Evidence for widespread translation of 5' untranslated regions.

Ribosome profiling experiments support the translation of a range of novel human open reading frames. By contrast, most peptides from large-scale proteomics experiments derive from just one source, 5' untranslated regions. Across the human genome we find evidence for 192 translated upstream regions, most of which would produce protein isoforms with extended N-terminal ends. Almost all of these N-terminal extensions are from highly abundant genes, which suggests that the novel regions we detect are just the tip of the iceberg. These upstream regions have characteristics that are not typical of coding exons. Their GC-content is remarkably high, even higher than 5' regions in other genes, and a large majority have non-canonical start codons. Although some novel upstream regions have cross-species conservation - five have orthologues in invertebrates for example - the reading frames of two thirds are not conserved beyond simians. These non-conserved regions also have no evidence of purifying selection, which suggests that much of this translation is not functional. In addition, non-conserved upstream regions have significantly more peptides in cancer cell lines than would be expected, a strong indication that an aberrant or noisy translation initiation process may play an important role in translation from upstream regions.

Start codon variant in LAG3 is associated with decreased LAG-3 expression and increased risk of autoimmune thyroid disease.

Autoimmune thyroid disease (AITD) is a common autoimmune disease. In a GWAS meta-analysis of 110,945 cases and 1,084,290 controls, 290 sequence variants at 225 loci are associated with AITD. Of these variants, 115 are previously unreported. Multiomics analysis yields 235 candidate genes outside the MHC-region and the findings highlight the importance of genes involved in T-cell regulation. A rare 5'-UTR variant (rs781745126-T, MAF = 0.13% in Iceland) in LAG3 has the largest effect (OR = 3.42, P = 2.2 × 10-16) and generates a novel start codon for an open reading frame upstream of the canonical protein translation initiation site. rs781745126-T reduces mRNA and surface expression of the inhibitory immune checkpoint LAG-3 co-receptor on activated lymphocyte subsets and halves LAG-3 levels in plasma among heterozygotes. All three homozygous carriers of rs781745126-T have AITD, of whom one also has two other T-cell mediated diseases, that is vitiligo and type 1 diabetes. rs781745126-T associates nominally with vitiligo (OR = 5.1, P = 6.5 × 10-3) but not with type 1 diabetes. Thus, the effect of rs781745126-T is akin to drugs that inhibit LAG-3, which unleash immune responses and can have thyroid dysfunction and vitiligo as adverse events. This illustrates how a multiomics approach can reveal potential drug targets and safety concerns.

Upstream open reading frame-introducing variants in patients with primary familial brain calcification.

More than 50% of patients with primary familial brain calcification (PFBC), a rare neurological disorder, remain genetically unexplained. While some causative genes are yet to be identified, variants in non-coding regions of known genes may represent a source of missed diagnoses. We hypothesized that 5'-Untranslated Region (UTR) variants introducing an AUG codon may initiate mRNA translation and result in a loss of function in some of the PFBC genes. After reannotation of exome sequencing data of 113 unrelated PFBC probands, we identified two upstream AUG-introducing variants in the 5'UTR of PDGFB. One, NM_002608.4:c.-373C>G, segregated with PFBC in the family. It was predicted to create an upstream open reading frame (ORF). The other one, NM_002608.4:c.-318C>T, was found in a simplex case. It was predicted to result in an ORF overlapping the natural ORF with a frameshift. In a GFP reporter assay, both variants were associated with a dramatic decrease in GFP levels, and, after restoring the reading frame with the GFP sequence, the c.-318C>T variant was associated with a strong initiation of translation as measured by western blotting. Overall, we found upstream AUG-introducing variants in the 5'UTR of PDGFB in 2/113 (1.7%) undiagnosed PFBC cases. Such variants thus represent a source of putative pathogenic variants.

JUN mRNA translation regulation is mediated by multiple 5' UTR and start codon features.

Regulation of mRNA translation by eukaryotic initiation factors (eIFs) is crucial for cell survival. In humans, eIF3 stimulates translation of the JUN mRNA which encodes the transcription factor JUN, an oncogenic transcription factor involved in cell cycle progression, apoptosis, and cell proliferation. Previous studies revealed that eIF3 activates translation of the JUN mRNA by interacting with a stem loop in the 5' untranslated region (5' UTR) and with the 5' -7-methylguanosine cap structure. In addition to its interaction site with eIF3, the JUN 5' UTR is nearly one kilobase in length, and has a high degree of secondary structure, high GC content, and an upstream start codon (uAUG). This motivated us to explore the complexity of JUN mRNA translation regulation in human cells. Here we find that JUN translation is regulated in a sequence and structure-dependent manner in regions adjacent to the eIF3-interacting site in the JUN 5' UTR. Furthermore, we identify contributions of an additional initiation factor, eIF4A, in JUN regulation. We show that enhancing the interaction of eIF4A with JUN by using the compound Rocaglamide A (RocA) represses JUN translation. We also find that both the upstream AUG (uAUG) and the main AUG (mAUG) contribute to JUN translation and that they are conserved throughout vertebrates. Our results reveal additional layers of regulation for JUN translation and show the potential of JUN as a model transcript for understanding multiple interacting modes of translation regulation.

N6-methyladenosine in 5' UTR does not promote translation initiation.

The most abundant N6-methyladenosine (m6A) modification on mRNAs is installed non-stoichiometrically across transcripts, with 5' untranslated regions (5' UTRs) being the least conductive. 5' UTRs are essential for translation initiation, yet the molecular mechanisms orchestrated by m6A remain poorly understood. Here, we combined structural, biochemical, and single-molecule approaches and show that at the most common position, a single m6A does not affect translation yields, the kinetics of translation initiation complex assembly, or start codon recognition both under permissive growth and following exposure to oxidative stress. Cryoelectron microscopy (cryo-EM) structures of the late preinitiation complex reveal that m6A purine ring established stacking interactions with an arginine side chain of the initiation factor eIF2α, although with only a marginal energy contribution, as estimated computationally. These findings provide molecular insights into m6A interactions with the initiation complex and suggest that the subtle stabilization is unlikely to affect the translation dynamics under homeostatic conditions or stress.

A seamless connection from the burst sequence to the start codon is essential for polyhedrin hyperexpression in alphabaculoviruses.

Alphabaculoviruses produce a large number of occlusion bodies (OBs) in host cells during the late stage of infection. OBs are mainly composed of polyhedrin (POLH), and high-level transcription of the polh gene has been exploited to express foreign proteins in insect cells. While making Bombyx mori nucleopolyhedrovirus (BmNPV) polh mutants using a conventional transfer vector-based method, we noticed that a virus with a short sequence insertion just before the polh start codon produces fewer very small OBs. Detailed analysis of several BmNPV mutants revealed that insertions between the burst sequence and start codon markedly decrease POLH accumulation and polh transcription. We further confirmed this decrease using recombinant viruses expressing a reporter gene driven by the polh promoter. These findings underscore the critical importance of a seamless connection from the burst sequence to the start codon for baculovirus polh hyperexpression.

N-terminal alanine-rich (NTAR) sequences drive precise start codon selection resulting in elevated translation of multiple proteins including ERK1/2.

We report the discovery of N-terminal alanine-rich sequences, which we term NTARs, that act in concert with their native 5'-untranslated regions to promote selection of the proper start codon. NTARs also facilitate efficient translation initiation while limiting the production of non-functional polypeptides through leaky scanning. We first identified NTARs in the ERK1/2 kinases, which are among the most important signaling molecules in mammals. Analysis of the human proteome reveals that hundreds of proteins possess NTARs, with housekeeping proteins showing a particularly high prevalence. Our data indicate that several of these NTARs act in a manner similar to those found in the ERKs and suggest a mechanism involving some or all of the following features: alanine richness, codon rarity, a repeated amino acid stretch and a nearby second AUG. These features may help slow down the leading ribosome, causing trailing pre-initiation complexes (PICs) to pause near the native AUG, thereby facilitating accurate translation initiation. Amplification of erk genes is frequently observed in cancer, and we show that NTAR-dependent ERK protein levels are a rate-limiting step for signal output. Thus, NTAR-mediated control of translation may reflect a cellular need to precisely control translation of key transcripts such as potential oncogenes. By preventing translation in alternative reading frames, NTAR sequences may be useful in synthetic biology applications, e.g. translation from RNA vaccines.

Initiation of translation is essential for protein synthesis. A crucial step is the correct choice of the start AUG, which leads to the production of the fully functional polypeptide. To date, nucleotide composition next to the AUG has been considered the only determinant of start codon selection. Our work identifies a large family of proteins whose start codon choice is determined by an N-terminal alanine-rich sequence (NTAR) that enables efficient protein translation. Many of these proteins are encoded by housekeeping genes. Among them, the NTARs of the pivotal kinases ERK1 and ERK2 are highly optimized in humans, shaping ERK signal transduction by increasing the kinase quantity. Our findings could be useful for applied biology, especially for mRNA-based therapeutics.

Analysis of a non-lethal biallelic frameshift mutation in ZMPSTE24 reveals utilization of alternative translation initiation codons.

Restrictive dermopathy (RD) is a lethal condition caused by biallelic loss-of-function mutations in ZMPSTE24, whereas mutations preserving residual enzymatic activity of the ZMPSTE24 protein lead to the milder mandibuloacral dysplasia with type B lipodystrophy (MADB) phenotype. Remarkably, we identified a homozygous, presumably loss-of-function mutation in ZMPSTE24 [c.28_29insA, p.(Leu10Tyrfs*37)] in two consanguineous Pakistani families segregating MADB. To clarify how lethal consequences are prevented in affected individuals, functional analysis was performed. Expression experiments supported utilization of two alternative translation initiation sites, preventing complete loss of protein function consistent with the relatively mild phenotypic outcome in affected patients. One of these alternative start codons is newly formed at the insertion site. Our findings indicate that the creation of new potential start codons through N-terminal mutations in other disease-associated genes should generally be taken into consideration in the variant interpretation process.

Chloroplasts evolved an additional layer of translational regulation based on non-AUG start codons for proteins with different turnover rates.

Chloroplasts have evolved from photosynthetic cyanobacteria-like progenitors through endosymbiosis. The chloroplasts of present-day land plants have their own transcription and translation systems that show several similarities with prokaryotic organisms. A remarkable feature of the chloroplast translation system is the use of non-AUG start codons in the protein synthesis of certain genes that are evolutionarily conserved from Algae to angiosperms. However, the biological significance of such use of non-AUG codons is not fully understood. The present study was undertaken to unravel the significance of non-AUG start codons in vivo using the chloroplast genetic engineering approach. For this purpose, stable transplastomic tobacco plants expressing a reporter gene i.e. uidA (GUS) under four different start codons (AUG/UUG/GUG/CUG) were generated and ß-glucuronidase (GUS) expression was compared. To investigate further the role of promoter sequences proximal to the start codon, uidA was expressed under two different chloroplast gene promoters psbA and psbC that use AUG and a non-AUG (GUG) start codons, respectively, and also showed significant differences in the DNA sequence surrounding the start codon. Further, to delineate the role of RNA editing that creates AUG start codon by editing non-AUG codons, if any, which is another important feature of the chloroplast transcription and translation system, transcripts were sequenced. In addition, a proteomic approach was used to identify the translation initiation site(s) of GUS and the N-terminal amino acid encoded when expressed under different non-AUG start codons. The results showed that chloroplasts use non-AUG start codons in combination with the translation initiation site as an additional layer of gene regulation to over-express proteins that are required at high levels due to their high rates of turnover.

Conserved Double Translation Initiation Site for Δ160p53 Protein Hints at Isoform's Key Role in Mammalian Physiology.

p53 is the most commonly mutated gene in human cancers. Two fundamental reasons for this are its long protein isoforms protect from cancer, while its shorter C-terminal isoforms can support cancer and metastasis. Previously, we have shown that the Δ160p53 protein isoform enhances survival and the invasive character of cancer cells. Here, we identified a translation initiation site nine codons downstream of codon 160-the known initiation codon for the translation of Δ160p53-that is recognized by the translation machinery. When translation failed to initiate from AUG160 due to mutation, it initiated from AUG169 instead, producing similar levels of a similar protein, Δ169p53, which promoted cell survival as efficiently as Δ160p53 following DNA damage. Interestingly, almost all mammalian species with an orthologue to human AUG160 also possess one for AUG169, while none of the non-mammalian species lacking AUG160 have AUG169, even if that region of the p53 gene is well conserved. In view of our findings, we do not believe that Δ169p53 acts as a different p53 protein isoform; instead, we propose that the double translation initiation site strengthens the translation of these products with a critical role in cell homeostasis. Future studies will help verify if this is a more general mechanism for the expression of essential proteins in mammals.

Kozak Similarity Score Algorithm Identifies Alternative Translation Initiation Codons Implicated in Cancers.

Ribosome profiling and mass spectroscopy have identified canonical and noncanonical translation initiation codons (TICs) that are upstream of the main translation initiation site and used to translate oncogenic proteins. There have previously been conflicting reports about the patterns of nucleotides that surround noncanonical TICs. Here, we use a Kozak Similarity Score algorithm to find that nearly all of these TICs have flanking nucleotides closely matching the Kozak sequence. Remarkably, the nucleotides flanking alternative noncanonical TICs are frequently closer to the Kozak sequence than the nucleotides flanking TICs used to translate the gene's main protein. Of note, the 5' untranslated region (5'UTR) of cancer-associated genes with an upstream TIC tend to be significantly longer than the same region in genes not associated with cancer. The presence of a longer-than-typical 5'UTR increases the likelihood of ribosome binding to upstream noncanonical TICs, and may be a distinguishing feature of a number of genes overexpressed in cancer. Noncanonical TICs that are located in the 5'UTR, although thought by some to be disadvantageous and suppressed by evolution, may translate oncogenic proteins because of their flanking nucleotides.

Human UFSP1 translated from an upstream near-cognate initiation codon functions as an active UFM1-specific protease.

Ubiquitin-fold modifier 1 (UFM1) is a recently identified ubiquitin-like posttranslational modification with important biological functions. However, the regulatory mechanisms governing UFM1 modification of target proteins (UFMylation) and the cellular processes controlled by UFMylation remain largely unknown. It has been previously shown that a UFM1-specific protease (UFSP2) mediates the maturation of the UFM1 precursor and drives the de-UFMylation reaction. Furthermore, it has long been thought that UFSP1, an ortholog of UFSP2, is inactive in many organisms, including human, because it lacks an apparent protease domain when translated from the canonical start codon (445AUG). Here, we demonstrate using the combination of site-directed mutagenesis, CRISPR/Cas9-mediated genome editing, and mass spectrometry approaches that translation of human UFSP1 initiates from an upstream near-cognate codon, 217CUG, via eukaryotic translation initiation factor eIF2A-mediated translational initiation rather than from the annotated 445AUG, revealing the presence of a catalytic protease domain containing a Cys active site. Moreover, we show that both UFSP1 and UFSP2 mediate maturation of UFM1 and de-UFMylation of target proteins. This study demonstrates that human UFSP1 functions as an active UFM1-specific protease, thus contributing to our understanding of the UFMylation/de-UFMylation process.

An exceptional biallelic N-terminal frame shift mutation in ZMPSTE24 leads to non-lethal progeria due to possible utilization of a downstream alternative start codon.

Biallelic mutations in ZMPSTE24 are known to be associated with autosomal recessive mandibuloacral dysplasia with type B lipodystrophy (MADB) and lethal restrictive dermopathy (RD), respectively. Disease manifestation is depending on the remaining enzyme activity of the mutated ZMPSTE24 protein. To date, complete loss of function has exclusively been reported in RD cases. In this study, we identified a novel N-terminal homozygous frameshift mutation (c.28_29insA) in a consanguineous family segregating with MADB. An in-depth analysis of the mutated sequence revealed, that the one base pair insertion creates a novel downstream in-frame start codon, which supposedly serves as an alternative translation initiation site (TIS). This possible rescue mechanism would explain the relatively mild clinical outcome in the studied individuals. Our findings demonstrate the necessity for careful interpretation of N-terminal variants potentially effecting translation initiation.

C9orf72 ALS/FTD dipeptide repeat protein levels are reduced by small molecules that inhibit PKA or enhance protein degradation.

Intronic GGGGCC (G4C2) hexanucleotide repeat expansion within the human C9orf72 gene represents the most common cause of familial forms of amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD) (C9ALS/FTD). Repeat-associated non-AUG (RAN) translation of repeat-containing C9orf72 RNA results in the production of neurotoxic dipeptide-repeat proteins (DPRs). Here, we developed a high-throughput drug screen for the identification of positive and negative modulators of DPR levels. We found that HSP90 inhibitor geldanamycin and aldosterone antagonist spironolactone reduced DPR levels by promoting protein degradation via the proteasome and autophagy pathways respectively. Surprisingly, cAMP-elevating compounds boosting protein kinase A (PKA) activity increased DPR levels. Inhibition of PKA activity, by both pharmacological and genetic approaches, reduced DPR levels in cells and rescued pathological phenotypes in a Drosophila model of C9ALS/FTD. Moreover, knockdown of PKA-catalytic subunits correlated with reduced translation efficiency of DPRs, while the PKA inhibitor H89 reduced endogenous DPR levels in C9ALS/FTD patient-derived iPSC motor neurons. Together, our results suggest new and druggable pathways modulating DPR levels in C9ALS/FTD.

High-risk human papillomavirus-18 uses an mRNA sequence to synthesize oncoprotein E6 in tumors.

Cervical cancer is the fourth most common cause of cancer in women worldwide in terms of both incidence and mortality. Persistent infection with high-risk types of human papillomavirus (HPV), namely 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68, constitute a necessary cause for the development of cervical cancer. Viral oncoproteins E6 and E7 play central roles in the carcinogenic process by virtue of their interactions with cell master proteins such as p53, retinoblastoma (Rb), mammalian target of rapamycin (mTOR), and c-MYC. For the synthesis of E6 and E7, HPVs use a bicistronic messenger RNA (mRNA) that has been studied in cultured cells. Here, we report that in cervical tumors, HPV-18, -39, and -45 transcribe E6/E7 mRNAs with extremely short 5' untranslated regions (UTRs) or even lacking a 5' UTR (i.e., zero to three nucleotides long) to express E6. We show that the translation of HPV-18 E6 cistron is regulated by the motif ACCaugGCGCG(C/A)UUU surrounding the AUG start codon, which we term Translation Initiation of Leaderless mRNAs (TILM). This motif is conserved in all HPV types of the phylogenetically coherent group forming genus alpha, species 7, which infect mucosal epithelia. We further show that the translation of HPV-18 E6 largely relies on the cap structure and eIF4E and eIF4AI, two key translation initiation factors linking translation and cancer but does not involve scanning. Our results support the notion that E6 forms the center of the positive oncogenic feedback loop node involving eIF4E, the mTOR cascade, and p53.

Human oncoprotein 5MP suppresses general and repeat-associated non-AUG translation via eIF3 by a common mechanism.

eIF5-mimic protein (5MP) is a translational regulatory protein that binds the small ribosomal subunit and modulates its activity. 5MP is proposed to reprogram non-AUG translation rates for oncogenes in cancer, but its role in controlling non-AUG initiated synthesis of deleterious repeat-peptide products, such as FMRpolyG observed in fragile-X-associated tremor ataxia syndrome (FXTAS), is unknown. Here, we show that 5MP can suppress both general and repeat-associated non-AUG (RAN) translation by a common mechanism in a manner dependent on its interaction with eIF3. Essentially, 5MP displaces eIF5 through the eIF3c subunit within the preinitiation complex (PIC), thereby increasing the accuracy of initiation. In Drosophila, 5MP/Kra represses neuronal toxicity and enhances the lifespan in an FXTAS disease model. These results implicate 5MP in protecting cells from unwanted byproducts of non-AUG translation in neurodegeneration.

Start codon disruption with CRISPR/Cas9 prevents murine Fuchs' endothelial corneal dystrophy.

A missense mutation of collagen type VIII alpha 2 chain (COL8A2) gene leads to early-onset Fuchs' endothelial corneal dystrophy (FECD), which progressively impairs vision through the loss of corneal endothelial cells. We demonstrate that CRISPR/Cas9-based postnatal gene editing achieves structural and functional rescue in a mouse model of FECD. A single intraocular injection of an adenovirus encoding both the Cas9 gene and guide RNA (Ad-Cas9-Col8a2gRNA) efficiently knocked down mutant COL8A2 expression in corneal endothelial cells, prevented endothelial cell loss, and rescued corneal endothelium pumping function in adult Col8a2 mutant mice. There were no adverse sequelae on histology or electroretinography. Col8a2 start codon disruption represents a non-surgical strategy to prevent vision loss in early-onset FECD. As this demonstrates the ability of Ad-Cas9-gRNA to restore the phenotype in adult post-mitotic cells, this method may be widely applicable to adult-onset diseases, even in tissues affected with disorders of non-reproducing cells.

Improved Identification of Small Open Reading Frames Encoded Peptides by Top-Down Proteomic Approaches and De Novo Sequencing.

Small open reading frames (sORFs) have translational potential to produce peptides that play essential roles in various biological processes. Nevertheless, many sORF-encoded peptides (SEPs) are still on the prediction level. Here, we construct a strategy to analyze SEPs by combining top-down and de novo sequencing to improve SEP identification and sequence coverage. With de novo sequencing, we identified 1682 peptides mapping to 2544 human sORFs, which were all first characterized in this work. Two-thirds of these new sORFs have reading frame shifts and use a non-ATG start codon. The top-down approach identified 241 human SEPs, with high sequence coverage. The average length of the peptides from the bottom-up database search was 19 amino acids (AA); from de novo sequencing, it was 9 AA; and from the top-down approach, it was 25 AA. The longer peptide positively boosts the sequence coverage, more efficiently distinguishing SEPs from the known gene coding sequence. Top-down has the advantage of identifying peptides with sequential K/R or high K/R content, which is unfavorable in the bottom-up approach. Our method can explore new coding sORFs and obtain highly accurate sequences of their SEPs, which can also benefit future function research.

An analysis of the Kozak consensus in retinal genes and its relevance to gene therapy.

Purpose: The classic Kozak consensus is a critical genetic element included in gene therapy transgenes to encourage the translation of the therapeutic coding sequence. Despite optimizations of other transgene elements, the Kozak consensus has not yet been considered for potential tissue-specific sequence refinement. We screened the -9 to -1 region relative to the AUG start codon of retina-specific genes to identify whether a Kozak consensus that is different from the classic sequence may be more appropriate for inclusion in gene therapy transgenes that treat inherited retinal disease. Methods: Sequences for 135 genes known to cause nonsyndromic inherited retinal disease were extracted from the NCBI database, and the -9 to -1 nucleotides were compared. This panel was then refined to 75 genes with specific retinal functions, for which the -9 to -1 nucleotides were placed in front of a GFP transcript sequence and RNAfold predictions performed. These were compared with a GFP sequence with the classic Kozak consensus (GCCGCCACC), and sequences from retinal genes with minimum free energy (MFE) predictions greater than the reference sequence were selected to generate an optimized Kozak consensus sequence. The original Kozak consensus and the refined retina Kozak consensus were placed upstream of the Renilla luciferase coding sequence, which were used to transfect retinoblastoma cell lines Y-79 and WERI-RB-1 and HEK 293T/17 cells. Results: The nucleotide frequencies of the original panel of genes were determined to be comparable to the classic Kozak consensus. RNAfold analysis of a GFP transcript with the classic Kozak sequence in the 5' untranslated region (UTR) generated an MFE prediction of -503.3 kcal/mol. RNAfold analysis was then performed with a GFP transcript containing each -9 to -1 Kozak sequence of 75 retinal genes. Thirty-eight of the 75 genes provided a greater MFE value than -503.3 kcal/mol and exhibited an absence of stable secondary structures before the AUG codon. The -9 to -1 nucleotide frequencies of these genes identified a Kozak consensus of ACCGAGACC, differing from the classic Kozak consensus at positions -9, -5, and -4. Applying this sequence to the GFP transcript increased the MFE prediction to -500.1 kcal/mol. The newly identified retina Kozak sequence was also applied to Renilla luciferase plus the REP1 and RPGR transcripts used in current clinical trials. In all examples, the predicted transcript MFE score increased when compared with the current transcript sequences containing classic Kozak consensus sequences. In vitro transfections identified a 7%-9% increase in Renilla activity when incorporating the optimized Kozak sequence. Conclusions: The Kozak consensus is a critical element of eukaryotic genes; therefore, it is a required feature of gene therapy transgenes. To date, the classic sequence of GCCRCC (-6 to -1) has typically been incorporated in gene therapy transgenes, but the analysis described here suggests that, for vectors targeting the retina, using a Kozak consensus derived from retinal genes can provide increased expression of the target product.

Role of Hakai in m6A modification pathway in Drosophila.

N6-methyladenosine (m6A), the most abundant internal modification in eukaryotic mRNA, is installed by a multi-component writer complex; however, the exact roles of each component remain poorly understood. Here we show that a potential E3 ubiquitin ligase Hakai colocalizes and interacts with other m6A writer components, and Hakai mutants exhibit typical m6A pathway defects in Drosophila, such as lowered m6A levels in mRNA, aberrant Sxl alternative splicing, wing and behavior defects. Hakai, Vir, Fl(2)d and Flacc form a stable complex, and disruption of either Hakai, Vir or Fl(2)d led to the degradation of the other three components. Furthermore, MeRIP-seq indicates that the effective m6A modification is mostly distributed in 5' UTRs in Drosophila, in contrast to the mammalian system. Interestingly, we demonstrate that m6A modification is deposited onto the Sxl mRNA in a sex-specific fashion, which depends on the m6A writer. Together, our work not only advances the understanding of mechanism and regulation of the m6A writer complex, but also provides insights into how Sxl cooperate with the m6A pathway to control its own splicing.