Working Group 4: Brain, conservation, evolution

Tissue and cell types

• most of the data available for brain is tissue - more primary cell types would be good to have. In addition including Macaque (primary cells + tissue) would make evolutionary analysis much stronger.
  • Alka - cortical neurons can be FACS sorted from adult mice
  • mouse neurons are commercially available; Silvia will talk to Stefan (?) about options
• Peter - fresh tissue for an experiment on the effects of post-mortem delay.
• Silvia - frozen tissue, can't isolate cells from this, but it is immediately post-mortem
• Alka - neuron sorting method, will send details to Silvia, Peter-done 25.02.2011
• time series of tissue from post-mortems if neonatal material will become available in time.
  • Al - feasibility, replicates, etc.

Regions of interest

• Silvia and Peter discussed specific brain regions associated w/ disease matched to CAGE libraries
  • network analysis required
  • genes responding in unusual states
• is tissue decomposition possible using the primary cell libraries? e.g. brain tissue region = how much percent astrocytes, ...?

Disease

• Alzheimer's disease, Parkinson's disease, frontal temporal dementia, Huntington's disease
  • induced pluripotent stem cells available for PD, something in the works for FTD, Silvia will talk to collaborators about these
• Silvia - non-coding RNA and pathology
  • Huntington's disease - mouse model available
  • schizophrenia patients
• Alka - non-coding RNA regulating 3 genes in Rett Syndrome

Timecourses

• Peter - we have fetal and adult brain, poss. neonatal brain
  • tissue timecourse
• Thomas and Dan are doing something similar in mouse

Evolution in the context of brain

• See this page for the analysis of human / macaque brain CAGE data: monkey brains
• expression and genomic variation in the context of evolution
• missing from F5 is another primate species - Peter has tissue for brain for human, macaque, mouse
• which promoters change behaviour in primates? in humans?
• positive selection in the human brain
  • potential to test with CAGE data
  • need to compare against another tissue e.g. hepatocytes, aortic smooth muscle, mesenchymal bone marrow which we already have for the 5 species, but we would like to have the other primate
  • Martin - group has experience with sequence-based measures of evolution
  • need measures of selection on the expression level, a couple of Nature papers mentioned
  • signatures of expression in different tissues, how these change b/w species
    • tissue partitioning

Liver evolution

• Martin Taylor, Alison Meynert
• Hepatocyte ChIP-seq for HNF1A and CEBPA in human/mouse/dog/chicken Schmidt et al, Science 2010
• combine this with the CAGE data for those species
• The questions we can address with this study are:
  • Are conserved binding sites more likely than non-conserved sites to elicit a local, hepatocyte specific ranscriptional response? (Use CEBPA non-expressing cells to generate a background model of proximal transcriptional responses). This could be used to estimate "functional turnover" as opposed to the "binding turnover" as reported by Duncan Odom.
  • Does hepatocyte specific expression (around binding sites) segregate through species lineages with the experimentally defined binding site?
  • If there is apparent turn-over of binding sites, is the pattern of local responsive transcription conserved?
  • Do we see conservation of hepatocyte specific transcriptional responses even in the absence of binding site conservation?

Networks

• identification of highly conserved/divergent sub-networks of regulation
• Network analysis across organisms & evolution of regulatory networks, in particular of developmental networks. Are there any subnetworks particularly conserved between organisms? What does this tell us about the functional importance and relevance of specific subnetworks? Do we see any recurring patterns in the network (Uri Alon-type feed-forward loops)? What are the conservation patterns and rates of divergence of transcription factors and specific regulatory relations? Do we see turnover of TFBSs, or do we see conservation of TFBSs in alignments? This can be applied specifically to brain, or more generally to all CAGE samples. TFBS prediction in Neanderthal compared to Homo sapiens. [Michiel].
• TFBS identification has to be independent of evolutionary analysis if you're going to use them in this context
  • Martin Frith's algorithm can be used for this purpose (Clover?)
• perturbed networks in disease, Al - focus is on cell level rather than tissue, cellular composition of tissue is a major complicating factor
  • a single gene perturbation e.g. Huntington's disease would be easiest

Promoter shapes

• James and Colin - epigenetics initially, but also looking at how shapes change over evolutionary time and between cell types
• how the sequence divergence of the promoter is related
• Terry - how quickly do promoters change b/w strains of mice (Sanger sequencing of mouse inbred strains)
  • mosaicism is a big issue here, the data is not very clean, wild mice would be better

People

• Peter Heutink
• Cesare Furlanello
• Thomas Jonghyun Ha
• Silvia Zucchelli
• Michiel De Hoon
• Alka Saxena
• Terry Meehan
• John Kenneth
• Alessandro Bonetti
• James Prendergast
• Alison Meynert
• Martin Taylor
• Al Forrest
• Morana Vitezic

Proposed satellite papers

The evolution of gene expression in vertebrates

• Martin Taylor, Peter Heutink, Alison Meynert ...?
• Using well matched tissues over 5 (or 6 with Macaque) species.
• Transcriptional constraint, sequence evolution.

The evolution of biological systems and tissues

• Martin Taylor, Peter Heutink, .....?
• Investigate evolution of transcription regulatory sequences for genes partitioned by cell/tissue and tissue specificity.
• Focus on one expression in reference organism (Human) and relate to mammalian scope sequence evolution.

Functional TFBS turnover and regulatory evolution

• Martin Taylor, Alison Meynert
• Investigate TFBS turnover in liver in the context of proximal TF expression correlated regulation. I.e. Are more conserved experimentally validated TFBS's more likely to show evidence of function.

Disease

• Peter Heutink, Silvia Zucchelli,.....
• Comparing control with disease state for Mendelian neurodegenerative disorders (Fronto Temporal Dementia, Huntington's, Alzheimer's, Parkinson's?)
• Several levels of analysis: Human post-mortem brain, transgenic mouse models and/or cellular, iPS cells of patients.
• Validate identified networks in cellular systems using variety of techniques.
• Transcriptional mis-regulation in the mecp2 null mouse visual cortex at 3 critical stages of development (Alka Saxena, with Michela Fagiolini, Erik Arner)
• Genomic architecture of genes involved in Rett Syndrome (Promoter and isoform usage in various tissues, expression levels, if new isoforms are discovered, there is a possibility of screening mutation negative Aussie Rett patients for pathogenic mutations within new isoforms)