Figure 3 - uploaded by Yamila N. Torres Cleuren
Content may be subject to copyright.
uORFs reduce the number of 43S PICs scanning across 5' UTRs. A-C) The impact of the number of uORFs on (A) scanning subunits on 5' UTR, (B) the translational efficiency of the 5' UTR, and (C) the translational efficiency of the protein (*** = p-values < 0.001). D) Coverage of small subunit (40S) footprints (upper, blue) and ribo-seq 80S complex footprints (lower, orange) in fixed windows of 100 nt up-and downstream of the first ATG uORF. E) Heatmaps showing the rate of scanning subunit consumption as measured by the ratio of small subunit reads upstream versus downstream of all uORFs stratified by surrounding Kozak score and start codon. F) Same as E, but with ranking of start and stop codon.
Source publication
Translation initiation is often attributed as the rate determining step of eukaryotic protein synthesis and key to gene expression control. Despite this centrality the series of steps involved in this process are poorly understood. Here we capture the transcriptome-wide occupancy of ribosomes across all stages of translation initiation, enabling us...
Contexts in source publication
Context 1
... mRNAs have a strong enrichment of small subunit footprints coinciding with the 5' end of transcripts (Fig. S2). The 5' peak is not present in non-coding RNAs arguing that it is a feature only of translated RNA molecules and not an artifact of the method (Fig. S3). The start of these footprints all coincide with the transcription start site, and have a wide range of read lengths from the lower detection limit (~15 nt) up to about 80nt which is slightly longer than scanning 43S PICs (Fig. 2B and Fig. S2, ...
Context 2
... very short 5' UTRs only threading is expected to be able to initiate translation as slotting would deposit the small subunit too far downstream to scan the start codon 2,14 . Consistently, we observed a strong 5' peak reduction in response to eIF4E inhibition in transcripts initiated through the Translation Initiator of Short 5' UTR (TISU) motif (Fig. S3E- F), in line with previous reports that these transcripts are eIF4E-sensitive 14 . Collectively, this suggests threading is dependent on eIF4E and is a common recruitment pathway during early ...
Context 3
... transcripts containing uORFs, the CDS is translated either from ribosomes that fail to recognise the often sub-optimal uORF TIS 7,21,22,23 , or by reinitiating ribosomes that continue scanning after translating the uORF 24,25 . Therefore, uORFs typically lead to reduced protein synthesis by consuming scanning 43S PICs (Fig. 3A-C) 26 . Consistent with our global estimates, we find a local decline of 43S PIC footprints coinciding . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in ...
Context 4
... copyright holder for this preprint . http://dx.doi.org/10.1101/811810 doi: bioRxiv preprint first posted online Oct. 21, 2019; with an increase in 80S footprints at uORF TIS' (Fig. 3D). The ratio of the 43S PIC density upstream vs downstream of a uORF TIS can therefore quantify to what extent uORFs consume scanning 43S PICs (Fig. 3D). As expected, uORFs starting with an ATG start codon ( Fig. 3E-F) and with a TIS context similar to the Kozak sequence (Fig. 3E) have the highest 43S PIC ...
Context 5
... copyright holder for this preprint . http://dx.doi.org/10.1101/811810 doi: bioRxiv preprint first posted online Oct. 21, 2019; with an increase in 80S footprints at uORF TIS' (Fig. 3D). The ratio of the 43S PIC density upstream vs downstream of a uORF TIS can therefore quantify to what extent uORFs consume scanning 43S PICs (Fig. 3D). As expected, uORFs starting with an ATG start codon ( Fig. 3E-F) and with a TIS context similar to the Kozak sequence (Fig. 3E) have the highest 43S PIC ...
Context 6
... doi: bioRxiv preprint first posted online Oct. 21, 2019; with an increase in 80S footprints at uORF TIS' (Fig. 3D). The ratio of the 43S PIC density upstream vs downstream of a uORF TIS can therefore quantify to what extent uORFs consume scanning 43S PICs (Fig. 3D). As expected, uORFs starting with an ATG start codon ( Fig. 3E-F) and with a TIS context similar to the Kozak sequence (Fig. 3E) have the highest 43S PIC ...
Context 7
... Oct. 21, 2019; with an increase in 80S footprints at uORF TIS' (Fig. 3D). The ratio of the 43S PIC density upstream vs downstream of a uORF TIS can therefore quantify to what extent uORFs consume scanning 43S PICs (Fig. 3D). As expected, uORFs starting with an ATG start codon ( Fig. 3E-F) and with a TIS context similar to the Kozak sequence (Fig. 3E) have the highest 43S PIC ...
Context 8
... we found the ability of the small subunit to resume scanning after uORF translation to be highly dependent on the choice of stop codon. For proteins, TAA and TGA have been reported as the most and least efficient termination codons, respectively 27 . Consistently, uORFs with TGA have the greatest reduction of downstream scanning small subunits ( Fig. 3F), the highest density of downstream 80S footprints ( Fig. S12A) and the lowest ratio of small subunit to 80S complexes directly over their stop codons (Fig. S12B). This less efficient stop codon recognition globally leads to a small, but significant effect on the TE of the downstream CDS (Fig. S12C). This suggests that failure to ...
Context 9
... to the known zebrafish Kozak sequence 34 (Fig. 4A). This model, however, considers positions independently and therefore only reflects an average over sequences with high IR and not the efficiency of any particular sequence. To obtain this, we grouped all genes with identical sequence context and ranked these sequences by their median IR (Fig. 4B, S13). The resulting ranking was consistent with a previous assessment of a small number of sequences in zebrafish 34 , but surprisingly revealed that the Kozak sequence is not the optimal context. The highest scoring sequence was AGGCATG which differs by two bases (G at -3 and -2). More surprisingly, several sequences that differ strongly ...
Context 10
... which makes it unlikely that the abundant 5' reads (suggesting a frequent occurrence) are due to back-sliding. The second possibility is that these reads are simply 3'-to-5' degradation intermediates. However, two observations argue against this possibility: first, non-coding RNAs have very few 5' reads arguing for a translation-dependent origin (Fig. S3G), and second, sequencing reads from degradation intermediates (and other possible artifacts) would be expected to increase when ribosome scanning is inhibited. Instead, upon eIF4E inhibition we observe that these short 5' fragments disappear together with fragments derived from SSU scanning (Fig. 2C). We therefore conclude that ...
Context 11
... plots (Fig. 3) . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in ...
Context 12
... mRNAs have a strong enrichment of small subunit footprints coinciding with the 5' end of transcripts (Fig. S2). The 5' peak is not present in non-coding RNAs arguing that it is a feature only of translated RNA molecules and not an artifact of the method (Fig. S3). The start of these footprints all coincide with the transcription start site, and have a wide range of read lengths from the lower detection limit (~15 nt) up to about 80nt which is slightly longer than scanning 43S PICs (Fig. 2B and Fig. S2, ...
Context 13
... very short 5' UTRs only threading is expected to be able to initiate translation as slotting would deposit the small subunit too far downstream to scan the start codon 2,14 . Consistently, we observed a strong 5' peak reduction in response to eIF4E inhibition in transcripts initiated through the Translation Initiator of Short 5' UTR (TISU) motif (Fig. S3E- F), in line with previous reports that these transcripts are eIF4E-sensitive 14 . Collectively, this suggests threading is dependent on eIF4E and is a common recruitment pathway during early ...
Context 14
... transcripts containing uORFs, the CDS is translated either from ribosomes that fail to recognise the often sub-optimal uORF TIS 7,21,22,23 , or by reinitiating ribosomes that continue scanning after translating the uORF 24,25 . Therefore, uORFs typically lead to reduced protein synthesis by consuming scanning 43S PICs (Fig. 3A-C) 26 . Consistent with our global estimates, we find a local decline of 43S PIC footprints coinciding . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in ...
Context 15
... copyright holder for this preprint . http://dx.doi.org/10.1101/811810 doi: bioRxiv preprint first posted online Oct. 21, 2019; with an increase in 80S footprints at uORF TIS' (Fig. 3D). The ratio of the 43S PIC density upstream vs downstream of a uORF TIS can therefore quantify to what extent uORFs consume scanning 43S PICs (Fig. 3D). As expected, uORFs starting with an ATG start codon ( Fig. 3E-F) and with a TIS context similar to the Kozak sequence (Fig. 3E) have the highest 43S PIC ...
Context 16
... copyright holder for this preprint . http://dx.doi.org/10.1101/811810 doi: bioRxiv preprint first posted online Oct. 21, 2019; with an increase in 80S footprints at uORF TIS' (Fig. 3D). The ratio of the 43S PIC density upstream vs downstream of a uORF TIS can therefore quantify to what extent uORFs consume scanning 43S PICs (Fig. 3D). As expected, uORFs starting with an ATG start codon ( Fig. 3E-F) and with a TIS context similar to the Kozak sequence (Fig. 3E) have the highest 43S PIC ...
Context 17
... doi: bioRxiv preprint first posted online Oct. 21, 2019; with an increase in 80S footprints at uORF TIS' (Fig. 3D). The ratio of the 43S PIC density upstream vs downstream of a uORF TIS can therefore quantify to what extent uORFs consume scanning 43S PICs (Fig. 3D). As expected, uORFs starting with an ATG start codon ( Fig. 3E-F) and with a TIS context similar to the Kozak sequence (Fig. 3E) have the highest 43S PIC ...
Context 18
... Oct. 21, 2019; with an increase in 80S footprints at uORF TIS' (Fig. 3D). The ratio of the 43S PIC density upstream vs downstream of a uORF TIS can therefore quantify to what extent uORFs consume scanning 43S PICs (Fig. 3D). As expected, uORFs starting with an ATG start codon ( Fig. 3E-F) and with a TIS context similar to the Kozak sequence (Fig. 3E) have the highest 43S PIC ...
Context 19
... we found the ability of the small subunit to resume scanning after uORF translation to be highly dependent on the choice of stop codon. For proteins, TAA and TGA have been reported as the most and least efficient termination codons, respectively 27 . Consistently, uORFs with TGA have the greatest reduction of downstream scanning small subunits ( Fig. 3F), the highest density of downstream 80S footprints ( Fig. S12A) and the lowest ratio of small subunit to 80S complexes directly over their stop codons (Fig. S12B). This less efficient stop codon recognition globally leads to a small, but significant effect on the TE of the downstream CDS (Fig. S12C). This suggests that failure to ...
Context 20
... to the known zebrafish Kozak sequence 34 (Fig. 4A). This model, however, considers positions independently and therefore only reflects an average over sequences with high IR and not the efficiency of any particular sequence. To obtain this, we grouped all genes with identical sequence context and ranked these sequences by their median IR (Fig. 4B, S13). The resulting ranking was consistent with a previous assessment of a small number of sequences in zebrafish 34 , but surprisingly revealed that the Kozak sequence is not the optimal context. The highest scoring sequence was AGGCATG which differs by two bases (G at -3 and -2). More surprisingly, several sequences that differ strongly ...
Context 21
... which makes it unlikely that the abundant 5' reads (suggesting a frequent occurrence) are due to back-sliding. The second possibility is that these reads are simply 3'-to-5' degradation intermediates. However, two observations argue against this possibility: first, non-coding RNAs have very few 5' reads arguing for a translation-dependent origin (Fig. S3G), and second, sequencing reads from degradation intermediates (and other possible artifacts) would be expected to increase when ribosome scanning is inhibited. Instead, upon eIF4E inhibition we observe that these short 5' fragments disappear together with fragments derived from SSU scanning (Fig. 2C). We therefore conclude that ...
Similar publications
It is generally accepted that the presence of ORFs in the 5′ untranslated region of eukaryotic transcripts modulates the production of proteins by controlling the translation initiation rate of the main CDS. In trypanosomatid parasites, which almost exclusively depend on post-transcriptional mechanisms to regulate gene expression, translation has b...
Citations
... uORFs can regulate gene expression via the biological activity of the uORF peptide, but they also often cis-regulate translation of the downstream main ORF 6,7 . Despite having poor initiation sequence contexts, many eukaryotic uORFs repress main ORF translation 1,3,4,[7][8][9][10][11] . uORF mutations are implicated in several human diseases via changes to main ORF translation 12,13 . ...
... Longer uORFs offer more time for elongating ribosomes to catch up, hit, and knock off 3′ scanning ribosomes. Nevertheless, most eukaryotic uORFs only weakly initiate translation and are short 1,3,4,[8][9][10][11]31 . UL4 uORF2 is 22 codons long, and we estimate reinitiation to be frequent (Table 1). ...
... Long, well-initiating uORFs that do not re-initiate well allow buffering (Fig. 3B, left panel, yellow-green line, Fig. S2A, yellow-green line) in the 80S-hit model (Fig. 1B), but these requirements are at odds with the typically short and poorly initiating nature of known uORFs 1,3,4,[8][9][10][11] . Consequently, when we use parameters speci c to UL4 uORF2 for the 80S-hit model ( Table 1), namely that uORF2 initiates poorly, re-initiates well, and is not very long, buffering is no longer predicted (Fig. S2B). ...
Upstream open reading frames (uORFs) are present in over half of all human mRNAs. uORFs can potently regulate the translation of downstream open reading frames by several mechanisms: siphoning away scanning ribosomes, regulating re-initiation, and allowing interactions between scanning and elongating ribosomes. However, the consequences of these different mechanisms for the regulation of protein expression remain incompletely understood. Here, we performed systematic measurements on the uORF-containing 5′ UTR of the cytomegaloviral UL4 mRNA to test alternative models of uORF-mediated regulation in human cells. We find that a terminal diproline-dependent elongating ribosome stall in the UL4 uORF prevents decreases in main ORF translation when ribosome loading onto the mRNA is reduced. This uORF-mediated buffering is insensitive to the location of the ribosome stall along the uORF. Computational kinetic modeling based on our measurements suggests that scanning ribosomes dissociate rather than queue when they collide with stalled elongating ribosomes within the UL4 uORF. We identify several human uORFs that repress main ORF translation via a similar terminal diproline motif. We propose that ribosome stalls in uORFs provide a general mechanism for buffering against reductions in main ORF translation during stress and developmental transitions.
