The EMBO Journal

Selective autophagy restrains innate immune signalling to maintain tissue homeostasis, yet how this repression is rapidly relieved during infection remains unclear. Here, we show that under basal conditions the inhibitor of {kappa}B kinase {gamma} (IKK{gamma}) Kenny is sequestered at autophagosomes through Atg8 and the selective autophagy receptor Ref(2)P, thereby silencing Imd pathway activity. Bacterial infection disrupts this interaction, releasing the IKK complex to enable immune signalling. Mechanistically, we identify the initiator caspase Dredd as a direct interactor of the IKK{gamma} Kenny and show that Dredd binds and cleaves Kenny in a ubiquitination-dependent manner during infection. This cleavage removes an N-terminal LC3-interacting region, uncoupling the IKK complex from autophagosomal degradation. Dredd-mediated processing of Kenny stabilises the IKK complex and is required for activation of the NF-{kappa}B transcription factor Relish, robust antibacterial responses, and host survival following infection. Together, these findings uncover a mechanism by which caspase-mediated cleavage intersects with selective autophagy to dynamically control NF-{kappa}B signalling during bacterial infection. Short summaryBacterial infection activates NF-{kappa}B signalling by triggering caspase-dependent cleavage of the IKK subunit Kenny, releasing the IKK complex from autophagosomal repression to enable effective innate immune responses.

Surveillance of mRNA translation relies on a suite of ribosome-associated quality control pathways. Recently, a novel pathway induced by trapping of translation factors in the ribosomal A-site has been described, involving ubiquitination of RPS27A/eS31 by the human E3 ubiquitin ligase RNF25. Here, we show that not only ribosome-stalling by low doses of translation inhibitors, but also amino acid starvation induces RPS27A/eS31 ubiquitination, identifying a natural trigger of RNF25 activation. Even under optimal growth conditions, RNF25 senses and resolves transient ribosome stalls. RPS27A/eS31 ubiquitination specifically depends on the ribosome collision sensor GCN1, a known cofactor of GCN2 involved in the integrated stress response. RNF25 and GCN2 both possess a GCN1-binding RWD domain, indicating a competitive relationship, with GCN2 acting as a negative regulator of RNF25 activation. Although both RNF25 and GCN2 respond to amino acid starvation, RPS27A/eS31 ubiquitination by RNF25 is not required for GCN2 activation, showing that both act in independent pathways. We propose that the RNF25 pathway acts as a first line of defence to resolve ribosome collisions, outcompeted by GCN2 binding to GCN1 under acute stress.

Stress granules form in response to diverse cellular perturbations to sequester translation components until the stress is resolved. Stress granules are composed of RNA-protein assemblies in membrane delimited structures and must be rapidly disassembled to release components to allow translation to resume. Disassembly of stress granules formed in response to heat stress is dependent on ubiquitiylation of stress granule components such as G3BP1. Ubiquitylation of stress granule proteins recruits the AAA-ATPase p97 (also known as VCP) to enable ubiquitin-dependent disassembly of these structures. Loss of p97 activity leads to the persistence of stress granules and is implicated in several age-related neurodegenerative diseases. Here we show that p97 recruitment to stress granules is dependent on its ubiquitin binding co-factor p47. p47 translocates to stress granules in response to a variety of cellular stressors and is required for the recruitment of p97 to stress granules. Loss of p47 leads to an inhibition in stress granule disassembly. We further show that p47 associates with G3BP1 in response to heat stress in a ubiquitin-dependent manner. Taken together our data adds to the growing list of p97 adaptors that are implicated in the recruitment of p97 for dissolution of stress granules.

Molecular chaperones are known for their role in preventing protein aggregation and assisting proteins in reaching their structurally functional state. DNAJB6, a J-domain protein that partners with Hsp70s and nucleotide exchange factors, is very potent in preventing amyloid formation of proteins with large intrinsically disordered regions (IDRs), including several disease-associated proteins. Complementary to this, we recently demonstrated a role for DNAJB6 in surveilling FG-Nucleoporins (FG-Nups) phase transitions and highlighted its role in nuclear pore complex assembly. We expand on this by showing that this activity of phase state surveillance is directed to several FG-Nups and shared with the closely related DNAJB2 and DNAJB8. We demonstrate that the surveillance mechanism of DNAJB6 is encoded in an unusually highly conserved IDR that promotes the formation of stable, gel-like assemblies of the chaperone itself. These assemblies likely provide a stable environment that can outcompete stable homotypic FG-Nup interactions and instead favors multivalent heterotypic chaperone:FG-Nup interactions. The evolutionary conservation of the DNAJB6-IDR, mutant analyses from both experimental in vitro and in cell data, and multiscale molecular dynamics simulations suggest that the sequence space for encoding stable gel-like assemblies is narrow and optimized to avoid self-aggregation while providing potent anti-amyloidogenic capacity.

Human LonP1 is an ATP-dependent mitochondrial protease that degrades damaged or redundant proteins. Indiscriminate proteolysis by LonP1 is limited through tight coordination of substrate recognition, unfolding, translocation and catalytic cleavage, yet the role of ATP hydrolysis in these individual steps remains unclear. Here, we show that LonP1 binds substrates and cleaves peptide bonds without ATP hydrolysis, whereas degradation of folded proteins strictly depends on ATP-driven unfolding and translocation. Initial substrate binding opens a closed ADP-bound resting state, enabling nucleotide exchange and stimulating ATPase activity. The opening also increases accessibility of the proteolytic chamber, modestly enhancing peptidase activity. Maximal peptidase activity is observed in a transition-state mimic stabilised by ADP{middle dot}AlF, in which substrate is engaged within the translocation channel. Cryo-EM analysis reveals that in this state the proteolytic active sites are no longer occluded, linking ATP-driven substrate translocation to full proteolytic activation. Together, these findings reveal how LonP1 prevents indiscriminate proteolysis during substrate selection by ensuring that efficient proteolysis occurs only in substrate-translocating states. Model of the conformational landscape and functional cycle of LonP1Schematic overview of LonP1 states and their inter-conversion. State transitions are modulated by substrate, nucleotide occupancy, temperature, and inhibitors. Key distinguishing features include the presence or absence of the lateral gap, nucleotide state, substrate engagement within the A-tunnel, and the handedness of the ATPase (A) domains. Additional indicators include the compactness of the proteolytic (P) domain and the presence of substrate density within the N-terminal (N) domain or at the coiled-coil domain (CCD) as well as the position of a loop within the catalytic centre. The depicted cryo-EM structures represent a model of a continuous conformational landscape and correspond to the closest matching biological states and positions within the reaction cycle, but may also capture transient intermediates or conformations stabilised by experimental conditions. The shown atomic models correspond to the states highlighted in larger font (R-state: PDB 7NGL; P1-state: PDB 7NFY; P2-state: PDB 7NGC; closed LonP1-ADP-substrate: PDB 9CC1). O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=118 SRC="FIGDIR/small/733973v1_ufig1.gif" ALT="Figure 1"> View larger version (59K): org.highwire.dtl.DTLVardef@16e0491org.highwire.dtl.DTLVardef@1ee02b1org.highwire.dtl.DTLVardef@f2b47aorg.highwire.dtl.DTLVardef@26f6b2_HPS_FORMAT_FIGEXP M_FIG C_FIG

Abstract/SummaryStresses like starvation trigger degradation of mature 40S ribosomes, requiring the coordinated breakdown of large and stable RNA-protein complexes. The atypical kinase RIOK3 orchestrates degradation by binding ubiquitylated 40S ribosomes and promoting rRNA decay. However, the mechanisms and factors that mediate rRNA decay remain unknown. Here we find that in response to starvation, RIOK3 recruits the terminal uridylyl-transferase TUT7 and the exonuclease DIS3L2 to 40S ribosomes. Sequencing analyses show that TUT7 adds oligo(uridine) tails to the 3' end of the 18S rRNA in these ribosomes. DIS3L2 subsequently recognizes uridylated 18S rRNA and carries out 3'-5' decay. We identify major decay intermediates that undergo further uridylation in a process of iterative uridylation and decay. Loss of DIS3L2 impairs 18S rRNA decay during starvation and leads to accumulation of uridylated 18S rRNA. Together these findings define a mechanism for ribosome degradation in which 3' oligo(uridine) tailing drives decay of rRNA from ribosomes.

Mitochondria are subcellular organelles responsible for energy production, and a hub for several cellular signaling pathways that ultimately control cellular processes ranging from cell death to innate immunity. Genotoxic stress, including cellular irradiation, has been shown to cause the mitochondrial-dependent activation of innate immunity via release of mitochondrial nucleic acids in the cytosol. Yet, how the other cellular events triggered by genotoxic stress affects mitochondria and mitochondrial immunity is largely unexplored. Nuclear DNA damage responses (DDR) are a set of well-described responses to genotoxic stressors that allow the cells to, through characterized mechanisms including transcriptional responses and checkpoint activation, survive or die. Whether canonical DDR directly influence mitochondrial structure and activation of downstream pathways of immunity remains unclear. Here, we identify mito-blobs: enlarged TOMM20-positive mitochondrial structures induced by genotoxic stress that are enriched for TFAM-marked mitochondrial DNA nucleoids, FASTKD2-positive mitochondrial RNA granules, and immunogenic double-stranded RNA. While structurally resembling other mitochondria stress responsive rearrangements, mito-blobs carry the distinctive feature of being induced by nuclear DNA double-stranded breaks alone. Strikingly, mitochondrial double-stranded breaks failed to induce mito-blobs, indicating nuclear-to-mitochondrial signaling rather than an autonomous response to mitochondrial genome damage. Mechanistically, we show that mito-blobs formation strictly requires MFN1/2- and OPA1-dependent fusion machinery, while nuclear DNA damage invokes a classical ATM-p53 response, which lead to cell cycle block in the G1 phase that reduce DRP1 S616 phosphorylation and shifting mitochondrial morphology toward a low-fission pro-fusion state. Strikingly, the use of the standard of care CDK4/6 inhibitor palbociclib, was sufficient to trigger mito-blobs without nuclear DNA damage. Considering the mito-blobs high content in nucleic acids, we additionally investigated if affecting their life cycle could perturb inflammatory and interferon-associated gene expression downstream of genotoxic stress. We describe how autophagic-lysosomal clearance triggered upon cell cycle block limited mito-blob persistence, and blocking autophagy disposal unmasked a strong inflammatory response. Taken together, these findings suggest mito-blobs are the product of an active mitochondrial remodeling process elicited in response to nuclear genotoxic stress and that concentrate immunogenic mitochondrial nucleic acids in defined structures for their correct disposal via autophagy and avoid aberrant activation of innate immunity. HighlightsO_LIGenotoxic stress induces enlarged mitochondrial structures called mito-blobs. C_LIO_LIMito-blobs concentrate mitochondrial DNA, RNA granules, and double-stranded RNA. C_LIO_LINuclear DNA damage, but not mitochondrial, is sufficient to induce mito-blobs. C_LIO_LICDK4 Inhibitors Trigger Mito-blob Formation. C_LIO_LIMito-blob formation and clearance alter inflammatory gene expression. C_LI eTOC blurbYu et al. identify mito-blobs as enlarged mitochondrial structures that form after genotoxic stress. They show that nuclear DNA breaks - as opposed to mitochondrial DNA breaks - trigger mito-blob formation through nuclear-to-mitochondrial signaling involving altered DRP1 phosphorylation and mitochondrial fusion machinery. Mito-blob persistence is limited by autophagic clearance and is associated with inflammatory and interferon-associated gene expression after stress.

Alongside their canonical function as adaptors in translation, tRNAs are precursors of tRNA-derived fragments that can regulate diverse aspects of gene expression. Queuosine (Q), present at position 34 of eukaryotic tRNAAsn/Asp/His/Tyr, has been implicated in suppressing tRNA fragmentation. In vertebrates, Q34 of tRNAAsp and tRNATyr is further modified by mannosylation and galactosylation, respectively, catalyzed by QTMAN and QTGAL. However, the interplay between these glycosylations and other anticodon loop modifications, and their impact on tRNA fragmentation, have remained unclear. Here, we define a modification circuit in human tRNAAsp in which Q34 stimulates DNMT2-dependent m5C38 formation, while subsequent Q34 mannosylation does not impact m5C38 installation; reciprocally, m5C38 inhibits Q34 incorporation by the tRNA-guanine transglycosylase TGT. By contrast, anticodon loop modifications of tRNATyr are installed independently, although our data support a hierarchical pathway in which TRMT5-mediated m1G37 formation precedes queuosinylation and galactosylation. Alongside demonstrating that Q34 glycosylation enhances protein synthesis, our data reveal that mannosylation of Q34 protects tRNAAsp from stress-induced cleavage, thus expanding the relevance of Q glycosylation beyond translation.

Autophagosome formation begins at phosphatidylinositol 3-phosphate-enriched endoplasmic reticulum (ER) subdomains termed omegasomes. DFCP1/ZFYVE1 is recruited to omegasomes through its FYVE domains and has recently been shown to function as an ATPase involved in omegasome constriction and autophagosome biogenesis. However, the structural basis of DFCP1 ATPase activity and how nucleotide-dependent conformational states regulate omegasome dynamics remain unknown. Here, we determined the crystal structures of the DFCP1 ATPase domain in complex with either ADP or the non-hydrolyzable ATP analogue, AppNHp at near atomic resolution. We employed structure-guided mutagenesis to define residues required for DFCP1 function in cells. The structures reveal that the active site contains a trans-acting Arg271 finger coordinating the {psi}-phosphate and trans-acting His323 that stacks on the ATP adenine ring. ATP hydrolysis results in a conformational switch in the vicinity of Arg271, while contacts with His323 persist in the presence of ADP. Biochemical analyses of DFCP1 and its mutants show that DFCP1 forms ATP-dependent microdomains on membranes. Disruption of His323 and Arg271 uncouples nucleotide binding from ATP hydrolysis and alters the oligomeric state of the protein. Live-cell imaging of DFCP1 knockout cells reconstituted with wild-type or mutant DFCP1 further demonstrates that these biochemical defects translate into distinct omegasome phenotypes. Together, our data provides a structural-function framework for DFCP1 ATPase activity and reveals how distinct catalytic elements control omegasome dynamics in vivo. We propose that nucleotide-dependent assembly and hydrolysis-driven conformational changes enable DFCP1 to regulate omegasome formation and its progression toward autophagosome closure. Significance StatementDFCP1 organizes the ER subdomains known as omegasomes, which are sites of autophagosome biogenesis. DFCP1 is an ATPase whose catalytic activity is required for function, but the precise role of its ATPase activity is unknown. Crystal structures of the DFCP1 ATPase domain in the presence of a non-hydrolyzable ATP analogue and ADP reveal an Arg finger that operates in trans, such that dimerization is required for ATP hydrolysis, and ATP hydrolysis destabilizes dimers. Dimerization is also promoted by an adenine base-stacking interaction in trans with H323. These residues are important for DFCP1 clustering on membranes and omegasome constriction, clarifying that the role of ATP is to promote the dimerization of the ATPase domain and thereby control the organization of DFCP1 on membranes.

Mitochondrial homeostasis is maintained by multiple quality control pathways, including mitophagy, which targets dysfunctional mitochondria for degradation. During receptor-mediated mitophagy, the outer membrane proteins BNIP3 and NIX directly recruit autophagy machinery to the mitochondrial surface, though their precise regulation is still unclear. In recent years, new BNIP3- and NIX-interacting proteins have been identified that influence mitophagic flux. PPTC7 and FBXL4 target BNIP3 and NIX for proteasomal turnover to keep levels of the receptors low, whereas TMEM11 is proposed to spatially control mitophagy by interacting with receptors at active mitophagy sites. However, it is unclear how each of these interactions is controlled and how they interplay with each other. Here, we identify a repressor of mitophagy, ARMC1, which forms a complex with TMEM11, BNIP3, and NIX. During mitophagy activation, ARMC1 dissociates from the complex, freeing the receptors to initiate mitophagy. We find that TMEM11 then acts in an antagonistic relationship with PPTC7, protecting the receptors from proteasomal degradation. Our data are consistent with a two-stage model. At steady state, a population of sentinel receptors is repressed and primed to respond to mitochondrial dysfunction. Once mitophagy is activated, TMEM11 protects BNIP3 and NIX, ensuring a sustained mitophagic response. Our findings provide a framework for understanding how two key regulatory pathways intersect to modulate receptor-mediated mitophagy.

The ATG8 conjugation machinery supports both canonical autophagy, in which ATG8 proteins are lipidated on forming autophagosomes, and CASM, in which ATG8 proteins are conjugated to stressed single-membrane compartments. How cells allocate this shared machinery between these competing membrane programs remains unclear. Here, we identify ATG16L2 as a specialized regulator of V-ATPase-responsive ATG8 lipidation (atg8ylation). ATG16L2 does not form stable homodimers and is inactive in the absence of ATG16L1, but it becomes functional through heterodimerization with ATG16L1. The ATG16L1-ATG16L2 heterodimer constitutes a VAIL-competent E3-like complex that is excluded from WIPI2-positive autophagic membranes but readily recruited to stressed lysosomes. ATG16L2 also confers VAIL activity onto the otherwise VAIL-deficient ATG16L1 isoform. We propose that ATG16L1 homodimers and ATG16L1-ATG16L2 heterodimers represent alternative assemblies of the ATG8 conjugation machinery: ATG16L1 homodimers are dedicated to WIPI2-positive autophagic membranes, whereas ATG16L1-ATG16L2 heterodimers containing either ATG16L1 isoform form a VAIL-selective assembly.

Mitochondrial poly-A polymerase (MTPAP) is essential for mitochondrial mRNA (mt-mRNA) polyadenylation, a critical step in mt-mRNA maturation. Mutations in MTPAP have been reported to cause a mitochondrial cytopathy. Here, we provide evidence of enhanced type I interferon (IFN) signalling in the blood of patients carrying mutations in MTPAP. Further, deletion of MTPAP in a fibroblast cell model led to abnormalities in mitochondrial respiration and mtRNA processing, and an upregulation of type I IFN signalling. Notably, both in patients fibroblast and in MTPAP-deleted cells, we observed the accumulation of non-coding mtRNA and mitochondrial double-stranded RNA (mt-dsRNA) within enlarged mitochondrial RNA granules. Cytosolic release of mt-dsRNA led to type I IFN induction mediated primarily by the RNA sensor MDA5 and its adaptor MAVS. Our findings reveal a novel consequence of MTPAP dysfunction, highlighting how impaired mtRNA maturation can drive innate immune system activation.

Lysosomal membrane damage triggers a multi-stage repair response essential for cellular homeostasis. Here we identify the oxysterol-binding protein-related protein ORP3 as a critical mediator of late-stage lysosomal membrane repair. Following lysosomal damage induced by L-leucine-leucine methyl ester (LLOME) or cationic amphiphilic drugs (CADs), ORP3 is phosphorylated and recruited to ER-lysophagosome contact sites via a signaling cascade initiated by lysosomal membrane ubiquitination, TAK1, p38 MAPK, and, to a lesser extent, IKK. p38-dependent phosphorylation promotes direct interaction between ORP3 and LC3B, which together with PI(4,5)P2 binding, is required for autophagic lysosome recruitment. ORP3 depletion impairs late-stage lysosomal recovery, elevates lysosomal lipid peroxidation, and reduces cell survival. A lipid transfer-deficient ORP3 mutant fails to restore lysosome function despite normal recruitment, indicating that ER-to-lysophagosome transfer of phosphatidylcholine by ORP3 is functionally required. ORP3 activity is subsequently terminated by VCP/p97-mediated deubiquitination of lysosomes. These findings define ORP3 as a MAPK regulated lipid transfer protein during the late autophagic phase of the endolysosomal damage response. SummaryLysosomal membrane damage triggers ubiquitination that activates a TAK1-p38 signaling cascade, phosphorylating the lipid transfer protein ORP3 and recruiting it to damaged lysosomes via LC3B interaction. ORP3-mediated phosphatidylcholine transfer from the ER is essential for late-stage lysosomal repair and cell survival. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=121 SRC="FIGDIR/small/731146v1_ufig1.gif" ALT="Figure 1"> View larger version (30K): org.highwire.dtl.DTLVardef@49f300org.highwire.dtl.DTLVardef@f0a90dorg.highwire.dtl.DTLVardef@1eaa560org.highwire.dtl.DTLVardef@f4de4_HPS_FORMAT_FIGEXP M_FIG C_FIG

ESCRT proteins remodel membranes at many cell organelles, including the endoplasmic reticulum (ER). Here, we investigate whether ESCRTs in budding yeast participate in stress-induced ER reorganisation. We find that ER stress triggers the formation of tubular ER subdomains that recruit various ESCRT proteins. Recruitment of the major ESCRT-III protein Snf7 is mediated by the ESCRT-associated protein Bro1, a homologue of human ALIX, in a manner that is mechanistically distinct from Bro1 function at endosomes. ESCRT-containing ER subdomains are derived from ceramide-rich ER exit sites and form contacts with the Golgi. Furthermore, Bro1 helps to concentrate the tethering and lipid transfer protein Tcb3, a homologue of human extended synaptotagmins, at these organelle contacts and contributes to cellular fitness when lipid metabolism is perturbed. These results indicate that specialised ER exit sites can be repurposed for contacting the Golgi directly and uncover ESCRTs as organisers of stress-inducible ER-Golgi contacts that help maintain cell homeostasis.

Mitochondrial {beta}-barrel proteins fulfill essential functions for cell viability including transport of proteins, ions and metabolites across the outer membrane. The translocase of the outer membrane (TOM complex) and the sorting and assembly machinery (SAM complex) transport these proteins into the outer membrane. Although {beta}-barrel proteins are essential for cell viability, the quality control mechanisms that monitor their import into the outer membrane remain unknown. We uncovered that Msp1 and Ubx2-recruited Cdc48 cooperate to remove a SAM-arrested Tom40 variant. Prior to degradation, the {beta}-barrel protein is deposited in cytosolic protein storages, termed MitoStores, to minimize accumulation of aberrant proteins in the cell. The cytosolic AAA ATPase Hsp104 disaggregates these proteins to facilitate their proteasomal degradation. Altogether, mitochondrial and cytosolic factors cooperate to degrade translocation-arrested mitochondrial {beta}-barrel proteins.

The microtubule plus-end tracking proteins (+TIPs) CLIP-170/Bik1 and EB/Bim1 form a condensate, the +TIP body, at the plus-end of most microtubules in vivo. Remarkably, however, these +TIP bodies typically impart different dynamics and interaction profiles to distinct microtubules, according to their cellular function. The molecular mechanisms underlying the functional versatility of the +TIP body are unknown. Here, we show that the +TIP Kar9 utilizes repeats of a lysine-aspartate-lysine (KDK)-centered short linear motif (SLiM) to interact with Bik1 on a restricted subset of cytoplasmic microtubules during yeast mitosis. Furthermore, these multivalent Kar9-Bik1 interactions tune the material behavior of the +TIP body to specify proper microtubule function. Indicating that KDK serves as generic Bik1-interaction motif, similar motifs are also present in Kip2, where they mediate Bik1-dependent recruitment of Kip2 to the +TIP body. Together, our study provides insights into how low-affinity Bik1 interactors diversify microtubule function by locally specializing the content and behavior of +TIP bodies.

During meiotic prophase I, the chromosome axis orchestrates programmed DNA double-strand break formation, repair and synapsis between homologous chromosomes. In mammals, the axis is assembled from the coiled-coil elements SYCP2 and SYCP3 that come together with the HORMA-domain proteins HORMAD1 and HORMAD2, but how these components associate into a coherent structural unit remains incompletely understood. Combining recombinant reconstitution, mass photometry, SEC-MALS, AlphaFold modelling and crosslinking mass spectrometry, we show that the HORMA domains of HORMAD1 and HORMAD2 form a selective pseudosymmetric heterodimer independently of either proteins own closure motif, with interface determinants conserved across vertebrates. We identify a previously unrecognised second closure motif (CM2) in SYCP2 that preferentially binds HORMAD1, distinct from the previously described HORMAD2-binding closure motif (CM1). Together, these results revise the current model of mammalian axis assembly and define a tripartite SYCP2-HORMAD1-HORMAD2 module as a fundamental structural unit of the mammalian meiotic chromosome axis.

Abstract/SummaryPrecise regulation of enzyme recruitment during Okazaki fragment maturation (OFM) is essential for faithful and efficient lagging-strand DNA synthesis. Emerging evidence suggests that PARP1 contributes to OFM yet its specific functions remain unclear. Here, we define context-dependent functions of PARP1 during OFM. Under physiological conditions, PARP1 co-localizes with PCNA in early S phase and restrains Pol {delta}-PCNA- mediated strand-displacement DNA synthesis, thereby preventing the formation of long 5' flaps, which is refractory to FEN1 cleavage. On the other hand, in LIG1-deficient cells, in which DNA nicks and unexpectedly long 5' flaps accumulate, PARP1 promotes the recruitment of LIG3 to catalyze OF ligation and DNA2 to facilitate long 5' flap processing. Collectively, our findings uncover previously unrecognized roles of PARP1 in regulating 5' flap dynamics to ensure efficient OFM and cell viability. HighlightsO_LIPARP1 plays context-dependent regulatory functions in Okazaki fragment maturation (OFM). C_LIO_LIPARP1 controls strand displacement DNA synthesis by the PCNA-Pol{delta} complex to dictate generation of short over long RNA-DNA flaps during canonic OFM. C_LIO_LIPARP1 senses unligated Okazaki fragments in DNA Ligase 1 deficient cells and suppresses unwanted conversion of DNA nicks into 5 flaps. C_LIO_LIProcessing of unligated nicks or flaps by DNA ligase 3 or DNA2, respectively in LIG1 deficient cells depends on PARP1. C_LIO_LIPARP1 inhibitors induce synthetic lethality with DNA ligase 1 or DNA2 inhibition. C_LI

N{superscript 1}-methyladenosine at position 58 (m{superscript 1}A58), installed by the TRMT6/TRMT61A complex, is a highly conserved structural modification implicated in tRNA stability and human disease. In yeast, loss of m{superscript 1}A58 disrupts initiator methionine tRNA (tRNA) metabolism through defects in precursor processing and rapid tRNA decay, but whether these mechanisms operate in human cells has remained unclear. Here, using acute degron-mediated depletion of TRMT6 in human HCT116 cells combined with precursor-resolved tRNA sequencing, we show that m{superscript 1}A58 controls two distinct steps of tRNA metabolism. TRMT6 depletion selectively reduces mature tRNA while causing accumulation of precursor species retaining both 5'-leader and 3'-trailer sequences, consistent with impaired maturation. In parallel, the nuclear 5'[->]3' exonuclease XRN2 selectively degrades the residual mature hypomodified tRNA pool without affecting precursor accumulation. Loss of mature tRNA activates the integrated stress response and impairs proliferation, which is restored by XRN2 co-depletion. Together, our findings identify m{superscript 1}A58 as a coordinator of human initiator-tRNA maturation and surveillance, establishing a two-step pathway that maintains tRNA homeostasis in human cells.

Microtubules are polarised polymers that assemble into highly specialised networks in a cell-specific manner. This is controlled in part by microtubule organising centres (MTOCs), which concentrate factors necessary for microtubule nucleation and the organisation of microtubule minus ends. Neurons rely on oppositely polarised microtubule networks, with axons containing mostly plus-end-out microtubules, and dendrites contain many minus-end-out microtubules. How minus-end-out microtubule polarity is established in dendrites remains an important question. Here, we identify a new type of MTOC within the dendrites of Drosophila class I dendritic arborisation neurons, a common model for the neuronal cytoskeleton. We show that membrane swellings distributed intermittently along dendrite shafts, which we term "dendritic varicosities", contain the principal component of the microtubule nucleating complex and repeatedly generate microtubules whose plus ends grow back towards the soma. Varicosities located specifically in distal regions also contain MTOC proteins implicated in minus end anchoring, and this correlates with the accumulation of minus ends specifically in distal varicosities. Depletion of these MTOC proteins leads to major defects in minus end organisation, with microtubule buckles and loops deforming the neuronal membrane. Thus, dendritic varicosities are an important new type of neuronal MTOC that contribute to the generation and organisation of the minus-end-out microtubule network within dendrites.