Abstract:
The fruit fly Drosophila melanogaster has been used in studies of genetics since
the early 1900s, and is a powerful model system for both experimental evolution
studies based on laboratory selection and studies focussing on the genetic control of
developmental processes. It is, thus, an ideal system with which to address questions
pertaining to the developmental and molecular biological underpinnings of adaptive
evolutionary change in life-history related traits, an approach often termed
developmental evolutionary biology. Laboratory selection experiments also provide an
opportunity to address the reversibility, or lack thereof, of microevolutionary
trajectories. In this thesis, I present results from two lines of investigation I carried out
on a set of replicate D. melanogaster populations subjected to selection for rapid preadult
development and early reproduction for over 250 generations. One one hand, I
studied the evolutionary trajectories of several life-history related traits in these
populations when subjected to 54 generations of reverse selection. In a separate set of
experiments, I examined the expression levels of certain developmentally important
genes in specific life-stages or tissues, as well as genome-wide expression levels in
larvae, pupae and young adults of the selected populations and their ancestral controls.
When I started my work, the four replicate selected populations of D.
melanogaster (FEJ1-4) had already undergone 250 generations of selection for faster
pre-adult development and early reproduction, and had diverged substantially for a
variety of traits from the four matched ancestral control populations (JB1-4) that were
maintained on a 21 day discrete generation cycle with no conscious selection on
development time and early reproduction. Briefly, relative to the JBs, the FEJs showed
reductions in the duration of all pre-adult life-stages, larval survivorship, body size and
ix
dry weight, lipid and glycogen content, adult lifespan and starvation and dessication
resistance, and early life as well as lifetime fecundity. The FEJs also showed
significantly reduced larval feeding rate and growth rate, foraging path length, digging
propensity, pupation height and urea tolerance. Relative to the JBs, the FEJs had higher
fecundity per unit dry weight early in life and took a longer time from eclosion to first
mating.
The reversibility of evolution has been debated extensively but rarely studied
empirically except for a couple of studies on D. melanogaster and E. coli by M. R.
Rose and R. E. Lenski, and colleagues, respectively. The reversibility of evolved
phenotypes depends on different factors that could have changed during the course of
forward selection, such as the availability of genetic variation, complexity and pattern
of epistatic interactions, and accumulation of mutations. I derived four populations
(RF1-4) from the FEJs, returned them to the ancestral JB maintenance regime and
studied the trajectories of several traits over 54 generations of reverse selection. I found
that larval and egg-to-adult survivorship, egg duration and early-life and middle-life
fecundity converged back to ancestral control levels, whereas larval, pupal and egg-toadult
duration and dry weight at eclosion did not converge completely. During the
terminal few assays of the RFs, the correspondence between development time and dry
weight at eclosion was parallel to that seen in the first 20 generations or so of forward
selection in the FEJs, suggesting that despite incomplete convergence, the joint
trajectory of these traits was similar under both forward and reverse selection. I also
observed that the response to reverse selection with respect to durations of different
pre-adult life stages was similar: the response was slow in the beginning up to
generation 5 of reverse selection and hastened up thereafter and was fast till generation
x
25, and after that again slowed down. My observations on development time and dry
weight at eclosion are consistent with those of M. R. Rose and colleagues who used
flies from the same ancestry and subjected them to selection for rapid development in
much the same way as us. However, in their study, fecundity did not converge back to
ancestral levels, and this difference in our results is probably due to the “early
reproduction” part of the selection protocol being very different between the two sets of
studies. Overall, the degree and mode of convergence I observed for the traits studied
suggests (a) no erosion of genetic variation for these traits over 250 generations of
forward selection in the FEJs, and (b) that it is unlikely that novel patterns of epistasis
or new mutations have accumulated in the FEJs over the course of forward selection.
My results also suggest that the broad contours of reverse evolution trajectories may be
quite repeatable across studies if the past selection history and starting genetic material
have been similar.
Regulating gene expression is a key step by which an organism activates the
information encoded in its genome to effect developmental changes, and differences in
this regulation can cascade through development resulting in different morphological or
physiological character states. Keeping this in view, I studied the gene expression
through different methods in FEJs in comparison to the JB controls. Drosophila
neuropeptide F (dnpf) is a homolog of mammalian NPY gene which is involved in
food/foraging-related behaviors in mammals. dnpf is expressed in the central nervous
system of Drosophila and plays a major role in the maintenance of foraging behavior.
Its expression is high at foraging stage (early third instar) and low in the wandering
stage (late third instar) in wild type larvae, and dnpf downregulation has been shown to
act as a switch between foraging and pupation behavior in Drosophila. In a gene
xi
expression study done through semi-quantitative RT-PCR method, I found that dnpf
expression in JBs was as expected (i.e. high at early third instar and low at late third
instar), whereas dnpf expression in FEJs was low right from early third instar larva and
it did not change till late third instar. This change in temporal pattern of dnpf expression
could be an important causal factor underlying the huge reduction in larval third instar
duration observed in the FEJs.
Precise spatial and temporal expression of genes is important for proper pattern
formation during development. In FEJs, some leg and wing malformations had been
observed from about the 100th generation of selection. Therefore, to check if there was
any change in the expression pattern of developmentally important proteins, I studied
the expression patterns of some such proteins in the embryos as well as in the wing
discs of third instar larvae by antibody staining technique. I observed no significant
difference in the spatial expression pattern of these proteins in FEJs compared to their
JB counterparts, suggesting that the expression patterns of these developmentally
important proteins have not changed in FEJs over the course of selection.
I also examined cell number and cell size in FEJs relative to the to JBs by
staining wing discs of third instar larvae with antibody against the protein Armadillo,
whereby one can mark the cell borders, count the cells and estimate their sizes. Using
this technique, I found that FEJ wing discs had less number of bigger cells whereas JBs
had more number of smaller cells. The reason for this is not clear at this time, but it
may be that the FEJs have evolved a reduction in the number of cell divisions as part of
a strategy to conserve energy.
xii
I further subjected one replicate population each of the FEJs and JBs to
microarray analysis to examine differences in genome-wide patterns of gene expression
between selected and control larvae, pupae and young adult males and females. I found
that expression level of a few hundred genes was changed in FEJs in different lifestages
used for the analysis. These changes were in both the directions i.e, many genes
were up-regulated and many were down-regulated in FEJs in comparison to JBs, and
many genes were consistently differentially expressed in FEJs across all life-stages
studied. Genes related to epigenetic control were up-regulated in all the stages studied
suggesting that changes in expression of many genes are possibly mediated by
epigenetic mechanisms in the FEJs. Further, gene ontology (GO) term enrichment
analysis using DAVID online bioinformatics tool showed that among the up-regulated
genes were many eclusters of genes related to translation, developmental processes,
phagocytosis etc., all of which are related to development. The down-regulated genes
were related to glutathione metabolism which consist genes such as glutathione-Stransferase
which is involved in oxidative stress mechanism. FEJs are less resistant to
different stresses compared to JBs. This could be because of the down-regulation of the
genes involved in glutathione metabolism. Further, it was observed that the genes
involved in the insulin signaling pathway are down-regulated and that of ecdysone
action were up-regulated in the FEJs. The final body size of Drosophila is known to be
greatly affected by an antagonistic interaction of insulin signaling and ecdysone action,
and these results suggest that the faster development of FEJS, and their smaller body
size, could be mediated by the evolution of higher basal levels of ecdysone and reduced
levels of insulin signalling. Though preliminary in nature, the gene expression results
indicate several avenues of further research that are likely to enhance our understanding
xiii
of the molecular genetic and developmental underpinnings of the rapid development
phenotype in the FEJs.