Engineering TALE-linked deaminases to facilitate precision adenine base editing in mitochondrial DNA.

DddA-derived cytosine base editors (DdCBEs) and transcription activator-like effector (TALE)-linked deaminases (TALEDs) catalyze targeted base editing of mitochondrial DNA (mtDNA) in eukaryotic cells, a method useful for modeling of mitochondrial genetic disorders and developing novel therapeutic modalities. Here, we report that A-to-G-editing TALEDs but not C-to-T-editing DdCBEs induce tens of thousands of transcriptome-wide off-target edits in human cells. To avoid these unwanted RNA edits, we engineered the substrate-binding site in TadA8e, the deoxy-adenine deaminase in TALEDs, and created TALED variants with fine-tuned deaminase activity. Our engineered TALED variants not only reduced RNA off-target edits by >99% but also minimized off-target mtDNA mutations and bystander edits at a target site. Unlike wild-type versions, our TALED variants were not cytotoxic and did not cause developmental arrest of mouse embryos. As a result, we obtained mice with pathogenic mtDNA mutations, associated with Leigh syndrome, which showed reduced heart rates.

Discovery of deaminase functions by structure-based protein clustering.

The elucidation of protein function and its exploitation in bioengineering have greatly advanced the life sciences. Protein mining efforts generally rely on amino acid sequences rather than protein structures. We describe here the use of AlphaFold2 to predict and subsequently cluster an entire protein family based on predicted structure similarities. We selected deaminase proteins to analyze and identified many previously unknown properties. We were surprised to find that most proteins in the DddA-like clade were not double-stranded DNA deaminases. We engineered the smallest single-strand-specific cytidine deaminase, enabling efficient cytosine base editor (CBE) to be packaged into a single adeno-associated virus (AAV). Importantly, we profiled a deaminase from this clade that edits robustly in soybean plants, which previously was inaccessible to CBEs. These discovered deaminases, based on AI-assisted structural predictions, greatly expand the utility of base editors for therapeutic and agricultural applications.

Targeted A-to-G base editing in human mitochondrial DNA with programmable deaminases.

Mitochondrial DNA (mtDNA) editing paves the way for disease modeling of mitochondrial genetic disorders in cell lines and animals and also for the treatment of these diseases in the future. Bacterial cytidine deaminase DddA-derived cytosine base editors (DdCBEs) enabling mtDNA editing, however, are largely limited to C-to-T conversions in the 5'-TC context (e.g., TC-to-TT conversions), suitable for generating merely 1/8 of all possible transition (purine-to-purine and pyrimidine-to-pyrimidine) mutations. Here, we present transcription-activator-like effector (TALE)-linked deaminases (TALEDs), composed of custom-designed TALE DNA-binding arrays, a catalytically impaired, full-length DddA variant or split DddA originated from Burkholderia cenocepacia, and an engineered deoxyadenosine deaminase derived from the E. coli TadA protein, which induce targeted A-to-G editing in human mitochondria. Custom-designed TALEDs were highly efficient in human cells, catalyzing A-to-G conversions at a total of 17 target sites in various mitochondrial genes with editing frequencies of up to 49%.

Co-targeting RNA Polymerases IV and V Promotes Efficient De Novo DNA Methylation in Arabidopsis.

The RNA-directed DNA methylation (RdDM) pathway in plants controls gene expression via cytosine DNA methylation. The ability to manipulate RdDM would shed light on the mechanisms and applications of DNA methylation to control gene expression. Here, we identified diverse RdDM proteins that are capable of targeting methylation and silencing in Arabidopsis when tethered to an artificial zinc finger (ZF-RdDM). We studied their order of action within the RdDM pathway by testing their ability to target methylation in different mutants. We also evaluated ectopic siRNA biogenesis, RNA polymerase V (Pol V) recruitment, targeted DNA methylation, and gene-expression changes at thousands of ZF-RdDM targets. We found that co-targeting both arms of the RdDM pathway, siRNA biogenesis and Pol V recruitment, dramatically enhanced targeted methylation. This work defines how RdDM components establish DNA methylation and enables new strategies for epigenetic gene regulation via targeted DNA methylation.

Discovery of cytosine deaminases enables base-resolution methylome mapping using a single enzyme.

Deaminases have important uses in modification detection and genome editing. However, the range of applications is limited by the small number of characterized enzymes. To expand the toolkit of deaminases, we developed an in vitro approach that bypasses a major hurdle with their toxicity in cells. We assayed 175 putative cytosine deaminases on a variety of substrates and found a broad range of activity on double- and single-stranded DNA in various sequence contexts, including CpG-specific deaminases and enzymes without sequence preference. We also characterized enzyme selectivity across six DNA modifications and reported enzymes that do not deaminate modified cytosines. The detailed analysis of diverse deaminases opens new avenues for biotechnological and medical applications. As a demonstration, we developed SEM-seq, a non-destructive single-enzyme methylation sequencing method using a modification-sensitive double-stranded DNA deaminase. The streamlined protocol enables accurate, base-resolution methylome mapping of scarce biological material, including cell-free DNA and 10 pg input DNA.

A DddA ortholog-based and transactivator-assisted nuclear and mitochondrial cytosine base editors with expanded target compatibility.

Bacterial double-stranded DNA (dsDNA) cytosine deaminase DddAtox-derived cytosine base editor (DdCBE) and its evolved variant, DddA11, guided by transcription-activator-like effector (TALE) proteins, enable mitochondrial DNA (mtDNA) editing at TC or HC (H = A, C, or T) sequence contexts, while it remains relatively unattainable for GC targets. Here, we identified a dsDNA deaminase originated from a Roseburia intestinalis interbacterial toxin (riDddAtox) and generated CRISPR-mediated nuclear DdCBEs (crDdCBEs) and mitochondrial CBEs (mitoCBEs) using split riDddAtox, which catalyzed C-to-T editing at both HC and GC targets in nuclear and mitochondrial genes. Moreover, transactivator (VP64, P65, or Rta) fusion to the tail of DddAtox- or riDddAtox-mediated crDdCBEs and mitoCBEs substantially improved nuclear and mtDNA editing efficiencies by up to 3.5- and 1.7-fold, respectively. We also used riDddAtox-based and Rta-assisted mitoCBE to efficiently stimulate disease-associated mtDNA mutations in cultured cells and in mouse embryos with conversion frequencies of up to 58% at non-TC targets.

DNA mismatch and damage patterns revealed by single-molecule sequencing.

Mutations accumulate in the genome of every cell of the body throughout life, causing cancer and other diseases1,2. Most mutations begin as nucleotide mismatches or damage in one of the two strands of the DNA before becoming double-strand mutations if unrepaired or misrepaired3,4. However, current DNA-sequencing technologies cannot accurately resolve these initial single-strand events. Here we develop a single-molecule, long-read sequencing method (Hairpin Duplex Enhanced Fidelity sequencing (HiDEF-seq)) that achieves single-molecule fidelity for base substitutions when present in either one or both DNA strands. HiDEF-seq also detects cytosine deamination-a common type of DNA damage-with single-molecule fidelity. We profiled 134 samples from diverse tissues, including from individuals with cancer predisposition syndromes, and derive from them single-strand mismatch and damage signatures. We find correspondences between these single-strand signatures and known double-strand mutational signatures, which resolves the identity of the initiating lesions. Tumours deficient in both mismatch repair and replicative polymerase proofreading show distinct single-strand mismatch patterns compared to samples that are deficient in only polymerase proofreading. We also define a single-strand damage signature for APOBEC3A. In the mitochondrial genome, our findings support a mutagenic mechanism occurring primarily during replication. As double-strand DNA mutations are only the end point of the mutation process, our approach to detect the initiating single-strand events at single-molecule resolution will enable studies of how mutations arise in a variety of contexts, especially in cancer and ageing.

Structural insights into DNMT5-mediated ATP-dependent high-fidelity epigenome maintenance.

Epigenetic evolution occurs over million-year timescales in Cryptococcus neoformans and is mediated by DNMT5, the first maintenance type cytosine methyltransferase identified in the fungal or protist kingdoms, the first dependent on adenosine triphosphate (ATP), and the most hemimethyl-DNA-specific enzyme known. To understand these novel properties, we solved cryo-EM structures of CnDNMT5 in three states. These studies reveal an elaborate allosteric cascade in which hemimethylated DNA binding first activates the SNF2 ATPase domain by a large rigid body rotation while the target cytosine partially flips out of the DNA duplex. ATP binding then triggers striking structural reconfigurations of the methyltransferase catalytic pocket to enable cofactor binding, completion of base flipping, and catalysis. Bound unmethylated DNA does not open the catalytic pocket and is instead ejected upon ATP binding, driving high fidelity. This unprecedented chaperone-like, enzyme-remodeling role of the SNF2 ATPase domain illuminates how energy is used to enable faithful epigenetic memory.

Mechanism and function of oxidative reversal of DNA and RNA methylation.

The importance of eukaryotic DNA methylation [5-methylcytosine (5mC)] in transcriptional regulation and development was first suggested almost 40 years ago. However, the molecular mechanism underlying the dynamic nature of this epigenetic mark was not understood until recently, following the discovery that the TET proteins, a family of AlkB-like Fe(II)/α-ketoglutarate-dependent dioxygenases, can oxidize 5mC to generate 5-hydroxymethylcytosine (5hmC), 5-formylcytosine (5fC), and 5-carboxylcytosine (5caC). Since then, several mechanisms that are responsible for processing oxidized 5mC derivatives to achieve DNA demethylation have emerged. Our biochemical understanding of the DNA demethylation process has prompted new investigations into the biological functions of DNA demethylation. Characterization of two additional AlkB family proteins, FTO and ALKBH5, showed that they possess demethylase activity toward N(6)-methyladenosine (m(6)A) in RNA, indicating that members of this subfamily of dioxygenases have a general function in demethylating nucleic acids. In this review, we discuss recent advances in this emerging field, focusing on the mechanism and function of TET-mediated DNA demethylation.

Programming and inheritance of parental DNA methylomes in mammals.

The reprogramming of parental methylomes is essential for embryonic development. In mammals, paternal 5-methylcytosines (5mCs) have been proposed to be actively converted to oxidized bases. These paternal oxidized bases and maternal 5mCs are believed to be passively diluted by cell divisions. By generating single-base resolution, allele-specific DNA methylomes from mouse gametes, early embryos, and primordial germ cell (PGC), as well as single-base-resolution maps of oxidized cytosine bases for early embryos, we report the existence of 5hmC and 5fC in both maternal and paternal genomes and find that 5mC or its oxidized derivatives, at the majority of demethylated CpGs, are converted to unmodified cytosines independent of passive dilution from gametes to four-cell embryos. Therefore, we conclude that paternal methylome and at least a significant proportion of maternal methylome go through active demethylation during embryonic development. Additionally, all the known imprinting control regions (ICRs) were classified into germ-line or somatic ICRs.

Increased mutation and gene conversion within human segmental duplications.

Single-nucleotide variants (SNVs) in segmental duplications (SDs) have not been systematically assessed because of the limitations of mapping short-read sequencing data1,2. Here we constructed 1:1 unambiguous alignments spanning high-identity SDs across 102 human haplotypes and compared the pattern of SNVs between unique and duplicated regions3,4. We find that human SNVs are elevated 60% in SDs compared to unique regions and estimate that at least 23% of this increase is due to interlocus gene conversion (IGC) with up to 4.3 megabase pairs of SD sequence converted on average per human haplotype. We develop a genome-wide map of IGC donors and acceptors, including 498 acceptor and 454 donor hotspots affecting the exons of about 800 protein-coding genes. These include 171 genes that have 'relocated' on average 1.61 megabase pairs in a subset of human haplotypes. Using a coalescent framework, we show that SD regions are slightly evolutionarily older when compared to unique sequences, probably owing to IGC. SNVs in SDs, however, show a distinct mutational spectrum: a 27.1% increase in transversions that convert cytosine to guanine or the reverse across all triplet contexts and a 7.6% reduction in the frequency of CpG-associated mutations when compared to unique DNA. We reason that these distinct mutational properties help to maintain an overall higher GC content of SD DNA compared to that of unique DNA, probably driven by GC-biased conversion between paralogous sequences5,6.

Single-cell DNA methylome and 3D multi-omic atlas of the adult mouse brain.

Cytosine DNA methylation is essential in brain development and is implicated in various neurological disorders. Understanding DNA methylation diversity across the entire brain in a spatial context is fundamental for a complete molecular atlas of brain cell types and their gene regulatory landscapes. Here we used single-nucleus methylome sequencing (snmC-seq3) and multi-omic sequencing (snm3C-seq)1 technologies to generate 301,626 methylomes and 176,003 chromatin conformation-methylome joint profiles from 117 dissected regions throughout the adult mouse brain. Using iterative clustering and integrating with companion whole-brain transcriptome and chromatin accessibility datasets, we constructed a methylation-based cell taxonomy with 4,673 cell groups and 274 cross-modality-annotated subclasses. We identified 2.6 million differentially methylated regions across the genome that represent potential gene regulation elements. Notably, we observed spatial cytosine methylation patterns on both genes and regulatory elements in cell types within and across brain regions. Brain-wide spatial transcriptomics data validated the association of spatial epigenetic diversity with transcription and improved the anatomical mapping of our epigenetic datasets. Furthermore, chromatin conformation diversities occurred in important neuronal genes and were highly associated with DNA methylation and transcription changes. Brain-wide cell-type comparisons enabled the construction of regulatory networks that incorporate transcription factors, regulatory elements and their potential downstream gene targets. Finally, intragenic DNA methylation and chromatin conformation patterns predicted alternative gene isoform expression observed in a whole-brain SMART-seq2 dataset. Our study establishes a brain-wide, single-cell DNA methylome and 3D multi-omic atlas and provides a valuable resource for comprehending the cellular-spatial and regulatory genome diversity of the mouse brain.

BARBEKO'ing in the lab: Versatile CRISPR screens with barcoded base editors.

Xu et al. (2021) describe a novel two-pronged CRISPR screen, termed BARBEKO, by coupling cytosine base editors and internally barcoded sgRNAs to eliminate double-stranded break-induced toxicity, enable high multiplicities of infection, and ensure experimental reproducibility.

Ferritin heavy chain supports stability and function of the regulatory T cell lineage.

Regulatory T (TREG) cells develop via a program orchestrated by the transcription factor forkhead box protein P3 (FOXP3). Maintenance of the TREG cell lineage relies on sustained FOXP3 transcription via a mechanism involving demethylation of cytosine-phosphate-guanine (CpG)-rich elements at conserved non-coding sequences (CNS) in the FOXP3 locus. This cytosine demethylation is catalyzed by the ten-eleven translocation (TET) family of dioxygenases, and it involves a redox reaction that uses iron (Fe) as an essential cofactor. Here, we establish that human and mouse TREG cells express Fe-regulatory genes, including that encoding ferritin heavy chain (FTH), at relatively high levels compared to conventional T helper cells. We show that FTH expression in TREG cells is essential for immune homeostasis. Mechanistically, FTH supports TET-catalyzed demethylation of CpG-rich sequences CNS1 and 2 in the FOXP3 locus, thereby promoting FOXP3 transcription and TREG cell stability. This process, which is essential for TREG lineage stability and function, limits the severity of autoimmune neuroinflammation and infectious diseases, and favors tumor progression. These findings suggest that the regulation of intracellular iron by FTH is a stable property of TREG cells that supports immune homeostasis and limits the pathological outcomes of immune-mediated inflammation.

Genome-wide profiling of 5-formylcytosine reveals its roles in epigenetic priming.

TET proteins oxidize 5-methylcytosine (5mC) to 5-hydroxymethylcytosine (5hmC), 5-formylcytosine (5fC), and 5-carboxylcytosine (5caC). 5fC and 5caC are excised by mammalian DNA glycosylase TDG, implicating 5mC oxidation in DNA demethylation. Here, we show that the genomic locations of 5fC can be determined by coupling chemical reduction with biotin tagging. Genome-wide mapping of 5fC in mouse embryonic stem cells (mESCs) reveals that 5fC preferentially occurs at poised enhancers among other gene regulatory elements. Application to Tdg null mESCs further suggests that 5fC production coordinates with p300 in remodeling epigenetic states of enhancers. This process, which is not influenced by 5hmC, appears to be associated with further oxidation of 5hmC and commitment to demethylation through 5fC. Finally, we resolved 5fC at base resolution by hydroxylamine-based protection from bisulfite-mediated deamination, thereby confirming sites of 5fC accumulation. Our results reveal roles of active 5mC/5hmC oxidation and TDG-mediated demethylation in epigenetic tuning at regulatory elements.

Genome-wide analysis reveals TET- and TDG-dependent 5-methylcytosine oxidation dynamics.

TET dioxygenases successively oxidize 5-methylcytosine (5mC) in mammalian genomes to 5-hydroxymethylcytosine (5hmC), 5-formylcytosine (5fC), and 5-carboxylcytosine (5caC). 5fC/5caC can be excised and repaired to regenerate unmodified cytosines by thymine-DNA glycosylase (TDG) and base excision repair (BER) pathway, but it is unclear to what extent and at which part of the genome this active demethylation process takes place. Here, we have generated genome-wide distribution maps of 5hmC/5fC/5caC using modification-specific antibodies in wild-type and Tdg-deficient mouse embryonic stem cells (ESCs). In wild-type mouse ESCs, 5fC/5caC accumulates to detectable levels at major satellite repeats but not at nonrepetitive loci. In contrast, Tdg depletion in mouse ESCs causes marked accumulation of 5fC and 5caC at a large number of proximal and distal gene regulatory elements. Thus, these results reveal the genome-wide view of iterative 5mC oxidation dynamics and indicate that TET/TDG-dependent active DNA demethylation process occurs extensively in the mammalian genome.

Dynamic readers for 5-(hydroxy)methylcytosine and its oxidized derivatives.

Tet proteins oxidize 5-methylcytosine (mC) to generate 5-hydroxymethyl (hmC), 5-formyl (fC), and 5-carboxylcytosine (caC). The exact function of these oxidative cytosine bases remains elusive. We applied quantitative mass-spectrometry-based proteomics to identify readers for mC and hmC in mouse embryonic stem cells (mESC), neuronal progenitor cells (NPC), and adult mouse brain tissue. Readers for these modifications are only partially overlapping, and some readers, such as Rfx proteins, display strong specificity. Interactions are dynamic during differentiation, as for example evidenced by the mESC-specific binding of Klf4 to mC and the NPC-specific binding of Uhrf2 to hmC, suggesting specific biological roles for mC and hmC. Oxidized derivatives of mC recruit distinct transcription regulators as well as a large number of DNA repair proteins in mouse ES cells, implicating the DNA damage response as a major player in active DNA demethylation.

MicroRNA-antagonism regulates breast cancer stemness and metastasis via TET-family-dependent chromatin remodeling.

Tumor cells metastasize to distant organs through genetic and epigenetic alterations, including changes in microRNA (miR) expression. Here we find miR-22 triggers epithelial-mesenchymal transition (EMT), enhances invasiveness and promotes metastasis in mouse xenografts. In a conditional mammary gland-specific transgenic (TG) mouse model, we show that miR-22 enhances mammary gland side-branching, expands the stem cell compartment, and promotes tumor development. Critically, miR-22 promotes aggressive metastatic disease in MMTV-miR-22 TG mice, as well as compound MMTV-neu or -PyVT-miR-22 TG mice. We demonstrate that miR-22 exerts its metastatic potential by silencing antimetastatic miR-200 through direct targeting of the TET (Ten eleven translocation) family of methylcytosine dioxygenases, thereby inhibiting demethylation of the mir-200 promoter. Finally, we show that miR-22 overexpression correlates with poor clinical outcomes and silencing of the TET-miR-200 axis in patients. Taken together, our findings implicate miR-22 as a crucial epigenetic modifier and promoter of EMT and breast cancer stemness toward metastasis.

Mitochondrial base editor induces substantial nuclear off-target mutations.

DddA-derived cytosine base editors (DdCBEs)-which are fusions of split DddA halves and transcription activator-like effector (TALE) array proteins from bacteria-enable targeted C•G-to-T•A conversions in mitochondrial DNA1. However, their genome-wide specificity is poorly understood. Here we show that the mitochondrial base editor induces extensive off-target editing in the nuclear genome. Genome-wide, unbiased analysis of its editome reveals hundreds of off-target sites that are TALE array sequence (TAS)-dependent or TAS-independent. TAS-dependent off-target sites in the nuclear DNA are often specified by only one of the two TALE repeats, challenging the principle that DdCBEs are guided by paired TALE proteins positioned in close proximity. TAS-independent off-target sites on nuclear DNA are frequently shared among DdCBEs with distinct TALE arrays. Notably, they co-localize strongly with binding sites for the transcription factor CTCF and are enriched in topologically associating domain boundaries. We engineered DdCBE to alleviate such off-target effects. Collectively, our results have implications for the use of DdCBEs in basic research and therapeutic applications, and suggest the need to thoroughly define and evaluate the off-target effects of base-editing tools.

Mitochondrial RNA modifications shape metabolic plasticity in metastasis.

Aggressive and metastatic cancers show enhanced metabolic plasticity1, but the precise underlying mechanisms of this remain unclear. Here we show how two NOP2/Sun RNA methyltransferase 3 (NSUN3)-dependent RNA modifications-5-methylcytosine (m5C) and its derivative 5-formylcytosine (f5C) (refs.2-4)-drive the translation of mitochondrial mRNA to power metastasis. Translation of mitochondrially encoded subunits of the oxidative phosphorylation complex depends on the formation of m5C at position 34 in mitochondrial tRNAMet. m5C-deficient human oral cancer cells exhibit increased levels of glycolysis and changes in their mitochondrial function that do not affect cell viability or primary tumour growth in vivo; however, metabolic plasticity is severely impaired as mitochondrial m5C-deficient tumours do not metastasize efficiently. We discovered that CD36-dependent non-dividing, metastasis-initiating tumour cells require mitochondrial m5C to activate invasion and dissemination. Moreover, a mitochondria-driven gene signature in patients with head and neck cancer is predictive for metastasis and disease progression. Finally, we confirm that this metabolic switch that allows the metastasis of tumour cells can be pharmacologically targeted through the inhibition of mitochondrial mRNA translation in vivo. Together, our results reveal that site-specific mitochondrial RNA modifications could be therapeutic targets to combat metastasis.