Saturday, February 27, 2016

The latest in gene regulation: a tribute to François Jacob

A conference titled "Genetic Control of Development and Evolution" was held at the Pasteur Institute in Paris in the autumn of 2015. The conference was a tribute to François Jacob for his contribution to the study of gene regulation. With the help of QCBS funding I was able to present my work at this conference.

My poster at the conference was about endosymbiotic bacteria of ants of the genus Camponotus. We show that the symbiont has driven major rearrangements in early egg development in order to ensure stable endosymbiosis and transmission to the next generation. I received excellent feedback at the conference that helped understand the data in a better light.

The conference was a broad update on gene regulation with reference to development and evolution. Since Jacob and Monod a lot has been learnt about gene regulation. Enhancers, the DNA elements that regulate the levels, timing and spatial extent of gene expression was one of the major themes of the conference. Many presenters showed how suboptimal binding sites make optimal enhancers wherein weak binding sites and ‘half’ binding sites play a major role in specificity of gene expression. Michael Levine of University of Berkley, uses artificial enhancers generated by mutating transcription factor binding sites and the spacing between them. He showed that native enhancers are weak in comparison to artificially generated optimized strong enhancers. The native enhancers are more specific, the stronger enhancers show mis-expression. Kenneth Zaret of the University of Pennsylvania showed that the recognition sites of Pioneer Factors (proteins that initiate release of chromatin bound DNA making it active) are half of the recognition sites of other transcription factors. This feature allows them to bind while allowing the other side of DNA to remain bound to histones. This binding is not sufficient and requires cooperativity and recruitment of other factors meaning many of these bindings are abortive, hence improved specificity. Justin Crocker from Howard Hughes Medical Institute - Ashburn, showed that in the shaven baby locus of Drosophila the relative affinity and specificity of binding sites have an inverse relationship implying that high affinity binding sites may actually bind other genes and decrease specificity. When he changes a low affinity site into a high affinity site it results in mis-expression. He also showed that in different species low affinity binding sites are poorly conserved but are functional in their respective contexts. Marion Guéroult-Bellone from CNRS Montpellier showed how in vitro expression assays are different from in vivo data in terms of levels of expression and reiterated a known fact that spacing between binding sites regulates levels of expression. François Spitz of EMBL Heidelberg showed that long distance enhancers act dynamically activating several genes at the same time where an appropriate TATA box is found at specific optimal distances from it. They use a reporter assay with one enhancer and test for all expressed regions of about 200 kbp region in both directions. If the enhancer was flipped, it affected a reporter at a similar distance in the opposite direction and resulted in a pattern mirroring its activity albeit only when appropriate TATA boxes fell in its range of activity. This pattern of activation is consistent with Topologically Associating Domains; meaning in 3D when a piece of DNA falls in the vicinity of another piece of DNA they interact regardless of their base-pair distance. 

Another major theme of the conference was chromatin remodeling for gene regulation.  Edith Heard from Curie Institute in Paris talked about mono-allelic gene expression in mammals. Repression of the X-chromosome copies in females is well-known, she showed that the X-chromosome that gets inactivated is randomly chosen such that two females, even if identical twins, are different. In addition there are two inactivation waves initiated by the Xist locus. The first one initiates early during development and is maintained throughout development, this one is initiated by the paternal locus. The second one is initiated at the blastocyst stage that initially kicks in both alleles and then stabilizes to persist in one allele inactivating it and not does not do so in the other allele that stays active.  This means that females are mosaics due to random X-inactivation. Gerarld Crabtree of Stanford talked about BAF complexes that are polycomb repressor complexes (PRCs), which seem to be opposing both PRC1 and PRC2 regulating both assembly and disassembly of chromatin.  Bluma Lesch of Whitehead Institute Cambridge, USA showed that developmental regulatory genes remain in the poised state in the Germline in which they contain opposing histone states H3K4me3 and H3K27me3 and are transcriptionally repressed. They continue in the poised state in the gamete stage and this phenomenon is conserved from drosophila to mammals.

A couple of interesting talks used transcriptome sequence analysis to establish relatedness of cells or tissues. Jacob Musser of EMBL Heidelberg used this method to test for relatedness between dinosaurs feathers/scales and bird feathers using scales on the feet of chickens. Principle component analysis of transcriptome sequences of chicken feathers, chicken scales (of feet), chicken nails, alligator scales, and alligator nails show relatedness between bird feathers (not scales of feet) with alligator scales. Alexander Van Oudenaarden of Hubrecht Institute, Utrecht has managed to find a way to separate single cells from the bone marrow blood precursors and RNA sequence them one cell at a time to look for shared expressed sequences. He then correlates their spatial proximity or distance with the number of shared genes. This analysis confirms some of known relationships in tissue precursors and points to spatial information in precursor cells of bone marrow, a site for multiple stem cells. Shahragim Tajbakhsh of Pasteur Institute – Paris showed that the shared program for development of the craniofacial muscles and esophagus muscles is not conserved in birds (birds do not need to chew and swallow). But interestingly basal animals (Ciona being one) have this program. He suggested that the birds might have lost it and more interestingly the craniofacial muscle developmental program may actually have been coopted from the esophageal program. 

Patricia Wittkopp of University of Michigan does experimental evolution in Drosophila and yeast from different wild and lab sources. She has developed a system with two different cis-alleles for the same gene, each driving a different FP (red/yellow) in the same organism; meaning the same trans-acting background driving two different cis-alleles. Using this system she shows that there is higher cis-regulatory divergence in an evolving population, implying that cis regulatory divergence may be favored by natural selection. She also uses induced C->T and G->A mutations and compares them with natural occurring polymorphisms, which showed that natural polymorphism form a subset of induced mutations in terms of noise indicating that noise is selected against in natural selection.

Overall great fun and highly inspiring!