Systems Biology of C. elegans Stuart Kim Stanford University
Overview Cell lineage Genome Mutant Phenotypes Protein interaction Gene Expression
Early cell divisions
B A S M E C D 4P C. elegans zygote lineage of 959 cells
Overview Cell lineage Genome Mutant Phenotypes Protein interaction Gene Expression
100 Mb ~19735 genes 1998
94% genes unknown (~1998) 1998
94% genes unknown (~1998) interactions for 92% of genes (~2001) 1998
Overview Cell lineage Genome Mutant Phenotypes Protein interaction Gene Expression
RNA mediated interference (RNAi) ds RNA exogenous cleaved to 22 NT pieces mrna + endogenous causes specific degradation of mrna mrna protein mutant phenotype
RNAi in C. elegans Feed worms strains of E. coli expressing ds RNA
Download movie of mcm-5(rnai) http://mpi-web.embl-heidelberg.de/dbscreen http://www.wormbase.org www.rnai.org http://watson.genes.hib.ac.jp/db/rnai_s/index.html
Mutant cell divisions
RNAi screen for 417 fat genes
RNAi screens in C. elegans (last two years) No mapping/cloning etc. Order bacterial strain set and do experiments yourself. Examples of screens: ~500 longevity 417 fat genes Genes needed for RNAi Embryonic polarity synaptic function
Overview Cell lineage Genome Mutant Phenotypes Protein interaction Gene Expression
Global Yeast Two Hybrid Subclone all ORFs into Yeast two hybrid system Systematically describe potential protein-protein interactions
Global Yeast Two Hybrid X Y 2898 nodes connected by 5460 edges
Protein interaction network et al. Vidal, 2004
Overview Cell lineage Genome Mutant Phenotypes Protein interaction Gene Expression
Gene expression Networks Eran Segal (Weizmann) Josh Stuart (UCSC)
C. elegans gene expression database 553 DNA microarray expts X ~19,000 genes C. elegans expression database (# hybridizations)
Cluster of co-expressed genes from DNA microarray experiments 553 DNA microarray experiments Expressed together Which are functionally related? Which are physiologically relevant? ~20,000 genes
Evolutionary conservation implies functional interaction Gene A Gene B Gene C A and B are functionally related C is co-expressed, but not functionally related During evolution, all promoter sites are mutated Only enhancers of functionally related genes are preserved Gene A Gene B Gene C
Conserved co-expression human Functionally related to YFG? Selective advantage for co-expression YFG Evolutionarily conserved May see co-expression in: 78 Fly 335 YFG Mouse 74 Worm Yeast 1209 YFG 167 234 YFG 1760 754
worm human gene 1 gene 2 gene 3 gene 1 gene 2 gene 3 Microarray data exp 1 exp 2 exp 3 exp 1 exp 2 exp 3 multi-species topomap Gene correlation matrices gene 1 gene 2 gene 3 gene 1 gene 2 gene 3 gene 1 gene 2 gene 3.8.3.2 gene 1 gene 2 gene 3.9.1.2 gene 1 gene 2 gene 3 gene 1 gene 2 gene 3 Gene rank matrices gene 1 meta-gene 2D scatter plot gene 2 gene 3 4 1023 8 2438 983 1508 gene 1 gene 2 gene 3 10 3231 21 2081 3373 2175 meta-gene p-value matrix meg 1 meg 2 meg 3 meg 1 meg 1 meg 2 meg 3.001.5.02.3.4.2 meta-gene network meg 3 meg 2
Gene Expression Network
Gene Expression Topomap 6886 meta-genes >1000 unknown
Ribosomal subunits
A gene expression microscope Plan A = microarrays Flo Pauli Plan B = GFP reporters Xiao Liu (Fuhui Long, Hanchuan Peng, Gene Myers)
Convert image to a data table description Age Age Age Intestines high level in pharynx High level Muscles in body wall muscle Expressed Neurons in some hypodermal cells Gene Gene 1 Gene 2 3 genes 959 cells genes quantitation 959 cells
A large number of GFP-reporter transgenic worms 3-D confocal data stack 959 cells Automatic cell lineage analyzer genes Quantitative measurement of GFP expression at the level of single cells
Long Term Plan Construct a imaging system to automate cell identification of C. elegans Generate GFP-reporter strains in a large scale Use this GFP analysis system to study aging
GFP = BWM (landmarks) RFP = expression DAPI = all nuclei
Automatic GFP analyzer Registration = align the image to a standard virtual worm model Segmentation = find the nuclei Recognition = name the nuclei Expression = measure GFP levels in each nucleus
Registration = align image to a standard virtual worm model
DR DL DR DL VR VL VR VL
Segmentation = find the nuclei
manual annotation to recognize nuclei 1 6 9 11 17 25 25 28 29 31 36 55 61 66 79 cell slice Peak Val Mean Val Size Mass cell name explanation 1 92 108.00 62.14 482 29951 bun_1_1 bwmvl1 2 32 165.00 103.54 892 92360 bun_3_1 bwmdr1 3 32 128.00 74.43 1895 141054 bun_2_1 bwmvr1 4 91 152.00 87.37 1353 118212 bun_4_1 bwmdl1. 79 68 130.00 70.34 2258 158834 bun_3_22 bwmvr24
Gene expression profile at the level of single cells Body Wall Muscle Intestine myo-3 Tissue Specificity
Gene expression profile at the level of single cells Body Wall Muscle Intestine myo-3 mir-64 Gradient of gene expression
Gene expression profile at the level of single cells Body Wall Muscle Intestine Different levels of transcriptional noise myo-3 mir-64 pha-4
Gene expression profile at the level of single cells Body Wall Muscle Intestine myo-3 mir-61 pha-4
~200 GFP-reporter transgenic worms/yr 3-D confocal data stacks ~200 genes 6 conditions ~300 cells 6 developmental time points Automatic cell lineage analyzer Quantitative measurement of GFP expression at the level of single cells
Systems Biology + OR
Systems Biology of C. elegans Conserved gene expression modules Josh Stuart Eran Segal (Daphne Koller) Automatic GFP lineage analyzer Xiao Liu Fuhui Long Hanchuan Peng Gene Myers