Central postgenomic. Transcription regulation: a genomic network. Transcriptome: the set of all mrnas expressed in a cell at a specific time.

Transcription

1 Transcription regulation: a genomic network Nicholas Luscombe Laboratory of Mark Gerstein Department of Molecular Biophysics and Biochemistry Yale University Understand Proteins, through analyzing populations Structures (Motions, Composition) Functions (Locations, Interactions) Evolution (Pseudogenes) Work on a diverse range of data Gerstein lab: Haiyuan Yu Snyder lab: Christine Horak Teichmann lab: Madan Babu Microarrays Sequences Structures Google Hits Pub- Med Hits 1st Pub- Med Hit Year Genome ~ Proteome ~63, Transcriptome Physiome Metabolome Phenome Morphome Interactome Glycome Secretome Ribonome Orfeome Regulome Cellome Operome Unknome Functome PubMed Hits Central postgenomic terms Transcriptome Proteome Transcriptome: the set of all mrnas expressed in a cell at a specific time Microarrays: -expression - ChIp-chip Transcriptome Initial Step: genome sequence, genes and genomic features XX XX XXX X Overview Overview 1. ChIp-chip experiments for 11 transcription factors -SBF & MBF 1. ChIp-chip experiments for 11 transcription factors -SBF & MBF 2. Computational studies of yeast regulatory network 2. Computational studies of yeast regulatory network 1

2 SBF and MBF are cell cycle transcription factors SBF = SCB Binding Factor Swi6p Swi4p SCB G1 cyclin genes SCB = Swi4p-Swi6p cell cycle box Regulator at the G1/S phase of cell cycle Only five known target genes MBF = MCB Binding Factor Mbp1 Swi6p SCB CGCGAAAA MCB ACGCGN SBF binds MCBs and MBF binds SCBs SBF MBF MCB S-phase cyclin genes DNA synthesis genes? MCB = Mlu1Cell Cycle Box SCB MCB ChIp-chip Chromatin Immunoprecipitation and DNA chip technology Crosslink Lyse Sonicate IP Reverse X-links Label Hybridize Yeast ChIp-chip Epitope-tagged Untagged 2

3 Examples of Swi4 Targets Swi4: 163 intergenic regions iydr309c Mbp1: 87 intergenic regions 5' GIC2 5' SUM1 43 were bound by both Swi4: 183 iylr341w genes YLR341w 5' MBP: 98 genes FKS1 5' iypl2 5 6 c [Horak, Luscombe et al (2002), Genes & Dev, BBP116: 3017 ] 5 ' Fraction of targets with consensus binding Sites 81% of SBF selected intergenic regions contain at least 1 binding site SCB Expression periodicity of SBF targets 52% G1/S 76% of MBF selected intergenic regions contain at least 1 binding site 14% G2/M 44% None MCB G1 S G2 M 6% 7% Functions of SBF and MBF Gene Targets 8% 11% 24% Cell Wall SBF Transcription 22% 22% DNA synthesis Proteolysis 4% 1% 2% 14% 24% Morphogenesis Other MBF 5% 9% Cell Cycle Unknown 41% M G1 S G2 WTM1 SWI4 SPT21 HCM1 YOX1 TOS4 PLM2 NDD1 YHP1 POG1 SBF and MBF target other transcription Factors TOS8 YAP5 MAL33 RGT1 PDR1 TYE7 GAT2 SOK2 3

4 Genomic distribution of TF binding sites Genomic distribution of TF binding sites multiple transcriptional regulators Hcm1YDL129w VCX1 Swi4 Tos8 Plm2 PCL2 Yox1 Yox1 CDC48 HNT1 Hcm1 Tos4 Tos8 Bud Growth Cell Cycle Control DNA Synthesis and Repair Pog1 Yap5 Yhp1 Yox1 Plm2 TF s have distinct Functional Roles Protein Synthesis and Metabolism Chromosome Segregation and Mitosis Tye7 Yeast transcriptional regulatory network Overview Comprehensive TF dataset 1. ChIp-chip experiments for 11 transcription factors -SBF & MBF Manual collection ChIp-chip experiments Dataset TRANSFAC Kepes' dataset ChIp-chip data by Snyder's lab ChIp-chip data by Young's lab Authors Wingender, E., et al Guelzim, N., et al Horak, C. E., et al Lee, T. I., et al URL suppinfo/ng873_s1.html download.html # of genes # of regulations Computational studies of yeast regulatory network 142 transcription factors 3,420 target genes 7,074 regulatory interactions [Yu, Luscombe et al (2003), Trends Genet, 19: 422] 4

5 Comprehensive Yeast TF network Transcription Factors Comprehensive Yeast TF network Transcription Factors Very complex network Very complex network But we can simplify: Topological measures Network motifs But we can simplify: Topological measures Network motifs Target Genes Target Genes [Barabasi, Alon] [Barabasi, Alon] Topological measures Indicate the gross topological structure of the network Topological measures Number of incoming and outgoing connections Incoming degree = 2.1 each gene is regulated by ~2 TFs Outgoing degree = 49.8 each TF targets ~50 genes Degree Path length Clustering coefficient Degree [Barabasi] [Barabasi] Scale-free distribution of outgoing degree Topological measures Number of intermediate TFs until final target Starting TF Indicate how immediate a regulatory response is 1 intermediate TF Regulatory hubs >100 target genes Dictate structure of network Average path length = 4.7 Final target Path length= 1 Most TFs have few target genes Few TFs have many target genes [Barabasi] [Barabasi] 5

6 Topological measures Ratio of existing links to maximum number of links for neighbouring nodes Comprehensive Yeast TF network Transcription Factors Measure how inter-connected the network is Average coefficient = neighbours 1 existing link 6 possible links Clustering coefficient = 1/6 = 0.17 Very complex network But we can simplify: Topological measures Network motifs Target Genes [Barabasi] [Alon] Network motifs SIM = Single input motifs Regulatory modules within the network HCM1 SIM MIM FBL FFL ECM22 STB1 SPO1 YPR013C [Alon] [Alon; Horak, Luscombe et al (2002), Genes & Dev, 16: 3017 ] MIM = Multiple input motifs FBL = Feed-back loops SBF MBF MBF SBF SPT21 HCM1 Tos4 [Alon; Horak, Luscombe et al (2002), Genes & Dev, 16: 3017 ] [Alon; Horak, Luscombe et al (2002), Genes & Dev, 16: 3017 ] 6

7 FFL = Feed-forward loops Comprehensive Yeast TF network Transcription Factors SBF Very complex network Yox1 Pog1 But we can simplify: Topological measures Network motifs Tos8 Plm2 Target Genes [Alon; Horak, Luscombe et al (2002), Genes & Dev, 16: 3017 ] [Barabasi, Alon] Essentiality of regulatory hubs Hubs dictate the overall structure of the network Represent vulnerable points How important are hubs for viability? Integrate gene essentiality data Transcription Factors % TFs that are essential Essentiality of regulatory hubs Essential genes are lethal if deleted from genome 1,105 yeast genes are essential Which TFs are essential? All TFs non-hubs (<100 targets) hubs (<100 targets) Hubs tend to be more essential than non-hubs [Yu et al (2004), Trends Genet, 20: 227 [Yu et al (2004), Trends Genet, 20: 227 Influence of network on gene expression Transcription Factors Target Genes How does the regulatory network affect gene expression? Yeast expression data [Brown] [Yu, Luscombe et al (2003), Trends Genet, 19: 422] Clustering the yeast cell cycle mrna expression level (ratio) RPL19B TFIIIC [Brown, Botstein, Church] Time-> Microarray timecourse of 1 ribosomal protein 7

8 Clustering the yeast cell cycle Clustering the yeast cell cycle 4 3 RPL19B TFIIIC RPL19B RPS6B Random relationship from 18M Close relationship from 18M (Associated Ribosomal Proteins) Simultaneous Local Clustering algorithm identifies further (reasonable) types of expression relationships Traditional Global Correlation TimeShifted (Algorithm adapted from local sequence alignment) Inverted [Qian et al] Multiple input motifs Single input motifs [Yu, Luscombe et al (2003), Trends Genet, 19: 422] Multiple input motifs % correlated pairs MIM SIM [Yu, Luscombe et al (2003), Trends Genet, 19: 422] [Yu, Luscombe et al (2003), Trends Genet, 19: 422] 8

9 Feed forward motifs Summary of expression relationships Target genes tend to be co-expressed Co-expression is higher when more TFs are involved TFs and target genes have complex relationships Often expression profiles are time-shifted [Yu, Luscombe et al (2003), Trends Genet, 19: 422] [Yu, Luscombe et al (2003), Trends Genet, 19: 422] Co-expression of target genes Expression of TFs and target genes Log ratio with respect to random expectation Log ratio with respect to random expectation correlated time-shifted Target genes tend to be co-expressed Co-expression is higher with more TFs [Yu et al, Luscombe et al] TFs and targets have complex expression relationship Often time-shifted indicating delayed regulation [Yu et al, Luscombe et al] Dynamic Yeast TF network Transcription Factors Analysed network as a static entity Gene expression data Genes that are differentially expressed under five cellular conditions Target Genes But network is dynamic Different sections of the network are active under different cellular conditions Integrate more gene expression data Cellular condition No. genes Cell cycle 437 Sporulation 876 Diauxic shift 1,876 DNA damage 1,715 Stress response 1,385 Assume these genes undergo transcription regulation 9

10 Backtracking to find active sub-network Network usage under cell cycle complete network cell cycle sub-network Define differentially expressed genes Identify TFs that regulate these genes Identify further TFs that regulate these TFs Active regulatory sub-network 142 TFs 3,420 genes 7,074 interactions 70 TFs 280 genes 550 interactions Network usage under different conditions Network usage under different conditions Cell cycle Sporulation Diauxic shift DNA damage Stress Cell cycle Sporulation Diauxic shift DNA damage Stress How do the networks change? topological measures network motifs How do the networks change? topological measures network motifs Our expectation Literature: Network topologies are perceived to be invariant [Barabasi] Scale-free, small-world, and clustered Different molecular biological networks Different genomes Random expectation: Sample different size sub-networks from complete network and calculate topological measures incoming degree path length clustering coefficient outgoing degree Outgoing degree Binary conditions greater connectivity Multi-stage conditions lower connectivity random network size Measures should remain constant Multi-stage: Controlled, ticking over of genes at different stages Binary: Quick, large-scale turnover of genes 10

11 Incoming degree Path length Binary conditions smaller connectivity less complex TF combinations Multi-stage conditions larger connectivity more complex TF combinations Binary conditions shorter path-length faster, direct action Multi-stage conditions longer path-length slower, indirect action Multi-stage Binary Multi-stage Binary Clustering coefficient Network usage under different conditions Binary conditions smaller coefficients less TF-TF inter-regulation Cell cycle Sporulation Diauxic shift DNA damage Stress Multi-stage Binary Multi-stage conditions larger coefficients more TF-TF inter-regulation How do the networks change? topological measures network motifs Our expectation Literature: motif usage is well conserved for regulatory networks across different organisms [Alon] Network motifs Random expectation: sample sub-networks and calculate motif occurrence Motifs SIM Cell cycle 32.0% Sporulat ion 38.9% Diauxic shift 57.4% DNA damage 55.7% Stress response 59.1% single input motif multiple input motif feed-forward loop MIM 23.7% 16.6% 23.6% 27.3% 20.2% FFL 44.3% 44.5% 19.0% 17.0% 20.7% random network size Motif usage should remain constant 11

12 Summary of sub-network structures Network hubs multi-stage conditions binary conditions Regulatory hubs fewer target genes longer path lengths more inter-regulation between TFs more target genes shorter path lengths less inter-regulation between TFs Do hubs stay the same or do they change over between conditions? Do different TFs become important? Our expectation Literature: current opinion is divided: Hubs are permanent features of the network regardless of condition Hubs vary between conditions Random expectation: sample sub-networks from complete regulatory network Random networks converge on same TFs 76-97% overlap in TFs classified as hubs ie hubs are permanent cell cycle sporulation diauxic shift DNA damage stress response all conditions Swi4, Mbp1 Ime1, Ume6 Some permanent hubs house-keeping Unknown functions functions transitient transient hubs Most are transient hubs Different TFs become key regulators in the network Msn2, Msn4 Implications for condition-dependent vulnerability of network permanent hubs Network shifts its weight between centres Regulatory circuitry of cell cycle time-course cell cycle permanent hubs sporulation Multi-stage conditions have: longer path lengths more inter-regulation between TFs diauxic shift stress How do these properties actually look? examine TF inter-regulation during the cell cycle DNA damage 12

13 Regulatory circuitry of cell cycle time-course transcription factors used in cell cycle phase specific ubiquitous early G1 late G1 S G2 M Regulatory circuitry of cell cycle time-course transcription factors used in cell cycle phase specific ubiquitous 1. serial inter-regulation 2. parallel inter-regulation Phase-specific TFs show serial inter-regulation drive cell cycle forward through time-course Ubiquitous and phase-specific TFs show parallel inter-regulation Many ubiquitous TFs are permanent hubs Channel of communication between house-keeping functions and cell cycle progression Summary 1 Use of microarrays for ChIp-chip experiments Identification of genomic binding sites for 11 TFs Large-scale Direct in vivo binding Summary 2 Incorporation of many data sources to build a comprehensive network of regulatory interactions 142 TFs 3,420 target genes 7,074 regulatory interactions Very complex network of inter-regulation Can be simplified Topological measures that describe the overall structure Network motifs that describe regulatory building blocks Summary 3 Essentiality of TFs TFs with many target genes tend to be essential Study of regulatory influence on gene expression Regulatory targets are often co-expressed TF-target expression relationships are complex Summary 4 Dynamic usage of the regulatory network Different sections of the network are used in different cellular conditions Underlying topology of the network changes Network parameters Network motifs Network shifts weight between hubs Serial and parallel inter-regulation between TFs to drive cell cycle progression 13

14 Snyder group and collaborators Michael Snyder Christine Horak Lara Umansky Kevin madden Susana Vidan Ghia Euskirchen Rebecca Goetsch John Rinn Madhuparna Roychowdry Gillian Hooker Mike McEvoy Stacy Piccirillo Sherman Weissman Milind Mahajan Patrick Brown Vishwanath Iyer David Botstein Charles Scafe Kenneth Nelson Yale Microarray Facility Gerstein group Mark Gerstein Haiyuan Yu Sarah Teichmann Madan Babu 14