, 5% glycerol, 5% SDS, 50 mM dithiothreitol, 5 mM EDTA, 0.01% bromphenol blue). The samples were treated with Endo H (0.01 Units Roche Diagnostics, In c.) and incubated at 37C for 2 hours. Control samples were treated and incubated similarly but without Endo H. Solublized proteins were separated on 6.5% SDS-polyacrylamide gels, transferred onto nitrocellulose membranes, and probed with the primary antibody (rabbit polyclonal IgG HAprobe (Y-11), SC-805, Santa Cruz Biotechnology, Inc.) diluted 1:500 in PBS containing 1% BSA. The secondary antibody used was horse radish peroxidase-con jugated goat anti-rabbit IgG-HRP (1:10000, SC-2004, Santa Cr uz Biotechnology, Inc.). Summation In this chapter, different t echniqu es were used to try to determine PAQR localization (using GFP and immunofluorescence) and PAQR topology (using immunofluorescence and the DTR). As described in the Re sults and Discussion, numerous pr oblems were encountered which prevented strong conclusions from being made. For localization, these problems included a lack of GFP expression detection (by Western blot and fluorescence microscopy) and a lack of optimized conditions for immunofluorescence. Fo r topology, based on the intermediate growth on histidinol plates and a lack of glycosylation of the DTR, it wa s concluded that the C-terminus of PAQR6 and PAQR7 are located in the cytoplas m, which is different from the Class I and Class III PAQRs, (Yamauchi et al., 2003; Kim et al., 2006; Daley et al., 2005).
84 A B C Figure 3-1. The yeast dual-topology reporter (DTR). The DTR is a tag on the C-terminal end of a membrane protein. In panel A, the DTR structure is shown with a 3x-HA tag for detection of expression via Western blotting and with glycosylati on sites on the Suc2 and His4C portions indicated by Y-shaped att achments. In panels B and C, the DTR is indicated by the blue colored square. In panel B, if the DTR passes through the ER lumen during protein folding, it becomes glycosylated, but his4 cells can not grow on histidinol. In panel C, if the DT R passes does not pass through the ER lumen during protein folding, it will not become gl ycosylated, but His4C will be able to convert histidinol to histidine to allow growth on histidinol-containing plates. [The figure in panel A was modified from Se ngstag C (2000) Using SUC2-HIS4C reporter domain to study topology of membrane proteins in Saccharomyces cerevisiae Methods Enzymol 327:175-190 with permission from El sevier (Figure 1, page 179).]
85 vector GFP-PAQR5 GFP-PAQR7 Beta-galactosidase activity % of untreated control 0 20 40 60 80 100 120 140 Untreated Progesterone (100 nM) Figure 3-2. Overexpression of N-terminally GFP-tagged PAQRs causes repression of FET3lacZ Yeast cells doubly transformed with the p FET3-lacZ vector and GAL1 -GFPPAQR vector and -galactosidase assays were perfor med. As described in chapter 2, PAQR5 caused constitutive repression of FET3-lacZ while PAQR7 required the addition of progesterone, indi cating that the N-terminal GFP tag does not affect PAQR activity. A B C Figure 3-3. Human Class II PAQR s do not fluoresce when tagged with GFP at the N-terminus. For each panel, the image of th e cells in the bright field was taken (left side of each panel) and the fluorescence image of the same cells was taken (right side of each panel). In panel A, soluble GFP was expre ssed and is visible th roughout the cell. In panel B and panel C, cells expressing GFP-PAQR5 and GFP-PAQR7, respectively, did not fluoresce.
86 1 2 3 4 5 6 7 8 Figure 3-4. Western blot of G FP-tagged PAQRs expressed in yeast. An anti-GFP antibody was used. On the left, the top arrow indicates the position of the 25 kDa molecular weight marker band, and the bottom arrow indicat es the position of the 20 kDa molecular weight marker band. The lanes are as follows: GFP, soluble fraction (lane 1), GFP membrane fraction (lane 2), GFP-PAQR5, soluble fraction (lane 3), GFP-PAQR5 membrane fraction (lane 4), GFP-PAQR7, soluble fraction (lane 5), GFP-PAQR7 membrane fraction (lane 6), PAQR8-GFP soluble fraction (lane 7), PAQR8-GFP membrane fraction (lane 8). vectorPAQR5-GFPPAQR6-GFPPAQR7-GFPPAQR8-GFP Beta-galactosidase activity % of untreated control 0 20 40 60 80 100 120 140 Untreated Progesterone (100 nM) Figure 3-5. Overexpression of C-terminally GFP-tagged PAQRs causes repression of FET3lacZ Yeast cells doubly transformed with the p FET3-lacZ vector and GAL1 -PAQRGFP vector were grown in LIM containing 1 M Fe3+ and 2% galactose and galactosidase assays were performed as de scribed in the Materials and Methods. As described in chapter 2, PAQR5 and PAQR 8 caused constitutive repression of FET3lacZ while PAQR6 and PAQR7 required the ad dition of progesterone, indicating that the C-terminal GFP tag does not affect PAQR activity.
87 1 2 3 4 5 6 Figure 3-6. Western blot of G FP-tagged PAQRs expressed in yeast. An anti-GFP antibody was used. On the left, the top arrow indicates the position of the 50 kDa molecular weight marker band, and the bottom arrow indicat es the position of the 37 kDa molecular weight marker band. The lanes are as follows: PAQR5-GFP, membrane fraction (lane 1), PAQR5-GFP, soluble fraction (l ane 2), PAQR7-GFP, membrane fraction (lane 3), PAQR7-GFP, soluble fraction (l ane 4), PAQR6-GFP, membrane fraction (lane 5), PAQR6-GFP, soluble fraction (lane 6).
88 A B C D E F G Figure 3-7. Immunofluorescence of GFP-tagged and HA-tagged PAQR s. For panels A-D, the anti-GFP antibody was used and for panels E-G, the anti-HA antibody was used. For each panel, the image of the ce lls in the bright field was taken (left side of each panel) and the fluorescence image of the same cells wa s taken (right side of each panel). For negative control cells staine d with the anti-GFP primary antibody (panel A) or the anti-HA primary antibody (panel E), there a ppears to be minimal fluorescence. For soluble GFP, the fluorescence appears to be distributed throughout the whole cell (panel B). Only a few spheroplasts were visible for PAQR5-GFP (panel C), PAQR6GFP (panel D), HA-PAQR5 (panel F) and HA-PAQR6 (panel G) and the fluorescence appears to outline the perimeter of the cell, suggesting a possible plasma membrane localization for these proteins.
89 Figure 3-8. Immunofluorescence of soluble GFP in unpermeabilized spheroplasts. The image of the cells in the bright field was taken (l eft side) and the fluorescence image of the same cells was taken (right si de). These experiments were used as a control to test the unpermeabilized spheroplasts for topol ogy analysis. Soluble GFP was expressed in yeast cells and spheroplasts were generated. Immunofluores cence conditions were used without intentionally permeabilizing the cells; however, permeabilization still occurred leading to immunofluorescent staining of GFP inside the cell. The conditions could not be improved, so this method of topology analysis was not used. vector PAQR6-DTRPAQR7-DTR Beta-galactosidase activity % of untreated control 0 20 40 60 80 100 120 140 160 Untreated Progesterone (100 nM) Figure 3-9. Overexpression of C-terminally DTR-tagged PAQRs causes repression of FET3lacZ Yeast cells doubly transformed with the p FET3-lacZ vector and TPI -PAQRDTR vector were grown in LIM containing 1 M Fe3+ and 2% galactose and galactosidase assays were performed as de scribed in the Materials and Methods. The vector control here expresses an ER membrane-bound subunit of yeast oligosaccharyltransferase (Ost4p-DTR) (Kim et al., 2006) because an empty vector was not available. As described in chapter 2, PAQR5 and PAQR8 caused constitutive repression of FET3-lacZ while PAQR6 and PAQR7 required the addition of progesterone, indi cating that the C-terminal DTR tag does not affect PAQR activity.
90 A B Figure 3-10. Histidinol growth assays of DTR-tagged PAQRs. For both panels, the samples are as follows from top to bottom: Ost4-DTR, PAQR6-DTR, PAQR7-DTR, PAQR6-HA, and PAQR7-HA. Spots from 5 L aliquots of cultures that were adjusted to an OD600 of 0.01 are shown here. In panel A, a plat e containing histidine was used to ensure that each culture of cells wa s viable. In panel B, the pl ate containing histidinol was used to assess whether the His4C portion of the DTR was located in the cytoplasm. 1 2 3 4 Figure 3-11. Endo H digestion of DTR-tagged PAQR6 and PAQR7. A Western blot was performed and an anti-HA antibody was use d. No shift in migration occurred for undigested PAQR6-DTR (lane 1) versus digested PAQR6-DTR (lane 2) or for undigested PAQR7-DTR (lane 3) versus digested PAQR7-DTR (lane 4). 1 2 Figure 3-12. Endo H digestion of RNase B. To ensure that the conditions used for Endo H digestion were allowed for activity of E ndo H, digestion of RNase B was performed and the samples were run on a polyacrylamide gel. A shift in migration occured of the RNase B band occurred for undigested RNas e B (lane 1) versus digested RNase B (lane 2). The gel was stained with Coomassie Blue dye.
91 Table 3-1. Primers used for cloning PAQRs in to pGREG575, pGREG600, and pJK90 (listed from 5 3) Primer name Vector Primer sequence Fwd PAQR5 pGREG575, pGREG600 GAATTCGATATCAAGCTTATCGATACCGTCGACAATGCTGAGCCTGA AGCT Rev PAQR5 pGREG575 GCGTGACATAACTAATTACATGAC TCGAGGTCGACTCATGTTTCTTTT TTATGTAATTCTGG Rev PAQR5 pGREG600 GTTCTTCTCCTTTACTCATTCTCGAGGTCGATGTTTCTTTTTTATGTAA TTCTGG Fwd PAQR6 pGREG600 GAATTCGATATCAAGCTTATCGATA CCGTCGACAATGCTCAGTCTCA AGCTGC Rev PAQR6 pGREG600 GTTCTTCTCCTTTACTCATTCTCGAGGTCGACTGTTGTTTGGCCTGGG TAC Fwd PAQR7 pGREG575, pGREG600 GAATTCGATATCAAGCTTATCGATACCGTCGACAATGGCCATGGCCC AGAAA Rev PAQR7 pGREG575 GCGTGACATAACTAATTACATGACTCGAGGTCGACTCACTTGGTCTT CTGATCAA Rev PAQR7 pGREG600 GTTCTTCTCCTTTACTCATTCTCGAGGTCGACTTGGTCTTCTGATCAA GTTT Fwd PAQR8 pGREG600 GAATTCGATATCAAGCTTATCGATACCGTCGACAATGACGACCGCCA TCTTGG Rev PAQR8 pGREG600 GTTCTTCTCCTTTACTCATTCTCGAGGTCGACGGAATCTTTCTTGGTC AGTCT Fwd PAQR6/90 pJK90 GTGGTTTGTTACGCATGCAAGCTTGATATCGAAATGTTCAGTCTCAA GCTGCC Rev PAQR6/90 pJK90 TGGTCTAGAGGTGTAACCACTTGAGTTCTTAGGCTGTTGTTTGGCCTG GGTAC Fwd PAQR7/90 pJK90 GTGGTTTGTTACGCAT GCAAGCTTGATATC GAAATGTCCATGGCCCA GAAACT Rev PAQR7/90 pJK90 TGGTCTAGAGGTGTAACCACTTGAGTTCTTAGGCTTGGTCTTCTGATC AAGTTT
92 CHAPTER 4 HEMOLYSIN III PROTEINS: BACTERIAL HOMOLOGUES OF PAQR PROTEINS Introduction to Bacterial PAQRs The PAQR fam ily of proteins has evolutiona ry origins in eubacteria with potential bacterial virulence factors called the Hemolysin III (HlyIII) protei ns (Tang et al., 2005; Thomas et al., 2007). Thus far, results of par tial characterizations of HlyIII-1 from Bacillus cereus (Baida and Kuzmin, 1995; Baida and Kuzmin, 1996), HlyIII from Bacillus anthracis (called anthralysin III, or AnlIII) (K lichko et al., 2003), and HlyIII from Vibrio vulnificus (Chen et al., 2004) (referred to as v-HlyIII) indicate that thes e proteins are involved in hemolysis (lysis of human erythrocytes) (Baida and Kuzmin, 1995; Baida and Kuzmin, 1996; Klichko et al., 2003; Chen et al., 2004), and may thus be important for pathogenicity. Roles in Virulence? Although th e role of HlyIII in vivo during pathogenesis is unc lear (Klichko et al., 2003; Chen et al., 2004), a V. vulnificus mutant for hlyIII had a 16-fold increase in the LD50 (dose that is lethal to 50% of test animals) upon infection in mice, suggesting that th is protein plays a role in the pathogenesis of V. vulnificus (Chen et al., 2004). This is especia lly interesting since little is understood about the virulence mechanisms of th is pathogen, which can cause death within 24 hours of infection and has mortality rates as hi gh as 50% for patients with wound infections or 75% for patients with septicemia (the presence of bacteria in the blood system) (Gulig et al., 2005). Whether the HlyIIIs are directly or indirectly in volved in hemolysis is unclear. In the case of HlyIII from B. cereus (called HlyIII-1), Baida and Kuzmin (1996) suggested that this protein functions as a hemolysin by forming pores in the cell membranes of erythrocytes. However, this conclusion was based on indirect evidence be cause crude cell extracts from recombinant
93 Escherichia coli expressing HlyIII were used in the hemo lysis assays (Baida and Kuzmin, 1996). In fact, all of the hemolytic activity characterizations of the HlyIII homologues involved the use of crude cell extracts from recombinant E. coli expressing these protei ns (Baida and Kuzmin, 1996; Klichko et al., 2003; Chen et al., 2004), thus raisi ng the possibility that these proteins are indirectly involved in cell lysis. Hemolysis by purified HlyIIIs has ye t to be demonstrated. Other Possible Functions Interestingly, the HlyIII hom ologues are found in a vari ety of bacteria, including E. coli K12, a non-hemolytic bacterium, leading to the hypot hesis that these proteins may be involved in other, non-pathogenic roles in bacteria. Bioinforma tic predictions of transc ription factor binding sites in E. coli (McCue et al., 2001) have le d to the hypothesis that the E. coli HlyIII homologue (called YqfA) is regulated by a transc ription factor cal led FabR (for F atty A cid B iosynthesis R egulator) and that YqfA is potentially involved in unsaturated fatty acid metabolism (McCue et al., 2001). However, DNA microarra y analysis of a FabR knockout mutant strain showed no significant change in yqfA expression (Zhang et al., 2002), so the role of FabR in regulating yqfA expression is unclear. Interestingly, the HlyIII pr oteins have homologues that affect lipid metabolism in Saccharomyces cerevisiae (Karpichev et al., 2002; Lyons et al., 2004), (called Implicated in Zinc Homestasis protein, or Izhp) and humans (called progestin/adipoQ receptors, or PAQRs) (Yamauchi et al., 2002; Tang et al., 2005). IZH transcription can be induced by different fatty acids (Karpichev et al., 2002; Lyons et al., 2004) and some Izh proteins seem to be involved in altering sterol content (Villa, 2007). The adiponectin receptors also seem to be involved in fatty acid metabolism, as they are receptors for a protein, adiponectin, that cause fatty acid oxidation (Yamauchi et al., 2002; Yamauchi et al., 2003).
94 Mechanisms of Action? The roles of the HlyIII protei ns are no t well-established a nd neither is the mechanism by which they function. As suggested by Baida a nd Kuzmin (1996), these proteins may function by forming pores in cell membranes. Alternatively, the HlyIII proteins may function similarly to their human PAQR relatives by acting as membrane receptors (Tang et al., 2005). Structurally, the HlyIII proteins and the PAQRs share the same predicted seven transmembrane (7TM) architecture (Kyte and Doolittle, 1982) with an intracellular N-terminus and an extracellular Cterminus. These proteins also share four sm all, highly conserved amino acid sequence motifs that are predicted to cluster on the cytoplasmic f ace of the membrane (Lyons et al., 2004; Tang et al., 2005) and may be important for their intrace llular molecular mechanism. If acting as membrane receptors, the HlyIII proteins could activ ate a cytoplasmic protein that is responsible for the hemolytic activity that is associated with HlyIII. Assays conducted in S. cerevisiae for the human PAQRs and all of the yeast IZHs indicate that overexpression of any of these proteins results in repr ession of the promoter for the high affinity iron uptake gene FET3 (Kupchak et al., 2007). Importantly, expression of the uncharacterized PAQR11, which is the closest human homologue of the HlyIIIs, responds like the other, established membrane receptor PAQRs (Figure 2-2). This response is hypothesized to result from an intracellular molecular mechan ism common to all PAQRs (Kupchak et al., 2007). Because the HlyIIIs share the c onserved motifs with the IZHs and PAQRs (Figure 1-1), it was hypothesized that the HlyIIIs can cause similar intracellular signaling in yeast to affect FET3 expression. Although HlyIII homologues can be found in a wide range of eubacterial genomes, these proteins and their human homol ogues have not been well character ized. Because their role in pathogenesis and potential function in lipid metabolism are not well understood, further
95 characterizations of the HlyIII homologues are needed to better understand this family of proteins. This work describes the cloning and expression of two HlyIII homologues from B. cereus (HlyIII-1, HlyIII-2) and one HlyIII homologue from E. coli (YqfA) in S. cerevisiae as well as their effect on the repression of the S. cerevisiae FET3 promoter. In addition, attempts to overexpress these proteins in E. coli for further characterizati on of function were made. Results and Discussion The HlyIII p roteins are an uncharacterized group of proteins that are related to a family of proteins in humans called the PAQRs (Tang et al., 2005; Thomas et al., 2007) and related proteins in yeast called the IZHs (Lyons et al., 2004). It was previously discovered that, in low iron media, the IZHs and human PAQRs can negatively affect expression of FET3 (Kupchak et al., 2007). This is interesting because FET3 encodes a high affinity iron uptake protein and is typically upregulated in low iron media (Askw ith et al., 1994; Yamaguc hi-Iwai et al., 1995). All PAQRs and their homologues share sim ilar amino acid sequence motifs that are predicted to be located intracellularly (Figure 1-1). Thus, it was hypothesized that all of these proteins share the same signaling mechanism. In addition, because there is no evidence in the literature that shows direct invo lvement of HlyIIIs in hemolysi s, it was hypothesized that the HlyIIIs do not directly cause hemolysis. Instead like the PAQRs, the HlyIIIs may be capable of signaling, and this signaling could cau se increased expression or activ ity of a hemolytic protein. Overexpression of some bacterial PAQRs causes repres sion of FET3 To begin examining the possibility that the HlyIII proteins are capable of causing a nonhemolytic physiological response similar to that of the PAQRs and the IZHs, the HlyIIIs we cloned and expressed in S. cerevisiae Overexpression of the E. coli HlyIII (YqfA) protein reduced expression of FET3-lacZ so that the level of -galactosidase activity was 20-25% of that of the vector control (Figure 4-1) This result is similar to that which was observed for the IZHs
96 and PAQRs (Kupchak et al., 2007), which suggests th at YqfA may signal by a mechanism that is similar to that of the IZHs and PAQRs. In addition, it was observed that overexpression of B. cereus HlyIII-2 consistently affected expression of FET3 lacZ so that -galactosidase activity was reduced to 50-60% of that of the vector control (Figure 4-2). B. cereus HlyIII-1 did not seem to affect FET3 lacZ (Figure 4-1). It is possible that the B. cereus HlyIIIs do not properly fold or are not expressed as well as the E. coli YqfA protein in S. cerevisiae. Alternatively, it has been shown that overexpression of some of the human PAQRs can not cause repression of FET3 lacZ unless the ligand for the protein is added during growth in LIM (Kupcha k et al., 2007). So it is possibl e that some HlyIII proteins require a ligand for activati on to cause repression of FET3 lacZ. Because the human Class I and Class II PAQR proteins have known ligands, but th e bacterial HlyIIIs do not, the characterization of the effect of PAQRs on FET3 was refocused on the human Class III proteins. In these studies, progesterone was tested as a potential ligand for PAQR10 and P AQR11, but activation of these proteins did not occur (F igure 2-3 and Figure 2-4). Expression of HlyIIIs in E. coli In addition to expressions in S. cerevisiae, we atte mpted to express the HlyIIIs in E. coli Figures 4-2 and 4-3 show the results of the protein expression for HlyIII-1 (25.5 kDa) and HlyIII-2 (26 kDa), respectively. No protein e xpression has been observed for YqfA (data not shown). Figure 4-2 shows the results of a pilot expr ession of HlyIII-1. In Figure 4-2A, a broad, indistinct band between 20 kDa a nd 30 kDa is present in the indu ced, 4 h HlyIII-1 sample (lane 4), that was not detected in the uninduced 4 h Hl yIII-1 sample (lane 3), the 0 h HlyIII-1 sample (lane 2) or the induced, 4 h pKM260 empty vector sample (lane 1). The same gel was then
97 stained for total protein using C oomassie Blue (Figure 4-2B), but the difference in expression of the between 20 kDa and 30 kDa was not detected. Figure 4-3 shows the results of a pilot expres sion of HlyIII-2. A broad, indistinct band between 20 kDa and 30 kDa is present in the induced, 4 h HlyIII-2 sample (lane 3), but is not detected in the uninduced 4 h Hly III-2 sample (lane 4), the 0 h Hl yIII-2 sample (lane 6) or the induced, 4 h pKM260 empty vector sample (lane 1) The same gel was then stained for total protein using Coomassie Blue (Figure 4-3B), but the difference in expression of the protein at that appears between 20 kDa and 30 kDa was not detected. Because expressions consistently yielded more protein for HlyIII-2 than HlyIII-1, further expression analyses focused on HlyIII-2. The solubility of expressed HlyIII-2, was determined as described in the Materials and Methods. In Figure 4-4A, a comparison of the insol uble cytoplasmic sample and the soluble cytoplasmic sample for the HlyIII-2 expressions (lanes 1 and 2, respectiv ely) shows a distinct band between 20 kDa and 30 kDa in the insoluble sa mple, but not in the soluble sample. This band does not appear in the insoluble cytoplas mic fraction from the negative control sample (C43(DE3) containing pKM260, lane 3). These results indicate that a protein between 20 kDa and 30 kDa appears to be expressed in the HlyIII-2 samples as an insoluble protein. Es timation of the molecular weight of this band is 23 kDa based on the migration distances of this band and the migration distances of the 15-40 kDa BenchmarkTM His-tagged Standard protein bands (the linear range of separation of 15% polyacrylamide gels is 15-43 kDa) (Sambrook a nd Russell, 2001). The same gel was stained for total protein using Coomassie Blue (Figure 4-4B), but the difference in expression of the protein at 23 kDa was not detected.
98 Estimation of the amount of protein that is present at 23 kDa in the insoluble cytoplasmic sample (lane 1, Figure 4-4A), could not be made based on the amount of protein present in each of the BenchmarkTM His-tagged Standard protein bands becau se this marker is not recommended by the manufacturer for quantitat ion (Invitrogen). Because the 23 kDa protein in the insoluble cytoplasmic sample can be detected with the InVision His-tag In-gel Stain (lane 1, Figure 45A) but not Coomassie Blue staining (lane 1, Fi gure 4-4B), a rough estimate of the quantity of protein in this band was made based on the det ection limits of the InVision His-tag In-gel Stain (~15 ng for a 23 kDa protein, Invitrogen ) and Coomassie Blue staining (~100 ng of polypeptide in a single band) (Sambrook and Ru ssell, 2001). Based on these differences in detection limit, the quantity of the 23 kDa protein per liter of cell culture was estimated to be between 10-60 g. The quantity of expressed HlyIII was very low (10-60 g per liter) and the proteins seemed to be expressing as insoluble inclusion bodies, which makes purification and refolding extremely difficult. Thus we decided to focus our characterization of the PAQRs and their homologues on other areas, including the e ffects of their expression on FET3 in S. cerevisiae, as described in Chapter 2. Materials and Methods Cloning of the Bacterial Hemolysin III Genes To clone hly III-1 and hlyIII-2 Bacillus cereus strain 10987 genomic DNA was obtained from the American Type Culture Collection (ATCC). To clone yqfA a colony of Eschericia coli TOP10 (Invitrogen) was used directly in PCR reactions. For yeast expression plasmids, the amplified gene sequences were cloned into the pRS316 plasmid (Sikorski and Hieter, 1989) via homologous recombination. After PCR amplifi cation of the genes and restriction enzyme
99 digestion of the pRS316 vector (BamHI, New England Biolabs R0136 and SacI New England Biolabs R0156), the PCR product and cut vector were transformed into the wild-type S. cerevisiae strain BY4742 (mat ) via the standard lithium acetate protocol (Guthrie and Fink, 1991) for homologous recombination. The DNA was rescued from the cells and transformed into the TOP10 strain of E. coli (Invitrogen). The Wizard Pl us Minipreps DNA Purification System [Promega A7500] was used to prepare the plasmid DNA from the E. coli cells. The DNA sequence was confirmed by the DNA Sequencing Co re Facility [Univers ity of Florida]. For E. coli expression plasmids, hlyIII-1 hlyIII-2 and yqfA were cloned into pKM260 and pilot expressions were performed. The pKM260 vector derived from the pET3b vector, has a T7 promoter and is constructed in such a way to allow N-terminal 6x-histid ine tagging of expressed proteins (Melcher, 2000). A tobacco etch virus (TEV) protease cleavage site is located between the 6x-histidine tag and the N-terminus of the expressed protein. To clone hlyIII-1 and hlyIII-2 Bacillus cereus strain 10987 genomic DNA was obtained from the American Type Culture Collection (ATCC). To clone yqfA a colony of Eschericia coli TOP10 (Invitrogen) was used directly in PCR reactions. The amplified ge ne sequences were cloned into pKM260 using standard restriction enzyme and ligation met hods (Sambrook and Russell, 2001). The plasmids were transformed into competent TOP10 E. coli cells (Invitrogen) by standard procedures (Sambrook and Russell, 2001), and the sequences of the cloned genes were confirmed by the ICBR DNA sequencing core laboratory at the University of Florida. Expression plasmids constructed with pKM260 that contain gene in serts have been named as follows: pJLS1 (hlyIII-1 ), pJLS2 ( hlyIII-2 ), and pJLS3 ( yqfA).
100 Expression of the Hemolysin III Proteins in S. cerevisiae The wild-type BY4742 (m at ) strain of S. cerevisiae was double-transformed with pFET3 lacZLEU2 and either pRS316GAL1 URA3 or pRS316GAL1 -HlyIIIURA3 Yeast overnight cultures were routinely grown at 30 oC in synthetic dextrose (SD) media with the appropriate selection of amino aci ds. Overnight cultures were re inoculated into low iron media (LIM) (Eide and Guarente 1992), containing 1 M FeCl3 for iron-defficient media or 1 mM FeCl3 for iron-replete media. Galactose (2%) was used to fully induce expression of GAL1 driven genes. galactosidase Reporter Assays Transformants containing p FET3 -397 and overexpression plasmids were used in galactosidase assays as previously described (Kupchak et al., 2007). Cells were grown to midlog phase in LIM and -galactosidase assays were performed on permeabilized cells as described (Guarente, 1983). Production of the lacZ gene product ( -galactosidase) was determined by measuring the activity of the enzyme. Upon hydrolysis of o-nitrophenyl-galactopyranoside (colorless) to o-nitrophenol (yellow) and galact ose, the absorbance at 420 nm was measured. Activity (in Miller Units) was determined as follows: (A420 x 1000)/(mL of culture x reaction time(minutes) x A600). Expression of the HlyIII Homologues in Escher icia coli For expression of each gene, the C43(DE3) strain of E. coli was used, as it is more amenable to membrane proteins expressions (Miroux and Walker, 1996). Also, HlyIII-1 seems to have toxic effects when expressed in E. coli (Baida and Kuzmin, 1995) so procedures for expression of toxic genes in pET vectors were used (Novagen 2002). An overnight culture of the transformed cells was grown at 30oC in 5 mL of LB/Amp (200 g/mL) and 1% glucose. The
101 culture was re-inoculate d using a 1:100 dilution into fresh media containing ampicillin (500 g/mL) and incubated at 30oC with shaking until the OD600 reached 0.6-0.8. The cells were collected by centrifugatio n and resuspended in fresh LB/Amp (500 g/mL). Expression was induced with 0.4 mM IPTG for 4-16 hours at 30oC. Aliquots of cells (1 mL each) were removed at different time poi nts between 0-16 h post-induction. The aliquots we re prepared for analysis by SDS-PAGE as described above except that after sample buffer was added, the mixture was incubated at 42oC for 30 minutes. These conditions were used for all subsequent protein expressions and analyses. Each gel was first stained using the InVision His-tag In-gel Stain (Invitrogen) according to the manufacturers protocol This stain consists of a fluorescent dye conjugated to Ni2+:nitriloacetic acid (NTA ) complex. The same gel was then stained for total protein using Coomassie Blue. To determine the solubility of expressed HlyIII-2, procedures outlined in the manufacturers pET System Manual (Novagen 2002) were followed for soluble versus insoluble (inclusion body) cytoplasmic fractions. After 4 hour expressions, pellets from 40 mL cultures were lysed using buffer A (4 mL) [50 mM Tris -HCl (pH 8), 5% glycer ol, 100 mM NaCl, 10 mM 3-[(3-Cholamidopropyl) dime thylammonio]-1-propanesulfonate (CHAPS), 1 mM Ethylenediamine Tetraacetic Acid (EDTA),1 mM Phenylmethylsulfonyl fluoride (PMSF) and egg white lysozyme (0.2 mg/mL)]. The mixtures were incubated at 37oC for 20 minutes and then placed on ice while sonica tion was performed. Samples were centrifuged at 3000 rpm to remove unlysed cells, and the supernatants were centrifuged at 16,000 x g (14,500 rpm) for 30 minutes at 4oC to isolate inclusion bodies in the pellet. The incl usion body pellet was washed with buffer A twice, then resuspended in 0.25 mL of 1% SDS with heating (42oC for 30 minutes)
102 and mixing (1400 rpm). The supernatant was sa ved for subsequent analysis by SDS-PAGE. SDS-PAGE samples were pr epared by mixing a 10 L aliquot of the soluble protein fraction or solubilized inclusion bodies w ith an equal aliquot of 2X sample buffer. The mixture was incubated at 42oC for 30 minutes prior to SDS-PAGE analysis. Summation The HlyIII p roteins have been proposed as pore-forming toxins that cause lysis of red blood cells (Baida and Kuzmin, 1995; Baida and Kuzmin, 1996; Klichko et al., 2003; Chen et al., 2004). Alternatively, the HlyIII proteins may func tion similarly to their human PAQR relatives by acting as membrane receptors (Tang et al., 2005). Here, we showed th at, like the human and yeast PAQRs, the HlyIII proteins can cause repression of FET3-lacZ when expressed in yeast. Whether this occurs because of activation of a common signaling pathway remains to be determined. Further characterizations of the Hl yIIIs were attempted by e xpressing these proteins in E. coli, but only low levels of expr ession could be obtained.
103 Vector HlyIII-1 HlyIII-2 YqfA Beta-galactosidase activity % of vector control 0 20 40 60 80 100 120 Figure 4-1. Overexpression of some bacterial PAQR homo logues causes repression of FET3lacZ HlyIII-1 and HlyIII-2 from B. cereus and YqfA from E. coli were tested. Yeast cells doubly transformed with the p FET3-lacZ vector and protein expression vector were grown in LIM containing 1 M Fe3+ and 2% galactose and -galactosidase assays were performed as described in th e Materials and Methods. The proteins do not have epitope tags and an tibodies to the proteins do not exist, so expression was not confirmed. Also, there are no known lig ands for the bacterial PAQRs that could be tested for ligand-dependent repression of FET3-lacZ .
104 Figure 4-2. SDS-PAGE of HlyIII-1 expressions from the pKM260 vector. The same gel was stained with the InVision Hi s-tag In-gel Stain (Invitrogen) (A), and then Coomassie Blue stain (B). The contents of the lanes are as follows: pKM260 (lane 1), 4 h induced; HlyIII-1, 0 h time point (lane 2) ; HlyIII-1, uninduced 4 hour time point (lane 3); HlyIII-1, induced 4 hour time point (lan e 4); HlyIII-1, induced 2 hour time point (lane 5); BenchmarkTM His-tagged Standard (lane 6, I nvitrogen). This was a 12% polyacrylamide gel. Figure 4-3. SDS-PAGE of HlyIII-2 expressions using the pKM260 vector. The same gel was stained with the InVision Hi s-tag In-gel Stain (Invitrogen) (A), and then Coomassie Blue stain (B). The contents of the lanes are as follows: pKM260 (lane 1), 4 h induced; BenchmarkTM His-tagged Standard (lane 2); HlyIII-2, induced 4 h time point (lane 3); HlyIII-2, uninduced 4 hour tim e point (lane 4); HlyIII-2, induced 2 h time point (lane 5); HlyIII-2, 0 h time point (lane 6). This was a 12% polyacrylamide gel.
105 Figure 4-4. SDS-PAGE of solubl e cytoplasmic samples and insoluble cytoplasmic samples for HlyIII-2 in pKM260. Expressions were at 4 h at 30oC. The same gel was stained with the InVision His-tag In -gel Stain (Invitrogen) (A), and then Coomassie Blue stain (B). The contents of the lanes are as follows: HlyIII-2 insoluble cytoplasmic sample (lane 1, the band for HlyIII-2 is indi cated with an arrow); HlyIII-2 soluble cytoplasmic sample (lane 2); pKM260 in soluble cytoplasmic sample (lane 3); BenchmarkTM His-tagged Standard (lane 4).
106 Table 4-1. Primers used for cloning HlyIIIs (listed from 5 3) Primer name Vector Primer sequence Fwd yqfA pRS316 TACTTCTTATTCCTCT ACCGGATCCCGCTCGAGGTCGACATGGTTCAGAA GCCCCTC Rev yqfA pRS316 TGAGCGCGCGTAATAC GACTCACTATAGGGCGAA TTGGAGCTCTTACGC CTGCCCAATATAC Fwd hlyIII-1 pRS316 TACTTCTTATTCCT CTACCGGATCCCGCTCGAGGTCGACATGACACAATT TGTGAAAGAA Rev hlyIII-1 pRS316 TGAGCGCGCGTAAT ACGACTCACTATAGGGCGAATTGGAGCTCTTATGA TGCTGTAGGTAGG Fwd hlyIII-2 pRS316 TACTTCTTATTCCT CTACCGGATCCCGCTCGA GGTCGACATGAATGCTTA TGTAAGGGA Rev hlyIII-2 pRS316 TGAGCGCGCGTA ATACGACTCACTATAGGGCGAATTGGAGCTCTTAAAT TACGTAACAATATACA Fwd yqfA pKM260 CACCATCACCATGCTAG CGAGAATCTTTATTTTCAGGGCGCCATGGTTCA GAAGCCCCTC Rev yqfA pKM260 ATATCTGCAGAATTCCAGCACACTGGCGGCCGTTACTAGTGGATCCTTAC GCCTGCCCAATATAC Fwd hlyIII-1 pKM260 CACCATCACCATGCTAGCGAGAATCTTTATTTTCAGGGCGTCATGACACA ATTTGTGAAAGAA Rev hlyIII-1 pKM260 ATATCTGCAGAATTC CAGCACACTGGCGGCCGTTACTAGTGGATCCTTAT GATGCTGTAGGTAGGA Fwd hlyIII-2 pKM260 CACCATCACCATGCT AGCGAGAATCTTTATTTTCAGGGCGCCATGGATGC TTATGTAAGGGAA Rev hlyIII-2 pKM260 ATATCTGCAGAATTC CAGCACACTGGCGGCCGTTACTAGTGGATCCTTAA ATTACGTAACAATATACAC
107 CHAPTER 5 FUTURE DIRECTIONS FOR CLA SS II PAQR CHARACT ERIZATIONS Introduction This dissertation has described the characte rization of the Class II PAQRs in yeast, including the use of a prom ote r-reporter assay to study signaling and structure/function studies. These studies indicate that the Class II PAQRs, including the uncharacter ized proteins PAQR6 and PAQR9, are able to sense and respond to pr ogestins and do not requ ire the presence of Gproteins (Chapter 2). In addi tion, attempts to determine the localization and topology of these proteins when expressed in yeast were also perf ormed, but were generally inconclusive (Chapter 3). The characterization of the bacterial Class III PAQRs was pe rformed in yeast and indicates that some of these proteins are able to transduc e signals in this organism as well (Chapter 4). Still, much work remains in order to unde rstand the PAQR family of proteins. In this Chapter, ideas for future Clas s II PAQR studies are described, including identification of novel protein-pr otein interactions between PAQRs and other proteins and study of a possible intrinsic enzymatic activity of the PAQRs. In addition, some preliminary work for other future studies is describe d, including the establishment of an expression system using Sf9 insect cells and ligand binding assays conducte d with membranes from yeast expressing the PAQRs. Identification of Novel Interactions Betw een PAQRs and Other Proteins While we were able to demonstrate that th e presence of progestins causes a physiological effect in cells that express the Class II P AQRs, several questions remain regarding the mechanism of signaling initiated by the Class II PAQRs and whether this mechanism is conserved for all PAQR family members.
108 By expressing the Class II PAQRs in yeast in the absence of any other human proteins, we have demonstrated that human G-proteins are not necessary for signaling. In addition, we showed that the Class II PAQRs can signal in the absence of yeast G-proteins as well (Chapter 2). Thus, the previous characte rization of the human Class II PAQRs as GPCRs (Thomas et al., 2007) should be re-evaluated. First, the evidence for the human Class II PAQRs as GPCRs includes bioinformatic predictions of structural simila rity, with seven predicted tran smembrane domains (Zhu et al., 2003a); however, this evidence is not concrete. Other analyses ha ve predicted that the Class II PAQRs have eight transmembrane domains (Lyons et al., 2004; Fernandes et al., 2005). Also, no researchers have performed a detailed experiment al analysis of the number of transmembrane domains or the topology of the proteins in the membrane. Other evidence for the Class II PAQRs as GPCRs includes an increase in [35S]GTP -S bound to membranes of cells transfected with sea trout PAQR7 or human PAQR7 in the presence of 17,20,21-trihydroxy-4-pregnen-3-one (20 -S) or progesterone, respectively (Thomas et al., 2007). In an incomplete characte rization of these two proteins, Thomas et al. (2007) demonstrated that increased [35S]GTP -S binding can be abolished for a truncated sea trout PAQR7 protein, but a truncated human PAQR7 was not tested. Finally, Thomas et al. (2007) claim that PAQR7 co-immunoprecipite d with G-proteins when an anti-Gi/o antibody was used for immunoprecipita tion, although the evidence presented is unclear, as the Western blot bands for PAQR7 prot eins were faint and no control was shown for untransfected cells or similar co-immunoprecipitati on experiments with a membrane protein that is not a GPCR.
109 It is possible that the Class II PAQRs are coupled to G-proteins; however, it is also possible that their ability to sense and res pond to the presence of progestins can occur independently of G-proteins. If the Class II PAQRs are able to function independently of Gproteins, then how are they able to initiate a signaling re sponse in the presence of progesterone? One approach to try to answer this ques tion lies in the contin ued use of yeast to characterize these proteins. Pr eviously, Mao et al. (2006) used the cytoplasmic domain of PAQR1 in the yeast two-hybrid syst em to screen a human cDNA library to identify interactions that take place between PAQR1 a nd other human proteins. In th is study, an interaction between PAQR1 and APPL1 (adaptor pr otein containing pleck strin homology domain, phosphotyrosine binding (PTB) domain and leucine zipper motif) was identified and this interaction was verified in mammalian cells (Mao et al., 2006). Mao et al. (2006) also observed that APPL1 and a small GTPase (Rab5) interacted more upon treatment of cells with adiponectin, leading to translocation of the glucose transport 4 (GLUT4 ) protein to the membrane. It is possible that the Class II PAQRs also in teract with the APPL1 protein to lead to downstream signaling effects. Interestingly, th e yeast homologue of APPL1, is required for Izh2p-, PAQR1-, and PAQR2-dependent repression of FET3 (Kupchak et al., 2007). Amongst other proteins, Kupchak et al. (2007) also determined that the yeast and human Class I PAQRs required the yeast AMPK subunits SIP1 and SNF4 for FET3-lacZ repression. Interestingly, in human cells, it has been demonstrated that AM PK is involved in PAQR1 and PAQR2 signaling (Yamauchi et al., 2003). By using a genetic muta tional approach similar to that described in Chapter 2 and by Kupchak et al. (2007), the FET3-lacZ assay could be used in combination with several mutants to determine if the Class II PAQRs require the same yeast proteins as the yeast and human Class I PAQRs for PAQR-dependent FET3-lacZ repression.
110 Similarly to Mao et al. (2006), an approach using the yeast two-hybrid system could be used to identify novel interactions between the Class II PAQRs and other human proteins. Because the locations (cytoplamic or extracytopl asmic) of different portions of the Class II PAQRs are unclear, a yeast two-hyb rid approach in which the whol e PAQR protein is used could be better than just using a so luble portion of the protein, as Mao et al. (2006) did. One such system that can be used for iden tification of membrane protein-pr otein interactions is the splitubiquitin membrane yeast two-hybrid sy stem (MbYTH) (Iyer et al., 2005). For the MbYTH, the Class II PAQRs would be fuse d at either their Nor Cterminus to the C-terminal half of ubiquitin (Cub) and a tran scription factor (Cub-TF ). The Class II PAQRs would serve as the bait wh ile a library of human cDNAs encoding fusions between human proteins and the N-terminal half of ubiquitin (N ub) would serve as the prey. Upon interaction of a PAQR with a bait protein, the two halves of ubiquitin would be reconstituted and recognized by ubiquitin-specific proteases leading to cleavage of the transc ription factor (I yer et al., 2005). The transcription factor then activates transcription of a reporter gene (Iyer et al., 2005). The prey protein from clones which express the reporte r can be identified and the interaction between the PAQR and the prey protein can be confirmed by another method, such as coimmunoprecipitation. Because the Cub-TF must be located in the cytoplasm for inter actions to be detected (Iyer et al., 2005) and the topology of the human Cl ass II PAQRs is debated (Thomas et al., 2007; Tang et al. 2005), the ability to identify proteins that interact with the Nor C-termini using this system could help in topology determination; however, difficulty identifying protein-protein interactions could be due to a topology that results in the ex tracytoplasmic localization of Cub-
111 TF. In this case, truncated Class II PAQRs that are still functional could be used. Functionality could be tested by the FET3-lacZ assay described in Chapter 2. Possible Intrinsic Enzymatic Activity of PAQR Proteins In addition to identifying potential protei n-protein interactions involved in PAQR signaling, another possibility is th at the PAQRs have an intrinsic enzym atic activity that leads to a signaling cascade and ev entually repression of FET3-lacZ PSI-BLAST searches indicated that the PAQR proteins are distantly related to the alkaline ceramidase family of proteins (Lyons lab, unpublished). Alkaline ceramidases in yeast are Ypc1p and Ydc1p and catalyze the hydrolysis of yeast ceramides to produce sphingosine and fa tty acids, with the catalytic activity of Ypc1p being reversible (Mao et al ., 2000a; Mao et al., 2000b). Ceramide and sphingosine serve as important signaling molecules (Mao et al., 2000a; Mao et al., 2000b). So how might sphingolipids affect FET3 ? Based on experimental evidence, Kupchak et al. (2007) proposed that Izh2p overexpr ession leads to repression of FET3-lacZ by activating the Nrg1p/Nrg2p repressors via AMPK and/or inactivating the Msn2p/Msn4p activator s via PKA. If the PAQRs are able to alter levels of sphingolipids, signa ling events may occur, such as activation of AMPK or PKA, that ultimately lead to repression of FET3-lacZ Some experimental evidence from yeast for this hypoth esis includes alleviat ion of Izh mediated FET3lacZ repression in the presence of Derythro -2-(N-myristoylamino)-1-phenyl-1-propanol (Kupchak, 2008), a known alkaline ceramidas e inhibitor (Bielawska et al., 1996). Further evidence that sphingolipid metabolis m is affected by PAQRs was described by Villa (2007) in experiments in which the levels of sphingoid bases increased when yeast PAQRs were overexpressed. Interestingly, in a global study of signal trans duction pathways affected in yeast by PAQR7 overexpression, the yeast proteins Pil1p and Lsp1p were identified as having increased phosphorylation (Regalla, 2007). Pil1p (phosphorylation is inhibited by long chain
112 bases) and Lsp1p (long chain bases stimulate phos phorylation) are regulat ed by the sphingolipid long-chain base-sensing kinases Pkh1p and Pkh2p (Z hang et al., 2004). It was determined that PAQR-mediated repression of FET3-lacZ is dependent on the presence of Pkh1p and Pkh2p (Kupchak, 2008). Although direct enzymatic activity assays for isolated PAQRs still need to be carried out and optimized, the evidence thus far supports that the PAQRs could affect sphingolipid metabolism. Similar experiments in yeast could be c onducted for the Class II PAQRs to examine whether or not they too may be involved in sp hingolipid metabolism. Of course, the results would need to be verified in mammalian cells befo re claims of this nature could be made. Expression of Class II PA QRs in Sf9 Insect Ce lls Because there is much conflict in the literat ure regarding the existence of membrane-bound steroid receptors (see Chapter 2) a model organism that is s impler than mammalian cells could be used to clarify the role of certain proteins in st eroid signaling. While the majority of this thesis has focused on the use of yeast to charact erize the Class II PAQRs as steroid receptors, cells from the insect Spodoptera frugiperda (or Sf9 insect cells) are al so an attractive option for Class II PAQR studies for several reasons. Sf9 insect cells have been useful for the func tional expression of many eukaryotic proteins (Altmann et al., 1999), including multi-transmem brane domain proteins such as GPCRs (McIntire et al., 2002) and th e human PAQR1 and PAQR2 prot eins (Dr. Chasta Parker, Winthrop University, personal communication). Expression of PAQR1 and PAQR2 in insect cells caused activation of AMPK in the presence of adiponectin (Dr. Chasta Parker, Winthrop University, personal communication). Because there has been success with the functional expression of some PAQRs, it was expected that Sf9 insect cells could also be used for characterizations of the Class II PAQRs. Also, Sf9 insect cells are an especially attractive
113 organism to use for Class II PAQR characteriza tions because insects do not have Class II PAQR homologues (Thomas et al., 2007) and may not have nuclear progesterone re ceptors (Maglich et al., 2001) that could interfere with individual huma n Class II receptor characterizations. Ligand Binding Experiments While our yeast system for studying the Class II PAQR response to progestins seems to indicate that these proteins sense and respond to these steroids, liga nd binding assays to membranes from cells expressing the PAQRs shoul d be performed to further support that the PAQRs are receptors for progestins. So far, diffe rent groups have obtained conflicting results for progesterone binding to membranes from bacterial or mammalian cells expressing PAQR7 (Zhu et al., 2003a; Krietsch et al., 2006; Thomas et al., 2007). Most of the methods used to demonstrat e progesterone binding to Class II PAQRs involved radiolabeled progesterone. In these assays, crude membra ne preparations or purified plasma membrane preparations were used (K rietsch et al., 2006; T homas et al., 2007). Membrane samples were incubated with tritiated progesterone and then the samples were filtered to remove unbound progesterone. Thus, we conducted some preliminary ligand binding experiments with cell membranes isolated from yeast cells overexpressing the PAQRs. While our experiments were limited by the lack of appropriate equipment, the attempts to demonstrate binding are presented in this ch apter with recommendations to improve similar experiments in the future. Results and Discussion This chapter, as well as Chapter 2, descri bes the use of non-m amma lian eukaryotic model organisms (yeast and Sf9 insect cells) for the expre ssion of the Class II PAQRs. In this chapter, ideas for future studies of the Class II PAQRs ar e described and preliminary results for some of
114 these ideas, including the use of insect cells for expression and ligand binding assays, are presented. Like yeast, insect cells do not have endoge nous Class II PAQRs (Thomas et al., 2007) and may not have nuclear progesterone receptors (Mag lich et al., 2001). These features make these two organisms attractive for the study of these proteins because interference from endogenous progesterone receptors does not exist. While the ma jority of our characterizations of the Class II PAQRs were performed in yeast, expression of these proteins in Sf9 insects cells were initiated as described in this chapter. In addition, at tempts to demonstrate binding to membranes from yeast cells expressing the Class II PAQRs are also described. Expression of Class II PAQRs in Insect Cells Initial attempts to de tect the expression of PAQR5 and PAQR6 were not successful. However, one attempt yielded successful expression of a protein that is close to the expected size of PAQR5 and PAQR6 (~42 kDa) (Figure 5-1). Because detectable expression of the PAQRs was inconsistent and we had some success with yeast expressions of Class II PAQRs, efforts were focused on the use of yeast. Because some success was achieved with expressing the Class II PAQRs in insect cells, this system can be considered for future studies, including topology studies. In fact, insect cells may be more amenable than yeast for topology studies by immunofluorescence because insect cells do not have cell walls that need to be removed. As descri bed in Chapter 3, the procedure for removal of yeast cell walls (called spheroplasting) causes permeabilization of the cells. Permeabilization of the cells allows access of the antibody to the epitope, whether it is located inside or outside the cell, making topology determination difficult.
115 In addition to topology studies, the Sf9 insect cells can be used to further explore the possibility that these proteins are coupled to Gproteins, as this system has been used for the characterization of GPCRs (M cIntire et al., 2002). In addition, because preliminary work with the Class I PAQRs indicates that these proteins could be expressed sufficiently in the Sf9 insect cell system for purification (Dr. Chasta Parker, Winthrop University, personal communication), it is possible that the Class II PAQRs could also be expressed in this system and purified for future biochemical characterizations, including ligand binding, signaling pathway st udies, and structural studies. Ligand Binding Experiments We wanted to dem onstrate that progesterone binds to membranes containing the PAQRs. Because our expressions in yeast were consiste ntly more successful than our insect cell expressions, we used yeast to isolate membrane s containing the PAQRs. Because of equipment limitations, we chose to pursue a single-point bindi ng assay to try to dem onstrate higher specific binding to membranes from cells expressing PAQR s versus vector control cells. The vacuum manifold to which we had access was capable of f iltering twelve samples per time point and did not have a pressure regulator. With this system, the nonspecific binding (NSB) was high compared to total binding (TB), and we found a rather high error when trying to determine total binding and nonspecific binding for triplicate samples. The results for PAQR5 and PAQR6 are presented in Figure 5-2. Previous single point bindi ng assays conducted by Thomas et al. (2007) demonstrated that progesterone bound to plas ma membranes of transfected human cells expressing PAQR7 about 2.5 times more than plas ma membranes from untransfected cells. In our case, triplicate samples (for PAQR5, Figure 52) and duplicate samples (for PAQR7, data not shown) did not seem to indica te that there was significant bi nding of progesterone to cells
116 expressing either of these proteins (data not shown). For PAQR6, when specific binding was calculated, there was a hi gh error (Figure 5-3). There are several possible reasons that we are observing poor specific binding and poor reproducibility of progesterone to the membranes of cells expressing the PAQRs. First, steroids are considered to be very sticky, which can ca use them to bind to various proteins and other biological structures, such as membranes (Wehli ng et al., 2007). This stickiness would cause high nonspecific binding. In addition, according to Thomas et al. (2002), the Class II PAQRs have rapid association and dissociation rates. This is especially impor tant because rapidly dissociating ligand-receptor complexes require a method for removal of unbound ligand that minimizes dissociation, such as rapid separation times and wash steps (Qume, 1999). With our system, the filtration was not uniform for the twelve different sample holders as some sample holders seemed to allow more rapid filtration than others. Variations in sample filtering times could cause variation in dissociation of specifically bound pr ogesterone during washes. Also, with the system we used, the vacuum pressure could not be regulated, so the pressure could have fluctuated between filtration of samples and during the washes. Fluctuations in vacuum pressure can cause fluctuations in the amount of time that it take s to filter samples and the amount of dissociated ligand that can be washed away. These pr oblems could be improved by using a 96-well cell harvester that allows for simultaneous filtering and washing of binding reactions at controlled vacuum pressures. Finally, it is possible that membrane solubili zation causes loss of liga nd binding activity. It has been noted that different preparation proc edures have caused loss of ligand binding activity of some receptors (Thomas et al., 2002). Op timization of parameters such as protein
117 concentration may also improve the results. If the protein concentrati on is too low, specific binding would be lower. If the protein con centration is too high, nonspecific binding to nonPAQR proteins and membrane components could in crease. Until a better sample filtering system can be used, it is difficult to assess these possibilities. Materials and Methods Plasmids. For Class II PAQR expr ession in insect cells, the genes for these proteins were cloned into pIEx-4 (Novagen) in collaboration with Dr. Chasta Parker (Winthrop University). This plasmid has an Autographa californica nucleopolyhedrovirus (AcNPV) derived hr5 enhancer and immediate early promoter (IE1) and allows for C-terminal 6x-histidine tagging of proteins. Primers were designed to have an Nco I or Pci I s ite at the beginning of th e gene and a Not I site at the end of the gene (Table 5-1) for cloning into the pIEx-4 vector. Expression of PAQR5 and PAQR6 in Sf9 Insect Cells Sf9 insect cells (71104-3, Novagen) were passaged and expressions were perform ed according to the manufacturers instructions us ing serum-free BacVector Insect Cell Medium (70590-3, Novagen). Rapidly growing cells (1 x 107) were used in 10 mL suspension cultures. Approximately 20 g of plasmid DNA was diluted with 1 mL of serum-free medium. Separately, 100 L of Insect GeneJuice Transfection Reagent (71259-4, Novagen) was diluted with 1 mL of serum-free medium and the DNA was added to this mixture dropwise. The transfection mixture was added to the cells in a 125 mL Erlenmeyer flask. The cells were grown at 28oC with shaking for 48 hours. The cells were ha rvested and the pellets were stored at -80oC. To detect PAQR protein expr ession, the Insect PopCultureTM Reagent (71187-3, Novagen) was used according to the manufacturers inst ructions. Briefly, 0.05 culture volume of Insect PopCultureTM Reagent and 100 Units of RQ1 RNase-Free DNase (M610A, Promega) were
118 added to thawed cell pellets. The sample was mixed by inversion and incubated for 15 minutes at room temperature. To detect expression of 6xhistidine tagged PAQRs, the SuperSignal WestHisProbeTM Kit (15168, Pierce) was used accord ing to the manufacturers in structions. Briefly, equal amounts of protein (in g) were loaded onto a 10% polyacr ylamide gel and electrophoresis was performed according to standard procedures (S ambrook and Russell, 2001). Western blots were performed using nitrocellulose according to stan dard procedures (Sambrook and Russell, 2001). Transfer buffer was Tris-glycine, pH 8.3, 10% methanol, and 0.1% SDS. The transfer was performed at 80 volts for 60 minutes. Membranes were blocked overnight at 4oC with 1% bovine serum albumin (BSA) in BupHTM Tris Buffered Saline containing 0.05% Tween-20 (TBST). Membranes were incubated with the HisProbeTM-HRP (diluted 1:5000 in TBST). The blot was incubated with SuperSignal West Pico Substrate Working So lution (Pierce) for 5 minutes, exposed to film and developed. Ligand Binding Assays for PAQRs Expressed in Yeast Ligand binding was tested with total m embrane extracts from cells expressing the PAQRs. Briefly, yeast cells transformed with expression vectors for the HA-tagged PAQRs (described in Chapter 2) were grown in synthe tic dextrose (SD) media and re inoculated into LIM containing 2% galactose. Total membrane pellets were prep ared as described in Chapter 2 with membrane isolation buffer (MIB) and ultracentrifugation at 130,000 x g for 90 minutes. To detect expression of HA-tagged proteins, Western blots were performed as described in Chapter 2 with the primary antibody (rabbit polyclonal Ig G HA-probe (Y-11), SC-805, Santa Cruz Biotechnology, Inc.) diluted 1:500 in PBS contai ning 1% BSA. The secondary antibody used was horse radish peroxidase-conjugated goat anti-rabbit IgG-HRP (1:10000, SC-2004, Santa
119 Cruz Biotechnology, Inc.). The blot was incuba ted with SuperSignal West Pico Substrate Working Solution (Pierce) for 5 minutes, exposed to film and developed. For binding reactions, samples for total bind ing were set up to have 5 nM tritiated progesterone (3H-PG, P5050, Sigma), an amount similar to that which was previously used in Class II PAQR single-point bindi ng assays (Karteris et al., 2006) in ethanol, while nonspecific binding samples had 1000 fold excess (5 M) cold progesterone in a ddition to 5 nM tritiated progesterone. The ethanol concentration was adjusted to 2% when necessary. In 500 L of YP (yeast extract and peptone) to reduce nonspecific binding (Blumer et al., 1988), either 40 ng/ L total protein or 110 ng/ L total protein was used, similar to amounts described in Windh and Manning (2002) for radioligand binding assays usi ng tritiated ligands for GPCRs. Samples were incubated at room temperature before being fi ltered onto GF/B filters (Whatman) that were presoaked in polyethyleinimine (0.5%) (Thomas et al., 2007; Krietsch et al., 2006). A 1225 Sampling Manifold (Millipore) was used to filter the samples. The vacuum was supplied by an aspirator vacuum pump without a pressure regulator. A 100 or 200 L aliquot of sample was filtered and washed three times with 1 mL of ice cold YP. After the filters dried, they were transferred to scintillation vials and 20 mL of Scintisafe 30% co cktail (SX23-5, Fisher Scientific) was added to the vial. The vials were mixed vigor ously and allowed to sit for a 2-3 hours before radioactivity was determined. Liquid scintillation counting was performed using in a Packard 1600 TR instrument. Each sample was counted for 5 minutes. The counts per minute (cpm) were converted to dpm as follows: dpm is cpm/(c ounter efficieny), where the counter efficiency is 0.6. The dpm per amount of protein filtered was calculated fo r the data plots. Specific binding was calculated as the difference between total binding (TB) and nonspecific binding (NSB).
120 Summation The focus of this dissertation has been on the use of S. cerevis iae as a model organism to study the human PAQR family of pr oteins. In particular, the f unction of the Class II PAQRs in sensing and responding to progestin steroids was examined. While we were able to demonstrate that the presence of prog estins causes a physiological effect in cells that express the Class II PAQRs, several questions remain regarding the mechanism of signaling initiated by the Class II PAQRs and whether this mechanism is cons erved for all PAQR family members. Here, to try to elucidate the way by which th e Class II PAQRs signal, it was proposed that novel protein-protein interactions could be explored using yeast as a tool. In addition, it was proposed that an intrinsic enzymatic activity could also be expl ored as a possibility. Although preliminary experiments have not been conducted for these experiments, preliminary work was performed for the expression of the Class II PAQRs in insect cells and to demonstrate binding of progesterone to these proteins. While different attempts were made to expre ss the Class II PAQRs in insect cells, these attempts yielded inconsistent success, perhaps due to inexperience working with this expression system. Because we saw expression of PAQR5 w ith this system, it could be very useful for future studies, including protein topo logy analysis and protein purification. In addition, while it is importa nt to demonstrate binding of pr ogesterone to the PAQRs, we had difficulty doing so. Whether the problems we had can be solved with the use of a better binding reaction filtering system will have to be determined once access to such a system is available.
121 1 2 3 Figure 5-1. Western blot of PAQR 5 and PAQR 6 expressed in Sf 9 insect cells. There is no band in the empty vector sample (lane 1) but a very stong band is visible in the sample for PAQR5 around 42 kDa) (lane 2) and a faint band is present in the sample for PAQR6 (lane 3) around the same size. The cells used for this expression were maintained by Ibon Garitaonandia.
122 A 0 200 400 600 800 1000 1200 1400 TB NSBdpm/ g of protein vector PAQR5 B 0 50 100 150 200 250 300 350 400 450 500 TB NSBdpm/ g of protein vector PAQR6 Figure 5-2. Binding of 3H-PG in a single-point binding assay with total membranes isolated from yeast cells expressing PAQR5 (panel A) or PAQR6 (panel B). Specific binding was calculated as the difference between total binding (TB) a nd nonspecific binding (NSB), as presented in Figure 5-2. Conditi ons used for binding are described in the Material and Methods.
123 0 50 100 150 200 250 300 350 400 vectorPAQR6specific binding (dpm/ g protein) Figure 5-3. Specific binding of 3H-PG in a single-point binding assay with total membranes isolated from yeast cells expressing PAQR 6. Specific binding was calculated as the difference between total bi nding and nonspecific binding. Table 5-1. Primers used for cloning P AQRs into pIEx-4 (listed from 5 3) Primer name Primer sequence Fwd PAQR5 CCAAGTGACCA TGGTGAGCCTGAAGCTC Rev PAQR5 AGAAGATGCGGCCGCTGTTTCTTTTTTATGTAATTCTG Fwd PAQR6 CCAAGTGACAT GTTCAGTCTCAAGCTGCC Rev PAQR6 AGAAGATGCGGCCGCCTGTTGTTTGGCCTGGGTAC
124 LIST OF REFERENCES Alberts B, Johnson A, Lewis J, Ra ff M, Roberts K and W alter P (2002) Molecular Biology of the Cell Garland Science, New York, NY. Altmann F, Staudacher E, Wilson IB and Marz L (1999) Insect cells as hosts for the expression of recombinant glycoproteins. Glycoconj J 16:109-123. Askwith C, Eide D, Van Ho A, Bernard PS, Li L, Davis-Kaplan S, Sipe DM and Kaplan J (1994) The FET3 gene of S. cerevisiae encodes a multicopper oxidase required for ferrous iron uptake. Cell 76:403-410. Baggett JJ, Shaw JD, Sciambi CJ, Watson HA a nd Wendland B (2003) Fluor escent labeling of yeast. Curr Protoc Cell Biol Chapter 4 :Unit 4 13. Baida GE and Kuzmin NP (1995) Cloning and prim ary structure of a new hemolysin gene from Bacillus cereus Biochim Biophys Acta 1264:151-154. Baida GE and Kuzmin NP (1996) Mechanism of action of hemolysin III from Bacillus cereus. Biochim Biophys Acta 1284:122-124. Baldi E, Casano R, Falsetti C, Krausz C, Ma ggi M and Forti G (1991) Intracellular calcium accumulation and responsiveness to progester one in capacitating human spermatozoa. J Androl 12:323-330. Beck CA, Weigel NL, Moyer ML, Nordeen SK and Edwards DP (1993) The progesterone antagonist RU486 acquires agonist activity upon stimulation of cAMP signaling pathways. Proc Natl Acad Sci U S A 90:4441-4445. Beukers MW and IJzerman AP (2 005) Techniques: how to boos t GPCR mutagenesis studies using yeast. Trends Pharmacol Sci 26 :533-539. Bielawska A, Greenberg MS, Perry D, Ja yadev S, Shayman JA, McKay C and Hannun YA (1996) (1 S,2 R )-Derythro-2-(N -Myristoylamino)-1-phenyl-1-pr opanol as an inhibitor of ceramidase. J Biol Chem 271:12646-12654. Blumer KJ, Reneke JE and Thorner J (1988) The STE2 gene product is the ligand-binding component of the -factor receptor of Saccharomyces cerevisiae J Biol Chem 263: 1083610842. Brauer AU, Nitsch R and Savaskan NE (2004) Id entification of m acrophage/microglia activation factor (MAF) associated with late en dosomes/lysosomes in microglial cells. FEBS Lett 563:41-48. Castilla JA, Gil T, Molina J, Hortas ML, Rodriguez F, TorresMunoz J, Vergara F and Herruzo AJ (1995) Undetectable expression of genomic progesterone receptor in human spermatozoa. Hum Reprod 10:1757-1760.
125 Chen HC and Farese RV (1999) Steroid hormones: Interactions with membrane-bound receptors. Curr Biol 9:R478-481. Chen YC, Chang MC, Chuang YC and Jeang CL (2004) Characterizati on and virulence of hemolysin III from Vibrio vulnificus. Curr Microbiol 49:175-179. Cherry JM, Adler C, Ball C, Chervitz SA, Dwi ght SS, Hester ET, Jia Y, Juvik G, Roe T, Schroeder M, Weng S and Botstein D (1998) SGD: Saccharomyces Genome Database. Nucleic Acids Res 26:73-79. Daley DO, Rapp M, Granseth E, Melen K, Drew D and von Heijne G (2005) Global topology analysis of the Escherichia coli inner membrane proteome. Science 308 :1321-1323. Deshaies RJ and Schekman R (1987) A yeast muta nt defective at an ea rly stage in import of secretory protein precursors into the endoplasmic reticulum. J Cell Biol 105 :633-645. Dohlman HG, Song J, Ma D, Courchesne WE and Thorner J (1996) Sst2, a negative regulator of pheromone signaling in the yeast Saccharomyces cerevisiae : expression, localization, and genetic interaction and physic al association with Gpa1 (t he G-protein alpha subunit). Mol Cell Biol 16 :5194-5209. Dosiou C, Hamilton AE, Pang Y, Overgaard MT, Tulac S, Dong J, Thomas P and Giudice LC (2008) Expression of membrane progestero ne receptors on huma n T lymphocytes and Jurkat cells and activation of G-proteins by progesterone. J Endocrinol 196:67-77. Dowell SJ and Brown AJ (2002) Yeast assa ys for G-protein-coupled receptors. Receptors Channels 8:343-352. Dressing GE and Thomas P (2007) Identification of membrane progestin receptors in human breast cancer cell lines and bi opsies and their potential involvement in breast cancer. Steroids 72 :111-116. Eide D and Guarente L (1992) Increased dosage of a transcriptional activator gene enhances iron-limited growth of Saccharomyces cerevisiae J Gen Microbiol 138 :347-354. Evaul K, Jamnongjit M, Bhagavath B and Hammes SR (2007) Testosterone and progesterone rapidly attenuate plasma membrane Gbetagamma-mediated signaling in Xenopus laevis oocytes by signaling through classical steroid receptors. Mol Endocrinol 21:186-196. Falkenstein E, Tillm ann HC, Christ M, Feuring M and Wehling M (2000) Multiple actions of steroid hormones--a focus on rapid, nongenomic effects. Pharmacol Rev 52:513-556. Falkenstein E and Wehling M (2000) Nongenomically initiated steroid actions. Eur J Clin Invest 30 Suppl 3:51-54. Farach-Carson MC and Davis PJ (2 003) Steroid hormone interactions with target cells: cross talk between membrane and nuclear pathways. J Pharmacol Exp Ther 307:839-845.
126 Feng L, Xie X, Ding Q, Luo X, He J, Fan F, Liu W, Wang Z and Chen Y (2007) Spatial regulation of Raf kinase signaling by RKTG. Proc Natl Acad Sci U S A 104:14348-14353. Fernandes MS, Brosens JJ and Gellersen B (2007) Honey, we need to talk about the membrane progestin receptors. Steroids Fernandes MS, Pierron V, Michalovich D, Astle S, Thornton S, Peltoketo H, Lam EW, Gellersen B, Huhtaniemi I, Allen J and Brosens JJ (2005) Regulated expression of putative membrane progestin receptor homologues in hu man endometrium and gestational tissues. J Endocrinol 187:89-101. Garitaonandia I (2008) Progestin and AdipoQ Receptors, Identification of PAQR3 as a New Possible Adiponectin Receptor, in Unpublished doctoral dissertation University of Florida, Gainesville. Gass EK, Leonhardt SA, Nordeen SK and Edwards DP (1998) The antagonists RU486 and ZK98299 stimulate progesterone r eceptor binding to deoxyribonucleic acid in vitro and in vivo, but have distinct eff ects on receptor conformation. Endocrinology 139 :1905-1919. Gitan RS and Eide D (2000) Zinc-regulated ubiq uitin conjugation signals endocytosis of the yeast ZRT1 zinc transporter. Biochem J 346: 329-336. Graham JD and Clarke CL (1997) Physiological action of progesterone in target tissues. Endocr Rev 18:502-519. Greenwald P, Barlow JJ, Nasca PC and Burn ett WS (1971) Vaginal cancer after maternal treatment with synthetic estrogens. N Engl J Med 285:390-392. Guarente L (1983) Yeast promoters and lacZ fusions designed to study expression of cloned genes in yeast. Methods Enzymol 101:181-191. Gulig PA, Bourdage KL and Starks AM (2005) Molecular Pathogenesis of Vibrio vulnificus J Microbiol 43 Spec No:118-131. Guo M, Aston C, Burchett SA, Dyke C, Fields S, Rajarao SJ, Uetz P, Wang Y, Young K and Dohlman HG (2003) The yeast G protein alpha subunit Gpa1 transmits a signal through an RNA binding effector protein Scp160. Mol Cell 12:517-524. Hanna R, Pang Y, Thomas P and Zhu Y (2006) Cell-surface expression, progestin binding, and rapid nongenomic signaling of zebra fish membrane progestin receptors alpha and beta in transfected cells. J Endocrinol 190 :247-260. Harashima T and Heitman J (2002) The Galpha pr otein Gpa2 controls yeast differentiation by interacting with kelch repeat prot eins that mimic Gbeta subunits. Mol Cell 10 :163-173. Heijne GV (1986) The distribution o f positively ch arged residues in bacterial inner membrane proteins correlates with th e trans-membrane topology. EMBO J 5:3021-3027.
127 Heim R, Cubitt AB and Tsien RY (1995) Improved green fluorescence. Nature 373:663-664. Herbst AL, Ulfelder H and Poskanzer DC (1971) Adenocarcinoma of the vagina. Association of maternal stilbestrol therapy w ith tumor appearance in young women. N Engl J Med 284:878-881. Huh WK, Falvo JV, Gerke LC, Carroll AS, Ho wson RW, Weissman JS and O'Shea EK (2003) Global analysis of protein localization in budding yeast. Nature 425:686-691. Iyer K, Burkle L, Auerbach D, Thaminy S, Di nkel M, Engels K and Stagljar I (2005) Utilizing the split-ubiquitin membrane yeast two-hybr id system to identify protein-protein interactions of integral membrane proteins. Sci. STKE 2005: pl3. Jansen G, Wu C, Schade B, Thomas DY and Wh iteway M (2005) Drag&Drop cloning in yeast. Gene 344:43-51. Josefsberg Ben-Yehoshua L, Lewellyn AL, Thom as P and Maller JL (2007) The role of Xenopus membrane progesterone receptor beta in medi ating the effect of progesterone on oocyte maturation. Mol Endocrinol 21:664-673. Kadowaki T and Yamauchi T (2005) Adiponectin and adiponectin receptors. Endocr Rev 26:439-451. Karpichev IV, Cornivelli L and Small GM (2002) Multiple regulatory roles of a novel Saccharomyces cerevisiae protein, encoded by YOL002c, in lipid and phosphate metabolism. J Biol Chem 277:19609-19617. Karteris E, Zervou S, Pang Y, Dong J, Hillhouse EW, Randeva HS and Thomas P (2006) Progesterone signaling in human myometri um through two novel membrane G proteincoupled receptors: potential role in functional progeste rone withdrawal at term. Mol Endocrinol 20:1519-1534. Kim H, Melen K, Osterberg M and von Heijn e G (2006) A global topology map of the Saccharomyces cerevisiae membrane proteome. Proc Natl Acad Sci U S A 103:1114211147. Kim H, Melen K and von Heijne G (2003) Topology models for 37 Saccharomyces cerevisiae membrane proteins based on C-terminal reporter fusions and predictions. J Biol Chem 278:10208-10213. Klichko VI, Miller J, Wu A, Popov SG and Alibek K (2003) Anaerobic induction of Bacillus anthracis hemolytic activity. Biochem Biophys Res Commun 303:855-862. Kreft SG, Wang L and Hochstrasser M (2006) Me m brane topology of the yeast endoplasmic reticulum-localized ubiquitin ligase Doa10 a nd comparison with its human ortholog TEB4 (MARCH-VI). J Biol Chem 281:4646-4653.
128 Krietsch T, Fernandes MS, Kero J, Losel R, He yens M, Lam EW, Huhtaniemi I, Brosens JJ and Gellersen B (2006) Human homologs of the putative G protein-coupled membrane progestin receptors (mPRalpha, beta, and ga mma) localize to the endoplasmic reticulum and are not activated by progesterone. Mol Endocrinol 20 :3146-3164. Kupchak BR (2008) The PAQR Family of Membrane Proteins in Saccharyomyces cerevisiae : Their Role in the Signal Transduction Pathway of FET3 and Flow Cytometric Analysis of AdipoR1 Binding to Adiponectin and TNF-Alpha in Unpublished doctoral dissertation University of Florida, Gainesville. Kupchak BR, Garitaonandia I, Villa NY, Mulle n MB, Weaver MG, Regalla LM, Kendall EA and Lyons TJ (2007) Probing the mechanism of FET3 repression by Izh2p overexpression. Biochim Biophys Acta 1773:1124-1132. Kyte J and Doolittle RF (1982) A simple method for displaying the hydropathic character of a protein. J Mol Biol 157 :105-132. Ladds G, Goddard A and Davey J (2005) Functi onal analysis of heterologous GPCR signalling pathways in yeast. Trends Biotechnol 23:367-373. Losel R and Wehling M (2003) Nongenom ic actions of steroid hormones. Nat Rev Mol Cell Biol 4:46-56. Lyons TJ, Villa NY, Regalla LM, Kupchak BR, Vagstad A and Eide DJ (2004) Metalloregulation of yeast membrane steroid receptor homologs. Proc Natl Acad Sci U S A 101:5506-5511. Maglich JM, Sluder A, Guan X, Shi Y, McK ee DD, Carrick K, Kamdar K, Willson TM and Moore JT (2001) Comparison of complete nuclear receptor sets from the human, Caenorhabditis elegans and Drosophila genomes. Genome Biology 2 :0029.1-0029.7. Mao C, Xu R, Bielawska A and Obeid LM ( 2000a) Cloning of an alkaline ceramidase from Saccharmomyces cerevisiae An enzyme with reverse (CoA-independent) ceramide synthase activity. J Biol Chem 275:6876-6884. Mao C, Xu R, Bielawska A, Szulc ZM and Obeid LM (2000b) Cloning and characterization of a Saccharmomyces cerevisiae alkaline ceramidase with spec ificity for dihydroceramide. J Biol Chem 275:31369-31378. Mao X, Kikani CK, Riojas RA, Langlais P, Wang L, Ramos FJ, Fang Q, Christ-Roberts CY, Hong JY, Kim RY, Liu F and Dong LQ (2006) A PPL1 binds to adiponectin receptors and mediates adiponectin signalling and function. Nat Cell Biol 8 :516-523. Martinez S, Pasten P, Suarez K, Garcia A, Nualart F, Montecino M, Hinrichs MV and Olate J (2007) Classical Xenopus laevis progesterone receptor associ ates to the p lasma membrane through its ligand-binding domain. J Cell Physiol 211:560-567.
129 McCue L, Thompson W, Carmack C, Ryan MP, Liu JS, Derbyshire V and Lawrence CE (2001) Phylogenetic footprinting of transcription fact or binding sites in prot eobacterial genomes. Nucleic Acids Res 29:774-782. McEwan IJ (2001) Bakers yeast rises to th e challenge: reconstitution of mammalian steroid receptor signalling in S. cerevisiae. Trends Genet 17:239-243. McIntire WE, Myung CS, MacCleery G, Wang Q and Garrison JC (2002) Reconstitution of G protein-coupled receptors w ith recombinant G protein alpha and beta gamma subunits. Methods Enzymol 343:372-393. Melcher K (2000) A modular set of prokaryot ic and eukaryotic expression vectors. Anal Biochem 277:109-120. Mentesana PE, Dosil M and Konopka JB (2002) Functional assays for mammalian G-proteincoupled receptors in yeast. Methods Enzymol 344:92-111. Milligan G (2003) Constitutive activity and inve rse agonists of G protein-coupled receptors: a current perspective. Mol Pharmacol 64:1271-1276. Miroux B and Walker JE (1996) Ov er-production of proteins in Escherichia coli : mutant hosts that allow synthesis of some membrane proteins and globul ar proteins at high levels. J Mol Biol 260:289-298. Morrill GA and Kostellow AB (1999) Progesterone induces meiotic division in the amphibian oocyte by releasing lipid second messengers from the plasma membrane. Steroids 64 :157167. Narasimhan ML, Coca MA, Jin J, Yamauchi T, Ito Y, Kadowaki T, Kim KK, Pardo JM, Damsz B, Hasegawa PM, Yun DJ and Bressan RA (2005) Osmotin is a homolog of mammalian adiponectin and controls apopt osis in yeast through a homol og of mammalian adiponectin receptor. Mol Cell 17 :171-180. Newbold RR, Padilla-Banks E and Jefferson WN (2006) Adverse effects of the model environmental estrogen diethylstilbestrol are transmitted to subsequent generations. Endocrinology 147:S11-17. Norman AW, Mizwicki MT and Norman DP (2004) Steroid-hormone ra pid actions, membrane receptors and a conformational ensemble model. Nat Rev Drug Discov 3:27-41. Oettel M and Mukhopadhyay AK (2004) Progester one: the forgotten hormone in men? Aging Male 7 :236-257. Pausch MH (1997) G-protein -coupled receptors in Saccharomyces cerevisiae : high-throughput screening assays for drug discovery. Trends Biotechnol 15:487-494. Pedelacq JD, Cabantous S, Tran T, Terwillig er TC and Waldo GS (2006) Engineering and characterization of a superfol der green fluorescent protein. Nat Biotechnol 24:79-88.
130 Pringle JR, Adams AE, Drubin DG and Haarer BK (1991) Immunofl uorescence methods for yeast. Methods Enzymol 194:565-602. Qume M (1999) Overview of ligand-receptor binding techniques. Methods Mol Biol 106:3-23. Rapp M, Granseth E, Seppala S and von Heijne G (2006) Identification and evolution of dualtopology membrane proteins. Nat Struct Mol Biol 13:112-116. Regalla, LM (2007) Characterization of the PAQR family of membrane proteins in Saccharomyces cerevisiae: examination of their role in metal and lipid metabolism and mass spectrometric-based analysis of signal transduction pathways, in Doctoral dissertation, University of Florida, Gainesville. Rehli M, Krause SW, Schwarzfischer L, Kreutz M and Andreesen R (19 95) Molecular cloning of a novel macrophage maturation-associated tr anscript encoding a protein with several potential transmembrane domains. Biochem Biophys Res Commun 217:661-667. Robertson KP, Smith CJ, Gough AM and Rocha ER (2006) Characterization of Bacteroides fragilis hemolysins and regulation and synergis tic interactions of HlyA and HlyB. Infect Immun 74:2304-2316. Rodriguez AM, Monjo M, Roca P and Palou A ( 2002) Opposite actions of testosterone and progesterone on UCP1 mRNA expressi on in cultured brown adipocytes. Cell Mol Life Sci 59:1714-1723. Sf A, Johansson M, Wallin E and von Heijn e G (1999) Divergent evolution of membrane protein topology: the Escherichia coli RnfA and RnfE homologues. Proc Natl Acad Sci U S A 96:8540-8544. Sambrook J and Russell DW (2001) Molecular Cloning: A Laboratory Manual Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Sengstag C (2000) Using SUC2-HIS 4C reporter domain to study t opology of membrane proteins in Saccharomyces cerevisiae. Methods Enzymol 327:175-190. Severance S, Chakraborty S and Kosman DJ (2004) The Ftr1p iron permease in the yeast plasma membrane: orientation, topology and structure-function relationships. Biochem J 380:487496. Shaner NC, Steinbach PA and Tsien RY (2005) A guide to choosing fluorescent proteins. Nat Methods 2 :905-909. Sikorski RS and Hieter P (1989) A system of shuttle vectors a nd yeast host strains designed for efficient manipulation of DNA in Saccharomyces cerevisiae Genetics 122 : 19-27. Slessareva JE and Dohlman HG (2006) G protein signaling in yeast: new components, new connections, new compartments. Science 314:1412-1413.
131 Sturgeon CM, Kemmer D, Anderson HJ and Roberge M (2006) Yeast as a tool to uncover the cellular targets of drugs. Biotechnol J 1:289-298. Suzuki T, Ganesh S, Agarwala KL, Morita R, Sugimoto Y, Inazawa J, Delgado-Escueta AV and Yamakawa K (2001) A novel gene in the ch romosomal region for juvenile myoclonic epilepsy on 6p12 encodes a brain-specif ic lysosomal membrane protein. Biochem Biophys Res Commun 288:626-636. Tang YT, Hu T, Arterburn M, Boyle B, Bright JM, Emtage PC and Funk WD (2005) PAQR proteins: a novel membrane receptor family defined by an ancient 7-transmembrane pass motif. J Mol Evol 61:372-380. Thomas P, Pang Y, Dong J, Groenen P, Kelder J, de Vlieg J, Zhu Y a nd Tubbs C (2007) Steroid and G protein binding characte ristics of the seatrout a nd human progestin membrane receptor alpha subtypes and their evolutionary origins. Endocrinology 148 :705-718. Thomas P, Zhu Y and Pace M (2002) Progestin me mbrane receptors involved in the meiotic maturation of teleost oocytes: a review with some new findings. Steroids 67 :511-517. Tiberi M and Caron MG ( 1994) High agonist-independent activity is a distinguishing feature of the dopamine D1B receptor subtype. J Biol Chem 269:27925-27931. Tokumoto T, Tokumoto M and Thomas P (2007) Interactions of diethylstilbestrol (DES) and DES analogs with membrane progestin recep tor-alpha and the correlation with their nongenomic progestin activities. Endocrinology 148 :3459-3467. Villa NY (2007) Characterization of novel yeast receptors implicated in metal and lipid metabolic pathways, in Doctoral dissertation, University of Florida, Gainesville. Villalobos A, Ness JE, Gustafsson C, Minshull J and Govindarajan S (2006) Gene Designer: a synthetic biology tool for constr ucting artificial DNA segments. BMC Bioinformatics 7:285-292. von Heijne G (1989) Control of topology and mo de of assembly of a polytopic membrane protein by positively charged residues. Nature 341 :456-458. Wehling M, Schultz A and Losel R (2007) To be or not to be (a receptor). Steroids 72 :107-110. Windh RT and Manning DR (2002) E xpression of G protein-coupled receptors and G proteins in Sf9 cells: analysis of coupling by radioligand binding. Methods Enzymol 343:417-429. Wooding S and Pelham HR (1998) The dynamics of golgi protein traffic visualized in living yeast cells. Mol Bio l Cell 9:2667-2680. Yamaguchi-Iwai Y, Dancis A and Klausner RD (1995) AFT1: a mediator of iron regulated transcriptional control in Saccharomyces cerevisiae. EMBO J 14:1231-1239.
132 Yamauchi T, Kamon J, Ito Y, Tsuchida A, Yokomi zo T, Kita S, Sugiyama T, Miyagishi M, Hara K, Tsunoda M, Murakami K, Ohteki T, Uc hida S, Takekawa S, Waki H, Tsuno NH, Shibata Y, Terauchi Y, Froguel P, Tobe K, Koyasu S, Taira K, Kitamura T, Shimizu T, Nagai R and Kadowaki T (2003) Cloning of adipon ectin receptors that mediate antidiabetic metabolic effects. Nature 423:762-769. Yamauchi T, Kamon J, Minokoshi Y, Ito Y, Waki H, Uchida S, Yamashita S, Noda M, Kita S, Ueki K, Eto K, Akanuma Y, Froguel P, Foufelle F, Ferre P, Carling D, Kimura S, Nagai R, Kahn BB and Kadowaki T (2002) Adiponectin stimulates glucose utilization and fatty-acid oxidation by activating AMP-activated protein kinase. Nat Med 8:1288-1295. Yang F, Moss LG and Phillips GN, Jr. (1996) Th e molecular structure of green fluorescent protein. Nat Biotechnol 14 :1246-1251. Zhang X, Lester RL and Dickson RC. (2004) Pil1p and Lsp1p negatively regulate the 3phosphoinositide-dependent protein kinase-like kinase Pkh1p and downstream signaling pathways Pkc1p and Ypk1p. J Biol Chem 279:22030-22038. Zhang YM, Marrakchi H and Rock CO (2002) The FabR (YijC) transcription factor regulates unsaturated fatty acid biosynthesis in Escherichia coli J Biol Chem 277:15558-15565. Zhou XC, Dowdy SC, Podratz KC and Jiang SW (2007) Epigenetic considerations for endometrial cancer prevention, diagnosis and treatment. Gynecol Oncol 107 :143-153. Zhu Y, Bond J and Thomas P (2003a) Identifica tion, classification, and pa rtial characterization of genes in humans and other vertebrate s homologous to a fish membrane progestin receptor. Proc Natl Acad Sci U S A 100:2237-2242. Zhu Y, Rice CD, Pang Y, Pace M and Thomas P (2003b) Cloning, expression, and characterization of a membrane progestin recep tor and evidence it is an intermediary in meiotic maturation of fish oocytes. Proc Natl Acad Sci U S A 100:2231-2236.
133 BIOGRAPHICAL SKETCH Jessica L. Sm ith was born and raised in a rural area of Western Pennsylvania. While pursuing her B.S. with a double major in biology and chemistry at the Clarion University of Pennsylvania, she developed an interest in molecular biology. She attended Texas A&M University for two semesters in the biochemistry graduate program and then transferred to the Department of Chemistry at the University of Fl orida, where she pursued her doctorate with Dr. Thomas Lyons. Jessica will join the research group of Dr. Vinay Pathak in the National Cancer Institutes HIV Drug Resistance Program to work as a postd octoral fellow and to pursue her interest in studying infectious diseases at the molecular level.