An in vivo drug screen in zebrafish reveals that cyclooxygenase 2-derived prostaglandin D2 promotes spinal cord neurogenesis

ORIGINAL ARTICLE An in vivo drug screen in zebrafish reveals that cyclooxygenase 2-derived prostaglandin D 2 promotes spinal cord neurogenesis Laura González-Llera 1 | Daniel Sobrido-Cameán 1 | Ana Quelle-Regaldie 2 | Laura Sánchez 2,3 | Ant on Barreiro-Iglesias 1 1 Department of Functional Biology, CIBUS, Faculty of Biology, Universidade de Santiago de Compostela, Santiago de Compostela, Spain 2 Department of Zoology, Genetics and Physical Anthropology, Faculty of Veterinary Science, Universidade de Santiago de Compostela, Lugo, Spain 3 Preclinical Animal Models Group, Health Research Institute of Santiago de Compostela (IDIS), Santiago de Compostela, Spain Correspondence Ant on Barreiro-Iglesias, Department of Functional Biology, CIBUS, Faculty of Biology, Universidade de Santiago de Compostela, 15782 Santiago de Compostela, Spain. Email: [email protected]s Present addresses Daniel Sobrido-Cameán, Department of Zoology, University of Cambridge, Cambridge, UK; and Ana Quelle-Regaldie, Translational Research for Neurological Diseases, Institut Imagine, INSERM UMR 1163, Université Paris Cité, Paris, France. Funding information Xunta de Galicia, Grant/Award Number: ED 431C 2021/18; European Molecular Biology Organization, Grant/Award Number: ALTF 62-2021; Agencia Estatal de Investigaci on, Grant/Award Number: PID2020-115121GB-I00 Abstract The study of neurogenesis is essential to understanding fundamental developmental processes and for the development of cell replacement therapies for central nervous system disorders. Here, we designed an in vivo drug screening protocol in developing zebrafish to find new molecules and signalling pathways regulating neurogenesis in the ventral spinal cord. This unbiased drug screen revealed that 4 cyclooxygenase (COX) inhibitors reduced the generation of serotonergic interneurons in the developing spinal cord. These results fitted very nicely with available single-cell RNAseq data revealing that floor plate cells show differential expression of 1 of the 2 COX2 zebrafish genes (ptgs2a). Indeed, several selective COX2 inhibitors and two different morpholinos against ptgs2a reduced the number of serotonergic neurons in the ventral spinal cord and led to locomotor deficits. Single-cell RNAseq data and different pharmacological manipulations further revealed that COX2-floor plate-derived prostaglandin D 2 promotes neurogenesis in the developing spinal cord by promoting mitotic activity in progenitor cells. Rescue experiments using a phosphodiesterase-4 inhibitor suggest that intracellular changes in cAMP levels underlie the effects of COX inhibitors on neurogenesis and locomotion. Our study provides compelling in vivo evidence showing that prostaglandin signalling promotes neurogenesis in the ventral spinal cord. 1|INTRODUCTION Studying neurogenesis, the process by which new neurons are produced from progenitor cells is essential to understand central nervous system (CNS) development and for the implementation of cell replacement therapies for CNS disorders. Research on the spinal cord has been particularly valuable in uncovering the molecular mechanisms that control neurogenesis and the determination of cell fate from progenitor cells (for a review see Reference [1]). Research on the spinal cord of various vertebrate species in recent decades has demonstrated that signalling molecules influence progenitor cells to develop distinct identities along the dorsoventral axis of the neural tube. For example, sonic hedgehog (Shh) signalling coming from the floor plate (FP) induces ventral identities in the spinal Received: 26 September 2023 Revised: 27 November 2023 Accepted: 18 December 2023 DOI: 10.1111/cpr.13594 This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2023 The Authors. Cell Proliferation published by Beijing Institute for Stem Cell and Regenerative Medicine and John Wiley & Sons Ltd. Cell Prolif. 2023;e13594. onlinelibrary.wiley.com/journal/13652184 1of14 https://doi.org/10.1111/cpr.13594 cord. 2,3 Shh induces the expression of specific transcription factors in progenitors of the ventral domains [from dorsal to ventral: the p0-2, pMN, and p3 or lateral FP (LFP) domains]. Each of these progenitor domains gives rise to distinct neuronal subtypes, which are also characterised by the expression of specific transcription factors (see Reference [1]). For example, in zebrafish, serotonergic interneurons, which are generated from the LFP in the ventral spinal cord, 4 are characterised by the expression of the transcription factor pet1 (fev 5 ). Shh, and other morphogens, also play a role in regulating neurogenesis. For instance, in the pMN domain, high levels of Shh signalling initially drive symmetrical cell division, while decreasing Shh levels leads to differentiation of progenitor cells. 6 Apart from the analysis of highly studied signalling pathways, like the Shh pathway, it would be of interest to find new molecules/signalling pathways regulating neurogenesis in the spinal cord. Deeper examination of the signalling pathways responsible for generating the diversity of neuronal subtypes from progenitor cells could uncover new targets for developing innovative treatments for movement disorders or spinal cord (or brain) injuries. The zebrafish serves as a valuable model for uncovering new molecules and signalling pathways involved in spinal cord neurogenesis. This can be achieved thanks to the implementation of unbiased small molecule drug screens and thanks to available single-cell RNAseq (scRNAseq) data from developing animals (e.g., Reference [7]). Zebrafish provide an ideal model for drug screens due to their rapid development, transparency of early embryos and larvae. Additionally, drugs can be applied in the water from where they are easily taken through the zebrafish skin (see Reference [8]). Here, we developed a two-step screening protocol to find new small molecules regulating spinal cord neurogenesis in developing zebrafish. For this, we decided to use as a model the population of serotonergic interneurons of the ventral spinal cord. This neuronal population offered a good model system to perform a drug screen due to the late appearance of these neurons during early development. Serotonergic neurons start to differentiate from 68 h post-fertilisation (hpf), 5 whereas, for example, most motor neurons are generated between 14 and 48 hpf. 9 This allowed us to apply drugs at 48 hpf after the earlier developmental period, which supports the specificity of the drug screen results (the drugs will not affect early gross embryo morphogenesis). Our drug screen revealed that several cyclooxygenase (COX) inhibitors decreased the generation of serotonergic interneurons in the developing zebrafish spinal cord. COXs (COX1 or COX2) are the rate-limiting enzymes in the generation of prostanoids [e.g., prostaglandins (PGs)]. COXs convert arachidonic acid to PGH 2 , which is then metabolised to other PGs, like PGE 2 or PGD 2 , by specific PG synthases (see Reference [10]). In the CNS, PGs are mainly known for their roles in pain, fever, apoptosis, or inflammatory processes (see Reference [11]), but there is very limited knowledge on their possible role in neurogenic processes in vivo, especially during early developing periods (see Reference [12]). Our drug screen results fitted very nicely with available scRNAseq data showing that FP cells express 1 of the 2 COX2 zebrafish genes (ptgs2a). 7 Indeed, specific COX2 inhibitors and ptgs2a morpholinos also reduced the number of serotonergic neurons in the spinal cord and led to locomotor deficits. ScRNAseq data also showed that FP spinal cord cells express 2 PGD 2 synthase (PGDS) genes (ptgdsb.1 and ptgdsb.2). Treatments with PGDS inhibitors also reduced the number of serotonergic interneurons in the spinal cord. Moreover, PGD 2 rescued the effect of a COX inhibitor. Analyses of cell death and mitotic activity revealed that the reduction in serotonergic neurons after inhibiting PG synthesis is caused by decreased mitotic activity in progenitor cells and not by increased cell death. Rescue experiments using a phosphodiesterase-4 (PDE4) inhibitor suggest that intracellular changes in cAMP levels underlie the effects of COX inhibitors on neurogenesis and locomotor deficits. Overall, our results indicate that COX2-FP-derived PGD 2 promotes neurogenesis in the ventral spinal cord by promoting the proliferation of progenitor cells. Our study is the first to provide compelling in vivo evidence for a role of PGs in the regulation of the neurogenic process in developing animals. Furthermore, our study provides a new drug screen protocol to find small molecules regulating spinal cord neurogenesis in developing vertebrates. 2|MATERIALS AND METHODS 2.1 |Animals All zebrafish lines were kept and raised under standard conditions 13 in the fish facilities of the Department of Genetics of the University of Santiago de Compostela (code of the facility: AE-LU-003, ES270280346401). All experiments were approved by the Bioethics committee of the University of Santiago de Compostela and the Xunta de Galicia (project license no.: 01/20/LU-003) and were carried out in accordance with EU Directive 2010/63/EU for animal experiments. For experimental analyses, we used wild-type or transgenic Tg(-3.2fev: EGFP) [referred also as pet1:gfp; obtained from the European Zebrafish Resource Center (EZRC), EZRC code ne0214Tg] larvae [up to 4 days post-fertilisation (dpf)]. Embryos were collected from the breeding tanks and were divided into Petri dishes at a density of maximally 100 embryos per dish until they were 2 dpf (when they were used for drug treatments, see below), but no formal randomisation method was used. For this study, a total of 3460 zebrafish embryos were used. The specific number of animals used for each experiment is indicated in the figure legends or in the supplementary files. 2.2 |Drug screen We carried out an unbiased drug screen of the LOPAC ® 1280—Small Scale library (Sigma; Cat#LO4200-1EA; International Version). We screened drugs from racks 9 and 10 of the library (160 drugs in total). In the library, drugs are diluted in DMSO at stock concentration of 10 mM. For treatments, 2 μL of the stock solution was diluted in 2 mL of fish water (reverse osmosis-purified water; working dilution of 10 μM). Drugs were tested at this single concentration of 10 μM because this corresponds to the compromised concentration between activity and toxicity for many compounds in the field of in vivo drug 2of14 GONZ ´ ALEZ-LLERA ET AL. 13652184, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/cpr.13594 by Universidade de Santiago de Compostela, Wiley Online Library on [06/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License testing in zebrafish (for a review see Reference [14]). Larvae were incubated from 2 dpf to 4 dpf in groups of 5 animals per well in 24 well-plates (Supplementary Figure 1). Control animals were always treated with DMSO alone. We also tested selective COX2 inhibitors present in other racks of the LOPAC ® library: Etodolac, Rofecoxib, Nimesulide, Niflumic Acid and DFU. These treatments were carried out as with the other drugs of the library in the unbiased drug screen. 2.3 |Treatments with specific drugs in whole larvae Specific drugs (see Table 1) were diluted in DMSO at stock concentration of 10 mM. For treatments, 2 μL of the stock solution were diluted in 2 mL of fish water (working dilution of 10 μM). Larvae were incubated from 2 dpf until 3 or 4 dpf in groups of 5 animals per well in 24 well-plates (Supplementary Figure 1A). Control animals were always treated with DMSO alone. 2.4 |Behavioural analyses Locomotor performance of 4 dpf zebrafish larvae after drug treatments was quantified with the Zebralab software using a Zebrabox (Viewpoint; Civrieux, France). The quantification software measures the distance moved by each fish in a certain period. For this analyses, 4 dpf larvae were transferred to 96 well plates (1 animal per well). Before measuring locomotor activity, the larvae were left in clean fish water without drugs or DMSO (control group) for 1 h. Total larval movement was measured for 1 h, alternating 10-min periods of light and dark conditions. Larvae were kept at a constant standard temperature of 28.5C while measuring locomotor activity. The locomotor activity of each larva was calculated based on the distance moved (in cm) in the 6 10-min periods. 2.5 |Morpholino treatments Morpholinos (Gene Tools, LLC) diluted in nuclease-free water were injected (approximately 3 nL; 1 mM) into the yolk at the one-cell stage of development. Control animals were injected with the Standard Control morpholino (50CCTCTTACCTCAGTTACAATTTATA-30) from Gene Tools. To knockdown the expression of ptgs2a, animals were injected with previously designed ptgs2a translation blocking (50-AACCAGTTTATTCATTCCAGAAGTG-30 15 ; ZFIN ID: ZDB-MRPHLNO-050722-5; ZFIN name: MO1-ptgs2a) or splicing (50-ATTCAACTTACACAACAGGATATAG-30 16 ;ZFINID:ZDBMRPHLNO-110427-4; ZFIN name: MO5-ptgs2a) morpholinos. After morpholino administration, larvae were left until 4 dpf in fish water. 2.6 |Whole-mount anti-serotonin (5-HT) or anti-gfp immunofluorescence After drug or morpholino treatments, 4 dpf larvae were euthanized by tricaine methanesulfonate (Sigma) overdose and then fixed with 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS; pH 7.4) for 2 h at 4C. After washes in PBS, 4 dpf larvae were incubated in Proteinase K from Tritirachium album (Sigma; Cat#: P4850; ≥800 U/mL; 1 μL per ml of PBS) for 35 min at room temperature and then in glycine [50 mM in PBS with 0.2% Triton X-100 (PBST)] for 10 min at room temperature. Then, the larvae were incubated with rabbit anti-5-HT (Immunostar, Still Water, MN, USA; Cat#: 20080; dilution 1:2500; RRID:AB_572263) or chicken anti-gfp (Abcam, Cambridge, UK; Cat#: ab13970; dilution 1:500; RRID:AB_300798) antibodies overnight at 4C. Then, they were rinsed in PBST and incubated overnight at 4C with Cy3-conjugated goat anti-rabbit (Jackson ImmunoResearch; Cat#: 111-165-144; dilution 1:500; RRID:AB_2338006) or Alexa Fluor 488-conjugated goat anti-chicken (Thermo Fisher Scientific; Waltham, MA, USA; Cat#: A-11039; dilution 1:500; RRID:AB_2534096) antibodies. Antibodies were always diluted in PBST with 1% DMSO, 1% normal goat serum, and TABLE 1 Drugs used in specific experiments. Drugs Company Cat.# Molecular Weight [g/mol] Vehicle Application Cyclopamine hydrate Sigma C4116 411.62 DMSO Sonic hedgehog (Shh) inhibitor (S)-(+)-Ibuprofen Sigma 375,160 206.28 DMSO COX inhibitor Nimesulide Sigma N1016 308.31 DMSO Selective COX-2 inhibitor Rolipram Sigma R6520 275.34 DMSO Phosphodiesterase-4 (PDE4) inhibitor Meloxicam sodium salt hydrate Sigma M3935 373.38 DMSO Selective COX-2 inhibitor hPGDS-IN-1 MedChemExpress HY-12791 416.43 DMSO Prostaglandin D Synthase (hPGDS) inhibitor AT-56 Sigma SML0856 397.52 DMSO Lipocalin-type prostaglandin D synthase (L-PGDS) inhibitor Prostaglandin D 2 methyl ester Cayman Chemical 10,008,385 366.5 DMSO Prostaglandin D 2 analogue Note: Note that drugs from the LOPAC library are not included in this table. GONZ ´ ALEZ-LLERA ET AL.3of14 13652184, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/cpr.13594 by Universidade de Santiago de Compostela, Wiley Online Library on [06/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 1% bovine serum albumin. Larvae were mounted with 70% glycerol in PBS. Control and treated animals were always processed in parallel and the same antibody solutions were used for both control and treated animals in each experiment. 2.7 |Cryostat sections 3 dpf larvae were euthanized by tricaine methanesulfonate (Sigma) overdose and then fixed with PFA in PBS (pH 7.4) for 2 h at 4C. After rinsing in PBS, larvae were cryoprotected overnight with 30% sucrose in PBS, embedded in Neg-50™(Thermo Scientific, Kalamazoo, MI, USA), and frozen with liquid nitrogen-cooled isopentane. Transverse sections of the body starting from the caudal fin (16-μm-thickness) were obtained on a cryostat and mounted on Superfrost Plus slides (Menzel-Glasser, Madison, WI, USA). 2.8 |TUNEL labelling on cryostat sections TUNEL staining was performed according to the manufacturer's protocol with minor modifications (In Situ Cell Death Detection Kit, TMR red; catalogue number 12156792910; Roche, Mannheim, Germany). Briefly, the sections were incubated in methanol for 15 min at 20C to permeabilize the lipid membranes, followed by brief washes in PBS and another incubation in 0.01 M citrate buffer pH 6.0 for 30 min at 70C. After several washes in PBS, sections were incubated in the TUNEL reaction mix, containing the Labelling Solution (TMR red labelled nucleotides) and the Enzyme Solution (terminal deoxynucleotidyl transferase), for 90 min at 37C. Slides were washed in PBS and distilled water, allowed to dry for 30 min at 37C, and mounted with MOWIOL ® 4-88 (Calbiochem, Darmstadt, Germany). Negative controls were obtained by incubating sections only with the Labelling Solution (without terminal deoxynucleotidyl transferase). Positive controls were generated by incubating some sections from control 3 dpf animals with recombinant DNAse I (400 U/mL; Roche) for 20 min at room temperature before the incubation in the reaction mix. Sections from control and drug-treated animals were always processed in parallel and the same TUNEL labelling solution was used for sections from control and drug treated animals. 2.9 |Immunofluorescence on cryostat sections Sections were first treated with 0.01 M citrate buffer pH 6.0 for 30 min at 90C for heat-induced epitope retrieval, allowed to cool for 10 min in cold water, and then rinsed in 0.05 M Tris-buffered saline (TBS) pH 7.4 for 20 min. Then, the sections were incubated overnight at RT with a rabbit polyclonal anti-pH3 antibody (1:500; Sigma; Cat#: H0412; RRID: AB_477043). Sections were then rinsed 3 times in TBS for 15 min each and incubated for 1 h at room temperature with a Cy3-conjugated goat anti-rabbit IgG antibody (Jackson ImmunoResearch; Cat#: 111-165-144; dilution 1:500; RRID:AB_2338006). All antibody dilutions were made in TBS containing 15% normal goat serum (Millipore) and 0.2% Triton X-100 (Sigma). Finally, sections were rinsed 3 times in TBS for 15 min each and in distilled water for 20 min, allowed to dry for 30 min at 37C, and mounted in MOWIOL ® 4-88 (Calbiochem). Sections from control and treated animals were always processed in parallel for each antibody staining and the same antibody solutions were used for sections of control and treated animals in each experiment. 2.10 |Imaging and cell counting in whole-mounted larvae and spinal cord sections After anti-5-HT immunofluorescence experiments in whole-mounted 4 dpf larvae, confocal photomicrographs were taken at the level of the caudal fin with TCS-SP2 spectral or Stellaris 8 confocal laser microscopes (Leica Microsystems) with a 20objective. For the quantification of serotonergic neurons, the total number of 5-HT-immunoreactive (ir) or pet1+interneurons located at the level of the caudal fin was quantified manually going through the stack of confocal optical sections. The experimenter was blinded during quantifications. After immunofluorescence experiments or TUNEL labelling in cryostat transverse sections, confocal photomicrographs were taken with the Stellaris 8 confocal laser microscope (Leica Microsystems). We quantified the number of labelled cells in 1 out of each 3 consecutive spinal cord transverse sections starting from the caudal end of the spinal cord and moving rostrally. Nine sections were quantified in each animal and then the mean number of labelled cells per section was calculated for each animal. For pH3 immunolabelling we quantified separately the number of positive cells in the ventral and dorsal portions of the spinal cord. After cell quantifications, figures were prepared with Adobe Photoshop 2022 (San Jose, CA, USA) with minor adjustments of brightness and contrast of the confocal images. Schematic drawings were also generated with Adobe Photoshop 2022. 2.11 |Statistical analysis Experiments with locomotor analyses or cell quantifications were carried out in a minimum of 3 different clutches of animals for each different drug or morpholino treatment, and a minimum of 15 animals were included in each experimental group. Each dot in the graphs represents one animal and n numbers for each experimental group are indicated in the figure legends or in the supplementary files. Statistical analyses were performed with Prism 9 (GraphPad software, La Jolla, CA, USA). Normality of the data was determined with the D'Agostino & Pearson test. For groups with low nnumber in the drug screen, we used a Shapiro–Wilk normality test. To determine statistically significant differences (p≤0.05) between two groups of normally distributed data we used an unpaired (Student's) t-test (two-tailed). To determine statistically significant differences between two groups of non-normally distributed data we used a Mann–Whitney U test 4of14 GONZ ´ ALEZ-LLERA ET AL. 13652184, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/cpr.13594 by Universidade de Santiago de Compostela, Wiley Online Library on [06/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License (two-tailed). To determine statistically significant differences between three groups of non-normally distributed data we used a Kruskal– Wallis test and post hoc Dunn's multiple comparison test. In the figures, significance values were represented by a different number of asterisks in the graphs: *p-value between 0.01 and 0.05, **p-value between 0.001 and 0.01, ***p-value between 0.001 and 0.0001, ****p-value <0.0001. Exact p-values are given in the figure legends. 3|RESULTS AND DISCUSSION 3.1 |An unbiased drug screen reveals that several COX inhibitors reduce the number of serotonergic cells in the spinal cord Spinal cord serotonergic cells begin to differentiate in developing zebrafish at around 60 hpf, when pet1 (fev) expression is first observed. The first mature cells expressing tryptophan hydroxylase 2 (the ratelimiting enzyme for 5-HT synthesis) can be observed at 68 hpf and a fully mature 5-HT-ir population can be observed at 4 dpf. 5 Based on this developing/differentiating timing we decided to apply drugs in our drug screen at 2 dpf, and for 2 days (Supplementary Figure 1A, B), aiming to identify drugs that could affect the behaviour of progenitor cells and/or the neuronal differentiation process. As indicated above, applying the drugs at this time point allowed us to avoid the earlier gross developmental period and other major ventral neurogenic events like the earlier period of motor neuron generation. We designed a two-step drug screening protocol (see Supplementary Figure 1B). In the primary screen, each drug from the LOPAC ® library was tested (at 10 μM) in 10 larvae (2 wells with 5 animals per well). In the secondary screen, hits from the primary screen (drugs that significantly changed the number of serotonergic interneurons as compared to DMSO controls) were further tested in groups of 15 to 30 larvae (Supplementary Figure 1B). First, we used Cyclopamine (a Shh signalling inhibitor) as a positive control to confirm the validity of our drug screening protocol. There was no previous data on the role of Shh signalling in the generation of spinal serotonergic neurons in developing zebrafish. However, the ventral location and origin from the LFP of these neurons 4,5,17 suggest that their generation is probably influenced by Shh signalling. Moreover, after spinal cord injury in adult zebrafish, these serotonergic neurons are also regenerated from the p3/LFP domain and a Cyclopamine treatment inhibits their regeneration. 18 Indeed, following this drug treatment protocol we observed a significant reduction in the number of 5-HT-ir neurons in cyclopamine-treated larvae as compared to DMSO controls (Figure 1A). This indicates that Shh signalling regulates the generation of serotonergic neurons in developing animals as during neuronal regeneration in adult zebrafish, 18 and confirms the validity of our screening protocol to find new small molecules affecting ventral neurogenesis in the developing spinal cord. We screened 160 compounds of the LOPAC ® library (racks 9 and 10) in the primary screen (Supplementary File 1). Cyclopamine was included as a positive control in all the drug screen experiments and always caused a significant reduction in numbers of serotonergic neurons (Supplementary File 1). Seven drugs killed more than 30% of the animals and were discarded for the secondary screen (Supplementary File 1). Of the remaining 153 compounds, 40 drugs significantly reduced the number of 5-HT-ir neurons at 4 dpf as compared to DMSO controls (Supplementary File 1). In the secondary screen, with larger groups of animals, we confirmed that 26 of these 40 compounds significantly reduce the number of 5-HT-ir neurons (Supplementary File 2). We found one discrepancy between the primary and secondary screens with the drug TMPH Hydrochloride (an antagonist of nicotinic acetylcholine receptors), which caused a significant increase in 5-HT-ir neurons in the secondary screen (Supplementary File 2). Interestingly, among the 26 drugs reducing the number of 5-HT-ir neurons in the secondary screen, there was another Shh signalling inhibitor (SANT-2; a smoothened receptor antagonist; Figure 1A), which confirms that Shh signalling regulates the generation of serotonergic spinal cord interneurons. In addition, 3 of the 26 drugs reducing the number of 5-HT-ir neurons have anti-dopaminergic activity (Perlapine, Molindone Hydrochloride, and L-741,626; Figure 1B), which fits nicely with previous data showing that dopamine promotes motor neuron generation in the developing zebrafish spinal cord. 19 Overall, these results confirm the validity of our screening protocol and reveal that Shh and dopamine regulate the generation of serotonergic neurons as previously shown for spinal cord motor neurons. Among the other 22 drugs reducing the number of 5-HT-ir neurons, there were 4 COX inhibitors (Ketorolac Tris Salt, Loxoprofen, Meloxicam Sodium, and S(+)-Ibuprofen; Figure 1C; Supplementary File 2). These unbiased drug screen results pointed to a previously unknown role of COXs and prostanoids in promoting neurogenesis in the ventral spinal cord. However, to confirm that COX inhibitors inhibit neurogenesis and not 5-HT production in existing cells, we repeated the S (+)-Ibuprofen treatments in a pet1:gfp transgenic zebrafish line. The S (+)-Ibuprofen (10 μM) treatment starting at 2 dpf significantly reduced the number of pet1:gfp +cells in the spinal cord of 4 dpf zebrafish (Figure 1D). Therefore, COX-derived prostanoids promote the generation of serotonergic neurons in the ventral spinal cord. Interestingly, the unbiased drug screen results also suggested that the reduction in serotonergic neurons could be related to COX2 inhibition (as opposed to COX1 inhibition). S(+)-Ibuprofen (the active enantiomer of ibuprofen), Ketorolac, and Loxoprofen are nonselective COX inhibitors, but Meloxicam Sodium shows 300-fold selectivity for COX2. In addition, among the non-significant drugs in the primary screen, there were 2 selective COX1 inhibitors: Ketoprofen (p=0.0977) and Indomethacin (p=0.5940) (Supplementary File 1). Overall, our drug screen results pointed to a possible role for COX2 in promoting neurogenesis in the ventral spinal cord. 3.2 |Selective COX2 inhibitors reduce the number of serotonergic cells in the spinal cord Based on the unbiased drug screen results suggesting a role for COX2 in promoting spinal cord neurogenesis, we decided to analyse the GONZ ´ ALEZ-LLERA ET AL.5of14 13652184, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/cpr.13594 by Universidade de Santiago de Compostela, Wiley Online Library on [06/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License FIGURE 1 An unbiased drug screen reveals new small molecules and signalling pathways that control neurogenesis in the ventral spinal cord. (A) Treatments with the Shh signalling inhibitor Cyclopamine (6.487 ± 0,224 cells, n=117; Mann–Whitney test; p-value <0.0001) or the smoothened receptor antagonist SANT-2 (7.682 ± 0.402 cells, n=22; Unpaired t-test; p-value =0.0003) reduced the numbers of serotonergic neurons (red fluorescence) in the ventral spinal cord as compared to DMSO controls (Cyclopamine control: 11.24 ± 0.264 cells, n=101; SANT-2 control: 10.23 ± 0.496 cells, n=22). (B) Treatments with anti-dopaminergic drugs from the LOPAC library (Perlapine: 7.222 ± 0.375 cells, n=27; Unpaired t-test; p-value <0.0001; molindone hydrochloride: 8.842 ± 0.384 cells, n=19; Unpaired t-test; p-value =0.0142; L-741,626: 7.563 ± 0.376 cells, n=21; Unpaired t-test; p-value =0.0067) reduced the numbers of serotonergic neurons in the ventral spinal cord as compared to DMSO controls (Perlapine control: 10.23 ± 0.496 cells, n=22; molindone hydrochloride control: 10.91 ± 0.581 cells, n=32; L-741,626 control: 9.429 ± 0.486 cells, n=16). (C) Treatments with COX inhibitors from the LOPAC library (Ketorolac: 8.125 ± 0.340 cells, n=16; Unpaired t-test; p-value =0.0022; Loxoprofen: 7.615 ± 0.789 cells, n=13; Unpaired t-test; p-value =0.0058; Meloxicam: 6.813 ± 0.579 cells, n=16; Mann– Whitney test; p-value =0.0005; S(+)-Ibuprofen: 7.000 ± 0.532 cells, n=24; Unpaired t-test; p-value <0.0001) reduced the numbers of serotonergic neurons in the ventral spinal cord as compared to DMSO controls (Ketorolac control: 10.91 ± 0.581 cells, n=32; Loxoprofen control: 10.23 ± 0.496 cells, n=13; Meloxicam control: 9.429 ± 0.486 cells, n=21; S(+)-Ibuprofen control: 10.23 ± 0.496 cells, n=22). (D) A treatment with the COX inhibitor S(+)-Ibuprofen (8.200 ± 0.582 cells, n=30; Unpaired t-test; p-value <0.0001) reduced the numbers of serotonergic pet1+neurons in the ventral spinal cord of pet1:gfp fish as compared to DMSO controls (11.84 ± 0.542 cells, n=19). Note the perfect colocalization of serotonin (red) and gfp (green) immunofluorescence signals in serotonergic cells. Rostral is to the right and dorsal to the top in all photomicrographs. Scale bars: 25 μm. 6of14 GONZ ´ ALEZ-LLERA ET AL. 13652184, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/cpr.13594 by Universidade de Santiago de Compostela, Wiley Online Library on [06/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License effect of 5 selective COX2 inhibitors (Etodolac, Rofecoxib, Nimesulide, Niflumic Acid, and DFU) available in other racks of the LOPAC ® library and that were not included in the unbiased drug screen (Supplementary File 3). Treatments with Etodolac, Rofecoxib, Nimesulide, and DFU (all at 10 μM) starting at 2 dpf significantly reduced the number of 5-HT-ir neurons in the spinal cord of 4 dpf zebrafish (Figure 2A; Supplementary File 3). Niflumic Acid also reduced the number of 5-HT-ir neurons at 4 dpf, but it caused a high mortality (Supplementary File 3), which could be related to other effects of the drug as a chloride or calcium channel inhibitor. 20 These results with selective COX2 inhibitors confirmed that COX2 activity promotes neurogenesis in the ventral spinal cord. 3.3 |ScRNAseq data from developing zebrafish reveals differential expression of a COX2 gene (ptgs2a) in FP cells An obvious cell population as a possible source of prostanoids for the regulation of neurogenesis in the ventral spinal cord is the FP, a specialised glial structure of the ventral midline mainly known for its role in ventral nervous tissue differentiation through Shh secretion (see ‘Introduction’). To determine if zebrafish FP cells express any of the zebrafish COX genes (ptgs1,ptgs2a,orptgs2b), we used an available single-cell transcriptome whole-body atlas from developing zebrafish. 7 This atlas was generated from 1, 2, and 5 dpf zebrafish, which nicely covers the developmental period for the generation of serotonergic neurons. In the atlas, cell cluster 176 was identified as the population of spinal cord FP cells based on the differential expression of several well-known FP marker genes (e.g., shha,shhb,gfap,slit1a,slit2,slit1b, foxj1a,wnt4b,spon1a,spon1b,orctgfa). 7 Importantly, one of the genes showing differential expression in this cluster was ptgs2a (p-value =7.38E-56; adjusted p-value =2.4E-51). 51 out of the 94 cells (54.25%) assigned to the spinal cord FP cluster (cluster number 176 of the atlas) show ptgs2a expression (Supplementary Figure 2A). Therefore, ptgs2a is differentially expressed in a cell population that is in an optimal location to influence the generation of serotonergic cells in the developing spinal cord, as with Shh signalling coming also from the FP (see above). Based on the single cell transcriptomic data, we decided to manipulate ptgs2a expression by using well-characterised translation 15 and splicing 16 morpholinos against the zebrafish ptgs2a transcript. These morpholinos had been previously used to knock down ptgs2a expression in developing zebrafish (e.g., 21,22 ). For example, ptgs2a knockdown efficacy with the splicing morpholino in zebrafish was previously demonstrated by RT-PCR. 22 Administration of translation blocking or splicing morpholinos significantly reduced the numbers of 5-HT-ir neurons in the ventral spinal cord at 4 dpf as compared to animals that received the control morpholino (Figure 2B). This confirmed that the effects of the COX2 inhibitors are probably caused by ptgs2a inhibition and that ptgs2a-derived prostanoids promote neurogenesis in the ventral spinal cord. In previous work, only a few studies reported that COX2 inhibition affects adult neurogenesis in rodents ( 23–25 ; for a review see Reference [12]). Meloxicam and Nimesulide treatments decreased the appearance of new neurons in the olfactory bulb of 6-week-old mice. 23 A Celecoxib (a COX2 inhibitor) treatment also reduced numbers of doublecortin+ neuroblasts in the dentate gyrus of 9-week-old mice. 25 COX2 knockout 8-week-old mice also exhibit a significant reduction in doublecortin+neuroblasts of the dentate gyrus. 24,25 Thus, our study is the first to reveal a role for COX2 (ptgs2a) in promoting neurogenesis in the spinal cord and during early developmental periods. 3.4 |Neurogenesis inhibition by non-steroidal anti-inflammatory drugs (NSAIDs) leads to locomotor deficits 8 different COX inhibitors and ptgs2a knock down with two different morpholinos caused a clear reduction in numbers of serotonergic spinal cord neurons of 4 dpf zebrafish (see above). Recent work has shown that serotonergic signalling from these intrinsic spinal cord neurons regulates locomotion by reducing spinally produced motor-bursting in zebrafish. 17 Consequently, it is of interest to test whether the reduction in the generation of serotonergic neurons has an impact on the locomotor activity of the 4 dpf larva. We measured locomotor activity during 1 h (6 10-min periods alternating light and dark conditions) in control and S(+)-Ibuprofen or Nimesulide treated (2–4 dpf) zebrafish. 4 dpf zebrafish were left in 96-well plates (1 larva per well) without DMSO or NSAIDs for 1 hour before measuring locomotor performance with the Zebrabox system (see ‘Material and Methods’). First, we confirmed that the drugs (these were newly purchased and not from the LOPAC ® library) also caused a significant reduction in numbers of 5-HT-ir spinal cord neurons in 4 dpf zebrafish as compared to DMSO controls (Figure 2C). Notably, these treatments led to a significant reduction in locomotor activity during the hour of swim tracking (Figure 2D). Thus, the decrease in spinal cord neurogenesis due to COX2 inhibition leads to behavioural (locomotor) deficits in 4 dpf zebrafish. We should consider that this locomotor deficit could also be caused by changes in other mature cell types apart from serotonergic neurons. For example, during this developmental period, pMN progenitors begin to generate oligodendrocytes after the initial developmental period dedicated to motor neuron production. 26–28 Interestingly, a recent study has shown that PGDS derived from oligodendrocyte progenitor cells promotes oligodendrocyte development in mice. 29 Thus, future work should investigate whether oligodendrogenesis is also regulated by COX2 signalling in the ventral spinal cord, which could also contribute to the locomotor deficits observed after COX2 inhibition. 3.5 |PGD 2 promotes spinal cord neurogenesis COX2 converts arachidonic acid to PGH 2 , which is then metabolised to other PGs by specific PG synthases. To identify the specific PG or GONZ ´ ALEZ-LLERA ET AL.7of14 13652184, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/cpr.13594 by Universidade de Santiago de Compostela, Wiley Online Library on [06/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License FIGURE 2 COX2 (ptgs2a) inhibition reduces the numbers of serotonergic neurons (red fluorescence) in the ventral spinal cord and leads to locomotor deficits. (A) Treatments with the selective COX2 inhibitors Etodolac (7.917 ± 0.499 cells, n=36; Unpaired t-test; p-value =0.0017), Rofecoxib (7.185 ± 0.403 cells, n=27; Unpaired t-test; p-value <0.0001), Nimesulide (6.652 ± 0.46 cells, n=23; Unpaired t-test; p-value <0.0001) and DFU (6.72 ± 0.38 cells, n=25; Unpaired t-test; p-value <0.0001) reduced the numbers of serotonergic neurons in the ventral spinal cord as compared to DMSO controls (10.05 ± 0.424 cells, n=38). (B) Translation (4.244 ± 0.395 cells, n=45; Unpaired t-test; p-value =0.0007) and splicing (4.881 ± 0.338 cells, n=59; Unpaired t-test; p-value =0.0015) morpholinos (MO) against the ptgs2a mRNA reduced the numbers of serotonergic neurons in the ventral spinal cord as compared to zebrafish treated with the control morpholino (6.250 ± 0.411 cells, n=40). (C) Treatments with the COX inhibitors S(+)-Ibuprofen (4.512 ± 0.375 cells, n=43; Unpaired t-test; p-value <0.0001) and Nimesulide (4.241 ± 0.519 cells, n=29; Unpaired t-test; p-value =0.0003) reduced the numbers of serotonergic neurons in the ventral spinal cord as compared to DMSO controls (7.367 ± 0.563 cells, n=49). (D) Animals treated with S(+)-Ibuprofen (294.2 ± 69.20 cm, n=49; Mann–Whitney test; p-value <0.0001) and Nimesulide (155.5 ± 50.63 cm, n=42; Mann–Whitney test; p-value <0.0001) (see C) showed significant locomotor deficits as compared to DMSO controls (1499 ± 163.1 cm, n=60). Examples of 10-min swim tracks (with light) are shown to the left. Total locomotor activity, which was recorded for 1 h (6 10-min periods alternating light and dark conditions), is shown in the graphs. Rostral is to the right and dorsal to the top in all photomicrographs. Scale bars: 25 μm. 8of14 GONZ ´ ALEZ-LLERA ET AL. 13652184, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/cpr.13594 by Universidade de Santiago de Compostela, Wiley Online Library on [06/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License FIGURE 3 Legend on next page. GONZ ´ ALEZ-LLERA ET AL.9of14 13652184, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/cpr.13594 by Universidade de Santiago de Compostela, Wiley Online Library on [06/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License