Neurotrophic Modulation Restores Motor and Developmental Defects in Zebrafish Models of ints11 Deficiency
Anna Pistocchi, Elena Chiricozzi, Matilde Molteni, Gaia Galassi, Laura Mauri, Francesca Balistreri, Stefania Magri, Anna Marozzi, Franco Taroni, Alex Pezzotta
TL;DR
This study uses zebrafish to model INTS11 deficiency and shows that neurotrophic treatments can restore motor and developmental defects linked to this genetic disorder.
Contribution
The first in vivo zebrafish model of INTS11-associated neurodevelopmental dysfunction is established, revealing conserved mechanisms and potential therapeutic strategies.
Findings
ints11 deficiency in zebrafish causes motor and behavioral deficits similar to human patients.
Pharmacological treatments like BDNF and OligoGM1 rescue the observed neurodevelopmental defects.
ints11 loss leads to dysregulated gene expression affecting neuronal and glial maturation.
Abstract
Mutations in INTS11, the catalytic subunit of the Integrator complex essential for RNA processing and transcriptional termination, have been linked to neurodevelopmental disorders (NDDs), yet the underlying mechanisms remain poorly understood. To address this gap, we developed and characterized a novel ints11 loss‐of‐function zebrafish model using CRISPR/Cas9 and morpholino‐based approaches, which recapitulates key phenotypic traits observed in human patients, including motor and behavioral deficits. ints11 deficiency led to marked impairments in locomotor activity and visual motor response, consistent with the neurological manifestations reported in INTS11‐mutated patients. These behavioral abnormalities were paralleled by significant dysregulation of neurodevelopmental gene expression, including decreased expression of islet1, map2, gfap, and mag, and upregulation of the progenitor…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
FIGURE 1
FIGURE 2
FIGURE 3
FIGURE 4- —NEXT Generation EU missione 4 componente 1
- —Piano sostegno alla ricerca
- —Ministero della salute10.13039/501100003196
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsRNA Research and Splicing · Williams Syndrome Research · Neurogenetic and Muscular Disorders Research
Introduction
1
The Integrator (Int) complex is a multiprotein structure of 15 subunits (INTS1–INTS15) organized into functional modules with enzymatic and regulatory activities (Offley et al. 2023). Int is recruited by RNA polymerase II (RNAPII) to non‐coding RNA loci, such as U snRNAs and enhancer RNAs, where it ensures proper transcription termination (Lai et al. 2015). Its absence leads to transcriptional readthrough, accumulation of unprocessed transcripts, and widespread transcriptional dysregulation (Davidson et al. 2020; Ezzeddine et al. 2011). Beyond this canonical role, Int contributes to the maturation or stability of other RNA species, including miRNAs, piRNAs, TERC, and the lncRNA NEAT1, and can also modulate protein‐coding gene expression in a context‐dependent manner (Barra et al. 2020). Depending on the genomic context, Int can promote productive elongation, enforce transcriptional pausing, or trigger the complete release of RNAPII from the gene (Fianu et al. 2024; Gardini et al. 2014; S. Hu, Peng, et al. 2023).
Within Integrator, INTS11 is the catalytic endonuclease that executes RNA cleavage. Its N‐terminal β‐CASP/MBL domains form the active site and coordinate two Zn^2+^ ions required for catalysis (Wu et al. 2017; Davidson et al. 2020). By contrast, the C‐terminal region primarily mediates protein–protein interactions, binding INTS9, which acts as a structural scaffold, and enabling subsequent association with INTS4 (Offley et al. 2023). The resulting INTS11–INTS9–INTS4 trimer, known as the Integrator Cleavage/Endonuclease Module (ICM), forms an electropositive RNA‐binding groove that captures the RNA substrate and guides it into the INTS11 catalytic center, thereby facilitating efficient cleavage and downstream transcriptional outcomes (Sabath et al. 2024).
The identified mutations in INTS11, most of which result in loss of function, are located within the catalytic MBL and β‐CASP domains, although some variants map to the C‐terminal domain, thereby disrupting the interaction of INTS11 with INTS9 and INTS4 (Kuang et al. 2023; Tepe et al. 2023). In a recent study of 15 individuals from 10 unrelated families, 19 biallelic variants in the catalytic domains were associated with a neurodevelopmental syndrome characterized by developmental and language delay, intellectual disability, impaired motor development, and brain atrophy (Tepe et al. 2023). Another homozygous variant that weakens the association of the INTS11–INTS9 heterodimer with RNAPII has been linked to severe growth restriction, profound cognitive and motor impairment, microcephaly, and progressive cerebellar atrophy (Kuang et al. 2023). INTS11 also forms a trimeric complex with INTS9 and BRCA1‐Associated ATM activator 1 (BRAT1), essential for the activation of neurogenic programs during neuronal differentiation and required for the supportive role of glial cells (Dokaneheifard et al. 2024; Tepe et al. 2023).
Studies in animal models have shown that INTS11 is essential for early embryonic development, with its loss leading to severe developmental defects or lethality due to widespread transcriptional and RNA‐processing dysregulation (Fianu et al. 2024; Tepe et al. 2023; Wu et al. 2017).
The high degree of evolutionary conservation, the rapid development, the optical transparency, and the ease of genetic manipulation make zebrafish, Danio rerio , an excellent experimental model for neurodevelopmental studies (Nelson and Granato 2022). To date, investigations of ints11 function in this model have focused mainly on its role in hematopoiesis, where its loss of function via morpholino alters Smad signaling and impairs the transition from primitive to definitive hematopoiesis (Tao et al. 2009). In the present study by combining the use of morpholino and CRISPR/Cas9‐mediated approaches, we demonstrate that loss of ints11 leads to marked impairments in locomotor activity and visual‐motor response alteration, consistent with the neurological manifestations reported in INTS11‐mutated patients. These behavioral abnormalities are accompanied by significant dysregulation of neurodevelopmental gene expression. Remarkably, expression of the zebrafish ints11 mRNA and pharmacological administration of brain‐derived neurotrophic factor (BDNF) (Lei et al. 2024) or the GM1 ganglioside‐derived oligosaccharide (OligoGM1) (Chiricozzi et al. 2019; Fazzari et al. 2023) both ameliorated motor and morphological abnormalities and restored neuronal marker expression. Together, these findings establish the first in vivo zebrafish model of INTS11‐associated neurodevelopmental dysfunction, providing a valuable platform to dissect disease mechanisms and to explore therapeutic strategies targeting RNA processing pathways and neurotrophic support in neurodevelopmental disorders.
Material and Methods
2
Zebrafish Declaration Guidelines
2.1
Zebrafish from wild‐type AB strain were maintained and bred according to standard procedures and national guidelines (Italian decree “4 March 2014, n. 26”). All experiments were approved by the Italian Ministry of Health and performed under the supervision of the institutional organism for animal welfare. Adult fish used for eggs collection belong to the Facility of University of Milan (authorization 0035597/23 dated 28/02/2023). Zebrafish larvae were used and processed within 5 days post‐fertilization (≤ 5 dpf). At these developmental stages, larvae are not considered a protected animal model under applicable European and Italian legislation (EU Directive 2010/63/EU and the Italian Legislative Decree 4 March 2014, n. 26). Embryos were collected by natural spawning, staged according to Kimmel and colleagues (Kimmel et al. 1995) and raised at 28°C in fish water (Instant Ocean, 0.1% Methylene Blue) in petri dishes, according to established techniques. For WISH assay, from 24 hpf, the embryos were grown in E3 medium supplemented with 0.003% 1‐phenyl‐2‐thiourea (PTU, Sigma‐Aldrich, Saint Louis, MO; cat. no P7629) to prevent pigmentation.
ints11 Loss of Function With Knock‐Down and Knock‐Out Approaches
2.2
The antisense ints11 splicing‐morpholino (sMO; Gene Tools, Philomath, OR, USA) previously described (Tao et al. 2009) was injected at the dose of 0.25 pmoL/embryo or 0.125 pmoL/embryo into 1‐cell stage embryos. As control, we used embryos injected with the standard control oligo (std‐MO) with no target in zebrafish (Stainier et al. 2017). All embryos/larvae were handled in accordance with institutional and national guidelines and were anesthetized with a 0.016% tricaine solution (ethyl 3‐aminobenzoate methane‐sulfonate salt; Sigma‐Aldrich) prior to manipulation for total RNA extraction. At the experimental endpoint, zebrafish embryo/larvae were properly anesthetized and fixed in 4% PFA and/or processed for total RNA extraction. The knock‐down efficacy was confirmed via RT‐PCR (Figure S1) after total RNA extraction (NucleoZOL; Macherey‐Nagel, Germany; cat. no 740404.200) and cDNA synthesis (cDNA Synthesis Kit; Promega, USA; cat.no A5000) following manufacturers' instructions. MOs and primers sequences are listed in the Table S1. The protocol adopted for the PCR assay is indicated in the Table S2.
The genome editing at the ints11 locus was achieved using the CRISPR‐Cas9 technique. Briefly, 20‐bp long crRNAs specific to S. pyogenes Cas9 were designed using CRISPOR tools (Concordet and Haeussler 2018) selecting the GRCz11 genome assembly. Synthetic guide RNAs were purchased from IDT as Alt‐R CRISPR‐Cas9 crRNA (2 nmol/mL) and Alt‐R CRISPR‐Cas9 tracrRNA (5 nmol/mL). The injection solution was prepared as described (Brix et al. 2024) and 3 nL were injected into 1‐cell stage embryos. As control, embryos were injected with the Cas9 protein assembled with the tracrRNA. To assess genome editing efficiency and specificity, genomic DNA (gDNA) was extracted from pools of 20 embryos using 200 μL of Tris‐EDTA buffer with 1 μg/μL Proteinase K (PK) for 4 h at 55°C. After PK inactivation for 15 min at 95°C, gDNA was used for PCR amplification. The efficiency of each sgRNA was evaluated via T7 endonuclease I (T7EI) with the EnGen Mutation Detection Kit (New England Biolabs, Ipswich, UK; cat.no E3321S). Reaction products were resolved on 2% agarose gel to detect digested PCR fragments, indicating the presence of CRISPR‐induced genomic alterations (Figure S2). The sequences of sgRNAs and PCR primers are listed in Table S1. The protocols for the PCR assays are described in Table S2.
mRNA Injection
2.3
The coding sequence of the zebrafish ints11 was synthesized by GeneArt (Thermo Fisher Scientific, Waltham, MA, USA) into pc‐Globin vector. Once linearized with ApaI (Promega, USA), mRNA was in vitro synthesized using the T7 mMESSAGE mMACHINE (Life Technologies, UK; cat. no AM1344). After purification with SigmaSpin Sequencing Reaction Clean‐Up (Merck; Germany; cat. no S5059), an aliquot of the mRNA was run into 1% agarose gel to assess the integrity. 120 pg/embryos were co‐injected for rescue experiments.
Touch Response (TR) Evoke Assay
2.4
A tactile stimulus was applied to the tail of 2.5 dpf larvae using a micro loader and the number of stimulations required to elicit the escape response was counted (Sztal et al. 2016). Larvae that responded within a maximum of two stimulations, exhibiting a rapid and linear escape trajectory directed away from the stimulus source, were classified as having a physiological touch response. In contrast, larvae requiring more than two stimulations and displaying irregular or non‐linear escape trajectories were classified as affected.
Pharmacological Treatments
2.5
At the 50% epiboly stage, ints11‐loss‐of‐function embryos and their controls were treated with either 100 ng/mL brain‐derived neurotrophic factor (BDNF; #450‐02, Peprotech, Rocky Hill, NJ, USA) or 100 μM OligoGM1, synthesized as previously reported (Fazzari et al. 2023). As OligoGM1 was dissolved in nuclease‐free water, embryos treated with 0.1% BSA in PBS, the vehicle used for BDNF resuspension, were used as the control group. For BDNF‐treated embryos, the chorion was manually pierced using a needle, as previously described (Barbereau et al. 2020) to facilitate compound uptake. Treatments were carried out using a ratio of 8 embryos per 1 mL of final solution and were refreshed every 24 h until 5 dpf.
Whole Mount In Situ Hybridization and Relative Quantification
2.6
Whole‐mount in situ hybridization (WISH) assays were performed using DIG‐labeled antisense RNA probes, following standard protocols as previously described (Thisse and Thisse 2008). Hybridization signal assessment was performed as a qualitative evaluation of signal intensity, classifying embryos/larvae into two categories (normal signal versus altered/reduced signal). For hybridization, embryos were incubated with the probe(s) overnight at 65°C in hybridization buffer containing 60% formamide. Prior to hybridization, embryos were permeabilized with Proteinase K (Sigma‐Aldrich, 10 μg/mL final concentration) for 15 min at 24 hpf and 30 min at 2.5 dpf. Images were acquired using a MZFLIII stereomicroscope equipped with a DC200 digital camera, and processed with LAS software version 4.13.0 (Leica Microsystems, Wetzlar, Germany).
Expression Analyses
2.7
For real‐time quantitative PCR (RT‐qPCR) analysis, total RNA was extracted from at least 30 zebrafish embryos/larvae using NucleoZOL (Macherey‐Nagel, Germany; cat. no 740404.200) according to the manufacturer's instructions. For each sample, 1 μg of total RNA was treated with RNase‐free DNase I (Roche Diagnostics, Basel, Switzerland) to eliminate genomic DNA contamination, and subsequently reverse‐transcribed into cDNA using the GoScript Reverse Transcription System (Promega, Madison, WI, USA; cat. no A5000). qPCR reactions were carried out with iQ SYBR Green Supermix (Promega; Madison, WI, USA; cat. no A6002) on a QuantStudio 5 Real‐Time PCR System (Applied Biosystems, Waltham, MA, USA). Gene expression levels were normalized to the zebrafish rpl8 housekeeping gene and calculated using the 2^−ΔΔCt^ comparative method. Accordingly, control (ctrl) mean has been set at 1. Primer sequences used in this study are listed in Table S1.
Behavior Analyses
2.8
Behavioral assays were performed using the MCAM Kestrel (Ramona Optics, Durham, NC, USA) system, which enabled simultaneous, high‐resolution video acquisition of zebrafish larvae in multi‐well plates (Jamison et al. 2025). A locomotor assay based on acute light/dark transitions, referred to as the visual motor response (VMR), was used to quantify stimulus‐evoked changes in activity following abrupt variations in illumination. Larvae were recorded at 45 frames per second for a total duration of 42 min, with a 10‐min acclimation period. During the experiment, alternating illumination conditions were applied as follows: a 300‐s pulse of white light (8000 lx) followed by a 300‐s period of darkness (0 lx). This sequence was repeated every 10 min throughout the recording period to elicit repeated lights‐on/lights‐off responses and quantify VMR‐associated changes in locomotor activity.
Study Design and Experimental Rigor
2.9
No adult animals were used in this study, except for breeding to obtain embryos for model generation. Sample sizes were based on previous zebrafish larval studies by our and other groups and on established experience with the experimental pipeline, rather than on an a priori sample‐size calculation (Brix et al. 2024; Pezzotta et al. 2022; Pezzotta et al. 2023; Gebhardt et al. 2019; Tzung et al. 2023). Outcome measurements were performed by different operators. During microinjection, embryos were not pre‐selected a priori; injections were performed on consecutively collected embryos obtained from mass matings, thus precluding any prior selection of embryos from specific adult fish. In rescue experiments combined with pharmacological treatment, injected embryos were stochastically (arbitrarily) allocated to the treatment groups (BDNF or vehicle; oligoGM1 or vehicle). For all experiments, embryos/larvae displaying overt morphological abnormalities (e.g., cardiac edema and/or body axis/spinal curvature) were excluded from subsequent analyses.
Statistical Analyses
2.10
Data are presented as means ± standard error of the mean (SEM), as specified in the figure captions. In all the datasets, no analyses of outliers were performed. Touch response (TR) outcomes were treated as categorical data (physiological vs. affected) and group comparisons were carried out using the chi‐squared (χ ^2^) followed by correction with Holm–Bonferroni. For continuous variables, statistical significance among multiple groups was assessed by one‐way analysis of variance (one‐way ANOVA) with Holm–Sidak's post hoc or the relative Kruskal–Wallis non‐parametric test with Dunn's post hoc in case dataset failed the Shapiro–Wilk normality test (alpha = 0.05). In the figure legends statistical significance was reported as p‐values of < 0.05 (), < 0.01 (), and < 0.001 (). For raw statistical data with actual p‐value refer to Tables S3 and S4. All analyses were conducted using GraphPad Prism version 9.0.0 (GraphPad Software, San Diego, CA, USA). All the experiments were performed at least in biological triplicates, otherwise specified.
Results
3
ints11 Morpholino‐ and CRISPR‐Cas9‐Mediated Deficiency Impairs Zebrafish Motor Function
3.1
In zebrafish, ints11 downregulation achieved with the injection of 2.5 ng of a splice‐blocking morpholino (sMO) disrupted normal hematopoiesis reducing scl and gata1 expression (Tao et al. 2009). We performed a dose–response assay and selected 0.125 pmol/embryo as the working dose, which minimized mortality while maintaining effective ints11 knock‐down (Figure S1). In parallel, a complementary CRISPR‐Cas9 model was generated and validated by injecting two sgRNAs targeting exon 4 and exon 5, which encode for the catalytic domain, to produce a non‐functional ints11 protein (Figure S1). ints11 loss‐of‐function models, hereafter ints11‐LoF, did not display gross circulatory defects. However, a smaller body size was consistently observed (Figure 1A–C). Touch‐evoked response assay (TR) was performed at 2.5 dpf taking into consideration the possible developmental delay observed in ints11‐LoF model. ints11‐LoF embryos exhibited an altered motor response compared to controls (Figure 1D).
*ints11 deficient embryos displayed developmental delay and altered motor response. (A–C) Representative images of 4 dpf controls (ctrl) injected with (A) either standard morpholino (std‐MO) or Cas9/tracrRNA and larvae injected with (B) ints11‐sMO and (C) ints11‐sgRNA. Scale bar indicates 300 μm. (D) Touch‐evoked escape response assay performed at 2.5 dpf in controls (n = 40), ints11‐sMO (n = 38) and ints11‐sgRNA (n = 42) embryos. Proportion of embryos with normal TR: Ctrl 39/40; ints11‐sMO: 13/38; ints11‐sgRNA: 12/42. Proportion of embryos with altered TR: Ctrl 1/40; ints11‐sMO: 25/38; ints11‐sgRNA: 30/42. In the histograms the percentage of embryos is expressed as mean ± SEM. Chi‐squared test with Holm–Bonferroni. Refer to Table S3 for raw statistical data. **p < 0.001. ns not significant.
ints11‐LoF Affects Larval Development and Altered Visual‐Motor Response
3.2
To confirm that the altered motor function was caused by ints11 loss‐of‐function, we performed rescue experiments by co‐injecting ints11 mRNA into ints11‐LoF embryos. We observed a significant improvement in motor performance, with a marked reduction in the proportion of ints11‐LoF embryos displaying abnormal touch‐evoked responses at 2.5 dpf (Figure 2A). Because early locomotor behavior in zebrafish primarily depends on the proper establishment of spinal and hindbrain motor circuits, we next employed complementary behavioral and morphological analyses to further characterize the phenotype. Morphological analyses performed at 5 dpf confirmed a consistent reduction in overall body size, with a significant shortening of the body axis and smaller head and tail dimensions in ints11‐LoF embryos (Figure 2B–D). Co‐injection of ints11 mRNA partially rescued these defects, restoring morphology toward control levels. To explore whether ints11 deficiency also impacts stimulus‐evoked sensorimotor behavior, we quantified larval locomotor responses to repeated light‐to‐dark and dark‐to‐light transitions using a high‐throughput imaging platform (Ramona Optics MCAM; Movie S1). At 5 dpf, ints11‐LoF larvae displayed a markedly reduced visual motor response (VMR), with a pronounced loss of responsiveness to illumination changes and failure to appropriately modulate locomotor activity across lights‐off/lights‐on transitions (Figure 2E). In contrast, rescued larvae displayed activity profiles comparable to controls. Notably, ints11‐LoF larvae remained largely immobile throughout the recording, suggesting an impairment in locomotor output and/or in the processing of visual/light cues (Figure 2F; Movie S1).
*ints11‐LoF causes morphological defects and impairs light–dark–dependent motor behavior in zebrafish larvae. (A) Touch‐evoked escape response assay performed at 2.5 dpf in controls (n = 25), ints11‐LoF (n = 26) and ints11‐rescue injected with ints11 mRNA (n = 30) embryos. Proportion of embryos with normal TR: ctrl 24/25; ints11‐LoF: 5/26; rescue: 26/30. Proportion of embryos with altered TR: Ctrl 1/25; ints11‐LoF: 21/26; rescue: 4/30. In the histograms the percentage of embryos is expressed as mean ± SEM. Chi‐squared test with Holm–Bonferroni. ***p < 0.001. ns not significant. Refer to Table S3 for raw statistical data. (B–D) Morphological analysis of 5 dpf larvae obtained from the Ramona Optics MCAM system. Each dot represents the value (expressed in millimeters, mm) of each embryo and mean ± SEM are depicted by black bars: (B) fish length (ctrl: 2832 ± 0.03363; ints11‐LoF: 2529 ± 0.05823; rescue: 2789 ± 0.02823), (C) head length (ctrl: 0.6881 ± 0.01165; ints11‐LoF: 0.6428 ± 0.01355; rescue: 0.6848 ± 0.08033) and (D) tail length (ctrl: 2144 ± 0.02499; ints11‐LoF: 1913 ± 0.03386; rescue: 2105 ± 0.02340). ctrl: N = 24; ints11‐LoF: N = 24; rescue (ints11‐LoF injected with ints11 mRNA): N = 24. Biological duplicate (two independent experiments, in each n = 12/category). Kruskal–Wallis non‐parametric test with Dunn's post hoc (B, D) and ordinary one‐way ANOVA, Holm‐Sidak's post hoc (C). ***p < 0.001; p < 0.05; ns not significant. Refer to Table S4 for raw statistical data. (E) Locomotor activity traces recorded under alternating light and dark periods in 5 dpf controls (green), ints11‐LoF (red) and rescue larvae (blue). Each dot represents an individual time point and reflects the mean locomotor activity calculated from 24 embryos. (F) Tracking of zebrafish larval movements using a speed qualitative color scale.
Altered Neuronal Marker Expression Underlies Motor Defects in ints11‐LoF Embryos
3.3
Given the observed behavioral abnormalities, we next examined whether ints11 deficiency impacts nervous system development by analyzing the expression of neuronal markers. We analyzed pax2a and krox20 expression by WISH using a mixture of antisense RNA probes, as their distinct, well‐defined expression territories provide an informative readout of antero–posterior central nervous system patterning, including midbrain–hindbrain boundary (MHB) specification and hindbrain segmentation. At 24 hpf, the pax2a probe labeled the midbrain‐hindbrain boundary (MHB) (Tan et al. 2022) and the optic regions, while krox20 (Nikolaou et al. 2009) allowed the identification of rhombomeres III and V and the neural tube (Figure 3A). The hybridization signal was weaker in ints11‐LoF embryos, in particular the level of the optic and rhombomere regions (Figure 3B), while co‐injection of the ints11 mRNA partially restored the expression pattern to levels comparable to controls (Figure 3C). We further examined the expression of islet1, a marker of motoneurons, at 2.5 dpf, when motor impairments become evident. In control embryos islet1 expression was clearly detectable in the epiphysis and in dorsal view, in the inner cell layer of the eye and the retinal ganglion cells, as well as in the forebrain and cranial nerve regions (Figure 3D) (Luo et al. 2016). In *ints11‐*LoF (Figure 3E), all these anatomical regions exhibited a reduction in markers expression. Co‐injection of ints11 mRNA partially restored signal intensity to levels comparable to those of control embryos (Figure 3F). To gain temporal insight into how ints11 deficiency affects neural development, we extended our analysis to two key developmental stages, 2.5 and 4 dpf, corresponding respectively to early neurogenesis and subsequent neuronal and glial maturation. At 2.5 dpf, when sensory motor circuits are first established, nestin expression, a marker of neural progenitors, was significantly upregulated, while runx1, a transcription factor required for sensory and motor neuron differentiation (Dooley et al. 2005), was markedly reduced (Figure 3G,H). By 4 dpf, a stage associated with neuronal network consolidation, nestin remained elevated, and markers of neuronal differentiation and function (gfap, map2, mag) were significantly downregulated (Bushell et al. 2008; Faustini et al. 2022; M. Hu, Bodnar, et al. 2023; Luo et al. 2016) (Figure 3I–K). In contrast, sox10, which regulates glial lineage specification (Kunke et al. 2025), remained unchanged at 2.5 dpf but decreased significantly at 4 dpf (Figure 3L), indicating delayed or impaired glial maturation. Notably, co‐injection of ints11 mRNA partially restored the expression of these markers to near‐control levels, except for runx1, which exhibited a higher expression level in the rescue condition (Figure 3G–L). These findings indicate that ints11 deficiency disrupts both early neuronal differentiation and later glial maturation, leading to persistent neurodevelopmental defects.
*ints11 deficiency impairs nervous system development and neuronal marker expression. (A–C) Representative WISH for pax2a and krox20 of 24 hpf (A) controls, (B) ints11‐LoF, and (C) rescue (ints11‐LoF injected with ints11 mRNA) embryos. On the top‐right the number of embryos with similar hybridization signal. Scale bar indicates 100 μm. (D–F) Representative WISH for islet1 of 2.5 dpf (D) controls, (E) ints11‐LoF, and (F) rescue (ints11‐LoF injected with ints11 mRNA) embryos. On the top‐right the number of embryos with similar hybridization signal. Scale bar indicates 200 μm. (G–L) RT‐qPCR expression analyses at 2.5 and 4 dpf for (G) nestin, (H) runx1, (I) gfap, (J) mag, (K) map2, and (L) sox10 in the different categories. Data are presented as mean ± SEM after analyses with the 2−ddCt method. According to 2−ddCt method, ctrl is set at mean 1. nestin 2.5 dpf (ints11‐LoF: 1.672 ± 0.2015; rescue: 1.226 ± 0.2294); nestin 4 dpf (ints11‐LoF: 1621 ± 0.1508; rescue: 0.8567 ± 0.2164); runx1 2.5 dpf (ints11‐LoF: 0.4780 ± 0.1324; rescue: 0.7421 ± 0.3036); runx1 4 dpf (ints11‐LoF: 1026 ± 0.0038; rescue: 1493 ± 0.0570); gfap 2.5 dpf (ints11‐LoF: 0.5843 ± 0.1378; rescue: 1.086 ± 0.2703); gfap 4 dpf (ints11‐LoF: 0.8523 ± 0.0028; rescue: 1.358 ± 0.0212); mag 2.5 dpf (ints11‐LoF: 0.2227 ± 0.1005; rescue: 1675 ± 1036); mag 4 dpf (ints11‐LoF: 0.4137 ± 0.0861; rescue: 0.8724 ± 0.0775); map2 2.5 dpf (ints11‐LoF: 0.6524 ± 0.0716; rescue: 1.006 ± 0.1221); map2 4 dpf (ints11‐LoF: 0.6867 ± 0.0316; rescue: 0.9025 ± 0.0126); sox10 2.5 dpf (ints11‐LoF: 0.9946 ± 0.1900; rescue: 0.9902 ± 0.1396); sox10 4 dpf (ints11‐LoF: 0.6712 ± 0.0265; rescue: 0.9346 ± 0.0015). Each dot represents an independent biological batch/experiment and is the mean of three technical replicates. Kruskal–Wallis non‐parametric test with Dunn's post hoc. p < 0.05, ns: Not significant. Refer to Table S4 for raw statistical data. Cn, cranial nerve; e, Epifisis; fb, forebrain; hb, hindbrain; icl, inner cell layer; mhb, midbrain‐hindbrain boundary; ot, optic tectum; r3, rhombomere 3; r5, rhombomere 5; rgc, retinal ganglion cell; sc, spinal cord.
Administration of Neurotrophic Factors Ameliorates the Phenotype of ints11‐LoF Embryos
3.4
Since in ints11‐LoF model we observed both motor alterations and dysregulation in several neuronal markers, we tested whether molecules with neurotrophic activity could rescue these defects. Embryos were treated with BDNF (Brain‐Derived Neurotrophic Factor) or OligoGM1, starting from the stage of 50% epiboly up to 5 days post fertilization. Through TRD assay performed at the stage of 2.5 dpf we observed a statistically significant improvement in the motor response of ints11‐LoF embryos treated with BDNF or OligoGM1, with no difference between the two drug treatments (Figure 4A). Morphological analysis performed using the automated system indicated that treatment with BDNF or OligoGM1 led to a partial recovery of the morphological defects observed in ints11‐LoF larvae (Figure 4B–D). Moreover, visual‐motor response assays revealed a partial restoration of light–dark–dependent locomotor activity. Larvae treated with BDNF, and more markedly those treated with OligoGM1, exhibited improved motor responses during light–dark transitions (Figures 4E and S2). Given the neurotrophic properties of BDNF and OligoGM1, we next assessed the expression of neuronal markers previously found dysregulated in ints11‐LoF embryos. Both treatments promoted a consistent recovery of gene expression, with a more pronounced effect observed following OligoGM1 administration (Figure 4F–H).
*Pharmacological administration of BDNF and OligoGM1 rescues the phenotype of ints11‐LoF model. (A) Touch‐evoked escape response assay performed at 2.5 dpf in controls (n = 65), ints11‐LoF (n = 80) and ints11‐rescue treated with BDNF (n = 74) or OligoGM1 (n = 73). Proportion of embryos with normal TR: ctrl 63/65; ints11‐LoF: 25/80; ints11‐LoF+BDNF: 55/74; ints11‐LoF+OligoGM1: 59/73. Proportion of embryos with altered TR: ctrl 2/65; ints11‐LoF: 55/80; ints11‐LoF+BDNF: 19/74; ints11‐LoF+OligoGM1: 14/73. In the histograms the percentage of embryos is expressed as mean ± SEM. ***p < 0.001; **p < 0.01; *p < 0.05; ns not significant. Chi‐squared test with Holm–Bonferroni. Refer to Table S3 for raw statistical data. (B–D) Morphological analysis of 5 dpf larvae obtained from the Ramona Optics MCAM system. Each dot represents the value (expressed in millimeters, mm) of each embryo and mean ± SEM are depicted by black bars. (B) Fish length (ctrl: 2.849 ± 0.0289; ints11‐LoF: 2.713 ± 0.0324; ints11‐LoF+BDNF: 2.877 ± 0.0221; ints11‐LoF+OligoGM1: 2.833 ± 0.0349) (C) head length (ctrl: 0.6841 ± 0.0076; ints11‐LoF: 0.6534 ± 0.0098; ints11‐LoF+BDNF: 0.7946 ± 0.0130; ints11‐LoF+oligoGM1: 0.6841 ± 0.0112) and (D) tail length (ctrl: 2.161 ± 0.0230; ints11‐LoF: 2.044 ± 0.0229; ints11‐LoF+BDNF: 2.173 ± 0.0183; ints11‐LoF+oligoGM1: 2.148 ± 0.0281). ctrl: N = 19; ints11‐LoF: N = 22; rescue (BDNF): N = 23; rescue (OligoGM1): N = 22. Biological duplicate. Ordinary one‐way ANOVA, Tukey post hoc. ***p < 0.001; *p < 0.001; *p < 0.05; ns not significant. (E) Cumulative distance traveled by larvae from the different experimental categories in response to light and dark variations. (F–H) RT‐qPCR expression analyses at 4 dpf for (F) gfap, (G) mag, and (H) map2 in the different categories. Data are presented as mean ± SEM after analyses with the 2−ddCt method. According to 2−ddCt method, ctrl is set at mean 1. Each dot represents the mean of three technical replicates from one biological batch. gfap (ints11‐LoF: 0.7503 ± 0.0876; ints11‐LoF+BDNF: 0.7504 ± 0.0203; ints11‐LoF+OligoGM1: 1.183 ± 0.0279); mag (ints11‐LoF: 0.7687 ± 0.0297; ints11‐LoF+BDNF: 1.166 ± 0.0238; ints11‐LoF+OligoGM1: 1.480 ± 0.1047); map2 (ints11‐LoF: 0.8318 ± 0.0231; ints11‐LoF+BDNF: 0.9431 ± 0.0776; ints11‐LoF+OligoGM1: 1.252 ± 0.0455). In 4F‐H, GM1 represents the abbreviation of OligoGM1. Kruskal–Wallis non‐parametric test with Dunn's post hoc. p < 0.05, ns: Not significant. Refer to Table S4 for raw statistical data.
Discussion
4
By combining morpholino‐ and CRISPR/Cas9‐mediated approaches, our study established complementary loss‐of‐function models of ints11 in zebrafish, demonstrating that its deficiency reproduces, at least in part, the pathological features observed in patients carrying INTS11 mutations, including developmental delay, locomotor impairment, and behavioral abnormalities (Kuang et al. 2023; Tepe et al. 2023).
The motor dysfunctions observed in ints11‐LoF embryos can be partly attributed to the dysregulation of key regulatory genes involved in neural development. Indeed, the altered expression of specific neuronal and glial markers supports the hypothesis that ints11 loss impairs hindbrain patterning, motoneuron differentiation, and glial maturation. The downregulation of krox20 likely reflects unstable rhombomere identity and hindbrain segmentation defects (Nikolaou et al. 2009). Likewise, pax2a, expressed in the eye and neural crest–derived periocular mesenchyme, was reduced in ints11‐LoF embryos, suggesting a potential role for ints11 in sensory neural patterning and ocular programs (Tan et al. 2022).
Reduced map2 expression may indicate dendritic stress or atrophy and impaired activity‐dependent protein synthesis, which could compromise neuronal connectivity and signal transmission (Shin'ya et al. 2025). Similarly, decreased islet1 expression points to defective motoneuron specification and early axogenesis, consistent with the sensory and locomotor deficits observed (Liang et al. 2011). The concurrent downregulation of sox10, mag, and gfap, together with the upregulation of nestin, suggests defective astroglial and oligodendroglia differentiation and an accumulation of immature progenitors. Such alterations are expected to reduce axonal conduction efficiency and weaken metabolic and trophic support within neural networks. Moreover, runx1 downregulation may reflect neuronal stress or systemic developmental perturbations, potentially contributing to the increased mortality observed in the morpholino model and the embryonic lethality reported in other ints11 mutant organisms (Wang and Stifani 2017).
Another distinctive feature of our model is the alteration of locomotor behavior in response to acute changes in illumination. Importantly, the light–dark paradigm used here does not measure endogenous circadian rhythmicity; rather, it captures the visual motor response (VMR) to repeated lights‐off/lights‐on transitions. In larvae, abrupt darkness elicits a stereotyped increase in locomotor activity often interpreted as a “light‐search”/dark photokinesis program, engaging defined sensory and neural pathways (Horstick et al. 2017). In this framework, the blunted transition‐evoked modulation and overall hypoactivity observed in ints11‐LoF larvae are more consistent with impaired stimulus processing (visual and/or non‐visual photic input) and/or downstream sensorimotor output. Consistent with this interpretation, the reduced islet1 signal in the epiphysis (a key source of melatonin and a hub for light‐dependent signaling) (Dekens et al. 2022), together with diminished pax2a/islet1 expression in retinal ganglion cells and the optic tectum, points to altered development and/or function of circuits involved in photic input and visuomotor processing. Looking forward, it would also be interesting to directly assay circadian rhythmicity in this model using established long‐duration paradigms (e.g., entrainment followed by recordings under constant conditions), as our current VMR assay cannot address endogenous clock‐driven behavior. Although ints11 is not currently linked to canonical clock genes, the Integrator complex participates in RNA 3′‐end processing and transcription termination, processes that can influence neuronal gene‐expression programs, in other systems, perturbations in RNA processing have been associated with changes in clock gene transcript dynamics and behavioral outputs (Beckwith et al. 2017). Future work using dedicated circadian paradigms could clarify whether ints11 has any role, direct or indirect, in shaping circadian gene‐expression programs and associated behavioral rhythms.
To note, mRNA‐mediated re‐expression of ints11 and pharmacological treatments with BDNF or OligoGM1 significantly ameliorated both morphological and behavioral phenotypes. Given that BDNF promotes synaptogenesis and activity‐dependent plasticity (Toader et al. 2025), and OligoGM1 supports neuronal differentiation and neuroprotection modulating the neurotrophic signaling (Chiricozzi et al. 2021), these findings suggest that ints11 deficiency may impair neurotrophic signaling downstream of RNA‐processing defects. The partial reversibility of the phenotype highlights therapeutic potential for interventions targeting neurotrophic and RNA‐metabolic pathways in INTS11‐associated neurodevelopmental disorders.
An important consideration when interpreting our CRISPR/Cas9 loss‐of‐function data is that F0 zebrafish crispants are intrinsically mosaic, such that both the fraction of edited cells and the repertoire of edited alleles can vary among embryos and across injection series (Vogan 2015). Despite the expected mosaicism, the phenotypes reported here were consistent and reproducible across independent experimental replicates, indicating that the observed outcomes are robust. Moreover, the strong concordance between the phenotypes obtained with morpholino knock‐down and CRISPR/Cas9 crispants, together with phenotypic rescue in complementary approaches (co‐injection of ints11 mRNA and pharmacological modulation with OligoGM1 and BDNF), supports the specificity of the ints11 loss‐of‐function effects and argues against method‐dependent artifacts. This integrated experimental design aligns with current zebrafish literature and guidelines, emphasizing that morpholinos should not be used as a standalone tool, given reported discordance between morpholino‐induced and mutant phenotypes (Lawson 2016). At the same time, published work indicates that stable germline mutants can, in some instances, display attenuated phenotypes due to genetic compensation/transcriptional adaptation, potentially masking loss‐of‐function effects over developmental time (El‐Brolosy et al. 2019; Rossi et al. 2015). Although stable ints11 mutant lines were not generated or analyzed in the present work, systematic evaluation of germline mutants, together with assessment of potential compensatory transcriptional responses will be an important next step to further refine the mechanistic interpretation of INTS11 function during neurodevelopment.
To summarize, the convergence of two independent genetic strategies, combined with successful rescue by mRNA reintroduction and pharmacological treatment, strongly supports a crucial role for ints11 in neural development. Future studies employing stable mutant lines and cell type–specific rescue models will be instrumental in elucidating the tissue‐specific functions of ints11 in brain development and neurophysiology.
Author Contributions
Anna Pistocchi: investigation, resources, supervision, conceptualization, formal analysis, visualization, funding acquisition. Elena Chiricozzi: visualization, resources. Matilde Molteni: software, methodology, validation, formal analysis. Gaia Galassi: methodology, visualization. Laura Mauri: methodology, visualization. Francesca Balistreri: methodology, visualization. Stefania Magri: visualization, methodology. Anna Marozzi: supervision. Franco Taroni: funding acquisition. Alex Pezzotta: conceptualization, data curation, project administration, writing – original draft, supervision, validation.
Funding
Anna Pistocchi, Stefania Magri and Franco Taroni were supported by Ministero della Salute RF‐2016‐02361285. Elena Chiricozzi was supported by the Next Generation EU Missione 4 Componente 1 for project n. 2022 FLAZEC_001, CUP G53D23004490006 Europe Union. Alex Pezzotta was supported by Piano Sostegno alla ricerca PSR2023 University of Milan. The PhD student G.G. was supported by the PhD program in Experimental Medicine of the University of Milan.
Supporting information
Data S1: jnc70408‐sup‐0001‐FigureS1‐S2‐TableS1‐S4.pdf.
Video S1: jnc70408‐sup‐0002‐VideoS1.mp4.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Barbereau, C. , A. Yehya , M. Silhol , et al. 2020. “Neuroprotective Brain‐Derived Neurotrophic Factor Signaling in the TAU‐P 301L Tauopathy Zebrafish Model.” Pharmacological Research 158, no. 104: 865. 10.1016/j.phrs.2020.104865.32417505 · doi ↗ · pubmed ↗
- 2Barra, J. , G. S. Gaidosh , E. Blumenthal , et al. 2020. “Integrator Restrains Paraspeckles Assembly by Promoting Isoform Switching of the lnc RNA NEAT 1 .” Science Advances 6, no. 27: 9072. 10.1126/sciadv.aaz 9072.PMC 745549432923585 · doi ↗ · pubmed ↗
- 3Beckwith, E. J. , C. E. Hernando , S. Polcowñuk , et al. 2017. “Rhythmic Behavior Is Controlled by the S Rm 160 Splicing Factor in Drosophila melanogaster .” Genetics 207, no. 2: 593–607. 10.1534/genetics.117.300139.28801530 PMC 5629326 · doi ↗ · pubmed ↗
- 4Brix, A. , L. Belleri , A. Pezzotta , et al. 2024. “ADA 2 Regulates Inflammation and Hematopoietic Stem Cell Emergence via the A 2b R Pathway in Zebrafish.” Communications Biology 7, no. 1: 615. 10.1038/s 42003-024-06286-3.38777862 PMC 11111730 · doi ↗ · pubmed ↗
- 5Bushell, K. M. , C. Söllner , B. Schuster‐Boeckler , A. Bateman , and G. J. Wright . 2008. “Large‐Scale Screening for Novel Low‐Affinity Extracellular Protein Interactions.” Genome Research 18, no. 4: 622–630. 10.1101/gr.7187808.18296487 PMC 2279249 · doi ↗ · pubmed ↗
- 6Chiricozzi, E. , E. Di Biase , G. Lunghi , et al. 2021. “Turning the Spotlight on the Oligosaccharide Chain of GM 1 Ganglioside.” Glycoconjugate Journal 38, no. 1: 101–117. 10.1007/s 10719-021-09974-y.33620588 PMC 7917043 · doi ↗ · pubmed ↗
- 7Chiricozzi, E. , L. Mauri , G. Lunghi , et al. 2019. “Parkinson's Disease Recovery by GM 1 Oligosaccharide Treatment in the B 4galnt 1+/− Mouse Model.” Scientific Reports 9, no. 1: 19330. 10.1038/s 41598-019-55885-2.31852959 PMC 6920361 · doi ↗ · pubmed ↗
- 8Concordet, J.‐P. , and M. Haeussler . 2018. “CRISPOR: Intuitive Guide Selection for CRISPR/Cas 9 Genome Editing Experiments and Screens.” Nucleic Acids Research 46, no. W 1: W 242–W 245. 10.1093/nar/gky 354.29762716 PMC 6030908 · doi ↗ · pubmed ↗