Proc. Natl. Acad. Sci. USA Vol. 91, pp. 4475-4479, May 1994 Biochemistry Delineation of a human histone H4 cell cycle element in vivo: The master switch for H4 gene transcription ANNA RAMSEY-EWING*, ANDRE J. VAN WUNEN, GARY S. STEIN, AND JANET L. STEIN Department of Cell Biology and Comprehensive Cancer Center, University of Massachusetts Medical Center, 55 Lake Avenue North, Worcester, MA 01655 Communicated by Sheldon Penman, November 3, 1993 ABSTRACT Histone gene expression is cell cycle regulated the H4 cell cycle regulatory element has not previously been at the transcriptional and the post-tnscriptional levels. Upon obtained. entry into S phase, histone gene trnscription is stimulated 2- In this study we demonstrate that an 11-bp proximal to 5-fold and peaks within 1-3 hr of the initiation of DNA promoter element mediates cell cycle-regulated transcription synthesis. We have delineated the proximal promoter element of the F0108 human H4 histone gene in vivo. The cell cycle responsible for cell cycle-dependent transcription of a human element (CCE) is located in the distal region of site II, one of histone H4 gene in vivo. Our results indicate that H4 cell two protein-DNA interaction domains identified by in vivo cycle-dependent transcriptional regulation is mediated by an genomic footprinting (17). The CCE functions as a master 11-base-pair element, the cell cycle element (5'-CTTTCG- switch for transcription of the H4 gene, as demonstrated by GTTTT-3'), that resides in the in vivo protein-DNA interaction transient expression assays and in vivo competition analysis. site, site H (nucleotides -64 to -24). The H4 cell cycle element We suggest that the H4 CCE functions via binding of its functions as a master switch for expression ofthe F0108 human cognate factor, HiNF-M, and that this DNA-protein inter- histone H4 gene in vivo; mutations within the H4 cell cycle action is a rate-limiting step in S-phase activation ofthis gene element drastically reduce the level of expression as well as in vivo. abrogate cell cycle-regulated transcription. Furthermore, these mutations result in a loss of binding in vitro of the cognate nuclear factor HiNF-M. In vivo competition analysis indicates MATERIALS AND METHODS that the cell cycle element mediates specific competition for a Cell Culture and Synchronization. HeLa S3 cells were DNA-binding factor, presumably HiNF-M, that is a rate- grown and maintained in suspension at 3-6 x 105 cells per ml limiting step in transcription of this H4 gene. in Joklik-modified minimum essential medium supplemented with 7% calf serum, 100 units of penicillin per ml, 100 pg of A predetermined sequence of requisite events ensures that a streptomycin per ml, and 2 mM glutamine. Exponentially cell divides only after it has completely replicated the ge- growing cells were synchronized by a double thymidine block nome, duplicated all subcellular organelles, and reached a procedure (18) and synchrony was monitored by [3H]thymi- critical mass. Consequently there is an ordered hierarchy of dine incorporation. events that results from differential gene expression in re- Transfection Experiments. Calcium phosphate-DNA co- sponse to extracellular and intracellular signals. Replication precipitation (19) was used to transfect HeLa S3 cells grow- of DNA occurs during the S phase of the cell cycle and ing on plastic culture dishes. To determine the amount of involves DNA synthesis and packaging into chromatin. This plasmid DNA taken up by transfected cells, episomal DNA process requires stringent coupling of histone gene expres- was isolated (20) from cells harvested 48 hr after transfection. sion to DNA synthesis and is regulated at the transcriptional For selection of stable transformants, monolayer cultures and the post-transcriptional levels (reviewed in refs. 1 and 2). were cotransfected with the DNA of interest and pSV2-neo As cells enter S phase, histone gene transcription is stimu- at 20:1 ratio, respectively, and refed 36-48 hr after glycerol lated 2- to 5-fold and peaks within 1-3 hr of the initiation of shock with medium containing G418 (Geneticin) at 500 pg of DNA synthesis (3-6). active antibiotic per ml of medium. Expression of fusion Studies ofthe promoter sequences involved in regulation of genes containing the bacterial chloramphenicol acetyltrans- histone gene transcription have implicated several discrete ferase (CAT) gene was monitored essentially as described by proximal promoter elements in cell cycle-regulated control Gorman et al. (19). Total protein was quantitated spectro- (7-10). It has been directly demonstrated that the distal H1 photometrically using Coomassie blue (Pierce or Bio-Rad). subtype-specific element, the AC-box (5'-AAACACA-3'), Site-Directed Mutagenesis. Mutation of the F0108 H4 site mediates cell cycle-dependent transcription of a chicken H1 II was performed using a modification of the polymerase gene in HeLa cells (7). The analogous element in a human H1 chain reaction (PCR) method of Higuchi et al. (21). Muta- gene has been implicated in preferential transcription in vitro genesis was carried out using the NEN Repliprime kit. After in S-phase extracts from HeLa cells (10). Growth-regulated restriction enzyme digestion, the DNA fragments of interest transcription of an H3 gene in a temperature-sensitive Chi- were isolated from polyacrylamide gels and subcloned into nese hamster fibroblast line requires a 32-nucleotide region pUC19. Mutations were confirmed by sequence analysis. located -150 bp upstream of the TATA box (8). The H2b Gene Expression Assays. Nuclear run-on transcription was subtype-specific consensus element, which contains the core performed essentially as described in Baumbach et al. (3) octanucleotide 5'-ATTTGCAT-3', has been implicated in using 1 x 107 nuclei per reaction. DNA slot blots on either cell cycle-dependent transcription of a human H2b gene (9). nitrocellulose or Zeta-probe nylon membranes contained Although much work has been done on the sequences in- denatured DNA the radiola- volved in transcriptional regulation of human H4 genes in fragments complementary to vitro (11-13) and in vivo (14-16), definitive identification of Abbreviations: CCE, cell cycle element; CAT, chloramphenicol acetyltransferase. The publication costs ofthis article were defrayed in part by page charge *Present address: Laboratory ofViral Diseases, National Institute of payment. This article must therefore be hereby marked "advertisement" Allergy and Infectious Diseases, National Institutes of Health, in accordance with 18 U.S.C. §1734 solely to indicate this fact. Bethesda, MD 20892. 4475 Downloaded by guest on September 24, 2021 4476 Biochemistry: Ramsey-Ewing et al. Proc. Natl. Acad Sci. USA 91 (1994) beled RNA transcripts of interest and various control DNA fusion gene and the endogenous H4 genes were elevated 2- to fragments. After hybridization and washing, the blots were 3-fold during early S phase (Table 1 and data not shown). air dried and exposed to preflashed XAR-5 film with a These results are consistent with previous studies showing Lightning Plus intensifying screen at -700C. Gel retardation that transcription of the endogenous H4 gene peaks during analysis was performed as described in van Wijnen et al. (22). early S phase, preceding the maximal rate of DNA synthesis (3). Thus, the H4 promoter confers cell cycle regulation on CAT gene transcription, independent of chromosomal posi- RESULTS tion. The H4 Histone Proximal Promoter Is Sufficient for Cell We then prepared a series of 5' promoter deletions of the Cyde-Regulated Transcription. To determine ifthe 5' flanking PH4CAT gene to determine the minimal promoter necessary region of the F0108 H4 histone gene supports cell cycle- for cell cycle-dependent transcription (Fig. 1A). Each dele- dependent transcription, we analyzed expression ofPH4CAT tion construct was assayed at various points during the cell (construct A3 in ref. 16), which contains =1 kb of sequences cycle in transient, synchronized HeLa cell cultures and in (nucleotides -1039 to -11 relative to the mRNA initiation site) synchronized stably transfected HeLa cell lines. CAT ex- fused to the bacterial CAT gene (Fig. 1A). Synthesis of pression was directly measured by run-on transcription in properly initiated CAT mRNA was confirmed by primer nuclei isolated from the synchronized cells. extension analysis of poly(A)+ RNA from two stably trans- Deletion of nucleotides -1039 to -216 does not alter cell fected HeLa cell lines (data not shown). Nuclear run-on cycle-dependent transcription from the FO108 promoter analysis revealed that the rates oftranscription ofthe PH4CAT (Fig. 1C, Table 1). Nevertheless, we observed a significant C H4 *up_ H2B-A - CAT #We __n4 120 O q X C 12. 02 80 I-, R _ U) C. 60 30, IC - - 1... A P114 SM 72Th c. <; 40 L .I I' 11 i C=AT OS ! pmM pmC 20 : IA at A' OCCOCOC(I 10G I11 [CMIC[IGGC AA [C1IT1GIATA CAC)CCUA. CCI BOY /M :iCTCCATAC1C11lCTAlAi(1A(C(1(AA 0.0 2.0 B CM Ac SV2CAT D H4 __ H2B-Ar CAT 120 qA PH4CAT __ a d~~~Reie. It -1 - 1 00 - nf 0 - 586CAT *e 0 co _ 60 -21SCAT U>2I0 00 CM Ac -21SCAT 0 ASICAT 0.0 2.0 -i c( 6.06) 1 0. (2.0 AC FIG. 1. Analysis of the level of expression and cell cycle transcription of F0108 histone H4CAT promoter deletion mutants. (A) Schematic diagram (not drawn to scale) showing deletions and point mutations of the PH4CAT parental construct. PH4CAT was made by fusing -1 kb of F0108 H4 5' flanking sequences to the bacterial CAT gene transcription unit derived from pSV2CAT (16). (B) Analysis of transient expression in HeLa cells. Each sample represents CAT activity in lysates from a pooled cell population from 10 independent transfections. Shown are autoradiographs of thin-layer chromatography plates on which [14C]chloramphenicol (CM) and its acetylated products (Ac) were resolved.