rps5-T38A
Mutant
After recruitment of the 43S PIC to the 5’ cap of the mRNA, the 43S complex scans along
the 5’ UTR until it encounters a start codon in the ribosomal P-site. Here, not only the discrimination against non-AUG codons but also the sequence context of the AUG are
crucial for efficient start codon selection (see Section 1.4.1).
First, the discrimination between AUG and non-AUG start codons was addressed using the dual luciferase reporter pRaugFFuug encoding AUG-Renilla and UUG-Firefly luciferase
reporter mRNAs (Cheung et al., 2007). This reporter plasmid allows to assess in parallel the initiation from AUG and UUG start codons at steady state in vivo by dual luciferase measurement. The absolute expression levels of AUG-Renilla and UUG-Firefly luciferase reporter were both reduced in rps5-T38A mutant cells compared to RPS5 wild-type. Interestingly, the effect of therps5-T38Amutation was much stronger on the AUG-Renilla (33 ± 11 %) than on the UUG-Firefly reporter (54 ± 15 %), thus increasing the UUG vs.
Figure 2.11: Therps5-T38Amutation alters the association of initiation factors. (A+B) RPS5 wt (A) and rps5-T38A (B) mutant cells were treated with cycloheximide prior to harvesting. WCEs were separated by sucrose density gradient centrifugation, and gradient fractions were subjected to Western blot analysis using antibodies against the indicated initiation factors and ribosomal proteins. The signal of each eIF in the 40S or 80S/polysomes fractions (boxed) was normalized to the corresponding Rps8 signals. (C) The relative levels of rps5-T38Ato wild-type were calculated and plotted (mean±SD; n= 3).
AUG ratio almost twofold (Fig. 2.12A). This finding suggests that therps5-T38Amutation impairs start codon recognition in vivo.
To assess the discrimination against different start codon consensus sequences SUI1-lacZ fusion reporters were used inrps5-T38A mutant andRPS5wild-type cells (Martin-Marcos et al., 2011). The native poor initiation context of the SUI1 gene with -3CGU-AUG is
reported to be unfavorable and to control Sui1 expression via its own function in start codon selection (Ivanov et al., 2010). The poorSUI1context was modified to perfectly match the yeast optimal consensus context of-3AAA-AUG (SUI1-opt-lacZ) and to exacerbate the poor
SUI1 context by introducing-3UUU-AUG (SUI1-UUU-lacZ) as an even more unfavorable
initiation codon context (Fig. 2.13 A) (Shabalina et al., 2004). Since the different reporters encode lacZ fusions the context dependent initiation can be measured by β-galactosidase assay in vivo(Miller, 1972). As expected inRPS5wild-type cells the SUI1-opt-lacZ fusion was expressed more than twice as much as the wild-type SUI1-lacZ fusion, whereas the
Figure 2.12:Therps5-T38Amutation leads to increased initiation from non-AUG start codons. (A) Schematic illustration of the dual luciferase reporter pRaugFFuug encoding AUG-Renilla (i) and UUG-Firefly (ii) luciferase. (B)RPS5wt andrps5-T38Amutant strains harboring the pRaugFFuug plasmid were analyzed by dual luciferase assay. Absolute expression levels of AUG-Renilla [RLU] and UUG-Firefly [FLU] in RPS5 wt and rps5-T38A mutant cells were plotted (graph) and normalized to wild-type (table below; mean ±SD;n= 4). (C) Ratios of expression levels of UUG vs. AUG reporter were calculated (graph) and normalized to wild-type levels (table below).
SUI1-UUU-lacZ was expressed at a level 4.4 times below the SUI1-opt-lacZ (Fig. 2.12 B). Intriguingly, these differences in the expression level between optimal, poor and highly unfavorable start codon context in SUI1-opt-, SUI1- and SUI1-UUU-lacZ, respectively, was completely lost by the rps5-T38A mutation (Fig. 2.12 B). Thus, phosphorylation of Rps5 at T38 is needed for the discrimination between good and poor start codon context.
A powerful assay to investigate scanning processivity was developed by Berthelot et al.
using a series of Firefly luciferase encoding constructs with 5’ UTRs that range in their length between the 5’cap and the start codon from 43 to 607 nucleotides (Fig. 2.14 A)
(Berthelot et al., 2004). The in vitro transcribed mRNAs were translated in extracts of
RPS5 wt and rps5-T38A mutant cells. The time 40S subunits need to locate the start codon was estimated by comparing the time courses of protein synthesis observed with the mRNAs carrying the different length of 5’ UTR. Considering the different delay times
Figure 2.13: Therps5-T38Amutation eliminates discrimination between different AUG context sequences in SUI1-lacZ reporters. (A) Representation of differentSUI1-lacZ reporter plasmids: (i) wild-typeSUI1context; (ii) optimal consensus sequence SUI1-opt-lacZ and (iii) highly unfavorable context SUI1-UUU-lacZ. (B) Expression of differentSUI1-lacZ reporter plasmids depicted in (A) inRPS5wt andrps5-T38Amutant cells was measured byβ-galactosidase activity (miller units U). Ratios of SUI1-opt-lacZ/SUI1-lacZ orSUI1-opt-lacZ/ SUI1-UUU-lacZ were calculated (table below; mean±SD;n= 4).
for different length eliminated the contribution of the early, undefined phase of loading and activation of scanning competent 40S subunits (Lorsch and Herschlag, 1999). The
comparison between the time courses of luciferase synthesis in RPS5 wt and rps5-T38A
mutant translation active extracts showed that the rps5-T38A mutant ribosomes need almost one minute longer to pass the early, undefined phase (Fig. 2.14 B). However, the
rps5-T38A mutant ribosomes have the same in vitro scanning rate as RPS5 wild-type ribosomes of approximately 9 nucleotides per second (Fig. 2.14 B). Thus, the rps5-T38A
mutation does not confer a scanning processivity defect.
Taken together, the rps5-T38A mutation impairs context dependent start codon recognition and increases initiation from non-AUG codons in vivo.