Basic HTML Version
Full Abstracts – REGULATION OF GENE EXPRESSION II
137
49
Why do transcriptional repressors recruit more than one corepressor?
Priyanka Upadhyai, Gerard Campbell. Biological Sciences, University of
Pittsburgh, Pittsburgh, PA.
Transcriptional repressors function primarily by recruiting corepressors (CoRs), accessory proteins that antagonize transcription by modifying chromatin
structure. Although a single CoR might be sufficient, many repressors, including the
Drosophila
protein Brinker, recruit multiple CoRs, with Brk recruiting
the CoRs, CtBP and Groucho, in addition to possessing a third repression domain (3R). Possible reasons for this are: (a) Quantitative: multiple CoRs provide
more activity; (b) Qualitative: different CoRs have distinct activities; (c) Availability: one CoRs may be inactive or not expressed in some cells, (d) Quality
control: more than one protects against stochastic events. Previous studies indicated Gro is sufficient for Brk to repress targets in the wing, questioning why
it should need to recruit CtBP, a ’short-range’ CoR compared to Gro which can function over longer distances. To resolve this question we have generated a
series of
brk
mutants in which the CtBP interaction motif (CiM), Gro interaction motif (GiM) and 3R are mutated individually or in combination. This was
achieved by generating a
brk
knockout, replacing the coding region with an
attP
ΦC31 bacteriophage integration site. Modified/mutated forms of
brk
were
then integrated into the
attP
site essentially replacing the endogenous gene with these forms. Analysis of these mutants reveals that recruitment of Gro alone
is sufficient for Brk to make a morphologically wild-type fly, but in the absence of a CiM and 3R most animals die as embryos. Gro, however, is not
sufficient for full fertility and CtBP or the 3R domain is required for normal oogenesis. Thus, Gro is not sufficient for Brk activity in specific situations
outside of the wing. These results can be explained by recent studies showing that, although it is ubiquitously expressed, Gro activity is downregulated by
Receptor Tyrosine Kinase activity which occurs in cells in which Brk functions during embryogenesis and oogenesis. Our results are consistent with Brk
needing to recruit more than one CoR because its primary CoR, Gro, is not available in all cells.
50
Transcriptional arithmetic during gene regulatory evolution.
Albert J. Erives. Dept of Biology, University of Iowa, Iowa City, IA, USA.
Morphogen gradients allow cells to infer their positions in a multi-cellular body. These morphogenic systems require two components: (1) a concentration
gradient of a substance over a field of cells; and (2) a mechanism whereby different genes can be induced when the concentration exceeds a gene-specific
threshold level, which is encoded in
cis
at each responding locus. We previously documented an evolutionary molecular mechanism by which concentration
threshold-specific responses to the Dorsal morphogen are encoded in Neurogenic Ectoderm Enhancers (NEEs) during
Drosophila
evolution. Here, I consider
direct consequences of this model given known laws of additivity for transcriptional enhancers, which may act collectively at a locus. I enumerate all
evolutionary paths by which an existing threshold response may be modified with a single mutational step, and find that these possible pathways depend on
the directionality of threshold selection (higher versus lower concentration thresholds). This unexpected asymmetry suggests that partially-redundant
enhancer activities will be frequently generated as an indirect byproduct of the evolutionary maintenance of threshold-sensitive regulatory DNAs. These
results show that the arithmetic logic of transcriptional integration is sufficient to explain many apparent cases of redundant enhancer activities at a locus
even when selection for robustness is minimal.
51
Ancestral sequence reconstruction of the
even-skipped
stripe 2 enhancer in
Drosophila
.
Carlos Martinez
1,2
, Ah-Ram Kim
1,3
, Joshua Rest
4
, Kenneth
Barr
5
, Michael Ludwig
1,2
, Kevin White
2
, John Reinitz
1,2,6,7
. 1) Ecology & Evolution, UC, Chicago, IL; 2) Chicago Center for Systems Biology, UC, Chicago,
IL; 3) Biochemistry & Cell Biology, SUNY, Stony Brook, NY; 4) Ecology & Evolution, SUNY, Stony Brook, NY; 5) Genetics, Genomics, & Systems
Biology, UC, Chicago, IL; 6) Statistics, UC, Chicago, IL; 7) Molecular Genetics & Cell Biology, UC, Chicago, IL.
We have developed a novel computational approach for the custom design of complex
cis
-regulatory sequences capable of expressing in arbitrary patterns,
as well as methods for predicting the evolution and putative ancestral sequences of extant enhancers in
Drosophila
. Our methodology involves the use of a
feed-forward transcriptional model, capable of predicting gene expression patterns directly from enhancer sequence, as a tool for enhancer design and to
provide a functional constraint for predicting enhancer evolution. In the former case, enhancer design was achieved through the use of simulated annealing in
conjunction with a transcriptional model in order to efficiently search the sequence space for novel enhancers having the desired expression pattern. For the
latter, Bayesian inference was used to generate a set of possible ancestral
eve
stripe 2 enhancer (S2E) sequences for a subset of the internal nodes of the
Drosophila
phylogenetic tree. Candidate ancestral sequences were selected for synthesis and experimental validation by checking the model predicted
expression patterns of each sequence against a reference
eve
stripe 2 expression pattern from
D. melanogaster
. In addition, we synthesized and tested
putative ancestral sequences predicted to lie along neutral evolutionary pathways between functionally conserved enhancer elements, such that at all points
along the path the
eve
stripe 2 pattern is maintained.