Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Characterization of 32b1, a partial cDNA encoding novel sequence having high similarity with the immediate… Chow, Cynthia Hsueh-Ying 1997

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1997-0223.pdf [ 6.2MB ]
JSON: 831-1.0099165.json
JSON-LD: 831-1.0099165-ld.json
RDF/XML (Pretty): 831-1.0099165-rdf.xml
RDF/JSON: 831-1.0099165-rdf.json
Turtle: 831-1.0099165-turtle.txt
N-Triples: 831-1.0099165-rdf-ntriples.txt
Original Record: 831-1.0099165-source.json
Full Text

Full Text

Characterization of 32b 1, a partial cDNA encoding novel sequence having high similarity with the immediate early gene, Cyr61 by CYNTHIA HSUEH-YING CHOW B.Sc. (Biochemistry), University of New Brunswick,1993 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTERS OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Biochemistry and Molecular Biology) We accept this thesis as conforming to the required standard The University of British Columbia April, 1997 © Cynthia Hsueh-Ying Chow, 1997 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication . of this thesis for financial gain shall not be allowed without my written permission. 1 Department of j^U&driMniM?^ %r C'lH0ff&a>li^  , The University of British Columbia Vancouver, Canada Date 31042-4-DE-6 (2788) A B S T R A C T The 32b 1 c D N A clone was isolated from a Xgtll library made from a Jurkat T lymphocyte cell line during a search for a transactivating factor for the H I V - L T R that was responsible for Ras responsiveness. The clone was 1.3kb in length and the full-length clone, postulated from Northern blot experiments done in human tissues, was estimated to be approximately 2.6 kb. This led to a comprehensive search of different libraries of human blood leukocytes, lung and brain cell lines to find the missing 5'end. Although the original segment of c D N A clone was isolated many times, novel sequence was never found experimentally. A search in the N C B I database revealed new 5' end sequence as wel l as evidence linking 32bl with wound repair. Further examination of the gene induction kinetics were carried out via Northern Hybridizaiton analysis. Unexpectedly, the gene was not found in peripheral blood leukocytes and none of the cell lines that the c D N A clone was isolated from was seen to express it. The only cell line that d id express it was a Rat-2 fibroblast cell line where a message having high sequence similarity with 32b 1 was seen to be induced by serum, EGF but not by the phorbol ester, P M A . The stability of the rat m R N A transcript was also studied from induction patterns after actinomycin D treatment with and without cycloheximide. It was found to have an transient half life of less than 30 minutes. The induction of the 32b 1 m R N A message was prolonged after treatment wi th cycloheximide which implicated that translation had an effect on its m R N A stability. These results allowed certain conclusions to be made about 32bl that can be a basis in further research into its function in cell proliferation. The focus of this research was to complete the c D N A ii sequence, find the missing 5' end of the cloned cDNA and to characterize the expression of the gene. i i i T A B L E OF C O N T E N T S Abstract i i Table of Contents iv List of Tables v i i i List of Figures ix List of Abbreviations x i Units of Measurement xiv Acknowledgements xv Literature Review 1 1. The discovery of 32bl 1 1. Cel l growth 2 2. Immediate Early Genes 7 3. C C N Family and its Characteristics 9 Materials and Methods 18 2.1 Vectors 18 2.2 Oligonucleotides 18 2.3 Manual and Automated D N A Sequencing 18 2.4 Library Screening 21 2.4.1 Touchdown PCR 21 2.4.2 D N A Fragment Purification 22 iv 2.4.3 Ligation and Digestion 23 2.5 Southern Blotting 24 2.6 Using the 5' R A C E System ...24 2.7 Cel l Culture and Mitogenic Stimulation 25 2.8 Extraction of R N A 27 2.9 Northern Blots 28 2.10 Radioactive Labelling of the Probe 29 Results 30 1. The 32bl Sequence a. Analysis of the 32bl sequence 30 b. Search for the 5' end 34 c. The discovery of novel 5' end sequence of 32bl from F A S T A search 36 2. Expression of 32bl m R N A 39 a. Tissue distribution of 32bl m R N A in Adul t Humans 39 b. Effect of Mitogenic Stimulation on the Expression of 32b 1 m R N A in different cell lines 41 3. Analysis of the Induction Kinetics for 32bl through Northern Hybridization....41 a. Time Course of 32bl induction in Rat-2 fibroblasts treated wi th Fetal Bovine Serum 43 b. Time Course of 32b 1 induction in Rat-2 fibroblasts treated with Phorbol 12-myristate 13-acetate (PMA) 43 c. Induction of 32b 1 m R N A in Rat-2 fibroblasts with V Epidermal Growth Factor (EGF) 46 4. Effect of Transcription and Translation Inhibitors on Stability of 32bl m R N A a. Time Course of decay of 32bl m R N A in Rat-2 fibroblasts treated with Actinomycin D 46 b. Time Course of 32b 1 decay in*Rat 2 cells treated with Actinomycin D and Cycloheximide 46 Discussion 51 1. What can the D N A sequence of the 32bl c D N A tell us about the gene products? 52 2. What does the novel 5' sequence tell us? 53 3. W h y was the search for the 5' end of 32b 1 unsuccessful? 55 4. What does the human tissue distribution of 32bl tell us when compared with other C C N family members? 57 5. Is 32bl the human homologue of cyr61? 59 6. What can be learned about the induction kinetics of 32blR after stimulation with various mitogens? 61 7. What is the effect of inhibition of protein synthesis and translation on the stability of the immediate early gene 32blR? 63 8. Is 32blR an immediate early gene? 64 9. What is the likelihood that 32bl is RBF-1? 65 10. A model for 32bl .65 11. Where do we go next? 67 vi Conclusion References LIST OF TABLES Table 1. Oligonucleotides designed for primer specific sequencing of 32b 1 Table 2. Tissue distribution of 32bl and various other C C N family members viii LIST OF FIGURES Figure 1. The cell cycle in a mammalian cell having a generation time of 16 hours Figure 2. Summary of two pathways initiated by EGF ligand binding of the receptor protein tyrosine kinase Figure 3. Characteristic features of the CCN family members Figure 4. Dendrogram of the amino acid sequence from representative members o the CCN family Figure 5. Protocol for the 5' RACE System for Rapid Amplification of cDNA ends Figure 6A. Map of the sequencing strategy for the 32b 1 cDNA Figure 6B. The partial cDNA sequence of the cloned 32bl cDNA Figure 7. Amino acid sequence similarity compared between 32bl and cyr61 Figure 8A. Representative DNA sequences showing high sequence similarity with 32b 1 found in a FAST A search Figure 8B. 32b 1 and new 5' end searched for new cut sites Figure 8C. The alignment of 32b 1 with the novel 5' end sequence found from the FASTA search and its translated amino acids Figure 9. Human Tissue Blot probed with 32bl Figure 10. Induction of 32b 1 Figure 11. Time Course of Rat- 2 fibroblasts treated with fetal bovine serum (FBS) and probed with 32b 1 Figure 12. Time Course of induction of 32b 1 RNA accumulation in Rat-2 fibroblasts treated with Phorbol 12-myristate 13-acetate (PMA) Figure 13. Time Course of Rat-2 fibroblasts treated with Epidermal Growth Factor (EGF) and probed with 32b 1 Figure 14A. A Time Course of decay of 32bl m R N A in Rat-2 fibroblasts following treatment with Actinomycin D Figure 14B. A Time Course observing 32bl m R N A hybridization patterns following treatment of Rat-2 fibroblasts with Actinomycin D and Cycloheximide Figure 15. Identified elements found in the 3'UTR of 32bl LIST OF ABBREVIATIONS A c t D Actinomycin D AIDS Acquired Immunodeficiency Syndrome BSA Bovine Serum Albumin C E F Chicken Embryo Fibroblasts C C N CEF-10/cyr61, C T G F and nov C T G F Connective Tissue Growth Factor D A G 1,2-Diacylglycerol depc diethyl-pyrocarbonate D M S O Dimethylsulfoxide DTT Dithiothreitol E C M Extracellular Matrix E D T A Disodium ethylene diamine tetraacetate EGF Epidermal Growth Factor EGFR Epidermal Growth Factor Receptor E R K 1 / E R K 2 Extracellular-signal-related kinases FBS Fetal Bovine Serum FGF Fibroblast Growth Factor fisp-12 fibroblast-inducible secreted protein-12 G D P Guanine diphosphate GTP Guanidine triphosphate L I S T O F A B B R E V I A T I O N S Hepes N-2-hydroxyethylpiperazine-Nl-2-Ethanesulfonic acid HIV-1 Human Immunodeficiency Virus type-1 H U V E Human umbilical vein endothelial cells M E K M A P K / E R K kinase M A P K Mitogen-Activated Protein Kinase M O P S 3-(N-morpholino)propanesulfonic acid M W Molecular Weight nov nephroblastoma-overexpressed gene O.D. Optical Density O L B Oligolabelling buffer P D G F Platelet Derived Growth Factor P C R Polymerase chain reaction P L C Phospholipase C P K C Protein Kinase C P M A Phorbol 12-myristate 13-acetate P V P Polyvinylpyrollidone RBE Ras-responsive binding element RBF Ras-responsive binding factor R T K Receptor Tyrosine Kinase SDS Sodium dodecyl sulfate SSC Standard saline citrate xii LIST OF A B B R E V I A T I O N S s s D N A Salmon sperm D N A TBE Tris borate E D T A T E M E D N i ^ i ^ i ^ i - t e t r a m e t h y l e t h y l e n e d i a m i n e TGF-(3 Transforming Growth Factor- beta T P A phorbol 12-tetradecanoate 13-acetate Tris Tris(hydroxymethyl)aminomethane tsg Twisted gastrulation U T R Untranslated Region xiii U N I T S O F M E A S U R E M E N T Ci curie cpm counts per minute Da Dalton dpm disintegrations per minute g gram k kilo 1 liter M Molar mol mole m milli (1 micro n nano p pico pfu plaque forming units % percent U units xiv A C K N O W L E D G E M E N T S Out of all the people who helped me to finish my Masters degree, the person who helped me the most was Dr George Mackie. I'd like to thank h im from the bottom of my heart for his time and effort in looking at the many revisions of my thesis and for creating a better situation for me to return to Vancouver for my defense. I am also very grateful to Dr. Michel Roberge who was extremely helpful with his revision suggestions and words of encouragement. It was wonderful to know that there were kind people who didn't mind taking time out of their busy schedules to help me. A s well , there are a great many other people that have made the path to my degree a little less rocky. I would like to thank Neena for teaching me how to do Northern Blots properly, for answering my never-ending questions and for lending me her book for over a year to help me write my thesis. I would also like to thank Elham for being such a big sister to me and for showing me all the ins and outs for making R N A . I would like to thank Wesley who was always wil l ing to help me out no matter how busy he was. I also want to tell Logan how grateful I am for all the delicious coffees and for going that extra mile in helping me when I really needed it. I want to thank John B. for his helpful advice and for being a good friend to me. I would like to thank Brendan for all the rides he gave me at 2 o'clock in the morning and for the enlightening conversations. I would like to thank Esther for her kindness and smiles that could always cheer me up. A s well , thanks to the rest of the people in the lab: Mario, John, Claudia, A m y and Martin, for each in their own way have done their share in helping me out. I haven't forgotten all the neat people in the department that helped make coming to work "fun". I wanted to thank Dr. Ivan Sadowski for giving me the opportunity to work in his lab. I would also like to thank the members of my defense committee, Dr. Louise Glass, Dr Michel Roberge and Dr Peter Candido, for taking time to come to my defense. Last but not least, I want to thank my family: M o m , Dad, Mike , A J and Ginger, for their love, patience, support and for never giving up on me. X V Literature Review 1. The discovery of 32bl The human immunodeficiency virus type-1 (HTV-1) has been previously identified as the causative agent for the Aquired Immunodeficiency Syndrome (AIDS). Within its nucleoid core, there are two single stranded RNA (ssRNA) molecules encoding its genome with a reverse transcriptase attached to one end of each. Once transmitted into the human host through bodily fluids, it finds its receptor, the CD4 molecule on the surface of Helper T Lymphocytes and monocytes and becomes internalized into the cell through binding and endocytosis. Here, the reverse transcriptase will transcribe the ssRNA molecules into double stranded DNA (dsDNA) which then migrates into the nucleus and integrates intself into the human host's genome as a latent provirus. It remains in this state for an undetermined amount of time before it is reactivated by the binding of cellular transcription factors to its promoter, the 5' Long Terminal Repeat (LTR). After this, the virus retranscribes its mRNA which then travels back into the cytoplasm where it is translated into viral proteins during assembly into a viable virion which then lyses the cell or buds out to infect other cells (Greene et al., 1991). Extensive research has been carried out in the area of cellular activation of HIV-1 from latency. Various cytokines have been found to effect HIV-1 gene expression both positively and negatively (Garcia et al., 1994). A cytokine activated pathway that has been implicated in the signal transduction of extracellular signals into the nucleus is the Ras/ERK pathway (Baldari et al., 1992). 1 Recently, four binding sites on the H I V - L T R have been defined as responsible for Ras responsiveness of the promoter, referred to as Ras-responsive binding elements I -IV (RBE I - IV) (Bell et al., unpublished). These sites are dispersed throughout the HIV-LTR and bind to two transcription factors, Ras responsive binding factors-1 and 2 (RBF-1 and RBF-2). This discovery led to Southwestern screening of a Jurkat T lymphocyte ^gtl l c D N A library using oligonucleotide A (comprising -159 to -138 of the H I V - L T R and encompassing RBE IV) to find the presumptive transcription factor, RBF-1, that bound to this site. This resulted in the isolation of six partial c D N A clones, the largest of which was 32bl. The c D N A fragment, 32bl, was cloned into the vector pGEM5fz(+) and analysis revealed that it contained novel D N A sequence. A sequence homology search uncovered resemblance to members of an immediate early gene family, referred to as C C N , named after members of the family. Immediate early genes are the primary genes induced during the transition from Go to the G i state of the cell cycle. 2. Cell growth In vivo or in culture, the cell is always found in one of two stages, either growing or quiescent. Growing cells are always increasing in mass and travelling through the cell cycle which eventually leads them to cell division. The cell cycle is categorized into four stages, seen in Figure 1: G i , the gap between nuclear division and D N A synthesis; S, the period of D N A synthesis; G2, the gap between D N A duplication and nuclear division, and M , the period of mitosis, during which the cell becomes two daughter cells (Darnell et al, 1990). 2 F i g u r e 1 . T h e c e l l c y c l e i n a m a m m a l i a n c e l l h a v i n g a g e n e r a t i o n t i m e o f 1 6 h o u r s . The cycle is made up of four stages: the Gl (the first gap) phase; the S (synthetic) phase, the G2 (second gap) phase and the M (Mitosis) phase (Darnell, 1990). 3 In contrast, quiescent cells are in a suspended state as they do not pass through the cell cycle nor do they increase in cell mass. They are in a state referred to as G 0 (see Figure 1). In cell culture, such cells can be induced to reinitiate growth by the addition of fresh serum. Serum is thought to supply the cells with trace amounts of growth factors and hormones which signal each cell to start events leading to cell growth and division. These ligands bind to specific receptors found on the cell suface to commence signal cascades leading to transcription of immediate early genes in the nucleus which mediate cell proliferation (Hil l and Triesman, 1995). Many growth factors, such as EGF and P D G F belong to a structurally diverse class of ligands that bind to receptors referred to as receptor tyrosine kinases (RTK). These specific receptors consist of an extracellular domain which binds the ligand, a single transmembrane domain, and an intracellular (cytoplasmic) domain with tyrosine kinase activity. RTKs are involved in many functions including the signaling of cell proliferation and differentiation depending on where they are expressed (Marshall, 1995). Figure 2 shows how these receptors link into two known pathways, the phosphatidylinositol signaling pathway and the Ras pathway (Berridge, 1993). Once a specific growth factor has bound to its RTK, the receptor becomes activated through dimerization at which time one receptor phosphorylates a tyrosine residue on the other in a process termed "autophosphorylation". This phosphorylated tyrosine acts as a docking site for phospholipase C-y l (PLC-yl) and adapter proteins, such as Grb2, which bind with high affinity through their src-homology-2 (SH2) domains. In the phosphatidylinositol pathway, the RTKs activate P L C - y l by phosphorylating it on specific residues which allow it to hydrolyse phosphatidylinositol 4 ( E G F ) Go G D P G i Ras GTP Grb-2 phosphatidylinositol 4,5-bisphosphate A Raf-l inositol 1,4,5-trisphosphate M E K M A P K 1,2-Diacylglycerol P K C | [Ca2+] Transcription factors Immediate Early Gene Expression Figure 2 Summary of two pathways initiated by EGF ligand binding of receptor protein tyrosine kinases. EGF has been found to initiate two pathways, phosphatidylinositol and ras signal transduction pathways. EGF, Epidermal Growth Factor; P L C , phopholipase C; P K C , Protein Kinase C; G D P , Guanine diphosphate; GTP, guanine triphosphate. 5 4,5-bisphosphate and create the products 1,2-diacylglycerol and inositol 1,4,5-trisphosphate. Inositol trisphosphate then plays a role in the increase of intracellular calcium (Ca^+) while the primary function of diacylglycerol is to activate a family of plasma membrane protein kinases called protein kinase C (PKC). P K C also requires Ca^+ ions which cause it to bind to the cytoplasmic side of the plasma membrane where it is then activated by D A G . P K C plays an important role in the mediation of growth in many cells (Darnell, 1990). P K C has also been found to phosphorylate the EGF receptor to decrease its affinity for EGF and thus down-regulates EGF's growth-stimulating ability. The tumor promoters, phorbol esters such as P M A activate P K C by acting as a D A G analog (Borner et a l , 1995). Through P K C , P M A regulates the activity of many transcription factors to alter the expression of cellular genes (Huang et al., 1995). Different P K C isoforms show different susceptibility to this treatment seen as a function of the cell type it is expressed in (Bi and Mamrack, 1994; Borner et al., 1995). Prolonged treatment with phorbol esters lead to eventual downregulation of the P K C pathway due to proteolysis (Bi and Mamrack, 1994; Borner et al., 1995). In the R a s / E R K M A P K pathway, the Grb2 protein which binds to the phosphotyrosine on the R T K is complexed with the Sosl protein, a common guanine exchanger through its src homology 3 (SH3) domains which have an affinity for proline rich sequences found on Sosl (Songyang and Cantley, 1995). The binding of Grb2 brings SoSl in close vicinity to Ras, a guanine nucleotide-binding protein localized on the inner membrane, allowing them to interact. SoSl activates the latter by catalysing exchange of the G D P bound to Ras with GTP (Figure 2). This GTP-bound Ras binds to the N-terminus of a Raf family member bringing it to the membrane. A t this point, it is 6 thought that an as yet unidentified protein activates Raf through phosphorylation. Raf then phosphorylates kinases such as M E K 1 / M E K 2 which in turn phosphorylate and activate E R K 1 / E R K 2 , types of M A P kinases. A fraction of these kinases translocates into the nucleus and phosphorylates transcription factors. Some M A P K s phosphorylate other protein kinases which then migrate into the nucleus and also phosphorylate transcription factors (Hunter, 1995). Another kinase referred to as M E K kinase ( M E K K ) also requires Ras for activation but works on the same level as c-Raf in a parallel cascade leading to M E K activation (Cano and Mahadevan, 1995; Crews and Erikson, 1993). 3. Immediate Early Genes As previously mentioned, the first genes induced as part of the rapid response through signaling cascades following stimulation with growth factors have been termed immediate early (IE) genes (Lau and Nathans, 1987). Their expression is rapidly activated within minutes to a few hours following the initial signal without requiring de novo protein synthesis. Immediate early genes have three defining characteristics: (1) many m R N A products of IE genes are nearly undetectable in quiescent cells but are usually present and functional at low levels allowing them to be rapidly accumulated within one hour of stimulation by growth factors (Kelly and Siebenlist, 1995; Lau and Nathans, 1991); (2) their induction occurs without de novo protein synthesis and (3) they are transcriptionally activated upon growth factor stimulation (Lau and Nathans, 1991). Once induced, immediate early gene expression is attenuated by rapid degradation and a cessation of transcription within a short time, causing the m R N A 7 expression to be transient (Lau and Nathans, 1987). This built-in negative feedback loop is carried out by the induced proteins encoded by the immediate early genes (Kelly and Siebenlist, 1995). These proteins downregulate the primary activating signalling pathways at different levels as seen with the activation of IKB-OC, the inhibitor of N F - K B immediately upon induction of the N F - K B gene (Kelly and Siebenlist, 1995). As well , this negative regulatory system is seen when FOS represses transcription of its own c-fos gene leading to message degradation (Herschman, 1991). These immediate early genes are also often superinduced in the presence of protein translation inhibitors such as cycloheximide (Simmons et al., 1989). This superinduction is thought to occur for two reasons. First, transcription is prevented from being shut off, and secondly, m R N A stability increases (Almendral et al, 1988). The expression of these early genes can lead to a variety of outcomes, depending on the distinct cellular environment that receives the signals (Kelly and Siebenlist, 1995). These responses may include differentiation, shape changes, locomotion and proliferation (Rozengurt, 1986). A subset of these genes may also be involved in regulation of complicated differentiation and wound healing processes involving cell division (Ryseck et al, 1991). Due to their numerous functions, immediate early genes can be divided into three categories according to their function and their pattern of induction upon stimulation by serum. The first set of genes, group I, is detectable within 5 minutes of growth factor stimulation, peak in concentration levels within 30 - 60 minutes and then become undetectable within approximately 2 hours. Identified genes of this group have been found to encode various transcription factors, such as c-fos and Jun-B (Lau and Nathans, 1991; Almendral et al, 1988). Group II m R N A s appear somewhat later and reach peak levels approximately two hours after 8 stimulation. These genes decay much slower and become undetectable between 4 to 8 hours after mitogen induction. These genes encode for proteins such as c-myc and JE. Finally, the last category, group III, have extremely long half-lives and encode many structural proteins (Lau and Nathans, 1991). 4. C C N Family and its characteristics In recent years, a group of immediate early genes containing related sequences has been discovered. Collectively, these are referred to as the " C C N " family, named after members of the family: CEF-10/ cyr61, CTGF and nov (Bork, 1993). The protein products of these genes contain a putative secretory signal at their amino termini suggesting that they are secreted proteins. The most prominent characteristic of this family is that each member contains 38 completely conserved cysteine residues, making up approximately 10% of the total sequence. These cysteines are dispersed throughout the amino and carboxy termini whereas the central region contains a divergent, acidic segment lacking any cysteines. Some members of the " C C N " family exhibit a new set of immediate early gene induction kinetics. Certain genes of this family accumulate their m R N A for as long as 8 hours but the m R N A itself only has a half life of 10 - 20 minutes (O'Brien et al., 1990) indicating that the m R N A is constantly transcribed. The members of this family have four conserved domains that have homology with other known proteins (Figure 3A). Domain 1 encodes a low-molecular-weight insulin-like growth factor-binding protein (IBP) consensus motif, G C G C C X X C , in the amino terminus of these family members. Insulin-like growth factor functions as a growth-stimulating peptide with mitogenic properties and travels in plasma tightly bound to 9 these specific IBPs (Brinkman et al., 1988). Domain II has identity wi th a von Willebrand factor type C repeat (VWC) region which is thought to participate in oligomerization. Domain III, first found in thrombospondin (type 1 repeat) (TSP1), has previously been involved in binding to both soluble and matrix macromolecules. Finally, domain IV, the C-terminal domain (CT module), has resemblance to the slit gene in Drosophila which may be involved in dimerization or receptor binding (Bork, 1993; X i n et al., 1996) . The family members are found in many species which implies that the members of this family play an important cellular function. Figure 3B shows a representative amino acid sequence alignment of the C C N family. Figure 4 shows a dendrogram displaying amino acid divergence among the family members. CEF-10 was the first member of this family to be discovered (Simmons et al., 1989). It was found to be induced at the permissive temperature in chicken embryo fibroblasts infected with the Rous sarcoma virus containing a temperature-sensitive mutation for v-src. CEF-10 was considered an immediate early gene because it exhibited characteristics similar to other IE genes such as expression as a rare transcript in control cells, increase of m R N A expression within an hour of mitogen stimulation, and superinduction in the presence of cycloheximide (Simmons et al., 1989). Comparison of the amino acid sequence of CEF-10 with a more recently found murine clone called cyr61 revealed approximately 80 per cent similarity suggesting that cyr61 may be a mouse homologue of CEF-10 (O'Brien et al., 1990). The cyr61 c D N A was isolated from quiescent mouse 3T3 fibroblasts induced with serum or P D G F . A n identical c D N A clone termed p I G - M l was also independently cloned from a c D N A library constructed wi th R N A extracted from a mouse embryo cell line that was treated 10 B cyrSl CEF10 FispL2 w Cyr61 CEFIO Fispl2 CTGF Nov Cyr61, CEFIO Fi.spl2 CTGF Nov Cyr61 CEFIO Fi.spl2 CTGF Nov Cyr61 CEF10 FLspl2 CTGF Nov Cyr61 CEFIO Fispl2 CTGF Nov K- -SSSTFRTLAVAVTIXHL- -TRLALST- CPAAC- -HCPLE - APKCAPGVCLVRDGCCCC1CVCAK 58 M - - C S A C A R P - A L A A A L L C L - - ARLALCSPCPAVC- -QCPAA-APQCAPGvCLVPDCCGCCKVCAK 58 H - - LAS VAC PIS LA - LVLLA LCTRPATCQDCS AQC - - Q C A A E A A P H C P A G V S L.VLDCCCCCRVCAK 61 M- -TAASMGPVRVAFWLLALCSRPAVCQNCSGPC- -RCPDEPAPRCPACVSLVLDGCGCCRVCAK 62 HETCCGQCLPVLLLLLLLLRPCEVSGREAACPRPCCCRCPAEP-PRCAPGVPAVIDGCGCCLVCAR 65 * * * * * * * * QLNEDCSKTQPCDHTKGLECNFGASSTALKGICRAQSEGRPCEYNSRIYQNGES FCJPNCKHQCTCI 124 QLNEDCSRTQPCDHTKCLECNFGAS PAATNGICRAQSECRPCEYNSKIYQNGESFQPNCKHQCTCI 124 QLCELCTERDPCDPHKCLF CDFGSPANRXIGVCTAK- DGAPCVFGGSVYRSCESFQSSCiCYQCTCL 126 QL.CELCTERDPCDPH1CGLFCDFGS PANRKICVCTAK-DCAPCIFCGTVYRSGES FQSSCKYQCTCL 127 QRCESCSPLLPCDESCGLYCDRGPEDCCCAGICHVL-EGDNCVFDCMIYRNGETFQPSCKYQCTCR 130 * • * ** * DGAVGCIPLCPQELSLPHLCCPNPRLVKVSCQCCEEWVCDEDSIK-DSLDD-QDDLLGLDASEVEL 188 DCAVCCIPLCPQELSLPNLCCPSPRLVKVPCQCCEEUVCDESKDALEELEGFFSKEFCLDASECEL 190 DCAVGCVPLCSHDVRLPSPDCPFPRRVKLPCKCCKEWVCDEPKDRTAVC? ALAAYRLED 185 DGAVGCMPLCSHDVRLPSPDCPFPRS.VKLPGKCCEEWVCDEPKDQTVVGP ALAAYRLED 186 DCQICCLPRCNLGLLLPCPDCPFPRKIEVPGECCEKyVCDPRDEVLLCGF AHAAYRQEA 189 . * * * TRNNELIAIGKGSSLKRLPVr GTEPRVLFNPLHAHGQKCIVQTTSWSQCSKSCGTGISTRVTNDNP 254 TRNNELIAIVKCC - LXMLPVFGSEPQ SRAFENPKCIVQTTSUSQCSKTCGTGISTRVTNDNP 251 T FGPDP TKHPJUiCLVQTTEWSACSKTCCMCISTRVTNDNT 225 T FGPDP- - TMIRANCLVQTTEUSACSKTCCMGISTRVTNDNA 226 T- - LGIDV SDSSANCIEQUEWSACSKSCCMCFSTRVTNRNQ 229 * * * * . * * * * ECRLVKETRICEVRPCCQP'/YSSLKKGKKCSKTKXSPEPVRFTYACCSSVIOCYRPKYCGSCTDGRC 320 DCKLIKETRICEVRPCGQPSYASIJCKCKKCTKTKKSPSPVRFTYAGCSSVKKYKP1CYCGSCVDGRC 317 FCRLEKQSRLCHVRPCEADLEENIKKGKKCIRTPKIAKPVKFELSGCTSVKTYRAKFCGVCTDCRC 291 SCRLEKQSR1-CMVRPCEADLEENIKKCKXCIRTPKISKPIKFELSGCTSHKTYRAKFCGVCTDGRC 292 QCEWKQTRLCHHRPCEN-EEPSDKKGKKCIQTKKSHKAVRFEYXNCTSVQTYKPRYCCLCNDGRC 294 * * * * * CTPLQTRTVKMRFRCEDGEMFSKNVMM- IQSCKC>rYNCPHPNEASFRLY- - SLFND1HKFRD 379 CTPQQTRTVKIP.FTICDDCETFTKSVHM- 1QSCRCNYNCPHANEA-YPFY- -RLVNDIHKFRD 376 CTPHRrTTLPVEreCPDCEIHKKN -MMFIKTCACHYNCPGDNDIFESLYYRKMYCDHA 348 CTPHRTTTLPVEFXCPOGEVHKKiJ -MMFIKTCACHYNCPGDNDIFESLYYRKMYGDMA 349 C T P H N T K T I Q V E F R C P Q C K F L K K P - M M L I N T C V C H C N C P Q S N N A F F Q P L D P M S S E A K I 351 Figure 3 . Characteristic features of the C C N family members. A, The four modules found in the CCN family showing sequence similarity with other proteins. The numbers indicate amino acid positions where the domains are found in cyr61. IBP, Insulin like growth factor binding protein, VWC, Von Willebrand factor type C repeat, TSP1, thrombospondin (type 1 repeat), CT, C-terminal domain. B, An amino acid alignment showing representative members of the CCN family. An asterisk is placed above each of the 38 conserved cysteines. Numbers at the right of the sequences refer to the last amino acid of that row. (O'Brien and Lau, 1992) 11 Xnov Nov NovH C T G F F i s p l 2 CeflO C y r 6 l Figure 4. Dendrogram of the amino acid sequences from representative members of the CCN family. This tree was generated using the PILEUP program and is a representation of the sequence similarity seen among the CCN family members reported from published papers (Ying and King, 1996) 12 with TGF-(3 and cycloheximide (Brunner et al., 1991). Their results showed that cyr61/ p I G - M l can be expressed when cells are stimulated to divide with serum, as wel l as when cells become growth arrested after treatment with TGF-p. Both the CEF-10 and cyr61 3' region contained at least two copies of the octanucleotide sequence T T A T T T A T found to confer instability in m R N A transcripts (Shaw and Kamen, 1986) as well as exhibiting similar induction kinetics after treatment with serum (Simmons et al., 1989; O ' Brien et al., 1990). Despite this, a significant difference in induction patterns was seen after treatment of the cell lines with the phorbol ester, phorbol 12-tetradecanoate 13-acetate (TPA). The m R N A for CEF-10 was immediately induced at time 0 but disappeared within one hour. Within two hours following initial exposure, the m R N A levels rebounded to show steadily increasing m R N A amounts beyond twenty-four hours. This time course correlated with the translocation, activation, and degradation pattern of protein kinase C (PKC) after activation suggesting that P K C may somehow repress the expression of CEF-10 (Simmons et al., 1989). Instead of being repressed like CEF-10, cyr61 was induced by T P A (O'Brien et a l , 1990). These results suggest that although cyr61 and CEF-10 are considered homologues, they may perform some different functions that cause dissimilar induction patterns after treatment wi th various mitogens. Further studies performed with cyr61 revealed that in actively growing cells, its protein was secreted and bound to both the extracellular matrix (ECM) and the cell surface wi th a half life of greater than 24 hours. Alternatively, in quiescent cells that were treated with serum to induce passage from Go to G i stage, cyr61 associates with the cell surface and has a short half life of 30 minutes (Yang et al., 1991). The cyr61 13 protein has similar functions to Wnt-1 (Yang et al., 1991). Like many growth factors, it has a chemotactic function but unlike growth factors, it also promotes cell attachment and spreading. From these and other results it was suggested that cyr61 may be an extracellular matrix signaling molecule. Other experiments also show that although it does not have detectable mitogenic activity on its own, it plays a role in enhancing the effect of other growth factors on fibroblasts and thus may play a regulatory role in cell proliferation and differentiation during development (Kireeva et al., 1996). Further clues to cyr61 function were seen when cyr61 was one of 12 genes isolated from murine Swiss 3T3 cells during a search for genes encoding critical inflammatory regulators which were attenuated by glucocorticoids as part of its anti-inflammatory effect. The product of these genes were thought to play a part in the mediation of intercellular interactions. The protein products of these genes were thought to function mostly in a paracrine manner and have functions that are intercellular or extracellular rather that within the cell (Smith and Herschman, 1995). During the isolation of (3IG-M1 (cyr61), a related gene was found to be induced simultaneously and was referred to as (3IG-M2 (Brunner et al., 1991). The same gene was isolated by another reseach group in serum-induced mouse fibroblasts and was referred to as fibroblast-inducible secreted protein-12 (fisp-12) (Ryseck et al., 1991). The gene's predicted protein has approximately 65% sequence identity to CEF-10/cyr61 (Ryseck et a l , 1991) and 94% similarity to Connective Tissue Growth Factor (CTGF) (Bradham et al., 1991). The CTGF was isolated from human umbilical vein endothelial cells (HUVE) as a secreted competitor for one of the platelet-derived growth factor (PDGF) receptors. Lke PDGF, CTGF functions as a mitoattractant in the endothelial 14 media but contains very little sequence homology with the P D G F dimers (Bradham et al., 1991). Further studies showed that in primary cultures of human foreskin fibroblasts, TGF-p induced autocrine production of C T G F which was not seen after stimulation with other growth factors. TGF-P must regulate C T G F through a unique pathway not used by the growth factors tested here: PDGF, EGF, FGF (Igarashi et al., 1993). C T G F has also been found to substantially increase during the response of scleroderma fibroblasts to TGF-p 1, a protein previously found to be important in scleroderma progression (Kikuchi, K . et a l , 1995). As well , C T G F distribution was increased in tissue sections from patients diagnosed with systemic and localized sclerosis as wel l as various other fibrotic skin disorders in comparison to normal tissue (Igarashi, A . et al., 1995; Igarashi, A et al', 1996). C T G F expression was also induced after scrape wounding of quiescent high density green monkey renal epithelial cells and found throughout the entire culture ( Pawar, S. et al., 1995). Calcium oxalate monohydrate crystal treatment of the same cells also stimulated C T G F expression. The presence of these crystals is a symptom of hyperoxaluria, a disease which lead to fibrosis, implicating C T G F as a causative agent (Hammes, M . et al., 1995). Together, the results from these experiments suggest that C T G F is an important factor in wound repair. Experiments done with cyr61 similar to those seen wi th C T G F imply that although they have 45% sequence identity, they act in completely different ways, C T G F as a growth factor and Cyr61 as an extracellular matrix signaling molecule (Kireeva et a l , 1996). A proto-oncogene was also found to be related to the " C C N " family. A n infectious retroviral strain of myeloblastosis-associated virus ( M A V ) type-1 was isolated 15 that induced formation of nephroblastomas after injection into chickens (Joliot et al., 1992). Upon progression of these polyclonal kidney tumors, a highly expressed novel gene called 'nov' (nephroblastoma-overexpressed gene) was discovered (Joliot et al., 1992). Unlike other members of the C C N family, the m R N A of this gene d id not contain the octanucleotide sequence (TTATTTAT) in its 3' U T R that promotes m R N A instability. Rather, it was found to associate with the extracellular matrix (ECM) (Perbal, 1994) and an amino terminus truncated form of nov was seen to induce morphological transformation in cells (Joliot et al., 1992). The human, mouse and Xenopus laevis orthologs have all been previously found and referred to as n o v H , n o v M and xnov respectively (Martinerie et al., 1994, Martinerie et al., 1996, Ying, Z . and M . King , 1996). The n o v H promoter is negatively regulated by WT-1 proteins and nov, as wel l as the other family members, are considered by this research group to be a novel type of insulin like growth factor binding protein (IBP) (Martinerie, C. et al., 1996, X i n et al, 1996; Kiefer et al., 1991). Previously identified IBPs did have similar features with the C C N family such as the 12 conserved cysteines within the first 100 base pairs which align wi th the members; as well , the middle domain lacked homology or cysteine residues and the last domain which also contained six conserved cysteine residues (Kiefer et a l , 1991). N o v H also participated in experiments similar to C T G F for characteristic comparative (Xin et al., 1996). Results indicated that their promoter sequences were completely lacking similarity with each other which suggested that the two genes were regulated in different ways (Xin et al, 1996). Evidence suggests that n o v H is a negative regulator of cell growth while CTGF has mitogenic properties (Xin et al., 1996) 16 A Drosophila gene encodes a protein referred to as "twisted gastrulation" (tsg) (Mason et al., 1994). It functions as one of seven genes that act as determinants to specify the fate of dorsal cells in Drosophila embryos. It contains a motif similar to that found in the C C N family where the cysteine residues can be completely aligned. It spans the region encompassing amino acid residues 26 to 78 in the C C N family members, which is in the same vicinity as domain I that has sequence similarity to insulin-like growth factor binding protein (IBP). The significance of this similarity is unknown. As well , domain II of cyr61 was shown to have sequence similarity with another Drosophila gene called short gastrulation which plays a role in dorsal patterning (Francois, V . et a l , 1994, Kireeva et a l , 1996). Experimental results seen with this family of immediate early genes suggest that although they have similar sequences and all seem to play a role in mediating cell differentiation, they have distinct patterns of induction and functions in different cellular environments. 17 Materials and Methods 2.1 Vectors The 32bl c D N A insert was cloned into the vector, pGEM-5Zf(+) (New England Biolabs) previous to my arrival in the lab. The vector contains 3003 bp and encodes a binding site for m l 3 forward sequencing primer, M13 reverse sequencing primer, T7 promoter and SP6 promoter. It also contains multiple cloning sites. The 32bl c D N A insert was cloned into the N O T I site. 2.2 Oligonucleotides Oligonucleotides were designed and synthesized for use in manual sequencing and the R A C E protocol (Further explained in Section 2.6 in Materials and Methods below). They are listed in Table 1 in the order of synthesis. 2.3 Manual and Automated DNA Sequencing Cloned c D N A s of interest were sequenced in both directions using a SequenaseTM kit (United States Biochemistry Corporation) based on the dideoxynucleotide chain termination method from Sanger et al. (1977). The enzyme used was a modified form of bacteriophage T7 D N A Polymerase (version 1.0). Potential cloned c D N A s were also sequenced using an Appl ied Biosystems (ABI) Mode l 373 Stretch Automated Sequencer found in the Nucleic A c i d Service Laboratory at U B C . 18 Table 1. Specific oligonucleotides made for Sequencing and RACE methods. The oligonucleotides are listed according to their order of synthesis. Orientation refers to the direction in which the primer was made with respect to the universal priming sites. c >liiy i n.iini' Orientation Use- Sequence CC1 Reverse Sequencing 5' TAG TAT TGG A G C A C A TGT TA 3' CC2 Forward Sequencing 5' A C A CCT GAT TTC TGT TTG A C A 3' CC3 Reverse Sequencing 5' GGG ATA TCA CTA GTG CGG CC 3' CC5 Reverse Sequencing 5' AGT ACT GCG GTT CCT GCG TG 3' CC6 Forward Sequencing 5' CTT A A A A C A CCT GAT TTC TG 3' CC7 Reverse Sequencing 5' CCT G A A GGG GGA C A C TCC AT 3' CC8 Forward Sequencing 5' A A G C A G TAA CAT GTG CTC C A A 3' CC9 Forward Sequencing 5' C A A CTC C A C A A G CTC C A A A A 3' GSP1 Reverse RACE 5' GGCGGCACTCAGGGTTGTCATTGGTAACTCG 3' GSP2 Reverse RACE 5' GGCCTTGTAAAGGGTTGTATAGGATGCGAGG 3' RACE 3' Forward RACE 5' C C T C G T G T A C A A T G A C G A A G T A A A A C C T C G 3' U A P Forward RACE 5' CTACTACTACTAGGCCACGCGTCGACTAGTAC 3' CC10 Forward Sequencing 5' G A A C A A A C A GAT C C A CAC GG 3' e c u Forward Sequencing 5' CCC A G C TCC CCC G A A TTC ATC 3' CC12 Reverse Sequencing 5' TGT GCC CTG G A A A C C C A G GT 3' CC13 Reverse Sequencing 5' AGT CGT ATT A A T TTC GAT A A 3' CC14 Reverse Sequencing 5' ATT TCT TGG TCT TGC TGC ATT T 3' 19 This method employs Taq D N A polymerase in ABI's Ampli taq DyeDeoxyTM Terminator Cycle Sequencing chemistry. The c D N A templates were prepared according to the protocol given wi th the 373 D N A Sequencer, "Preparation of plasmid DNA: A modified mini alkaline-lysis /PEG precipitation procedure." This procedure was followed by cycle sequencing on a PERKIN ELMER D N A Thermal Cycler according to the protocol with the 373 D N A Sequencer, "Sequencing dsDNAfrom plasmid using Taq Terminator chemistry", using 6 pmol of primer in each reaction. The tubes were placed in a thermal cycler preheated to 96 ° C and the following cycle was carried out: Rapid thermal ramp to 96 °C , 96 ° C for 15 sec Rapid thermal ramp to 50 °C , 50 ° C for 1 sec, Rapid thermal ramp to 60 °C , 60 °C for 4 minutes, 25 cycles total. Rapid thermal ramp to 4 ° C and hold. Before the sample was loaded on the gel, it was extracted with phenol/chloroform according to the 373 Sequencer protocol to remove excess, unincorporated dideoxy terminators. The pellet was applied to a polyacrylamide gel using the Appl ied Biosystems Automated D N A Sequencer in the N A P S unit. A hard copy or disk copy of the D N A sequence was returned for further analysis. Some D N A sequencing was completely performed by the Nucleic A c i d Service Laboratory at U B C after submission of the miniprep D N A and specific primers. Nucleotide and amino acid sequence analyses were carried out using Dnasis Pro Version 3.2. 20 Sequencing done on the LICOR automated synthesizer followed procedures in the protocol "Thermo Sequenase fluorescent labelled primer cycle sequencing kit with 7-deaza-dGTP" (Amersham Life Science). 2.4 Library Screening 2.4.1 Touchdown P C R A X g t l l c D N A library of a U937 monocyte cell line treated with P M A for 3.5 days and an additional A,gtll Jurkat T cell library (Clontech) as well as libraries (from I. Sadowski or R. MacGillivray) derived from the Hela cell line, human lung tissue and human brain tissue were searched. The method used for screening required annealing temperatures from 65 ° C to 55 ° C in the "touchdown" PCR reactions as described by Don et al (1991). The template was first denatured for 5 minutes at 94 ° C and then 2 cycles at each annealing temperature were carried out for 1 min with an immediate denaturing at 94 ° C (1 min) before and extension at 72 ° C (2 min) afterwards. A t the final annealing temperature, 20 cycles were carried out. The reaction finished with a 10 minute extension at 72°C. Each screening used a 1:10 dilution of the library (stock: 1-9 x 109 p f u / m l when the concentration was known) and was amplified using Vent D N A polymerase (New England Biolabs). Each PCR reaction contained 1 x Vent polymerase buffer ( lOmM KC1, l O m M (NH4)2SC>4, 20mM Tr is -HCl (pH8.8,@ 25 °C), 2 m M MgSC>4, 0.1% Triton X-100) , 30mM dNTPs, lOOmM of each primer, lOOmM MgSC>4, 100 L i g / m l 21 BSA, 2 U vent polymerase. Each reaction used a vector-directed primer found on either side of the cloning site (A,gtll arms forward or reverse) and a gene specific primer. 2.4.2 D N A fragment purification Upon completion of each reaction, the PCR products were separated by electrophoresis on a 0.7%-l% agarose gel. Distinct bands were purified on a N A 4 5 D E A E membrane (Schleicher & Schuell) (S&S NA45) using a method based on an application update from Schleicher & Schuell. Once the desired sized band was specified, a slit was placed before and after the band into the agarose. The N A 4 5 D E A E membrane was placed into these slits using sterile tweezers. The gel was then placed back into the gel apparatus and run for 5 minutes at 150V or 20 minutes at 120V. The N A 4 5 D E A E membrane below the band into which the D N A had run was placed in a microfuge tube and 300|il of N E T buffer was added. The tube was incubated at 65 ° C for 30 minutes. A n y ethidium bromide was extracted with 750 ui water-saturated n-butanol by vortexing and brief centrifugation. Following this, the membrane was removed from the tube using tweezers. After centrifugation, the solution in the tube separated into two layers, the lower layer containing D N A and the N E T buffer. This lower layer was transferred to a new microfuge tube and the D N A was precipitated by addition of 750 ul of 95% ethanol and placement at -70 ° C for 30 minutes or -20 ° C overnight. The precipitated D N A was then pelleted via centrifugation for 10 minutes at high speed. The pellet was washed with 70% ethanol and dried under vacuum. The pellet was resuspended in 20 Ltl of distilled water. A small amount of the D N A solution 22 (usually 2 ul) was applied and run on an agarose gel for estimation of the amount of D N A recovered. 2.4.3 Ligation and Digestion Purified "blunt-end" D N A fragments were ligated into a vector in a reaction containing: the treated vector (cut with either Hinc II or Sma I and dephosphorylated) (Maniatis et al, 1992), the purified fragment (2 ul of a concentrated D N A miniprep), T4 D N A ligase (1 (il, New England Biolabs), l x ligase buffer (supplied by the manufacturer containing 50mM Tr is -HCl (pH 7.8), 10 m M MgCl2 ,10 m M dithiothreitol (DTT), I m M A T P , 50 jig/ml BSA) and 4% P E G 8000 (Sigma). Following an overnight ligation at 15 ° C , the reaction was transformed into competent D H 5 a cells and grown overnight on an L B ampicillin plate. Drug resistant colonies were picked and then the plasmid D N A was extracted according a protocol found in Maniatis et al (1992). Plasmids containing the desired insert were found by restriction enzyme digestion on either side of the cloning site. Digestions contained 1 ul (20U) of each enzyme in appropriate l x buffer (supplied by the manufacturer) with the miniprep D N A (2 ul). The reaction was then incubated at the recommended temperature for at least 1 hour before electrophoresis on a 0.7-1% w:v agarose gel. Each gel was run in l x TBE (0.89 M Tris, 0.89 M boric acid, 1 m M E D T A (pH 8.0)) containing 0.1 ( ig /ml ethidium bromide. Gels were run by electrophoresis at 10V/cm with a size marker, X D N A digested with Hind III (Sigma). After separation, bands were detected by U V illumination and photographed. D N A sequence of positive clones was determined allowing further analysis to be carried out. 23 2.5 Southern Blots P C R products and digests of putative cloned c D N A s were separated on a 0.7% w:v agarose gel (as described in 2.4.3) and then transferred onto a piece of nylon membrane (Zeta-Probe GT Genomic Tested Blotting Membrane; BIO-RAD) using a vacuum blotter, trans-Vac TE 80 (Hoefer Scientific Instruments). The method employed for transfer was specified by the manufacturer. Each blot was then exposed to 300 mj U V irradiation in a GS Gene L i n k e r ™ U V Chamber (BIO-RAD) to crosslink the D N A to the blot. Blots were prehybridized for 4 hours in hybridization solution (50% formamide, 6X SSC, 0.1% polyvinylpyrollidone, 0.1% ficoll 400), 50mM sodium phosphate (pH 6.5), 0.5 % sodium dodecylsulfate (SDS), 100 Hg /ml salmon sperm D N A (ssDNA)) at 42 °C . The s sDNA was denatured in 0.15 M N a O H at 37 ° C for 5 minutes and neutralized with one equivalent of acetic acid immediately prior to addition to the hybridization solution. Following prehybridization, overnight hybridization was carried out in the same conditions after a change to 20 m l of fresh solution. The probe was added to the s s D N A to be denatured simultaneously before addition to the hybridization solution. The blot was washed in approximately 250 m l of 0.1 X SSC, 0.1% SDS for three washes at 30 minutes each at 50 ° C . The blot was then wrapped in saran wrap and exposed to film (Kodak Scientific Imaging F i lm X - O M A T A R 8X10). 2.6 Us ing the 5' R A C E System A n alternative method was used to search for the missing fragment of the 32bl c D N A clone using a R A C E protocol. A kit called the 5' RACE System for Rapid 24 Amplification of cDNA ends was purchased from GIBCO BRL. An overview of its protocol is shown in figure 5. The method used for screening followed the protocol given with the kit. The RNA used in this procedure was extracted using method described in section 2.8. 2.7 Cell Culture and Mitogenic Stimulation The Jurkat T cell line, U937 promonocytic cell line (human histiocytic lymphoma) and Hela (epitheloid carcinoma of the cervix) cell line were cultured in RPMI-1640 medium (with L-glutamine and without sodium bicarbonate; Gibco BRL). HL-60 cells (promyelocytic leukemia cell line from peripheral blood) were grown in Iscove's modified Dulbecco's medium (IMDM with L-glutamine, with 25mM Hepes buffer, without sodium bicarbonate; Gibco BRL). The Rat-2 fibroblast cell line was grown in Dulbecco's modified Eagle medium (D-MEM with low glucose, with L-glutamine, with 110 mg/L sodium pyruvate, without sodium bicarbonate). Each medium used was supplemented with 10% fetal bovine serum (FBS; Gibco BRL) and 1% streptomycin (10000 Lig/ml) and penicillin (10000 U/ml) (Gibco BRL). Before induction, the Rat 2 fibroblasts and CNA7 cells were rendered quiescent through density arrest as described by Jahner et al (1991). The various stimulants were used in the following final concentrations: epidermal growth factor (EGF; Calbiochem, receptor grade), 20ng/ml; phorbol 12-myristate 13-acetate (PMA; Gibco BRL), lOOng/ml; fetal bovine serum (FBS; Gibco BRL), 20%; actinomycin D (Calbiochem Corporation), l|ig/ml; cycloheximide (Sigma), 10|ig/ml. The agents were dissolved in media and then added directly to the cell culture. 25 mRNA 5' (A)n-3' <=iGSP1 ' Anneal first strand primer, GSP1, to mRNA. cDNA Copy mRNA into cDNA •(A)n"3 with S U P E R S C R I P T II RT. 3'-(dC)n Anchor Primer 3 '-(dC) n-\ <=GSP2 Degrade mRNA template • ( A ) n " 3 ' with RNase H . Purify cDNA with G L A S S M A X Spin Cartridge Tail purified cDNA with dCTP and TdT. Amplify dC-tailed cDNA by PCR using the Anchor Primer and nested GSP2. PCR product ready for: agarose gel electrophoresis/size enrichment nested primer amplifications U D G cloning into vector pAMP1 Figure 5. Protocol for the 5' Race System for Rapid Ampl i f ica t ion of c D N A ends. This procedure is based on reverse transcription of a sample of R N A from a specific cell line shown to express the desired message. A gene specific primer (GSP1) is used as a start site for the reverse transcriptase and once completed, the c D N A fragment is purified and a 3' poly (dC) tail is added. The fragment is then amplified using polymerase chain reaction (PCR) wi th an anchor primer with affinity for the d C tail. From here, theoretically, one can clone and sequence the 5' fragment and religate it to the original clone, (figure from protocol with R A C E kit) 26 2.8 Extraction of R N A Total R N A was isolated from the various cell lines according to the protocol included with the "Ultraspec R N A Isolation System" (I. D . Lab Inc.)- The method is based on the procedure described in Chomczynski et al (1987). Caution was used during each extraction procedure. Gloves were worn at all times and separate freshly autoclaved tips and tubes were used for R N A only. The solutions used, (eg. the deionized distilled water) were treated with diethyl-pyrocarbonate (depc) and autoclaved. Tissue culture cells were grown in the specified media for several days until a total cell count reached 5-10 mill ion cel ls /ml for cells grown in suspension or 1.5 to 2.3 mil l ion cells per plate of monolayer cells. Cells grown in suspension were sedimented by centrifugation at 12,000 x g for 5 minutes. The cell pellet was suspended in 1 m l of Ultraspec R N A solution and transferred to a 1.5 m l tube. For plated cells, the growth medium was discarded, 1 ml of Ultraspec R N A solution was added, the cells were scraped off the plate and the cell suspension was transferred to a 1.5 m l microfuge tube. After homogenization, the cells were stored at 4 ° C for 5 minutes to allow complete dissociation of nucleoprotein complexes. 0.2 ml of chloroform was added per m l of Ultraspec R N A solution used. The mixture was vigorously mixed for at least 15 seconds, stored on ice at 4 ° C for 5 minutes, and centrifuged at 12,000 x g at 4 ° C for 15 minutes. The upper aqueous phase containing the R N A was transferred to a fresh tube. A n equal amount of isopropanol was added and the samples were stored for no more than 10 minutes at 4 °C . Tubes were then centrifuged at 12,000 x g at 4 ° C for 10 minutes. The R N A formed at the bottom of the tube as a white pellet. The supernatant was removed and the pellet covered with 1 m l of cold 70% ethanol followed by 27 centrifugation at high speed for 5 minutes at 4 °C . The wash solution was then carefully removed using a sterile tip and the tube left open to air dry for 10 minutes. Pellets were resuspended in 50 ul of depc-treated ddEbO and the R N A concentration of each sample was quantified by U V spectroscopy and stored at -70 °C . 2.9 Northern Blots Samples containing 20ug of R N A were separated by electrophoresis in a 1% agarose gel containing 5% formaldehyde and l x M O P S E D T A buffer (20mM M O P S p H 7.0, 5 m M sodium acetate, 1 m M E D T A ; depc-treated and autoclaved). Each R N A sample was adjusted to a volume of 5 LLI with depc-treated deionized distilled water and 25 ul of R N A loading buffer (56% formamide, l x M O P S E D T A buffer, 2.2 M formaldehyde, 7.5% glycerol, 0.3% bromophenol blue) was added. The samples were heated at 70 ° C for 15 minutes before addition of 1 ul of ethidium bromide (1 mg/ml) . The samples were then loaded and the gel was run from 50V to 100V in l x M O P S -E D T A buffer until the bromophenol blue dye was approximately 1 cm from the bottom. Visibili ty on a U V box after electrophoresis ensured intact R N A . Overnight transfer onto nylon paper (Zeta-Probe GT Genomic Tested Blotting Membrane; BIO-RAD) was done using capillary action in 10X SSC. The membrane was irradiated wi th U V and hybridized as for the Southern blot explained previously. Following detection of a message, the Northern blots were also stripped by immersing the nylon membrane in approximately 500 m l of boiling 0.1% SDS solution for 20 minutes. Complete stripping was ensured by reexposure of the blot to film 28 overnight to detect any remaining signal. The stripped blots were then reprobed with G A P D H , a control, to ensure equal loading of R N A . 2.10 Radioactive Labelling of the Probe D N A to be labelled was purified by electrophoresis on a 0.7% agarose gel and the appropriate band cut out. The gel slice was placed inside an autoclaved 0.7 ml tube containing a plug of siliconized glass wool. A hole was pierced through the bottom of the tube using a 21G needle. This smaller tube was placed within a larger 1.5 m l tube and the assembly spun at high speed for 5 minutes. The liquid that flowed through to the larger tube contained template D N A for labelling. Randomly labelled probes were used in the Northern and Southern hybridizations. Approximately 600 ng (estimated by visualization on a gel stained with ethidium bromide under U V light) of D N A was added to distilled water to a final volume of 68(il. The D N A was denatured at 100 ° C for 10 minutes and then immediately placed on ice for 1 to 5 minutes. The following reagents were added: 20 ul oligolabelling buffer (Pharmacia), 10 ul [a 3 2 P ] - d C T P label (100 uCi), 3 i l l D N A polymerase I (Klenow fragment) (3.3U/ui). This reaction was placed in a 37°C water bath for at least 1 hour. The probe was then precipitated at -20 ° C for 4 to 24 hours after addition of an equal volume of 5 M ammonium acetate and 4 volumes of 95% ethanol at -20 ° C . The labelled probe was recovered by centrifugation at high speed in a microfuge for 10 minutes and the supernatant removed. The pellet was resuspended in 50 ui dH.20 and 1 ul was counted to ensure adequate labelling. If the specific activity was judged to be too low, the probe was relabelled by repeating the above protocol. 29 Results I. The 32b 1 sequence a. Analysis of the 32bl sequence The 32bl insert was cloned into the pGem5fz(+) vector prior to my arrival in the lab. It was partially sequenced by another technician but the entire sequence was repeated and checked for any discrepancies. The majority of the 32bl clone was sequenced manually but sections were also sequenced using a L I C O R automated sequencer for increased accuracy. Figure 6A is a map showing the sequencing strategy used (refer to sections 2.2 and 2.3 in Materials and Methods for primer sequences and sequencing methods). The insert in the 32bl c D N A clone was largely sequenced in both directions. A section of D N A at the 5' end spanning 245 bp as wel l as a small section at the 3' end were sequenced numerous times, but only in one direction. Because of the numerous times it was repeated, the D N A sequence is most likely accurate. Also, careful inspection of chromatograms from the automated sequencers ensured that only clear sharp peaks were used in the sequencing anaylsis. Figure 6B shows the D N A sequence of 32bl translated into its amino acid sequence. Another indication of the accuracy of 32b 1 was the ability to translate the D N A sequence without encountering stop codons in the coding region which resulted in 91% similarity with the amino acid sequence of cyr61 (Figure 7). The sequence of the 32bl c D N A contains an open reading frame (ORF) encoding a predicted protein with cysteines at positions characteristic of the " C C N " family 30 m13rev.5r2a 32b1Ra cynrev2 5R.CC3A 32b1rev frebl 1 131 245 388 500 32b1for cynfor2 nfrea cc6.5f6ha CCl.5Ra1 32b1Fa cc7.5r2 762 871 1,029 1,263 I s yaws* I Multiple fragments same direction j He! Both strands I 1*3 Both strands plus I i c' f ' Start codon frame 1 i 1 . <->-Stop codon frame 2 Bumps on fragments show mot i f s , hollow rectangles show features! Figure 6A Map of the sequencing strategy for the 32bl cDNA. Manual and automated sequencing techniques were used to sequence the cloned 32bl cDNA. In addition to ml3univ and ml3rev primers binding to the vector, specific oligonucleotides were designed and synthesized to sequence the rest (refer to Materials and Methods section 2.2 and 2.3 for oligonucleotide sequences and sequencing procedures). 31 1 AATTCGGGGGAGCTGGGATTCGATGCCTCCGAGGTGGAGTTGACGAGAAACAATGAATTG 1 AsnSerGlyGluLeuGlyPheAspAlaSerGluValGluLeuThrArgAsnAsnGluLeu 59 20 60 ATTGCAGTTGGAAAAGGCAGCTCACTGAAGCGGCTCCCTGTTTTTGGAATGGAGCCTCGC 119 21 IleAlaValGlyLysGlySerSerLeuLysArgLeuProValPheGlyMetGluProArg 4 0 12 0 ATCCTATACAACCCTTTACAAGGCCAGAAATGTATTGTTCAAACAACTTCATGGTCCCAG 17 9 41 IleLeuTyrAsnProLeuGlnGlyGlnLysCysIleValGlnThrThrSerTrpSerGln 60 18 0 TGCTCAAAGACCTGTGGAACTGGTATCTCCACACGAGTTACCAATGACAACCCTGAGTGC 239 61 CysSerLysThrCysGlyThrGlylleSerThrArgValThrAsnAspAsnProGluCys 8 0 24 0 CGCCTTGTGAAAGAAACCCGGATTTGTGAGGTGCGGCCTTGTGGACAGCCAGTGTACAGC 2 99 81 ArgLeuValLysGluThrArglleCysGluValArgProCysGlyGlnProValTyrSer 100 300 AGCCTGAAAAAGGGCAAGAAATGCAGCAAGACCAAGAAATCCCCCGAACCAGTCAGGTTT 35 9 101 SerLeuLysLysGlyLysLysCysSerLysThrLysLysSerProGluProValArgPhe 12 0 360 ACTTACGCTGGATGTTTGAGTGTGAAGAAATACCGGCCCAAGTACTGCGGTTCCTGCGTG 419 121 ThrTyrAlaGlyCysLeuSerValLysLysTyrArgProLysTyrCysGlySerCysVal 14 0 4 20 GACGGCCGATGCTGCACGCCTCAGCTGACCAGGACTGTGAAGATGCGGTTCCGATGCGAA 4 7 9 141 AspGlyArgCysCysThrProGlnLeuThrArgThrValLysMetArgPheArgCysGlu 160 4 80 GATGGGGAGACATTTTCCAAGAACGTCATGATGATCCAGTCCTGCAAATGCAACTACAAC 539 161 AspGlyGluThrPheSerLysAsnValMetMetlleGlnSerCysLysCysAsnTyrAsn 18 0 54 0 TGCCCGCATGCCAATGAAGCAGCGTTTCCCTTCTACAGGCTGTTCAATGACATTCACAAA 5 99 181 CysProHisAlaAsnGluAlaAlaPheProPheTyrArgLeuPheAsnAspIleHisLys 200 600 TTTAGGGACTAAATTGCCTACCTGGGTTTCCAGGGCACACCTAGACAAACAAGGGAGAAG 65 9 201 PheArgAsp*** 203 660 AGTGTCAGAATCAGAATCATGGAGAAAATGGGCGGGGGTGGTGTGGGTGATGGGACTCAT 719 720 TGTAGAAAGGAAGCCTTGCTCATTCTTGAGGAGCATTAAGGTATTTCGAAACTGCCAAGG 77 9 7 8 0 GTGCTGGTGCGGATGGACACTAATGCAGCCACGATTGGAGAATACTTGCTTCATAGTATT 83 9 84 0 GGAGCACATGTTACTGCTTCATTTTGGAGCTTGTGGAGTTGATGACTTCTGTTTTCTGTT 8 99 900 TGTAAATTATTTGCTAAGCATATTTTCTCTAGGCTTTTTTCCTTTTGGGGTTCTACAGTC 959 960 GTAAAAGAGATAATAAGATTAGTTGGACAGTTTAAAGCTTTTATTCGTCCTTTGACAAAA 1019 102 0 GTAAATGGGAGGGCATTCCATCCCTTCCTGAAGGGGGACACTCCATGGAGTGTCTGTGAG 107 9 1080 ACGGCAGTCTATCTGCACTCTAAACTAAACTGCAAACAGAAATCAGGTGTTTTAAGACTG 113 9 114 0 AATGTTTTATTTATCAAAATGTAGCTTTTGGGGAGGGAGGGGAAATGTAATACTGGAATA 1199 1200 ATTTGTAAATGATTTTAATTTTATATCAGTGAAAAGATTTTATTTATGGAATTAACCATT 125 9 12 60 TAATACCCGAATTCC 127 4 Figure 6B. The partial c D N A sequence of the cloned 32bl c D N A . This sequence encompasses 1.3kb. The open reading frame was found after comparison to the cyr61 D N A coding sequence and careful inspection of experimental data. 32 SCORES I n i t l : 820 I n i t n : 985 Opt: 9 92 91.0% i d e n t i t y i n 201 aa overlap + 10 20 30 20a022 NSGELGFDASEVELTRNNELIAVGKGSSLKRLPV I h l l l l l l l l l l i l l l h l l l l l l l l l l l Cyr6_M RLVKVSGQCCEEWVCDEDSIKDSLDDQDDLLGLDASEVELTRNNELIAIGKGSSLKRLPV 150 160 170 180 190 200 40 50 60 70 80 90 2 0a022 FGMEPRILYNPL—QGQKCIVQTTSWSQCSKTCGTGISTRVTNDNPECRLVKETRICEVR II I I I = I : I I I = I I I I I I I I I I I I I I I I : I I I I I I I I I I I I I I I I I I I I I I I II I I I Cyr6_M FGTEPRVLFNPLHAHGQKCIVQTTSWSQCSKSCGTGISTRVTNDNPECRLVKETRICEVR 210 220 230 240 250 260 100 110 120 130 140 150 2 0a022 PCGQPVYSSLKKGKKCSKTKKSPEPVRFTYAGCLSVKKYRPKYCGSCVDGRCCTPQLTRT I I I II I I II I I I I I II I I I II II I II II I I I I I I II II I I I I I I II I I I I II I I III Cyr6_M PCGQPVYSSLKKGKKCSKTKKSPEPVRFTYAGCSSVKKYRPKYCGSCVDGRCCTPLQTRT 270 280 290 300 310 320 160 170 180 190 200 2 OaO22 VKMRFRCEDGETFSKNVMMIQSCKCNYNCPHANEAAFPFYRLFNDIHKFRD I I I I I I I I M I i ! ! M M 1 i I i M ! i j [ | ! : i | ! : 1 : : I : ! [ i M M M I Cyr6_M VKMRFRCEDGEMFSKNVMMIQSCKCNYNCPHPNEASFRLYSLFNDIHKFRD 330 340 350 360 370 Figure 7 Amino acid sequence similarity between 32bl and cyr61. The amino acid sequence was aligned and calculated to have a 91% sequence similarity. 33 allowing it to be aligned with other members. The cloned c D N A contained 1274 nucleotides and from comparison with cyr61, the sequence with which it is most similar, the coding sequence contained 203 amino acids ending at nucleotide #617. The amino acid sequences of cyr61 and 32bl have a 91% similarity and their D N A sequences are 84% identical. The remaining 646bp of 32b 1 was considered to be the 3' UTR. According to a cyr61 sequence comparison, 32bl is still missing 175 amino acids of the peptide sequence which includes an amino terminus similar to cyr61 and part of an acidic highly diverse central portion. b. Search for the 5' end The cloned 32b 1 c D N A was considered to be missing its 5' end because in comparison to the full-length c D N A encoding cyr61, approximately 500 basepairs were missing. As well , no initiation codon was found for the translated version of 32bl that was similar in position to that in cyr61. In Northern blots of R N A s from human tissues, the sizes of the R N A s hybridizing to a probe for 32bl were approximately 4.4kb and 2.6kb (see Figure 9). These sizes are much larger than that of the probe itself (1.3 kb). In order to characterize 32bl accurately, it was important to find the full length c D N A . Conventional hybridization techniques used to rescreen the original Jurkat T lymphocyte A,gtll library yielded three 200 bp c D N A fragments making them unattractive to clone and sequence. Touchdown PCR techniques used to screen two Jurkat T cell libraries revealed that the 5' end of 32b 1 was undetectable here (refer to Materials and Methods section 2.4 for details). Cloned c D N A s identical to the cloned 34 32b 1 c D N A were isolated numerous times but no additional sequence information was acquired. In another effort to isolate the extreme 5' end of the 32bl c D N A , R A C E was attempted (see section 2.6 in Materials and Methods) but was also unsuccessful in this Jurkat T cell line. In exploring various reasons for this result, Northern Hybridization was performed in an attempt to assess the quality of the T cell m R N A used in this procedure. Although the total R N A looked intact in a stained gel, probing a blot with a labelled 32b 1 c D N A probe revealed that there was no detectable message being expressed in this cell line. Accordingly, the search for the 5' end was extended to other libraries. A library derived from U937 monocytes treated with P M A for 3.5 days (Section 2.4.1 in Materials and Methods) was found to contain fragments that hybridized to the 32bl c D N A clone in a Southern blot analysis. D N A sequencing determination of the largest fragment revealed that it contained a c D N A sequence identical to that in the 32b 1 c D N A but not any novel sequence. Northern blot analysis revealed similar results to that of the Jurkat T cell library. Although the 32bl c D N A clone could be found using PCR, Northern blot analysis could not detect expression of 32bl m R N A in this cell line with or without P M A induction. The repeated isolation of 32bl in the leukocyte derived libraries using PCR suggested that perhaps a mitogen was required to stimulate significant expression in these cell lines. Thus, various cultures of Hela cells, HL-60 cells, Jurkat T cells and U937 cells were stimulated with the phorbol ester P M A for 30 minutes and 2 days and probed wi th a labelled 32b 1 c D N A in a Northern blot. There was no detection of a hybridizing species. A s well , serum and D M S O added to the selected cell culture media for various amounts of time did not 35 induce 32b 1 m R N A expression (refer to section 2.7 in Materials and Methods for details). Other libraries made from human lung tissue, Hela cell lines and human brain tissue were probed but no sequence was found that even allowed binding of the gene specific primer to create a PCR product. c. The discovery of novel 5' end sequence of 32bl from Fasta search A F A S T A search revealed numerous matches with 32bl from various sources. Most matches were fragments of D N A derived from the WashU-Merck EST project (Hillier, L . et al., 1995). Figure 8A shows representative D N A sequences found, referred to here by their accession numbers in the nucleotide database (NCBI) for convenience. Some representative D N A sequences found within the known 32bl sequence were: T92167 and Y09859 (Hillier, L . et al, 1995, Pilarsky, C. et al, 1996). Three other sequences: W51901, N86328 and Z50168 were found to align with 32bl and reveal novel sequence in the 5' end (Genini, M , et al., 1996, Hill ier , L . et al, 1995, Liew, C.C. , 1996) (Figure 8A). The largest 5' fragment obtained from the search, Z50168, was 304 bases long and overlapped with 32bl with a 100% match aligning 85 bases. The remaining 221 bases was novel sequence from the 5' end of 32bl. This z50168 sequence was connected with 32bl (referred to as z5-32bl for convenience) and translated in frame to produce an amino acid sequence lacking any stop codons and containing three cut sites (Figure 8B). The novel amino acid sequence aligned with 32b 1 is shown in figure 8C. In an alignment wi th Cyr61, there are still approximately 100 amino acids missing. 36 A Z50168 n86328 W51901 32b1fini t92167 y 0 9 8 5 9 3 *1 89 161 305 391 632 960 1,l"35 1,270 1,495 B Drall (187) Xmalll (637) Hindin (1209) Mapping all cutsites. Cutters : Drall, Hindm & Xmalll Non-Cutters : Apal, BamHI, Bspl06, BstXI. EcoRI, EcorV, Kpnl, NotI, PstI, Sad, SacII, Sail, Smal, Spel, Xbal & Xhol Figure 8 Representative DNA sequences showing high sequence similarity with 32bl found in a FASTA search. A, The sequences found are referred to here by their accession number in the NCBI nucleotide database for convenience. B, The sequence that aligned with 32b 1 revealing new 5' end sequence was checked for new cut sites. 37 HSA3 3 2 1 0 - 5 ' C C A A A A C G G G G A A A G T T T C C A G C C C A A C T G T A A A C A T C A G T G C A C A T G T A #1 C C A A A A C G G G G A A A G T T T C C A G C C C A A C T G T A A A C A T C A G T G C A C A T G T A Q N G E S F Q P N C K H Q C T C I H S A 3 3 2 1 0 - 5 ' T T G A T G G C G C C G T G G G C T G C A T T C C T C T G T G T C C C C A A G A A C T A T C T C T C #51 T T G A T G G C G C C G T G G G C T G C A T T C C T C T G T G T C C C C A A G A A C T A T C T C T C D G A V G C I P L C P Q E L S L H S A 3 3 2 1 0 - 5 1 C C C A A C T T G G G C T G T C C C A A C C C T C G G C T G G T C A A A G T T A C C G G G C A G T G #101 C C C A A C T T G G G C T G T C C C A A C C C T C G G C T G G T C A A A G T T A C C G G G C A G T G P N L G C P N P R L V K V T G Q C HSA3 3 2 1 0 - 5 ' C T G C G A G G A G T G G G T C T G T G A C G A G G A T A G T A T C A A G G A C C C C A T G G A G G C T G C G A G G A G T G G G T C T G T G A C G A G G A T A G T A T C A A G G A C C C C A T G G A G G C E E W V C D E D S I K D P M E D HSA3 3 2 1 0 - 5 ' A C C A G G A C G G C C T C C T T G G T A A G G A G C T G G G A T T C G A T G C C T C C G A G G T G 3 2 b l f i n i G G A G C T G G G A T T C G A T G C C T C C G A G G T G #201 #251 A C C A G G A C G G C C T C C T T G G T A A G G A G C T G G G A T T C G A T G C C T C C G A G G T G Q D G L L G K E L G F D A S E V HSA3 3210-5' G A G T T G A C G A G A A A C A A T G A A T T G A T T G C A G T T G G A A A A G G C A G C T C A C T 3 2 b l f i n i GAGTTGACGAGAAACAATGAATTGATTGCAGTTGGAAAAGGCAGCTCACT G A G T T G A C G A G A A A C A A T G A A T T G A T T G C A G T T G G A A A A G G C A G C T C A C T E L T R N N E L I — A V G K G S S L Figure 8C The alignment of 32bl w i th the novel 5' end sequence found from the F A S T A search and its traslated amino acids. The three D N A sequences showing novel 32b 1 5' end were connected with 32b 1 forming z5-32bl. 38 2. Expression of 32bl mRNA a. Tissue distribution of 32bl mRNA in adult humans A Northern blot was performed using the cloned 32b 1 c D N A insert to probe different human tissues in order to find out where the m R N A was expressed for comparison wi th other C C N family members. It was also helpful in estimating the approximate size of full-length 32bl m R N A in humans. A blot containing 2.0 ug of poly A + R N A extracted from one of 8 different human tissues in each lane was probed with 32bl (refer to Materials and Methods section 2.9). Figure 9 shows that a major R N A species of 2.6 kb was detected in 7 of 8 tissues.. The 2.6 kb m R N A was expressed at the highest levels in the ovary and colon (Figure 9, lanes 5 and 7). Moderate amounts of m R N A were found in the small intestine, testis, prostate, and spleen (Figure 9, lanes 6, 4, 3 and 1). A low level of 32b 1 was transcribed in the thymus (figure 9, lane 2). The only tissue that did not express the 2.6 kb m R N A was peripheral blood leukocytes (Figure 9 lane 8). A n additional 4.4 kb species was most prominent in R N A extracted from the spleen (figure 9, lane 1) although it was seen in all the tissues shown to express the 2.6 kb m R N A (figure 9, lane 1-7). There was also a dark smear seen in the R N A extracted from the colon tissue and a faint smear in the spleen R N A implying that 32b 1 may also be found in smaller sizes in these areas or imperfections in the purchased Clontech blot. 39 1 2 3 4 5 6 7 8 7.5 kb > 4.4 kb " 2.4 kb • 1.3 kb • G A P D H Figure 9. Human Tissue Blot probed wi th 32bl . In order to examine where the 32bl gene was expressed, a Human Tissue blot was purchased from Clontech and was probed with 32bl as described in section 2.9 of Materials and Methods. Each lane contained 2 jig of poly A+ m R N A . The blot was stripped and reprobed with labelled G A P D H as a control to ensure equal amounts of R N A was loaded in each lane. The lanes contained R N A extracted from the following tissues: 1, spleen; 2, thymus; 3, prostate; 4, testis; 5, ovary; 6, small intestine; 7, colon; 8, peripheral blood leukocyte. 40 b. Effect of mitogenic stimulation on expression of 32bl mRNA in different cell lines The induction kinetics of the 32b 1 m R N A expression in different cell lines were deduced in order to acquire additional information about when and where expression of the gene was stimulated in an immediate early gene manner and thus allow comparison with other C C N family members. Figure 10 shows a representative blot of total R N A from cell lines probed with 32bl c D N A . A number of cell lines of human lymphocyte origin were examined for their ability to express the putative 32b 1 m R N A under various conditions. Figure 10 lanes 1-3 shows that cells grown for at least 2 days in serum including Jurkat T cells (lane 1), U937 (lane 2) and HL-60 (lane 3) failed to express the 2.6 kb m R N A . These cell lines were subjected to a number of treatments in an attempt to induce expression of 32bl m R N A : treatment wi th P M A for 30 minutes or 2 days; D M S O for 2 days; and starvation followed by serum treatment. In no case was the 2.6 kb m R N A detected, (data not shown). Unexpectedly, this 32bl probe hybridized wi th a 2.6 kb transcript in Rat 2 fibroblasts that is induced in cells treated with epidermal growth factor (EGF) (Figure 10, lanes 4-6). A 2.6 kb message appears after 30 minutes of induction (Figure 10, lane 5) but is less intense after 90 minutes of induction (Figure 10, lane 6). 3. Analysis of the Induction Kinetics for 32bl mRNA through Northern Hybridization The results from the previous Northern hybridization experiment showed that a 2.6 kb species wi th high sequence identity to the 32bl gene was-transiently induced after stimulation with EGF in Rat-2 cells. This experiment identified a gene displaying 41 1 2 3 4 5 6 Figure 10. Induction of 32blR. 20 [ig of total R N A was loaded in each lane of a formaldehyde gel, separated electrophoretically, blotted to a nylon membrane, and then probed with labelled 32bl c D N A , as in section 2.9 in Materials and Methods. The R N A in the first three lanes was extracted from cells growing in serum for approximately 2 days. R N A in the last three lanes was extracted from Rat-2 fibroblasts that were rendered quiescent (as described in section 2.7 in Materials and Methods) and then treated with 20 n g / ml EGF for various lengths of time. The cell lines and treatments are as follows: 1, Jurkat T cell; 2, U937; 3, HL-60; 4, Quiescent Rat-2 fibroblasts; 5, 30' EGF induction of quiescent Rat-2 fibroblasts; 6, 90' EGF induction of quiescent Rat-2 fibroblasts. 42 immediate early gene induction characteristics in rodent fibroblasts and suggested that further study of its induction kinetics would be necessary to characterize its role in cell proliferation. To prevent confusion, the Rat-2 hybridizing species w i l l be referred to as 32blR for "rat". a. Time Course of 32bl induction in Rat-2 fibroblasts treated with Fetal Bovine Serum In order to define the kinetics of induction of 32blR m R N A with serum, medium containing 20% fetal bovine serum (FBS) was added to Rat 2 cells which were rendered quiescent (refer to section 2.7 in Materials and Methods). The data in Figure 11 show that a 2.6 kb m R N A appeared after 45 minutes and reached a peak concentration at one hour (Figure 11, lanes 4 and 5). The size of the message became slightly smaller at 90 minutes (Figure 11, lane 6) and completely disappeared within 4 hours. b. Time course of 32bl induction in Rat-2 cells treated with Phorbol 12-myristate 13-acetate (PMA) Phorbol esters were added to density arrested Rat-2 fibroblasts to determine whether the expression of 32blR m R N A could be induced. After P M A addition, R N A was extracted at various time points and analysed on a Northern blot. Unexpectedly, a 2.6 kb R N A could not be detected (Figure 12A). This experiment was repeated several times with the same result. As a control, the same blot was probed with c-fos (Figure 12B). The c-fos m R N A was detected within 30 minutes following induction but was no longer detectable 60 minutes following induction. 43 1 2 3 4 5 6 7 8 9 3 2 b l G A P D H Figure 11. Time Course of Rat-2 fibroblasts treated with Fetal Bovine Serum (FBS) and probed with 32bl. The Rat-2 fibroblasts were grown in DMEM media supplemented with 10% Fetal Bovine Serum and 1% antibiotics. They were rendered quiescent after confluency through density arrest and then induced with 20% FBS. RNA was extracted at various time points and analysed by Northern Hybridization (Refer to sections 2.7, 2.8, and 2.9 in Material and Methods for details). The RNA extractions were organized on the gel in the following order: 1, Quiescent cells; 2,15 min; 3, 30 min; 4, 45 min; 5, lh; 6, 90 min; 7, 2h; 8, 4h; 9, 8h. The same blot was stripped and reprobed with a control DNA probe, GAPDH, to ensure that a consistent amount of RNA was loaded in each lane. 44 Figure 12. Time Course of induction of 32bl R N A accumulation i n Rat-2 fibroblasts treated wi th Phorbol 12-myristate 13-acetate (PMA) . The Rat-2 fibroblasts were grown in D M E M medium supplemented with 10% fetal bovine serum and 1% antibiotics. They were rendered quiescent after confluency through density arrest and then induced with 100 n g / m l of P M A (Section 2.7 in Materials and Methods). Cells were harvested at appropriate times and total R N A was extracted (section 2.8, Materials and Methods). Samples containing 20 | ig of R N A were denatured, resolved by electrophoresis, and probed with a 32b 1 c D N A (section 2.9, Materials and Methods). A , R N A extraction followed P M A addition at the following time points: 1, Quiescent cells (0 min); 2, 15 min; 3, 30 min; 4, 45 min; 5, l h ; 6, 90 min; 7, 2h; 8, 4h; 9, 8h. B, The same Northern blot was stripped and reprobed with a c-fos probe. C, The same blot was stripped and reprobed with a control probe, G A P D H , to check the loading of R N A in each lane. 45 c. Induction of 32bl mRNA in Rat-2 cells with EGF A time course was done in which Rat-2 cells were treated wi th 20 n g / m l of EGF at time 0 and R N A was extracted at various time points thereafter. This experiment was attempted in order to obtain a more detailed picture of the kinetics of induction by EGF with a higher number of time points over a shorter period of time (compared to Figure 10). A Northern blot probed with a 32bl c D N A revealed that EGF induced an R N A of 2.6 kb, which was detectable only 30 minutes after treatment (Figure 13, lane 3). 4. Effect of transcription and translation inhibitors on stability of 32bl mRNA a. Time course of decay of 32bl mRNA in Rat-2 fibroblasts treated with Actinomycin D In order to estimate the half-life of 32b 1 m R N A , Rat-2 fibroblasts were induced with serum for 1 hour and then treated with actinomycin D , an inhibitor of transcription. R N A was extracted subsequently at 30 minute intervals. A 2.6 kb m R N A was detected at the 30 minute time point (Figure 14A, lane 2), as was a smaller species at 60 minutes (Figure 14A, lane 3). N o hybridizing species was detectable at 90 minutes (Figure 14A, lane 5). b. Time course of 32bl decay in Rat-2 fibroblasts treated with Actonomycin D and Cycloheximide Cycloheximide, an inhibitor of translation was added with Actinomycin D in order to qualitatively measure the stability of 32bl m R N A . Cycloheximide was added 46 Figure 13. Time Course of Rat-2 fibroblasts treated wi th Epidermal Growth Factor (EGF) and probed wi th 32bl. The Rat-2 fibroblasts were grown in D M E M media supplemented with 10% Fetal Bovine Serum and 1% antibiotics. They were rendered quiescent after confluency through density arrest and then induced with 20ng/ml EGF (Section 2.7 in Materials and Methods). Cells were harvested at appropriate times and total R N A was extracted (section 2.8, Materials and Methods). Samples containing 20 ug of R N A were denatured, resolved by electrophoresis, and probed with a 32bl c D N A (section 2.9, Materials and Methods). R N A extraction followed EGF addition at the following time points: 1, Quiescent cells; EGF was added and R N A extraction followed at these time points: 2, 15 min; 3, 30 min; 4, 45 min; 5, l h ; 6, 90 min; 7, 2h; 8, 4h; 9, 8h. The same blot was stripped and reprobed with a control probe, G A P D H , to check the loading of R N A in each lane. 47 1 2 3 4 5 6 32bl - 2.6kb G A P D H - 0.7kb Figure 14A. A time course of decay of 32bl m R N A in Rat-2 fibroblasts fol lowing treatment wi th Act inomycin D . Rat-2 fibroblasts were rendered quiescent via density arrest and treated with 20% FBS for l h (Section 2.7 in Materials and Methods). Cells were then treated with actinomycin D ( lng/ml) and samples were taken for R N A extraction (Section 2.8, Materials and Methods). 20 ug of each R N A sample was separated in each lane, transferred to nylon blotting membrane and then probed with a 32b 1 c D N A probe (section 2.9, Materials and Methods). Lane 1, quiescent Rat-2 fibroblasts; 2, quiescent Rat-2 cells treated with FBS for 1 hour. The samples in the following lanes were treated with FBS for 1 hour, at which time Actinomycin D was added (time 0) and samples were taken for R N A extraction at the following time points: 3, ActD 30 min; 4, Ac tD 60 min; 5, A c t D 90 min; 6, Ac tD 120 min . G A P D H was used as a control for equal sample loading. 48 simultaneously with Actinomycin D in a manner similar to the above experiment. A strong signal was seen 60 minutes after addition of the inhibitors (figure 14B, lane 3). A t succeeding times, the signal became fainter but was clearly visible after 2 hours (Figure 14B, lane 5). However, the signal disappeared within eight hours after the inhibitors were added (not shown). 49 1 2 3 4 5 32bl G A P D H Figure 14B. A Time Course observing 32bl m R N A hybridization patterns fol lowing treatment of Rat-2 fibroblasts wi th Act inomycin D and Cycloheximide. Rat-2 fibroblasts were rendered quiescent via density arrest and treated with 20% FBS for l h . Cells were then treated with l n g / m l actinomycin D (ActD) and lOng /ml cycloheximide, simultaneously (section 2.7, 2.8 in Materials and Methods). 20 u.g of each R N A sample was separated in each lane and probed with a 32bl c D N A (described in section 2.9, Materials and Methods). Lane 1, quiescent Rat-2 cells induced 1 hour with 10% FBS; 2, Ac tD & Cycloheximide (Cyclo), 30 min; 3, Ac tD & Cyclo, 60 min; 4, Ac tD & Cyclo, 90 min; 5, Ac tD & Cyclo, 120 min. G A P D H was used as a control. 50 Discussion For clarification purposes, I have divided this section into two halves. The first part discusses 32bl as a c D N A product of its human gene. This w i l l describe conclusions drawn from its coding sequence and its 3' untranslated region that may predict its induction kinetics. There w i l l also be discussion of information derived from its tissue distribution and lack of expression in various human cell lines seen through Northern hybridization experiments. This project took an interesting turn when the 5' end of 32bl could not be found in various libraries derived from human cell lines. The only cell line seen to express the m R N A product after treatment with serum and various growth factors was in another species, a Rat-2 fibroblast cell line. Various time course experiments were done in these Rat-2 cells to produce an induction profile on what is presumably an m R N A transcribed from a rat gene with very high sequence identity with 32b 1. As seen in the results, to prevent any confusion, the rat m R N A that hybridizes to the 32bl probe in Northern blots w i l l be referred to as 32blR m R N A . The cross-species hybridization was carried out in Rat-2 fibroblasts because in other cloning projects, the detection of m R N A species under stringent conditions, that is, similar conditions to the Northern blot used with 32b 1, led to discovery of new genes encoding c D N A sequences of higher than 90% similarity at the amino acid level (Calalb, M . et al., 1992, Stacker, S. et a l , 1993). Further analysis of the 32blR species was deemed interesting for two reasons. First, it was interesting because it was a transcript novel to this species; secondly, because it allowed comparison of functional differences of two highly similar genes, cyr61 and 32blR in mouse and rat fibroblasts respectively. 51 A n y differences in results could later be further analysed through studying any sequence or environmental divergence of the genes and their expression. It was hypothesized that studies in the fiboblasts cell lines that expressed cyr61 and 32blR would most likely be similar seeing that they were both derived from rodent cells. Al though wi th different cell lines, previous clones have been seen to act in a similar manner after transfection into both rat and mouse fibroblasts, as seen with five cosmids involved in commitment of fibroblasts to adipogenesis (Colon-Teicher, L . et al., 1993). Regardless of the outcome, the results of the comparison may lead to clues to the function of 32b 1. 1. What can the DNA sequence of the 32bl cDNA tell us about the gene products? The predicted amino acid sequence of the 1.3 kb fragment of 32bl is 91% identical to cyr61 and contains the high number of cysteine residues in the colinear positions that are characteristic of this family. Unfortunately, because the putative consensus polyadenylation signal ( A A T A A A ) was not found, a part of the 3' end is still considered missing. In other genes, including cyr61, this putative signal was found within 20bp of the poly A tail (O'Brien et al., 1990; Ryseck et a l , 1991). A 49 bp direct repeat was found in the 3'UTR of cyr61 but not in 32bl. This may be due to missing D N A from the 3' end or simply to its absence in 32bl. The 3 'UTR seen in the 32bl clone is made up of 60% A and T residues and contains many in-frame stop codons. This A U -rich region contains several U-rich regions and three A U U U A pentamers at positions 1148, 1242 and 1259 (see Figure 15). While the pentamers, referred to as A U - R i c h Elements ( A U R E s / A R E s ) , have previously been found to facilitate degradation of 52 m R N A , nearby U-rich elements aid in m R N A destabilization and deadenylation, the first step in the degradation pathway (Ross, 1995; Chen et al, 1994; Shaw and Kamen, 1986). This is similar to cyr61 which also contains multiple copies of these pentamers (O'Brien et al., 1990). The first two pentamers in 32bl m R N A are part of an octanucleotide consensus sequence (TTATTTAT) also shown by Shaw and Kamen (1986) to destabilize the G M - C S F m R N A . Thus, it can be predicted that the 32bl m R N A should be very transient and short lived; my data in Rat-2 fibroblasts confirm this prediction. The portion of 32bl 3'UTR which may aid in conferring instability is shown underlined in Figure 15. As well, in the 3' UTR of 32bl, there were two copies of a 6mer, TTTGTA, and one 7mer, T T T A T A , which were previously found to be responsible, together with a 5' element, for serum inducibility in immediate early genes (Freter et al., 1992). It is a feature commonly found in certain types of IEGs such as trancription factors, for example c-fos and c-myc, but not in the cytoplasmic signal transduction proteins, structural proteins or cell surface receptors (Freter et al., 1992). These results provide evidence suggesting that 32b 1 gene is immediate early. The bolded italic regions in Fig. 15 represent these sequences. 2. What does the novel 5' sequence tell us? The significance of the novel 32b 1 sequence lies in the circumstances under which it was found. The partial c D N A , z50168, was discovered during a subtractive hybridization search for proteins downregulated in skeletal muscle tumors referred to as rhabdomyosarcomas (RMS) (Genini, M . et a l , 1996). These small-cell tumors are insufficiently differentiated into muscle cells. Since z50168 (the 5' end of 32bl) was 53 600 TTTAGGGACTAAATTGCCTACCTGGGTTTCCAGGGCACACCTAGACAAACAAGGGAGAAG 65 9 201 PheArgAsp*** 203 660 AGTGTCAGAATCAGAATCATGGAGAAAATGGGCGGGGGTGGTGTGGGTGATGGGACTCAT 719 72 0 TGTAGAAAGGAAGCCTTGCTCATTCTTGAGGAGCATTAAGGTATTTCGAAACTGCCAAGG 77 9 7 8 0 GTGCTGGTGCGGATGGACACTAATGCAGCCACGATTGGAGAATACTTGCTTCATAGTATT 8 3 9 8 4 0 GGAGCACATGTTACTGCTTCATTTTGGAGCTTGTGGAGTTGATGACTTCTGTTTTCTG7T 8 99 900 r G T A A A T T A T T T G C T A A G C A T A T T T T C T C T A G G C T T T T T T C C T T T T G G G G T T C T A C A G T C 959 960 GTAAAAGAGATAATAAGATTAGTTGGACAGTTTAAAGCTTTTATTCGTCCTTTGACAAAA 1019 1020 GTAAATGGGAGGGCATTCCATCCCTTCCTGAAGGGGGACACTCCATGGAGTGTCTGTGAG 107 9 108 0 ACGGCAGTCTATCTGCACTCTAAACTAAACTGCAAACAGAAATCAGGTGTTTTAAGACTG 1139 114 0 AATGTTTTATTTATCAAAATGTAGCTTTTGGGGAGGGAGGGGAAATGTAATACTGGAATA 1199 1200 ATTTGTAAATGATTTTAATTTTATATCAGTGAAAAGATTTTAlTTj^TGGAATTAACCA/TT 1259 12 60 TAATACCCGAATTCC 12 7 4 Figure 15 Identified elements found in the 3' UTR of 32bl. The underlined bases at positions 1148, 1242 and 1259 are thought to confer instability to the m R N A message (Shaw and Kamen, 1986). The bolded, italic sequences at positions 897, 1201 and 1219 were identified by Freter et al. (1992) as elements characteristic of immediate early genes. 54 found in normal cells but was absent in the tumors, these results suggest that 32b 1 plays a significant role in differentiation, perhaps as a tumor suppressor gene. It was also important to look for clues to 32bl's function in the other cells from which it was isolated. The many partial 32b 1 sequences derived from the WashU-Merck EST project revealed the gene to be expressed in adult female uterus (AA044451, Hil l ier ,L. et al., 1995), normal lung tissue from a 72 year old male (T92167, Hiller , 1. et al., 1995), and human fibroblasts (W51901, Hill ier, L . et al., 1995). 32bl was also isolated in different projects from the fetal heart (N86328, Liew, C, 1996) and from benign tissue of prostate cancer patients (Y09859, Pilarsky, C. et al., 1996). This was especially interesting because the reference, although not yet published, also refers to 32bl as a partial c D N A sequence (226bp) which codes for a putative tumor suppressor gene for prostate cancer (Pilarsky, C. et al., 1996). This evidence strongly suggests that 32b 1 plays an important regulatory function in humans, consistent wi th the evidence from other family members. 3. Why was the search for the 5' end of 32bl unsuccessful? There are many possible explanations why the 5' end could not be found in the various c D N A libraries looked at. The libraries searched were several years old and the missing fragment could have been simply lost over time. It was also conceivable that the library became contaminated with 32b 1 and that it was not normally present in this source. Since 32bl encodes a novel sequence, another possibility is that it may only be found as this 1.3 kb fragment in these particular cell lines. 55 The lack of detection seen in the Northern blots using R N A from leukocytic cell lines implies that 32b 1 m R N A is an extremely rare transcript. Northern blot analysis requires approximately 1 x 10 ^ target transcripts in order to be detectable on the gel (Hengen, 1995). These results were supported by the lack of expression seen in blood leukocytes on the human tissue blot (Figure 9 seen below). It is interesting to note that P M A failed to induce 32b 1R expression in Rat-2 fibroblasts as well as in the human blood leukocyte derived cell lines tested. Perhaps a different mitogen may induce 32b 1 or it may simply not be expressed in these cell lines. Another possible reason for lack of induction in these cell lines may be because they are oncogenic in nature. They are less dependent on serum for cell growth, therefore they may continue to proliferate after serum starvation in which case they would be less responsive to addition of serum (Yu et al, 1993). It is also important to note that the PCR fragments cloned were all truncated to the same nucleotide. This gene could have a secondary structure that does not allow a reverse transcriptase to read through the m R N A . This would yield a truncated message such as the original 32b 1 clone because most of the isolation methods attempted were based on reverse transcription. Other researchers have found similar results where a c D N A could be isolated using P C R but not with Northern blot hybridization (Ashworth, 1993). They concluded that the isolation of a PCR product from a particular c D N A sample does not necessarily indicate that this gene is significantly expressed in that tissue. Rather the extreme sensitivity of P C R led to its isolation and Northern blotting should be used to confirm its meaningful expression (Ashworth, 1993). 56 4. What does the human tissue distribution of 32bl tell us when compared with other C C N family members? The human 32b 1 m R N A was detected only in a purchased tissue distribution blot. Here, the probe 32bl hybridized to two transcripts: a smaller 2.6kb band and a 4.4kb fragment. A similar larger R N A has been seen in tissue blots wi th other transcripts such as G a b l , where the larger transcript was thought to be representative of alternative splicing or to be derived from a related gene (Holgado-Madruga, 1996). As seen in Table 2, most of the other family members were strongly expressed in the lung but the purchased Northern tissue blot did not have lung so no comparison could be made. Cyr61, the mouse homologue displays a different tissue distribution than 32bl. There were only 4 common tissues probed with cyr61 (O'Brien et al, 1990): both messages were expressed in the ovary and testis; but while 32bl was expressed in the small intestine and spleen, cyr61 was not. As mentioned above, no message was detected in the peripheral blood leukocytes. The expression found in the thymus, which contains a high number of immature T cells, correlated with this result by having a weaker expression relative to the other tissues. Northern blot analyses also echoed this result when the blood leukocyte cell lines, such as Jurkat T cells, U937 cells, and HL-60 cells, tested did not express the 32b 1 transcript even after treatment wi th various agents. Information about the tissues that express 32bl w i l l eventually lead to clues about its general and tissue-specific functions in comparison to other C C N family members. 57 Table 2. Tissue distribution of 32bl and various other CCN family members. The information compiled in this table was derived from a 32bl probed Northern blot of Human tissue and from published literature. Low, lowest, moderate (mod), high, highest and undetectable (undetect) concentrations of mRNA expression seen were indicated with respect to the tissues tested for each family member. " " represents data not available. Tissues 32bl cyr61 CEF-10 fisp-12 nov spleen mod undetect low mod low thymus low undetect prostate mod testis mod low lowest ovary high low intestine mod undetect low undetect colon high blood leukocyte undetect heart mod undetect mod undetect kidney low low mod v. low liver undetect undetect undetect undetect adrenal gland low brain low low mod mod uterus mod muscle mod undetect undetect pancreas undetect lung highest highest highest highest bursa low stomach low 58 5. Is 32bl the human homologue of cyr61? In order to determine whether 32bl is the human homologue of cyr61, it is first important to clearly define the word "homology". The word has acquired numerous different meanings over the years that have caused much confusion in the scientific literature when used in comparisons between different species. In biological terms, homology has traditionally been defined as "having a common evolutionary origin" but in recent years, it has also been loosely used in the context of "possessing similarity or being matched" (Reeck et al., 1987). To prevent confusion, it has been suggested by various researchers that the term 'homology' used in this context be replaced by the terms "similarity","identity", "isology" or not used at all (Reeck, 1987; Dover, 1987; Aboit iz , 1987; Wegnez, 1987). Wi th respect to biological evolution, homology itself is a quality. A gene is either homologous or not. A t present, it can only be hypothesized that the species compared have a common ancestor. The protein product of the homologous genes may not serve the same function or be similarly regulated depending on its environment. Examining sequence similarity or identity between the two genes can help determine the probability that the two genes are homologus. In the case of the genes coding for 32bl and cyr61, wi th 91% sequence similarity, the real question is, what is the likelihood that these two genes became this similar from two divergent genes. The convergence of two sequences can happen in two different ways; first, two different proteins may be trying to fulfill the same function in similar cellular environments. The second way may be through genomic changes, that is, genes become crossed over, have point mutations, etc. to arrive at similar sequences. It is highly unlikely that through these two ways, a 59 gene would converge to have such high sequence similarity (Dover, 1987). Therefore, there is a high probability that 32bl is the human homologue of cyr61. The cross-species hybridization of the 32bl probe to an m R N A fragment found in Rat-2 cells under highly stringent washing conditions implies that there is also a high sequence similarity between 32bl and 32blR. Based on the arguments seen above, it is also reasonable to assume that 32blR is also the rat homologue of cyr61. Despite high sequence similarity, these genes do not seem to be functional homologues. Comparison of the induction kinetic profiles of 32blR with its mouse homologue cyr61 indicate that the m R N A products of both genes function at different times of the cell cycle. The m R N A product of 32blR is very transient when induced by serum or EGF while the expression of cyr61 following serum stimulation continues for up to 8 hours. Cyr61 is also induced by phorbol esters while 32blR does not respond to P M A stimulation. Perhaps the fact that 32bl is lacking the 49 bp repeat found in the cyr61 3 'UTR contributes to the differences seen in their induction patterns. These regulatory differences are also seen when other family members are compared (Kireeva et a l , 1996, X i n et a l , 1996). Cellular environment may play a role in the induction of m R N A expression of the two genes causing them to become differentially induced but from the results obtained here, cyr61 and 32blR likely perform different functions in the cell lines in which they are expressed. 60 6. What can be learned about the induction kinetics of 32blR after stimulation with various mitogens? In comparing the induction kinetics of 32blR with similar experiments done for other C C N family members, it must be noted that minor differences in conditions may cause a varied response in gene expression. Methods used to render the cells quiescent have an impact on the immediate early response of the fibroblasts to various growth factors and cytokines (Jahner et al, 1991a). As well, immortal cells and primary cells do not carry out completely normal cell functions. The age and type of cell line used in experiments may also give various results depending on the length of the cell cycle and the proteins expressed. Finally, it is vital to remember that the majority of the experiments are done using c D N A from a human gene to probe m R N A extracted from a Rat-2 fibroblast cell line. The results may not reflect how 32bl is expressed in humans. The primary reason for doing the three induction time course experiments was to establish a context in which the gene is induced and the length of time it is expressed. From our knowledge of events that follow treatment wi th these agents, one can infer the distinct signal transduction pathways that involve the transcription of 32blR in Rat-2 fibroblasts. The m R N A of 32blR was expressed after serum and EGF, but not P M A , was added to quiescent cells. These results indicate that there may be a serum response element (SRE) element present in the promoter of 32blR, as seen with cyr61 (Latinkic et al, 1991). This element binds to the serum response factor (SRF) and plays an important role in the serum induction and transcriptional repression of c-fos (Latinkic et al, 1991). The induction kinetics of the 32blR m R N A and c-fos are similar in their transient 61 appearance but in contrast, cyr61 transcription is mantained for 8 hours following serum induction (Latinkic et al, 1991). Other factors, such as cell culture environment or the presence of a regulatory factor which inhibits or induces m R N A transcription, must play a role in serum induced expression to exhibit the different kinetics. One growth factor present in serum is EGF which is needed for growth by almost all cells in culture. EGF activates 32blR in a manner which is more transient than expression seen with serum which implies that serum contains additional regulatory proteins that also play a role in induction. EGF is an activator of a receptor tyrosine kinase (RTK) which couples to both the phosphatidylinositol pathway, leading to activation of the Protein Kinase C (PKC) and an increase in the intracellular (Ca2+) concentration (Boonstra, 1995), as well as to the Ras pathway which sends messages to the nucleus via the M A P kinases (Karnitz, 1995). To test the role of the P K C and calcium concentration in the induction of 32blR and for comparison to other family members, quiescent Rat-2 cells were treated with a phorbol ester P M A . The main biological response to treatment with P M A is the induction of the phospholipid dependent Protein Kinase C (PKC) pathway by substituting for D A G (Darnell et al, 1990). The effect of P M A addition on the expression of 32blR was very important because CEF-10 and cyr61, the chicken and mouse homologue displayed different patterns of induction. Unexpectedly, 32blR was uninduced by phorbol esters. The validity of the experiment was further tested by probing the same Northern blot with c-fos, a gene known to be phorbol ester inducible. Its pattern of induction was the same as previously published (Huang et al, 1995; Bi et al, 1994; Borner et al, 1995). Because the phosphatidylinositol pathway failed to induce 32bl m R N A transcription, it is most 62 likely that EGF induces 32blR through the Ras pathway or through another pathway that has not been previously described. 7. What is the effect of inhibition of protein synthesis and translation on the stability of the immediate early gene 32blR? Actinomycin D was used to shut off transcription to qualitatively estimate the half life of 32bl. Although a higher frequency of time points would have given a more accurate estimation, the half-life of 32bl was approximated to be less than 30 minutes. As well , cycloheximide was used in conjunction with actinomycin D in order to test the effect of translation on the stability of the m R N A transcript. The superinducing effect of cycloheximide may be due to two independent factors, an increase in stability of the m R N A message and a prolonged transcription of the genes. The ability of cycloheximide to stabilize the 32b 1 m R N A message suggests one of two things: either a newly synthesized or labile nuclease is required for degradation or inhibition of the m R N A transcript is linked to translation (Lau and Nathans, 1991). In the latter circumstance, perhaps there is a site that must be translated in order to initiate the m R N A degradation process. Treatment with cycloheximide increased the stability of the 32bl message at least 3-fold. Previous studies with c-myc revealed that mammalian cells treated with cycloheximide extend m R N A expression due to a retardation of deadenylation ability but the actual degradation after this primary step is not affected (Ross, 1995). From the experiments done with 32bl and assuming the smaller band seen just prior to disappearance of the hybridizing band is due to deadenylation, the results do not concur with those from c-myc. Although the deadenylation step does 63 take a longer period to complete, the time required for the degradation step is also considerably lengthened. This supports a possibility that either the smaller band seen is not due to deadenylation or there are other factors influencing the stability of the 32bl m R N A message besides translation. These results confirm that translation plays an important role in increasing the stability of the message. 8. Is 32blR an immediate early gene? Immediate early genes (IEGs) derive their name from their analogy with primary transcriptional response of D N A viruses (Lau and Nathans, 1987). A gene is referred to as being " immediate early " when it meets the three defining criteria previously discussed in the Literature Review (Lau and Nathans, 1991). Based on these requirements, 32b IR is most likely an immediate early gene meeting two of the three characteristics. Besides being induced immediately after stimulation with various mitogens without de novo protein synthesis, the 32blR m R N A is a rare transcript in control cells and becomes superinduced in the presence of transcription inhibitors such as cycloheximide; these are all general traits of IE genes (Simmons et al, 1989). The 32blR m R N A species induced by different agents is expressed in a very transient manner and has a very short half-life of less than 30 minutes, estimated using Act D , consistent wi th the behavior of group I members of the immediate early gene group (Lau and Nathans, 1991; Lau and Nathans, 1987). The final experiment required in order to prove without a doubt that 32blR is an immediate early gene would be a nuclear run-on assay. The results from this experiment would indicate if the gene is transcriptionally activated as a result of mitogen stimulation or simply stabilized. 64 9. What is the likelihood that 32bl is RBF-1? Although further studies must be done in order to fully understand 32bl's role in the cell, there are many results that support the probability that 32bl is not RBF-1. The RBF-1 and RBF-2 factors are thought to be required for Ha-Ras responsiveness in Jurkat T lymphocytes, with or without P M A induction, and in macrophages. From electrophoretic mobility shift assays (EMSAs) of the mutated RBE binding sites done wi th Jurkat nuclear extracts, it was determined which areas were important for RBF binding in response to Ras induction (Bell, B. et al., unpublished). The major contradictory result found for 32bl was that it appeared to be a very rare transcript in this same Jurkat T lymphocyte cell line, and in most blood leukocytes, seen through lack of detection in Northern blots. Perhaps 32b 1 was isolated in the Southwestern blot screening because it contained many modules for binding and oligomerization (Bork, 1993). Results seem to suggest that 32bl and its family members play a role in the mediation of cell proliferation and wound repair. 10. A model for 32bl Experimental results with 32bl were unexpected because the 5' end could never be found in the Jurkat T cell line from which it was originally isolated. The c D N A was repeatedly isolated from blood leukocyte derived cell lines using P C R but never detected in Northern blot analysis. This implies that the 32bl gene is not significantly expressed here. The numerous times that 32bl was isolated during P C R in both Jurkat T cells and U937 cells treated with P M A for 3.5 days suggests that it is not likely to be a 65 coincidence. In gathering clues from the accumulated results, it seems that in these cell lines, 32b 1 could only be found in this 1.3 kb length as a rare transcript without the 5' end seen in its mouse homologue, cyr61. The 32bl transcript is encoded by a gene with homology wi th the C C N family which members are important in mediating cell differentiation. The F A S T A results indicate that 32bl may be a tumor suppressor gene, from the findings of two separate research groups where 32b 1 was found in normal cells but not in tumors (Genini et a l , 1996, Pilarsky et al., 1996). This also supports a regulatory function for 32b 1. It has been previously found in nov, another C C N family member, that there is a regulatory function in the amino terminus of the protein. Ful l length nov negatively regulates cell growth but an amino truncated version can transform CEF cells (Joliot et al., 1992). A n interesting hypothesis may be that since the Jurkat T cell line is oncogenic and 32bl is a tumor suppressor gene, perhaps 32bl is similar to nov in that its regulatory portion is found in the amino terminus while the carboxy terminus, without this control, contributes to oncogenicity. The 5' end of 32bl could be absent in these leukocyte cell lines because this truncated version of 32bl contributes to the uncontrolled growth in these cells albeit at low levels. This would explain the absence of the 5' end. A model for 32bl function may be that it functions in a similar way to cyr61 because of their high sequence similarity and any differences seen here may be due to cellular environment. In this case, 32b 1 may act as an extracellular matrix molecule as well as a tumor suppressor gene. 32bl would be lacking its regulatory domain in the amino terminus and may actually be signaling cell proliferation in an uncontrolled manner as a signaling molecule bound to the extracellular matrix causing cancerous 66 growth. Under normal circumstances, the 5' end of 32bl would regulate the signaling function of 32bl and promote cell proliferation in a controlled manner, perhaps through the binding of an insulin-like growth factor with domain I of its protein which has sequence similarity with IBPs. 11. Where do we go next? The induction kinetics profile of 32blR hybridizing with a human c D N A probe has left some interesting hypotheses to be tested. A t this point, 32bl and 32blR should be separated into two distinct sets of experiments and eventually converged to allow further speculation on the homologous background of the two genes. In studying 32bl, the human c D N A clone, the first step would be to find a human cell line that expresses the m R N A transcript. Because some of the C C N families were found in fibroblasts, a human fibroblast cell line may be a reasonable place to start. Before wasting time on looking in a library for a very rare species, induce quiescent cells wi th serum and extract R N A in a time course experiment. A Northern blot of the various time points using 32b 1 as a probe w i l l allow estimation of when (or if) 32b 1 is expressed. From here, R A C E can be used on that R N A sample to isolate the 5' end of 32bl, or alternatively a library can be made derived from R N A extracted at the time point where 32bl is expressed at its highest level. This library can then be searched using Touchdown P C R or conventional screening methods. Once the full length clone is found, then its induction kinetics can be studied in that particular cell line. It can also be used to probe previously tested Rat-2 Northern time courses or to see if a different induction pattern is seen. It can be further tested by doing serum induction 67 experiments in the 3T3 mouse fibroblast cell line to see if it reacts similarly to cyr61. That is, is the difference in induction kinetics and tissue expression a function of the cellular environment? If dissimilar, further studies could be done to find any significant differences in the two gene products that may play a role in serum induction kinetics. Some interesting experiments that would answer some questions that have arisen in comparison with other family members would be a Northern blot testing the tissue distibution of 32b 1 m R N A in human lung relative to other tissues. A s seen wi th many other C C N members, is it most highly expressed in the lung? Another extremely interesting experiment would be to repeat a similar experiment to the one described for nov where its 5' truncated transcript was tested to see if it induced transformation in cells. This would be interesting in the leukocyte cell lines where the truncated c D N A clone was only found missing the 5' end. Wi th respect to 32blR, it would also be important to return to the Rat-2 fibroblast cell line and find the rat homologue of 32b 1. Since we know where and when it is expressed, it would seem like a simple task to return to the R N A extractions expressing 32bIR in Northern blots and use the techniques discussed in this thesis to find the full-length clone. Another important experiment would be to fully define 32blR as an immediate early gene by doing a nuclear run-on assay to confirm that transcription is activated during mitogen stimulation. For other comparisons which may give clues to the function and regulation of 32blR, further time course experiments using other growth factors such as fibroblast growth factor (FGF), Platelet derived growth factor (PDGF) and Transforming growth factor beta (TGF-p) could be carried out. D N A 68 sequencing of both c D N A clones w i l l allow further comparison and would be seen as the discovery of two novel members of the C C N family. 69 Conclusion The fact that these " C C N " family members are found in a wide range of species implies their importance in regulating cell proliferation. Many of the genes were discovered in fibroblasts, cells involved in this function and were seen to be induced by growth factors, such as TGF-f3, PDGF and EGF that have been previously discovered to play a significant role in wound repair. Although there have been many results supporting this general function for C C N family members, the specific members when compared in similar experiments display completely different regulation as wel l as different roles in cell differentiation. Cyr61 has been hypothesized as an extracellular matrix signalling molecule, C T G F is a growth factor, nov is a negative regulator for growth and 32bl as a tumor suppressor. This is a relatively new group of immediate early genes on which further study must be carried out, especially to determine the significance of the 38 aligned cysteines in their amino acid structure and their specific roles in vivo. During this term of research on 32bl, two novel C C N family members were discovered. 32bl, discovered from human Jurkat T lymphocytes, and another 32bl homologue in Rat-2 fibroblasts (32b 1R). Although the complete 32bl gene was never found, much information was found with respect to its induction patterns and R N A stability which may help in further research in the future. 70 References Adamson, E.D. 1993. Growth Factors and Their Receptors in Development. Developmental Genetics 14: 159-164. Akimoto, K v R. Takahashi, S. Moriya. 1996. E G F or P D G F receptors activate atypical FKCX through phophatidylinositol 3-kinase. E M B O J. 15: 788-798. Almendral, J.M., D. Sommer, H. MacDonald-Bravo, J. Burckhardt, J. Perera, R. Bravo.1988. Complexity of the Early Genetic Response to Growth Factors in Mouse Fibroblasts. Mol.Cel l .Biol . 8: 2140-2148. Ashworth, A. 1993. Cloning transcription factors by homology. Transcription Factors. A Practical Approach. D . Latchmam (Ed.) pl25 -142. Attisano, L. J. Wrana, F. Lopez-Casillas, J. Massague. 1994. TGF-P receptors and actions. Biochim. et Biophys. Acta 1222: 71-80 Avruch, J., X. Zhang, and J. Kyriakis. 1994. Raf meets Ras: completing the framework of signal transduction pathway. TIBS 19 Baldari, C, G. Macchia, A. Massone and I. Telford. 1992. p21ras contributes to HIV-1 activation in T-cells. FEBS 304: 261-264 Bell, B. and I. Sadowski. 1996. The novel factors RBF-1 and RBF-2 are required for Ras responsiveness of the HIV-1 LTR. (Accepted by Oncogene). Berridge, M. J. 1993. Inositol trisphosphate and calcium signalling. Nature 361: 315-324. Berridge, M. J. 1995. Calcium signalling and cell proliferation. Bioessays 17: 491-500. Bi, N. and M. Mamrack. 1994. P M A Inhibits the Growth of Human Fibroblasts after the Induction of Immediate-Early Genes. Exp. Cell Res. 212:105-112. Boonstra, J., P Rijken, B. Humbel, F. Cremers, A. Verkleij, P. van Bergen en Henegouwen. 1995. The epidermal growth factor. Cel l Biol Int 19: 413-430. Bork, P. 1993. The modular architecture of a new family of growth regulators related to connective tissue growth factor. FEBS 327: 125-130. Borner, C, M. Veffing, S. Jaken, P. Parker, I. Weinstein. 1995. Two closely related Isoforms of Protein Kinase C Produce Reciprocal Effects on the Growth of Rat Fibroblasts. The J. Biol . Chem. 270: 78-86. Bos, J.L. 1995. p21ras: an oncoprotein functioning in growth factor-induced signal transduction. Eur J Cancer 31:1051-1054. 71 Boudewijn, M v and J. Bos. 1995. Regulation of Ras-mediated signalling: more than one way to skin a cat. TIBS 20:18-22. Bradham, D., A. Igarashi, R. Potter, and G. Grotendorst. 1991. Connective Tissue Growth Factor: a Cysteine-rich Mitogen Secreted by Human Vascular Endothelial Cells is Related to the SRC-induced Immediate Early Produce CEF-10. J. Cel l Biol . 114: 1285-1292. Brinkman, A., C. Groffen, D.J. Kortleve, A. Geurts can Kessel, and S.L.S Drop. 1988. Isolation and characterization of a c D N A encoding the low molecular weight insulin -like growth factor binding protein (IBP-1). E M B O J. 7: 2417-2423. Brunner, A., J. Chinn, M. Neubauer, and A. Purchio. 1991. Identification of a Gene Family Regulated by Transforming Growth Factor (3. D N A Cel l Biol . 10: 293-300. Bunone, G., P. Briand, R. Miksicek, D. Picard. 1996. Activation of the unliganded estrogen receptor by EGF involves the M A P kinase pathway and direct phosphorylation. E M B O J. 15: 2174-2183. Calalb, M., D. Fox, and S. Hanks. 1992. Molecular Cloning and Enzymatic Analysis of the Rat Homolog of "PhK-yT," an Isoform of Phosphorylase Kinase Catalytic Subunit. J. Biol. Chem. 267:1455-1463. Cano, E. and L. Mahadevan. 1995. Parallel signal processing among mammalian M A P K s . TIBS 20:117-122. Chavrier, P., M. Zerial, P. Lemaire, J. Almendral, R. Bravo, and P. Charnay. 1988. A gene encoding a protein with zinc fingers is activated during G o / G l transition in cultured cells. E M B O J. 7: 29-35. Chen, C.Y., A.B. Shyu. 1994. Selective degradation of early-resonse-gene mRNAs : functional analyses of sequence features of the AU-r i ch elements. M o l . Cel . Biol . 14: 8471-8482. Chomczynski, P., and N. Sacchi. 1987. Single-Step Method of R N A Isolation of A c i d Guanidinium Thiocyanate-Phenol-Chloroform Extraction. Analytical Biochemistry 162: 156-159. Coffer, P., B Burgering, M. Peppelenbosch, J.L. Bos, W. Kruijer. 1995.. U V activation of receptor tyrosine kinase activity. Oncogene 11: 561-569. Cohen, G., R. Ren, D. Baltimore. 1995. Modular Binding Domains in Signal Transduction Proteins. Cel l 80: 237-248. Crews, C, and R. Erikson, 1993. Extracellular Signals and Reversible Protein Phosphorylation: What to Mek of it all. Cel l 74: 215-217. 72 Daub, H., F. Weiss, C. Wallasch, A. Ulrich. 1996. Role of transactivation of the EGF receptor in signalling by G-protein-coupled receptors. Nature 379: 557-560. Don, R.H., P. Cox, B. Wainwright, K. Baker, and J. Mattick. 1991. Touchdown ' PCR to circumvent spurious priming during gene amplification. N A R 19: 4008. Dover, G, F. Aboitiz, and M. Wegnez. 1987. Nonhomologous Views of a Terminology Muddle . Cel l 51:515-516. Estable, M.C., B.Bell, A. Merzouki, J. Montaner, M. Shaughnessy, and I. Sadowski. 1996. HIV-1 LTR variants derived from patients at all stages of infection display a wide range of sequence polymorphism and transcription activity. J.Virol. 70: 4053- 4062. Fang, M.A., G.M. Noguchi, S. McDougall. 1995. Epidermal Growth Factor Induces Egr-1 Messenger R N A and Protein in Mouse Osteoblastic Cells. Calcif. Tissue Int. 57: 450-455. Francois, V . , M. Solloway, J. W. O'Neill, J. Emery and E. Bier. 1994. Dorsal-ventral patterning of the Drosophila embryo depends on a putative negative growth factor encoded by the short gastrulation gene. Genes Dev. 8: 2602-2616. Freter, R., J. Irminger, J. Porter, S. Jones, C. Stiles. 1992. A Novel 7 Nucleotide Motif in 3' Untranslated Sequences of the Immediate Early Gene Set Mediates Platelet-Derived Growth Factor Induction of the JE Gene. M o l . Cell . Biol . 12: 5288-5300. Garcia, J., R. Gaynor. 1994. Regulatory mechanisms involved in the control of HIV-1 gene expression. AIDS 8: S3 - S17. Genini, M., P. Schwalbe, F. Scholl and B. Schafer. 1996. Isolation of Genes differentially expressed in Human Primary Myoblasts and Embryonal Rhabdomyosarcoma. Int. J. Cancer 66: 571-577. Greene, W. The molecular biology of Human Immunodeficiency Virus Type I Infection. Ther N e w England Journal of Medicine. 324: p308 - 317. Hammes, M., J. Lieske, S. Pawar, B. Spargo, F. Toback. 1995. Calcium oxalate monohydrate crystals stimualte gene expression i in renal epithelial cells. Kidney Int. 48: 501-509. Heldin, C-H. 1995. Dimerization of Cell Surface Receptors in Signal Transduction. Cel l 80: 213-223. Hengen, P. Quantitative PCR: an accurate measure of m R N A ? TIBS 20: 476-477. 73 Herschman, H. 1991. Primary Response Genes Induced by Growth Factors and Tumor Promoters. Annu . Rev. Biochem, 60: 281- 319 Hill, C , and R. Triesman. 1995. Transcriptional Regulation by Extracellular Signals: Mechanisms and Specificity. Cel l 80:199-211. Hillier, L., N. Clark, T Dubuque, K. Elliston, M. Hawkins, M. Holman, M Hultman, T. Kucaba, M. Le, G. Lennon, M. Marra, J. Parsons, L. Rifkin, T. Rohlfing, M. Soares, F. Tan, E. Trevaskis, R. Waterston, A . Williamson, P. Wohldmann, and R. Wilson. 1995. The WashU-Merck EST Project. Unpublished. Hipskind, R.A., D. Buscher, A . Nordheim, M. Baccarini. 1994. R a s / M A P kinase-dependent and -independent signaling pathways target distinct ternary complex factors. Genes Dev 8424:1803-1816. Holgado-Madruga, M., D. Emlet, D. Moscatello, A . Godwin, A . Wong. 1996. A Grb2-associated docking protein in EGF- and insulin-receptor signalling. Nature 379: 560-564. Huang, T., J. Duyster, J. Wang. 1995. Biological response to phorbol ester determined by alternative G l pathways. Proc. Natl , Acad. Sci. U S A 92: 4793-4797. Hunter, T., M. Karin. 1992. The Regulation of Transcription by Phosphorylation. Cel l 70: 375-387. Hunter, T. 1995. Protein Kinases and Phosphatases: The Y i n and Yang of Protein Phosphorylation and Signaling. Cel l 80: 225-236. Igarashi, A . , H. Okochi, D. Bradham, and G.Grotendorst. 1993. Regulation of Connective Tissue Growth Factor Gene Expression in Human Skin Fibroblasts and During Wound Repair. Molecular Biology of the Cel l 4: 637-645. Igarashi, A . , K. Nashiro, K. Kikuchi, S. Sato, H. Ihn, G. Grotendorst, K. Takehara. 1995. Significant Correlation Between Connective Tissue Growth Factor Gene Expression and Skin Sclerosis in Tissue Sections from Patients with Systemic Sclerosis. J. Invest. Dermatol. 105: 280-284. Igarashi, A . , K. Nashiro, K, Kikuchi, S.Sato, H. Ihn, M. Fujimoto, G. Grotendorst, and K. Takehara. 1996. Connective Tissue Growth Factor Gene Expression in Tissue Sections From Localized Scleroderma, Keloid, and Other Fibrotic Skin Disorders. J. Invest. Dermatol. 106: 729-733. Jacobson, A . , and S. Peltz. 1996. Interrelationships of the pathways if m R N A decay and translation in eukaryotic cells. Annu . Rev. Biochem. 65: 693-739 74 Jahner, D., and T. Hunter. 1991a. The Ras-Related Gene rhoB Is an Immediate Early Gene Inducible by v-Fps, Epidermal Growth Factor, and Platelet-Derived Growth actor in Rat Fibroblasts. M o l . Cel l Biol. 11: 3682-3690. Jahner, D., and T. Hunter. 1991b. The Stimulation of quiescent rat fibroblasts by v-src and v-fps oncogenic protein-tyrosine kinases leads to the induction of a subset of immediate early genes. Oncogene. 6:1259-1268. Joliot, V., C. Martinerie, G. Dambrine, G. Plassiart, M. Brisac, J. Crochet, and B. Perbal. 1992. Proviral Rearrangement and Overexpression of a N e w Cellular Gene (nov) in Myeloblastosis-Assiciated Virus Type-1 Induced Nephroblastomas. Mol .Cel l .Biol . 12:10-21. Joseph, L., M. Le Beau, G. Jamieson, S. Acharya, T. Shows, J. Rowley and V. Sukhatme. 1988. Molecualr cloning, sequencing and mapping of EGR2, a human early growth response gene encoding a protein with "zinc-binding finger" structure. Proc. Natl . Acad. Sci. U S A 85:7164-7168. Karnitz, L.M. and R. Abraham. 1995. Cytokine receptor and signalling mechanisms. Current Opinion in Immunology 7: 320-326. Kayne, P. 1994. Knocking D o w n a Pathway to build it up. TIGS 10:145-149. Kelly, K., U . Siebenlist. 1995. Immediate-early genes induced by antigen receptor stimulation. Curr.Opin.Immunol. 7: 327-332. Kiefer, M., R. Ioh, D. Bauer, J. Zapf. 1991. Molecular Cloning of a New Human Insulin-like Growth Factor Binding Protein. Biochem. Biophys. Res. Comm. 176: 219 - 225 Kikuchi, K., T. Kadono, H. Ihn, S. Sato, A. Igarashi, H. Nakagawa, K. Tamaki, and K. Takehara. 1995. Growth Regulation in Scleroderma Fibroblests: Increased Response to Transforming Growth Factor-(3l. J. Invest. Dermatol. 105:128-132. Kireeva, M.L., F. Mo, G. Yang, L. Lau. 1996. Cyr61, a Product of a Growth Factor-Inducible Immediate-Early Gene, Promotes Cell Proliferation, Migration and Adhesion M o l . Cel l . Biol . 16:1326-1334. Korez, A., D. Soprano, K. Soprano. 1995. Tyrosine protein kinase expression in long-term quiescent WI-38 cells following growth factor simulation. J Cel l Biochem 59: 42-52. Latinkic, B., T. O'Brien, and L. Lau. 1991. Promoter Function and structure of the growth factor-inducible immediate early gene cyr61. NAR19: 3261-3267. Lau L. and D. Nathans. 1987. Expression of a set of growth-related immediate early genes in B A L B / c 3T3 cells: Coordinate regulation wi th c-fos or c-myc. Proc.Natl.Acad.Sci. 84:1182-1186. 75 Lau, L., D. Nathans. 1991. Genes Induced By Serum Growth Factors. Molecular Aspects of Cellular Regulation. V o l . 6. Cohen and Foulkes (Ed.) p257 - 293. Lemaire, P., O. Revelant, R. Bravo, and P. Charnay. 1988. Two mouse genes encoding potential transcription factors with identical DNA-b ind ing domains are activated by growth factors in cultured cells. Proc. Natl . Acad. Sci. U S A 85: 4691-4695. Liew, C.C. 1996. c D N A s from the fetal heart. Unpublished. Lodish, H., D. Baltimore, A . Berk, S. Zipursky, P. Matsudaira, J. Darnell. 1990. Molecular Cel l Biology, 2nd ed. Scientific American Books, W . H . Freeman and Company, N . Y . Manak, J.R., and R. Prywes. 1993. Phosphorylation of serum response factor by casein kinase II: evidence against a role in growth factor regulation of fos expression. Oncogene 8: 703-711. Marshall, C. J. 1995. Specificity of Receptor Tyrosine Kinase Signaling: Transient versus Sustained Extracellulr Signal-Regulated Kinase Activation. Cel l 80:179-185. Martinerie, C , E. Veigas-Pequignot, I. Guenard, B. Dutrillaux, V. Nguyen, A . Bernheimn and B. Perbal. 1992. Physical mapping of human loci homologous to the chicken nov proto-oncogene. Oncogene 7: 2529-2534. Martinerie, C, V. Huff, I. Joubert, G. Badzioch, S.L. Saunders, and B. Perbal. 1994. Structural Analysis of the human nov proto-oncogene and expression in Wilms tumors. Oncogene 9: 2728-2732. Martinerie, C , G. Chevalier, F. Rauscher III, B. Perbal. 1996. Regulation of nov by WT-1: a potential role for nov in nephrogenesis. Oncogene 12:1479-1492. Martinerie, C , and B. Perbal. 1996. Short Communication: Genomic Structure and Chromosome mapping mouse nov gene. Genomics 38:425-429. Mason, E., C. Konrad, C. Webb and J. Marsh. 1994. Dorsal Midl ine fate in Drosophila embryos requires twisted gastrulation, a gene encoding a secreted protein related to human connective tissue growth factor. Genes Dev. 8:1489-1501. McCormick, F. 1993. H o w receptors turn Ras on. Nature 363:15-16. Montgomery, R. B., D. Moscatello, A . Wong, J. Cooper, W. Stahl. 1995. Differential modulation of mitogen-activated protein (MAP) kinase/extracellular signal-related kinase kinase and M A P kinase activities by a mutant epidermal growth factor receptor. J Biol Chem 270: 30562-30566. 76 Moodie, S v and A . Wolfman. 1994. The 3Rs of life: Ras, Raf and growth regulation. TIGS 10:44-48. O'Brien, T., G. Yang, L. Sanders, and L. Lau. 1990. Expression of cyr61, a Growth Factor-Inducible Immediate Early Gene. Mol.Cell .Biol . 10: 3569-3577. O'Brien, T., and L. Lau. 1992. Expression of the Growth Factor-Inducible Immediate Early Gene cyr61 Correlates with Chondrogenesis during Mouse Embryonic Development. Cel l Growth Diff. 3: 645-654. Pawar, S., S. Kartha, F. Toback. 1995. Differential Gene Expression in Migrating Renal Epithelial Cells After Wounding. J. Cell . Physiol. 165: 556-565. Perbal, B. 1994. Contribution of MAV-l - Induced Nephroblastoma to the Study of Genes Involved in Human Wilms ' Tumor Development. Crit. Rev. Oncogenesis 5: 589 - 613 Pilarsky, C, U. Schmidt, C. Elrich, A . Bergmann, M.P. Wirth. 1996. A partial c D N A -sequence for a putative tumor suppressor gene for prostate cancer identified by differential-display-PCR. Unpublished. Pronk, G., J.L. Bos. 1994. The role of p21ras in receptor tyrosine kinase signalling. Biochim. et Biophys. Acta 1198:131-147. Robinson, C. J. 1993. Growth Factors: Therapeutic Advances in Wound Healing. Annals of Medicine 25: 535 - 538. Reeck, G., C. de Haen, D. Teller, R. Doolittle, W. Fitch, R. Dickerson, P. Chambon, A . McLachlan, E. Margoliash, T. Jukes, and E. Zuckerland. 1987. "Homology" in Proteins and Nucleic Acids: A Terminology Muddle and a Way out of it. Cel l 50: 667. Ross, J. 1995. m R N A Stability In Mammalian Cells. Microbiological Reviews 59: 423-450. Rozengurt, E. 1986. Early Signals in the Mitogenic Response. Science 232:161-166. Ryder, K., L. Lau, and D. Nathans. 1988. A gene activated by growth factors is related to the oncogene v-jun. Proc. Natl . Acad. Sci. U S A 85:1487-1491. Ryseck, R., H. MacDonald-Bravo, M. Mattei, and R. Bravo. 1991. Structure, Mapping and Expression of fisp-12, a Growth Factor-Inducible Gene Encoding a Secreted Cysteine-rich Protein. Cel l Growth Diff. 2: 225-233. Sassone-Corsi, P. 1994. Goals for signal transduction pathways: l inking up with transcriptional regulation. E M B O J. 13: 4717-4728. 77 Scholz, G., C. Martinerie, B. Perbal, and H. Hanafusa. 1996. Transcriptional D o w n Regulation of the nov Proto-Oncogene in Fibroblasts Transformed by p60v-src. M o l . Cel l . Biol . 16: 481-486. Shaw, G., and R. Kamen. 1986. A Conserved A U Sequence from the 3' Untranslated Region of G M - C S F m R N A Mediates Selective m R N A Degradation. Cel l 46: 659- 667. Silvennoinen, O., C. Schindler, J. Schessinger, D . Levy. 1993. Ras-independent Growth Factor Signaling by Transcription Factor Tyrosine Phosphorylation. Science 261:1736-1739. Simmons, D . , D . Levy, Y.Yannoni, and R. Erikson. 1989. Identification of a phorbol ester-repressible v-src inducible gene. Proc. Natl . Acad. Sci. U S A 86:1178-1182. Smith, J., and H. Herschman. 1995. Glucocorticoid-attenuated Response Genes Encode Intercellular Mediators, Including a New C-X-C Chemokine. J. Biol . Chem. 270: 16756-16765. Songyang, Z., L. Cantley. 1995. Recognition and Scecificity in protein tyrosine kinase-mediated signalling. TIBS 20: 471-475. Sporn, M., and A. Roberts. 1993. A Major Advance in the Use of Growth Factors to Enhance Wound Healing. J. Cl in . Invest. 92: 2565 - 2566 Stacker, S., C. Hovens, A. Vitali, M. Pritchard, E. Baker, G. Sutherland, and A.Wilks. 1993. Molecular cloning and chromosomal localisation of the human homologue of a receptor related to tyrosine kinases (RYK). Oncogene 8:1347-1356. Taylor. G. 1991. Polymerase chain reaction: basic principles and automation. PCR A practical Approach, M J McPherson, P. Quirke and G.Taylor (Ed.) p i -14. White, M., C. Nicolette, A. Minden, A. Polverino, L. Van Aelst, M. Karin, and M. Wigler. 1995. Multiple Ras Functions Can Contribute to Mammalian Cell Transformation. Cel l 80: 533-541. Xin, L.W., C. Martinerie, M. Zumkeller, M. Westphal, B. Perbal. 1996. Differential expression of n o v H and C T G F in human glioma cell lines. J. Cl in . Pathol: M o l . Path. 49: M91 - M97 Yang, G., and L. Lau. 1991. Cyr61, Product of Growth Factor-inducible Immediate Early Gene, Is Associated with the Extracellular Matrix and the Cel l Surface. Cel l Growth Diff. 2: 351-357. Ying, Z., and M. King. 1996. Isolation and characterization of xnov, a Xenopus laevis ortholog of the chicken nov gene. Gene 171: 243-248. 78 Yu, C, E, Prochownik, M. Imperiale, and R. Jove. 1993. Attenuation of Serum Inducibility of Immediate Early Genes By Oncoproteins in Tyrosine Kinase Signalling Pathways. Mol .Cel l . Biol. 13: 2011-2019. Zerial, M., L. Toschi, R. Ryseck, M. Schuermann, R. Muller and R. Bravo. 1989. The product of a novel growth factor activated gene, fos B, interacts wi th J U N proteins enhancing their D N A binding activity. E M B O J. 8: 805-813. Zumkeller, W., R. Biernoth, S. Kocialkowski, C. Martinerie, P. Schofield, B. Perbal, M. Westphal. 1996. Expression and synthesis of insulin-like growth factor (IGF)-L-II and their receptors in human glioma cell lines. Int. J. Oncol. 9: 983 - 992 79 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items