Determinants of Base Editing Outcomes from Target Library Analysis and Machine Learning.

Although base editors are widely used to install targeted point mutations, the factors that determine base editing outcomes are not well understood. We characterized sequence-activity relationships of 11 cytosine and adenine base editors (CBEs and ABEs) on 38,538 genomically integrated targets in mammalian cells and used the resulting outcomes to train BE-Hive, a machine learning model that accurately predicts base editing genotypic outcomes (R ≈ 0.9) and efficiency (R ≈ 0.7). We corrected 3,388 disease-associated SNVs with ≥90% precision, including 675 alleles with bystander nucleotides that BE-Hive correctly predicted would not be edited. We discovered determinants of previously unpredictable C-to-G, or C-to-A editing and used these discoveries to correct coding sequences of 174 pathogenic transversion SNVs with ≥90% precision. Finally, we used insights from BE-Hive to engineer novel CBE variants that modulate editing outcomes. These discoveries illuminate base editing, enable editing at previously intractable targets, and provide new base editors with improved editing capabilities.

Regulation of Co-transcriptional Pre-mRNA Splicing by m6A through the Low-Complexity Protein hnRNPG.

N6-methyladenosine (m6A) modification occurs co-transcriptionally and impacts pre-mRNA processing; however, the mechanism of co-transcriptional m6A-dependent alternative splicing regulation is still poorly understood. Heterogeneous nuclear ribonucleoprotein G (hnRNPG) is an m6A reader protein that binds RNA through RRM and Arg-Gly-Gly (RGG) motifs. Here, we show that hnRNPG directly binds to the phosphorylated carboxy-terminal domain (CTD) of RNA polymerase II (RNAPII) using RGG motifs in its low-complexity region. Through interactions with the phosphorylated CTD and nascent RNA, hnRNPG associates co-transcriptionally with RNAPII and regulates alternative splicing transcriptome-wide. m6A near splice sites in nascent pre-mRNA modulates hnRNPG binding, which influences RNAPII occupancy patterns and promotes exon inclusion. Our results reveal an integrated mechanism of co-transcriptional m6A-mediated splicing regulation, in which an m6A reader protein uses RGG motifs to co-transcriptionally interact with both RNAPII and m6A-modified nascent pre-mRNA to modulate RNAPII occupancy and alternative splicing.

CMT2N-causing aminoacylation domain mutants enable Nrp1 interaction with AlaRS.

Through dominant mutations, aminoacyl-tRNA synthetases constitute the largest protein family linked to Charcot-Marie-Tooth disease (CMT). An example is CMT subtype 2N (CMT2N), caused by individual mutations spread out in AlaRS, including three in the aminoacylation domain, thereby suggesting a role for a tRNA-charging defect. However, here we found that two are aminoacylation defective but that the most widely distributed R329H is normal as a purified protein in vitro and in unfractionated patient cell samples. Remarkably, in contrast to wild-type (WT) AlaRS, all three mutant proteins gained the ability to interact with neuropilin 1 (Nrp1), the receptor previously linked to CMT pathogenesis in GlyRS. The aberrant AlaRS-Nrp1 interaction is further confirmed in patient samples carrying the R329H mutation. However, CMT2N mutations outside the aminoacylation domain do not induce the Nrp1 interaction. Detailed biochemical and biophysical investigations, including X-ray crystallography, small-angle X-ray scattering, hydrogen-deuterium exchange (HDX), switchSENSE hydrodynamic diameter determinations, and protease digestions reveal a mutation-induced structural loosening of the aminoacylation domain that correlates with the Nrp1 interaction. The b1b2 domains of Nrp1 are responsible for the interaction with R329H AlaRS. The results suggest Nrp1 is more broadly associated with CMT-associated members of the tRNA synthetase family. Moreover, we revealed a distinct structural loosening effect induced by a mutation in the editing domain and a lack of conformational impact with C-Ala domain mutations, indicating mutations in the same protein may cause neuropathy through different mechanisms. Our results show that, as with other CMT-associated tRNA synthetases, aminoacylation per se is not relevant to the pathology.

Detection and quantification of glycosylated queuosine modified tRNAs by acid denaturing and APB gels.

Queuosine (Q) is a conserved tRNA modification in bacteria and eukaryotes. Eukaryotic Q-tRNA modification occurs through replacing the guanine base with the scavenged metabolite queuine at the wobble position of tRNAs with G34U35N36 anticodon (Tyr, His, Asn, Asp) by the QTRT1/QTRT2 heterodimeric enzyme encoded in the genome. In humans, Q-modification in tRNATyr and tRNAAsp are further glycosylated with galactose and mannose, respectively. Although galactosyl-Q (galQ) and mannosyl-Q (manQ) can be measured by LC/MS approaches, the difficulty of detecting and quantifying these modifications with low sample inputs has hindered their biological investigations. Here we describe a simple acid denaturing gel and nonradioactive northern blot method to detect and quantify the fraction of galQ/manQ-modified tRNA using just microgram amounts of total RNA. Our method relies on the secondary amine group of galQ/manQ becoming positively charged to slow their migration in acid denaturing gels commonly used for tRNA charging studies. We apply this method to determine the Q and galQ/manQ modification kinetics in three human cells lines. For Q-modification, tRNAAsp is modified the fastest, followed by tRNAHis, tRNATyr, and tRNAAsn Compared to Q-modification, glycosylation occurs at a much slower rate for tRNAAsp, but at a similar rate for tRNATyr Our method enables easy access to study the function of these enigmatic tRNA modifications.

CMT disease severity correlates with mutation-induced open conformation of histidyl-tRNA synthetase, not aminoacylation loss, in patient cells.

Aminoacyl-transfer RNA (tRNA) synthetases (aaRSs) are the largest protein family causatively linked to neurodegenerative Charcot-Marie-Tooth (CMT) disease. Dominant mutations cause the disease, and studies of CMT disease-causing mutant glycyl-tRNA synthetase (GlyRS) and tyrosyl-tRNA synthetase (TyrRS) showed their mutations create neomorphic structures consistent with a gain-of-function mechanism. In contrast, based on a haploid yeast model, loss of aminoacylation function was reported for CMT disease mutants in histidyl-tRNA synthetase (HisRS). However, neither that nor prior work of any CMT disease-causing aaRS investigated the aminoacylation status of tRNAs in the cellular milieu of actual patients. Using an assay that interrogated aminoacylation levels in patient cells, we investigated a HisRS-linked CMT disease family with the most severe disease phenotype. Strikingly, no difference in charged tRNA levels between normal and diseased family members was found. In confirmation, recombinant versions of 4 other HisRS CMT disease-causing mutants showed no correlation between activity loss in vitro and severity of phenotype in vivo. Indeed, a mutation having the most detrimental impact on activity was associated with a mild disease phenotype. In further work, using 3 independent biophysical analyses, structural opening (relaxation) of mutant HisRSs at the dimer interface best correlated with disease severity. In fact, the HisRS mutation in the severely afflicted patient family caused the largest degree of structural relaxation. These data suggest that HisRS-linked CMT disease arises from open conformation-induced mechanisms distinct from loss of aminoacylation.

Queuosine modification protects cognate tRNAs against ribonuclease cleavage.

Eukaryotic transfer RNAs (tRNA) contain on average 13 modifications that perform a wide range of roles in translation and in the generation of tRNA fragments that regulate gene expression. Queuosine (Q) modification occurs in the wobble anticodon position of tRNAs for amino acids His, Asn, Tyr, and Asp. In eukaryotes, Q modification is fully dependent on diet or on gut microbiome in multicellular organisms. Despite decades of study, cellular roles of Q modification remain to be fully elucidated. Here we show that in human cells, Q modification specifically protects its cognate tRNAHis and tRNAAsn against cleavage by ribonucleases. We generated cell lines that contain completely depleted or fully Q-modified tRNAs. Using these resources, we found that Q modification significantly reduces angiogenin cleavage of its cognate tRNAs in vitro. Q modification does not change the cellular abundance of the cognate full-length tRNAs, but alters the cellular content of their fragments in vivo in the absence and presence of stress. Our results provide a new biological aspect of Q modification and a mechanism of how Q modification alters small RNA pools in human cells.

Dual function of C/D box small nucleolar RNAs in rRNA modification and alternative pre-mRNA splicing.

C/D box small nucleolar RNAs (SNORDs) are small noncoding RNAs, and their best-understood function is to target the methyltransferase fibrillarin to rRNA (for example, SNORD27 performs 2'-O-methylation of A27 in 18S rRNA). Unexpectedly, we found a subset of SNORDs, including SNORD27, in soluble nuclear extract made under native conditions, where fibrillarin was not detected, indicating that a fraction of the SNORD27 RNA likely forms a protein complex different from canonical snoRNAs found in the insoluble nuclear fraction. As part of this previously unidentified complex,SNORD27 regulates the alternative splicing of the transcription factor E2F7p re-mRNA through direct RNA-RNA interaction without methylating the RNA, likely by competing with U1 small nuclear ribonucleoprotein (snRNP). Furthermore, knockdown of SNORD27 activates previously "silent" exons in several other genes through base complementarity across the entire SNORD27 sequence, not just the antisense boxes. Thus, some SNORDs likely function in both rRNA and pre-mRNA processing, which increases the repertoire of splicing regulators and links both processes.

Autophosphorylation of alpha isoform of calcium/calmodulin-dependent kinase II regulates alcohol addiction-related behaviors.

The development of addiction is associated with a dysregulation of glutamatergic transmission in the brain reward circuit. α isoform of calcium/calmodulin-dependent kinase II (αCaMKII) is one of the key proteins that regulates structural and functional plasticity of glutamatergic synapses. αCaMKII activity can be controlled by the autophosphorylation of threonine 286. The role of this autophosphorylation in the regulation of addiction-related behaviors has been proposed but is still poorly understood. Here, using αCaMKII autophosphorylation-deficient mutant mice (T286A), we show that, in comparison with wild-type animals, they are less resistant to high doses of alcohol and do not show psychostimulant response neither to alcohol injections nor during voluntary alcohol drinking. T286A mutants are also less prone to develop alcohol addiction-related behaviors including an increased motivation for alcohol, persistent alcohol seeking during withdrawal and alcohol consumption on relapse. Finally, we demonstrate that αCaMKII autophosphorylation regulates also alcohol-induced remodeling of glutamatergic synapses in the hippocampus and amygdala. In conclusion, our data suggest that αCaMKII autophosphorylation-dependent remodeling of glutamatergic synapses is a plausible mechanism for the regulation of the alcohol addiction-related behaviors.

A new heavy lanthanide-dependent DNAzyme displaying strong metal cooperativity and unrescuable phosphorothioate effect.

In vitro selection of RNA-cleaving DNAzymes was performed using three heavy lanthanide ions (Ln(3+)): Ho(3+), Er(3+) and Tm(3+). The resulting sequences were aligned together and about half of the library contained a new family of DNAzyme. These DNAzymes have a simple loop structure, and they are active only with the seven heavy Ln(3+). Among the tested non-lanthanide ions, only Y(3+) induced cleavage and even Pb(2+) failed to cleave, suggesting a very high specificity. A representative DNAzyme, Tm7, has a sigmoidal metal binding curve with a Hill coefficient of 3, indicating that three metal ions are involved in the catalytic step. Its pH-rate profile has a slope of 1, suggesting a single deprotonation step is involved in the rate-limiting step. Tm7 has a cleavage rate of 1.6 min(-1) at pH 7.8 with 10 µM Er(3+). Phosphorothioate substitution at the cleavage junction completely inhibits the activity, which cannot be rescued by Cd(2+) alone, or by a mixture of Er(3+) and Cd(2+), suggesting that two interacting metal ions are involved in direct bonding to both non-bridging oxygen atoms. A new model involving three lanthanide ions is proposed based on this study. A biosensor is engineered using Tm7 to detect Dy(3+) down to 14 nM.

Biochemical Characterization of a Lanthanide-Dependent DNAzyme with Normal and Phosphorothioate-Modified Substrates.

A trivalent lanthanide (Ln(3+))-dependent RNA-cleaving DNAzyme, Ce13d, was recently isolated via in vitro selection. Ce13d is active in the presence of all Ln(3+) ions. Via introduction of a single phosphorothioate (PS) modification at the cleavage site, its activity with Ln(3+) decreases while all thiophilic metals can activate this DNAzyme. This property is unique to Ce13d and is not found in many other tested DNAzymes. This suggests the presence of a well-defined but general metal binding site. Herein, a systematic study of Ce13d with the PO substrate (using Ce(3+)) and the PS substrate (using Cd(2+)) is performed. In both the PO and PS systems, the highest activity was with ∼10 µM metal ions. Higher concentrations of Ce(3+) completely inhibit the activity, while Cd(2+) only slows the activity. A comparison of different metal ions suggests that the role of metal is to neutralize the phosphate negative charge. Both systems follow a similar pH-rate profile with a single deprotonation step, indicating similar reaction mechanisms. The activity difference between the Rp and Sp form of the PS substrate is <10-fold, which is much smaller than most known RNA-cleaving enzymes. Mutation studies identified eight highly conserved purines, among which the two adenines play mainly structural roles, while the guanines are likely to be involved in metal binding. Ce13d can serve as a model system for further understanding of DNAzyme biochemistry and bioinorganic chemistry.

Base editing rescue of spinal muscular atrophy in cells and in mice.

Spinal muscular atrophy (SMA), the leading genetic cause of infant mortality, arises from survival motor neuron (SMN) protein insufficiency resulting from SMN1 loss. Approved therapies circumvent endogenous SMN regulation and require repeated dosing or may wane. We describe genome editing of SMN2, an insufficient copy of SMN1 harboring a C6>T mutation, to permanently restore SMN protein levels and rescue SMA phenotypes. We used nucleases or base editors to modify five SMN2 regulatory regions. Base editing converted SMN2 T6>C, restoring SMN protein levels to wild type. Adeno-associated virus serotype 9-mediated base editor delivery in Δ7SMA mice yielded 87% average T6>C conversion, improved motor function, and extended average life span, which was enhanced by one-time base editor and nusinersen coadministration (111 versus 17 days untreated). These findings demonstrate the potential of a one-time base editing treatment for SMA.

tRNA Queuosine Modification Enzyme Modulates the Growth and Microbiome Recruitment to Breast Tumors.

BACKGROUND: Transfer RNA (tRNA) queuosine (Q)-modifications occur specifically in 4 cellular tRNAs at the wobble anticodon position. tRNA Q-modification in human cells depends on the gut microbiome because the microbiome product queuine is required for its installation by the enzyme Q tRNA ribosyltransferase catalytic subunit 1 (QTRT1) encoded in the human genome. Queuine is a micronutrient from diet and microbiome. Although tRNA Q-modification has been studied for a long time regarding its properties in decoding and tRNA fragment generation, how QTRT1 affects tumorigenesis and the microbiome is still poorly understood. RESULTS: We generated single clones of QTRT1-knockout breast cancer MCF7 cells using Double Nickase Plasmid. We also established a QTRT1-knockdown breast MDA-MB-231 cell line. The impacts of QTRT1 deletion or reduction on cell proliferation and migration in vitro were evaluated using cell culture, while the regulations on tumor growth in vivo were evaluated using a xenograft BALB/c nude mouse model. We found that QTRT1 deficiency in human breast cancer cells could change the functions of regulation genes, which are critical in cell proliferation, tight junction formation, and migration in human breast cancer cells in vitro and a breast tumor mouse model in vivo. We identified that several core bacteria, such as Lachnospiraceae, Lactobacillus, and Alistipes, were markedly changed in mice post injection with breast cancer cells. The relative abundance of bacteria in tumors induced from wildtype cells was significantly higher than those of QTRT1 deficiency cells. CONCLUSIONS: Our results demonstrate that the QTRT1 gene and tRNA Q-modification altered cell proliferation, junctions, and microbiome in tumors and the intestine, thus playing a critical role in breast cancer development.

Continuous evolution of SpCas9 variants compatible with non-G PAMs.

The targeting scope of Streptococcus pyogenes Cas9 (SpCas9) and its engineered variants is largely restricted to protospacer-adjacent motif (PAM) sequences containing G bases. Here we report the evolution of three new SpCas9 variants that collectively recognize NRNH PAMs (where R is A or G and H is A, C or T) using phage-assisted non-continuous evolution, three new phage-assisted continuous evolution strategies for DNA binding and a secondary selection for DNA cleavage. The targeting capabilities of these evolved variants and SpCas9-NG were characterized in HEK293T cells using a library of 11,776 genomically integrated protospacer-sgRNA pairs containing all possible NNNN PAMs. The evolved variants mediated indel formation and base editing in human cells and enabled A•T-to-G•C base editing of a sickle cell anemia mutation using a previously inaccessible CACC PAM. These new evolved SpCas9 variants, together with previously reported variants, in principle enable targeting of most NR PAM sequences and substantially reduce the fraction of genomic sites that are inaccessible by Cas9-based methods.

Quantification of Queuosine Modification Levels in tRNA from Human Cells Using APB Gel and Northern Blot.

Queuosine (Q) is a hypermodified base in the wobble anticodon position of tRNAs coding for the amino acids Tyr, His, Asn, and Asp. tRNA Q-modification is introduced by a queuine tRNA-ribosyltransferase (TGT) that replaces the guanine base at G34 at these tRNAs with the modified base. tRNA Q-modification is widely distributed among prokaryotic and eukaryotic organisms, but only bacteria synthesize Q-modified tRNA de novo. In mammals, tRNA Q-modifications strictly rely on the presence of gut microbiomes or diets to produce the queuine base. Despite decades of study, cellular roles of tRNA Q-modification are still not fully understood. Here we describe a method to quantify tRNA Q-modification levels in individual tRNAs from human cells based on the presence of a cis-diol in the Q modification. This cis-diol moiety slows modified tRNA migration through polyacrylamide gels supplemented with N-acryloyl-3-aminophenylboronic acid (APB) compared to the unmodified tRNA. This difference can be visualized by Northern blots using probes for specific tRNA.

Nuclear fate of yeast snoRNA is determined by co-transcriptional Rnt1 cleavage.

Small nucleolar RNA (snoRNA) are conserved and essential non-coding RNA that are transcribed by RNA Polymerase II (Pol II). Two snoRNA classes, formerly distinguished by their structure and ribonucleoprotein composition, act as guide RNA to target RNA such as ribosomal RNA, and thereby introduce specific modifications. We have studied the 5'end processing of individually transcribed snoRNA in S. cerevisiae to define their role in snoRNA biogenesis and functionality. Here we show that pre-snoRNA processing by the endonuclease Rnt1 occurs co-transcriptionally with removal of the m7G cap facilitating the formation of box C/D snoRNA. Failure of this process causes aberrant 3'end processing and mislocalization of snoRNA to the cytoplasm. Consequently, Rnt1-dependent 5'end processing of box C/D snoRNA is critical for snoRNA-dependent methylation of ribosomal RNA. Our results reveal that the 5'end processing of box C/D snoRNA defines their distinct pathway of maturation.