Baixe modelo matemático de processos celulares e outras Notas de estudo em PDF para Engenharia de Produção, somente na Docsity! REPORT Global effects of DNA replication and DNA replication origin activity on eukaryotic gene expression Larsson Omberg1,5, Joel R Meyerson1,6, Kayta Kobayashi2,7, Lucy S Drury3, John FX Diffley3,* and Orly Alter1,4,* 1 Department of Biomedical Engineering, University of Texas, Austin, TX, USA, 2 College of Pharmacy, University of Texas, Austin, TX, USA, 3 Cancer Research UK London Research Institute, Clare Hall Laboratories, South Mimms, Hertfordshire, UK and 4 Institutes for Cellular and Molecular Biology, and Computational Engineering and Sciences, University of Texas, Austin, TX, USA 5 Present address: Department of Biological Statistics and Computational Biology, Cornell University, Ithaca, NY 14853, USA 6 Present address: Center for Cancer Research, National Cancer Institute, Bethesda, MD 20892, USA 7 Present address: Pharmacy Department, Intermountain Medical Center, Murray, UT 84157, USA * Corresponding authors. O Alter, Department of Biomedical Engineering and Institutes for Cellular and Molecular Biology and Computational Engineering and Sciences, University of Texas, Austin, TX 78712, USA. Tel.: þ 1 512 471 7939; Fax: þ 1 512 471 2149; E-mail: orlyal@mail.utexas.edu and JFX Diffley, Cancer Research UK London Research Institute, Clare Hall Laboratories, South Mimms, Hertfordshire EN6 3LD, UK. Tel.: þ 44 1707 625 869; Fax: þ 44 1707 625 801; E-mail: john.diffley@cancer.org.uk Received 4.3.09; accepted 19.8.09 This report provides a global view of how gene expression is affected by DNA replication. We analyzed synchronized cultures of Saccharomyces cerevisiae under conditions that prevent DNA replication initiation without delaying cell cycle progression. We use a higher-order singular value decomposition to integrate the global mRNA expression measured in the multiple time courses, detect and remove experimental artifacts and identify significant combinations of patterns of expression variation across the genes, time points and conditions. We find that, first, B88% of the global mRNA expression is independent of DNA replication. Second, the requirement of DNA replication for efficient histone gene expression is independent of conditions that elicit DNA damage checkpoint responses. Third, origin licensing decreases the expression of genes with origins near their 30 ends, revealing that downstream origins can regulate the expression of upstream genes. This confirms previous predictions from mathematical modeling of a global causal coordination between DNA replication origin activity and mRNA expression, and shows that mathematical modeling of DNA microarray data can be used to correctly predict previously unknown biological modes of regulation. Molecular Systems Biology 5: 312; published online 13 October 2009; doi:10.1038/msb.2009.70 Subject Categories: functional genomics; cell cycle Keywords: a higher-order singular value decomposition (HOSVD); cell cycle; DNA replication origin licensing and firing; DNA microarrays; mRNA expression This is an open-access article distributed under the terms of the Creative Commons Attribution Licence, which permits distribution and reproduction in any medium, provided the original author and source are credited. Creation of derivative works is permitted but the resulting work may be distributed only under the same or similar licence to this one. This licence does not permit commercial exploitation without specific permission. Introduction DNA replication and transcription occur on a common template, and there are many ways in which these activities may influence each other. First, the passage of DNA replication forks offers an opportunity to change gene expression patterns. The transcription of capsid genes generally occurs late in most viral infections and is often dependent upon prior replication of the viral genome (Rosenthal and Brown, 1977; Thomas and Mathews, 1980; Toth et al, 1992). In bacterio- phage T4, there is a direct coupling of replication and transcription because the sliding clamp processivity factor gp45 acts as a mobile transcriptional enhancer (Herendeen et al, 1989). Such coupling may serve a regulatory role: DNA replication differentially affects the transcription of the embryonic and somatic 5S rRNA genes in the frog Xenopus laevis (Wolffe and Brown, 1986). Second, juxtaposed genes and replication origins can influence each other’s activity. The origin of the virus SV40 replication overlaps promoter and enhancer elements for both early and late gene expres- sion (Cowan et al, 1973; Alwine et al, 1977), and there are many examples of transcription factors influencing replica- tion origin function (DePamphilis, 1988). Moreover, induced transcription into a yeast Saccharomyces cerevisiae replica- tion origin can inactivate it (Snyder et al, 1988). Finally, clashes between replication and transcription machinery are & 2009 EMBO and Macmillan Publishers Limited Molecular Systems Biology 2009 1 Molecular Systems Biology 5; Article number 312; doi:10.1038/msb.2009.70 Citation: Molecular Systems Biology 5:312 & 2009 EMBO and Macmillan Publishers Limited All rights reserved 1744-4292/09 www.molecularsystemsbiology.com B o x 1 T he st ru ct ur e of th e da ta in th is st ud y is of an or de rh ig he rt ha n th at of a m at rix .G en es ,t im e po in ts an d co nd iti on s of M cm 2– 7 or ig in bi nd in g, ea ch re pr es en ta de gr ee of fr ee do m in a cu bo id ,t ha ti s, a th ird -o rd er te ns or . U nf ol de d in to a m at rix , th es e de gr ee s of fr ee do m ar e lo st an d m uc h of th e in fo rm at io n in th e da ta te ns or m ig ht al so be lo st . W e in te gr at e th es e da ta by us in g a te ns or H O S V D (S up pl em en ta ry in fo rm at io n S ec tio ns 2 an d 3, an d M at he m at ic a N ot eb oo k) .T hi s H O S V D un co ve rs in th e da ta te ns or (S up pl em en ta ry in fo rm at io n D at as et 3) pa tte rn s of m R N A ex pr es si on va ria tio n ac ro ss th e ge ne s, tim e po in ts an d co nd iti on s, as de pi ct ed in th e ra st er di sp la y of S up pl em en ta ry in fo rm at io n E qu at io n (1 )( le ft) w ith ov er ex pr es si on (r ed ), no ch an ge in ex pr es si on (b la ck )a nd un de re xp re ss io n (g re en ). T hi s H O S V D w as re ce nt ly re fo rm ul at ed in an al og y w ith th e m at rix si ng ul ar va lu e de co m po si tio n (S V D )( G ol ub an d V an Lo an ,1 99 6; A lte re ta l, 20 00 ;N ie ls en et al ,2 00 2; A lte ra nd G ol ub ,2 00 4; A lte r, 20 06 ;L ia nd K le ve cz ,2 00 6) su ch th at it se pa ra te s th e da ta te ns or in to a w ei gh te d su m of co m bi na tio ns of th re e pa tte rn s ea ch ,t ha ti s, ‘s ub te ns or s’ ,a s de pi ct ed in th e ra st er di sp la y of S up pl em en ta ry in fo rm at io n E qu at io n (7 )( rig ht ), w he re th e se co nd an d th ird H O S V D co m bi na tio ns an d th ei r co rr es po nd in g w ei gh ts ar e sh ow n ex pl ic itl y. In th es e ra st er di sp la ys ,e ac h ex pr es si on pa tte rn ac ro ss th e tim e po in ts is ce nt er ed at its tim e- in va ria nt le ve l. T he ge ne s ar e so rt ed by th ei r‘ an gu la r di st an ce s’ be tw ee n th e se co nd an d th ird H O S V D co m bi na tio ns (S up pl em en ta ry in fo rm at io n S ec tio n 2. 6) ,w hi ch re pr es en tt he un pe rt ur be d ce ll cy cl e ex pr es si on os ci lla tio ns (F ig ur e 2 an d S up pl em en ta ry F ig ur e 12 ). T he w hi te lin es se pa ra te th e ev en an d od d hy br id iz at io n ba tc he s, in di ca te d by bl ac k ar ro w s. T hi s re fo rm ul at io n of th e H O S V D w as sh ow n to en ab le its in te rp re ta tio n in te rm s of th e ce llu la rs ta te s, bi ol og ic al pr oc es se s an d ex pe rim en ta la rt ifa ct s th at co m po se th e da ta te ns or by de fin in g th e si gn ifi ca nc e of ea ch co m bi na tio n of pa tte rn s, an d th e op er at io n of ro ta tio n in a su bs pa ce of th es e co m bi na tio ns (O m be rg et al ,2 00 7) .I n th is st ud y, w e ex te nd th is an al og y to m at he m at ic al ly de fin e th e op er at io n of re co ns tr uc tio n in a su bs pa ce of co m bi na tio ns (S up pl em en ta ry in fo rm at io n S ec tio n 2. 4) , an d us e it to co m pu ta tio na lly re m ov e ex pe rim en ta l ar tif ac ts fr om th e gl ob al m R N A ex pr es si on m ea su re d in m ul tip le tim e co ur se s (S up pl em en ta ry in fo rm at io n S ec tio n 3. 3) . B o x 1 H ig h e r- o rd e r s in g u la r v a lu e d e c o m p o s it io n (H O S V D ) Global effects of replication on gene expression L Omberg et al 2 Molecular Systems Biology 2009 & 2009 EMBO and Macmillan Publishers Limited Figure 3 DNA replication-dependent and Mcm2–7 origin binding-dependent gene expression. Raster display, in which the expression of each gene is centered at its time-invariant level. (A) DNA replication is required for efficient histone gene expression. Raster display of histone gene expression shows that histone genes are overexpressed in the Cdc6þ /45þ control, relative to the Cdc6 condition, and to a lesser extent also relative to the Cdc45 condition, in a highly correlated manner. (B) Origin licensing decreases the expression of genes with origins near their 30 ends. Raster display of the expression of the 16 most significant genes in this class shows that these genes are overexpressed in the Cdc6 relative to the Cdc45 time courses, and to a lesser extent also relative to the control, in a manner less correlated than that of the histone genes. Global effects of replication on gene expression L Omberg et al & 2009 EMBO and Macmillan Publishers Limited Molecular Systems Biology 2009 5 largely independent of underlying cell cycle events, such as DNA replication (Simon et al, 2001). Second, we find that B3.5% of the overall expression of the 4270 genes depends on DNA replication but is independent of the method of origin inactivation. These replication-dependent perturbations in mRNA expression are represented by the fourth and sixth combinations, which correlate with over- expression in the averaged control and underexpression in both Cdc45 and Cdc6 time courses. The fourth combination also correlates with time-invariant underexpression and with expression variation across the genes that is enriched in underexpressed histone genes with the P-value o9.21013. The sixth combination correlates with overexpression of histone genes at the G1/S phase with the corresponding P-value o4.9104. Taken together, the time-averaged and G1/S expression of histone genes is reduced in both situations where DNA replication is prevented, indicating that DNA replication is required for efficient histone gene expression. To examine the joint effects of DNA replication and origin binding on global mRNA expression, we classified the 4270 genes into intersections of the fourth through seventh combinations of expression patterns (Supplementary informa- tion Dataset 6). Enrichment in overexpressed histone genes was also observed for the fifth combination with the P-value o1.5108. Among the 1294 genes that are underexpressed in the fourth and overexpressed in the fifth and sixth combinations, the four most significant genes, in terms of the fraction of mRNA expression that they capture in these combinations, are the histone genes HTA1, HTA2, HTB1 and HTB2. Six of the nine histone genes are among the ten most significant genes, an enrichment that corresponds to a P-value B2.11015. Overall, the histone genes are over- expressed in the control relative to the Cdc6 condition, in which the Mcm2–7 licensing of origins and subsequent DNA replication are prevented, and to a lesser extent also relative to the Cdc45 condition, in which DNA replication is prevented but only after the origins are licensed (Figure 3a). Previous work has shown that the coupling of histone mRNA levels to DNA replication is primarily due to transcrip- tional regulatory mechanisms (Lycan et al, 1987). Because in our study the Rad53 checkpoint kinase is not activated in either the Cdc6 or Cdc45 conditions as previously described (Piatti et al, 1995; Tercero et al, 2000), and because we did not observe any significant enrichment in DNA damage-induced genes (Jelinsky and Samson, 1999) in the fourth through seventh combinations of expression patterns, in which the corresponding P-values 49.4102, we suggest that these effects on histone gene expression are directly dependent on DNA replication status, independent of DNA damage check- point responses. Third, we find that B2.6% of the mRNA expression is affected by Mcm2–7 origin binding. The origin binding- dependent perturbations are represented by the fifth and seventh combinations, which correlate with overexpression in the Cdc6 and underexpression in the Cdc45 cultures. The fifth combination correlates with time-invariant underexpres- sion that is enriched in underexpressed genes with autono- mously replicating sequences (ARSs) adjacent to their 30 ends, defined as genes with at least one confirmed ARS at a distance of less than 100 nucleotides from their respective 30 ends (Cherry et al, 1997; Nieduszynski et al, 2007) with the P-value o1.9103. The seventh combination correlates with G2/M overexpression of genes with ARSs near their 30 ends, with the P-value o6.9104. Taken together, origin licensing de- creases time-averaged and G2/M expression of genes with origins near their 30 ends. We did not observe any significant enrichment in genes with ARSs near their 50 ends nor did we observe any significant enrichment in genes that overlap ARSs, where all the corresponding P-values were 41.2101. We suggest that origin licensing may affect the expression of adjacent genes by interfering with transcription elongation and/or pre-mRNA 30-end processing (Proudfoot, 2004; Gil- martin, 2005; Weiner, 2005; Rosonina et al, 2006), thus destabilizing the mRNA transcripts. The accumulation of mRNA transcripts of this class of genes throughout the Cdc6 relative to the Cdc45 time courses is consistent with the observed peak in their differential expression late in the time courses at the G2/M phase (Figure 3b). Of the 153 genes with ARSs near their 30 ends, 16 are among the 100 most significant of the 1412 genes that are over- expressed in the fourth and underexpressed in the fifth and seventh combinations, an enrichment that corresponds to a P-value o3.4107. No other significant enrichment was observed among these 100 genes, nor was an enrichment in gene ontology (GO) (Cherry et al, 1997) annotations observed among the 16 genes with ARSs near their 30 ends. Of these 153 and 16 genes, only 24 and 5, respectively, are cell cycle regulated. These 16 genes are overexpressed in the Cdc6 condition, in which the origins are unlicensed, relative to the Cdc45 condition, in which Mcm2–7 bind origins throughout the time course, and to a lesser extent also relative to the control, in which Mcm2–7 bind origins only during G1. The expression of these genes is not as highly correlated as that of the nine genes for histones, consistent with this class of genes being co-degraded but not necessarily co-transcribed. Previous studies have shown that transcription can interfere with the function of downstream origins (Snyder et al, 1988; Nieduszynski et al, 2005; Donato et al, 2006). Our results reveal that downstream origins can also interfere with the expression of upstream genes. This interference requires origin licensing but does not require origin firing. We suggest that cells may exploit this complex reciprocal arrangement in different contexts to regulate gene expression or origin activation. A global pattern of correlation between DNA binding of Mcm2–7 and reduced expression of adjacent genes, most of which are not cell cycle regulated, during the cell cycle phase G1 was discovered from previous mathematical modeling of DNA microarray data (Alter and Golub, 2004; Omberg et al, 2007), in which the mathematical variables and operations were shown to represent biological reality (Alter, 2006). The mathematical variables, patterns uncovered in the data, were shown to correlate with activities of cellular elements, such as regulators or transcription factors. The operations, such as classification, rotation or reconstruction in subspaces of these patterns, were shown to simulate experimental observation of the correlations and possibly even the causal coordination of these activities (Supplementary information Section 2). Of the 153 genes with ARSs near their 30 ends, the ARSs near 151 were identified in Mcm2–7 high-throughput binding assays, and for Global effects of replication on gene expression L Omberg et al 6 Molecular Systems Biology 2009 & 2009 EMBO and Macmillan Publishers Limited the ARSs near 139 of those, including all of the 16 significant genes, consensus sequence elements were identified (Wyrick et al, 2001; Xu et al, 2006). Our results, therefore, suggest that a causal relation underlies this correlation, that is, the binding of Mcm2–7 is responsible for diminished expression of adjacent genes, and show that mathematical modeling of DNA microarray data can be used to correctly predict previously unknown biological modes of regulation. Supplementary information Supplementary information is available at the Molecular Systems Biology website (www.nature.com/msb). Acknowledgements We thank BA Cohen and VR Iyer and his laboratory members, Z Hu, PJ Killion and S Shivaswamy, for technical assistance with DNA microarrays from the Washington University Microarray Core and at the University of Texas, respectively, and AW Murray for insightful comments. We also thank GH Golub for introducing us to matrix and tensor computations, and the American Institute of Mathematics in Palo Alto for hosting the 2004 Workshop on Tensor Decompositions where some of this study was carried out. This study was supported by Cancer Research UK (JFXD) and National Human Genome Research Institute R01 Grant HG004302 and National Science Foundation CAREER Award DMS-0847173 (OA). Conflict of interest The authors declare that they have no conflict of interest. References Alter O (2006) Discovery of principles of nature from mathematical modeling of DNA microarray data. Proc Natl Acad Sci USA 103: 16063–16064 Alter O, Brown PO, Botstein D (2000) Singular value decomposition for genome-wide expression data processing and modeling. Proc Natl Acad Sci USA 97: 10101–10106 Alter O, Golub GH (2004) Integrative analysis of genome-scale data by using pseudoinverse projection predicts novel correlation between DNA replication and RNA transcription. Proc Natl Acad Sci USA 103: 16577–16582 Alwine JC, Reed SI, Stark GR (1977) Characterization of the autoregulation of simian virus 40 gene A. J Virol 24: 22–27 Brewer BJ (1988) When polymerases collide: replication and the transcriptional organization of the E. coli chromosome. Cell 53: 679–686 Cherry JM, Ball C, Weng S, Juvik G, Schmidt R, Adler C, Dunn B, Dwight S, Riles L, Mortimer RK, Botstein D (1997) Genetic and physical maps of Saccharomyces cerevisiae. Nature 387: 67–73 Cowan K, Tegtmeyer P, Anthony DD (1973) Relationship of replication and transcription of Simian Virus 40 DNA. Proc Natl Acad Sci USA 70: 1927–1930 DePamphilis ML (1988) Transcriptional elements as components of eukaryotic origins of DNA replication. Cell 52: 635–638 Donato JJ, Chung SC, Tye BK (2006) Genome-wide hierarchy of replication origin usage in Saccharomyces cerevisiae. PLoS Genet 2: e141 Gerke JP, Chen CT, Cohen BA (2006) Natural isolates of Saccharomyces cerevisiae display complex genetic variation in sporulation efficiency. Genetics 174: 985–997 Gilmartin GM (2005) Eukaryotic mRNA 30 processing: a common means to different ends. Genes Dev 19: 2517–2521 Golub GH, Van Loan CF (1996) Matrix Computations, 3rd edn. Baltimore, Maryland, USA: Johns Hopkins University Press Herendeen DR, Kassavetis GA, Barry J, Alberts BM, Geiduschek EP (1989) Enhancement of bacteriophage T4 late transcription by components of the T4 DNA replication apparatus. Science 245: 952–958 Hu Z, Killion PJ, Iyer VR (2007) Genetic reconstruction of a functional transcriptional regulatory network. Nat Genet 39: 683–687 Jelinsky SA, Samson LD (1999) Global response of Saccharomyces cerevisiae to an alkylating agent. Proc Natl Acad Sci USA 96: 1486–1491 Klevecz RR, Li CM, Marcus I, Frankel PH (2008) Collective behavior in gene regulation: the cell is an oscillator, the cell cycle a developmental process. FEBS J 275: 2372–2384 Li CM, Klevecz RR (2006) A rapid genome-scale response of the transcriptional oscillator to perturbation reveals a period-doubling path to phenotypic change. Proc Natl Acad Sci USA 103: 16254–16259 Liu B, Alberts BM (1995) Head-on collision between a DNA replication apparatus and RNA polymerase transcription complex. Science 267: 1131–1137 Lycan DE, Osley MA, Hereford LM (1987) Role of transcriptional and posttranscriptional regulation in expression of histone genes in Saccharomyces cerevisiae. Mol Cell Biol 7: 614–621 Nieduszynski CA, Blow JJ, Donaldson AD (2005) The requirement of yeast replication origins for pre-replication complex proteins is modulated by transcription. Nucleic Acids Res 33: 2410–2420 Nieduszynski CA, Hiraga S, Ak P, Benham CJ, Donaldson AD (2007) OriDB: a DNA replication origin database. Nucleic Acids Res 35: D40–D46 Nielsen TO, West RB, Linn SC, Alter O, Knowling MA, O’Connell JX, Ferro M, Sherlock G, Pollack JR, Brown PO, Botstein D, van de Rijn M (2002) Molecular characterisation of soft tissue tumours: a gene expression study. Lancet 359: 1301–1307 Omberg L, Golub GH, Alter O (2007) Tensor higher-order singular value decomposition for integrative analysis of DNA microarray data from different studies. Proc Natl Acad Sci USA 104: 18371–18376 Orlando DA, Lin CY, Bernard A, Wang JY, Socolar JE, Iversen ES, Hartemink AJ, Haase SB (2008) Global control of cell-cycle transcription by coupled CDK and network oscillators. Nature 453: 944–947 Piatti S, Lengauer C, Nasmyth K (1995) Cdc6 is an unstable protein whose de novo synthesis in G1 is important for the onset of S phase and for preventing a ‘reductional’ anaphase in the budding yeast Saccharomyces cerevisiae. EMBO J 14: 3788–3799 Proudfoot N (2004) New perspectives on connecting messenger RNA 30 end formation to transcription. Curr Opin Cell Biol 16: 272–278 Rosenthal LJ, Brown M (1977) The control of SV40 transcription during a lytic infection: late RNA synthesis in the presence of inhibitors of DNA replication. Nucleic Acids Res 4: 551–565 Rosonina E, Kaneko S, Manley JL (2006) Terminating the transcript: breaking up is hard to do. Genes Dev 20: 1050–1056 Simon I, Barnett J, Hannett N, Harbison CT, Rinaldi NJ, Volkert TL, Wyrick JJ, Zeitlinger J, Gifford DK, Jaakkola TS, Young RA (2001) Serial regulation of transcriptional regulators in the yeast cell cycle. Cell 106: 697–708 Snyder M, Sapolsky RJ, Davis RW (1988) Transcription interferes with elements important for chromosome maintenance in Saccharomyces cerevisiae. Mol Cell Biol 8: 2184–2194 Spellman PT, Sherlock G, Zhang MQ, Iyer VR, Anders K, Eisen MB, Brown PO, Botstein D, Futcher B (1998) Comprehensive identification of cell cycle-regulated genes of the yeast Saccharomyces cerevisiae by microarray hybridization. Mol Biol Cell 9: 3273–3297 Tavazoie S, Hughes JD, Campbell MJ, Cho RJ, Church GM (1999) Systematic determination of genetic network architecture. Nat Genet 22: 281–285 Global effects of replication on gene expression L Omberg et al & 2009 EMBO and Macmillan Publishers Limited Molecular Systems Biology 2009 7