Gene recoding by synonymous mutations creates promiscuous intragenic transcription initiation in mycobacteria.

IMPORTANCE: Mycobacterium tuberculosis (Mtb) is the causative agent of tuberculosis, one of the deadliest infectious diseases worldwide. Previous studies have established that synonymous recoding to introduce rare codon pairings can attenuate viral pathogens. We hypothesized that non-optimal codon pairing could be an effective strategy for attenuating gene expression to create a live vaccine for Mtb. We instead discovered that these synonymous changes enabled the transcription of functional mRNA that initiated in the middle of the open reading frame and from which many smaller protein products were expressed. To our knowledge, this is one of the first reports that synonymous recoding of a gene in any organism can create or induce intragenic transcription start sites.

Gene recoding by synonymous mutations creates promiscuous intragenic transcription initiation in mycobacteria.

Each genome encodes some codons more frequently than their synonyms (codon usage bias), but codons are also arranged more frequently into specific pairs (codon pair bias). Recoding viral genomes and yeast or bacterial genes with non-optimal codon pairs has been shown to decrease gene expression. Gene expression is thus importantly regulated not only by the use of particular codons but by their proper juxtaposition. We therefore hypothesized that non-optimal codon pairing could likewise attenuate Mtb genes. We explored the role of codon pair bias by recoding Mtb genes ( rpoB, mmpL3, ndh ) and assessing their expression in the closely related and tractable model organism M. smegmatis . To our surprise, recoding caused the expression of multiple smaller protein isoforms from all three genes. We confirmed that these smaller proteins were not due to protein degradation, but instead issued from new transcription initiation sites positioned within the open reading frame. New transcripts gave rise to intragenic translation initiation sites, which in turn led to the expression of smaller proteins. We next identified the nucleotide changes associated with these new sites of transcription and translation. Our results demonstrated that apparently benign, synonymous changes can drastically alter gene expression in mycobacteria. More generally, our work expands our understanding of the codon-level parameters that control translation and transcription initiation. IMPORTANCE: Mycobacterium tuberculosis ( Mtb ) is the causative agent of tuberculosis, one of the deadliest infectious diseases worldwide. Previous studies have established that synonymous recoding to introduce rare codon pairings can attenuate viral pathogens. We hypothesized that non-optimal codon pairing could be an effective strategy for attenuating gene expression to create a live vaccine for Mtb . We instead discovered that these synonymous changes enabled the transcription of functional mRNA that initiated in the middle of the open reading frame and from which many smaller protein products were expressed. To our knowledge, this is the first report that synonymous recoding of a gene in any organism can create or induce intragenic transcription start sites.

Scaling gene expression for cell size control and senescence in Saccharomyces cerevisiae.

Cells divide with appropriate frequency by coupling division to growth-that is, cells divide only when they have grown sufficiently large. This process is poorly understood, but has been studied using cell size mutants. In principle, mutations affecting cell size could affect the mean size ("set-point" mutants), or they could affect the variability of sizes ("homeostasis" mutants). In practice, almost all known size mutants affect set-point, with little effect on size homeostasis. One model for size-dependent division depends on a size-dependent gene expression program: Activators of cell division are over-expressed at larger and larger sizes, while inhibitors are under-expressed. At sufficiently large size, activators overcome inhibitors, and the cell divides. Amounts of activators and inhibitors determine the set-point, but the gene expression program (the rate at which expression changes with cell size) determines the breadth of the size distribution (homeostasis). In this model, set-point mutants identify cell cycle activators and inhibitors, while homeostasis mutants identify regulators that couple expression of activators and inhibitors to size. We consider recent results suggesting that increased cell size causes senescence, and suggest that at very large sizes, an excess of DNA binding proteins leads to size induced senescence.

Differential Scaling of Gene Expression with Cell Size May Explain Size Control in Budding Yeast.

Yeast cells must grow to a critical size before committing to division. It is unknown how size is measured. We find that as cells grow, mRNAs for some cell-cycle activators scale faster than size, increasing in concentration, while mRNAs for some inhibitors scale slower than size, decreasing in concentration. Size-scaled gene expression could cause an increasing ratio of activators to inhibitors with size, triggering cell-cycle entry. Consistent with this, expression of the CLN2 activator from the promoter of the WHI5 inhibitor, or vice versa, interfered with cell size homeostasis, yielding a broader distribution of cell sizes. We suggest that size homeostasis comes from differential scaling of gene expression with size. Differential regulation of gene expression as a function of cell size could affect many cellular processes.

Extensive genomic recoding by codon-pair deoptimization selective for mammals is a flexible tool to generate attenuated vaccine candidates for dengue virus 2.

The four serotypes of dengue virus (DENV) are the leading etiologic agent of disease caused by arthropod-borne viruses (arboviruses) in the world, with billions at risk of DENV infection spread by infected mosquitoes. DENV causes illness ranging from dengue fever (DF) to life-threatening dengue hemorrhagic fever (DHF) and dengue shock syndrome (DSS). DENV proliferates well in two different host systems, an invertebrate mosquito vector and vertebrate primate host, which have a distinct difference in their preference of codon pairs (CP) for translation (different "codon pair bias"). Consequently, arboviruses must delicately balance the use of codon pairs between mammals and arthropods, which presents an Achilles' heel that we have exploited by specifically shifting the codon pair preference in the E and NS3 ORFs away from mammals while keeping the CPB favorable for mosquito ORFs. Here we report that recoding of the ORFs has led to variants that were over-attenuated in rhesus macaques although induction of protective antibodies in animals vaccinated with the smallest recoded ORF (E) was observed. The flexibility of our synthetic vaccine design (by decreasing the number of unfavorable CPs in the E ORF), allowed us to construct two new vaccine candidates (EhminA and EhminB) with intermediate attenuation in cell culture and neonatal mice, a result demonstrating proof of concept. New DENV vaccine candidates are being developed based on selective attenuation by dramatic recoding, with flexibility in balancing the attenuation and immunogenicity by marrying rational design and empirical modification.

mmi1 and rep2 mRNAs are novel RNA targets of the Mei2 RNA-binding protein during early meiosis in Schizosaccharomyces pombe.

The RNA-binding protein Mei2 is crucial for meiosis in Schizosaccharomyces pombe. In mei2 mutants, pre-meiotic S-phase is blocked, along with meiosis. Mei2 binds a long non-coding RNA (lncRNA) called meiRNA, which is a 'sponge RNA' for the meiotic inhibitor protein Mmi1. The interaction between Mei2, meiRNA and Mmi1 protein is essential for meiosis. But mei2 mutants have stronger and different phenotypes than meiRNA mutants, since mei2Δ arrests before pre-meiotic S, while the meiRNA mutant arrests after pre-meiotic S but before meiosis. This suggests Mei2 may bind additional RNAs. To identify novel RNA targets of Mei2, which might explain how Mei2 regulates pre-meiotic S, we used RNA immunoprecipitation and cross-linking immunoprecipitation. In addition to meiRNA, we found the mRNAs for mmi1 (which encodes Mmi1) and for the S-phase transcription factor rep2 There were also three other RNAs of uncertain relevance. We suggest that at meiotic initiation, Mei2 may sequester rep2 mRNA to help allow pre-meiotic S, and then may bind both meiRNA and mmi1 mRNA to inactivate Mmi1 at two levels, the protein level (as previously known), and also the mRNA level, allowing meiosis. We call Mei2-meiRNA a 'double sponge' (i.e. binding both an mRNA and its encoded protein).

Extensive recoding of dengue virus type 2 specifically reduces replication in primate cells without gain-of-function in Aedes aegypti mosquitoes.

Dengue virus (DENV), an arthropod-borne ("arbovirus") virus, causes a range of human maladies ranging from self-limiting dengue fever to the life-threatening dengue shock syndrome and proliferates well in two different taxa of the Animal Kingdom, mosquitoes and primates. Mosquitoes and primates show taxonomic group-specific intolerance to certain codon pairs when expressing their genes by translation. This is called "codon pair bias". By necessity, dengue viruses evolved to delicately balance this fundamental difference in their open reading frames (ORFs). We have undone the evolutionarily conserved genomic balance in the DENV2 ORF sequence and specifically shifted the encoding preference away from primates. However, this recoding of DENV2 raised concerns of 'gain-of-function,' namely whether recoding could inadvertently increase fitness for replication in the arthropod vector. Using mosquito cell lines and two strains of Aedes aegypti we did not observe any increase in fitness in DENV2 variants codon pair deoptimized for humans. This ability to disrupt and control DENV2's host preference has great promise towards developing the next generation of synthetic vaccines not only for DENV but for other emerging arboviral pathogens such as chikungunya virus and Zika virus.

Re-annotation of 12,495 prokaryotic 16S rRNA 3' ends and analysis of Shine-Dalgarno and anti-Shine-Dalgarno sequences.

We examined 20,648 prokaryotic unique taxids with respect to the annotation of the 3' end of the 16S rRNA, which contains the anti-Shine-Dalgarno sequence. We used the sequence of highly conserved helix 45 of the 16S rRNA as a guide. By this criterion, 8,153 annotated 3' ends correctly included the anti-Shine-Dalgarno sequence, but 12,495 were foreshortened or otherwise mis-annotated, missing part or all of the anti-Shine-Dalgarno sequence, which immediately follows helix 45. We re-annotated, giving a total of 20,648 16S rRNA 3' ends. The vast majority indeed contained a consensus anti-Shine-Dalgarno sequence, embedded in a highly conserved 13 base "tail". However, 128 exceptional organisms had either a variant anti-Shine-Dalgarno, or no recognizable anti-Shine-Dalgarno, in their 16S rRNA(s). For organisms both with and without an anti-Shine-Dalgarno, we identified the Shine-Dalgarno motifs actually enriched in front of each organism's open reading frames. This showed to what extent the Shine-Dalgarno motifs correlated with anti-Shine Dalgarno motifs. In general, organisms whose rRNAs lacked a perfect anti-Shine-Dalgarno motif also lacked a recognizable Shine-Dalgarno. For organisms whose 16S rRNAs contained a perfect anti-Shine-Dalgarno motif, a variety of results were obtained. We found one genus, Alteromonas, where several taxids apparently maintain two different types of 16S rRNA genes, with different, but conserved, antiSDs. The fact that some organisms do not seem to have or use Shine-Dalgarno motifs supports the idea that prokaryotes have other robust mechanisms for recognizing start codons for translation.

Prokaryotic coding regions have little if any specific depletion of Shine-Dalgarno motifs.

The Shine-Dalgarno motif occurs in front of prokaryotic start codons, and is complementary to the 3' end of the 16S ribosomal RNA. Hybridization between the Shine-Dalgarno sequence and the anti-Shine-Dalgarno region of the16S rRNA (CCUCCU) directs the ribosome to the start AUG of the mRNA for translation. Shine-Dalgarno-like motifs (AGGAGG in E. coli) are depleted from open reading frames of most prokaryotes. This may be because hybridization of the 16S rRNA at Shine-Dalgarnos inside genes would slow translation or induce internal initiation. However, we analyzed 128 species from diverse phyla where the 16S rRNA gene(s) lack the anti-Shine-Dalgarno sequence, and so the 16S rRNA is incapable of interacting with Shine-Dalgarno-like sequences. Despite this lack of an anti-Shine-Dalgarno, half of these species still displayed depletion of Shine-Dalgarno-like sequences when analyzed by previous methods. Depletion of the same G-rich sequences was seen by these methods even in eukaryotes, which do not use the Shine-Dalgarno mechanism. We suggest previous methods are partly detecting a non-specific depletion of G-rich sequences. Alternative informatics approaches show that most prokaryotes have only slight, if any, specific depletion of Shine-Dalgarno-like sequences from open reading frames. Together with recent evidence that ribosomes do not pause at ORF-internal Shine-Dalgarno motifs, these results suggest the presence of ORF-internal Shine-Dalgarno-like motifs may be inconsequential, perhaps because internal regions of prokaryotic mRNAs may be structurally "shielded" from translation initiation.

Assaying Glycogen and Trehalose in Yeast.

Organisms store carbohydrates in several forms. In yeast, carbohydrates are stored in glycogen (a multi-branched polysaccharide) and in trehalose (a disaccharide). As in other organisms, the amount of stored carbohydrate varies dramatically with physiological state, and accordingly, an assay of stored carbohydrate can help reveal physiological state. Here, we describe relatively easy and streamlined assays for glycogen and trehalose in yeast that can be applied either to a few samples, or in a moderately high-throughput fashion (dozens to hundreds of samples).

Relative contributions of the structural and catalytic roles of Rrp6 in exosomal degradation of individual mRNAs.

The RNA exosome is a conserved complex for RNA degradation with two ribonucleolytic subunits, Dis3 and Rrp6. Rrp6 is a 3'-5' exonuclease, but it also has a structural role in helping target RNAs to the Dis3 activity. The relative importance of the exonuclease activity and the targeting activity probably differs between different RNA substrates, but this is poorly understood. To understand the relative contributions of the exonuclease and the targeting activities to the degradation of individual RNA substrates in Schizosaccharomyces pombe, we compared RNA levels in an rrp6 null mutant to those in an rrp6 point mutant specifically defective in exonuclease activity. A wide range of effects was found, with some RNAs dependent mainly on the structural role of Rrp6 ("protein-dependent" targets), other RNAs dependent mainly on the catalytic role ("activity-dependent" targets), and some RNAs dependent on both. Some protein-dependent RNAs contained motifs targeted via the RNA-binding protein Mmi1, while others contained a motif possibly involved in response to iron. In these and other cases Rrp6 may act as a structural adapter to target specific RNAs to the exosome by interacting with sequence-specific RNA-binding proteins.

The non-homologous end-joining pathway of S. cerevisiae works effectively in G1-phase cells, and religates cognate ends correctly and non-randomly.

DNA double-strand breaks (DSBs) are potentially lethal lesions repaired by two major pathways: homologous recombination (HR) and non-homologous end-joining (NHEJ). Homologous recombination preferentially reunites cognate broken ends. In contrast, non-homologous end-joining could ligate together any two ends, possibly generating dicentric or acentric fragments, leading to inviability. Here, we characterize the yeast NHEJ pathway in populations of pure G1 phase cells, where there is no possibility of repair using a homolog. We show that in G1 yeast cells, NHEJ is a highly effective repair pathway for gamma-ray induced breaks, even when many breaks are present. Pulsed-field gel analysis showed chromosome karyotypes following NHEJ repair of cells from populations with multiple breaks. The number of reciprocal translocations was surprisingly low, perhaps zero, suggesting that NHEJ preferentially re-ligates the "correct" broken ends instead of randomly-chosen ends. Although we do not know the mechanism, the preferential correct ligation is consistent with the idea that broken ends are continuously held together by protein-protein interactions or by larger scale chromatin structure.

Cyclin-Dependent Kinase Co-Ordinates Carbohydrate Metabolism and Cell Cycle in S. cerevisiae.

Cyclin-dependent kinases (CDKs) control cell division in eukaryotes by phosphorylating proteins involved in division. But successful proliferation requires co-ordination between division and cellular growth in mass. Previous proteomic studies suggested that metabolic proteins, as well as cell division proteins, could potentially be substrates of cyclin-dependent kinases. Here we focus on two metabolic enzymes of the yeast S. cerevisiae, neutral trehalase (Nth1) and glycogen phosphorylase (Gph1), and show that their activities are likely directly controlled by CDK activity, thus allowing co-ordinate regulation of carbohydrate metabolism with cell division processes. In this case, co-ordinate regulation may optimize the decision to undertake a final cell division as nutrients are being exhausted. Co-regulation of cell division processes and metabolic processes by CDK activity may be a general phenomenon important for co-ordinating the cell cycle with growth.

Large-scale recoding of an arbovirus genome to rebalance its insect versus mammalian preference.

The protein synthesis machineries of two distinct phyla of the Animal kingdom, insects of Arthropoda and mammals of Chordata, have different preferences for how to best encode proteins. Nevertheless, arboviruses (arthropod-borne viruses) are capable of infecting both mammals and insects just like arboviruses that use insect vectors to infect plants. These organisms have evolved carefully balanced genomes that can efficiently use the translational machineries of different phyla, even if the phyla belong to different kingdoms. Using dengue virus as an example, we have undone the genome encoding balance and specifically shifted the encoding preference away from mammals. These mammalian-attenuated viruses grow to high titers in insect cells but low titers in mammalian cells, have dramatically increased LD50s in newborn mice, and induce high levels of protective antibodies. Recoded arboviruses with a bias toward phylum-specific expression could form the basis of a new generation of live attenuated vaccine candidates.

A new transcription factor for mitosis: in Schizosaccharomyces pombe, the RFX transcription factor Sak1 works with forkhead factors to regulate mitotic expression.

Mitotic genes are one of the most strongly oscillating groups of genes in the eukaryotic cell cycle. Understanding the regulation of mitotic gene expression is a key issue in cell cycle control but is poorly understood in most organisms. Here, we find a new mitotic transcription factor, Sak1, in the fission yeast Schizosaccharomyces pombe. Sak1 belongs to the RFX family of transcription factors, which have not previously been connected to cell cycle control. Sak1 binds upstream of mitotic genes in close proximity to Fkh2, a forkhead transcription factor previously implicated in regulation of mitotic genes. We show that Sak1 is the major activator of mitotic gene expression and also confirm the role of Fkh2 as the opposing repressor. Sep1, another forkhead transcription factor, is an activator for a small subset of mitotic genes involved in septation. From yeasts to humans, forkhead transcription factors are involved in mitotic gene expression and it will be interesting to see whether RFX transcription factors may also be involved in other organisms.

Measurement of average decoding rates of the 61 sense codons in vivo.

Most amino acids can be encoded by several synonymous codons, which are used at unequal frequencies. The significance of unequal codon usage remains unclear. One hypothesis is that frequent codons are translated relatively rapidly. However, there is little direct, in vivo, evidence regarding codon-specific translation rates. In this study, we generate high-coverage data using ribosome profiling in yeast, analyze using a novel algorithm, and deduce events at the A- and P-sites of the ribosome. Different codons are decoded at different rates in the A-site. In general, frequent codons are decoded more quickly than rare codons, and AT-rich codons are decoded more quickly than GC-rich codons. At the P-site, proline is slow in forming peptide bonds. We also apply our algorithm to short footprints from a different conformation of the ribosome and find strong amino acid-specific (not codon-specific) effects that may reflect interactions with the exit tunnel of the ribosome.

A developmentally regulated translational control pathway establishes the meiotic chromosome segregation pattern.

Production of haploid gametes from diploid progenitor cells is mediated by a specialized cell division, meiosis, where two divisions, meiosis I and II, follow a single S phase. Errors in progression from meiosis I to meiosis II lead to aneuploid and polyploid gametes, but the regulatory mechanisms controlling this transition are poorly understood. Here, we demonstrate that the conserved kinase Ime2 regulates the timing and order of the meiotic divisions by controlling translation. Ime2 coordinates translational activation of a cluster of genes at the meiosis I-meiosis II transition, including the critical determinant of the meiotic chromosome segregation pattern CLB3. We further show that Ime2 mediates translational control through the meiosis-specific RNA-binding protein Rim4. Rim4 inhibits translation of CLB3 during meiosis I by interacting with the 5' untranslated region (UTR) of CLB3. At the onset of meiosis II, Ime2 kinase activity rises and triggers a decrease in Rim4 protein levels, thereby alleviating translational repression. Our results elucidate a novel developmentally regulated translational control pathway that establishes the meiotic chromosome segregation pattern.

Deliberate reduction of hemagglutinin and neuraminidase expression of influenza virus leads to an ultraprotective live vaccine in mice.

A long-held dogma posits that strong presentation to the immune system of the dominant influenza virus glycoprotein antigens neuraminidase (NA) and hemagglutinin (HA) is paramount for inducing protective immunity against influenza virus infection. We have deliberately violated this dogma by constructing a recombinant influenza virus strain of A/PR8/34 (H1N1) in which expression of NA and HA genes was suppressed. We down-regulated NA and HA expression by recoding the respective genes with suboptimal codon pair bias, thereby introducing hundreds of nucleotide changes while preserving their codon use and protein sequence. The variants PR8-NA(Min), PR8-HA(Min), and PR8-(NA+HA)(Min) (Min, minimal expression) were used to assess the contribution of reduced glycoprotein expression to growth in tissue culture and pathogenesis in BALB/c mice. All three variants proliferated in Madin-Darby canine kidney cells to nearly the degree as WT PR8. In mice, however, they expressed explicit attenuation phenotypes, as revealed by their LD50 values: PR8, 32 plaque-forming units (PFU); HA(Min), 1.7 × 10(3) PFU; NA(Min), 2.4 × 10(5) PFU; (NA+HA)(Min), ≥3.16 × 10(6) PFU. Remarkably, (NA+HA)(Min) was attenuated >100,000-fold, with NA(Min) the major contributor to attenuation. In vaccinated mice (NA+HA)(Min) was highly effective in providing long-lasting protective immunity against lethal WT challenge at a median protective dose (PD50) of 2.4 PFU. Moreover, at a PD50 of only 147 or 237, (NA+HA)(Min) conferred protection against heterologous lethal challenges with two mouse-adapted H3N2 viruses. We conclude that the suppression of HA and NA is a unique strategy in live vaccine development.

Effects of the yeast RNA-binding protein Whi3 on the half-life and abundance of CLN3 mRNA and other targets.

Whi3 is an RNA binding protein known to bind the mRNA of the yeast G1 cyclin gene CLN3. It inhibits CLN3 function, but the mechanism of this inhibition is unclear; in previous studies, Whi3 made no observable difference to CLN3 mRNA levels, translation, or protein abundance. Here, we re-approach this issue using microarrays, RNA-Seq, ribosome profiling, and other methods. By multiple methods, we find that the whi3 mutation causes a small but consistent increase in the abundance of hundreds of mRNAs, including the CLN3 mRNA. The effect on various mRNAs is roughly in proportion to the density of GCAU or UGCAU motifs carried by these mRNAs, which may be a binding site for Whi3. mRNA instability of Whi3 targets may in part depend on a 3' AU rich element (ARE), AUUUUA. In addition, the whi3 mutation causes a small increase in the translational efficiency of CLN3 mRNA. The increase in CLN3 mRNA half-life and abundance together with the increase in translational efficiency is fully sufficient to explain the small-cell phenotype of whi3 mutants. Under stress conditions, Whi3 becomes a component of P-bodies or stress granules, but Whi3 also acts under non-stress condition, when no P-bodies are visible. We suggest that Whi3 may be a very broadly-acting, but mild, modulator of mRNA stability. In CLN3, Whi3 may bind to the 3' GCAU motifs to attract the Ccr4-Not complex to promote RNA deadenylation and turnover, and Whi3 may bind to the 5' GCAU motifs to inhibit translation.