\ F LIBRARY Michigan State University '\ 5. PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE ;_____ JUC » _J W MSU Is An Affirmative Action/Equal Opportunuy Institution chided-wean“! L__. _\ REGULATION OF THE HUMAN THYMIDINE KINASE GENE PROMOTER IN SERUM STIMULATED AND SV4O INFECTED CELLS BY Holger Harald Helmut Roehl A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Public Health 1991 ABSTRACT REGULATION OF THE HUMAN THYMIDINE KINASE GENE PROMOTER IN SERUM STIMULATED AND SV40 INFECTED CELLS \ BY Holger Harald Helmut Roehl Regulatory regions in the human thymidine kinase (TK) gene promoter that are responsible for activation of transcription at the 61/5 interface when quiescent cells are mitogenically induced to re-enter the cell cycle by either serum stimulation or infection with the oncogenic simian virus 40 (SV40) were identified. A set of promoter deletion mutants was generated, linked to the bacterial neomycin resistance (neo) gene and stably transfected into Rat3, murine NIH-3T3, and simian CV1 cells. Expression of the TR promoter-neo hybrid genes was studied in serum starved, serum stimulated and SV4O infected cells. Regions within the TR promoter conveying Gl/S phase specific regulation were identified, and evidence is presented suggesting that serum stimulation and SV4O infection may activate the TK promoter using partially overlapping or different pathways. A promoter fragment containing 135 base pairs (bp) upstream of the transcriptional start site is sufficient to confer serum and SV4O regulated expression to the neo reporter gene in all cell lines tested. A 67 bp human TK promoter fragment confers constitutive low expression to the neo gene in serum stimulated Rat3 and CV1 cells, induced expression in SV40 infected CV1 cells, and constitutive high expression in both serum stimulated and SV40 infected NIH-3T3 cells. These results suggest that the human TK promoter contains both negative and positive regulatory elements, some of which function in a cell type and/or mitogen specific manner. Based on these results, a model for regulation of the human TK gene promoter is presented. Fur Eugen Kern 18.4.1909 - 12.3.1989 iv ACKNOWLEDGEMENTS I would like to thank Dr. Susan E. Conrad for the advice, time and encouragment she offered during the course of this project. I would also like to thank my committee members for guidance and assistance: Drs. Jerry Dodgson, Ronald Patterson, Richard Schwartz, and John Wang. I also wish to thank past and present members of the Conrad laboratory for discussions and support, especially Michelle Anderson, Moriko Ito and Marc Carozza. Finally, I would like to thank my wife, Lisa, and my parents for their encouragement and continued support. TABLE OF CONTENTS LIST OF TABLES..... OOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOO Vii LIST OF FIGURESOOCOOOOOOCCOOOOOOOOOOOOOOOOOOOOOOOOOOO Viii CHAPTER 1: LITERATURE REVIEW THE CELL CYCLE........................... 1 S PHASE REGULATED GENES.................. 5 BIOCHEMICAL AND MOLECULAR PROPERTIES OF THYMIDINE KINASE......................... 10 REGULATION OF THYMIDINE KINASE EXPRESSION............................... 14 INTERACTION OF TK PROMOTER REGIONS WITH DNA BINDING FACTORS...................... 24 LITERATURE CITED.... ....... .............. 35 CHAPTER 2: ARTICLE: IDENTIFICATION OF A G1-S PHASE- REGULATED REGION IN THE HUMAN THYMIDINE KINASE GENE PROMTER...................... 46 ABSTRACT................................. 47 LITERATURE CITED......................... 66 CHAPTER 3: ARTICLE: MULTIPLE CIS-ACTING ELEMENTS ARE INVOLVED IN REGULATION OF THE HUMAN THYMIDINE KINASE PROMOTER IN SERUM STIMULATED AND SV4O INFECTED CELLS....... 69 ABSTRACT................................. 70 INTRODUCTION............................. 71 MATERIALS AND METHODS.................... 76 RESULTS.................................. 81 DISCUSSION........... .......... .......... 103 LITERATURE CITED......................... 111 SUMMARY AND CONCLUSIONS.. .............. .............. 116 vi CHAPTER 1 CHAPTER 2 CHAPTER 3 LIST OF TABLES Consensus sequences for transcription factor binding sites and homologous sequences in the human TK promoter...... 30 Transfection efficiencies of human TK promoter-neo constructs in Rat3 cells... 54 Effect of mitogenic stimulation on the expression of human TK promoter-neo hybrid genes in different stably transfected cell 1ines.................. 104 vii CHAPTER 1 CHAPTER 2 CHAPTER 3 LIST OF FIGURES The eukaryotic cell cycle............... De novo and salvage pathways to obtain purine nucleotides and thymidylate...... Sequence of the 5' region of the human TK gene....... ........ .................. Structure of human TK promoter-neo hybrid gene constructs ...... ............ Northern blot analysis of human TK promoter-neo gene and TK minigene constructs in pools of stably transfected Rat3 cells.................. Determination of the transcriptional start site in the human TK promoter-neo gene constructs with a primer extension assaYOOOOOOOOOOOOOOOOOOOO0.000.000.0000. Structure of human TK promoter-neo hybrid gene constructs.................. DNA synthesis in growth arrested NIH-3T3/a cells after serum stimulation or SV4O infection....................... Northern blot analysis of human TK promoter-neo gene constructs in pools of stably transfected NIH-3T3/a cells... Northern blot analysis of human TK promoter-neo gene constructs in pools of stably transfected NIH-3T3/b cells... Determination of the transcriptional start site in human TK promoter-neo gene constructs using a primer extension assay......................... Electrophoretic mobility shift analysis using nuclear extracts prepared from NIH-3T3/b cells at various times after mitogenic treatment..................... viii 11 13 52 57 62 84 86 89 92 95 99 Northern blot analysis of human TK promoter-neo gene constructs in pools and single colonies of stably transfected CV1 cells......... .......... 101 ix CHAPTER 1 LITERATURE REVIEW The Cell Cycle Normal eukaryotic cells exist in either a proliferative or a non-proliferative, quiescent state. Proliferative cells progress through a chain of complex and tightly controlled events called the cell cycle. During one round of the cell cycle a single parent cell prepares for and divides into two daughter cells. The eukaryotic cell cycle is composed of four phases, G1, S, 62 and M phase (92)(Fig. 1). During Gl phase, cells prepare for DNA synthesis, which occurs in S phase. 61 phase itself is divided into several sequential subphases termed competence, entry, progression and assembly (93). During competence, immediate early genes (i.e. genes that do not require de novo protein synthesis for their expression) are transcribed and the chromatin structure changes. During entry, polysomes increase, and progression is marked by rapid protein synthesis. Finally, assembly occurs just prior to S phase, and it is assumed that the enzymes needed for DNA synthesis are assembled into a functional complex in the nucleus. A critical regulatory point of the cell cycle, the restriction (R) point (91, 92, 93), lies in G1 phase. At 1 (30 state x"\ / 14 Figure 1.“. The eukaryotic cell cycle (taken from Darnell et a1. (13)). 3 the R-point, a cell decides either to progress to S phase, initiate DNA replication and complete the cell cycle or to cease proliferation. In tissue culture cells, this decision is dependent upon growth conditions, such as the availability of serum (133) or cell density (92). Once a cell has passed the R point it is irreveresibly committed to initiate DNA replication and progress through one round of cell division. The molecular and biochemical events involved in growth control at the R-point remain to be elucidated. In the budding yeast Saccharomyces cerevisiae a regulatory point called START has been defined. At START, a 61 phase cell determines whether or not it will proceed to S phase, and it has been suggested that START is equivalent to the R-point (61). In _S_. cerevisiae, transition from 61 to S phase is dependent upon activation of the protein kinase p34CDC28 (86, 97) by Gl specific cyclins (31, 98, 131). RecentLy , mammalian equivalents of the yeast G1 cyclins appear to have been isolated (73, 82). After a cell leaves 61, it enters S phase during which the entire DNA content of the nucleus is replicated (92). Additionally, the replicated DNA is assembled into nucleosomes and the chromosome scaffolds are duplicated (57). Upon completion of S phase, the cell enters GZ phase where it prepares for mitosis in M phase. Cell morphology changes and the cytoskeleton begins rearranging (92). 4 Entry into M phase is also tightly regulated. Regulation at this point is dependent on activation of the p34Cdcz protein kinase which requires association with cyclin (61, 87). In M phase, actual cell division takes place. During nuclear division, chromosomes which were duplicated in S phase are segregated (75). Nuclear division is followed by cytoplasmic division, resulting in two G1 phase daughter cells. After division, daughter cells can either continue to proliferate or enter a quiescent, non-proliferating state termed G0 (92, 93). In vivo, terminally differentiated cells such as neurons are in GO phase. Tissue culture cells can also enter G0 phase under adverse growth conditions such as serum depletion or contact inhibition (122). G0 phase cells decrease in size because many protein and RNA molecules used during proliferation are degraded and are not resynthesized. A new set of RNA and protein molecules appear after cells become quiescent, and it is thought that these proteins are responsible for maintaining the G0 state. G0 phase tissue culture cells can be mitogenically stimulated to re-enter the cell cycle in GI phase, and continue to proliferate. Tissue culture cells are an ideal model system for studying how 60 phase cells re-enter the cell cycle and how the cell cycle is controlled at its various stages. In cancer cells the normal regulatory control points are overridden and uncontrolled proliferation ensues. Thus, 5 the study of normal cells and their comparison to cancer cells will allow one to identify and understand mechanisms of growth control. § Phase Regulated Genes The mammalian cell cycle is tightly regulated, and its events occur sequentially in a cause/effect relation. Specific sets of genes must be induced and repressed at various stages to ensure proper progression through the cell cycle. The regulatory mechanism(s) governing the availability of macromolecules required for progression through a given cell cycle phase is unknown. For example, as cells enter S phase, the activities of a number of enzymes and molecules involved in DNA synthesis increase. Among' these :molecules are thymidine kinase (TK), thymidylate synthase (TS), dihydrofolate reductase (DHFR) and replication dependent histones. The interesting question of how these molecules are regulated is being investigated by a number of laboratories. In order to conduct such studies, it is necessary to synchronize cells. A widely used method involves growth arresting confluent cells in G0 phase of the cell cycle by serum starvation (23, 49, 92, 120). Another method involves using a chemical block such as sodium butyrate, which arrests cells in G1 phase (35). There are also temperature sensitive hamster cell lines available which arrest in 61 phase at the non-permissive temperature (45, 6 124). A method which has recently become more popular is synchronization by centrifugal elutriation (22). In this method continously growing cells are separated according to their size. Since the size of a cell increases as it proliferates and traverses the cell cycle, this separates cells based on their position in the cell cycle. The regulation of S phase specific genes has been studied in numerous laboratories using several of these synchronization methods. A number of S phase regulated genes have been studied and their regulation was shown to be complex. Regulation at the transcriptional as well as at the post- transcriptional level has been observed and results reported. have often lbeen conflicting: ‘When. the work presented in this dissertation was initiated, it was known that when cells synchronized by serum starvation are induced to re-enter the cell cycle by addition of fresh serum or infection with SV40, TK enzyme levels remain low throughout Gl phase, increase sharply at the Gl-S boundary, and remain elevated throughout S and G2 phase (5, 41). The level of TK enzyme is paralleled by the level of TR mRNA, which increases approximately 10-20 fold as cells enter S phase (114, 118). It was shown in this laboratory and by others that the TK gene is regulated at the transcriptional (45, 66, 118, 124,) as well as at the post-transcriptional level (10, 39, 66). Studies on thymidylate synthase (TS) enzyme showed that 7 its activity is very low in quiescent cells, but increases up to 20 fold during S phase when mouse fibroblasts (40, 84) or human fibroblasts (4) are stimulated to re-enter the cell cycle by serum stimulation. TS enzyme activity is paralleled by TS mRNA levels when cells are synchronized by serum starvation or mitotic shake off selection (4, 40, 83). A result contradicting these findings was reported using mouse cells synchronized by centrifugal elutriation (38). In that study, no changes in TS mRNA levels were detected during the cell cycle. TS enzyme activity was maximal in S phase, however, which suggested translational or post-translational control. No increase in transcription of the TS gene in resting versus growing cells was observed by Ayusawa et a1. (4), and a moderate 3 fold increase in transcription during S phase was reported by Jenh et a1. (40). Most published reports suggest that the increase of TS mRNA during the cell cycle is mainly due to regulation at the post-transcriptional level (4, 40, 43, 83). Interestingly, in Saccharomyces cerevisiae, the TS gene appears to be primarily controlled at the level of transcription, indicating that TS regulation might differ significantly between yeast cells and mammalian cells (74). Discrepancies in the results obtained might be explained by different regulatory mechanisms in different cell lines and by different methods of synchronizing cells. Regulation of dihydrofolate reductase (DHFR) gene 8 expression has been investigated in several laboratories. Cell lines containing amplified DHFR genes were used for these studies because DHFR expression in a single copy cell line is extremely low. Leys et a1. (64, 65) reported a 3 fold increase in DHFR mRNA after quiescent mouse cells were stimulated to re-enter the cell cycle. No increase in DHFR specific transcription and no change in DHFR mRNA half life was observed. It was suggested that DHFR expression is regulated mainly at the post-transcriptional level by changes in DHFR hnRNA stability in the nucleus. A different conclusion was reached by others, who reported that DHFR is transcriptionally regulated. An increase in DHFR gene transcription during S phase was observed in mouse cells after G0 phase cells were serum stimulated (105, 132). Farnham and Schimke (17) reported that the mouse DHFR transcription rate is low in G1 phase, increases transiently 7 fold at the beginning of S phase and decreases immediately thereafter. Again, the discrepancy in the different reports was postulated to be due to the use of different cell lines, which might have been in different physiological states, and to differences in cell synchronization procedures. Feder et al. (18) suggested that DHFR cannot be considered a cell cycle regulated enzyme since variations in growth conditions prior to centrifugal elutriation can lead to drastic increases in DHFR.:mRNA, regardless of the. cell cycle stage. Thus, the mechanisms of DHFR gene regulation in 9 continuously' cycling' cells ‘versus serum stimulated quiescent cells await further study. Replication dependent histone genes are the best characterized class of S phase dependent genes. They are expressed coordinately with DNA synthesis. Replication dependent histone mRNAs increase during S phase, and are highest in mid S phase when DNA replication is maximal (1, 33, 72). Expression of histone genes is regulated at both transcriptional and post-transcriptional levels. Transcription increases 3-5 fold as as cells enter S phase, and is reduced rapidly when DNA synthesis is inhibited (1, 33, 113). Both S phase specific regulatory regions within several histone gene promoters (3, 32, 54, 89, 109), and several putative S phase-specific trans- acting factors which are thought to interact with these regions, have been identified (9, 20, 32, 54). Comparison of the 5' flanking regions of different subtypes of replication. dependent. histone genes from. a ‘variety of organisms reveal only subtype specific conserved sequences. This suggests that the coordinate transcriptional regulation of different histone gene subtypes may be regulated by different mechanisms. Post-transcriptional regulation, primarily at the level of mRNA degradation, plays an important role in controlling levels of histone mRNA. The half life of histone mRNA increases 5 fold during S phase, and decreases rapidly when DNA synthesis is inhibited (33, 10 113). Degradation of histone mRNA is mediated by a stem loop structure at the 3' terminus, a structure which all replication dependent histone mRNAs have in common (60, 90). Histone mRNA. must be translated to the normal termination codon for efficient degradation to occur, and the distance between the termination codon and the stem loop is important for degradation (27, 102). It was also shown that the 3’ end of histone mRNA, including the stem loop structure, is important in post-transcriptional regulation at the level of RNA processing in the nucleus (70). By comparing the results reviewed above, it becomes apparent that no single mechanism has been identified that controls the concerted expression of S phase regulated genes. For this dissertation, the transcriptional regulation of the human TK gene was investigated. Elucidation of the mechanism of transcriptional control of the human. TK: gene will contribute to identifying and understanding pathways which might be common in regulating all S phase specific genes. fiigghemical and Molecular Properties of Thymidine Kinase Thymidine kinase is an enzyme in the salvage pathway for pyrimidine nucleotide biosynthesis (see Fig. 2). Thymidine kinase catalyzes the phosphorylation of thymidine (dT) to form thymidine-monophosphate (dTMP) (103). The same product can also be formed in the de novo 11 5-Phosph0nbosyl-1-pyrophospnate (PRPP) ' S-Ph05phoribova-1-amine F //- Slocuec 0v anxdoxares :./n’—CHO from ; tetrahydrofolare Formvlglvcinamlde ribonde tFGAR) i ! ‘lf‘ NH: from glutamlne Formvlglvcmamudme nbotnde (FGAM) , ,,--- Bvocueo bv EDTA’OHHCS Undylate «UMP! : ‘ I ( Bloc-co by L,/—_— CHO from I antltolates . ‘ tetrahydrofolate {NU-slag ,/ —CH3 from udelf: [ lnosinate (IMP) ' gags; tetrahydrofolate d8. —- (a nucleonde) ' / ’ 'D-b' ‘ ' l L ...NA. \\G ? ~ ( x' * / ' , / uanylate (GMP) HGPRT i Adenvlate (AMP) Thymndylate (TMP) ‘ » APRT ',. IHGPRT IDOSIOE E1 (adenme ADP (hvooxenlhme- (a nucleosme) H phosphoruoosy: "\ ATP Guanosine lgxzsgnheonbosv' ’ a]: transmittal- \\ {transterase} Ademne R'bose phoscmale unvmuo-ne G ‘\- R'bose h h //' . uamne -— I p 0st ate '— Hypoxanthme ' from PRPP Thymudme Figure 2. (a base) ". I De novo and salvage pathways to obtain purine nucleotides and thymidylate (15)). (taken from Darnell et al. 12 pathway using thymidylate synthase, which covalently attaches a methyl group to uridine-monophosphate (dUMP) to form dTMP (103). The molecular weight of the native human TK enzyme was determined to be 96,000 using gel filtration and glycerol gradient sedimentation (111). When the purified enzyme was electrophoresed under denaturing conditions, one predominant peptide with a molecular weight of 24,000 was identified. This suggests a tetrameric structure for the enzyme in solution. Even though TK enzyme activity was described several decades ago, the isolation and cloning of TK genes was not feasible before the availability of molecular biology techniques. Several groups independently isolated human TK genomic and cDNA clones (6, 7, 58, 67). Additionally, TK genes have been cloned from mouse (35, 68), hamster (62, 63), and chicken (77). Virally encoded TK genes from Herpes simplex virus (76), fowlpox virus (8), and vaccinia virus (129) have also been cloned. Flemington et a1. (19) completely sequenced a 12.9 kb human TK gene and its flanking regions. The gene contains seven exons, which code for a 1430 bp mRNA excluding polyadenylation. This is consistent with the length of 1.5 kb reported for mature TK mRNA (6, 58). Several putative transcription elements and two transcriptional start sites were identified. The sequence of the human TK promoter region is shown in Fig. 3 and putative regulatory -436 ATTCCAAACT -376 CCTACCTCTG -316 ATCCATCATG -256 CGCCTCGCTC -196 CCACCCCGAT -136 CGTCCCGGAT -76 CCGGGCGCTG -16 TTTAAAGCGG +45 GGTTCGTGAA Figure 3. Putative translational start sites are marked. transcriptional start sites, site. -426 AATAAATGAG -366 CAGACATCTT -306 GCGTCTACAG ~246 GCCCCTTTAA -186 CAGCCACGTC -126 TCCTCCCACG -66 ATTGGCCCCA ~6 TCGGCGCGGG +55 CTTCCCGGAG Sequence of the 5’ region of the human TK gene. regulatory -416 CTAACTCCGC -356 CTTCCAAGGA -296 CCGCATGGGC -236 ACTTGGTGGG -l76 CATCGCCCTG -116 AGGGGGCGGG *huum -56 TGGCGGCGGG +1 +5 AACCAGGGGC --> +65 GCGCAATGAG -¢> : inverted CCAAT boxes, elements 13 -406 CCCAGCCCCT ~346 ACCTTGCTTG -286 GTGCGTCCCT -226 CGGACCGAGG -166 ATTTCCAGGC -106 CTGCGGCCAA -46 GCCGGCTCGT +15 TTACTGCGGG --> +75 CTGCATTAAC and -396 -386 TAGTCCCTCC CTGCAATCCA -336 -326 GGCAAACCCA CACCAGACAC ~276 -266 CTGTTTATAT GGCCAGAGCC -216 -206 CGGGGCCTCA GATCCAGGCC -156 -146 CCTCCCAGTC CCTGGGCGCA -96 -86 ATCTCCCGCC AGGTCAGCGG -36 -26 GATTGGCCAG CACGCCGTGG -74 +25 +35 ACGGCCTTGG AGAGTACTCG +85 +95 TGCCCACTGT GCTGCCTTGC transcriptional and *fitfi: GC-box, TATA box, --->: translational start 14 elements in the region between -135 bp and +1 bp are marked. The transcriptional start sites are located at +1 bp and +7 bp. There is also a TATA box located at -25 bp, two inverted CCAAT boxes located at -75 bp and -44 bp, and a GC box located at -122 bp. The isolation of TK gene clones, and the determination of their sequence and flanking regions, provided the basis for studying the transcriptional and post-transcriptional regulation of the gene during the cell cycle. Regulation of Thymidine Kinase Expression There are several early reports concerning the regulation of thymidine kinase enzyme levels. In 1965, Eker (16) reported that there are different levels of TR activity in rapidly proliferating than in non- proliferating cells. Several years later, it was shown that when G0 phase cells are induced to re—enter the cell cycle by the addition of mitogenic agents such as fresh serum or infection with SV40, TK enzyme levels are low during G0 and G1 phase, increase as cells enter 8 phase, and stay elevated throughout G2 phase (5, 41, 47, 81). The use of RNA and protein synthesis inhibitors, such as actinomycin D and cycloheximide, demonstrated that the induction of TR enzyme activity is dependent on protein synthesis in serum stimulated cells (41, 47, 81). TK enzyme activity, whose induction is not sensitive to DNA synthesis inhibitors such as hydroxyurea or cytosine 15 arabinoside, increases at the beginning of S phase. TK enzyme activity decreases in M phase and this decrease is sensitive to DNA synthesis inhibitors (5). In the reports described above, cells growth arrested in GO phase by serum starvation were induced to synchronously re-enter the cell cycle by mitogenic stimulation. In the work presented in this dissertation, cells were also synchronized by serum starvation. G0 phase cells were then mitogenically stimulated by two different methods: i. serum stimulation and ii. infection with the DNA tumor virus SV40. Addition of fresh serum to contact inhibited cells increases first the rate of mRNA synthesis, then the rate of protein synthesis finally leading to DNA synthesis (49, 122). Others have shown, by directly measuring TK enzyme levels, that serum stimulation of G0 phase cells induces TK enzyme activity as cells enter S phase (5, 41). The increase in TK enzyme levels depends on transcription and de novo protein synthesis. As alluded. to above, infection. with. SV40 can also mitogenically stimulate G0 phase cells. SV40 is a well characterized. small DNA. tumor ‘virus (123) that. has a double stranded circular genome of 5243 bp. The genome encodes the early proteins large and small T antigen, and the structural late proteins VP1, VP2 and VP3. SV40 is oncogenic, and initiates a variety of alterations of cellular functions in infected cells. It can override 16 normal cellular proliferation control mechanisms. Thus, a very interesting field of study is the regulation of host genes by the virus. SV40 infection of G0 phase cells causes cells to re-enter the cell cycle, and induces cellular DNA synthesis in the absence of normal serum factors (24, 48, 49, 94, 99). SV40 infection also induces TK enzyme activity during S phase (49, 94). Using SV40 tsA mutants, Postel and Levine (95) showed that the SV40 transforming protein, large T antigen, is required for the induction of TX enzyme activity following viral infection. Large T antigen can specifically' bind. viral (14) and cellular' DNA (88), and influence the pattern. of both cellular transcription and enzymatic activity (107, 108). Supporting the involvement of large T antigen in TK enzyme induction is the observation that microinjection of large T antigen into tissue culture cells also induces cellular DNA synthesis (121). DNA synthesis in these experiments was measured by 3H-thymidine incorporation into DNA. This indirectly shows that TK enzyme is also induced, because 3I-I-dT needs to be phosphorylated by thymidine kinase in order to eventually' be incorporated into DNA, Taken together, these results demonstrate that SV40 large T antigen can induce cellular TK enzyme as cells enter S phase. Early reports of TR regulation were mainly descriptive due to limitations in the available techniques. With the advent of molecular biology, the regulation of TR gene l7 expression has been studied in a number of systems using several experimental approaches. Cloned TK genes have been transfected into TK- cells, and their regulation has been studied. The regulation of endogenous TK genes in TK+ cells has also been investigated. These studies provide evidence that TK gene expression is regulated at both transcriptional and post-transcriptional levels. Since this dissertation focuses on transcriptional regulation, the evidence for transcriptional regulation of the TR. gene will be reviewed in greater’ detail than evidence for post-transcriptional regulation. Some of the articles cited were published while this study was in progress, and will therefore confirm and overlap with data presented in Chapters 2 and 3. Kreidberg and Kelly (52) were the first to genetically analyze the human TK promoter and define regions required for activity. A series of promoter deletion mutants was generated and linked to a human genomic TK clone. The deletion mutants were tested for their ability to stably transform mouse L TK- cells to the TK+ phenotype and their transformation frequency ‘was expressed. relative ‘to 'the transformation frequency of a wild type construct which was set at 100%. Using this approach, a minimal promoter region was identified that contained 83 bp upstream of the transcriptional start site and had a transformation frequency of 76%. A construct containing 53 bp upstream of the transcriptional start site had lost nearly all 18 promoter activity and a transformation frequency of 1.7% was observed. In a similar study, Arcot et al. (2) defined human TK promoter regions required for maximal levels of expression in transiently transfected growing cells. Promoter deletions were linked to the bacterial chloramphenicol acetyltransferase (cat) gene, and cat activity was assayed in transiently transfected mouse L cells. A promoter fragment containing 139 bp upstream of the transcriptional start site retained 64% of activity relative to a "full length" 457 bp promoter fragment. The next two smaller promoter deletions contained 88 bp and 58 bp, and possessed 28% and 19% activity, respectively. At the onset of this study, several pieces of evidence indicated that the TR gene is transcriptionally regulated after mitogenic stimulation of quiescent cells. First, using simian CV1 cells, it was shown that endogenous TK mRNA levels are highly induced following serum stimulation or SV40 infection, and that the pattern parallels that of TR enzyme levels (118). Second, using nuclear run on transcription assays, a 6-7 fold increase in TX transcription was detected at the G1/S boundary in serum stimulated and a 3-4 fold increase in SV40 infected CV1 cells (114, 115). This result was confirmed by Coppock and Pardee (10), who reported that the rate of TR transcription increased 2-4 fold between G0 and S phase in serum stimulated BALE/c 3T3 cells. Lieberman et a1. (66), also using nuclear run on assays, showed that the rate of l9 transcription of the endogenous TK gene increased 11 fold in serum stimulated mouse L929 cells. However, they detected only a 2 fold increase in the rate of transcription of a cloned genomic murine TK gene which was transfected into mouse L-M TK- cells. The aim of the work described in this dissertation was to define gflgracting regions within the human TK promoter that are responsible for transcriptional activation of the TR gene at the G1/S boundary after mitogenic stimulation of quiescent cells. Results described here and by others have continued to provide evidence for transcriptional regulation by studying the human TK promoter linked to non-cell cycle regulated. genes, such. as the bacterial neomycin resistance (neo) or cat gene (45, 124). A human TK promoter fragment containing 444 bp upstream of the transcriptional start site is sufficient to convey Gl/S phase regulation to a linked reporter gene. In initial experiments done by Travali et al. (124), a 444 bp human TK promoter fragment was linked to the cat gene and stably transfected into BALB/c 3T3 derived cells. Cat activity, measured at various times after serum stimulation, increased drastically during S phase. Similar experiments were conducted in parallel in Rat3 cells this laboratory. The results reported by Travali et al. (124) could not be reproduced due to the fact that cat activity was very high even in quiescent cells before serum stimulation. .A possible explanation for this is that the bacterial cat 20 protein is very stable in the Rat3 cell line used in this laboratory and that GO phase Rat3 cells maintained in low serum die before cat protein decreases to basal levels. Kim et al. (45) showed that a 444 bp human TK promoter fragment linked to the neo gene confers cell cycle dependent regulation to the neo mRNA. These experiments were conducted using a temperature sensitive Chinese hamster cell line which was released from its temperature sensitive block in G1 phase. Deletion analysis revealed that the region between ~67 bp and -444 bp is responsible for conveying Gl/S phase specific expression to the linked neo gene. A triple fusion gene was then constructed linking the human TK promoter region from -67 bp to -444 bp to a truncated Herpes simplex virus TK promoter (which is not expressed in a cell cycle regulated manner) fused to the neo reporter gene. The human TK fragment was able to confer cell cycle regulation to the heterologous promoter in an orientation dependent manner. Using' a set. of human. TKI promoter' deletion. mutants linked. to the neo gene, the experiments described in Chapter 2 of this dissertation located the Gl/S phase regulatory region of the human TK promoter in serum stimulated Rat3 cells to a 68 bp fragment between -67 bp and -135 bp upstream of the transcriptional start site. This result was confirmed by others, who also reported that sequences from -67 bp to -135 bp are sufficient to confer Gl/S phase regulation on a heterologous promoter 21 (46) . Lipson et al. (69) reported that a human TK promoter fragment which contains 83 bp upstream of the transcriptional start site is partially Gl/S phase regulated. This suggests that sequences necessary for Gl/S phase specific regulation may be located both upstream and downstream of -83 bp. To date no laboratory has reported the identification of a Gl/S phase specific regulatory "motif" in the human TK promoter region. The results reviewed above describe transcriptional regulation of the TK gene when growth arrested cells are released to progress through the cell cycle. The question of whether or not the same regulatory mechanism(s) are responsible for periodic increases in TK enzyme activity in continously cycling cells was investigated by Sherley and Kelly (112). Synchronous populations of HeLa cells were obtained by centrifugal elutriation of exponentially growing cells. Under these conditions, the change of TR enzyme levels during the cell cycle appears to be due to post-transcriptional mechanisms. TK mRNA levels increased less than 3 fold between G1 and G2 phase while TK enzyme levels increased about 15 fold. TK mRNA is translated with 10 fold higher efficiency in S phase than in 61 phase. Additionally, TK protein stability decreases drastically at the time of cell division. These findings appear to contradict the evidence for transcriptional regulation of the TK gene described above and raise the issue of whether the gene is transcriptionally regulated 22 at all during a normal cell cycle. One possible explanation is that different TK gene regulatory mechanisms may be operating in quiescent cells which are serum stimulated and in continuously cycling cells. Alternatively, it is conceivable that there is a high basal level of TX mRNA in continuously cycling cells due to the relatively long half life of TK mRNA (10, 115). TK mRNA, not yet completely degraded from the previous cell cycle, might be present during G1 phase thus accounting for the less than 3 fold increase in TK mRNA levels during the cell cycle. This issue will only be conclusively resolved by performing nuclear run on transcription experiments using nuclei isolated from elutriated cells. Besides Sherley and Kelly’s results (112), cited above, there are additional reports providing evidence for post- transcriptional regulation of the TK gene. These studies indicate that there are multiple post-transcriptional mechanisms regulating the TK gene on several levels. Among them are nuclear processing, differential mRNA half life, translational regulation and changes in the rate of TK enzyme degradation. In BALB/c 3T3 cells, a drastic change in nuclear processing of TK heterogenous nuclear RNA was observed at the Gl/S boundary (30). The half life of mature TK mRNA also changes between GO/Gl and S phase by a factor of three (10). Evidence for translational control is provided by the fact that TK enzyme is always highest in S phase, even when there are high levels of TK 23 mRNA in G1 phase (39, 124). Translational control was also observed when mouse muscle cells transfected with the chicken TK gene undergo differentiation (29). There was an uncoupling between TK mRNA levels, which were high, and the rate of TK protein synthesis, which was reduced 10 fold, as myoblasts commit to terminal differentiation. Finally, the rate of degradation of TX enzyme starts to increase drastically at the end of M phase (112). The carboxyterminal 40 amino acids of the human TK protein appear to be responsible for the specific degradation of the protein at the end of mitosis (44). Taken together, these reports provide convincing evidence for post- transcriptional regulation of the TK gene. Thus, the increase in TK enzyme activity during S phase after mitogenic stimulation of quiescent cells is a combination of transcriptional and post-transcriptional regulation. As alluded to above, the aim of this dissertation was to define gig-acting elements in the TK promoter that convey Gl/S phase transcriptional regulation to the gene when quiescent cells are either serum stimulated or infected with SV40. When this study was initiated, it had recently been established that part of the regulation of the TK gene was at the level of transcription. Putative transcription elements had only been identified by virtue of their sequence homology to known regulatory elements, and no systematic analysis of the promoter had been reported. It was hoped that the results of this study 24 would contribute to the understanding of whether serum stimulation and SV40 infection activate and/or induce different transcription factors that interact with different regions of the promoter, or whether both mitogenic treatments use the same pathway for promoter activation. The results obtained by investigating TK gene transcription will contribute to our understanding of whether or not a general, underlying mechanism(s) exists by which S phase specific genes are regulated. Interaction of TK promoter regions with DNA binding factors Transcription of RNA polymerase II genes, such as the human TK gene, is dependent on specific interactions of transcription factors with DNA sequences within a promoter. Most, but not all genes transcribed by RNA polymerase II have an AT rich region which contains a TATAAA consensus sequence located between ~20 bp and ~30 bp upstream of the transcriptional start site (104). Interaction of the transcription initiation factor TFIID with the TATA box is reponsible for properly positioning RNA polymerase II so that transcription will be initiated at the transcriptional start site at +1 bp. A CCAAT box is located between ~60 bp and ~80 bp in many promoters. A variety of different CCAAT binding proteins (15), which regulate transcription by interacting with this sequence, have been identified. In recent years a large number of 25 less common sequence specific DNA binding proteins that activate or repress transcription in mammalian cells have been characterized (80). Detailed investigations into known transcription factors have yielded exciting results which are briefly summarized below (80). Binding properties of DNA binding proteins are usually assayed in vitro. Transcriptional activities of sequence specific DNA binding proteins are measured by transiently transfecting genes encoding wild type and mutant transcription factors into tissue culture cells. Promoters which have binding sites for the factor under investigation are linked to reporter genes and cotransfected. These analyses have revealed a great variety of transcriptional activation domains and DNA binding domains, which are separable from each other. There are at least three different classes of DNA binding domains in the transcription factors described to date. The first class contains a zinc finger motif, of which there are actually two distinct subclasses. The first subclass consists of a stretch of 30 amino acids with two cysteine and two histidine residues, which are 2+ ion in a tetrahedral arrangement. stabilized by a Zn The transcription factor Spl belongs to this class and it has three tandem zinc fingers at its carboxy terminus (11, 42). The second subclass of zinc fingers binds zinc through two pairs of conserved cysteines rather than two cysteines and two histidines. Examples of this subclass 26 are steroid receptors, such as the glucocorticoid receptor (21, 110). The amino acids in the zinc finger motif are highly conserved but different zinc finger proteins have different DNA binding specificities. Therefore, it is thought that zinc fingers provide a structural framework for binding and that the amino acids conveying binding specificity are positioned elsewhere. The second class of DNA binding domains is the leucine zipper (56). A leucine zipper consists of an idealized a-helix with four leucine residues at intervals of seven amino acids. This region is preceded by a 30 amino acid basic region. The leucine zipper is needed for homo- or heterodimerization, and subsequent DNA binding of transcription factors. The actual DNA binding domain appears to reside in the 30 amino acid basic domain. It has been proposed that the leucine zipper juxtaposes the basic regions of two polypeptides in a manner suitable for sequence specific recognition of DNA. The leucine zipper motif itself does not appear to be involved in DNA binding. An example of this class of DNA binding proteins is C/EBP, a DNA binding protein isolated from rat liver nuclei, which exists in solution as a stable dimer (55, 56). Another example is the c~fos/c~jun heterodimer, which forms via a leucine zipper (51, 128). Only the c~jun/c-fos heterodimeric complex can selectively bind to the AP1 recognition site (59), c~fos alone cannot bind. The third class of DNA binding domains is the homeo 27 domain. A homeo domain consists of a region of approximately 60 amino acids which contains a helix-loop- helix DNA binding motif. Examples of this class of DNA binding proteins are the ubiquitously expressed Octl, the lymphoid specific Oct2 and the mouse maternal factor Oct3 (34, 101, 106, 119). All three factors have a homeo domain and a POU-specific box which together constitute the POU domain (100). After a DNA binding protein is bound to its specific DNA recognition sequence, the protein's transcriptional activation domain can exert its influence on transcription. Three different types of transcriptional regulation domains have been identified. The first type consists of an acidic region which has a net negative charge and may form amphipathic a~helical structures. This acidic region was first identified in the yeast transcription factors GCN4 (36, 37, 96) and GAL4 (71, 96). GAL4 has two short acidic regions, both of which can activate transcription. Those two regions have no sequence homology to each other, or to the GCN4 acidic activation domain. GCN4 has an acidic region 19 amino acids in length, which functions as the transcriptional activation domain. It has also been reported that the jun protein has an acidic domain that can activate transcription in yeast (117). The second type of transcriptional activation domain consists of a glutamine rich region, first identified in 28 Spl (11, 12). The two regions of an Spl molecule mediating transcriptional activation to the greatest extent have a very high (25%) glutamine content and very few charged amino acids. Glutamine rich regions have also been found in Octl (119) and AP2 (130), but it is not clear whether these regions are involved in the transcriptional activation mediated by these factors. The third type of transcriptional activation domain consists of a proline rich region. This type of domain has been identified in CTF/NF-I, a family of related polypeptides that activate both transcription and DNA replication (78). 'The transcriptional activation domain resides in the carboxy terminal region of CTF/NF-I, which has a very high (25%) proline content. The transcriptional activation domain of c~jun consists of a negatively charged domain which is proline rich (116). It has also Jbeen reported. that certain steroid receptors contain proline rich sequences in regions of the molecule that function as transcriptional activators (28, 53). There are a number of other DNA binding proteins whose DNA binding domains and transcriptional activation domains have been investigated, but they do not have striking homology to any one of the domains described above. Transcriptional activation domains are defined somewhat loosely and it appears that their three dimensional structure, due to net negative charge or bends in the protein introduced by proline, might be more important for 29 activation than their actual amino acid sequence. Undoubtedly, new transcription factors will be identified and cloned in the future which will lead to the identification of presently unrecognized new DNA binding motifs and will define transcriptional activation domains more accurately. To fully understand the mechanism of transcriptional regulation of the human TK gene, it will be necessary to define both gig-acting DNA elements in the promoter, and sequence specific DNA binding proteins that interact with these sequences. The human TK promoter has two transcriptional start sites located at +1 bp and +7 bp (Fig. 3). Several putative transcriptional regulatory elements in the region between ~135 bp and +1 bp have been identified by sequence homology (Table 1) . As stated previously, there is a TATA box at ~25 bp, two inverted CCAAT boxes located at ~75 bp and ~44 bp, and a GC box (25, 26) at ~122 bp. There are eight sequences that have homology to the consensus sequence for SV40 large T antigen binding located between ~134 bp and ~48 bp, and there are six sequences that have homology to the GC~box consensus sequence located between ~134 bp and ~11 bp. Additionally, there is one sequence each which has homology to the consensus sequence for binding sites of AP2 (79) at ~70 bp and AP3 (79) at ~30 bp. Treisman (125, 126) has shown that a region of dyad symmetry in the c~fos promoter is required for 3O Aon-n\mnomuv hose .nuf 600 ul of DME per 10 cm plate. Infection was for one hour at 37°C. At the end of infection, virus was removed and the original low serum medium was added back to the cells. Plasmid Constructions, The plasmids used (444-Neo, 135-Neo, 67-Neo) were described previously (23). 77 DNA Transfggtions. CaPO4 transfections were carried out as described by Wigler (29). Stable transfectants were selected in medium containing 800 ug/ml G418 (NIH-3T3 cells) or 400 ug/ml G418 (CV1 cells). After resistant colonies were clearly visible, 10-50 colonies per construct were pooled and expanded for analysis. Prepazation of Total RNA. Total RNA was prepared from tissue culture cells by the following modification of the method of Favaloro et al. (7). Cells were washed once with PBS without calcium and magnesium, and were lysed with 1 ml of lysis buffer (100 mM Tris-HCl, pH 7.5, 12 mM EDTA, 150 mM NaCl, 1% SDS, 200 ug/ml proteinase K) per 10 cm plate. Lysed cells were washed off the plate and cellular DNA was sheared by several passages through a 22 gauge needle. The lysate was incubated at 37° C for 45 min, nucleic acids were extracted twice with phenol/chloroform and ethanol precipitated. The precipitate was resuspended in 400 ul TE. Four ul of 1 M MgClz, 0.5 ul of RNasin (40 U/ul) and 1 ul of 10 U/ul RNase free DNase were added and the solution was incubated at 37°C for 45 min. RNA was extracted twice with phenol/chloroform and ethanol precipitated. The precipitated RNA was resuspended in TE and its concentration was determined by reading the absorbance at 260 nm. \° ,- , ;_- -J. 578 11° -c - .c'o , . Toizatio 5 Using standard techniques (2), equal amounts of total RNA (20 ug) were separated in 1% agarose formaldehyde gels 78 containing 500 ng/ml ethidium bromide. After electrophoresis, gels were photographed to ascertain that equal amounts of RNA were loaded in each lane. RNA was transferred to nitrocellulose filters by blotting overnight. Prehybridization, hybridization and washes after hybridization were done as described by Stuart et al. (26). 32P-radiolabeled probe was prepared. with Boehringer Mannheim’s Random Primed DNA Labeling Kit according to the manufacturer's specifications. An internal 916 bp HindIII/NcoI fragment from the neo gene was used as a template to make radiolabeled probe to assay the neo- constructs. A 1169 bp HindIII SV40 early region fragment was used as a template to make radiolabeled probe to assay SV40 early mRNA expression in SV40 infected cells. Primer Extension Assays, Primer extension assays were essentially performed as described by Eisenberg et al. (6). The primer, a 25 base oligonucleotide (5’ CGGACTGGCTTTCTACGTGTTCCGC3 ’; +72 to +47 bp downstream of the HindIII site at the 5’ end of the Neo gene), was synthesized using an .Applied Biosystems 3803 DNA synthesizer (Macromolecular Structure Facility, Dept. of Biochemistry, Michigan State University). 100 ng of primer were labeled at 37°C for 30 min. using 2 ul of [1-32P]~dATP (6000 Ci/mmol, 150 mCi/ml) and 10 units of T4 polynucleotide kinase. The primer was then purified by anion exchange chromatography. 79 The following buffers were used for primer extension assays. 10X hybridization buffer contains 1.5 M KCl, 100 mM Tris-HCl pH 8.3, 10 mM EDTA. Extension reaction mix contains 20 mM Tris-HCl pH 8.3, 10 mM MgC12, 6 mM DTT, 0.3 mM of each dNTP, 150 ug/ml actinomycin D and 10 units AMV reverse transcriptase (Life Sciences Inc.). The hybridization was performed as follows: 20 ug of total RNA was resuspended in 10 ul H20. 5 ul of hybridization solution (3.5 ul or approximately 10,000 cpm of labeled primer, and 1.5 ul 10X hybridization buffer) was added and the mixture was incubated at 65°C for 1.5 hr and then cooled to room temperature. 30 ul of an extension reaction mix were added to the 15 ul hybridization reaction and the mixture was incubated at 37°C for 60 min. DNase free RNase A (30 ug/ml) was added and incubation was continued for 15 min at 37°C. Nucleic acids were extracted once with phenol/chloroform and ethanol precipitated. The pellet was resuspended in 4 ul 0f sequencing gel running buffer, heated to 90°C for 3 min. and electrophoresed on a 9% polyacrylamide/8 M urea gel in 1X TBE. Gel Mebility Shift Assays. NIH 3T3/b cells were grown to confluence, serum starved, and serum stimulated or SV40 infected as described above. At various times after Imitogenic treatment, nuclear extracts were prepared from NIH 3T3/b cells using the procedure of Dignam et al. (5). Nuclear extracts were frozen at ~70°C until ready for use. 80 Protein concentrations were determined using the Bradford assay. Binding reactions were performed for 20 min at room temperature in binding buffer (20 mM Hepes, pH 7.9, 20% glycerol, 0.1 M KCl, 0.2 mM EDTA, 0.5mM DTT, 5 mM M9012). Reactions contained 5-20 ug of extract protein, 0.7 ng of probe endfilled with [a-32P1-dCTP, and. 5 ug of non- specific competitor (either poly dI/dC or poly dA/dT). At the end of the binding reaction, the products were analyzed immediately on 5% nondenaturing polyacrylamide gels run in 0.4 X TBE. 81 RESULTS Qell Lines end Transfectipns. We previously localized the region required for regulation of the human TK promoter in serum stimulated Rat3 cells to between ~135 bp and -67 bp relative to the transcriptional start site using hybrid genes containing TK promoter fragments linked to the bacterial neo gene. Promoter fragments containing as little as 135 bp upstream of the cap site showed normal regulation of the gene, but a 67 bp promoter fragment resulted in constitutively low levels of neo mRNA after serum stimulation (23). These results were subsequently (confirmed in stably transfected hamster cell lines by Kim (and Lee, who also showed that sequences from -135 bp to ‘—67 bp are sufficient to confer Gl-S phase regulation to a laeterologous promoter (14). To compare the cis-acting sequences required for promoter regulation in serum stimulated and SV40 infected cells, we measured the level of neo mRNA expression at Vlarious times after SV40 infection of the pools of stably 1:ransfected Rat3 cells described previously. No induction <>f the TK-neo hybrid genes was observed under these Conditions. When the blots were reprobed with an SV40 early region fragment, however, no SV40 transcripts were detected (data not shown). We therefore conclude that SV40 either does not infect or express its early genes efficiently in serum starved Rat3 cells. To continue this study, TK-neo hybrid genes were 82 transfected into cell lines that are efficiently infected by SV40. The cell lines used were simian CV1 cells, which are permissive for SV40 replication and two sublines of NIH-3T3 cells (NIH-3T3/a and NIH-3T3/b), which are non- permissive for SV40 replication. Three TK-neo constructs were studied: 1., 444-Neo, which contains a 444 bp promoter fragment that appears to be fully active and regulated, 2., 135-Neo, which contains the smallest TK promoter fragment (135 bp) that was regulated by serum in our previous study in Rat3 cells and 3., 67-Neo, which contains a 67 bp promoter fragment that was not inducible in our previous study (Fig.1) . These constructs were stably transfected into CV1 cells and both sublines of NIH-3T3 cells, and G418 resistant colonies were selected. In all cases, transfection of 67-Neo was slightly (2-5 fold) less efficient than the other two constructs. Pools of 10 to 50 independent colonies obtained from all three constructs were expanded for analysis. In the case of 67-Neo in CV1 cells, single colonies were also expanded into cell 1 ines . Induction of Human 1K Eromoter-neo Hybrig Genes ln Engine NIH-3T3 C elle. We first investigated pools of NIH-3T3/a cells containing the TK-neo hybrid gene constructs. The onset of DNA synthesis in these cells was between 12 and 14 hours after serum stimulation or SV40 infection, as determined by 3H-thymidine incorporation into DNA (Fig.2). Note that the increase in thymidine 83 Fig. 13 Structure of human TK promoter-neo hybrid gene constructs. The restriction enzyme sites used to construct the deletion mutants, and their positions relative to the transcriptional start site (+1) are shown. The human TK gene sequence from -135 bp to +30 bp is shown, and putative regulatory elements and transcriptional start sites are marked. Symbols: ****, GC-box; ====, inverted CCAAT boxes; ----, TATA box; -->, transcriptional start sites. There are also 8 sequences that have homology to the consensus sequence for large T antigen binding sites located between -135 bp and -48 bp. Additionally, there are 6 sequences that have homology to the GC-box consensus sequence located between -134 bp and -11 bp. There is also one sequence each which has homology to the consensus sequence for binding sites for AP2 (-70 bp to -63 bp) and AP3 (-30 bp to -22 bp). 84 An: A: hc m T p. w m h“ N! 85 Fig. 2: DNA synthesis in growth arrested NIH-3T3/a cells after serum stimulation or SV40 infection. At the indicated times after mitogenic treatment, cells were labeled for one hour with 3 uCi [3H]-thymidine. Counts per minute incorporated per cell were determined by trichloroacetic acid precipitation of a given number of cells and subsequent scintillation counting. (Am) Serum stimulated cells. (B.) SV40 infected cells. :ells the were ounts d b)’ 3r of Serum 86 7— QOL x nan/was Time (hours) 87 incorporation is not as drastic in SV40 infected as in serum stimulated NIH-3T3 cells, although the TK promoter is induced to the same extent (see below). Although the reason for this is unknown, it might be due to the fact that an MOI of 30 is not high enough to stimulate DNA synthesis efficiently in SV40 infected cells. An MOI of 500 was used in the original report, which showed that DNA synthesis was stimulated approximately 25 fold in SV40 infected 3T3 cells (8). Neo mRNA levels were analyzed at 0 (G0), 3, 6 (G1), 14 (Gl-S boundary), 18 (S), and 24 hours following SV40 infection or serum stimulation. The results of this experiment (Fig.3 A and B) show that the transfected genes are regulated similarly by both mitogenic agents, and thus that the TK promoter is activated upon SV40 infection. The overall level of expression of 135-Neo is lower than that of 444-Neo, but the magnitude of induction is the same: both are induced approximately 4-5 fold between GO and S (18 hrs) phase in serum stimulated and SV40 infected cells. A maximum induction of 6-9 fold is seen by 24 hours after mitogenic treatment. This result is consistent with the Gl-S phase specific induction of these promoters in serum stimulated Rat3 cells (23). In contrast, the pattern of expression of 67-Neo is quite different. 67-Neo is slightly induced between 0 (G0) and 3 hours (G1), and is then expressed throughout the time course at similar levels to 135-Neo when it is induced at 18 and 24 88 Fig. 3: Northern blot analysis of human TK promoter-neo gene constructs in pools of stably transfected NIH-3T3/a cells. Total RNA was extracted at the indicated times (hours) after serum stimulation or SV40 infection as described in Materials and Methods. (A.) RNA from serum stimulated cells probed with a neo gene fragment. (B.) RNA from SV40 infected cells probed with a neo gene fragment. The lower panels in (A.) and (B.) show longer exposures of the Northern blots directly above them. (C.) The blot shown in (B.) rehybridized with an SV40 early region probe. 89 figsr‘sts e rtmtfi I: . u . o m... 1...... v.8. . .1..» in.“ {I .1.. ti E r... fl? sen: . .. .. .{Esigsiwgwrf. ; . . . . “Men .9“ w ... - I mu9soumuo onmu9eo um u99oumu9coumu9eo 002 Km 002 .mnp 002.??? 002 .50 002 $9. 0029:? 01>m Eamwm 90 hours. No further induction of 67-Neo is observed as cells pass the 61 boundary and enter S phase. Since SV40 large T antigen is required for the induction of TX enzyme levels in infected cells (22), we examined the timing of T antigen mRNA expression relative to the activation of the TR jpromoter (Fig. 3C). The Northern blot shown in Fig. BB was reprobed with a SV40 early region fragment. T antigen mRNA is first detected at some point between 6 and 14 hours, and is expressed at maximal levels by 18 hours. The high levels of expression of 67-Neo during 61 phase was intriguing, and different from our previous results in Rat3 cells (23). To insure that this pattern of expression was not peculiar to one particular pool of cells, a second subline of NIH-3T3 cells, NIH 3T3/b, was obtained from another laboratory, stably transfected, and investigated. Experiments with SV40 infected NIH-3T3/b cells confirmed the previous results, and are shown in Fig. 4A. 444-Neo and 135-Neo are both induced between GO and S phase. 67-Neo is expressed at high levels throughout the time course, even in GO phase cells. By reprobing the blot shown in Fig. 4A with an SV40 early region probe, it was shown that T antigen mRNA can be detected by 12 hours and is present at high levels by 18 hours after infection (Fig. 48). Together, the experiments in NIH-3T3 cells indicate that TK promoter sequences downstream of -135 bp are 91 Fig. 4: Northern blot analysis of human TK promoter-neo gene constructs in pools of stably transfected NIH-3T3/b cells. Total RNA was extracted at the indicated times (hours) after SV40 infection as described in Materials and Methods. (A.) RNA from SV40 infected cells hybridized with a neo gene probe. The lower panel in (A.) shows a longer exposune of the NOrthern blot directly above it. (B.) The blot shown in (A.) rehybridized with an SV40 early region probe. 92 444-Neo 135-Neo 67-N90 oeeaxoeeasofiefig 3‘; 93 sufficient to convey Gl-S phase specific regulation to a linked neo gene in both serum stimulated and SV40 infected cells. Sequences between -444 bp and -l35 bp are required for maximal expression of the linked neo gene, but are not required for Gl-S phase specific regulation. An interesting outcome of these experiments is the fact that 67-Neo is fully active throughout 61 phase in NIH-3T3 cells. This suggests a model where a negative regulatory element(s) that represses expression of the neo gene during 61 is localized between 135 bp and 67 bp upstream of the mRNA cap site. Deletion of the putative negative regulatory element(s) would cause neo to be expressed at high levels in both GI and S phase NIH-3T3 cells. Ma in 5' Ends of Re omb' K-n o mRNAs fro NIH- 3T3 Cells. Because 67-Neo was regulated differently in NIH-3T3 cells than in Rat3 cells, and because it gave rise to slightly lower transfection efficiencies than 444-Neo or 135-Neo, we were concerned that it might be integrating next to cellular promoters and not initiating at the TR cap site. To address this issue, primer extension assays were performed with the same RNA preparations used for the Northern analysis in Fig. 3A. A 25 bp oligonucleotide that hybridizes within the neo gene was used as a primer. The extended. products are expected to be 101 and 95 nucleotides long if transcription initiates at both of the start sites determined by Kreidberg and Kelly (17). The results of this experiment are presented in Fig. 5. All 94 Fig. 5: Determination of the transcriptional start site in human TK promoter-neo gene constructs using a primer extension assay. Total RNA from stably transfected NIH- 3T3/a cells was extracted at the indicated times (hours) after serum stimulation and 20 ug of RNA was analyzed per extension reaction. HpaII digested pBR322 was labeled with [a-32P1-dCTP and used as size marker (lanes M, in base pairs). Additional bands appear in lanes M are due to slight degradation of the size marker. 95 135 67 444 9. o 0.. a up :. 11 91 2,6 1 ,. 11 . .1. .1 . . . 96 of the constructs (444-Neo, 135-Neo, and 67-Neo) are initiating at the expected TK transcriptional start sites at both 3 (Gl) and 18 (S) hours. In agreement with the Northern blot analysis shown in Fig. 3A, 444-Neo and 135-Neo are induced between 61 (3 hr) and S phase (18 hr). In contrast, 67-Neo is expressed at both time points at levels equivalent, to 'the expression of 135-Neo at 18 hours. Analysis of Protein ginding to TK Pmomgter Sequences. The data presented above indicate that TK promoter sequences between -135 bp and -67 bp are required for the Gl-S phase induction of the promoter in NIH-3T3 cells, and that additional sequences downstream of -67 bp are also involved. in. activity. Specifically, we jpropose ‘that a negative regulatory element that represses the promoter in GO and 61 resides between -135 bp and -67 bp. If this is true, the negative element might function by binding a protein factor (a repressor) during G0 and Gl, and such a factor might be released at the Gl-S boundary. To look for such an interaction, we have performed gel mobility shift assays with human TK promoter fragments and nuclear extracts from serum starved, serum stimulated and SV40 infected NIH-3T3 cells. Non-transfected NIH-3T3/b cells were serum starved and stimulated as described in Materials and Methods. Control experiments using 3H-thymidine incorporation indicated that these cells were in S phase by 12 hours after serum 97 stimulation (data not shown). Nuclear extracts were prepared at 0 (GO), 3 (G1) and 13 (8) hours after serum stimulation, and used in binding reactions with either -135/+30 or -135/-67 human TK promoter probes (Fig. 6A). No difference in the banding pattern was detected at the different time points with either probe. The gel shift pattern was also compared in serum stimulated and SV40 infected cells. No difference was detected using either the -135/-67 probe (Fig. 6B) or the -135/+30 probe (data not shown). Thus, the changes in TK promoter activity during serum stimulation and SV40 infection of NIH-3T3 cells are not accompanied by detectable changes in the pattern of protein binding under these conditions. Induction of the Human TK Eromoter-neo Hybrid Genes in Simian CV1 Qells. The expression and regulation of 444-Neo, 137-Neo and 67-Neo were also examined in stably transfected simian CV1 cells. We previously demonstrated that S phase begins between 8 and 10 hours in serum stimulated CV1 cells and between 16 and 18 hours in SV40 infected CV1 cells (25). The time points chosen for analysis of the TK—neo hybrid genes were therefore 0, 2, 12, 18, and 24 hours following SV40 infection or serum stimulation. In serum stimulated cells (Fig. 7A), the results are very similar to those previously seen in Rat3 cells (23). Both 444-Neo and 135-Neo are induced approximately 4-fold between GO and S phase. The timing of induction (between 2 and 12 hours) is similar to that of 98 Fig. 6: Electrophoretic mobility shift analysis using nuclear extracts prepared from NIH-3T3/b cells at various times after ndtogenic treatment. (A.) Nuclear extracts were prepared at 0, 3, and 13 hours after serum stimulation as described in Materials and Methods . A human TK probe spanning from —135 bp to +30 bp (left panel) or from -135 bp to -67 bp (right panel) was incubated with either 5 or 20 ug of nuclear extract from each time point. The lanes marked P show the free DNA probes. (B.) Nuclear extracts were prepared at 0, 3, and 13 hours after serum stimulation (left panel) and at 0, 8, and 24 hours after SV40 infection (right panel). A human TK probe spanning from —135 bp to -67 bp was incubated with either 5 or 20 ug of nuclear extract from each time point. The lane marked P shows the free DNA probe. 99 9a a o a. n o 8 >m :33 m m m m A a. .eft V35 fro: 50.x on? 100 Fig. 7: Northern blot analysis of human TK promoter-neo gene constructs in pools and single colonies of stably transfected CV1 cells. Total RNA was extracted at the indicated times (hours) after serum stimulation (left panel) or SV40 infection (right panel). (A.) RNA from pools of serum stimulated cells hybridized with a neo gene probe. (B.) RNA from pools of SV40 infected cells hybridized with a neo gene probe. (C.) The blot shown in (B.) rehybridized with an SV40 early region probe. (D.) Two cell lines derived from independent single colonies stably transfected with 67-Neo were infected with SV40. A Northern blot of RNA from the single colony cell lines was hybridized with a neo gene probe. 101 u muzoumuzonmuzo oozfio Cozumnp oozé: oe>m amazon 002 .ho muzon 02.3.. . Sam—mm t mac 2-..: < 102 the endogenous TK gene (25) , although it is of somewhat lower magnitude. 67-Neo is expressed at low levels throughout the time course, and fails to be induced as cells enter S phase. In SV40 infected CV1 cells all three constructs, including 67-Neo, are induced between Gl and S phase (Fig. 7B). As expected, this induction follows the expression of T antigen mRNA, which appears between 12 and 18 hours Fig. 7C). The induction of 67-Neo between 18 and 24 hours was surprising, and was confirmed using two cell lines derived from independent single colonies (Fig. 7D). In both single colonies, 67-Neo is regulated and clearly induced between Gl and S phase. The fact that 67-Neo is induced by SV40 and not by serum suggests that SV40 induces and/or activates different transcription factors than serum stimulation, and that these factors are capable of interacting with TK promoter sequences downstream of -67 bp. 103 DISCUSSION Murine NIH-3T3 and simian CV1 cells containing stably integrated human TK promoter-neo hybrid genes were used to identify and compare cis-acting sequences involved in regulation of the human TK promoter in serum stimulated and SV40 infected cells. A comparison of the activity and regulation of promoter-deletion mutants in a number of cell lines indicates that the TR promoter contains both positive and negative regulatory elements, some of which function in a cell line and/or mitogen specific manner (see Table 1). Sequences upstream of -135 bp are required for maximal activity in some cell lines in both transient (1) and stable (this report) transfection assays, but are not required for Gl/S phase regulation. Promoter fragments conta ining 1 3 5 bp upstream of the transcriptional start site are fully regulated by both serum and SV40 in all cell lines tested. The pattern of expression from a 67 bp promoter fragment is complex, being high throughout Gl and S phase in NIH-3T3 cells, constitutively low in serum stimulated Rat3 and CV1 cells, and induced during 8 phase in SV40 infected CV1 cells. These results confirm our earlier findings that sequences between -135 bp and -67 bp are required for regulation of the TK promoter in serum stimulated cells, and further suggest a negative regulatory model that is outlined below. In addition, they provide the first conclusive evidence for transcriptional regulation of the TR promoter 104 .cuouuma OHHHoQO Guano mIHw m CH ooHMHsmou no: can sowmmmumxo Ho mHo>mH anmm 9 QCOU HOZU .cuovuna OHuHoonm manna mIHw s :H omumHsoou no: can conmoumxm Mo mHo>mH soao .cuouumo oHHHoQO manna mIHo m CH ooosocH conmoumxm n .Ammv muHsmmu omanHnsm aHms0H>mHmm H A m on oz on omzlho H H H H 92 H ODZImmH ow I H H H H 602 DH 2 «we o¢>m asumm ov>m asumm o¢>m Eauom mBnImHz sauna .mmcHH HHmo ompomumcmuu >Hnmum ucmummuwo :H mocmm Gwyn»: omZIumuoaoum &9 Gwen: Ho conmmuon mcu so cOHumHsBHum oHcomouHfi mo vommum "H «Home 105 during SV40 infection, and indicate that the virus can mediate an effect via different cis-acting sequences than serum stimulation. The activation of the TK promoter by SV40 reported here differs from previous work in which we were unable to demonstrate transcriptional induction of the endogenous TK gene in SV40 infected CV1 cells using nuclear run-on assays (25). We believe that the earlier negative result was due to the fact that the experiments were performed using TK cRNA transcribed from the T3 promoter in the vector pT7/T3 18. We have since determined that transcripts synthesized from this promoter cross-hybridize to CV1 cell 28S rRNA due to sequences between the T3 promoter and the polylinker cloning site (S. Conrad, unpublished results). The true TK signal in our experiments was therefore obscured by cross hybridization to rRNA» ‘When these experiments were repeated, using transcripts synthesized from the Sp6 promoter, a 3-4 fold induction of the TR gene was observed (C. Stewart, Ph.D. dissertation). This result agrees well with the data presented in the current report. A surprising result of the present studies is that 67-Neo is expressed at induced levels (i.e. levels at least as high as fully induced 135-Neo) throughout GI and S phase in serum stimulated and SV40 infected NIH-3T3 cells, although the identical construct is expressed at constitutively low levels after serum stimulation of CV1 106 and Rat3 (23) cells. In order to explain this difference, we propose that the region between -135 bp and -67 bp contains both positive and negative regulatory elements, and that additional positive element(s) are localized downstream of -67 bp. We postulate that transcription is repressed during 61 due to a negative element between -135 bp and -67 bp, and that this repression is relieved as cells enter 8 phase. Deletion of this putative negative element leads to high levels of expression in NIH-3T3 cells, but not in CV1 or Rat3 cells. The difference in the pattern of expression of 67-Neo in the three cell lines could be explained if NIH-3T3 cells contain high levels of a factor that binds to a positive element downstream of -67 bp, while Rat3 and CV1 cells do not. High levels of expression in Rat3 and CV1 cells would therefore be dependent upon additional positive elements between -135 bp and -67 bp. An alternative explanation for the pattern of expression of 67-Neo in NIH-3T3 cells would be that this truncated promoter is very weak, and that by requiring expression of the neo gene we selected a subset of transfectants where integration occurred next to strong constitutive enhancers. We do not believe that this is true for several reasons. First, if survival in. G418 required integration next to an enhancer, we would expect to see a larger effect than the modest 2-5 fold decrease in transfection efficiency observed with 67-Neo in NIH-3T3 cells. Second, transfection of CV1 cells and Rat3 cells 107 with 67-Neo occurs at lower efficiency relative to 444-Neo than does transfection of NIH-3T3 cells. This argues that the selection is even stronger in CV1 and Rat3 cells. In spite of this selection, 67-Neo is expressed at low levels in serum stimulated CV1 and Rat3 cells. Thus, the pattern of expression of 67-Neo in NIH-3T3 cells is unlikely to be an artifact generated by the selection method used. The model for negative regulation presented above suggests that a trans-acting repressor is bound to TX promoter sequences between -135 bp and -67 bp during 60 and 61. One mechanism by which such repression could be relieved would be by dissociation of the factor as the cells enter S phase. Our gel shift experiments do not support such a model, since we detect no changes in the banding pattern with Go, 61 and S phase extracts from NIH-3T3 cells. These results differ from one previous report ( 16) , where the gel shift pattern with the TR promoter changed between GO and Gl/S phase with extracts from BALB/c 3T3 cells. The reasons for this discrepancy are unknown, but our results are in agreement with a recent report that the pattern of protein binding to the TR promoter is unchanged during the Gl-S transition in hamster lung fibroblasts (14). Several models could explain a change in promoter activity in the apparent absence of changes in protein binding. First, the protein responsible for Gl-S phase regulation of the gene might not form a stable complex with the promoter under the 108 conditions of the gel shift assay. Second, the change in activity might be due to modification, rather than removal, of a protein, and such modification(s) might not be detectable by this (or any other DNA binding) assay. In this regard, it is interesting to note that the pattern of binding of Hela cell derived serum response factor to the c-fos promoter is the same in extracts from continuously growing cells, serum starved. cells, and. cells shortly after serum stimulation (28). A second intriguing result of these experiments is the difference in regulation of 67-Neo in serum stimulated and SV40 infected CV1 cells, which indicates that the virus is capable of activating 'the jpromoter ‘via sequences downstream of -67 bp, while serum stimulation is not. One possible mechanism for this would be by direct binding of T antigen to the promoter, but this seems unlikely for several reasons. First, we have previously noted an 8 hour lag between the appearance of large T and the induction of the endogenous TK gene in infected CV1 cells (26), and a similar lag is seen in the case of the TK-neo hybrid genes. Second, gel shift assays detected no unique bands in SV40 infected NIH-3T3 cells, suggesting that large T is not binding directly to the promoter (Fig. 68). Finally, using purified large T antigen, we have been unable to detect any binding by either gel shift or immune precipitation assays (data not shown). We therefore propose that viral infection induces and/or activates 109 transcription factors in CV1 cells that are not induced by serum stimulation, and that these factors are capable of interacting with promoter sequences downstream of -67 bp. Thus the study of mitogenic stimulation by SV40 may reveal novel pathways that regulate progression from 61 to S phase. The precise sequences involved in positive and negative regulation of the TR promoter remain unknown. It is attractive to propose that sequences located at -65 bp, -56 bp and -17 bp, which have homology to an Spl binding site, are involved in activation of the TR promoter in SV40 infected CV1 cells, since it has been shown that SV40 is capable of inducing both Spl mRNA and protein (24) and phosphorylation of Spl (10). It is also possible that this element is responsible for the activity of 67-Neo in NIH-3T3 cells, if Spl is expressed at constitutively high levels in these cells. A final resolution of these issues awaits additional biochemical and mutational analyses to provide a detailed understanding of the functional importance of cis-acting DNA sequences and trans-acting factors involved in TX promoter regulation. 110 ACKNOWLEDGEMENTS The authors would like to thank Drs. J. Dodgson, R. Patterson and R. Schwartz for helpful comments on the manuscript. We would also like to thank Dr. M. Luskey for a generous gift of purified large T antigen, Dr. L. Tack for providing a detailed gel mobility shift protocol and M. Carozza for SV40 virus preparations. This work was supported by NIH grant CA37144. H.H.R. was supported by a DuVall and a College of Natural Science Fellowship. 111 LITERATURE CITED Arcot, S. S., E. K. Flemington and P. L. Deininger. 1989. The human thymidine kinase gene promoter. J. Biol. Chem. 264: 2343-2349. Ausubel, F. M., R. Brent, R. E. Kingston, D. D. Moore, J. A. Smith, J. G. Seidman, and K. Struhl. 1987. Current protocols in molecular biology. Greene Publishing Associates and Wiley-Interscience. Brent, T. P., J. A. V. Butler, and A. R. Crathorn. 1965. Variations in phosphokinase activities during the cell cycle in synchronous populations of HeLa cells. Nature 207: 176-177. Coppock, D. L., and A. B. Pardee. 1987. Control of thymidine kinase mRNA during the cell cycle. Mol. Cell. Biol. 7: 2925-2932. Dignam, J. D., R. M. Lebovitz, and R. G. Roeder. 1983. Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nucleic Acids Res. 11: 1475-1489. Eisenberg, S. P., D. M. Coen, and S. L. McKnight. 1985. Promoter domains required for expression of plasmid-borne copies of the herpes simplex virus thymidine kinase gene in virus-infected mouse fibroblasts and microinjected frog oocytes. Mol. Cell. Biol. 5: 1940-1947. 10. 11. 12. 112 Favaloro, J., R. Treisman, and R. Kamen. 1980. Transcription maps of polyoma virus-specific RNA: Analysis by two-dimensional nuclease 81 gel mapping. Methods Enzymol. 65: 718-749. Gershon, D., L. Sachs, and E. Winocour. 1966. The induction of cellular DNA synthesis by simian virus 40 in contact-inhibited and x-irradiated cells. Proc. Natl. Acad. Sci. USA 56: 918-925. Ito, M., and S. E. Conrad. 1990. Independent regulation of thymidine kinase mRNA and enzyme levels in serum-stimulated cells. J. Biol. Chem. 265: 6954- 6960. Jackson, S. P., J. J. MacDonald, S. Lees-Miller, and R. Tjian. 1990. GC box binding induces phosphorylation of Spl by a DNA-dependent protein kinase. Cell 63: 155-165. Johnson, L. F., C. L. Fuhrman, and L. M. Weidemann. 1978. Regulation of dihydrofolate reductase gene expression in mouse fibroblasts during the transition from resting to growing state. J. Cell. Physiol. 97: 397-406. Johnson, L. F., L. G. Rao, and A. J. Muench. 1982. Regulation of thymidine kinase enzyme level in serum- stimulated mouse 3T6 fibroblasts. Exp. Cell. Research 138: 79-85. 13. 14. 15. 16. 17. 18. 113 Kim, Y. K., S. Wells, Y. F. C. Lau, and A. S. Lee. 1988. Sequences contained within the promoter of the human thymidine kinase gene can direct cell-cycle regulation of heterologous fusion genes. PNAS 85: 5894-5898. Kim, Y. K., and A. S. Lee. 1991. Identification of a 70-base pair cell cycle regulatory unit within the promoter of the human thymidine kinase gene and its interaction with cellular factors. Mol. Cell. Biol. 11: 2296-2302. Kit, 8., D. R. Dubbs, P. M. Frearson, and J. Melnik. 1966. Enzyme induction in SV40-infected green monkey kidney cultures. Virology 29: 69-83. Knight, G. B., J. M. Gudas, and A. B. Pardee. 1987. Cell-cycle specific interaction of nuclear DNA-binding proteins with a CCAAT sequence from the human thymidine kinase gene. Proc. Natl. Acad. Sci. USA 84: 8350-8354. Kreidberg, J. A., and T. J. Kelly. 1986. Genetic analysis of the human thymidine kinase gene promoter. M01. Cell. Biol. 6: 2903-2909. Lieberman, H. B., P. F. Lin, D. B. Yeh, and F. H. Ruddle. 1988. Transcriptional and posttranscriptional mechanisms regulate murine thymidine kinase gene expression in serum-stimulated cells. Mol. Cell. Biol. 8: 5280-5291. 19. 20. 21. 22. 23. 24. 25. 114 Lipson, K. E., S. T. Chen, J. Koniecki, D. H. Ku, and R. Baserga. 1989. S-phase-specific regulation by deletion mutants of the human thymidine kinase promoter. Proc. Natl. Acad. Sci. USA 86: 6848-6852. Navalgund, L. G., C. Rossana, A. J. Muench, and L. F. Johnson. 1980. Cell cycle regulation of thymidylate synthase gene expression in cultured mouse fibroblasts. J. Biol. Chem. 255: 7386-7390. Pardee, A. B., 1974. A restriction point for control of normal animal proliferation. Proc. Natl. Acad. Sci. 1 USA 71: 1286-1290. Postel, E. H., and A. J. Levine. 1976. The requirement of simian virus 40 gene A product for the stimulation of cellular tymidine kinase activity after viral infection. Virology 73: 206-215. Roehl, H. H., and S. E. Conrad. 1990. Identification of a Gl-S-phase-regulated region in the human tymidine kinase gene promoter. Mol. Cell. Biol. 10: 3834-3837. Saffer, J. D., S. P. Jackson, and S. J. Thurston. 1990. SV40 stimulates expression of the trans-acting factor Spl at the mRNA level. Genes & Development 4: 659-666. Stewart, C. J., M. Ito, and S. E. Conrad. 1987. Evidence for transcriptional and post-transcriptional control of the cellular thymidine kinase gene. Mol. Cell. Biol. 7: 1156-1163. 26. 27. 28. 29. 115 Stuart, P., M. Ito, C. Stewart, and S. E. Conrad. 1985. Induction of cellular thymidine kinase occurs at the mRNA level. Mol. Cell. Biol. 5: 1490-1497. Travali, S., K. E. Lipson, D. Jaskulski, E. Lauret, and R. Baserga. 1988. Role of the promoter in the regulation of the thymidine kinase gene. Mol. Cell. Biol. 8: 1551-1557. Treisman, R., 1986. Identification of a protein- binding site that mediates transcriptional response of the c-fos gene to serum factors. Cell 46: 567-574. Wigler, M., A. Pellicer, S. Silverstein, and R. Axel. 1978. Biochemical transfer of single copy eucaryotic genes using total cellular DNA as donor. Cell 14: 725- 731. SUMMARY AND CONCLUSIONS Rat3, murine NIH-3T3, and simian CV1 cells containing stably integrated human TK-neo hybrid genes were used to identify cis-acting sequences involved in regulation of the human TK promoter in serum stimulated and SV40 infected cells. A comparison of the regulation of promoter-deletion mutants indicates that the TK promoter contains both positive and negative regulatory elements, some of which function in a cell line and/or mitogen specific manner. Regions within the human TK promoter conveying G1/S phase specific regulation upon the promoter were identified, and evidence was presented suggesting that serum stimulation and SV40 infection may induce the TK promoter using partially overlapping or different pathways. It was shown that TK promoter elements between -444 bp and -l35 bp upstream of the transcriptional start site are required for maximal expression of the linked neo gene, but are not required for Gl-S phase specific regulation. Promoter fragments containing 135 bp upstream of the transcriptional start site are fully regulated by serum stimulation and SV40 infection in all cell lines tested. The pattern of expression from a 67 bp promoter fragment is complex. It is constitutively low in serum stimulated Rat3 and CV1 cells, induced during S phase in SV40 infected CV1 cells, and high throughout G1 and S phase in both serum stimulated and SV40 infected NIH-3T3 cells. 116 118 fully induced 135-Neo) in G1 phase in NIH-3T3 cells. 5. There is a positive element(s) downstream of -67 bp. This element is sufficient to convey high levels of expression to 67-Neo in NIH-3T3 cells, when the negative element described above is deleted. It is hypothesized that this element is inactive in serum stimulated CV1 and Rat3 cells, where 67-Neo is expressed at constitutive low levels. High levels of expression in serum stimulated CV1 and Rat3 cells would therefore be dependent upon additional positive elements between -135 bp and -67 bp. It is acknowledged that the element described under 3. and 5. might be overlapping or indeed be the same element. A challenge for the future will be to exactly define a gig-acting, S phase specific motif(s) in the TK promoter by investigating the regulation of a series of point mutants. Protein factors which interact with this region(s) in a Gl/S phase specific manner have to be identified next. Finally, the genes encoding these factors need to be isolated to allow a detailed characterization of the factors. This will allow one to determine whether serum stimulation and SV40 infection activate TK transcription by the same or by different pathways. Finally, it will be very interesting to compare the identified gig-acting motif(s) and the factor(s) interacting with them to other 8 phase regulated genes to search for an underlying mechanism that might regulate these genes in concert. 117 These results suggest a regulatory model outlined below. 1. There are TK promoter elements located between -444 bp and -135 bp upstream of the transcriptional start site that are required for' maximal expression of the linked neo gene. They are not required for Gl-S phase specific regulation when GO phase cells are induced to re- enter the cell cycle by either serum stimulation or SV40 infection. 2. There are regulatory elements located downstream of 135 bp that are sufficient for conveying Gl-S phase regulation to the linked neo gene when G0 phase cells are either SV40 infected or serum stimulated to re-enter the cell cycle. 3. There is a regulatory element(s) located downstream of -67 bp. This element is sufficient to activate the TK promoter in SV40 infected CV1 cells. In contrast, in serum stimulated CV1 cells, 67-Neo is expressed at constitutive low levels. It is hypothesized that SV40 infection induces and/or activates transcription factors that are not induced by serum stimulation and that these factors are capable of interacting with promoter sequences downstream of -67 bp. 4. There is a negative regulatory element(s) located between -135 bp and -67 bp that represses expression of the neo gene during G1 phase after serum stimulation or SV40 infection. This is shown by the fact that 67-Neo is expressed at high levels (i.e. levels at least as high as HICHIGRN STATE UNIV. LIBRARIES 1|”lWWIWWHlWlllHWWWI111”"?le 31293009011523