Rocio Espinosa Portero:
Regulation of the Escherichia coli protein synthesis machinery in response to phosphate starvation

Date: 13-05-2022    Supervisor: Sine Lo Svenningsen & Michael Askvad Sørensen




Protein synthesis is one of the most tightly regulated processes in Escherichia coli, due to its high energetic cost. During balanced growth, efficient protein synthesis carried out by the ribosomes and associated machinery ensures bacterial growth optimization. Research on the regulation of ribosomes, and in particular ribosomal RNA (rRNA), including synthesis, maturation, and degradation, during balanced growth is extensive in E. coli. However, outside of the steady-state growth scenario, such as under phosphate (Pi) starvation, protein synthesis regulation is not well known. Part of the E. coli stress response to nutrient starvation is degradation of rRNA. Starvation for phosphate was shown by our laboratory to be among the conditions that result in extensive rRNA degradation.

During phosphate starvation, the stringent response – a global stress response in E. coli – is triggered by synthesis of guanosine tetra- and pentaphosphate (collectively referred to as ppGpp). Downregulation of rRNA synthesis, along with an overall transcriptional profile change within the cell, is one of the most significant effects of ppGpp. SpoT, but not RelA – which are the two enzymes responsible of ppGpp synthesis – has been described to activate the stringent response upon phosphate limitation.

In this thesis, I expand the knowledge on how E. coli adapts its translation machinery to phosphate starvation. I quantify the degree to which DNA, RNA, and protein accumulation is reduced upon phosphate limitation. Importantly, I show that mRNA availability, rather than ribosomal levels, become limiting for protein synthesis when E. coli cells are starved for phosphate. rRNA degradation provides the nucleotides which are recycled to maintain basal level of transcription and protein synthesis.

I also show that the initial activation of the stringent response is RelA-dependent. ppGpp synthesis by RelA is not affected by the presence of amino acids in the medium. Moreover, nucleotide pools (adenosine triphosphate and guanosine triphosphate, ATP and GTP, respectively) are depleted upon phosphate starvation. We hypothesized that a decrease in ATP and the low binding affinity for ATP of aminoacyl tRNA synthetases could lower the charging level of one or more tRNAs, which in turn activated RelA-dependent ppGpp synthesis. In an attempt to narrow down the pool of tRNAs, I analyzed codon usage frequencies of the Pho regulon, which remain generally similar to steadystate growth. Thus, in the search for such tRNA, I quantified the charging levels of 14 tRNA isoacceptors. The charging levels of those tRNAs remained constant or increased slightly upon phosphate starvation. To complete the search, I performed a CLIP-seq (CrossLinking and ImmunoPrecipitation followed by deep sequencing) experiment during phosphate starvation by IV crosslinking RelA to RNA, purifying this RNA, and sequencing the results. The analysis of these results will potentially reveal one or more tRNAs that become uncharged upon phosphate starvation.

In addition, by using phosphate starvation as a condition to study rRNA degradation, I show that certain modifications of the rRNA do not affect rRNA stability. Moreover, because the carbon-stressinduced rRNA degradation pathway is initiated by the endoribonuclease RNase E, I constructed a plasmid which contains a mutated sequence of the 16S rRNA which do not include the initial cleavage sites by RNase E. Preliminary results show that an alternative pathway takes place in rRNA degradation during phosphate starvation when RNase E cleavage is impaired.