Compromised transcription-mRNA export factor THOC2 causes R-loop accumulation, DNA damage and adverse neurodevelopment

Bhattacharjee, Rudrarup; Jolly, Lachlan A.; Corbett, Mark A.; Wee, Ing Chee; Rao, Sushma R.; Gardner, Alison E.; Ritchie, Tarin; van Hugte, Eline J. H.; Ciptasari, Ummi; Piltz, Sandra; Noll, Jacqueline E.; Nazri, Nazzmer; van Eyk, Clare L.; White, Melissa; Fornarino, Dani; Poulton, Cathryn; Baynam, Gareth; Collins-Praino, Lyndsey E.; Snel, Marten F.; Nadif Kasri, Nael; Hemsley, Kim M.; Thomas, Paul Q.; Kumar, Raman; Gecz, Jozef

doi:10.1038/s41467-024-45121-5

Download PDF

Article
Open access
Published: 08 February 2024

Compromised transcription-mRNA export factor THOC2 causes R-loop accumulation, DNA damage and adverse neurodevelopment

Nature Communications volume 15, Article number: 1210 (2024) Cite this article

2038 Accesses
15 Altmetric
Metrics details

Subjects

Abstract

We implicated the X-chromosome THOC2 gene, which encodes the largest subunit of the highly-conserved TREX (Transcription-Export) complex, in a clinically complex neurodevelopmental disorder with intellectual disability as the core phenotype. To study the molecular pathology of this essential eukaryotic gene, we generated a mouse model based on a hypomorphic Thoc2 exon 37–38 deletion variant of a patient with ID, speech delay, hypotonia, and microcephaly. The Thoc2 exon 37–38 deletion male (Thoc2^Δ/Y) mice recapitulate the core phenotypes of THOC2 syndrome including smaller size and weight, and significant deficits in spatial learning, working memory and sensorimotor functions. The Thoc2^Δ/Y mouse brain development is significantly impacted by compromised THOC2/TREX function resulting in R-loop accumulation, DNA damage and consequent cell death. Overall, we suggest that perturbed R-loop homeostasis, in stem cells and/or differentiated cells in mice and the patient, and DNA damage-associated functional alterations are at the root of THOC2 syndrome.

Single-cell long-read sequencing-based mapping reveals specialized splicing patterns in developing and adult mouse and human brain

Article Open access 09 April 2024

Antisense oligonucleotide therapeutic approach for Timothy syndrome

Article Open access 24 April 2024

Messenger RNA transport on lysosomal vesicles maintains axonal mitochondrial homeostasis and prevents axonal degeneration

Article Open access 10 April 2024

Introduction

Human neurodevelopment is a highly orchestrated and complex process. The identification of naturally occurring gene variants leading to neurodevelopmental disorders (NDD) highlights the involvement of over 2300 different genes (https://www.omim.org)¹ and a myriad of cellular, molecular, and developmental mechanisms, only a few of which are being successfully targeted with precision therapies². Many essential biological processes are compromised by such NDD gene variants and often knockout of these genes cause embryonic lethality. One such example is variants in the X-chromosome gene THOC2 that cause NDDs with clinically heterogeneous presentations, now we call THOC2 syndrome^3,4,5. THOC2, a highly constrained gene, encodes the largest subunit of the THO subcomplex in the highly-conserved Transcription-Export (TREX) complex. The TREX complex is essential for transcription, mRNA processing and export, preventing DNA damage, and maintaining ESC self-renewal, pluripotency and differentiation during embryogenesis^6,7.

The TREX complex is composed of a THO sub-complex (THOC1-3, 5–7) and eight accessory proteins (UAP56/DDX39B, ALYREF/THOC4, UIF, CHTOP/SRAG, CIP29/SARNP, POLDIP3, ZC3H11A and LUZP4)^7,8,9. THOC2 is a scaffold protein critical for TREX assembly and function^10,11. The functional TREX is co-transcriptionally loaded onto the mRNAs and transfers the mature transcripts to the export adapters NXF1-NXT1 via ALYREF, eventually exporting the mRNAs to cytoplasm through the nuclear pores^7,12,13. In addition to its mRNA export function, the TREX complex has emerged as a multifunctional molecular machine with critical roles in transcription, 3’ mRNA processing, stress responses, mitotic progression, and genome stability^{7,14,15,16,17,18,19,20}.

Many studies in lower organisms have illuminated the essential roles of THOC2 in cellular and developmental processes. For example, yeast lacking Tho2 (mammalian THOC2 homolog) were cell growth inhibited¹⁰. C. elegans thoc2 knockouts were slow-growing, sterile, had defects in specific sensory neurons and died prematurely¹⁸. In Drosophila, THO2 depletion in S2 cells inhibited export of mRNA encoding heat shock protein (Hsp) and cell proliferation²¹. Zebrafish Thoc2 is essential for embryonic development²². In mice, studies in embryonic stem cells (mESC) show Thoc2 and Thoc5 regulate self-renewal by selectively binding and regulating the export of mRNAs encoding the Yamanaka factors (Oct3/4, Sox2, Klf4 and c-Myc)⁶. However, in vivo Thoc2 studies in mammalian models have been non-existent likely because of apparent embryonic lethality of knockout lines (International Mouse Phenotyping Consortium (IMPC, http://www.mousephenotype.org)²³. Indeed, complete THOC2 depletion causes severe nuclear mRNA retention in mammalian cells²⁴. Whilst these data highlight the essential role of THOC2 in cell and organism function, its involvement in human development and disease has also expanded. For example, a de novo translocation resulting in a PTK2-THOC2 gene fusion with THOC2 expression knockdown was implicated in cognitive impairment and cerebellar hypoplasia²⁵. Cytoplasmic aggregation of THOC2/THO proteins impairs nucleo-cytoplasmic transport and may be a common process explaining the pathology of neurodegenerative diseases, including Huntington’s, Alzheimer’s and amyotrophic lateral sclerosis²⁶. These findings, along with published^3,4,5 missense, microdeletion or splicing-defective THOC2 NDD variants suggest that THOC2 performs multiple critical roles with major impacts on neuronal function. We premised that the brain is sensitive to THOC2 function given its heightened requirements for and diversity of transcription and protein synthesis. However, the role of THOC2 and TREX complex in normal brain development and, importantly, in causing NDDs remains elusive largely due to our inability to create relevant in vivo models of this essential gene. So, the broader functional understanding of THOC2 in mammalian development remains unknown. Thus, we sought to generate a preclinical mouse model to delve deeper into the unresolved aspects of THOC2 biology with a view to understand the molecular and cellular roles in brain development as well as mechanism of the disease pathology in individuals with THOC2 pathogenic variants.

We reasoned that naturally occurring hypomorphic (i.e., partial loss of function) variants in THOC2 patients with NDDs and thus compatible with life offer the opportunity for in vivo investigations by overcoming lethality. We utilized one such hypomorphic THOC2 variant (NC_000023.11:g.123609724_123612089del, ClinVar: VCV000804360.1) we reported in a patient with intellectual disability (ID), speech problems, short stature, low body weight, sensorineural hearing loss and microcephaly⁵ to generate a mouse model with a microdeletion encompassing Exons 37 and 38. We show that Thoc2 is critical during early post-fertilisation development, but mice can survive into adulthood. Neurobehavioral testing shows that the Thoc2 exon 37–38 deletion male mice (called Thoc2^Δ/Y) have learning, memory, and sensorimotor deficits, recapitulating phenotypes observed in most of the patients with THOC2 syndrome. Our transcriptomic and proteomic data shows major dysregulation in pathways associated with transcription, splicing and higher order cognitive functions. We show Thoc2^Δ/Y mice have a reduced neural stem cell (NSC) pool, altered brain cytoarchitecture, and multiple impairments of neuron structure and function that lead to significantly compromised electrophysiological properties and synapse formation. However, to our surprise, we observe that mRNA export is not overtly affected in the Thoc2^Δ/Y mice, but instead results in accumulation of R-loops and concomitant DNA damage. This mechanism is also evident in THOC2 exon 37–38 deletion patient-derived dermal fibroblasts. Together, our data reveal a mechanism of THOC2 NDD originating early in development and affecting cell’s ability to resolve R-loops and consequently preventing DNA damage and causing cell death.

Results

THOC2 gene essentiality in eukaryotes²⁷, and absence of a knockout mammalian model suggest that its complete loss is incompatible with life²³. However, we have previously reported 22 THOC2 (19 missense, two splicing defective and one microdeletion) variants, 12 of which are hypomorphic and functionally validated in individuals with intellectual disability (ID) and variable expressivity of other traits including hypotonia, growth restriction and microcephaly^3,4,5. We postulated that these naturally hypomorphic THOC2 alleles could be utilized for generating a viable mouse model to gain insights into the THOC2/TREX biology as well as understanding the pathology of THOC2 syndrome.

Generation of a viable mouse model using a hypomorphic Thoc2 exon 37-38 deletion variant

Our initial attempts to generate knock-in mice carrying either the p.Leu438Pro or p.Ile800Thr THOC2 variant, that reduce the stability of THOC2 protein (and other THO subunits) in patient-derived cell lines³, failed to produce any live offspring (see Supplementary Data 1). Subsequently, we generated a mouse model based on a pathogenic deletion that includes exons 37 and 38 that code for the extreme COOH-terminus of the THOC2 protein using CRISPR/Cas9 editing (Fig. 1a). Four founder heterozygous females (Δ/+) and two hemizygous males with exon 37-38 deletion (Δ/Y) were recovered (Supplementary Fig. 1a, b). We confirmed the deletion of genomic sequences coding for Thoc2 exon 37-38 and absence of any off-target CRISPR events by whole genome sequencing. We further eliminated the off-target events by backcrossing the Thoc2^Δ/+ with WT C57BL/6JArc male mice for three generations. We confirmed the deletion of exon 37-38 sequences from the Thoc2 mRNA in 30 days old Thoc2^Δ/Y mice brain using RT-PCR and Sanger sequencing (Fig. 1b). Exon 37 and 38 deletion in the Thoc2 mRNA is in-frame, not susceptible to non-sense mediated decay (NMD) and predicted to translate into a C-terminally truncated ~172 kDa compared to normal ~175 kDa THOC2 protein (Supplementary Fig. 1c). We analysed the adult Thoc2^Δ/Y mice brain lysates by western blotting (WB) with two antibodies capable of distinguishing the full length and the truncated THOC2 proteins (Fig. 1c). As expected, we observed a slightly smaller THOC2 protein in the Thoc2^Δ/Y compared to the full-length protein in the Thoc2^+/Y mouse brains. The levels of truncated THOC2 protein were increased in Thoc2^Δ/Y mice brains (Fig. 1c) as also observed in the THOC2 exon 37-38 deletion patient fibroblasts and lymphoblastoid cell line (LCL)⁵. The Thoc2^Δ/+ female mouse brains also showed increased level of THOC2 protein (Supplementary Fig. 1d, lane 2). As expected, only the truncated THOC2 protein was present in the Thoc2^Δ/Δ female mouse brains (Supplementary Fig. 1d, lane 3).

**Fig. 1: Generation of *Thoc2* exon 37-38 deletion mouse model.**

While breeding animals for experimental work, we recovered ~15–16% Thoc2^Δ/Y animals (both embryos and postnatal) compared to the expected Mendelian genotypic ratio of 25% (Fig. 1d). We predicted reduced recovery of Thoc2^Δ/Y mice was likely due to early post fertilisation death. We investigated this by performing in vitro fertilization (IVF) with Thoc2^Δ/+ mouse oocytes and Thoc2^+/Y sperms and scoring the blastocysts’ for Thoc2^+/Y and Thoc2^Δ/Y genotypes, using our rapid gDNA extraction-multiplex PCR method, 4 days after the IVF procedure (Fig. 1e). We found that only ~14% blastocysts were Thoc2^Δ/Y (Fig. 1e), indicating that reduced recovery of the Thoc2^Δ/Y mice was due to the death of zygotes that are much before the blastocyst stage. THOC2 has higher expression at the blastocyst stage⁶. A recent report showed that THO subunits have higher expression during embryonic stages (E10.5 to E18.5) that declines sharply from postnatal day 3 (P3) onwards²⁸. We asked if THOC2 and other TREX proteins also show a similar trend and analysed the levels of THOC2 protein in the brains of Thoc2^+/Y and Thoc2^Δ/Y mice at embryonic (E14.5 and E18.5) and postnatal (P0, P10, P30 and P60) stages. We found that the THOC2 protein levels are high at the embryonic stages and decline sharply from postnatal day 10 onwards (Fig. 1f). This developmental decline of THOC2 and other TREX subunit proteins in the brains of Thoc2^+/Y and Thoc2^Δ/Y mice (Supplementary Fig. 1e) agreed with the publicly available²⁸ and our E14.5, E18.5 and P10 mouse brain RNA-seq data (Supplementary Fig. 1f). These results support a crucial role for Thoc2 in early development, mostly post fertilisation and at a stage of pluripotent embryonic stem cells.

We observed that the Thoc2^Δ/Y mice (embryo and adult) have significantly reduced weight (~15%) and smaller size compared to their Thoc2^+/Y littermates (Fig. 1g, h), a phenotype similar to the patient with the THOC2 exon 37-38 deletion variant (see Supplementary Data 2). The vital organs (e.g., brain, heart, lung, kidney, and liver) of the Thoc2^Δ/Y mice were smaller and lighter as compared to the Thoc2^+/Y mice, suggesting a proportional global reduction in their body size and weight (Supplementary Fig. 1g). The P30 Thoc2^Δ/Y mice showed a distinct shorter and flatter snout compared to the Thoc2^+/Y mice (Supplementary Fig. 1h). Finally, we did not observe difference in the weight of 30–120 days old Thoc2^+/+ and Thoc2^Δ/+ female mice (Supplementary Fig. 1i).

Together, our data showed that we have generated a viable Thoc2 exon 37-38 deletion mouse model that recapitulated smaller body size, presence of exon 37-38 deleted mRNA and C-terminally truncated protein as also observed in the THOC2 exon 37-38 deletion patient⁵.

Thoc2 ^Δ/Y mice have learning, memory and sensorimotor deficits

As the mouse model is based on a THOC2 pathogenic NDD variant, we subjected eight-week-old Thoc2^Δ/Y mice and Thoc2^+/Y littermates to a battery of neurobehavioral tests to assess cognition, motor coordination, sensorimotor function, gross and fine motor skill, and anxiety-like behaviour (Fig. 2 and Supplementary Fig. 2).

**Fig. 2: *Thoc2*^Δ/Y mice display multiple neurobehavioral deficits.**

Cognitive function assessment

We utilized the Morris Water Maze²⁹ and Barnes Maze³⁰ to assess spatial learning and memory and the Y-maze test³¹ to assess spatial working and reference memory. The Morris Water Maze results indicated significant learning deficits in Thoc2^Δ/Y mice, as these mice took longer time to locate the escape platform compared to their Thoc2^+/Y littermates, even after 5 days of “learning trials” (Fig. 2a). The Thoc2^Δ/Y mice also showed significantly fewer visits to the target quadrant and escape platform and spent much less time in the target quadrant or on the escape platform compared to their Thoc2^+/Y littermates, suggesting that they had never fully learned the task (Supplementary Fig. 2a). Consistent with these results, Thoc2^Δ/Y mice also showed significantly higher escape latencies than the Thoc2^+/Y mice on both the second and third day of “learning trials” on the Barnes Maze, as well as on probe trials (Fig. 2b). On probe trials, Thoc2^Δ/Y mice also made fewer visits to the old escape box location and spent much less time in the quadrant of the old escape box, providing additional support for the presence of spatial learning deficits in Thoc2^Δ/Y mice (Supplementary Fig. 2b). Interestingly, however, both the Thoc2^+/Y and Thoc2^Δ/Y mice showed comparable performance in finding the new escape box, indicating cognitive flexibility is intact in Thoc2^Δ/Y mice (Supplementary Fig. 2b). Given the significant learning deficits in Thoc2^Δ/Y mice, we chose to further probe spatial working and reference memory using the Y-maze task. Thoc2^Δ/Y mice showed a significantly reduced percentage of alternations between the maze arms compared to their Thoc2^+/Y littermates (Fig. 2c), indicating an impairment of working memory.

Motor function and sensorimotor control assessment

The motor coordination and gait analysis from Rotarod³² and Digigait³³ tests, respectively, did not show any difference between the Thoc2^Δ/Y and Thoc2^+/Y mice, suggesting no impairment in either gross motor coordination or overall gait (i.e., swing time, brake time) (Supplementary Fig. 2c, d). It is important to note, however, that Thoc2^Δ/Y mice showed a significant decrease in stride length in the hind-paws, as well as a concomitant decrease in propel time of the forepaws (Supplementary Fig. 2d), which could indicate gait asymmetries. Interestingly, in the open field test³⁴, Thoc2^Δ/Y mice travelled a significantly greater total distance as compared to their Thoc2^+/Y littermates, indicating a potentially hyperactive locomotor phenotype in the Thoc2^Δ/Y mice (Fig. 2d). Similarly, the Thoc2^Δ/Y mice showed an increased distance travelled compared to their Thoc2^+/Y littermates during both the elevated plus maze task³⁴ (Supplementary Fig. 2e) and the light-dark test³⁵ (Supplementary Fig. 2f). Thoc2^Δ/Y mice also had higher number of entries into all arms on the elevated plus maze task (Supplementary Fig. 2e) and made a greater number of transitions in the light-dark test (Supplementary Fig. 2f). However, there was no difference between the groups for the time spent in either the outer or inner zone of the open field during task performance (Fig. 2d).

Next, we tested fine motor coordination in Thoc2^Δ/Y mice using the beam walking test³⁶. We found that the Thoc2^Δ/Y mice showed a significantly higher number of foot slips while crossing the beam, suggesting that these mice had more difficulty in maintaining their balance and fine motor control during task performance (Fig. 2e). We then probed fine motor control using the pasta handling task³⁷. Interestingly, the Thoc2^Δ/Y mice took significantly longer time to consume one piece of pasta as compared to Thoc2^+/Y littermates (Fig. 2f). This was associated with more frequent drops of the pasta, as well as more frequent adjustments to grip when holding and eating the pasta, collectively considered as ‘atypical behaviours’ (Fig. 2f). Similarly, Thoc2^Δ/Y mice showed significantly increased time to contact and remove the tape when compared to their Thoc2^+/Y littermates in the adhesive removal test³⁸ (Fig. 2g). Importantly, however, this was not associated with gross changes in grip strength (Supplementary Fig. 2g). Taken together, these results suggest that impairments observed are more indicative of deficits in fine motor control and sensorimotor function than skeletal muscle alterations.

Anxiety assessment

Surprisingly, on the elevated plus maze³⁴, the Thoc2^Δ/Y mice showed a decrease in anxiety-like behaviour, as evidenced by increased time spent in open arm and centre of the maze and more frequent entries into the open arm, with a concomitant decrease in time spent in the closed arm (Supplementary Fig. 2e). Corroborating these observations, the Thoc2^Δ/Y mice also spent more time in the light compartment and less time in the dark compartment during the light-dark test and took less time to emerge from the dark chamber into the light chamber, as compared to their Thoc2^+/Y littermates (Supplementary Fig. 2f).

Taken together, the neurobehavioral studies indicate impairments in hippocampus-dependent spatial learning and memory, deficits in fine motor control and sensorimotor function, hyperactivity, and reduced anxiety in Thoc2^Δ/Y mice, phenotypes that are similar to the THOC2 exon 37-38 deletion patient (Supplementary Fig. 2h).

Thoc2 ^Δ/Y embryonic and adult mice brain show significant dysregulation in genes associated with transcription, cell cycle, cell death and cognitive function

As the TREX/THO plays a conserved and central role in the process of transcription and subsequent export of mRNAs^7,15,19, we asked if the truncated THOC2 protein would cause specific transcriptional and/or mRNA export perturbations. Surprisingly, we did not observe any overt nuclear retention of mRNAs in Thoc2^Δ/Y cultured primary neurons as well as THOC2 exon 37-38 deletion patient fibroblasts (Supplementary Fig. 3a). We investigated if there was any transcriptomic dysregulation by performing RNA-seq of embryonic (E14.5, E18.5) and post-natal day 10 (P10) Thoc2^Δ/Y and Thoc2^+/Y mouse brain total RNAs (specifically cortex and hippocampus; functionally compromised regions as indicated by the behavioural outcomes). We observed significant transcriptional dysregulation in Thoc2^Δ/Y brains at all three stages with the largest effect at E18.5 (658 dysregulated genes at FDR < 0.05 and 1299 genes at p < 0.01) (Fig. 3a–c and Supplementary Data 3). As there were only 15 significantly dysregulated genes at E14.5 (FDR < 0.05), we opted to use a cut off criteria of p < 0.01 at all-time points for rank-based gene ontology enrichment analysis using g:Profiler³⁹. The differentially expressed genes (with p < 0.01 cut off) in Thoc2^Δ/Y E14.5 brains were enriched in processes like transcription, histone methylation and embryonic development among others (Fig. 3d and Supplementary Fig. 3b). However, GO analysis of E18.5 dysregulated transcriptome showed enrichment of genes in processes of apoptosis, cell cycle regulation, response to DNA damage, learning and memory as well as RNA Polymerase II associated transcription (Fig. 3e and Supplementary Fig. 3c). Finally, the dysregulated genes in Thoc2^Δ/Y P10 brains showed GO term enrichment for RNA Polymerase II associated transcription, mitosis, and cell death; the terms also shared with the E14.5 and E18.5 stages (Fig. 3f and Supplementary Fig. 3d). Interestingly, however, the dysregulated genes from only E18.5 and P10 showed an enrichment for genes with higher GC content compared to all expressed genes in the mouse genome (Fig. 3g).

**Fig. 3: *Thoc2*^Δ/Y embryonic and adult mice brains show significantly dysregulated transcriptome.**

We found that the dysregulated genes were significantly over represented in a list of 1571 known NDD genes (SysNDD)⁴⁰ at all developmental stages; 29/225 in E14.5 (p < 0.001), 185/1299 in E18.5 (p < 1.9e–20) and 56/618 in P10 (p < 0.04) of Thoc2^Δ/Y brains (the p values represent hypergeometric test significance) (Supplementary Fig. 3e and Supplementary Data 3). We utilized mammalian phenotype ontology developed by Mouse Genome Informatics (MGI) and Knockout Mouse Phenotyping Program 2 (KOMP2) to further analyse the ontological enrichment of significantly dysregulated genes in E14.5, E18.5 and P10 Thoc2^Δ/Y mice brains. The dysregulated genes in E14.5 were enriched for genes associated with phenotypes such as completely penetrant pre-weaning lethality, abnormal lung and eye morphology, abnormal embryo size, decreased bone mass and body length and reduced fertility (Supplementary Fig. 3f). Similarly, E18.5 dysregulated genes showed enrichment for genes associated with phenotypes such as embryonic growth retardation and lethality during foetal growth, pre-weaning lethality, spermatogenesis defect, and decreased body weight among others (Supplementary Fig. 3g). Finally, the P10 dysregulated genes were enriched for processes like myelin sheath morphology, oligodendrocyte physiology, and abnormal Schwann cell morphology along with phenotypes such as decreased body weight, embryonic growth retardation and incompletely penetrant pre-weaning lethality (Supplementary Fig. 3h).

Together, our developmental transcriptome data highlighted that the THOC2 exon 37-38 truncated protein has a major impact on transcription, cell death, and cell cycle in Thoc2^Δ/Y mouse brains at all three developmental stages, with the E18.5 stage being the most affected. The phenotypic ontology terms, embryonic lethality, reduction in body weight along with learning and memory impairment are consistent with our IVF and neurobehavioral testing data.

Thoc2 ^Δ/Y E18.5 and P10 mouse brains show significant dysregulation of proteins involved in mRNA processing, synapse organization and cognitive processes

With significant transcriptional dysregulation in the Thoc2^Δ/Y E14.5, E18.5 and P10 brains, we set out to explore the impact of truncated THOC2 protein on the Thoc2^Δ/Y brain proteome using quantitative liquid chromatography coupled to tandem mass spectrometry analysis (LC-MS/MS). We observed significant protein alterations in both E18.5 (644 proteins) and P10 (396 proteins) Thoc2^Δ/Y mice brains with larger proportion of proteins down- (E18.5: 75% and P10: 67%) than up-regulated (E18.5: 25% and P10: 33%) (Fig. 4a, b and Supplementary Data 4). Gene ontology analysis showed that significantly dysregulated proteins at E18.5 were enriched for processes such as gene expression regulation, splicing and synapse organization among others (Fig. 4c, d and Supplementary Fig. 4a, b). At P10 stage, the dysregulated proteins were enriched for cellular functions involving spliceosome, mRNA surveillance, pre- and post-synapse organization, and cell junction organization (Fig. 4e, f and Supplementary Fig. 4c, d). Further ontological analysis of proteins significantly dysregulated in both the E18.5 and P10 Thoc2^Δ/Y mouse brains (using MGI mammalian phenotype and KOMP2 mouse phenotype as the reference databases) revealed an enrichment for processes involved in pre-weaning and embryonic lethality and abnormal behaviours including hyperactivity and spatial learning and memory among others (Fig. 4g–j). Strikingly, 112 dysregulated proteins in E18.5 and 63 in P10 brains were coded by the known NDD genes (Supplementary Fig. 4e, f and Supplementary Data 4)⁴⁰. Furthermore, genes encoding the dysregulated proteins in E18.5 and P10 Thoc2^Δ/Y mouse brain were enriched for higher GC content (Fig. 4k), a finding similar to that from our RNA-seq data (see Fig. 3g).

**Fig. 4: *Thoc2*^Δ/Y embryonic and adult mice brains show significantly dysregulated proteome.**

We also compared the changes in expression of proteins from E18.5 to P10 in the Thoc2^Δ/Y and Thoc2^+/Y mouse brains. Whereas the majority (1581, FDR < 0.05) of proteins showed similar trend of change at these two stages, 421 and 688 proteins behaved differently in Thoc2^Δ/Y and Thoc2^+/Y mouse brains, respectively (Supplementary Fig. 4g). Gene ontology analysis of the 688 unique Thoc2^+/Y proteins showed enrichment mainly for biological processes like protein dephosphorylation, inner mitochondrial membrane organization, cytokinesis, chemical synaptic transmission, and cytosolic transport (Supplementary Fig. 4h). However, 421 unique Thoc2^Δ/Y proteins showed enrichment for processes like translation, peptide biosynthesis and protein metabolism, gene expression and mRNA catabolic processes (Supplementary Fig. 4i), suggesting that proteins altered in the Thoc2^Δ/Y are associated with processes different to those in the Thoc2^+/Y mouse brain. Taken together, both embryonic and postnatal stages of Thoc2^Δ/Y mouse brains showed significant dysregulation of proteins enriched for pathways consistent with our neurobehavioral and transcriptome data.

Thoc2 ^Δ/Y E18.5 mice embryonic cortex shows reduction in ventricular zone thickness

As the E18.5 stage showed a major transcriptomic dysregulation and is a critical developmental stage representing the completion of neurogenesis in cortical development⁴¹ we performed histological studies on the Thoc2^+/Y and Thoc2^Δ/Y E18.5 embryonic brains. While haematoxylin and eosin (H&E) staining did not reveal any gross histological alterations (Supplementary Fig. 5a, b), a closer inspection of the cortical layers showed a significant reduction in the thickness of the ventricular zone (VZ) of the E18.5 Thoc2^Δ/Y mouse brain (Fig. 5a). We validated these observations by VZ NSC marker PAX6 immunofluorescence staining of E14.5 and E18.5 Thoc2^+/Y and Thoc2^Δ/Y mouse brains. We observed a significantly reduced PAX6 layer thickness in both the E14.5 and E18.5 Thoc2^Δ/Y mouse brain (Fig. 5b, c), confirming a reduction in the VZ size. We also performed H&E staining of Thoc2^+/Y and Thoc2^Δ/Y mouse brains at P30 and did not find any gross morphology alterations in the Thoc2^Δ/Y mouse brains compared to Thoc2^+/Y. However, a closer look at the cytoarchitecture of the cortices showed a significantly reduced overall thickness of the cortical plate and corpus callosum (Fig. 5d, e). Finally, we observed no change in the hippocampus structure or volume (Supplementary Fig. 5c, d). Overall, we found minor, but significant anatomical differences in the VZ of the E18.5 Thoc2^Δ/Y mouse brains suggesting compromised function of the NSCs.

**Fig. 5: Altered cytoarchitecture in *Thoc2*^Δ/Y mouse brain.**

Thoc2 ^Δ/Y primary NSCs have increased apoptosis, perturbed cell cycle properties and DNA damage

The VZ is the primary proliferative zone that harbors the NSCs⁴². NSC malfunction is frequently involved in NDDs such as primary microcephaly, autism spectrum disorders (ASD) or X-linked ID^43,44. Reduced VZ thickness and RNA-seq data showing transcriptionally dysregulated genes enriching for cell cycle and apoptotic pathways in E18.5 Thoc2^Δ/Y brains prompted us to investigate if the NSCs in these mice have altered cell death and/or cell cycle properties. To this end, we analysed the cell cycle kinetics of the primary NSCs from E18.5 Thoc2^+/Y and Thoc2^Δ/Y mouse brains using Hoechst staining. We observed a significant increase of Thoc2^Δ/Y NSCs in the G2/M phase with a concomitant reduction in the G1 phase (Fig. 6a), indicating a delay in the completion of cell cycle in the Thoc2^Δ/Y NSCs. TUNEL assay showed a significantly higher percentage of TUNEL positive cells indicating increased apoptosis of Thoc2^Δ/Y compared to Thoc2^+/Y NSCs (Fig. 6b). Consistent with the TUNEL results, our Annexin-V assay showed a significant increase in apoptotic cells in the Thoc2^Δ/Y compared to Thoc2^+/Y NSCs (Fig. 6c).

**Fig. 6: Perturbed cell cycle kinetics, increased cell death and elevated levels of DNA damage in *Thoc2*^Δ/Y mice.**

The G2/M checkpoint control senses unrepaired DNA damage and allows cells to repair before entering mitosis⁴⁵. We asked if the increase in Thoc2^Δ/Y NSCs in the G2/M phase could be a consequence of checkpoint invoked by accumulation of DNA damage in the preceding phase/s and could represent a pool of NSCs that, for unknown reasons, did not undergo apoptosis. We checked this possibility by determining the levels of classical DNA damage marker i.e., γ-H2AX protein (H2AX phosphorylated at Ser139)⁴⁶ in Thoc2^+/Y and Thoc2^Δ/Y NSCs by western blotting. We found significantly higher γ-H2AX levels in Thoc2^Δ/ than the Thoc2^+/Y NSCs although the total H2AX protein levels were unchanged (Fig. 6d). We validated this observation by analyzing the amount of DNA damage in NSCs using single cell gel electrophoresis comet assay⁴⁷. Consistent with increased γ-H2AX levels, a significantly higher number of Thoc2^Δ/Y NSCs showed comet tail formation compared to the Thoc2^+/Y NSCs, indicating accumulation of fragmented DNA resulting from DNA damage (Fig. 6e and Supplementary Fig. 6a, b). We next asked if the DNA damage is also present in vivo. To address this, we determined γ-H2AX and H2AX protein levels in Thoc2^Δ/Y and Thoc2^+/Y E14.5 and E18.5 (the stages that contain high number of NSCs) embryonic brain lysates. Similar to the in vitro NSCs, E14.5 and E18.5 Thoc2^Δ/Y mouse brains showed significantly increased γ-H2AX but not the H2AX protein levels (Fig. 6f). Lastly, we performed comet assay on the THOC2 exon 37-38 deletion patient fibroblasts to determine if these cells also have damaged DNA. Notably, compared to the unrelated human control fibroblasts, we observed comet tail formation in a significant number of patient fibroblasts, indicating an elevated DNA damage in patient cells (Fig. 6g and Supplementary Fig. 6c, d). Together, these data showed that THOC2 exon 37-38 deletion causes DNA damage both in the patient fibroblasts and the Thoc2^Δ/Y mouse NSCs.

Thoc2 ^Δ/Y NSCs and THOC2 exon 37-38 deletion patient fibroblasts have increased R-loops and DNA Damage

R-loops are three-stranded nucleic acid structures, composed of DNA:RNA hybrids and the associated non-template single-stranded DNAs that are highly susceptible to DNA breaks^48,49. To determine if the DNA damage in the Thoc2^Δ/Y NSCs was due to R-loop accumulation, we probed Thoc2^+/Y and Thoc2^Δ/Y NSCs with anti-RNA-DNA hybrid S9.6 monoclonal antibody to detect formation of R-loops by immunofluorescence. We observed a significantly increased accumulation of RNase H-sensitive R-loops with characteristic S9.6 fluorescent nuclear foci in Thoc2^Δ/Y compared to Thoc2^+/Y NSCs as well as patient compared to control fibroblasts (Fig. 7a, b). We observed similar results, albeit not as typical nuclear R-loop foci, using a different S9.6 antibody (Supplementary Fig. 7a, b). As RNase T1 treatment specifically degrades single-stranded RNAs, leaving RNA:DNA hybrids intact, we detected a clear increase in S9.6 nuclear signal in Thoc2^Δ/Y compared to Thoc2^+/Y RNase T1 treated NSCs as well as in patient compared to control fibroblasts, suggesting accumulation of RNA:DNA hybrids (Supplementary Fig. 7a, b)^50,51. The RNase H treatment, that specifically digests RNA in RNA:DNA hybrids, abolished S9.6 signal both in Thoc2^Δ/Y and Thoc2^+/Y NSCs and in control and patient fibroblasts (Fig. 7a, b and Supplementary Fig. 7a, b). We further confirmed these data by S9.6 dot-blot assay on the genomic DNA extracted (in RNase-free conditions) from mouse NSCs and patient fibroblasts (see Supplementary Fig. 7c for loading). While we observed no difference in S9.6 signal between the untreated Thoc2^Δ/Y and Thoc2^+/Y NSCs (Fig. 7c, top panel), RNase T1 treatment resulted in overall reduced but higher S9.6 signal in the Thoc2^Δ/Y than Thoc2^+/Y NSCs implying an accumulation of the R-loops in the Thoc2^Δ/Y NSCs (Fig. 7c, middle panel). As expected, S9.6-specific signal was abolished in the RNase H treated Thoc2^Δ/Y and Thoc2^+/Y NSCs (Fig. 7c, bottom panel). The untreated, RNase T1 or RNase H treated THOC2 exon 37-38 deletion patient and control fibroblasts showed S9.6 signals similar to those observed in the Thoc2^Δ/Y and Thoc2^+/Y NSCs, respectively, indicating R-loops also accumulate in the patient cells (Fig. 7d). We premised if reducing R-loop accumulation would be sufficient to rescue the DNA damage phenotype of Thoc2^Δ/Y NSCs. Indeed, RNase H1 overexpression resulted in significantly reduced R-loop accumulation (Fig. 7e) and comet tail formation (Fig. 7f) in Thoc2^Δ/Y NSCs, indicating that R-loop accumulation is likely responsible for DNA damage in these cells. Taken together, we suggest that THOC2 exon 37-38 deletion interferes with R-loop homeostasis and contributes to the increased DNA damage in the Thoc2^Δ/Y NSCs and patient fibroblasts.

**Fig. 7: R-loop accumulation and DNA damage in *Thoc2*^Δ/Y NSCs and *THOC2* ex37- 38 deletion patient fibroblasts.**

Thoc2 ^Δ/Y neurons show perturbed in vitro neural migration, electrophysiological properties and synapse formation

The molecular and cellular alterations in Thoc2^Δ/Y NSCs, together with the transcriptomic and proteomic data indicating multiple perturbations in neuronal processes, prompted us to investigate the effect of Thoc2 exon 37-38 deletion on neuron structure and function. We hypothesized that neurons, developing from the defective NSCs, might also show DNA damage. We addressed this by γ-H2AX immunofluorescence staining of the cortical/hippocampal Thoc2^Δ/Y E18.5 mouse primary neurons. We found that significant number of Thoc2^Δ/Y neurons show increased γ-H2AX foci, indicating increased DNA damage (Fig. 8a). As NSC perturbations can lead to altered neuronal migration^52,53, we asked if Thoc2^Δ/Y also impacts neuronal migration. We tested this by an in vitro neuronal migration assay using Thoc2^+/Y and Thoc2^Δ/Y neurospheres^52,54. We observed a significant reduction in distance the neurons migrated out of the seeded neurospheres from Thoc2^Δ/Y compared to those from Thoc2^+/Y mouse brain (Fig. 8b and Supplementary Fig. 8a). As the evidence suggests that different cell fates might be determined by the dynamics of the preceding rounds of NSC divisions⁵⁵, we asked if the altered cell cycle properties in Thoc2^Δ/Y NSCs impacts neural differentiation. To address this, we performed in vitro neural differentiation assay on E18.5 Thoc2^Δ/Y and Thoc2^+/Y NSCs⁵⁶. We then stained the cells in culture with cell specific markers (NSC: PAX6; Neurons: ß-III TUBULIN; Astrocytes: GFAP and Oligodendrocytes: CNPASE) (Fig. 8c) and scored the different cell types. We observed a significant increase in the differentiated cell population in the Thoc2^Δ/Y as compared to Thoc2^+/Y cultures, indicating a premature differentiation of the Thoc2^Δ/Y NSCs. Consistent with this observation, we observed a significant concomitant reduction in the number of progenitor cells in the Thoc2^Δ/Y as compared to Thoc2^+/Y cultures (Fig. 8c). We also analyzed the primary axon length, an indicator of the connectivity potential of neurons, of Thoc2^+/Y and Thoc2^Δ/Y E18.5 mouse neurons. While the Thoc2^Δ/Y neurons had 32% shorter primary axons as compared to the Thoc2^+/Y neurons (Fig. 8d), no significant changes were observed in the number of neurite termini between the Thoc2^+/Y and Thoc2^Δ/Y neurons (Supplementary Fig. 8b).

**Fig. 8: Structural and functional impairments in *Thoc2*^Δ/Y neurons.**

As we observed significant learning, memory, and sensorimotor problems as well as perturbations in neural maturation in the Thoc2^Δ/Y mice, we asked if these outcomes could be due to changes in synapse formation and/or function. We determined this by immunofluorescent staining for the pre- and post-synaptic markers Synapsin-1 (SYN1) and postsynaptic density-95 (PSD95) proteins, respectively, followed by colocalization quantification of excitatory synaptic puncta⁵⁷. We observed a significant reduction in SYN1-PSD-95 puncta in Thoc2^Δ/Y compared to the Thoc2^+/Y days in vitro 14 (DIV14) neurons, indicating an impairment in formation of functional synaptic connections among the Thoc2^Δ/Y DIV14 neurons (Fig. 8e). We also analysed in vivo synapse forming potential of Thoc2^Δ/Y mouse neurons using Golgi-Cox staining of the P30 Thoc2^Δ/Y mice brains and analysing the dendritic spines. While there was no obvious change in dendritic spine density between Thoc2^Δ/Y and Thoc2^+/Y neurons (Supplementary Fig. 8c), we observed a significant reduction in mature spines in the Thoc2^Δ/Y neurons (Fig. 8f) consistent with our in vitro synaptic puncta results Fig. 8e).

The morphological and synaptic impairments prompted us to investigate if there were any functional electrophysiological alterations in the Thoc2^Δ/Y mouse neurons. We tested this possibility using multi-electrode arrays (MEA) by determining the electrical activity of DIV7 and DIV14 primary cortical/hippocampal neuron cultures generated from E18.5 Thoc2^+/Y and Thoc2^Δ/Y mouse brains. DIV7 neurons, that are at the immature stage in culture, did not show any difference in electrophysiological parameters like mean firing rate, mean burst rate, mean burst spike rate and network burst rate between the Thoc2^Δ/Y and Thoc2^+/Y neuron cultures (Supplementary Fig. 8d). However, at DIV14, a more mature stage of neuronal development, we observed significant reduction in neural network activity as indicated by little to no network bursts in the Thoc2^Δ/Y neuronal networks (Fig. 8g and Supplementary Fig. 8e) in neurons plated at same density (Supplementary Fig. 8f). This observation was confirmed by an increase in random spike (isolated, a-synchronous spikes) frequency in Thoc2^Δ/Y cultures, indicative of failure to form mature neuronal network activity (Fig. 8h). Apart from the neural network activity, the overall electrical activity of the neurons cultured from Thoc2^Δ/Y mice brains was also significantly decreased, as indicated by the reduction in mean firing rate, mean burst rate and mean burst spike rate (Fig. 8i–l).

Overall, we conclude that Thoc2 exon 37-38 deletion causes multiple downstream impairments of NSC-derived neurons like migration, maturation, synapse formation and electrophysiology. Our observation of significant reduction in excitatory synaptic puncta formation is consistent with the data showing failure of Thoc2^Δ/Y E18.5 mouse neurons in establishing synchronous functional neural networks. These results were supported by our transcriptome and proteome data showing enrichment of dysregulated gene/proteins in synapse formation and function.

Discussion

Originally identified as mRNA export machinery, the TREX complex has emerged to have roles in multiple biological processes including transcription regulation, maintaining genomic stability, and regulating stem cell maintenance^6,7,16,58. Accumulating evidence has linked perturbed TREX function to neurodevelopmental and neurodegenerative disorders as well as cancer^{5,7,59,60,61,62}. As an essential TREX subunit, THOC2 has been the focus of many studies^{11,59,60,63,64,65,66,67}. While these studies have provided insights into varied roles of THOC2 protein, the findings were obtained mainly from in vitro knockdown cell or knockout lower organism models^{6,20,21,22,24,25,68}. The lack of a viable mammalian model has been a major barrier to investigating the biological significance of THOC2 (and more broadly the TREX complex) due to gene essentiality. Thus, to gain deeper insights into the biology of THOC2 and TREX in brain development, and more specifically, how compromised THOC2 function leads to NDDs, we generated and studied the mouse model of THOC2 syndrome based on a pathogenic exon 37-38 deletion patient variant⁵.

The yeast Tho2 C-terminal region has been shown to bind nucleic acids and facilitate recruitment of the THO to chromatin¹⁰. While Tho2 deleted for the C-terminal region did not affect TREX assembly, it did impact gene expression¹⁰. The Thoc2 exon 37-38 deletion also did not have a significant effect on the levels of THO/TREX subunits (see Supplementary Fig. 1e), but did impact gene expression and genome stability, suggesting a partial loss of function mechanism of THOC2 exon 37-38 deleted protein in the mouse brain. Overall, the Thoc2^Δ/Y mice phenocopied the core features of the THOC2 syndrome. Similar to the patient, the Thoc2^Δ/Y mice had reduced body weight and size. In addition, the birth rate of Thoc2^Δ/Y mice was reduced by about one third of the number expected. Therefore, we investigated if reduction in body size and weight across all organs as well as low birth rate was due to general and early developmental deficits in proliferation, differentiation, and likely cell death of stem cells.

IVF assays showed that the lower birth rate of Thoc2^Δ/Y mice is likely due to alterations before the blastocyst stage, when the majority of the cells in the embryo are multipotent stem cells. A previous report has also shown Thoc2 to be important for blastocyst development and pluripotency maintenance⁶. Together, these findings confirmed the crucial role of Thoc2 during early embryonic development, and likely in maintaining the stem cell niche. That THO subunit proteins are also essential for maintenance of haematopoiesis, intestinal stem cells and triple negative breast cancer stemness indicates broader developmental roles of the TREX complex^59,69,70. The importance of the THO complex in brain development was supported by our observations that THOC2 and virtually all other TREX subunit protein levels are highly abundant in brains at embryonic stages (E14.5 and E18.5) but decline sharply from post-natal day 10 (P10) to P60. (see Fig. 1 and Supplementary Fig. 1). The precise molecular events that allow some Thoc2^Δ/Y mouse embryos to survive to birth and beyond are unknown and need further investigations.

The Thoc2^Δ/Y mice that survived, however, generally recapitulated the phenotype of THOC2 syndrome^3,4,5. Multiple mouse models of ID and ASDs for genes like Syngap2, Fmr1, Mecp2 and Shank2/3 with converging cognitive, social, and emotional deficits have proved to be excellent tools for understanding the neuropathology (reviewed in⁷¹). Likewise, the Thoc2^Δ/Y mice showed significant deficit in learning the task on both the Morris Water Maze and Barnes maze, as well as impaired spatial working and reference memory on the Y-maze. Such cognitive impairments are clinical hallmarks for NDD/ID disease models⁷¹, and suggest that the Thoc2^Δ/Y mouse model has translational relevance for THOC2 syndrome, where patients have ID as the core clinical presentation⁵. Interestingly, the Thoc2^Δ/Y mice are also hyperactive and have decreased anxiety-associated behaviour, traits seen as comorbid features in a subset of THOC2 patients⁵. These observations are consistent with our proteomics data, which showed dysregulation of proteins associated with these behavioural phenotypes (see Fig. 4 and Supplementary Data 4). Moreover, we observed significant deficits in sensorimotor control and fine motor coordination noted on multiple behavioural measures, including the pasta handling task, the adhesive removal test and beam walking in Thoc2^Δ/Y mice. While the neural basis of this has yet to be explored, it may be due to impaired myelination in these mice. In line with this, a significant generalised delay in myelination and reduction in white matter volume, particularly in the trigone region was observed from MRI imaging in the THOC2 exon 37-38 deletion patient⁵. Importantly, however, overall gross motor functions in Thoc2^Δ/Y mice were not impaired, as indicated by performance comparable to wild-type mice on the rotarod, Digigait, and grip strength tests. This is similar to what is observed in THOC2 patients⁵, and is in line with imaging findings in the THOC2 exon 37-38 deletion patient, which indicated moderate plagiocephaly and brachycephaly, but no ex-vacuo dilation of the lateral ventricles, no heterotopic grey matter and normal-appearing subcortical and brainstem structures. Similarly, the E18.5 Thoc2^Δ/Y mouse brains had reduced VZ but no other gross morphological anomaly. However, our transcriptomic and proteomic data from the embryonic and adult Thoc2^Δ/Y mouse brains revealed significant dysregulation of genes and proteins that normally are major mediators of nervous system function and development. Enrichment of dysregulated genes/proteins in pathways like transcription, splicing, embryonic lethality, learning, memory, and synapse organization converge very well with our cellular and molecular findings, e.g., death of Thoc2^Δ/Y mice before blastocyst stage, the behavioural tests showing significant perturbation in learning and memory and cell cycle anomaly and cell death of the Thoc2^Δ/Y mouse NSCs. Interestingly, we observed significant transcriptional dysregulation of known NDD genes, like Syngap1, Huwe1, Shank3, Dlg4, Kdm5c, and Ctnnd1 (among many others) (see Supplementary Fig. 3), which indicates that Thoc2 acts as an important regulator of neurodevelopmental processes and its compromised function results in diverse neurological symptoms.

An increased THOC2 protein abundance during embryonic stages and significant death of Thoc2^Δ/Y embryos before blastocyst stage, together with the data showing that THOC2 and THOC5 are required for mESC maintenance⁶, strongly support a critical role of THOC2/TREX in early development. These findings, together with a reduced VZ thickness in the E14.5 and E18.5 Thoc2^Δ/Y mouse brains and neuropathological outcomes pointed towards likely aberrations in the neural progenitors. We noted deregulation of the cell cycle, particularly at the G2/M phase, in the Thoc2^Δ/Y NSCs. The G2/M checkpoint is activated upon DNA damage and prevents the cells from entering mitosis with damaged genomic DNA⁴⁵. Indeed, Thoc2^Δ/Y mouse NSCs had higher amounts of damaged DNA. The elevated DNA damage was followed by increased apoptosis in the Thoc2^Δ/Y NSCs. Note that a recent study has also reported an increased G2/M cell count in THOC2 and THOC5 depleted radioresistant triple negative breast cancer (TNBC) cells⁵⁹.

Impairment of NSCs has been linked to multiple NDD, neurodegenerative disorders, and to brain tumours [reviewed in ref. ⁴³]. The multiple molecular and cellular impairments observed in Thoc2^Δ/Y NSCs were correlated with a plethora of neuron structure and function anomalies, for example, presence of significant in vitro neural migration defects with premature differentiation of NSCs and axon shortening in the cultured Thoc2^Δ/Y primary neurons. Notably, we found a significant reduction in mature dendritic spines in the Thoc2^Δ/Y mice brains and a reduction in formation of synaptic puncta in cultured Thoc2^Δ/Y mouse primary neurons, which agree with the results showing an impaired neural network activity in the Thoc2^Δ/Y mouse neurons in DIV14 culture. Our results corroborate with synaptic disorganizations observed in NDDs like ID, ASD, and epilepsy^72,73 along with studies showing minimal alteration in proper synaptic activity causing significant disruption in the neuronal network formation^74,75. Interestingly, our proteomics data also showed significant dysregulation of proteins associated with synapse formation and function. Taken together, our data showing multiple functional deficits in Thoc2^Δ/Y mouse neurons suggest that THOC2 has critical roles in post-mitotic neurons. Literature also shows that TREX complex plays an essential role in cellular senescence⁷⁶. Furthermore, THOC2 was shown to mis-localize in Huntington’s Disease and Amyotrophic lateral sclerosis, potentially contributing to these neurodegenerative pathologies²⁶. Therefore, neuronal defects in our mouse model together with the other studies highlight a critical role for THOC2/TREX in neurons, stem cells and early stages of mouse development.

Studies have shown that increased DNA damage in THO subunit depleted cells was caused by unscheduled accumulation of triple-stranded R-loop structures^19,77, and that other TREX components were also close interactors of R-loops⁷⁸. In addition, with their co-transcriptional origin and increased susceptibility to DNA damage, accumulating recent evidence has recognized R-loops as potential regulators of gene expression^79,80. More recently, R-loops and associated DNA damage have been implicated in the etiology of neurological disease like spinal muscular atrophy (SMA) and ALS⁸¹. Based on these results and the evidence that the yeast Tho2 (orthologue of mammalian THOC2) interacts with nucleic acids via it’s C-terminal disordered region¹⁰, we premised that increased DNA damage and transcriptional dysregulation could be due to inefficient resolution of the R-loops in Thoc2^Δ/Y mice and THOC2 exon 37-38 deletion patient cells. Indeed, we observed increased accumulation of R-loops in the Thoc2^Δ/Y NSCs that was significantly reduced on RNase H1 overexpression and was accompanied by a concomitant decline in DNA damage, thus establishing R-loop accumulation as one of the major contributors to the DNA damage detected in Thoc2^Δ/Y NSCs. We observed similar R-loop accumulation and DNA damage in the THOC2 exon 37-38 deletion patient-derived fibroblasts, reiterating a conserved mechanism of perturbed THOC2 function related cellular anomalies.

Our data clearly showed C-terminally truncated THOC2 protein impacts R-loop homeostasis leading to increased DNA damage and transcriptional dysregulation in Thoc2^Δ/Y neurons/brain. However, THOC2 transient depletion results in bulk polyA+ mRNA nuclear retention²⁴. Still others have shown THOC1, 2, 5 or 6 subunits mediate nuclear export of specific mRNAs, e.g., those involved in pluripotency or haematopoiesis^{6,12,21,65,69}. Therefore, we cannot completely rule out an impact on the export of specific mRNAs in the THOC2 exon 37-38 deleted cells even though an obvious nuclear mRNA retention in Thoc2^Δ/Y E18.5 neurons and patient fibroblasts was not observed (see Supplementary Fig. 3a). RNA-seq of neuronal cytoplasmic and nuclear RNA fractions may facilitate identification of a subset of mRNAs whose export from the nucleus may be impacted in Thoc2^Δ/Y neurons and perhaps also in other cell types and stem cell niches. Additionally, THOC2 has been shown as a critical factor for cancer cell survival^59,62,63 and its mutations are positively correlated with increased survival of cancer patients⁶⁷. Therefore, we believe our Thoc2 mouse model can serve as a starting point towards exploration of unprecedented onco-therapeutic strategies.

In summary, our unsuccessful attempts to generate mouse model for p.Leu438Pro or p.Ile800Thr pathogenic patient variants suggested that reduced stability of THOC2 and other THO subunit proteins, possibly resulting in loss of functional TREX complex, is incompatible for survival in mice. However, the p.Leu438Pro or p.Ile800Thr reduced levels are compatible with life in humans but at the expense of causing NDDs due to altered neuronal development. These observations confirm the importance of THOC2 in healthy development and differentiation. Following on from this, we successfully generated a viable Thoc2 mouse model of a unique naturally occurring pathogenic THOC2 exon 37-38 deletion patient variant. Although THOC2 exon 37-38 truncated protein accumulated and did not impact the levels of other TREX subunits in Thoc2^Δ/Y and patient cells, it led to a functionally compromised TREX complex that is sufficient for survival of Thoc2^Δ/Y mice. The Thoc2^Δ/Y mice allowed us to investigate biological functions of THOC2 beyond its essential role in mRNA export. We also found that perturbed R-loop homeostasis and DNA damage contributes to many molecular, cellular, and organism-level anomalies in the Thoc2^Δ/Y mice and likely, also in the individuals living with THOC2 syndrome. Overall, our results establish Thoc2^Δ/Y mouse as an invaluable model for deeper understanding of the THOC2/TREX biology in stemness, development and differentiation as well as the clinical neuropathology of THOC2 syndrome in the future. A schematic diagram showing the molecular, cellular, and phenotypic impact of exon 37-38 deleted truncated THOC2 protein on Thoc2^Δ/Y mice is illustrated in Fig. 9.

**Fig. 9: Summary of the key results of studies presented here.**

Methods

Experimental animal model and ethics statement

C57BL/6JArc inbred genetic background mice (www.arc.wa.gov.au) were used for generating the Thoc2 exon 37-38 deletion mouse model. The mouse colony maintenance and breeding were carried out in Bioresource Facility (BRF), Laboratory Animal Services (LAS), the University of Adelaide. All breeding and experimental procedures were approved by Animal Ethics Committee (AEC) of the University of Adelaide. We abbreviated the Thoc2 exon 37-38 deletion male mice as Thoc2^Δ/Y, wild type (WT) males as Thoc2^+/Y, heterozygous females as Thoc2^Δ/+, WT females as Thoc2^+/+ and homozygous females as Thoc2^Δ/Δ. The animals were housed in individually ventilated cages (IVCs) under a 22–24 °C temperature and 40–70% relative humidity environment. The mice were given access to food and water ad libitum with a 12 hr light/dark cycle. The Thoc2^Δ/Y and Thoc2^+/Y mice (at 2 weeks of age and embryonic stages) were genotyped by our optimised protocol that involved rapid genomic DNA extraction (from ear notches of 2-week-old mice or a small amount of embryonic tissue) and multiplex PCR (Supplementary Fig. 1h). We maintained the colony and generated experimental animals by mating Thoc2^Δ/+ with Thoc2^+/Y mice (Supplementary Fig. 1b).

CRISPR/Cas9 editing for generating Thoc2 exon 37-38 deletion mice

Wild type (WT) SpCas9 was complexed with guide RNAs targeting introns 36 and 38 (see Supplementary Data 5) along with PAM killing ssDNA templates (Supplementary Data 5) and then microinjected into the cytoplasm of C57BL/6JArc zygotes. Injected zygotes were then transferred to the oviduct of pseudo-pregnant females for in-utero development. The two single-stranded oligos were utilized to insert PAM killing mutations at the intronic cut sites to reduce the proportion of alleles that may be repaired by Non-Homologous End Joining (NHEJ) between the two cut sites which would delete exons 37 and 38 (see Fig. 1a). After successful pregnancy and birth, pups were genotyped using a rapid PCR strategy (Supplementary Fig. 1j) and deletion of Exon 37 and 38 was confirmed by Sanger sequencing. For genotyping the founder animals, genomic DNA (gDNA) was isolated from a small ear notch (also served as an animal identifier) using High Pure PCR Template Preparation Kit (Roche, 11796828001) as per manufacturer’s instructions. PCR was performed with KAPA HiFi HotStart PCR mix (Roche, KK2601) (details in Supplementary Data 5).

Rapid gDNA extraction and multiplex genotyping PCR

For routine genotyping of the Thoc2 mice and especially embryos, a rapid gDNA extraction and PCR method was developed. gDNA was isolated from a small ear notch of the Thoc2 mice (also served as animal identifier) or embryo toe-tip using MyTaq Extract-PCR Kit (Meridian Bioscience, BIO-21127). PCR amplification was performed using MyTaq HS Red mix 2× supplied with the gDNA extraction kit (details in Supplementary Data 5).

RNA extraction and RT-PCR

RNA was isolated from mouse brain using RNeasy Mini Kit and on-column RNase-free DNase-treatment (Qiagen, 74004). cDNAs were generated from the total RNA (1 µg each) using Superscript IV reverse transcriptase (Invitrogen, 18091050). PCR amplification was performed using Phusion High-Fidelity DNA Polymerase (NEB, M0530S). PCR products were gel purified and subjected to Sanger sequencing.

In vitro fertilization and blastocyst genotyping

In vitro fertilization was performed to generate large batch of blastocysts using a published protocol⁸². Briefly, the donor females (8–12 weeks old) were first injected with 0.15 ml of pregnant mare’s serum gonadotropin (PMSG) (7.5 IU) and 48 h later with 0.15 ml of human chorionic gonadotropin (hCG) (7.5 IU) intraperitoneally. After 15–17 h, the superovulated females were sacrificed by cervical dislocation and oviducts were collected. Similarly, mature donor males (3-4 months old) were sacrificed and cauda epididymides harvested, cut and the clots of spermatozoa were collected in a drop of CARD FERTIUP-PM medium (Cosmo Bio, USA, KYD-002-05-EX) and allowed to capacitate for 60 min by placing the dish in an incubator (37 °C, 5% CO₂). Next, each oviduct ampulla was dissected, and cumulus-oocyte-complexes (COCs) were allowed to release and transferred into drops of CARD Medium (Cosmo Bio, USA, KYD-003-EX). The COCs were then placed in an incubator (37 °C, 5% CO₂) for 30–60 min before starting the in vitro fertilization process. After the incubation period, adequate amount of sperm suspension was added to the drops of COCs. After 3 h of incubation, the oocytes were washed three times using modified human tubal fluid (mHTF)⁸³ and cultured for further 3 h. After a total of 6 h of incubation with sperm in vitro, presumptive zygotes were assessed for evidence of fertilization (2 pronuclei), divided into groups of 20–30 zygotes per drop of mHTF and cultured in mHTF overnight. Next day, 2-cell embryos were assessed and grouped for further 72-hour culture allowing them to develop to blastocyst stage. The blastocysts were removed from the culture media and transferred into micro-drops of M2 media (Merck, M7167) at room temperature. Individual blastocysts were frozen in PCR tubes in 1 µl of M2 medium plus 9 µl sterile water at -20 °C for subsequent DNA extraction. Blastocyst gDNA was extracted by adding 10 µl of extraction buffer (185.5 mM Tris-HCl pH 8.3, 185.5 mM KCl, 0.008 % gelatin, 0.8 % Tween 20, 1.2 mg/ml baker’s yeast tRNA (Sigma, R8759) and 8.5 mg/ml Proteinase K (Thermo Scientific, EO0491) into the individual blastocyst containing PCR tubes followed by incubation at 56 °C for 10 min and then at 95 °C for 10 min. The blastocysts were then genotyped by same the PCR strategy used for genotyping the Thoc2 embryos and postnatal mice as described above.

Cell culture

Fibroblasts

Dermal fibroblasts from the THOC2 Ex37-38Del patient and unrelated male controls were maintained in DMEM (Gibco, 10566024) containing 2 mM L-glutamine, 1 mM sodium pyruvate, and 10% foetal bovine serum (FBS) at 37 °C with 5% CO₂.

Primary Mouse Neurons

Primary neuronal (cortical and hippocampal) cultures were prepared from E18.5 (18.5 days post coitum) embryonic brains isolated from the time-mated Thoc2^Δ/+ females as previously described with some modifications⁸⁴. Briefly, the brains were micro-dissected to remove midbrain and the meninges, trypsinised (0.05% w/v) for 20 min at 37 °C with gentle shaking every 5 min. The tissues were then washed with neuron-plating medium (Neurobasal A medium supplemented with 10% FBS, 1 mM sodium pyruvate, 2 mM L-glutamine and 1% penicillin/streptomycin). The cells were triturated using fire-polished glass pipettes and plated in Poly-L-Lysine-coated coverslips (Menzel Gläser, MENZBB018018A2) at a density of 3 × 10³ cells per well of 12-well culture plates for immunofluorescence (IF) studies or 1 × 10⁶ cells per well of 24-well MEA plate (Multichannel Systems, Germany, 24W300/30G-288) for multielectrode array studies, and maintained in neuron maintenance medium (Neurobasal A supplemented with 2% B-27 supplement, 2 mM Glutamax and 1% penicillin/ streptomycin) and cultured for the required duration with half change of media every 3^rd day.

Primary mouse neural stem cells (NSC)

NSCs were isolated from the dorsal cortex of E18.5 embryonic mouse brains and cultured as non-adherent neurospheres as described⁵⁶. Briefly, dorsal cortex was isolated by micro-dissecting midbrain and the meninges, and then trypsinised (0.05% w/v) for 20 min at 37 °C with gentle shaking every 5 min. The tissues were then washed with NSC medium (DMEM/F12 medium supplemented with 2% B27, 1% penicillin/streptomycin, 20 ng/ml EGF and 20 ng/ml bFGF). The cells were triturated using fire-polished glass pipettes and plated in T25 flasks with NSC medium at a density of 4 × 10⁶ cells/flask. The neurospheres were supplemented with fresh NSC medium every 3rd Day. For generating the adherent NSCs, passage 2 neurospheres were dissociated into single cell suspension using trypsin (0.05%) in Versene solution (Gibco, 15040066) and plated at a density of 1.8 × 10⁶ cells/9.5 cm² culture area in NSC medium. The media was half changed every 2^nd Day.

RNase H1 lentivirus generation

Mouse RNase H1 (NM_001286865.1) was PCR amplified from the mouse fibroblast cDNA using attB1-forward and attB2-Reverse primer (sequences in Supplementary Data 5) with Platinum SuperFi PCR Master Mix (Thermo Scientific, 12368010), gel purified and recombined into pDONR221 gateway plasmid. The resulting construct was recombined into pJS65 MND-DEST SV40-Tomato (kindly provided by Prof Richard Iggo, INSERM, Bordeaux, France). The RNase H1 expressing lentivirus was generated by co-transfecting pMD2.G, psPAX2 and pJS65 MND-DEST SV40-Tomato-Myc-mRNase H1 plasmids using published PEI protocol⁸⁵. The lentivirus was tested for RNase H1 expression by transducing the HEK293T cells (ATCC, 293T-CRL-3216) at different MOIs and western blotted with the mouse anti-c-Myc (9E10)-HRP antibody (1:2000) (See Supplementary Data 6) (Supplementary Fig. 8b).

Western blotting

Mouse tissue samples were lysed in 50 mM Tris–HCl pH 8.0, 150 mM NaCl, 1% IGEPAL CA-630, 0.5% sodium deoxycholate, 0.1% SDS, 1× Protease inhibitor/No EDTA and cultured cells (NSCs and Fibroblasts) in 50 mM Tris–HCl pH 7.5, 250 mM NaCl, 0.1% Triton X-100, 1 mM EDTA, 50 mM NaF, 0.1 mM Na₃VO₄ and 1× Protease inhibitor/No EDTA (Roche). The lysates were sonicated using microtip, 15 sec at 30% amplitude (3× times) for tissues and once for 10 sec at 25% amplitude for cells (Sonics Vibra-Cell VCX 130). The tissue/cell debris were removed by centrifugation at 15,000× g for 20 min at 4 °C and proteins assayed using Pierce BCA assay kit (Thermo Scientific, 23227). 8 µg protein of each sample was resolved on 4-12% Bis-Tris (Bio-Rad, 3450123) or 3–8% Tris-acetate protein gels (Invitrogen, EA03785BOX), transferred to nitrocellulose membranes and western blotted with appropriate primary (1:1000) and HRP conjugated secondary antibodies (1:1000) (see Supplementary Data 6). Chemiluminescence signal was detected by Clarity Western ECL Substrate (Bio-Rad, 1705060) and captured using a Bio-Rad ChemiDoc MP Gel Documentation System. Note: The anti-THOC2-I antibody epitope is located between amino acids 1097-1200 and the anti-THOC2-II antibody between amino acids 1543–1593. The two antibodies were used to distinguish full length and truncated THOC2 proteins from the Thoc2^+/Y and Thoc2^Δ/Y mouse brain or patient cell lysates, respectively. The unprocessed western blots are provided in the Source data file.

H&E staining

E18.5 and P30 mouse brains were fixed in 4% paraformaldehyde (PFA) and embedded in paraffin. 5 µm sections were cut and stained with Mayer’s Haematoxylin and Eosin in a Leica Autostainer XL/Leica CV5030 Coverslipper machine following standard protocols. Images were acquired using 3D Histech Pannoramic Scan II machine equipped with Carl Zeiss Plan-Apochromat ×20 objective with Point Grey Grasshopper 3 CCD monochrome camera and LED-based RGB illumination unit. The image measurements and annotations were performed in the 3D Histech SlideViewer (V2.6) software.

Golgi-Cox staining

Golgi-cox staining on P30 mouse brain was performed using a previously reported protocol with slight modification⁸⁶. Briefly, mouse brains were perfused using 4% PFA, removed from the skull and immersed in 4% PFA overnight at 4^oC. The brains were then incubated for 72 h in Golgi-Cox stain (5% potassium dichromate solution, 5% mercuric chloride and 5% potassium chromate) at 37 °C. The tissue blocks were then removed and placed into 30% sucrose solution for 3 days at room temperature. Sections (200 μm thick) were cut on a vibratome, and images were taken using an Olympus IX83 microscope (Olympus, Japan). Subsequently, analysis of the dendritic spines was carried out using Fiji software (NIH). All sample analysis was carried out blind to mouse genotype.

Immunofluorescence (IF)

Cultured cells

NSCs, primary mouse neurons and fibroblasts were cultured on Poly-L-Lysine (Sigma, P2636) coated glass coverslips and immunostained using a modified protocol reported earlier³. Briefly, cells were fixed with 4% PFA for 10 min, washed three times with 1 × PBS, permeabilized with 0.2% Triton X-100 in 1 × PBS for 10 min, washed three times with 1 × PBS, and blocked in 10% Horse Serum in 1 × PBS for 1 h at room temperature. Cells were then incubated with appropriate primary antibodies (1:1000) (see Supplementary Data 6) diluted in 10% Horse Serum in 1 × PBS overnight at 4 °C. The following day, cells were washed three times with 1 × PBS, incubated with appropriate Alexa Fluor secondary antibodies (1:1500) (see Supplementary Data 6) for 2 h at room temperature, washed three times with 1 × PBS and then mounted in ProLong Gold Antifade Mounting agent with DAPI (Invitrogen, P36935) with overnight curing.

Brain sections

The IF protocol for tissue sections was used as described earlier⁵⁷ with some modifications. Briefly, E18.5 embryonic mouse brains were fixed by immersion in 4% PFA for 24 h, cryoprotected in 30% sucrose and embedded in Optimal cutting temperature compound (OCT, Tissue-Tek, IA018). A Leica CM1900 cryostat was used to prepare 16 μm sections that were permeabilized using 0.5% Triton X-100 in 1 × PBS for 5 min at room temperature. The sections were washed three times with 1 × PBS, blocked with 10% Horse Serum, 1% BSA and 0.1% Triton X-100 in 1 × PBS for 1 hr at room temperature and incubated with appropriate primary antibodies (1:500 dilution) (see Supplementary Data 6) diluted in 10% Horse Serum, 1% BSA and 0.1% Triton X-100 in 1 × PBS overnight at 4 °C. Following day, sections were washed three times with 0.05% Triton X-100 in 1 × PBS, incubated with appropriate Alexa Fluor secondary antibodies (1:1000 dilution) (see Supplementary Data 6) for 3-4hrs at room temperature, washed three times with 1 × PBS and mounted in ProLong Gold Antifade Mounting agent with DAPI with overnight curing.

RNA-DNA hybrid immunofluorescence using S9.6 antibody

S9.6 IF was performed as per published protocols^87,88. Briefly, mouse NSCs and patient fibroblasts were fixed in ice-cold 100% methanol for 10 min at −20 °C and blocked using 1% BSA in 1× PBS for 1 h at room temperature. Cells were treated with RNase T1 (Thermo Scientific, EN0541) or RNase H (NEB, M0297S) in the presence of appropriate buffers supplied with the enzymes and then incubated with mouse anti-S9.6 antibody (Sigma, MABE1095 or Kerafast, AB01137-23.0) at 1:1000 in 1% BSA in 1× PBS overnight at 4 °C. Next day, cells were washed three times with 1× PBS, incubated with secondary anti-rabbit Alexa Fluor 555 conjugate (see Supplementary Data 6) diluted in 1× PBS with 1% BSA (1:2000) for 2 h at room temperature and finally, mounted in ProLong Gold Antifade Mounting agent with DAPI. Images were captured using a Zeiss Axioplan2 microscope (Carl Zeiss, Jena, Germany) equipped with an HBO 100 lamp (Carl Zeiss) using Axiocam Mrm camera and V.4.9.1.0 software (Axiovision, Carl Zeiss) at 63×63 magnification. For NSCs, image acquisition settings were calibrated using the Thoc2^+/Y NSCs, and for patient fibroblasts, image acquisition settings were calibrated using unrelated control fibroblasts. The same settings were then applied to assess the experimental samples i.e., Thoc2^Δ/Y NSCs and THOC2 exon 37-38 deletion patient fibroblasts. Fiji (ImageJ) software was used for measuring mean nuclear grey value for S9.6 fluorescence⁸⁹.

Fluorescence in situ hybridization (FISH)

FISH was used to detect polyA+ RNA in human fibroblasts and mouse neurons as described previously⁹⁰. Briefly, cells were washed once with 1× PBS, fixed with 4% PFA for 10 min, washed three times with 1 × PBS, and then permeabilized with 0.5% Triton X-100 in 1× PBS for 5 min, washed two times with 1× PBS and then once with 2× SSC for 10 mins at room temperature. An oligo dT (70) probe (HPLC purified) labelled with Alexa Fluor 546 at the 5’ end, was then added at a concentration of 1 ng/µl and incubation was performed overnight at 42 °C. The cells were washed two times with 2× SSC, for 15 min each, then once with 0.5× SSC for 5 min, and lastly, once with 1× PBS for 5 min at room temperature. Images were captured using a Zeiss Axioplan2 microscope (Carl Zeiss, Jena, Germany) equipped with an HBO 100 lamp (Carl Zeiss) using Axiocam Mrm camera and V.4.9.1.0 software (Axiovision, Carl Zeiss).

Flow cytometry

Apoptosis assay

Apoptotic NSCs were detected by annexin V staining using a published protocol⁹¹ with some modifications. Briefly, NSCs were harvested, washed in 1× PBS, and resuspended in 20 μl annexin V binding buffer (Hank’s balanced salt solution with 1% HEPES and 5 mmol/L CaCl₂) with 0.075 μg/ml annexin V-PE (Stem Cell Technologies, 100-0331) at 1 × 10⁵ cells/ml. The cells were stained for 20 min at 4 °C, diluted with 200 μl of ice-cold annexin V binding buffer, and analysed immediately on a LSRFortessa X20 Flow cytometer (BD Biosciences).

Cell cycle assay

For cell cycle assay, NSCs were resuspended in pre-warmed media at 1 × 10⁶ cells/ml. Hoechst 33342 (Sigma, B2261) added at a final concentration of 10 µg/ml, mixed thoroughly, and incubated at 37 °C for 45 min in dark. The cells were then pelleted and resuspended in 500 µL of cold PFE solution (1× PBS, 2% FBS, 2 mM EDTA) and analysed on a LSRFortessa X20 Flow cytometer (BD Biosciences). All Flow Cytometry data were analysed using FlowJo V. 10.0.8 software (FlowJo, LLC). The gating strategy has been exemplified in Supplementary Fig. 9 provided in the Supplementary information file.

TUNEL assay

TUNEL assay was performed on NSCs cultured on Poly-L-Lysine treated coverslips using Click-iT Plus TUNEL Assay kit (Invitrogen, C10617). The coverslips were then mounted using ProLong Gold Antifade Mounting agent with DAPI and imaged on a Zeiss Axioplan2 microscope using Axiocam Mrm camera and V.4.9.1.0 software at ×20 magnification. Percentage of apoptotic NSCs was determined by quantifying TUNEL-positive cells in five random microscopic fields capturing ≥100 cells per technical replicate.

Single cell electrophoresis assay

DNA damage in E18.5 mouse NSCs and patient fibroblasts was assessed using alkaline single cell electrophoresis or comet assay⁴⁷ with minor modifications. Briefly, NSCs were harvested and resuspended in 1% low-melting point agarose prepared in distilled water. The cell suspension was dropped on glass slides precoated with a thin layer of 1% agarose in distilled water and spread with a coverslip (22 mm × 22 mm) to make the electrophoresis mini gel. The cells were lysed using comet lysis buffer (2.5 M NaCl, 100 mM disodium EDTA, 10 mM Tris-HCl pH 10, 200 mM NaOH, 1% Triton X-100 and 1% sodium lauryl sarcosinate) for 1 hr at 4 °C. Electrophoresis was performed using an alkaline electrophoresis solution (200 mM NaOH, 1 mM disodium EDTA, pH >13). The slides were stained with SYBRGold Nucleic Acid Gel Stain (Invitrogen, S11494) and imaged using a Zeiss Axioplan2 microscope. The tail moments were quantified using the CaspLab software⁹².

Dot blot assay

Dot blot assay was performed using a published protocol⁶⁴ with some modifications. Briefly, gDNA was isolated from mouse NSCs or patient fibroblasts using DNeasy Blood and Tissue Kit (Qiagen, 69504) without the RNase treatment. Equal amount of gDNA of each sample (2 µg) was treated with RNase T1 (Thermo Scientific, EN0541), RNase H (NEB, M0297S) or buffer at 37 °C for 1 hr. The samples were spotted on a nitrocellulose membrane, cross-linked using a UV cross linker (GS Gene Linker, Bio-Rad) (120 mJ) and blocked in 5% milk in PBST (1× PBS, 0.1% Tween 20) overnight. The membrane was incubated with rabbit S9.6 antibody (1:1000) diluted in 3% milk, 1% BSA in PBST overnight at 4 °C. Next day, the membrane was washed three times with PBST, incubated with goat anti-rabbit-HRP antibody (Agilent, P0448) diluted in 3% milk, 1% BSA in PBST (1:1000) for 2 h at room temperature, washed three times with PBST and visualized using Clarity Western ECL Substrate on a Bio-Rad ChemiDoc MP Gel Documentation System.

Neural migration assay

In vitro neuronal migration assay was performed as reported⁵⁴. Neurospheres at Passage 3 were cultured for 5 days and then seeded on Poly-L-Lysine coated coverslips at 100 neurospheres/9.5 cm² in NSC medium without the growth factors. Neurons were allowed to migrate away from the seeded neurosphere for 48 h and then fixed with 4% PFA. Immunofluorescence was performed using anti-ß-Tubulin III (Tuj1) antibody (1:1000 dilution). A neuronal migration scoring system we published was used⁵². Neurospheres and the migrating neurons were imaged using a Zeiss Axioplan2 microscope (Carl Zeiss). The edge of contiguous nuclei was used as neurosphere periphery and its outline was used to generate concentric circles (bins) using Fiji software, outside of the neurosphere (which are the migration zones) with increasing size (each bin increases by 45 µm). Neural cell density was then calculated in each bin, which results in migration distance measurement independent of initial sphere size, and bin volumes.

Neural differentiation assay

Passage 3 neurospheres were cultured for 5 days, dissociated into single cell suspension, and plated on Poly-L-Lysine coated coverslips at 2.1 × 10⁴ cells/9.5 cm² culture area in NSC medium without the growth factors. The cells were allowed to differentiate for 72 h and then fixed with 4% PFA. Immunofluorescence was performed as above. Cell type specific primary antibodies used were: GFAP (for astrocytes), CNPASE (for oligodendrocytes), ß-III TUBULIN (for neurons), and PAX6 (for progenitors). The details of the primary (all diluted 1:1000) and secondary antibodies (all diluted 1:2000) are provided in the Supplementary Data 6.

Multielectrode array analysis

Spontaneous neural network activity was recorded using 24-well multielectrode arrays (MEAs) (Multichannel systems, MCS GmbH, Reutlingen, Germany) embedded with 12 gold electrodes with a diameter of 30 μm, spaced 300 µm apart as described⁵⁷ with some modifications. The wells were coated with Poly-L-Lysine and neurons were plated at 1 × 10⁶ cells/well. Before recording neuronal activity, the plates were equilibrated in the MEA recording chamber for 30 min at 37 °C with 5% CO₂. Neuronal activity was then recorded for 15 min. Recordings were sampled at 20,000 Hz and filtered using a high-pass 2nd order Butterworth filter with a 100 Hz cut-off, and a low-pass 4th order Butterworth filter with a 3500 Hz cut-off. The individual spike detection threshold was set at ±6.0 standard deviations. MEA data was analyzed using Multiwell analyzer software (Multichannel systems) and a MATLAB (The Mathworks, Natick, MA, USA) code to extract parameters describing the spontaneous network activity⁹³. The mean firing rate (MFR) (Hz) was calculated for each well individually by averaging the firing rate of each separate electrode by the total amount of active electrodes of the well. Bursts were detected by the Multiwell analyzer built-in burst detection algorithm. Bursts were defined with a maximum of 50 ms inter spike interval (ISI) to start a burst, a maximum of 50 ms ISI to end a burst, with a minimum of 4 spikes per bursts, and a minimum of 100 ms inter burst interval (IBI). Synchronous network bursts were detected when a burst was present in at least 50% of all electrodes. The percentage of random spikes (PRS) was calculated using the percentage of isolated spikes that did not belong to a burst. The IBI, and network burst IBI (NIBI) were calculated by subtraction of the time stamp of the beginning from the time stamp of the ending of each burst or network burst.

Synaptic puncta analysis

Quantification of excitatory synapses was performed as reported⁵⁷. Neurons were stained with dendritic MAP2 (purple) (1:1000), presynaptic SYN1 (red) (1:500) and postsynaptic PSD95 (green) (1:500) markers and imaged using Zeiss Axioplan2 microscope (Carl Zeiss) at ×63 magnification. Fiji software (NIH) was used for quantitative analysis of SYN1/PSD95 co-localisation puncta (yellow) by manual labelling using the multi-point tool. Antibody details are provided in the Supplementary Data 6.

Whole genome sequencing and analysis

gDNA was isolated using High Pure PCR Template Preparation Kit (Roche, 11796828001) as per manufacturer’s instructions. gDNA of each sample was initially quantified by NanoDrop (Thermo Fisher Scientific Inc.) and quality checked for purity and integrity by agarose gel electrophoresis, followed by final quantification using Qubit 2.0 fluorometer. Next generation sequencing libraries were prepared using NEBNext Ultra II DNA Library Prep Kit for Illumina with NEBNext Multiplex Oligos for Illumina (NEB E7645). All library clean-up steps were performed using Agencourt AMPure XP (Beckman Coulter). Paired-End 150 bp sequencing was performed using an Illumina NovaSeq 6000 system according to manufacturer’s instructions. Sequence reads were mapped to GRCm38 with BWA-MEM (0.7.15) and variants were called using GATKv4.1.6. All variants were annotated for allele frequency, locus identity and likely pathogenicity using SnpEff⁹⁴ and filtered for significance using SnpSift⁹⁵. The copy number variations were detected and characterized using CNVnator⁹⁶. The alignments for the sequence reads were viewed with the Integrative Genomics Viewer (IGV) v.2.14.0.

RNA sequencing and analysis

Total RNAs were extracted from the E14.5, E18.5 and P10 mouse brain dorsal cortex (we refer to it as ‘brain’ in the text for simplicity) by in-house protocol that uses combination of TRIzol Reagent (Invitrogen) and RNeasy Mini Kit (Qiagen) and quantified and qualified by NanoDrop (Thermo Fisher Scientific Inc.), Agilent 2100 Bioanalyzer (Agilent Technologies, USA), and 1% agarose gel. 1 μg total RNA from each sample with RNA integrity number (RIN) value ≥ 7 was used for library preparation. Next generation sequencing libraries were prepared using VAHTS Universal V8 RNA-seq Library Prep Kit for Illumina as per manufacturer’s protocol (Vazyme Biotech). Briefly, the poly(A) mRNAs were isolated using NEBNext Poly(A) mRNA Magnetic Isolation Module (NEB). The mRNA fragmentation and priming were performed using NEBNext First Strand Synthesis Reaction Buffer and Random Primer mix. The first strand cDNA was synthesized using NEB ProtoScript II Reverse Transcriptase and the second-strand cDNA using Second Strand Synthesis Enzyme Mix. The double-stranded cDNAs were purified using AxyPrep Mag PCR Clean-up (Axygen) and treated with NEBNext Ultra End Repair/dA-Tailing mix to convert fragmented DNA to repaired DNA and add dA-tails to both ends, followed by T-A ligation using NEBNxt Ultra II Ligation Module to add adaptors to both the ends. The adaptor-ligated DNAs were size selected for fragments of ~360 bp (with ~300 bp inserts) using AxyPrep Mag PCR Clean-Up kit (Axygen). Each sample was amplified by PCR for 11 cycles using P5 and P7 primers; both primers carried sequences that can anneal with flow cell to perform bridge PCR and P7 primer with a six-base index allowing for multiplexing. The PCR products were cleaned up using AxyPrep Mag PCR Clean-Up Kit (Axygen), validated using an Agilent 2100 Bioanalyzer (Agilent Technologies, USA), and quantified by Qubit 2.0 Fluorometer (Invitrogen, USA). The libraries with different indices were multiplexed and loaded on an Illumina NovaSeq 6000 (Illumina, USA). Strand-specific sequencing at 80 million 2 × 150 bp paired-end reads per sample was carried out. Sequence read pairs from each sample were quantified against the mouse transcripts using Salmon v1.6.0⁹⁷ with prebuilt indexes for the mouse mm10 genome build available from RefGenie (http://refgenomes.databio.org/v3/assets/splash/0f10d83b1050c08dd53189986f60970b92a315aa7a16a6f1/salmon_sa_index?tag=default)⁹⁸. Differential gene and transcript expression between Thoc2^+/Y and Thoc2^Δ/Y genotypes were assessed at each developmental time point separately (E14.5, E18.5 and P10) using the default options for the glmQLFit function on transcripts that were included after using the filterByExpr function from the edgeR v 3.40.2 package in the R v 4.2.2 environment⁹⁹. Lists of expressed genes or transcripts not excluded by the filterByExpr function at each time point were used as the background in subsequent gene-ontology analyses. Gene ontology term enrichments were analysed using the g:GOSt functional profiling tool available from the g:Profiler web service (https://biit.cs.ut.ee/gprofiler/gost)³⁹. ENSEMBL mouse gene ID were submitted as an ordered query ranked by ascending p-values to a cut off value < 0.01 for each time point. Significantly enriched terms (FDR < 0.05) identified from g:GOSt analysis were then clustered using the REVIGO web service (http://revigo.irb.hr/)¹⁰⁰. The percent GC content of the differentially expressed genes were compared to the mouse genome using ShinyGO v0.76¹⁰¹.

Total brain proteome analysis using LC-MS

Embryonic day 18.5 and adult Day 10 mouse brains were lysed in 500 µl 0.1 M triethylammonium bicarbonate (TEAB) using the Precellys tube homogeniser. The cell lysates were heat-treated for 5 min at 95 °C and further homogenized using a sonicator probe at 4 °C (80% output). Samples were centrifuged at 13,000 × g for 10 min at 4 °C to remove cell debris. Protein concentration in the supernatants was estimated using the Nanodrop (Thermo Fisher Scientific). A total of 100 µg protein from each sample was reduced using 10 mM tris(2-carboxyethyl)phosphine (TCEP) with 0.05% SDS, alkylated with 200 mM methyl methanethiosulfonate (Sigma Aldrich, Australia) and digested overnight with 50:1 trypsin (Pierce Thermo Fischer Scientific, Australia) at 37 °C. Trypsin digestion was stopped by incubation with 10% trifluoroacetic acid (TFA) at 37 °C for 15 min followed by centrifugation at 13,000 × g for 10 min at 4 °C The supernatants containing digested peptides were dried down and dissolved in 50 µl 0.1% trifluoroacetic acid and desalted using home-made POROS R3 (Life Technologies, Mulgrave, Victoria, Australia) micro-column. The clean peptides were then used for LC-MS/MS.

The peptide samples (200 ng per sample) were separated by nano-liquid chromatography (nLC) using the Dionex RSLC 3500 system (Thermo Fischer, Australia) coupled online to the trapped-ion mobility spectrometry Time-of-Flight (timsTOF) Pro mass spectrometer (Bruker Daltonics, Germany) using the Data Independent Acquisition-Parallel Accumulation Serial Fragmentation (DIA-PASEF) mode. Reversed-phase chromatography was done using a 25 cm, 75 µm ID Aurora C18 nano column (IonOpticks, Australia) with an integrated captive spray emitter. The peptides were eluted using a 125-minute gradient from 0% to 35% buffer B (acetonitrile in 0.1% formic acid) at a rate of 400 nL/minute. Buffer A consisted of 0.1% (v/v) formic acid.

The raw data files from each sample generated on the timsTOF Pro mass spectrometer were processed using the software package Spectronaut™ v16.3 (Biognosys)¹⁰². directDIA™ analysis was performed and the data were searched against the Uniprot Mus musculus FASTA database (17,056 entries; Year 2021). The following additional parameters were applied: variable modifications–deamidation (N/Q), oxidation (M); fixed modification—methylthio (C); enzyme—Trypsin/P; missed cleavages—2. All other parameters were set to default values. Proteins with a false discovery rate (FDR) of ≤1% were reported. Gene Ontology (GO) enrichment analysis was performed using Enrichr web-platform^103,104. For the Enricher plots, p value values for function clustering and multiple detection corrections were calculated, and the plots were sorted based on the Benjamini-Hochberg corrected p value represented as the length of the GO term bars. Volcano plots were generated using R-studio¹⁰⁵. Venn diagrams were generated using jvenn online platform¹⁰⁶. The percent GC content of the genes corresponding to dysregulated proteins were compared to the mouse genome using ShinyGO v0.76¹⁰¹.

Behavioural testing

Morris water maze

This test was used for assessing spatial and related forms of learning and memory. The 1.2 m diameter water maze pool was filled to a depth of 40 cm with 22 °C water, leaving 15 cm diameter submerged platform 1 cm below the water level. Approximately 500 ml of non-toxic white paint was added to the water to make it opaque and hide the platform from the animal. It also served as a contrast agent for tracking the dark-coloured mice. The mice were placed in the water (one at a time) at one of the cardinal points: North (N), South (S), East (E) and West (W) that created imaginary four equal quadrants and allowed 2 min to find the platform. If the mouse found the platform within this time it was allowed 10 sec on the platform before being removed, gently towelled down, and placed under a warming lamp. If the mouse failed to find the platform within the 2 min time window, it was guided to the platform and left for 10 sec on the platform before being removed, towelled down, and placed under a warming lamp. The routine was repeated four times per day, with an interval time of 20–30 min between swims, for 5 days. A probe trial was run at the end of reference memory trials to assess working memory. Video capture and quantification was performed using the CleverSys TopScan tracking software.

Y-maze

The Y-Maze (SD instruments), used to assess willingness to explore new environment, had three identical arms of 30 cm in length, which were symmetrical to each other. The walls of the apparatus were 14 cm high. For the first trial, the mouse was placed at the end of the home arm, facing away from the centre, and allowed to explore 2 of the 3 Y-Maze arms for a 10-minute period. A partition blocking off the novel arm of the maze was in place during this initial trial (Trial 1). Each of the three arms were marked by a unique cue attached to the end to differentiate it from the others. The cues used in the present study were 2-dimensional pictures of black and white symbols including a circle, stripes, and triangles. Two hours later, the test was repeated with the partition removed so that all arms, including the novel arm, were available to explore (Trial 2). The time spent in each of the three accessible arms was video recorded during 5-minute trial, and the data, including arm entries and time spent in each arm, were electronically recorded. Analysis was performed using TopScan rodent tracking software. Mice typically prefer to investigate the novel arm in the maze rather than returning to one that has been previously visited. Therefore, mice with intact working memory should remember not having visited the novel arm and should spend a greater proportion of time exploring that arm relative to the other arms, a parameter termed as spontaneous alteration and utilized for assessing the working memory of mice.

Barnes maze

The Barnes maze test, used for testing spatial learning, spatial reference memory and executive function, was performed using a modified method¹⁰⁷. The maze is an elevated, open circular platform with holes evenly distributed around the periphery. One of the holes is pre-determined as the escape hole, with a black escape box placed below the hole. The test was performed over the course of five days: three days of acquisition trials, a rest day, and a probe day. During the trials, animals were subjected to two trials spaced 15–20 min apart. Mice were placed in the centre of the Barnes maze in a brightly lit room and the time taken by the animals to find and enter the escape box recorded. On day 5, two trials were conducted 30 min apart. In both trials, the time taken to locate and enter both the location of the old escape box, as well as a novel escape box location, were recorded. Data was recorded using the ANY-maze Video Tracking System version 4.99 m (Stoelting Co.).

Rotarod test

This test was used for testing balance, grip-strength, and motor coordination. Mice were pre-trained on the rotarod (Ugo Basile, Italy) for three 5-minute trials. The first two trials were run at a constant speed of 4 rpm, and the 3rd one at an accelerating speed ranging from 4–40 rpm over the 5-minute period. The testing phase was carried out on the next day and involved three trials. Each trial involved the accelerating setting (4–40 rpm) across the 5-minute trial. The time a mouse stayed on the rotating rod before falling was taken as a measure of motor co-ordination. The interval between trials was 25–30 min.

DigiGait test

This test was used for in-depth assessment of gait. The mouse was placed inside the testing chamber positioned above a clear Perspex treadmill belt. The treadmill was started at a low speed (usually 10 cm/sec) until the animal was comfortable with walking on the treadmill. The speed was then slowly increased up to 25 cm/sec (depending on the motor functionality of the mice) and recordings of ventral view of mice were made using a high-speed camera positioned below. Approximately 15 steps were captured. The mouse was then returned to its home cage, with the instrument automatically computing the parameters of interest, such as Swing duration of the paws, Braking time, Propulsion time, and Stride length.

Open field test

The open field test was used to assess locomotor activity and anxiety, based on the innate but conflicting tendencies of mice to avoid bright light, and open spaces and engage in the exploration of a novel space. Mice were placed into a brightly lit open field (1 m × 1 m, grey arena) for 10 min. Open field activity was recorded and analysed using CleverSys tracking software. Anxiety-like behaviour was assessed by examining the time that mice spent in the perimeter relative to the centre of the maze.

Beam walking

This test was used for assessing fine motor coordination and balance in mice. A black acrylic beam 1 m in length and tapering in width from 3.5 cm to 0.5 cm was elevated ~50 cm above a padded cushion. Animals were placed on the wider end of the beam and required to walk up the incline to the narrow end, where an enclosed box was positioned. Animals underwent a training comprised of 3 trials at least 48 hr before testing. Mice were subjected to three consecutive test trials with a 30-minute interval between trials. The time to traverse the beam was recorded and test trials were filmed on both sides from the escape box end for retrospective analysis of foot slips. Recordings were analysed by a single observer blinded to the treatment condition. The results were averaged between the three test trials.

Pasta manipulation (Capellini Handling Test)

The test quantitatively measures forepaw dexterity and was performed as reported³⁷. In brief, capellini pasta pieces (1.0 mm diameter) were cut to 2.6 cm lengths and marked at 0.86 cm increments to allow for visualisation of pasta strand movement during testing. Animals were given capellini pasta in their home cage to eat for 7 days prior to the test and were simultaneously habituated in the plexiglass testing chamber for up 10 min each day. On the day prior to test day, food was removed from the home cage overnight (~10–12 hrs). On the test day, mice were placed in the Plexiglass cage, on top of and surrounded by mirrors, for 5 min. Video cameras, with a high frame rate and ×16 zoom lens, were set up on both sides to capture the reflection of the paws of mice in the mirror while they eat/manipulate the pasta. After 5 min of habituation, a piece of pasta was dropped into the cage and recording continued until a maximum of 3 pasta were consumed. The trial length was 10 min. Parameters such as average eating time per pasta, no of times pasta dropped, paw adjustments while eating and atypical behaviors were analysed from the captured video.

Adhesive removal test

This test was performed as described earlier³⁸ to assess sensorimotor function in the mice. During the experiments, a piece of adhesive tape (3 mm × 3 mm) was gently placed on the forepaw by using blunt-end forceps. Sensorimotor performance was investigated by measuring the latency to contact the adhesive (time-to-contact) and the latency to remove the adhesive from each paw (time-to-remove). The maximum trial length was 120 sec. Trials were conducted one paw per trial, two trials for each paw, for a total of four trials, and latency was recorded and averaged across the trials.

Elevated plus maze

The elevated plus maze test was used to measure anxiety. The maze was comprised of two open arms and two closed arms that extended from a common central platform elevated to a height of 40 cm above the floor. The closed arms provided protection via a 15 cm high wall with a passageway 4.5 cm wide. The open arms had no walls but were of the same width as the closed arms. Mice were placed on the centre square, facing an open arm, and allowed to freely explore the apparatus for a 10-minute period. Video capture and analysis for parameters including number of arm entries and time spent in the closed versus open arms was performed using the CleverSys TopScan tracking software.

Light–dark test

The light-dark test measures anxiety-like behaviour and was conducted in the TruScan locomotor system (Coulbourn Instruments, USA). The locomotor cells were 40.6 cm wide 40.6 cm deep × 40.6 cm high, encased by twin photo-optic arrays (sensor rings), set up to quantify and qualify movements in the horizontal plane. A black Perspex insert was used to create an area in which half the cell is dark, and half is light (750 lux). The mice were placed in the dark side and left to wander freely for 10 min. The photocells in the sensor rings measure movement from the dark side to the light side and back, and the amount of time spent in each zone was calculated by the TruScan software. The light/dark exploration test measures the conflict between the natural tendencies of a mouse to explore a novel environment versus their tendency to avoid a brightly lit open field. More anxious animals spend less time in the lit area.

Grip strength test

A grip strength meter (Ugo Basile, Italy) was used to test the forelimb grip strength of the mice. The technique relies on the inherent resistance of mice to a backward movement due to which they tend to grasp/grip any available object to counter against the pull-back applied by the researcher. The mouse (one at a time) was moved close to the T-shaped bar with its torso horizontal and only forepaws touching the bar. The mouse was then gently pulled backwards by its tail making sure that it gripped the frontal portion of the T-bar. The digital meter recorded the peak force applied by the mouse while grasping the T-bar. Three trials were performed per animal to generate the average grip strength measurement data.

Statistics and reproducibility

All results are presented as the mean values ± SEM with details of experimental and biological replicates provided in the figure legends. Comparisons between two groups were analysed by Student’s t test and multiple-groups by one/two-way ANOVA using GraphPad Prism 9. In all box plots, the boundary of the box closest to zero indicates the 25th percentile and the boundary farthest from zero indicates the 75th percentile. A line within the box shows the median, and the whiskers above and below the box indicate the maximum and minimum points of the dataset, respectively.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The whole genome sequencing and RNA-Seq data generated in this study have been deposited in NCBI’s Gene Expression Omnibus and are accessible through GEO Series accession number GSE245539. The proteomics data generated in this study have been deposited to the ProteomeXchange Consortium (http://proteomecentral.proteomexchange.org) via the PRIDE partner repository with the dataset identifier PXD040358. All remaining data supporting the findings in this study are provided in the manuscript files and/or the Supplementary Information files. Source data are provided with this paper.

References

Amberger, J. S., Bocchini, C. A., Schiettecatte, F., Scott, A. F. & Hamosh, A. OMIM.org: online mendelian inheritance in man (OMIM(R)), an online catalog of human genes and genetic disorders. Nucleic Acids Res. 43, D789–D798 (2015).
Article PubMed Google Scholar
Ozlu, C., Bailey, R. M., Sinnett, S. & Goodspeed, K. D. Gene transfer therapy for neurodevelopmental disorders. Dev. Neurosci. 43, 230–240 (2021).
Article CAS PubMed Google Scholar
Kumar, R. et al. THOC2 mutations implicate mRNA-export pathway in X-linked intellectual disability. Am. J. Hum. Genet. 97, 302–310 (2015).
Article CAS PubMed PubMed Central Google Scholar
Kumar, R. et al. Severe neurocognitive and growth disorders due to variation in THOC2, an essential component of nuclear mRNA export machinery. Hum. Mutat. 39, 1126–1138 (2018).
Article CAS PubMed PubMed Central Google Scholar
Kumar, R. et al. Expanding clinical presentations due to variations in THOC2 mRNA nuclear export factor. Front. Mol. Neurosci. 13, 12 (2020).
Article CAS PubMed PubMed Central Google Scholar
Wang, L. et al. The THO complex regulates pluripotency gene mRNA export and controls embryonic stem cell self-renewal and somatic cell reprogramming. Cell Stem Cell 13, 676–690 (2013).
Article CAS PubMed PubMed Central Google Scholar
Heath, C. G., Viphakone, N. & Wilson, S. A. The role of TREX in gene expression and disease. Biochem. J. 473, 2911–2935 (2016).
Article CAS PubMed Google Scholar
Folco, E. G., Lee, C. S., Dufu, K., Yamazaki, T. & Reed, R. The proteins PDIP3 and ZC11A associate with the human TREX complex in an ATP-dependent manner and function in mRNA export. PLoS One 7, e43804 (2012).
Article ADS CAS PubMed PubMed Central Google Scholar
Viphakone, N. et al. Luzp4 defines a new mRNA export pathway in cancer cells. Nucleic Acids Res. 43, 2353–2366 (2015).
Article CAS PubMed PubMed Central Google Scholar
Pena, A. et al. Architecture and nucleic acids recognition mechanism of the THO complex, an mRNP assembly factor. EMBO J. 31, 1605–1616 (2012).
Article CAS PubMed PubMed Central Google Scholar
Puhringer, T. et al. Structure of the human core transcription-export complex reveals a hub for multivalent interactions. Elife 9, e61503 (2020).
Article PubMed PubMed Central Google Scholar
Katahira, J. & Yoneda, Y. Roles of the TREX complex in nuclear export of mRNA. RNA Biol. 6, 149–152 (2009).
Article CAS PubMed Google Scholar
Meinel, D. M. et al. Recruitment of TREX to the transcription machinery by its direct binding to the phospho-CTD of RNA polymerase II. PLoS Genet. 9, e1003914 (2013).
Article PubMed PubMed Central Google Scholar
Cai, S. et al. Knockdown of THOC1 reduces the proliferation of hepatocellular carcinoma and increases the sensitivity to cisplatin. J. Exp. Clin. Cancer Res. 39, 135 (2020).
Article CAS PubMed PubMed Central Google Scholar
Castellano-Pozo, M., Garcia-Muse, T. & Aguilera, A. R-loops cause replication impairment and genome instability during meiosis. EMBO Rep. 13, 923–929 (2012).
Article CAS PubMed PubMed Central Google Scholar
Salas-Armenteros, I. et al. Human THO-Sin3A interaction reveals new mechanisms to prevent R-loops that cause genome instability. EMBO J. 36, 3532–3547 (2017).
Article CAS PubMed PubMed Central Google Scholar
Gomez-Gonzalez, B. et al. Genome-wide function of THO/TREX in active genes prevents R-loop-dependent replication obstacles. EMBO J. 30, 3106–3119 (2011).
Article CAS PubMed PubMed Central Google Scholar
Castellano-Pozo, M., Garcia-Muse, T. & Aguilera, A. The Caenorhabditis elegans THO complex is required for the mitotic cell cycle and development. PLoS One 7, e52447 (2012).
Article ADS CAS PubMed PubMed Central Google Scholar
Dominguez-Sanchez, M. S., Barroso, S., Gomez-Gonzalez, B., Luna, R. & Aguilera, A. Genome instability and transcription elongation impairment in human cells depleted of THO/TREX. PLoS Genet. 7, e1002386 (2011).
Article CAS PubMed PubMed Central Google Scholar
Yamazaki, T. et al. The closely related RNA helicases, UAP56 and URH49, preferentially form distinct mRNA export machineries and coordinately regulate mitotic progression. Mol. Biol. Cell 21, 2953–2965 (2010).
Article CAS PubMed PubMed Central Google Scholar
Rehwinkel, J. et al. Genome-wide analysis of mRNAs regulated by the THO complex in Drosophila melanogaster. Nat. Struct. Mol. Biol. 11, 558–566 (2004).
Article CAS PubMed Google Scholar
Amsterdam, A. et al. Identification of 315 genes essential for early zebrafish development. Proc. Natl. Acad. Sci. USA 101, 12792–12797 (2004).
Article ADS CAS PubMed PubMed Central Google Scholar
Dickinson, M. E. et al. High-throughput discovery of novel developmental phenotypes. Nature 537, 508–514 (2016).
Article CAS PubMed PubMed Central Google Scholar
Chi, B. et al. Aly and THO are required for assembly of the human TREX complex and association of TREX components with the spliced mRNA. Nucleic Acids Res. 41, 1294–1306 (2013).
Article CAS PubMed Google Scholar
Di Gregorio, E. et al. A de novo X;8 translocation creates a PTK2-THOC2 gene fusion with THOC2 expression knockdown in a patient with psychomotor retardation and congenital cerebellar hypoplasia. J. Med. Genet. 50, 543–551 (2013).
Article PubMed Google Scholar
Woerner, A. C. et al. Cytoplasmic protein aggregates interfere with nucleocytoplasmic transport of protein and RNA. Science 351, 173–176 (2016).
Article ADS CAS PubMed Google Scholar
Wang, T. et al. Identification and characterization of essential genes in the human genome. Science 350, 1096–1101 (2015).
Article ADS CAS PubMed PubMed Central Google Scholar
Cardoso-Moreira, M. et al. Gene expression across mammalian organ development. Nature 571, 505–509 (2019).
Article CAS PubMed PubMed Central Google Scholar
Morris, R. G., Garrud, P., Rawlins, J. N. & O’Keefe, J. Place navigation impaired in rats with hippocampal lesions. Nature 297, 681–683 (1982).
Article ADS CAS PubMed Google Scholar
Barnes, C. A. Memory deficits associated with senescence: a neurophysiological and behavioral study in the rat. J. Comp. Physiol. Psychol. 93, 74–104 (1979).
Article CAS PubMed Google Scholar
Kraeuter, A. K., Guest, P. C. & Sarnyai, Z. The Y-maze for assessment of spatial working and reference memory in mice. Methods Mol. Biol. 1916, 105–111 (2019).
Article CAS PubMed Google Scholar
Shiotsuki, H. et al. A rotarod test for evaluation of motor skill learning. J. Neurosci. Methods 189, 180–185 (2010).
Article PubMed Google Scholar
Sashindranath, M., Daglas, M. & Medcalf, R. L. Evaluation of gait impairment in mice subjected to craniotomy and traumatic brain injury. Behav Brain Res. 286, 33–38 (2015).
Article CAS PubMed Google Scholar
Carola, V., D’Olimpio, F., Brunamonti, E., Mangia, F. & Renzi, P. Evaluation of the elevated plus-maze and open-field tests for the assessment of anxiety-related behaviour in inbred mice. Behav Brain Res. 134, 49–57 (2002).
Article PubMed Google Scholar
Bourin, M. & Hascoet, M. The mouse light/dark box test. Eur. J. Pharmacol. 463, 55–65 (2003).
Article CAS PubMed Google Scholar
Luong, T. N., Carlisle, H. J., Southwell, A. & Patterson, P. H. Assessment of motor balance and coordination in mice using the balance beam. J. Vis. Exp. 49, 2376 (2011).
Tennant, K. A. et al. The vermicelli and capellini handling tests: simple quantitative measures of dexterous forepaw function in rats and mice. J. Vis. Exp. 41, 2076 (2010).
Bouet, V. et al. The adhesive removal test: a sensitive method to assess sensorimotor deficits in mice. Nat. Protoc. 4, 1560–1564 (2009).
Article CAS PubMed Google Scholar
Raudvere, U. et al. g:Profiler: a web server for functional enrichment analysis and conversions of gene lists (2019 update). Nucleic Acids Res. 47, W191–W198 (2019).
Article CAS PubMed PubMed Central Google Scholar
Kochinke, K. et al. Systematic phenomics analysis deconvolutes genes mutated in intellectual disability into biologically coherent modules. Am. J. Hum. Genet. 98, 149–164 (2016).
Article CAS PubMed PubMed Central Google Scholar
Toma, K., Wang, T. C. & Hanashima, C. Encoding and decoding time in neural development. Dev. Growth Differ. 58, 59–72 (2016).
Article PubMed Google Scholar
Martinez-Cerdeno, V. & Noctor, S. C. Neural progenitor cell terminology. Front. Neuroanat. 12, 104 (2018).
Article CAS PubMed PubMed Central Google Scholar
Liszewska, E. & Jaworski, J. Neural stem cell dysfunction in human brain disorders. Results Probl. Cell Differ. 66, 283–305 (2018).
Article CAS PubMed Google Scholar
Moortgat, S. et al. HUWE1 variants cause dominant X-linked intellectual disability: a clinical study of 21 patients. Eur. J. Hum. Genet. 26, 64–74 (2018).
Article CAS PubMed Google Scholar
Stark, G. R. & Taylor, W. R. Analyzing the G2/M checkpoint. Methods Mol. Biol. 280, 51–82 (2004).
CAS PubMed Google Scholar
Mah, L. J., El-Osta, A. & Karagiannis, T. C. gammaH2AX: a sensitive molecular marker of DNA damage and repair. Leukemia 24, 679–686 (2010).
Article CAS PubMed Google Scholar
Lu, Y., Liu, Y. & Yang, C. Evaluating in vitro DNA damage using comet assay. J. Vis. Exp. 128, 56450 (2017).
Allison, D. F. & Wang, G. G. R-loops: formation, function, and relevance to cell stress. Cell Stress 3, 38–46 (2019).
Article CAS PubMed PubMed Central Google Scholar
Hicks, T. et al. R-loop-induced irreparable DNA damage evades checkpoint detection in the C. elegans germline. Nucleic Acids Res. 50, 8041–8059 (2022).
Article CAS PubMed PubMed Central Google Scholar
Marabitti, V. et al. ATM pathway activation limits R-loop-associated genomic instability in Werner syndrome cells. Nucleic Acids Res. 47, 3485–3502 (2019).
Article CAS PubMed PubMed Central Google Scholar
Ramirez, P., Crouch, R. J., Cheung, V. G. & Grunseich, C. R-loop analysis by dot-blot. J. Vis. Exp. 167, 62069 (2021).
Homan, C. C. et al. Mutations in USP9X are associated with X-linked intellectual disability and disrupt neuronal cell migration and growth. Am. J. Hum. Genet. 94, 470–478 (2014).
Article CAS PubMed PubMed Central Google Scholar
Kato, M. Genotype-phenotype correlation in neuronal migration disorders and cortical dysplasias. Front. Neurosci. 9, 181 (2015).
Article PubMed PubMed Central Google Scholar
Durbec, P., Franceschini, I., Lazarini, F. & Dubois-Dalcq, M. In vitro migration assays of neural stem cells. Methods Mol. Biol. 438, 213–225 (2008).
Article CAS PubMed Google Scholar
Hardwick, L. J., Ali, F. R., Azzarelli, R. & Philpott, A. Cell cycle regulation of proliferation versus differentiation in the central nervous system. Cell Tissue Res. 359, 187–200 (2015).
Article CAS PubMed Google Scholar
Jolly, L. A., Homan, C. C., Jacob, R., Barry, S. & Gecz, J. The UPF3B gene, implicated in intellectual disability, autism, ADHD and childhood onset schizophrenia regulates neural progenitor cell behaviour and neuronal outgrowth. Hum. Mol. Genet. 22, 4673–4687 (2013).
Article CAS PubMed Google Scholar
Mincheva-Tasheva, S., Nieto Guil, A. F., Homan, C. C., Gecz, J. & Thomas, P. Q. Disrupted excitatory synaptic contacts and altered neuronal network activity underpins the neurological phenotype in PCDH19-clustering epilepsy (PCDH19-CE). Mol. Neurobiol. 58, 2005–2018 (2021).
Article CAS PubMed Google Scholar
Rondon, A. G., Jimeno, S., Garcia-Rubio, M. & Aguilera, A. Molecular evidence that the eukaryotic THO/TREX complex is required for efficient transcription elongation. J. Biol. Chem. 278, 39037–39043 (2003).
Article CAS PubMed Google Scholar
Bai, X. et al. THOC2 and THOC5 regulate stemness and radioresistance in triple-negative breast cancer. Adv. Sci. (Weinh) 8, e2102658 (2021).
Article PubMed Google Scholar
Gu, X. J., Li, Y. J., Wang, F. & Ye, T. MiR-30e-3p inhibits gastric cancer development by negatively regulating THO complex 2 and PI3K/AKT/mTOR signaling. World J. Gastrointest. Oncol. 14, 2170–2182 (2022).
Article PubMed PubMed Central Google Scholar
Mattioli, F. et al. Clinical and functional characterization of recurrent missense variants implicated in THOC6-related intellectual disability. Hum. Mol. Genet. 28, 952–960 (2019).
Article CAS PubMed Google Scholar
Zhou, X. et al. Knockdown THOC2 suppresses the proliferation and invasion of melanoma. Bioengineered 10, 635–645 (2019).
Article CAS PubMed PubMed Central Google Scholar
Alvarez-Mora, M. I. et al. Diagnostic yield of next-generation sequencing in 87 families with neurodevelopmental disorders. Orphanet. J. Rare Dis. 17, 60 (2022).
Article PubMed PubMed Central Google Scholar
Kamp, J. A. et al. THO complex deficiency impairs DNA double-strand break repair via the RNA surveillance kinase SMG-1. Nucleic Acids Res. 50, 6235–6250 (2022).
Article CAS PubMed PubMed Central Google Scholar
Guria, A. et al. Identification of mRNAs that are spliced but not exported to the cytoplasm in the absence of THOC5 in mouse embryo fibroblasts. RNA 17, 1048–1056 (2011).
Article CAS PubMed PubMed Central Google Scholar
Tamhankar, V., Tamhankar, P., Chaubal, R., Chaubal, J. & Chaubal, N. Novel consensus splice site pathogenic variation in THOC2 gene leads to recurrent arthrogryposis multiplex congenita phenotype: a case report. Cureus 13, e19682 (2021).
PubMed PubMed Central Google Scholar
Ye, Z., Bing, A., Zhao, S., Yi, S. & Zhan, X. Comprehensive analysis of spliceosome genes and their mutants across 27 cancer types in 9070 patients: clinically relevant outcomes in the context of 3P medicine. EPMA J. 13, 335–350 (2022).
Article PubMed PubMed Central Google Scholar
Yuan, X. et al. THO complex-dependent posttranscriptional control contributes to vascular smooth muscle cell fate decision. Circ. Res. 123, 538–549 (2018).
Article CAS PubMed Google Scholar
Mancini, A. et al. THOC5/FMIP, an mRNA export TREX complex protein, is essential for hematopoietic primitive cell survival in vivo. BMC Biol. 8, 1 (2010).
Article PubMed PubMed Central Google Scholar
Pitzonka, L. et al. The THO ribonucleoprotein complex is required for stem cell homeostasis in the adult mouse small intestine. Mol. Cell Biol. 33, 3505–3514 (2013).
Article CAS PubMed PubMed Central Google Scholar
Verma, V., Paul, A., Amrapali Vishwanath, A., Vaidya, B. & Clement, J. P. Understanding intellectual disability and autism spectrum disorders from common mouse models: synapses to behaviour. Open Biol. 9, 180265 (2019).
Article CAS PubMed PubMed Central Google Scholar
Bourgeron, T. A synaptic trek to autism. Curr. Opin. Neurobiol. 19, 231–234 (2009).
Article CAS PubMed Google Scholar
Robbins, E. M. et al. SynCAM 1 adhesion dynamically regulates synapse number and impacts plasticity and learning. Neuron 68, 894–906 (2010).
Article CAS PubMed PubMed Central Google Scholar
Lavi, A., Perez, O. & Ashery, U. Shaping neuronal network activity by presynaptic mechanisms. PLoS Comput. Biol. 11, e1004438 (2015).
Article ADS PubMed PubMed Central Google Scholar
Lavi, A., Sheinin, A., Shapira, R., Zelmanoff, D. & Ashery, U. DOC2B and Munc13-1 differentially regulate neuronal network activity. Cereb. Cortex 24, 2309–2323 (2014).
Article PubMed Google Scholar
Kim, S. Y. et al. Global transcriptional downregulation of TREX and nuclear trafficking machinery as pan-senescence phenomena: evidence from human cells and tissues. Exp. Mol. Med. 52, 1351–1359 (2020).
Article CAS PubMed PubMed Central Google Scholar
Luna, R., Rondon, A. G., Perez-Calero, C., Salas-Armenteros, I. & Aguilera, A. The THO complex as a paradigm for the prevention of cotranscriptional R-loops. Cold Spring Harb. Symp. Quant. Biol. 84, 105–114 (2019).
Article PubMed Google Scholar
Mosler, T. et al. R-loop proximity proteomics identifies a role of DDX41 in transcription-associated genomic instability. Nat. Commun. 12, 7314 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Skourti-Stathaki, K. & Proudfoot, N. J. A double-edged sword: R loops as threats to genome integrity and powerful regulators of gene expression. Genes Dev. 28, 1384–1396 (2014).
Article CAS PubMed PubMed Central Google Scholar
Crossley, M. P., Bocek, M. & Cimprich, K. A. R-loops as cellular regulators and genomic threats. Mol. Cell 73, 398–411 (2019).
Article CAS PubMed PubMed Central Google Scholar
Cuartas, J. & Gangwani, L. R-loop mediated DNA damage and impaired DNA repair in spinal muscular atrophy. Front. Cell Neurosci. 16, 826608 (2022).
Article CAS PubMed PubMed Central Google Scholar
Bunting, M. D. et al. Generation of gene drive mice for invasive pest population suppression. Methods Mol. Biol. 2495, 203–230 (2022).
Article PubMed Google Scholar
Kito, S. et al. Improved in vitro fertilization and development by use of modified human tubal fluid and applicability of pronucleate embryos for cryopreservation by rapid freezing in inbred mice. Comp. Med. 54, 564–570 (2004).
CAS PubMed Google Scholar
Kaech, S. & Banker, G. Culturing hippocampal neurons. Nat. Protoc. 1, 2406–2415 (2006).
Article CAS PubMed Google Scholar
Iggo, R. Lentiviral transduction of mammary epithelial cells. Methods Mol. Biol. 2471, 159–183 (2022).
Article CAS PubMed Google Scholar
Douglass, M. L. et al. Is SGSH heterozygosity a risk factor for early-onset neurodegenerative disease? J. Inherit. Metab. Dis. 44, 763–776 (2021).
Article CAS PubMed Google Scholar
Garcia-Rubio, M., Barroso, S. I. & Aguilera, A. Detection of DNA-RNA hybrids in vivo. Methods Mol. Biol. 1672, 347–361 (2018).
Article CAS PubMed Google Scholar
Smolka, J. A., Sanz, L. A., Hartono, S. R. & Chedin, F. Recognition of RNA by the S9.6 antibody creates pervasive artifacts when imaging RNA:DNA hybrids. J. Cell Biol. 220, e202004079 (2021).
Article CAS PubMed PubMed Central Google Scholar
Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat. Methods 9, 676–682 (2012).
Article CAS PubMed Google Scholar
Dias, A. P., Dufu, K., Lei, H. & Reed, R. A role for TREX components in the release of spliced mRNA from nuclear speckle domains. Nat. Commun. 1, 97 (2010).
Article ADS PubMed Google Scholar
Mrozik, K. M. et al. LCRF-0006, a small molecule mimetic of the N-cadherin antagonist peptide ADH-1, synergistically increases multiple myeloma response to bortezomib. FASEB Bioadv. 2, 339–353 (2020).
Article CAS PubMed PubMed Central Google Scholar
Konca, K. et al. A cross-platform public domain PC image-analysis program for the comet assay. Mutat. Res. 534, 15–20 (2003).
Article CAS PubMed Google Scholar
Frega, M. et al. Distinct pathogenic genes causing intellectual disability and autism exhibit a common neuronal network hyperactivity phenotype. Cell Rep. 30, 173–186 e6 (2020).
Article ADS CAS PubMed Google Scholar
Cingolani, P. et al. A program for annotating and predicting the effects of single nucleotide polymorphisms, SnpEff: SNPs in the genome of Drosophila melanogaster strain w1118; iso-2; iso-3. Fly (Austin) 6, 80–92 (2012).
Article CAS PubMed Google Scholar
Cingolani, P. et al. Using drosophila melanogaster as a model for genotoxic chemical mutational studies with a new program, SnpSift. Front. Genet. 3, 35 (2012).
Article PubMed PubMed Central Google Scholar
Abyzov, A., Urban, A. E., Snyder, M. & Gerstein, M. CNVnator: an approach to discover, genotype, and characterize typical and atypical CNVs from family and population genome sequencing. Genome Res. 21, 974–984 (2011).
Article CAS PubMed PubMed Central Google Scholar
Patro, R., Duggal, G., Love, M. I., Irizarry, R. A. & Kingsford, C. Salmon provides fast and bias-aware quantification of transcript expression. Nat. Methods 14, 417–419 (2017).
Article CAS PubMed PubMed Central Google Scholar
Stolarczyk, M., Reuter, V. P., Smith, J. P., Magee, N. E. & Sheffield, N. C. Refgenie: a reference genome resource manager. Gigascience 9, giz149 (2020).
Article PubMed PubMed Central Google Scholar
Chen, Y., Lun, A. T. & Smyth, G. K. From reads to genes to pathways: differential expression analysis of RNA-Seq experiments using Rsubread and the edgeR quasi-likelihood pipeline. F1000Res 5, 1438 (2016).
PubMed PubMed Central Google Scholar
Supek, F., Bosnjak, M., Skunca, N. & Smuc, T. REVIGO summarizes and visualizes long lists of gene ontology terms. PLoS One 6, e21800 (2011).
Article ADS CAS PubMed PubMed Central Google Scholar
Ge, S. X., Jung, D. & Yao, R. ShinyGO: a graphical gene-set enrichment tool for animals and plants. Bioinformatics 36, 2628–2629 (2020).
Article CAS PubMed Google Scholar
Bruderer, R. et al. Extending the limits of quantitative proteome profiling with data-independent acquisition and application to acetaminophen-treated three-dimensional liver microtissues. Mol. Cell. Proteomics 14, 1400–1410 (2015).
Article CAS PubMed PubMed Central Google Scholar
Kuleshov, M. V. et al. Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucleic Acids Res. 44, W90–W97 (2016).
Article CAS PubMed PubMed Central Google Scholar
Chen, E. Y. et al. Enrichr: interactive and collaborative HTML5 gene list enrichment analysis tool. BMC Bioinformatics 14, 128 (2013).
Article PubMed PubMed Central Google Scholar
RCoreTeam. R: a language and environment for statistical computing. (R Foundation for Statistical Computing, 2021).
Bardou, P., Mariette, J., Escudie, F., Djemiel, C. & Klopp, C. jvenn: an interactive Venn diagram viewer. BMC Bioinformatics 15, 293 (2014).
Article PubMed PubMed Central Google Scholar
Sunyer, B., Patil, S., Höger, H. & Lubec, G. Barnes maze, a useful task to assess spatial reference memory in the mice. https://assets.researchsquare.com/files/nprot-349/v1/6d50cbea-4031-4bb6-a009-75e4ef58eb5b.pdf (2007).

Download references

Acknowledgements

We would like thank Lottie Servin, Rosanna Duncan, and team from Laboratory Animal Services, The University of Adelaide for helping with mouse colony maintenance, breeding, and health monitoring. Prof. Richard Iggo, INSERM, Bordeaux, France, for his kind gift of pJS65 MND-DEST SV40-Tomato plasmid. Agatha Labrinidis and Jane Sibbons, Adelaide Microscopy, The University of Adelaide, for assisting with microscopy services. Giulia Parodi, Sergio Martinoia’s Bioengineering lab, University of Genova, for generating the raster plots from MEA results. Shreya Agarwala, The University of Adelaide, for generating illustrations of behavioural testing. Tina Cardamone and Aira Nuguid, Phenomics Australia, University of Melbourne, for providing Histopathology and Slide Scanning Service. A/Prof. John Finnie from discipline of Anatomy and Pathology, School of Medicine, University of Adelaide, for helping with histopathology reports. The authors acknowledge the facilities and the scientific and technical assistance of the South Australian Genome Editing (SAGE) Facility, the University of Adelaide, and the South Australian Health and Medical Research Institute. SAGE is supported by Phenomics Australia. Phenomics Australia is supported by the Australian Government through the National Collaborative Research Infrastructure Strategy (NCRIS) program. G.B. is supported by the Angela Wright Bennett Foundation, The McCusker Charitable Foundation via Channel 7 Telethon Trust, The Stan Perron Charitable Foundation and Mineral Resources Limited. J.E.N. was supported by the Veronika Sacco Clinical Cancer Research Fellowship from the Florey Medical Research Foundation. This study was financially supported by the National Health and Medical Research Council of Australia (NHMRC) Fellowship APP1155224 to J.G. and NHMRC project grant APP1163240 to J.G., R.K., and P.Q.T.

Author information

These authors jointly supervised this work: Raman Kumar, Jozef Gecz.

Authors and Affiliations

Adelaide Medical School, The University of Adelaide, Adelaide, SA, 5005, Australia
Rudrarup Bhattacharjee, Mark A. Corbett, Sushma R. Rao, Alison E. Gardner, Tarin Ritchie, Nazzmer Nazri, Clare L. van Eyk, Dani Fornarino, Marten F. Snel, Kim M. Hemsley, Raman Kumar & Jozef Gecz
Robinson Research Institute, The University of Adelaide, Adelaide, SA, 5005, Australia
Rudrarup Bhattacharjee, Lachlan A. Jolly, Mark A. Corbett, Alison E. Gardner, Tarin Ritchie, Sandra Piltz, Clare L. van Eyk, Melissa White, Dani Fornarino, Paul Q. Thomas, Raman Kumar & Jozef Gecz
School of Biomedicine, The University of Adelaide, Adelaide, SA, 5005, Australia
Lachlan A. Jolly, Sandra Piltz, Melissa White & Paul Q. Thomas
Discipline of Anatomy and Pathology, School of Biomedicine, The University of Adelaide, Adelaide, SA, 5005, Australia
Ing Chee Wee & Lyndsey E. Collins-Praino
Proteomics, Metabolomics and MS-imaging Core Facility, South Australian Health and Medical Research Institute, and Adelaide Medical School, The University of Adelaide, Adelaide, SA, 5005, Australia
Sushma R. Rao & Marten F. Snel
Department of Human Genetics, Radboudumc, Donders Institute for Brain, Cognition, and Behavior, Nijmegen, 6500, HB, the Netherlands
Eline J. H. van Hugte, Ummi Ciptasari & Nael Nadif Kasri
South Australian Health and Medical Research Institute, Adelaide, SA, 5000, Australia
Sandra Piltz, Melissa White & Paul Q. Thomas
School of Biomedicine, Faculty of Health and Medical Sciences, University of Adelaide and Precision Cancer Medicine Theme, Solid Tumour Program, South Australian Health and Medical Research Institute, Adelaide, SA, 5000, Australia
Jacqueline E. Noll
Childhood Dementia Research Group, College of Medicine and Public Health, Flinders Health & Medical Research Institute, Flinders University, Bedford Park, Adelaide, SA, 5042, Australia
Nazzmer Nazri & Kim M. Hemsley
Undiagnosed Diseases Program, Genetic Services of WA, King Edward Memorial Hospital, Subiaco, WA, 6008, Australia
Cathryn Poulton & Gareth Baynam
Western Australian Register of Developmental Anomalies, King Edward Memorial Hospital, Subiaco, WA, 6008, Australia
Gareth Baynam
Rare Care Centre, Perth Children’s Hospital, Nedlands, WA, 6009, Australia
Gareth Baynam

Authors

Rudrarup Bhattacharjee
View author publications
You can also search for this author in PubMed Google Scholar
Lachlan A. Jolly
View author publications
You can also search for this author in PubMed Google Scholar
Mark A. Corbett
View author publications
You can also search for this author in PubMed Google Scholar
Ing Chee Wee
View author publications
You can also search for this author in PubMed Google Scholar
Sushma R. Rao
View author publications
You can also search for this author in PubMed Google Scholar
Alison E. Gardner
View author publications
You can also search for this author in PubMed Google Scholar
Tarin Ritchie
View author publications
You can also search for this author in PubMed Google Scholar
Eline J. H. van Hugte
View author publications
You can also search for this author in PubMed Google Scholar
Ummi Ciptasari
View author publications
You can also search for this author in PubMed Google Scholar
Sandra Piltz
View author publications
You can also search for this author in PubMed Google Scholar
Jacqueline E. Noll
View author publications
You can also search for this author in PubMed Google Scholar
Nazzmer Nazri
View author publications
You can also search for this author in PubMed Google Scholar
Clare L. van Eyk
View author publications
You can also search for this author in PubMed Google Scholar
Melissa White
View author publications
You can also search for this author in PubMed Google Scholar
Dani Fornarino
View author publications
You can also search for this author in PubMed Google Scholar
Cathryn Poulton
View author publications
You can also search for this author in PubMed Google Scholar
Gareth Baynam
View author publications
You can also search for this author in PubMed Google Scholar
Lyndsey E. Collins-Praino
View author publications
You can also search for this author in PubMed Google Scholar
Marten F. Snel
View author publications
You can also search for this author in PubMed Google Scholar
Nael Nadif Kasri
View author publications
You can also search for this author in PubMed Google Scholar
Kim M. Hemsley
View author publications
You can also search for this author in PubMed Google Scholar
Paul Q. Thomas
View author publications
You can also search for this author in PubMed Google Scholar
Raman Kumar
View author publications
You can also search for this author in PubMed Google Scholar
Jozef Gecz
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

R.B., R.K., and J.G. designed the study and experiments. R.B. performed the majority of the experiments and analyses with substantial input from R.K. L.A.J. advised on the design and execution of primary neural stem cell and differentiated neuron studies. M.A.C. analysed the RNA-Seq data and helped with its presentation. R.B. and I.C.W. performed the behavioural studies and data analysis with supervision from L.C.P. S.R.R. and M.F.S. contributed to the proteomics experiment and analysis. A.E.G and M.A.C. performed WGS and proteome analyses. T.R. helped with patient-skin fibroblast culture and comet assay. E.J.H.vH. and U.C. performed multielectrode array data analysis with supervision from N.N.K. S.P. performed IVF. N.N. and RB performed Golgi-Cox staining with supervision from K.M.H. C.L.vE advised on RNA-Seq experiments. J.E.N. contributed to flow cytometry experiment and analysis. M.W. designed the ex37-38 del CRISPR/Cas9 editing strategy with supervision from P.Q.T. D.F. designed and helped with RT-qPCR assays. G.B. and C.P. contributed with patient clinical diagnosis and analysis. P.Q.T. supervised IVF, in vivo neurobiology assays, and behavioural testing. R.B., R.K., and J.G. wrote the original draft with input from L.A.J., M.A.C., L.C.P., and P.Q.T. All authors reviewed and edited the manuscript. R.K. and J.G. jointly contributed to overall supervision and administration of the project.

Corresponding author

Correspondence to Jozef Gecz.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks Guillaume Hautbergue and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Peer Review File

Description of Additional Supplementary Files

Supplementary Data 1

Supplementary Data 2

Supplementary Data 3

Supplementary Data 4

Supplementary Data 5

Supplementary Data 6

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Bhattacharjee, R., Jolly, L.A., Corbett, M.A. et al. Compromised transcription-mRNA export factor THOC2 causes R-loop accumulation, DNA damage and adverse neurodevelopment. Nat Commun 15, 1210 (2024). https://doi.org/10.1038/s41467-024-45121-5

Download citation

Received: 16 June 2023
Accepted: 15 January 2024
Published: 08 February 2024
DOI: https://doi.org/10.1038/s41467-024-45121-5

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.