| Divergent
Functions of Gene Duplicates during Eye Development and Network
Regulation
The
broad goal of my research is to determine how gene duplication
contributes to the architecture and maintenance of regulatory
cascades during development and evolution. Most developmental
genes in Drosophila melanogaster are pleiotropic and often
have similar interacting genes even when expressed in different
tissues. Gene duplication is a key process by which developmental
pleiotropy can be obviated. There are numerous duplication
events in the pathway, and examining the different roles of
these genes can provide insight into the way the compound
eye develops as well as the way in which gene networks and
morphogenetic pathways are regulated.
Many
duplicated genes become pseudogenes and are often rendered
nonfunctional or redundant. Two events can occur when having
an extra copy of a gene is in some way advantageous –
subfunctionalization or neofunctionalization. It is such rare
events that often provide the engine for speciation and adaptive
changes. These phenomena result in new players in a morphogenetic
pathway - each controlling a different aspect of development.
Gene duplication is an important event in the evolution of
the regulatory cascade that specifies eye development in Drosophila
melanogaster.
Over 70% of the RD network comprises paralog pairs that have
varying levels of function and connectivity within the cascade.
The duplicates that have been characterized and have been
shown to affect eye development in our current cascade are
eyeless and twin of eyeless, eyegone and twin of eyegone,
sine oculis and optix, teashirt and tiptop, and distal antenna
and distal antenna-related. I am interested in Tsh and Tio
from a developmental, genetic and evolutionary perspective.
Differential
selection in the RD network
A phylogenetic and genomic analysis of all the RD genes (with
a focus on paralog pairs) that was conducted across 12 Drosophila
species and three basal insects revealed that three of the
duplication events (eyg/toe; tsh/tio; dan/danr) occurred after
the diversification of the Drosophilid lineage (Figure
1). We compared dn/ds ratios of the paralog pairs
and found that each gene duplicate was diverging at a significantly
faster (or slower) rate than its sister gene. An examination
of selection signatures across the coding region of each gene
showed that conserved domains were more highly constrained
than non-conserved regions (Figure 2). Additionally,
dn/ds values of the non-conserved regions varied between each
paralog pair. We were able to identify regions of functional
change based on previous lab work that correlated with domains
with more relaxed selection. This kind of genomic and evolutionary
analysis can be used as a reasonable predictor to identify
areas of functional change, which can then be used as targets
for molecular dissection and analysis.
Teashirt
and Tiptop
Teashirt is a transcription factor with three distantly spaced
C2-H2 zinc finger motifs. It was first identified as a specifier
of trunk identity and segmentation in the trunk. Tsh has a
dual role in eye specification – it can act both as
an eye suppressor as well as an inducer of eye specification.
Misexpression of tsh has been shown to induce ectopic eye
formation in the antenna. Tiptop was identified in 2005 as
a regulator of Drosophila embryogenesis as a paralog of Tsh.
We know now that the basal copy of Tsh/Tio is structurally
similar to Tio. Tio is first detected at stage 10 in the posterior
embryo, with subsequent expression in the brain, gut, trunk
and CNS cells. In the eye-antennal disc, Tio is expressed
anterior to the MF. While Tsh mutants are embryonic lethal,
Tio mutants are homozygous-viable. The presence of Tsh can
compensate for the lack of Tio, and Tio represses Tsh in certain
regions during early embryogenesis. Tsh has three zinc fingers
as DNA binding regions, whereas Tio has four. Basal insects
like Tribolium castaneum, with only one copy of the gene,
have an ancestral copy with four zinc fingers, and the vertebrate
homologs also have a 4th zinc finger. This is particularly
interesting as Tsh and Tio are examples of fairly recent duplicates
with different structures and different mutant phenotypes.
Using
the GAL4/UAS system, Tsh and Tio were misexpressed in developing
tissues throughout fly development and flies were screened
for the presence of ectopic eyes. Tsh was capable of inducing
ectopic eyes with 4 GAL4 drivers, whereas Tio induced ectopic
eyes with 8 GAL4 lines (including those that gave ectopic
eyes with Tsh). We conclude that Tio is a more effective inducer
of ectopic eyes than Tsh. Additionally we find that (a) Tsh
and Tio are expressed at similar levels in the eye antennal
disc during development; (b) the genes can only induce ectopic
eyes in the developing antenna; (c) unlike other RD genes,
Tsh and Tio can coax two separate cell populations into a
retinal fate; (d) Tio can induce multiple ectopic eyes in
a single antennal disc while Tsh cannot, and (e) ectopic eyes
induced by Tio have a full morphogenetic furrow, while those
induced by Tsh have a partial furrow (Figure 3;
Figure 4).
We
are currently trying to elucidate the cause of such subtle
differences between two paralogs. Through a variety of genetic
screens, Y2H screens and structure-function experiments, we
are working on identifying the exact domains that confer functional
changes during the nascent stages of protein evolution. A
deficiency screen revealed that Tsh and Tio both interact
with Homothorax, and are able to suppress a dominant allele
of Wingless. We have also identified an interaction of Tsh
and Tio with the repressor C-terminal binding protein (CtBP),
and are currently working on elucidating (a) the mechanism
by which Tsh and Tio interact with CtBP; (b) whether the repressive
properties of Tsh and Tio are due to CtBP binding, and (c)
if the strength of suppression is different between Tsh and
Tio. The structure/function analysis involves a systematic
deletion of regions of the gene and subsequent crosses to
GAL4 lines. Results will be compared to the dn/ds analysis
to confirm our hypothesis about differential selection across
coding regions.
|
