Study of how catabolite repression and ribosome levels determine cell growth in batch cultures of Saccharomyces cerevisiae
Abstract
In response to environmental changes, cells launch new programmes of gene expression.
Here I address how cells with their finite proteomes regulate ribosomal levels as the
environment changes in batch cultures and, given two carbon sources, how cells prioritise
which to use. In the initial chapters, I describe how I measure the dynamics of both growth
and ribosome levels in S. cerevisiae using microplate readers. I show that growth only
enters a prolonged exponential phase if the initial sugar concentration is sufficiently high,
otherwise the specific growth rate peaks rather than plateaus. Nevertheless, Monod’s
law still holds for the initial sugar concentration and the maximum specific growth rate.
Using GFP-tagged ribosomes, I then measure the population’s total ribosome level and
demonstrate that the population’s growth rate is proportional to this level in the early
phases of growth. I go on to define the effective translation rate as the ratio between the
population’s growth rate and its ribosome level, and find that in exponential phase, the
effective translation rate and the ribosomal fraction are constant over time. Further, the
results of challenging cells with different stresses suggest an empirical upper limit to the
effective translation rate. To understand these findings, I develop a minimal self-replicator
model and analyse its behaviour both at and away from steady state. I extend this model
to include carbon and nitrogen and derive mechanistically a form of Monod’s law for
two substrates. In the final chapters, I investigate how cells prioritise the use of non-glucose sugars, specifically galactose and palatinose, which is little studied. I show that
cells exhibit diauxie in galactose-palatinose mixtures, prioritising galactose. In addition,
I demonstrate that constitutively active Gal4 in a gal80∆ mutant causes a long delay in
palatinose metabolism, and that this delay can be mostly alleviated by deleting the gene
for the galactose transporter. To investigate the cause of this effect, I perform an RNAseq
Gal4 signalling. With this discovery, I build a simple model to understand how the Gal4
signal affects the inducibility of the MAL network, which predicts that the isomaltases are
excessively expressed relative to MAL11 when Gal4 is constitutively active. I therefore
delete IMA1, which encodes one of the two isomaltases, from the gal80∆ mutant, and, as
expected, this mutant can use palatinose with little lag. My results therefore provide a
novel example of non-glucose catabolite repression in S. cerevisiae and how one network
can affect another by changing its inducibility.