Jnk1 and Downstream Signalling Hubs Regulate Anxiety-like Behaviours in a Zebrafish Larvae Phenotypic Screen

Jnk1 and Downstream Signalling Hubs Regulate Anxiety-like Behaviours in a Zebrafish Larvae Phenotypic Screen | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Jnk1 and Downstream Signalling Hubs Regulate Anxiety-like Behaviours in a Zebrafish Larvae Phenotypic Screen Ye Hong, Christel Sourander, Benjamin Hackl, Jedidiah Patton, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3830013/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 14 May, 2024 Read the published version in Scientific Reports → Version 1 posted 10 You are reading this latest preprint version Abstract Current treatments for anxiety and depression show limited efficacy in many patients indicating that research into new underlying mechanisms is needed. Inhibition of JNK1 has been shown to evoke an anxiolytic-and antidepressant-like phenotype in mice however the downstream effectors that elicit these behavioural effects are unknown. Here we employ a zebrafish ( D. Rerio ) larvae behavioural assay to identify an antidepressant-/anxiolytic-like phenotype based on 2759 measured stereotypic responses to clinically proven antidepressant and anxiolytic (AA) drugs. Employing machine learning, we classify an AA phenotype from behavioural features measured during and after a startle battery in fish exposed to AA drugs (fluoxetine, imipramine, diazepam, lithium chloride, ketamine). We demonstrate that structurally independent JNK inhibitors replicate the AA classification with high accuracy, consistent with findings in mice. We go on to identify signalling hubs downstream from JNK1 by comparing phosphoproteome data from wildtype and Jnk1-/- mouse brains, and test these hubs as possible mediators of the AA phenotype in zebrafish larvae. Among these, we find that AKT, GSK-3, 14-3-3ζ/ε and PKCε, when pharmacologically targeted, phenocopy clinically proven AA drugs. This assay shows promise as an early phase screening for compounds with anti-stress-axis/anxiolytic-like properties, and for mode of action analysis. Biological sciences/Biochemistry Biological sciences/Biological techniques Biological sciences/Cell biology Biological sciences/Drug discovery Biological sciences/Neuroscience Biological sciences/Systems biology Health sciences/Diseases Health sciences/Medical research Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 INTRODUCTION Anxiety and depression are highly prevalent mental disorders that involve abnormal neural function and dysregulation of circuits associated with threat and fear responses 1,2 . Together they account for over half of the global burden of mental disorders, and are among the largest of disease burdens overall 3,4 . Current treatments are lacking as not all patients achieve remission from their symptoms 5 , while side effects are common 6 . These disorders develop as a result of a complex interplay between genes and the environment 7,8 , and involve a wide variety of mechanisms 5,9 . One of the most consistent changes associated with depression is increased plasma cortisol and deregulation of hypothalamic-pituitary-axis homeostasis associated with desensitised cortisol receptor feedback mechanisms 10,11 . Clinically useful antidepressant drugs such as serotonin reuptake inhibitors (SSRIs) and tricyclic antidepressants have been shown to recover cortisol homeostasis 12 . A screening assay based on behaviour can potentially accelerate mode of action analysis and mechanistic understanding. Phenotypic screening represents a strategy that can identify molecules with the ability to alter behaviour 13 . While mouse models are used to model disease, the time and resources needed to screen compounds in mice would be prohibitive, thus simpler systems are needed when large numbers of compounds should be tested. The zebrafish larvae as a model organism is emerging as a viable approach for screening of phenotypic behaviours 14–16 . For analysis of stress-induced behaviour, the zebrafish hypothalamic-pituitary-interrenal (HPI) axis is cortisol responsive already by four days post fertilization, and like the HP-adrenal (HPA) axis in mammals, its dysregulation is associated with dysfunctional coping behaviours 17,18 . Also, zebrafish larva habenula and amygdala are involved in affective behaviours at an early stage 17 , and zebrafish brain neurotransmitters and corresponding receptors (including NMDA receptors) are expressed and functional at the larval stage. Zebrafish express orthologues for about 70% of human genes with 47% of these genes bearing a 1:1 relationship 19,20 , compared to 80% in mice 21 . The larvae display stereotypical behavioural responses such as thigmotaxis which is an evolutionarily conserved anxiety behavioural response 22 , and hyper-locomotor activity in response to neuroactive drugs 15,16,23 . For these reasons it is used for central nervous system drug screens at an early developmental stage which is suitable for scaleup 18,24 . The JNK1 pathway represents a potential regulator of anxiety and depression. In brain there are three JNK genes and 10 splice variants expressed. Among these, JNK1 is known to play important roles during morphogenesis and in axonal/dendritic architecture modelling during brain development 25–28 . In nervous tissue, JNK has been shown to be activated by endocrine stress leading to retraction of dendritic spines, and JNK-inhibitor treated or Jnk1-/- mice show reduced anxiety- and depressive-like behaviours 29–31 . Moreover, JNK signalling may contribute to glucocorticoid resistance and HPA-axis dysregulation underlying anxiety and depression 32 , as JNK phosphorylation of the glucocorticoid receptor blocks its nuclear translocation and subsequent gene regulation 33,34 . Also, JNK activity is reduced by the fast-acting anti-depressant ketamine when administered to hippocampal neurons in culture 31 . However, JNK1 is expected to phosphorylate a large number of downstream targets in brain 28 , and whether any of these downstream targets contribute to anxiety and depressive-like responses is not known. In this study we identify signalling hubs downstream of JNK1 by comparing the mouse brain phosphoproteome from wild-type and Jnk1-/- mice. We then utilize a zebrafish larvae behavioural assay to test whether any of these downstream hubs phenocopy known antidepressant and anxiolytic drugs, and JNK1 inhibition. This identifies downstream players from JNK1 that phenocopy AA-like stereotypic responses in a zebrafish larvae screen. RESULTS Establishing an acoustic/light stimulus battery to profile zebrafish larvae behavioural responses to anti-depressant and anxiolytic drugs We developed a multiparameter test battery to assess the behaviour of zebrafish larvae in response to light and acoustic stressors. For this, larvae were placed individually in square wells of a 96-well plate and exposed to blue (470 nm), red (635 nm), purple (a combination of red and blue lights), white light, or flickering light, combined with light or heavy tapping as indicated (Fig. 1A and B). Zebrafish larvae exhibited robust motility from 4 days post-fertilization (dpf) onwards for all stimuli, except for white light where older larvae were less responsive (Fig. 1C). We expanded the test to include five cycles, each followed by a 29-minute intermission post startle (Fig. 1D). This enabled us to monitor stress, recovery, and habituation responses. Classical and fast acting anti-depressant and anxiolytic drugs alter zebrafish larvae motility We next measured motility of zebrafish larvae following a 1-hour incubation with classical a nti-depressant or a nxiolytic (AA) drugs: fluoxetine (a SSRI), imipramine (an antagonist against the serotonin transporter > noradrenaline transporter > dopamine transporter), LiCl (which targets inositol monophosphatase and GSK3 35 ), diazepam (a Gamma Amino Butyric Acid receptor antagonist) or low dose (1 – 10 µM) ketamine; a fast-acting antidepressant, or high dose (100 µM) ketamine or MK801 (also known as dizocilpine) 36 ; NMDA receptor antagonists with psychotic activities (Fig. 1E-K; Table 1). Additional data for each drug treatment (i.e. individual fish tracking plots and quantitative output from individual cycles) is found in supplementary figures 1 to 6. The responses to the anti-psychotic drug Haloperidol responses are shown for comparison with the AA drugs (Supplementary figure 7). Motility responses during the 1-minute startle period were quantified from six peak clusters (P1 to P6) (Fig. 1E). Clinically relevant doses of the AA drugs reduced the distance moved during the startle period (Fig. 1F-K). Even lithium at 1 mM showed a response similar to other AA drugs, while at 10 mM (above the therapeutic range), LiCl increased distance moved, possibly due to engagement of additional targets. Diazepam, with its muscle relaxant properties, caused a more significant motility inhibition than other AA drugs. However, it induced an unexpected increase in motility in response to purple light plus tapping (P5) (Fig. 1H), resembling the paradoxical hyperactivity or agitation effect seen in some individuals in response to diazepam 16,37 . Overall the motility profiles from the AA drugs (at the clinically relevant dose) were similar. Ketamine elicits an antidepressant effect in humans at steady state plasma levels of around 1 µM, and an anaesthetic effect at an average dose of 9.3 µM in humans 38,39 . We therefore treated larvae with doses in this range. During the acute startle phase, ketamine decreased motility at doses of 10 µM and above for peaks P1-P3 and increased motility for peaks P4-P6. This closely matched the effect of MK801 and is therefore likely to be mediated through an NMDA receptor antagonist effect as both drugs share the phencyclidine site on (Fig. 1J-K). The response at lower doses of MK801 is consistent with the higher affinity of MK801 for NMDA receptor binding (Ki = 30.5 nM for MK801 verses 417 nM for ketamine) 38 . In summary, treatment of zebrafish larvae with classical AA drugs suppressed their motility responses during the test battery, at doses that are clinically relevant for an anti-depressant/anxiolytic effect in mammals. Distance, turning, pausing and spurting are altered by AA drugs during the startle period We next examined the behavioural syllables that make up the motility response. Thus, in addition to distance travelled, we also measured turning (sum of turning angles), pausing (total time spent pausing), spurt velocity, time thigmotaxis and distance thigmotaxis near the tank border (Fig. 2A). We characterised the effect of AA drugs on this expanded set of features using 411 zebrafish larvae behaviour tracks (Fig. 2B). Fluoxetine, imipramine, LiCl and diazepam elicited strikingly similar behavioural syllables during the startle period (Fig. 2B). All AA drugs increased turning behaviour and decreased thigmotaxis distance and time, and spurting behaviours at lower doses. Only diazepam (at 1 µM or higher) diverged somewhat in that thigmotaxis distance was increased. Ketamine mimicked MK801 at high doses (100 µM) in that both increased thigmotaxis, whereas turning increased with 10 µM ketamine or 1 µM MK801, consistent with the effect being mediated by the NMDA receptor for which MK801 has higher affinity. Notably, increased turning was a prominent feature of the AA-drug treated fish (Fig. 2B). MK801 induced a large increase in motility at 10 µM consistent with the known hyperlocomotor effect of NMDAR inhibition in rodents and fish 38,40-42 , however when measuring the startle phase only, the overall effect of ketamine on motility was minor, even though for stimuli it increased motility (Supplementary figure 6A, B). Interestingly, during the post-startle phase shown in figure 5, MK801 and ketamine responses were almost identical. Identification of JNK 1 pathway signalling hubs Having established phenotypic responses to clinically used AA drug classes in zebrafish larvae, we next turned our attention to the JNK1 pathway. As JNK1 regulates many physiological processes in brain, we were interested to identify downstream signalling hubs that could serve as alternative, possibly more specific AA drug targets than JNK1 itself. Initially we identified JNK1-regulated phosphoproteins by comparing the phosphoproteome from wild-type (WT) and Jnk1-/- mouse brains. Phosphoproteins that differed significantly by more than 1.5 fold in Jnk1 knockout brain are presented as a circular array (Fig. 3A). We next predicted the most highly connected phosphoproteins to the JNK1-regulated ones using the GeneMANIA physical interaction network database. The predicted proteins that physically interact with the JNK1-regulated phosphoproteins are displayed at the centre of the circle (Fig. 3A). From the identified interacting proteins, we determined 29 candidate hubs from the JNK1-regulated phosphoproteome based on interaction counts and interaction confidence levels. To assess the significance of these interactions, we compared interaction counts for JNK1-regulated phosphoproteins to 1000 randomly generated datasets derived from the entire brain phosphoproteome (Fig. 3B). These hubs demonstrated significantly higher connectivity in Jnk1-/- mouse brains compared to WT brains. From among the hubs, we further selected those for which small molecule drugs were available for testing in the zebrafish behavioural screen. The selected hubs included YWHAZ/E (14-3-3s), AKT1, GSK3B, BRAF, and PKCE/G. JNK 1 pathway hub compounds induce AA-like stereotypic behaviours during the startle period We next screened JNK1 pathway drugs during the startle period. JNK inhibitors SP600125 at 1 µM, and JNK-IN-8 at 10 nM mimicked the AA drug profiles (Fig. 4A). Downstream of JNK1, we targeted signalling hubs 14-3-3 and AKT using R18, an inhibitor of 14-3-3 client binding 43 and SC79 (an AKT activator). This also induced a profile similar to that obtained with clinically used AA drugs. In contrast AKTi, an inhibitor of AKT showed an opposite effect on distance and pausing and less effect on turning (Fig. 4A). GSK-3 inhibitor (SB216763) at 1 µM and LiCl at 1 mM, both known to inhibit GSK-3, produced similar responses. We also screened PKC-targeted drugs. PKC-epsilon activator (FR236924) 44 at 0.1 µM increased turning and pausing while decreasing distance (Fig. 4A). Bryostatin-1, a mixed PKC agonist, also increased turning, but differed from FR236924 in that it reduced spurting. Also at 10 nM (a dose that activates PKCa and PKCd isoforms, but antagonizes PKCg), it increased thigmotaxis and decreased pausing. Conversely, the pan-PKC inhibitor 3,4-Bis(3-indoyl)maleimide (BIM) increased turning and decreased thigmotaxis at 100 nM, a dose that inhibits PKCa more effectively than PKCe 45 , We also evaluated whether JNK regulated HPI axis stress in this model. To this end, we used saline stress which has been shown to elicit cortisol-dependent motility in zebrafish larvae that is HPI axis dependent 46 . This treatment increased the distance moved and spurting, and decreased turning (Fig. 4B, C). Notably, pre-treatment with JNK inhibitor SP600125 (10 µM) for 1 hour significantly reduced these stress-induced behaviours (Fig. 4C). This identifies turning, pausing, spurting and distance as features that are JNK and HPI axis dependent from zebrafish motility tracking. Drug behaviour profiles during the post-startle period The post-startle period is also relevant for stress responses and habituation 47 , therefore we examined zebrafish behaviour during this period. Clinically validated drugs produced behaviours similar to those during the startle period but with a more pronounced reduction in distance moved, increased pausing, and milder effects on thigmotaxis (Fig. 5A). Turning behaviour during the post-startle response period increased more linearly with dose, and at 10 µM, ketamine's behaviour profile resembled classical AA drugs in this period. Among the JNK1 pathway hub compounds, several induced anxiolytic-like profiles during the post-startle period (Fig. 5B). For example, the 14-3-3 inhibitor R18 (0.1 µM) mimicked classical AA drug profiles, reducing distance moved while increasing turning and pausing. The selective PKCε activator FR236924 (0.1 µM) showed a similar effect, decreasing distance and thigmotaxis while increasing turning and pausing. JNK inhibitor SP600125 (1 µM) reduced distance, decreased thigmotaxis, and increased pausing, while the less well characterised JNK-IN-8 at 0.01 µM, had no significant effect. The AKT activator SC79 (2 µM) produced a response similar to JNK inhibitor and AA drugs, while the AKT inhibitor AKTi (50 µM) induced an opposite response, possibly representing anxiogenic-like behaviours. GSK-3 inhibitor SB216763 mimicked the response of AKT activator SC79. The PKC activators FR236924 (0.1 µM) and PMA evoked similar responses, although Bryostatin-1, which activates some PKC isoforms, differed substantially in profile 45 . Notably, NMDA receptor antagonists MK801 and high-dose ketamine showed markedly different dose response profiles during and after the startle period (Fig. 5), with post-startle responses aligning more closely with AA drugs. AKTi, BIM, and Bryostatin-1 did not fit the AA phenotype, as expected, whereas haloperidol fit both “AA” and “other” classification. Machine learning identifies compounds with anti-depressant/anxiolytic classification In addition to traditional statistical analysis, we employed machine learning to create a comprehensive AA phenotype classification. For this we used Receiver Operating Characteristic (ROC) curves to assess the "fit" of test drugs to this classification. This approach has the advantage of overcoming the limitations and biases of regular statistical analysis and enhances the interpretation of complex data. We tested three supervised learning models: Random Forest classification-regression, GLMNET linear-regression, and SVM non-linear-regression. The SVM non-linear regression model achieved the highest training accuracy (>0.99) for both the startle and post-startle response periods with AA drugs, making it the chosen model to validate the test compounds. Using motility tracking from the startle response period, we foud that SP600125, JNK-IN-8, SC79, GSK-3 inhibitor (SB216763), PKC isoform activators (FR236924 and PMA), and the 14-3-3 inhibitor (R18) classified as "AA" drug types with an area under the curve (AUC) of >97% in the SVM ROC analysis. For JNK inhibitors, this aligned with mouse data where JNK1 inhibition exerts an anxiolytic effect (Fig. 6A, B) 29 . Conversely, inhibitors of the same molecules (BIM, AKTi, SB59085, and bryostatin-1) did not fit the "AA" category. Interestingly, Haloperidol, which is a D2 receptor antagonist used for its anti-psychotic drug effect, classified with both "AA" and "other" phenotypes. When using post-startle response data, the accuracy of predictions for test drugs was generally higher. Thus, SC79, R18, SB216763, PMA, and FR236924 scored highly for an AA phenotype in the post-startle period, even surpassing the scores of JNK inhibitors, suggesting that greater functional specificity is achieved by targeting downstream of JNK1. A summary of the multiparametric zebrafish larvae behaviour test and the impact of signalling hubs downstream of JNK on the AA phenotype are outlined in Figure 6C. DISCUSSION In this study, we establish a multi-parameter behavioural screen to classify a common AA phenotype in zebrafish larvae using clinically proven AA drugs. We use distinct groups of classical anti-depressant drugs; fluoxetine, representing a commonly used serotonin reuptake inhibitor; imipramine, a tricyclic anti-depressant, and ketamine, a fast-acting anti-depressant, as well as the mood stabilizing drug lithium chloride and diazepam, a commonly used anxiolytic of the benzodiazepine class. Despite having distinct pharmacological profiles, these drugs elicit similar effects on zebrafish larvae motility patterns. All AA drugs reduce distance moved, thigmotaxis and spurting, while they increase turning. This differs from the behavioural responses to psychosis-inducing drugs MK801 and high dose ketamine NMDA receptor antagonist drugs and anti-psychotic Haloperidol, and saline-induced HPI axis stress, all of which produce opposing behaviours in the same test battery. We employ machine learning to characterise an “AA phenotype” using all of these behavioural features, and go on to use the assay to evaluate the signalling hubs downstream of JNK1. The results uncover new regulators downstream of JNK1 that align with the classical AA-drug phenotype performing even better than JNK inhibitors, suggesting increased functional specificity. Dysregulation of the hypothalamic-pituitary-adrenal axis leading to hypercortisolemia is one of the main circuit irregularities contributing to anxiety disorders and major depression 48–51 . This hormonal change can trigger a cascade of events including the activation of JNKs. Accordingly, JNK is activated by cortisol and induces pro-inflammatory cytokines as well as amygdaloid dendritic hypertrophy 31,52–54 . Additionally, JNK plays a role in the homeostatic regulation of glucocorticoid receptor transcriptional activity 25,28 . In contrast, when JNK is inhibited in mouse brain, a reduction in anxiety- and depressive-like behaviours is observed 55 , as is the case with JNK inhibitor treatment or genetic deletion 29,56–58 . In contrast, chimeric JNK activation induces depressive-like behaviour and impaired assessment of risk/reward benefit 58,59 . In zebrafish brain at the larval stage, JNK1 orthologs ( mapk8a and mapk8b) are expressed 60 , and consistent with behavioural findings in rodents 54 , we find that treatment with structurally independent JNK1 inhibitors SP600125 and JNK-IN-8, induce an AA phenotype with A.U.C.s of 97.9% and 98.3% respectively in the startle period. Thus, JNK inhibitors elicit a behavioural response in zebrafish larvae that matches closely to that of clinically used AA drugs. A main goal of this study was to identify signalling hubs downstream of JNK1 influencing anxiety- and depressive-like behaviours. We discovered 14-3-3ζ (YWAZ) and 14-3-3ε (YWHAE), members of the 14-3-3 client binding family 61 , as key hubs in this process. JNK is known to phosphorylate 14-3-3ζ leading to client protein release 62 . Consistent with this, the interaction count for ζ and ε 14-3-3 isoforms increases in Jnk1-/- mouse brain, indicating that JNK1 regulates 14-3-3 binding dynamics in the CNS. In zebrafish larvae, we show that R18 peptide which disrupts 14-3-3 client binding 63 , recapitulates the AA-drug phenotype with A.U.C. of 93.5% and 97.2% in the startle and post-startle period respectively, consistent with the involvement of ζ or ε isoforms 61,64 . 14-3-3ε has also been shown to mediate chronic stress-induced depression 65 . Interestingly, fluoxetine rescues the behavioural deficit in zebrafish lacking 14-3-3ζ suggesting a shared mechanism with fluoxetine 66 . We also identify AKT as a potential signalling hub downstream of JNK1 based on mouse brain phosphoproteome interaction network analysis. AKT and 14-3-3 are mechanistically linked, thus inhibition of 14-3-3 client binding activates AKT 67,68 , potentially placing AKT regulation by JNK1 downstream of 14-3-3. We found that the AKT activator SC79 evoked an AA phenotype with A.U.C.s of 99.4% and 98.3% for the startle and post-startle response periods in the zebrafish larva screen. Conversely, the allosteric inhibitor of AKT (AKTi) induced an opposite response (A.U.C. 60.3% and 60.4% respectively). That AKT activation induces an AA phenotype fits with the well-documented anti-depressant action of AKT in animal models and depressed subjects, and its regulation of BDNF 69,70 . Consistent with this, we also identify that GSK-3β is a hub downstream of JNK1 in mouse brain. This is not surprising as GSK-3α/β isoforms are phosphorylated directly by AKT on S21 and S9 (α and β isoforms respectively), leading to GSK-3 inhibition 71 . In the zebrafish larva assay, the GSK-3α/β inhibitor SB216763, elicits an AA phenotype with A.U.C.s of 97.6% and 99.0% during the startle and post-startle response phases. B-Raf was another signalling hub identified downstream of JNK1. In the interaction screen, B-Raf displayed an increase in protein interactions (from “0” to “10”) in the phosphoproteome from Jnk1-/- mouse brain compared to wild-type. However, when we tested the B-Raf inhibitor SB590885, the resulting phenotype did not classify as “AA” either during the startle or post-startle period. Interestingly, B-Raf is also known to interact with 14-3-3, which functions to maintain its kinase domain in an inactive state 72 . The expectation therefore would be that activation of B-Raf upon disruption of 14-3-3 binding may allow it to contribute to an anxiolytic-like state, as observed with the 14-3-3 client binding inhibitor R18. However, it was not possible for us to test B-Raf further due to lack of activators for this molecule. The oncogenic nature of B-Raf activating mutations contraindicates the use of such molecules. Finally, we tested PKC activators and inhibitors in the zebrafish screen to explore their impact as possible effector hubs downstream of JNK1. We used the PKCε-activator FR236924 and the pan-PKC activator PMA, both of which evoked a clear anxiolytic-like effect (A.U.C.s of 97.8% and 95.7% respectively). Consistent with this, the broad specificity PKC inhibitor (BIM) had the opposite effect. Notably however, in some contrast with our findings in zebrafish larvae, BIM treatment attenuates corticotrophin releasing factor (CRF) facilitation of the acoustic startle response in mice, although it had no effect in the absence of CRF 73 . Moreover, PKCγ -/- mice exhibit low anxiety 74 . Interestingly also, that PKCε activator (FR236924) promotes an AA phenotype in zebrafish is supported by the response to byrostatin-1 at 100 nM, (a dose that activates PKCε 75 ) during the startle response, even though when analysed at the full dose response in our study the overall phenotype is “other”. While these studies suggest isoform-specific roles for PKC variants in anxiety-like behaviours, we identify that PKCε represents a signalling hub downstream from JNK1 which controls an AA phenotype in zebrafish larvae, indicating that further analysis of this isoform in the context of stress-related behaviours is warranted. The observation of overlapping behavioural phenotypes elicited by drugs with distinct pharmacologies merits discussion. The primary target of fluoxetine is to block serotonin reuptake, and that of imipramine is to block serotonin and norepinephrine reuptake, and to additionally antagonise dopamine D2 and acetylcholine receptors 76 . These drugs take weeks to elicit a notable effect on neurotransmitter levels in both responders and non-responders and other effects such as neuroplasticity and neurogenesis changes may be additionally needed for the therapeutic effect 76,77 . A recent study identified that the Tropomyosin receptor kinase (TrkB) is a common binding site for a range of anti-depressant drugs including fluoxetine, imipramine and ketamine. The authors demonstrated that binding of these drugs to TrkB occurs within minutes and is required for the anti-depressant-like effect on behaviour 78 . The authors rationalized that the well studied neuroplasticity effects of BDNF, mediated via TrkB binding, could explain the anti-depressant effect and its timeline. TrkB could represent the common mediator of drug responses measured here as zebrafish larvae indeed express TrkB orthologs ( trkB1 and trkB2 ) from 6 days post fertilization onwards 79 . Moreover, activation of TrkB leads within minutes to the activation of AKT 69 . We uncover that JNK1 signalling hubs converge on the AKT pathway with 14-3-3ζ/ε and GSK-3 in addition to AKT itself identified. Moreover, treatment of zebrafish larvae with the AKT activator (SC79) evokes an AA phenotype with closes match to the clinically used AA drugs with an A.U.C. of 99.4%. AKT pathway activation is the most likely explanation for the temporal and phenotypic overlap observed in response to AA and JNK1 hub drugs. Animal models are limited when it comes to modeling complex disorders such as anxiety and depression. They do not account for subjective experience, or complex genetic and environmental background, and the models that do exist in mice require large numbers to attain statistical power 80,81 . Here we present an assay that is more straightforward and scalable than rodent models. It measures stereotypic responses to acoustic and light stress, based on larval behavioural sequalae (distance, turning, pausing, spurting, thigomataxis distance, thigmotaxis time, dose and size effect). This may lack face and construct validity, and it is not designed to model a specific disease mechanism. Nonetheless, it does measure HPI axis stress responses and the neuroendocrine and monoamine control systems underlying the stress axis is conserved from chordates to mammals 82 , moreover locomotor activity in zebrafish larvae is inherently cortisol dependent 12 . A similar approach has been used to identify novel anti-psychotic compounds in zebrafish larvae 15 , and to investigate the paradoxical excitatory effects of GABA and serotonin 16 , for example. Interestingly, here we also observe a similar hyperactivity effect with the GABAR agonist diazepam when fish are exposed to violet light and acoustic stimuli together (Fig. 1 H). Our study in addition exploites machine learning to classify the desired phenotype, enabling use of complex combinations of features. The result is high accuracy prediction scores for test drugs that are known to be anxiolytic/anti-depressant in mice e.g. AKT activator and GSK-3 inhibitor 69,70,83 . This highlights the potential of the zebrafish larvae screen to potentially identify compounds that mimic AA drugs, for further testing in rodent models, which is rather the goal of a phenotypic screen. A possible explanation for the high prediction score could be linked to the dependence of the zebrafish larvae locomotor response on the HPI-axis 46 , and the drugs used regulate the HPA-axis 84 . Although the stereotypic behaviours measured here may not directly relate to AA conditions, JNK is activated by cortisol 30,85 leading to synapse withdrawal 31 , which may influence circuit behaviour and long-term mental state. Notably also, the phenotype of AA-drugs differ from those of psychotic compounds MK801 and ketamine. Similarly, the anti-psychotic D2 antagonist Haloperidol shows dual classification which is consistent with its known paradoxical side effect, whereby some patients can respond with worsening of symptoms such as increased agitation, insomnia and hallucinations in patients 12 . Consistent with this in the zebrafish larvae, we find that Haloperidol shows a distinct hyper-motility phenotype, which possibly accounts for its classification as “other”; yet it also increases turning and diminishes spurting, consistent with the “AA” phenotype (Fig. 2 ; Supplementary Fig. 7). This paradoxical effect to Haloperidol was also previously reported in zebrafish larvae 15 . A long-term advantage of the screen is the potential reduction in the number of mice needed for drug discovery, while expanding the utility of zebrafish larvae in neuropsychopharmacology studies. MATERIALS AND METHODS Zebrafish husbandry - WT zebrafish ( D. Rerio ) larvae were derived from a cross between the high-performance AB strain and the Tu short-fin wild type strain 20 . Fish were maintained in the Turku Bioscience Zebrafish Core facility at 28.5°C, under a 12 h day/night cycle. Embryos were cultured in E3 medium (5 mM NaCl, 0.17 mM KCl, 0.33 mM CaCl 2 , 0.33 mM MgSO 4 ) with 30 eggs per 10 cm diameter dish. Larvae of indeterminate sex were used from 3 to 8 dpf as indicated. Experiments were carried out in the afternoon to maximise consistency of locomotor behaviour 86 . Fish were bred in two tanks. At least two plates of fish were tested for every drug. There were minor differences in baseline motilities between plates therefore data was normalised to internal control fish measured on the same plate. Afterward the experiment, zebrafish larvae are anesthetized with 200mg/L Tricaine for 5min, and then euthanized by transferring to 1% chloramine ice-water solution. Behavioural experiments were carried out following the guidelines that were approved by the National Animal Experiment Board (ELLA) and the Regional State Administrative Office of Southern Finland (ESAVI) under licenses ESAVI/16343/2019 and ESAVI/30686/2022, which is the authority in Finland that ensures that animal experimentation is carried out following the law and thus follows EU ethical guidelines for animal experimentation. All all methods are performed and reported in accordance with the ARRIVE guidelines ( https://arriveguidelines.org ). Behavioural battery test − 7 days post fertilization (dpf) zebrafish larvae were transferred to square well 96-well plates and drug dosages were administered following which the plate was directly transferred to the DanioVision™ device (Noldus Information Technology, BV, Wageningen, Netherlands) and acclimatized for 1 h in the dark at 28.5°C. Larvae then underwent the test battery consisting of 5 cycles of a 60 s string of acoustic/vibratory (tapping) and light stimuli programmed using Ethovision13 XT software (Noldus Information Technology, BV, Wageningen, Netherlands). The 60 s stimuli were as follows; high (H) and low (L) intensity patterns (1.6 & 4.2 watt): 1 s pause, 6x L-tap (0.5 s break between individual taps), 4 s pause, 4x H-tap, 1 s pause, 8x L-tap, 8 s pause, 2x L- tap, 10x H- tap, 7 s pause, 2x H-tap, 13 s pause, 14x H-tap, 3 s pause. Visual stimuli evoked using red (635 nm, 192 Lm), blue (470 nm, 240 Lm), purple (red and blue together) and white LEDs, switched colour at 10 s intervals in the following order: darkness, red, blue, purple (red & blue), white, and sequential flickering light of all colours sequentially at 0.1 to 0.2 s intervals. After the stimuli period there was a 29-minute intermission period. This 30 min sequence was repeated 4 more times. Digital video tracking was performed using the Daniovision imaging unit using a pre-programmed protocol with the included Ethovision XT software. Acquisition was 60 fps and exported as csv file for automated analysis using customized software in R. The data analyst was agnostic to the drug properties. Stress response assay − 5 or 7 dpf larvae (as indicated) were transferred to a 48-well plate with 1 fish/well and acclimatized for 1 h following which they were exposed to stress: either with white light illumination (10,000 Lux) for 50 s or addition of 100 mM NaCl. Larvae were tracked for a further 30 min following stressor in the dark using infrared light. NaCl experiments were performed using ambient level (500 Lux) white light. Drug treatments – Larvae at the indicated ages were treated with increasing doses of drugs and compounds as follows: fluoxetine, diazepam, FR236924 and PMA (Phorbol 12-myristate 13-acetate) and ketamine were from Tocris Bioscience, Abingdon, UK; haloperidol, imipramine, LiCl, SP600125, SB590885, Bryostatin-1 and MK801 from MERCK Sigma Aldrich Solutions, Darmstadt, DE; R18 from Enzo Life Sciences distributed by AH Diagnostics, Helsinki, FI; SB216763, JNK-in-8 and AKTi from Selleck Chemicals LLC, Houston, Texas; bisindolylmaleimide I (BIM) from Santa Cruz Biotechnology Inc., Dallas, Texas and SC79 from R&D systems, McKinley Place NE, Minneapolis or NaCl in E3 medium. Treatments were for 1h at 28.5 o C before commencing the startle battery. Control fish were treated with equivalent volume of carrier. Behaviour data analysis – Motility : Raw motility tracking data was exported in csv format, where each file contained data from 1 experimental cycle, and each worksheet contained time series data from 1 fish. Relative motility index (M.I. R ) (plots of left side) represents the motility during a certain condition (M c x ) relative to the motility during the control condition (M c 0 ). The relative motility index was calculated as follows from all 5 cycles: $${\text{M}.\text{I}.}_{\text{R}}\left({\text{c}}_{\text{x}}\right)\left[\text{%}\right]=\frac{\left(M{\text{c}}_{\text{x}}\right)-\left(M{\text{c}}_{0}\right)}{\left(M{\text{c}}_{0}\right)}\text{*}100$$ “Relative Peak Analysis” (shown in Fig. 2 right side panels) represents the motility during peaks P1, P2, P3, P4, P5, P6 of a test condition relative to the same peaks from control fish in the same plate. The data scientist handling the analysis was agnostic to the drug treatments. Analysis was done in R. Analysis of other features : X and Y positions, heading, and distance features from the time series data were utilized to calculate behavioural features: distance, turning, pausing, spurt velocity, and thigmotaxis by distance and time. Distance – Distance was summed from all time points per fish and mean total distance travelled per treatment dose was calculated. Turning – Turning angles between consecutive time points were calculated from the heading parameter of the last and current time points, and converted to magnitude representing the smaller angle in 360 degrees. Relative turning angles were calculated by summing all time points for each fish, then normalizing to distance travelled, and presented as average values according to drug dose. Pausing - Time points with value 0 for distance were summed to give a value for total pause time for each fish. Spurt velocity - Time bouts between pauses were identified and the velocity for each time point was calculated from distance feature. The maximum velocity from each time bout was recorded and the mean maximum velocities from all time bouts in a single fish were calculated. Thigmotaxis - The well border coordinates from the 96-well plate were recorded from Ethovision graphic user interface (GUI). An inner border was (virtually) created that was 40% of the width and length of the original border, and placed in the centre of each well. The coordinates of these inner borders served as thresholds for thigmotactic behaviour. Time points where a fish travelled outside the inner border were identified and the ratio of near outer border travel versus total travel for each fish was calculated for time and distance. Anaysis was done in R. Machine learning analysis - Machine learning algorithms were applied to zebrafish larvae behaviour data during the 1 min startle period and during the 10 min of the recovery period. Specifically, the quantitative variables total pause duration, total distance traveled, relative turning angle, average spurt speed, total thigmotaxis time and total thigmotaxis distance were normalized against control (or zero-dosage fish) data as percentages. These normalized metrics taking into account effect sizes, along with dosage level variable, were incorporated as features for the machine learning models. The models were tasked to differentiate between “antidepressant or anxiolytic” drugs and “other” drugs. For the training dataset, drug treatments including diazepam, fluoxetine, imipramine, and lithium chloride (LiCl) were categorized as "antidepressant/anxiolytic" class, while ketamine and MK801 treatments were categorized as "other". Generalized Linear Model via penalized maximum likelihood (Glmnet), Support Vector Machine (SVM) with a polynomial kernel, and Random Forest algorithms were utilized to build 3 separate models each for startle response behaviour (SRB) data during the 1 min startle battery and post-startle response behaviour (PSRB) data during the 10 min recovery period. The constructed models were subsequently employed to predict the class category of other drug treatments. Machine learning was performed with “caret” package within the R environment. Statistics - All behavioural features were compared to control fish from the same experiment. For statistical testing, feature values were collected into a distribution, and this data according to drug dose was compared to respective control distributions using Wilcoxon signed-ranked test or one-way or two-way ANOVA with Tukey or Dunnett’s multiple comparison test as indicated. Calculations and plots were made using a customized pipeline in R programming language. Data analysis was automated and the data scientist was agnostic to the treatment drugs. Preparation of samples for mass spectrometry, SDS-PAGE and in-gel digestion – All the chemicals for digestion and mass spectrometry analysis were from Sigma-Aldrich (Stockholm, Sweden). Whole brain was isolated from WT and Jnk1-/- C57B6J mice 87 at embryonic day 15 (E15), post-natal day 0 (P0), post-natal day 21 (P21), and 8 months (Adult). We extracted brains from three mice for each age and genotype and snap froze them in liquid N 2 and then pulverized them in a Micro-dismembrator II (Braun Biotech International, Melsungen, Germany). We weighed the powder and stored aliquots at -80°C before use. SDS (1%), supplemented with protease and phosphatase inhibitors (Sigma-Aldrich, St. Louis, USA), was added to the brain-powder. Mechanical disruption followed using a syringe twice. We centrifuged samples for 1 hour at 4°C and collected the supernatant. We quantified protein concentration using the Total Protein Kit, Micro Lowry, Peterson's Modification (Sigma-Aldrich, St. Louis, USA). A total protein amount of 100 µg for each sample was separated on 12% Criterion gels (Bio-Rad Laboratories, Hercules, Ca, USA), following the manufacturer’s instructions. We washed the gel in milliQ, stained for 30 min with GelCode (Bio-Rad Laboratories) and washed overnight in milliQ before manually slicing each lane into 5 equal slices. We de-stained gel slices by washing 3 times in 25 mM ammonium bicarbonate (AMBIC)/50% acetonitrile and dried in a vacuum centrifuge (Speedvac). Samples were reduced (10mM DTT/100 mM AMBIC, 1 h at 56°C), alkylated (55 mM iodoacetamide (IAA)/100 mM AMBIC, 45 min at room temperature in the dark), washed 2 times with 100 mM AMBIC and dehydrated with ACN before being dried in the Speedvac. We rehydrated slices with 12.5 µg/ml modified porcine trypsin (Promega, Madison, WI, USA) in 50 mM AMBIC and digested overnight at 37°C. Peptides were eluted 2 times in 75% ACN/5% FA, dried 1 h in the Speedvac and dissolved in 0.1% formic acid. TiO 2 phospho-peptide enrichment – we incubated samples for 10 min with Buffer A (1 M GA, 80% ACN, 5% TFA) followed by incubation with the TiO2 Mag Sepharose (GE Healthcare Life Sciences) for 30 min in an Eppendorf Thermomixer at 800 rpm. Liquid was removed and washing steps were done using Buffer A and 2 times using Buffer B (80% ACN, 1% TFA). Phospho-peptides enriched were eluted 2 times using 100 mM ammonium hydroxide in water and dried 1h in the Speedvac. LC-MS/MS analysis- Dried peptides were resuspended in 0.1% formic acid and separated using an Eksigent nano-LC 2D system (Eksigent, Dublin, CA, USA) consisting of a solvent degasser, a nano-flow pump and a cooled auto-sampler. Peptide concentrations were measured by Nanodrop (ThermoFisher, Stockholm) and the concentration adjusted to the same amount in all samples. Eight µl of sample was loaded and washed for 15 min onto a pre-column (Acclaim PepMap 100, C18, 3 µm particle size, 50 µm diameter, Thermo Fisher Scientific, Hägersten, Sweden) at a constant flow of 5 µl/min solvent B (0.1% FA in ACN). We ran three biological replicates and two technical repeats for each of the two genotypes at the four time points. Phosphopeptides were loaded into a RP analytical column (10 µm fused silica emitter, 75 µm ID column, 16 cm Pico Tip, New Objective) packed in-house with C18 material ReproSil-Pur and separated using an eighty-minute gradient from 3–40% solvent B at a flow rate of 300 nl/min. The gradient was followed by 20 min column washing with 90% ACN, 0.1% FA and 15 min re- equilibration with 3% solvent B. The HPLC system coupled to an LTQ-Orbitrap XL mass spectrometer (ThermoFisher Scientific, Bremen, Germany) operated in Data Dependent Acquisition mode. Spray voltage was set to 1.90 kV and the temperature of the heated capillary was set to 200°C. The ten most intense ions from the survey scan performed by the Orbitrap were fragmented by collision-induced dissociation (CID) in the LTQ (normalized collision energy 35, parent mass selection window 0.5 Da, 30 ms activation time and minimum signal threshold for MS/MS scans set to 100). We excluded unassigned charge states and singly charged ions from fragmentation. The dynamic exclusion list was limited to a maximum of 500 masses with a retention time window of 2 minutes and a relative mass window of 10 ppm. The X-calibur software version 2.0.7 (Thermo Scientific) controlled the HPLC, mass spectrometer and data acquisition. Shotgun data analysis from LTQ Orbitrap – The raw data was uploaded into Progenesis LC-MS (Waters) for analysis. After raw mass spectra were aligned and normalisation across all runs was carried out by total ion current in the analysis area manually defined for the reference HPLC for the time and mass ranges. We submitted the raw files from the mass spectrometer to Thermo Proteome Discoverer (version 1.0.43.0) for protein identification by Mascot. Cysteine carbamidomethylation was set as a fixed modification, methionine oxidation and phosphorylation of serine, threonine, tyrosine (STY) were set as variable modifications. Data were searched against Uniprot Mus musculus database containing 170.598 sequences (release date April 2014). The mass tolerance was set to 10 ppm for the precursor ion and 0.8 Da for the fragment ions. Trypsin was set as the protease and a maximum 2 of missed cleavages were allowed. We ran raw data using an FDR of 5% and a score cutoff of 20. We discarded proteins marked as contaminants, reverse hits and “non-unique peptides”. We selected phospho-peptides with a Mascot score higher than 20 and verified using the default Mascot Delta score of 0.1 for comparative analysis between groups. Data merging and statistical analysis phosphoproteomics data – Downstream analysis of MS intensity outputs from Progenesis LC-MS (Waters) utilized a customized R (v.3.4.0) software pipeline. Phosphopeptide data were refined as follows: (i) all phosphopeptides from 5 gel slices were merged, (ii) identical phosphopeptides, i.e. those with identical sequence, UniProt identifier and number and position of phosphorylated residues, were merged by summing the intensities. (iii) If a peptide entry had > 1 missing value across 3 replicates, all entries were removed. For quality control of data, a distribution plot of phosphopeptide intensities for each unique genotype at given ages was visualized using boxplot and histogram analysis, using ggplot2 88 (Supplementary F1). “Phosphopeptide-centric” and phosphosite-centric analysis was performed as appropriate and is defined in figure legends. We used the Bioconductor package RankProd 89,90 for statistical analysis. We calculated mean ratios of phosphopeptide intensities [ Jnk1-/-/ WT] per entry (for both phosphopeptide-centric and protein-centric changes). Only phosphorylation intensity changes that passed a threshold Ratio Jnk1−/− /WT > 2, with p-value < 0.05 were included. Hub analysis – 29 candidate hubs (Fig. 2 H, I) that were significantly altered in Jnk1-/- mouse brain compared to wildtype brain and were part of the psychiatric disease associated genes according to the MetaCore™ database schizophrenia gene list were selected to analyse as potential signalling hubs acting downstream of JNK1 in regulating depressive/anxiety-like behaviour. The selection criteria included a threshold for phosphoproteins that showed either ≥ 7 physical interactions, or > 3 high weight (weight > 0.04) physical interactions from the physical interaction network built with GeneMANIA from among the significantly altered phosphoproteins derived from the wildtype and Jnk1-/- mouse brain phosphoproteomes. 29 proteins fulfilled these criteria. To control for hub protein's inherent functional plasticity variations, 1000 control phosphoprotein networks of the same size as the Jnk1-/- psychiatric disease phosphoprotein network were randomly generated from the entire brain phosphoproteome dataset using GeneMANIA (v.3.4.1). The number of interactions of the 29 proteins from the significantly altered phosphoproteome network in Jnk1-/- mouse brain was each compared to the distribution of interaction count from 1000 randomly generated control networks to determine the significance of the hubs. Statistical significance for the hubs was calculated using the hypergeometric test. P-values were adjusted using the Benjamini-Hochberg procedure. Receiver Operating Characteristic (ROC) analysis - ROC curves were constructed to visualise the predictive proficiency of the machine learning models. Individual curves were plotted for each test drug from the test dataset, paired with a control drug that was either classified as a "positive control" or a "negative control". Each treatment drug is assigned with a "true" class label that reflected its known properties and molecular interactions. Drugs of the "antidepressant or anxiolytic" class were coupled with the "negative control" drug ketamine, which provided the "true" class label of "other". Conversely, drugs classified under the "other" class were paired with the "positive control" drug fluoxetine for the "true" class label of "Antidepressant or anxiolytic" for the purposes of ROC curve calculation. The construction of the ROC curve was facilitated by the 'pROC' package within the R statistical programming environment. Declarations Data availability- The phosphoproteomics data generated during the current study is available in the PRIDE repository, (www.ebi.ac.uk/pride/) and can be accessed with the following username and password [email protected] ; Password: qNDO1jC9. AUTHOR CONTRIBUTIONS B.H., C.S. J.P. and J.J. carried out experimental work and data analysis. Y.H. carried out the phosphoproteomic and bioinformatics data analysis, machine learning. C.S., Y.H. and E.C. wrote the manuscript. All authors contributed to the manuscript and figure preparation. I.P. supervised the fish husbandry and contributed to data interpretation. E.C. conceptualized and supervised the project. FUNDING: Academy of Finland grant #135090 and 310583, Business Finland grant 1817/31/2015, Åbo Akademi University, ERASMUS Mundi, Åbo Akademi Foundation, Svenska Kulturfonden and K. Albin Johansson. ACKNOWLEDGEMENTS We thank Zebrafish Core (Turku Bioscience Centre, University of Turku and Åbo Akademi University) supported by Biocenter Finland for assistance. COMPETING INTERESTS The authors have nothing to disclose. References Giacobbe, P. & Flint, A. Diagnosis and Management of Anxiety Disorders. Continuum (Minneap Minn) 24 , 893-919 (2018). https://doi.org/10.1212/CON.0000000000000607 Penninx, B. W., Pine, D. S., Holmes, E. A. & Reif, A. Anxiety disorders. Lancet 397 , 914-927 (2021). https://doi.org/10.1016/S0140-6736(21)00359-7 Collaborators, G. M. D. Global, regional, and national burden of 12 mental disorders in 204 countries and territories, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet Psychiatry 9 , 137-150 (2022). https://doi.org/10.1016/S2215-0366(21)00395-3 Collaborators, G. D. a. I. Global burden of 369 diseases and injuries in 204 countries and territories, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet 396 , 1204-1222 (2020). https://doi.org/10.1016/S0140-6736(20)30925-9 Nemeroff, C. B. The State of Our Understanding of the Pathophysiology and Optimal Treatment of Depression: Glass Half Full or Half Empty? Am J Psychiatry 177 , 671-685 (2020). https://doi.org/10.1176/appi.ajp.2020.20060845 Griebel, G. & Holmes, A. 50 years of hurdles and hope in anxiolytic drug discovery. Nat Rev Drug Discov 12 , 667-687 (2013). https://doi.org/10.1038/nrd4075 Wang, T. et al. Polygenic risk for five psychiatric disorders and cross-disorder and disorder-specific neural connectivity in two independent populations. Neuroimage Clin 14 , 441-449 (2017). https://doi.org/10.1016/j.nicl.2017.02.011 Meier, S. M. & Deckert, J. Genetics of Anxiety Disorders. Curr Psychiatry Rep 21 , 16 (2019). https://doi.org/10.1007/s11920-019-1002-7 Garakani, A. et al. Pharmacotherapy of Anxiety Disorders: Current and Emerging Treatment Options. Focus (Am Psychiatr Publ) 19 , 222-242 (2021). https://doi.org/10.1176/appi.focus.19203 Keller, J. et al. HPA axis in major depression: cortisol, clinical symptomatology and genetic variation predict cognition. Mol Psychiatry 22 , 527-536 (2017). https://doi.org/10.1038/mp.2016.120 Perrin, A. J., Horowitz, M. A., Roelofs, J., Zunszain, P. A. & Pariante, C. M. Glucocorticoid Resistance: Is It a Requisite for Increased Cytokine Production in Depression? A Systematic Review and Meta-Analysis. Front Psychiatry 10 , 423 (2019). https://doi.org/10.3389/fpsyt.2019.00423 Ota, K. T. et al. REDD1 is essential for stress-induced synaptic loss and depressive behavior. Nat Med 20 , 531-535 (2014). https://doi.org/10.1038/nm.351310.1038/nm.3513. Epub 2014 Apr 13. Vincent, F. et al. Phenotypic drug discovery: recent successes, lessons learned and new directions. Nat Rev Drug Discov 21 , 899-914 (2022). https://doi.org/10.1038/s41573-022-00472-w Colwill, R. M. & Creton, R. Imaging escape and avoidance behavior in zebrafish larvae. Rev Neurosci 22 , 63-73 (2011). https://doi.org/10.1515/RNS.2011.008 Bruni, G. et al. Zebrafish behavioral profiling identifies multitarget antipsychotic-like compounds. Nat Chem Biol 12 , 559-566 (2016). https://doi.org/10.1038/nchembio.2097 McCarroll, M. N. et al. Zebrafish behavioural profiling identifies GABA and serotonin receptor ligands related to sedation and paradoxical excitation. Nat Commun 10 , 4078 (2019). https://doi.org/10.1038/s41467-019-11936-w de Abreu, M. S. et al. Unconventional anxiety pharmacology in zebrafish: Drugs beyond traditional anxiogenic and anxiolytic spectra. Pharmacol Biochem Behav 207 , 173205 (2021). https://doi.org/10.1016/j.pbb.2021.173205 Maximino, C. et al. Fingerprinting of psychoactive drugs in zebrafish anxiety-like behaviors. PLoS One 9 , e103943 (2014). https://doi.org/10.1371/journal.pone.0103943 Howe, K. et al. The zebrafish reference genome sequence and its relationship to the human genome. Nature 496 , 498-503 (2013). https://doi.org/10.1038/nature12111 Kalueff, A. V., Stewart, A. M. & Gerlai, R. Zebrafish as an emerging model for studying complex brain disorders. Trends Pharmacol Sci 35 , 63-75 (2014). https://doi.org/10.1016/j.tips.2013.12.002 Waterston, R. H. et al. Initial sequencing and comparative analysis of the mouse genome. Nature 420 , 520-562 (2002). https://doi.org/10.1038/nature01262 Schnörr, S. J., Steenbergen, P. J., Richardson, M. K. & Champagne, D. L. Measuring thigmotaxis in larval zebrafish. Behav Brain Res 228 , 367-374 (2012). https://doi.org/10.1016/j.bbr.2011.12.016 Basnet, R. M., Zizioli, D., Taweedet, S., Finazzi, D. & Memo, M. Zebrafish Larvae as a Behavioral Model in Neuropharmacology. Biomedicines 7 (2019). https://doi.org/10.3390/biomedicines7010023 Menezes, F. P., Kist, L. W., Bogo, M. R., Bonan, C. D. & Da Silva, R. S. Evaluation of age-dependent response to NMDA receptor antagonism in zebrafish. Zebrafish 12 , 137-143 (2015). https://doi.org/10.1089/zeb.2014.1018 Coffey, E. T., Hongisto, V., Dickens, M., Davis, R. J. & Courtney, M. J. Dual roles for c-Jun N-terminal kinase in developmental and stress responses in cerebellar granule neurons. J Neurosci 20 , 7602-7613 (2000). Chang, L., Jones, Y., Ellisman, M. H., Goldstein, L. S. & Karin, M. JNK1 is required for maintenance of neuronal microtubules and controls phosphorylation of microtubule-associated proteins. Dev Cell 4 , 521-533 (2003). https://doi.org/S1534580703000947 [pii] Komulainen, E. et al. JNK1 controls dendritic field size in L2/3 and L5 of the motor cortex, constrains soma size, and influences fine motor coordination. Front Cell Neurosci 8 , 272 (2014). https://doi.org/10.3389/fncel.2014.00272 Coffey, E. T. Nuclear and cytosolic JNK signalling in neurons. Nat Rev Neurosci 15 , 285-299 (2014). https://doi.org/10.1038/nrn3729 Mohammad, H. et al. JNK1 controls adult hippocampal neurogenesis and imposes cell-autonomous control of anxiety behaviour from the neurogenic niche. Mol Psychiatry 23 , 362-374 (2018). https://doi.org/10.1038/mp.2016.203 Hollos, P., Marchisella, F. & Coffey, E. T. JNK Regulation of Depression and Anxiety. Brain Plast 3 , 145-155 (2018). https://doi.org/10.3233/BPL-170062 Hollos, P., John, J. M., Lehtonen, J. V. & Coffey, E. T. Optogenetic Control of Spine-Head JNK Reveals a Role in Dendritic Spine Regression. eNeuro 7 (2020). https://doi.org/10.1523/ENEURO.0303-19.2019 Holsboer, F. The corticosteroid receptor hypothesis of depression. Neuropsychopharmacology 23 , 477-501 (2000). https://doi.org/10.1016/S0893-133X(00)00159-7 Jovicic, M. J., Lukic, I., Radojcic, M., Adzic, M. & Maric, N. P. Modulation of c-Jun N-terminal kinase signaling and specific glucocorticoid receptor phosphorylation in the treatment of major depression. Med Hypotheses 85 , 291-294 (2015). https://doi.org/10.1016/j.mehy.2015.05.015 Sevilla, L. M. et al. Glucocorticoid Resistance: Interference between the Glucocorticoid Receptor and the MAPK Signalling Pathways. Int J Mol Sci 22 (2021). https://doi.org/10.3390/ijms221810049 Gould, T. D. & Manji, H. K. Glycogen synthase kinase-3: a putative molecular target for lithium mimetic drugs. Neuropsychopharmacology 30 , 1223-1237 (2005). https://doi.org/10.1038/sj.npp.1300731 Huettner, J. E. & Bean, B. P. Block of N-methyl-D-aspartate-activated current by the anticonvulsant MK-801: selective binding to open channels. Proc Natl Acad Sci U S A 85 , 1307-1311 (1988). https://doi.org/10.1073/pnas.85.4.1307 Mancuso, C. E., Tanzi, M. G. & Gabay, M. Paradoxical reactions to benzodiazepines: literature review and treatment options. Pharmacotherapy 24 , 1177-1185 (2004). https://doi.org/10.1592/phco.24.13.1177.38089 Zanos, P. et al. Ketamine and Ketamine Metabolite Pharmacology: Insights into Therapeutic Mechanisms. Pharmacol Rev 70 , 621-660 (2018). https://doi.org/10.1124/pr.117.015198 Kim, J. W. & Monteggia, L. M. Increasing doses of ketamine curtail antidepressant responses and suppress associated synaptic signaling pathways. Behav Brain Res 380 , 112378 (2020). https://doi.org/10.1016/j.bbr.2019.112378 Carlsson, A., Waters, N. & Carlsson, M. L. Neurotransmitter interactions in schizophrenia-therapeutic implications. Eur Arch Psychiatry Clin Neurosci 249 Suppl 4 , 37-43 (1999). https://doi.org/10.1007/pl00014183 Carey, R. J., Dai, H. & Gui, J. Effects of dizocilpine (MK-801) on motor activity and memory. Psychopharmacology (Berl) 137 , 241-246 (1998). https://doi.org/10.1007/s002130050616 McDougall, S. A. et al. MK801-induced locomotor activity in preweanling and adolescent male and female rats: role of the dopamine and serotonin systems. Psychopharmacology (Berl) 237 , 2469-2483 (2020). https://doi.org/10.1007/s00213-020-05547-3 Wang, B. et al. Isolation of high-affinity peptide antagonists of 14-3-3 proteins by phage display. Biochemistry 38 , 12499-12504 (1999). https://doi.org/10.1021/bi991353h Tanaka, A. & Nishizaki, T. The newly synthesized linoleic acid derivative FR236924 induces a long-lasting facilitation of hippocampal neurotransmission by targeting nicotinic acetylcholine receptors. Bioorg Med Chem Lett 13 , 1037-1040 (2003). https://doi.org/10.1016/s0960-894x(03)00089-1 Keenan, C., Goode, N. & Pears, C. Isoform specificity of activators and inhibitors of protein kinase C gamma and delta. FEBS Lett 415 , 101-108 (1997). https://doi.org/10.1016/s0014-5793(97)01104-6 Lee, H. B. et al. Novel zebrafish behavioral assay to identify modifiers of the rapid, nongenomic stress response. Genes Brain Behav 18 , e12549 (2019). https://doi.org/10.1111/gbb.12549 Beppi, C., Straumann, D. & Bögli, S. Y. A model-based quantification of startle reflex habituation in larval zebrafish. Sci Rep 11 , 846 (2021). https://doi.org/10.1038/s41598-020-79923-6 Brown, E. S., Varghese, F. P. & McEwen, B. S. Association of depression with medical illness: does cortisol play a role? Biol Psychiatry 55 , 1-9 (2004). https://doi.org/10.1016/s0006-3223(03)00473-6 Binder, E. B. et al. HPA-axis regulation at in-patient admission is associated with antidepressant therapy outcome in male but not in female depressed patients. Psychoneuroendocrinology 34 , 99-109 (2009). https://doi.org/10.1016/j.psyneuen.2008.08.018 Shin, L. M. & Liberzon, I. The neurocircuitry of fear, stress, and anxiety disorders. Neuropsychopharmacology 35 , 169-191 (2010). https://doi.org/10.1038/npp.2009.83 Padilla-Coreano, N. et al. Direct Ventral Hippocampal-Prefrontal Input Is Required for Anxiety-Related Neural Activity and Behavior. Neuron 89 , 857-866 (2016). https://doi.org/10.1016/j.neuron.2016.01.011 Qi, A. Q., Qiu, J., Xiao, L. & Chen, Y. Z. Rapid activation of JNK and p38 by glucocorticoids in primary cultured hippocampal cells. J Neurosci Res 80 , 510-517 (2005). https://doi.org/10.1002/jnr.20491 Mitra, R. & Sapolsky, R. M. Acute corticosterone treatment is sufficient to induce anxiety and amygdaloid dendritic hypertrophy. Proc Natl Acad Sci U S A 105 , 5573-5578 (2008). https://doi.org/10.1073/pnas.0705615105 10.1073/pnas.0705615105. Epub 2008 Apr 7. Zhang, J. et al. Inhibition of JNK ameliorates depressive-like behaviors and reduces the activation of pro-inflammatory cytokines and the phosphorylation of glucocorticoid receptors at serine 246 induced by neuroinflammation. Psychoneuroendocrinology 113 , 104580 (2020). https://doi.org/10.1016/j.psyneuen.2019.104580 Mohammad, H. et al. JNK1 controls adult hippocampal neurogenesis and imposes cell-autonomous control of anxiety behaviour from the neurogenic niche. Mol Psychiatry 23 , 487 (2018). https://doi.org/10.1038/mp.2017.21 Zhao, H. B. et al. Xiao Yao San Improves the Anxiety-Like Behaviors of Rats Induced by Chronic Immobilization Stress: The Involvement of the JNK Signaling Pathway in the Hippocampus. Biol Pharm Bull 40 , 187-194 (2017). https://doi.org/10.1248/bpb.b16-00694 Stefanoska, K. et al. Neuronal MAP kinase p38α inhibits c-Jun N-terminal kinase to modulate anxiety-related behaviour. Sci Rep 8 , 14296 (2018). https://doi.org/10.1038/s41598-018-32592-y Zhou, X. et al. Stress-Activated Protein Kinase JNK Modulates Depression-like Behaviors in Mice. Mol Neurobiol 60 , 2367-2378 (2023). https://doi.org/10.1007/s12035-023-03209-x Openshaw, R. L., Pratt, J. A. & Morris, B. J. The schizophrenia risk gene Map2k7 regulates responding in a novel contingency-shifting rodent touchscreen gambling task. Dis Model Mech 15 (2022). https://doi.org/10.1242/dmm.049310 Santos-Ledo, A. et al. Alternative splicing of jnk1a in zebrafish determines first heart field ventricular cardiomyocyte numbers through modulation of hand2 expression. PLoS Genet 16 , e1008782 (2020). https://doi.org/10.1371/journal.pgen.1008782 Fu, H., Subramanian, R. R. & Masters, S. C. 14-3-3 proteins: structure, function, and regulation. Annu Rev Pharmacol Toxicol 40 , 617-647 (2000). https://doi.org/10.1146/annurev.pharmtox.40.1.617 Tsuruta, F. et al. JNK promotes Bax translocation to mitochondria through phosphorylation of 14-3-3 proteins. EMBO J 23 , 1889-1899 (2004). https://doi.org/10.1038/sj.emboj.7600194 Petosa, C. et al. 14-3-3zeta binds a phosphorylated Raf peptide and an unphosphorylated peptide via its conserved amphipathic groove. J Biol Chem 273 , 16305-16310 (1998). https://doi.org/10.1074/jbc.273.26.16305 Besser, J. et al. Expression analysis of the family of 14-3-3 proteins in zebrafish development. Gene Expr Patterns 7 , 511-520 (2007). https://doi.org/10.1016/j.modgep.2006.10.007 Zhao, Y. et al. Identification of 14-3-3 epsilon as a regulator of the neural apoptotic pathway for chronic-stress-induced depression. iScience 24 , 102043 (2021). https://doi.org/10.1016/j.isci.2021.102043 Antón-Galindo, E. et al. Deficiency of the ywhaz gene, involved in neurodevelopmental disorders, alters brain activity and behaviour in zebrafish. Mol Psychiatry 27 , 3739-3748 (2022). https://doi.org/10.1038/s41380-022-01577-9 Castañeda, A. et al. pVHL suppresses Akt/β-catenin-mediated cell proliferation by inhibiting 14-3-3ζ expression. Biochem J 474 , 2679-2689 (2017). https://doi.org/10.1042/BCJ20161097 Gómez-Suárez, M. et al. 14-3-3 Proteins regulate Akt Thr308 phosphorylation in intestinal epithelial cells. Cell Death Differ 23 , 1060-1072 (2016). https://doi.org/10.1038/cdd.2015.163 Varidaki, A., Mohammad, H., Coffey, E.T. Molecular mechanisms of depression . 1st edn, (Elsevier, 2016). Criado-Marrero, M. et al. Early Life Stress and High FKBP5 Interact to Increase Anxiety-Like Symptoms through Altered AKT Signaling in the Dorsal Hippocampus. Int J Mol Sci 20 (2019). https://doi.org/10.3390/ijms20112738 King, M. K. et al. Glycogen synthase kinase-3 inhibitors: Rescuers of cognitive impairments. Pharmacol Ther 141 , 1-12 (2014). https://doi.org/10.1016/j.pharmthera.2013.07.010 Gunderwala, A., Cope, N. & Wang, Z. Mechanism and inhibition of BRAF kinase. Curr Opin Chem Biol 71 , 102205 (2022). https://doi.org/10.1016/j.cbpa.2022.102205 Toth, M., Gresack, J. E., Hauger, R. L., Halberstadt, A. L. & Risbrough, V. B. The role of PKC signaling in CRF-induced modulation of startle. Psychopharmacology (Berl) 229 , 579-589 (2013). https://doi.org/10.1007/s00213-013-3114-9 Bowers, B. J., Collins, A. C., Tritto, T. & Wehner, J. M. Mice lacking PKC gamma exhibit decreased anxiety. Behav Genet 30 , 111-121 (2000). https://doi.org/10.1023/a:1001951104208 Sun, M. K. & Alkon, D. L. Dual effects of bryostatin-1 on spatial memory and depression. Eur J Pharmacol 512 , 43-51 (2005). https://doi.org/10.1016/j.ejphar.2005.02.028 Malhi, G. S. & Mann, J. J. Depression. Lancet 392 , 2299-2312 (2018). https://doi.org/10.1016/S0140-6736(18)31948-2 Alvarez, J. C. et al. Plasma serotonin level after 1 day of fluoxetine treatment: a biological predictor for antidepressant response? Psychopharmacology (Berl) 143 , 97-101 (1999). https://doi.org/10.1007/s002130050924 Casarotto, P. C. et al. Antidepressant drugs act by directly binding to TRKB neurotrophin receptors. Cell 184 , 1299-1313.e1219 (2021). https://doi.org/10.1016/j.cell.2021.01.034 Martin, S. C., Marazzi, G., Sandell, J. H. & Heinrich, G. Five Trk receptors in the zebrafish. Dev Biol 169 , 745-758 (1995). https://doi.org/10.1006/dbio.1995.1184 Strekalova, T. et al. Chronic mild stress paradigm as a rat model of depression: facts, artifacts, and future perspectives. Psychopharmacology (Berl) 239 , 663-693 (2022). https://doi.org/10.1007/s00213-021-05982-w Taschereau-Dumouchel, V., Michel, M., Lau, H., Hofmann, S. G. & LeDoux, J. E. Putting the "mental" back in "mental disorders": a perspective from research on fear and anxiety. Mol Psychiatry 27 , 1322-1330 (2022). https://doi.org/10.1038/s41380-021-01395-5 Ortega, V. A., Mercer, E. M., Giesbrecht, G. F. & Arrieta, M. C. Evolutionary Significance of the Neuroendocrine Stress Axis on Vertebrate Immunity and the Influence of the Microbiome on Early-Life Stress Regulation and Health Outcomes. Front Microbiol 12 , 634539 (2021). https://doi.org/10.3389/fmicb.2021.634539 Pan, J. Q. et al. AKT kinase activity is required for lithium to modulate mood-related behaviors in mice. Neuropsychopharmacology 36 , 1397-1411 (2011). https://doi.org/10.1038/npp.2011.24 Tafet, G. E. & Nemeroff, C. B. Pharmacological Treatment of Anxiety Disorders: The Role of the HPA Axis. Front Psychiatry 11 , 443 (2020). https://doi.org/10.3389/fpsyt.2020.00443 Solas, M., Gerenu, G., Gil-Bea, F. J. & Ramírez, M. J. Mineralocorticoid receptor activation induces insulin resistance through c-Jun N-terminal kinases in response to chronic corticosterone: cognitive implications. J Neuroendocrinol 25 , 350-356 (2013). https://doi.org/10.1111/jne.12006 Fitzgerald, J. A., Kirla, K. T., Zinner, C. P. & Vom Berg, C. M. Emergence of consistent intra-individual locomotor patterns during zebrafish development. Sci Rep 9 , 13647 (2019). https://doi.org/10.1038/s41598-019-49614-y Dong, C. et al. Defective T cell differentiation in the absence of Jnk1. Science 282 , 2092-2095 (1998). H, W., (Springer-Verlag New York, 2009). A language and environment for statistical computing. (R Foundation for Statistical Computing, Vienna, Austria., 2013). Del Carratore, F. et al. RankProd 2.0: a refactored bioconductor package for detecting differentially expressed features in molecular profiling datasets. Bioinformatics 33 , 2774-2775 (2017). https://doi.org/10.1093/bioinformatics/btx292 Table Table 1 is available in the Supplementary Files section. Additional Declarations No competing interests reported. Supplementary Files Supplementarymaterialandlegends.pdf Table1.docx Cite Share Download PDF Status: Published Journal Publication published 14 May, 2024 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 26 Feb, 2024 Reviews received at journal 12 Feb, 2024 Reviewers agreed at journal 27 Jan, 2024 Reviews received at journal 24 Jan, 2024 Reviewers agreed at journal 18 Jan, 2024 Reviewers invited by journal 08 Jan, 2024 Editor assigned by journal 07 Jan, 2024 Editor invited by journal 07 Jan, 2024 Submission checks completed at journal 07 Jan, 2024 First submitted to journal 02 Jan, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-3830013","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":265900020,"identity":"bc4ade79-0113-4ea2-b8bf-ab0aee3c2753","order_by":0,"name":"Ye Hong","email":"","orcid":"","institution":"Åbo Akademi University","correspondingAuthor":false,"prefix":"","firstName":"Ye","middleName":"","lastName":"Hong","suffix":""},{"id":265900021,"identity":"87e3ab12-0ee8-4755-be4b-c81dd32ca96f","order_by":1,"name":"Christel Sourander","email":"","orcid":"","institution":"Åbo Akademi University","correspondingAuthor":false,"prefix":"","firstName":"Christel","middleName":"","lastName":"Sourander","suffix":""},{"id":265900022,"identity":"dd00a47c-bfe0-42e6-912b-a3cc48382d6c","order_by":2,"name":"Benjamin Hackl","email":"","orcid":"","institution":"Medical University of Vienna","correspondingAuthor":false,"prefix":"","firstName":"Benjamin","middleName":"","lastName":"Hackl","suffix":""},{"id":265900023,"identity":"06b53cc8-f7fc-4288-8c7e-311f28528fff","order_by":3,"name":"Jedidiah Patton","email":"","orcid":"","institution":"University of Turku","correspondingAuthor":false,"prefix":"","firstName":"Jedidiah","middleName":"","lastName":"Patton","suffix":""},{"id":265900024,"identity":"83870816-7266-4a82-97db-31127f11c297","order_by":4,"name":"Jismi John","email":"","orcid":"","institution":"University of Turku","correspondingAuthor":false,"prefix":"","firstName":"Jismi","middleName":"","lastName":"John","suffix":""},{"id":265900025,"identity":"af11c036-d0ca-4214-acd3-3d9f71b6946c","order_by":5,"name":"Ilkka Paatero","email":"","orcid":"","institution":"University of Turku","correspondingAuthor":false,"prefix":"","firstName":"Ilkka","middleName":"","lastName":"Paatero","suffix":""},{"id":265900026,"identity":"c766e672-65cb-4487-830e-5944515cb1c7","order_by":6,"name":"Eleanor Coffey","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAtElEQVRIiWNgGAWjYJCCDxUMDHIMEkBWhQFxOhhnnGFgMAZrOUOKlsQGsBZi1POzHz7YcKDicHr/7OYDDAcKiNAi2ZOW2HDgzOHcGXeOJTAcIMZhBjd4zB9/bDuc23Ajx4D5AzFa7G/wGDYc/Hc4XR6ohUhbJEBaGg4nGBCtReIMyC/H0g03Av1ygCgt/O2gEKuxlpe73XzwwYE/RGiBgmYweYB4DQwMdaQoHgWjYBSMgpEGADz4QQcKydf9AAAAAElFTkSuQmCC","orcid":"","institution":"Åbo Akademi University","correspondingAuthor":true,"prefix":"","firstName":"Eleanor","middleName":"","lastName":"Coffey","suffix":""}],"badges":[],"createdAt":"2024-01-02 17:14:45","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3830013/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3830013/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-024-61337-3","type":"published","date":"2024-05-15T00:38:38+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":49384072,"identity":"f05e626f-e72a-4396-8b97-15ab87b70a94","added_by":"auto","created_at":"2024-01-09 19:45:23","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":213616,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMotility profiles for anxiolytic and anti-depressant drugs.\u003c/strong\u003e (A) A schematic of experimental setup shows the apparatus, square-well plate and representative motility traces from 1min tests from larvae treated with MK801. (B) Motility responses from zebrafish larvae at 3 to 8 days post fertilization (dpf) during the 1 min startle battery. \u0026nbsp;(C) Mean motility data of zebrafish larvae at 3-8 dpf. Larvae number per age group; 3 dpf (n=95), 4 dpf (n=94), 5 dpf (n=94), 6 dpf (n=89), 7 dpf (n= 89) and 8 dpf (n=97). (D) Example traces from the entire test battery showing mean motility of 7 dpf fish. 60 min acclimation is followed by 5 cycles (C1 to C5) of 1 min startle period with 29 min recovery between cycles. Measurements are from larvae treated with or without fluoxetine (0.1, 1.0, or 10 µM) for 1 h (n=22 to 24). (E) A trace of 1 min startle response shows motility of fish treated with high doses of ketamine or MK801 (100 µM; n=115). (F-K) Motility responses are shown for AA drugs according to stimuli clusters P1 to P6. The left-side panels depict the overall mean motility during the 5 cycles of 1 min startle. The right-side panels show the mean motility data within the peaksonly. Y-axis scaling differs to accommodate various size effects. One-way ANOVA with post-hoc Tukey test was performed on the original MI distributions to calculate the p-values of the left-side panels; Two-way ANOVA with post-hoc Dunnett test was performed on the original MI distributions to calculate the p-values of the right-side panels. * p-value≤0.05; ** p-value *≤ 0.01; ***p-value≤0.001; ****p-value≤0.0001. The number of fish measured for each treatment were as follows (the total number of repeats is shown in parenthesis): control: 136(680 observations), diazepam: 70 (350 observations), fluoxetine: 72 (360 observations), imipramine: 69 (345 observations), LiCl: 41 (205 observations), ketamine: 70 (350 observations), MK801: 70 (350 observations).\u003c/p\u003e","description":"","filename":"Ffig1.png","url":"https://assets-eu.researchsquare.com/files/rs-3830013/v1/3592623efa1fce53b13e5b24.png"},{"id":49384070,"identity":"1b2a12fc-1cf6-4e25-969c-2d000bd74797","added_by":"auto","created_at":"2024-01-09 19:45:23","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":109583,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTesting the effect of AA drugs on zebrafish larvae behavioural sequalae during the 1 min startle phase. \u003c/strong\u003e(A) The behavioural features extracted from zebrafish larvae tracking are shown. Distance, turning, pausing, spurting, time and distance thigmotaxis are extracted using R statistical computing platform. (B) Fish were exposed to AA drugs and ketamine, MK801 or Haloperidol as indicated and features (distance, turning, pausing, spurting, time and distance thigmotaxis) were extracted from the entire 1 min startle period. Averaged data indicates change relative to control for each of these behaviours according to color code. P-values were calculated by Wilcoxon Rank Sum test and adjusted with Benjamini-Hochberg procedure, where * p-value≤0.05; ** p-value *≤ 0.01; ***p-value≤0.001; ****p-value≤0.0001.\u003c/p\u003e","description":"","filename":"Ffig2.png","url":"https://assets-eu.researchsquare.com/files/rs-3830013/v1/5eb7214f958d7d2e44727ed0.png"},{"id":49384074,"identity":"f77de45f-2c4a-4959-b392-29b22fa2623e","added_by":"auto","created_at":"2024-01-09 19:45:23","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1721032,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIdentification of signalling hubs downstream of JNK1 in mouse brain. \u003c/strong\u003e(A) JNK1 regulated phosphoproteins from \u003cem\u003eJnk1-/-\u003c/em\u003e mouse brain are organised in a large outer circle with connections to the predicted interacting proteins in the centre. Warmer colours indicate a larger number of physical interactions, red being the highest. (B) Signalling hubs derived from the \u003cem\u003eJnk1-/-\u003c/em\u003e mouse brain phosphoproteome (labelled using gene names)represent the most highly connected phosphoproteins from among all phosphoproteins that are significantly altered in \u003cem\u003eJnk1-/-\u003c/em\u003e brain verses wild-type. For each hub, the distribution of physical interaction count from the 1000 background networks is represented with a boxplot.The number of physical interactions for the same hub in the \u003cem\u003eJnk1-/-\u003c/em\u003e mouse brain phosphoproteome (ie the phosphoproteins that are significantly altered in \u003cem\u003eJnk1-/-\u003c/em\u003e mouse brain phosphoproteome), is depicted with a red cross. The most highly ranked JNK1-regulated signalling hubs are shown.\u003c/p\u003e","description":"","filename":"Ffig3.png","url":"https://assets-eu.researchsquare.com/files/rs-3830013/v1/1b3f8e952b1606b6cce33fe9.png"},{"id":49386117,"identity":"6c95d49c-30c8-41a8-8d66-f43a84913898","added_by":"auto","created_at":"2024-01-09 19:53:23","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":99684,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTesting the effect of JNK1 pathway hub drugs on zebrafish larvae behaviour during the startle period. \u003c/strong\u003e(A) Zebrafish behaviours are shown during the 1 min startle period following treatment with pharmacological inhibitors or activators of JNK and downstream signalling hubs. As above, drug treatments were for 1 h before exposure to the startle battery. Mean distance, turning, pausing, spurting, time and distance thigmotaxis are shown for the following numbers of fish measurements (5 cycles per zebrafish larvae) JNK-IN-8: 346, SP600125: 224, Haloperidol: 220, SB216763: 358, FR236924: 360, R18: 354, SC79: 318, BIM: 216, PMA: 115, SB590885: 206, bryostatin-1: 359, AKTi: 360. P-values were calculated by Wilcoxon Rank Sum test and adjusted with Benjamini-Hochberg procedure and are indicated as follows * p-value≤0.05; ** p-value *≤ 0.01; ***p-value≤0.001; ****p-value≤0.0001. (B) Mean motility profiles of zebrafish larvae before and after milliQ (n=24) or NaCl (100 mM, n=24). (C) Mean distance, turning, pausing and spurting during the first 10 min following 100 mM NaCl was measured in the presence or absence of JNK inhibitor SP600125 (10 µM) or DMSO carrier in 7 dpf larvae. The number of fish per group were as follows: control = 47, NaCl + DMSO = 47, SP00125 + E3 = 48, SP00125 + NaCl = 45. Adjusted p-values are indicated as follows *p-value\u0026lt;0.05; ** p-value *\u0026lt; 0.01; ***p-value\u0026lt;0.005. n.s. = not significant. P-values were calculated using hypergeometric test and P-values were adjusted using the Benjamini-Hochberg procedure.\u003c/p\u003e","description":"","filename":"Ffig4.png","url":"https://assets-eu.researchsquare.com/files/rs-3830013/v1/c7c2b6aece6727219a8638bd.png"},{"id":49386116,"identity":"272d6b90-cccd-4ca3-872d-6071a8d64ec7","added_by":"auto","created_at":"2024-01-09 19:53:23","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":122870,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTesting the effect of AA drugs and JNK1 pathway hub drugs on zebrafish larvae behaviour during the post-startle period. \u003c/strong\u003e(A) Fish were exposed to AA drugs and ketamine or MK801 doses as indicated and new features (distance, turning, pausing, spurting, time and distance thigmotaxis) were extracted from the first 10 min following the 1 min startle period. Averaged data from the following fish measurement numbers: diazepam: 337, fluoxetine: 338, imipramine: 335, LiCl: 210, ketamine: 325, MK801: 323 are shown. P-values were calculated by Wilcoxon Rank Sum test and adjusted with Benjamini-Hochberg procedure and are indicated as follows: * p-value≤0.05; ** p-value *≤ 0.01; ***p-value≤0.001; ****p-value≤0.0001. (B) Zebrafish behaviours during the 10 min following the 1 min startle period are shown with or without treatment with pharmacological inhibitors or activators of JNK and downstream signalling hubs. As above, drug treatments were for 2 h before exposure to the startle battery. Mean distance, turning, pausing, spurting, time and distance thigmotaxis are shown for the following numbers of fish measurements (from 5 cycles per zebrafish larva): JNK-IN-8: 346, SP600125: 224, Haloperidol: 220, SB216763: 358, FR236924: 360, R18: 354, SC79: 318, BIM: 216, PMA: 115, SB590885: 206, bryostatin-1: 359, AKTi: 360. P-values were calculated by Wilcoxon Rank Sum test and adjusted with Benjamini-Hochberg procedure and are indicated as follows * p-value≤0.05; ** p-value *≤ 0.01; ***p-value≤0.001; ****p-value≤0.0001.\u003c/p\u003e","description":"","filename":"Ffig5.png","url":"https://assets-eu.researchsquare.com/files/rs-3830013/v1/2e997ad765bd71959391005e.png"},{"id":49384073,"identity":"1f3259e2-6122-42c2-911f-9c705da79303","added_by":"auto","created_at":"2024-01-09 19:45:23","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":132796,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMachine learning predicts compounds with high sensitivity and specificity for generating an AA phenotype.\u003c/strong\u003e (A)\u003cstrong\u003e \u003c/strong\u003eROC curves show the Support Vector Machine (SVM) results for JNK pathway hub compounds to predict an “AA” or “other” phenotype during the Startle Response Behaviour (SRB) and the Post Startle Response Behaviour (PSRB). The number of fish measurements used for testing the machine learning models were JNK-IN-8: 346, SP600125: 224, Haloperidol: 220, SB216763: 358, FR236924: 360, R18: 354, SC79: 318, BIM: 216, PMA:, SB590885: 206, Bryostatin-1: 359, AKTi: 360. (B) A table summary of the area under the curve (AUC) values from Random Forest, Glmnet and SVM machine learning models are shown. Bold: Ranked highest overall taken from highest score across all machine learning approaches. The number of zebrafish tracks used for ML model training for the SRB class label “AA” were- diazepam: 350, fluoxetine: 335, imipramine: 343, LiCl: 210, and for class label “other” – ketamine: 349 and MK801: fish. During the post-startle response behaviour (PSRB) were as follows: class label “antidepressant or anxiolytic” – diazepam: 337, fluoxetine: 338, imipramine: 335, LiCl: 210; class label “other” – ketamine: 325 and MK801: 323. (C) A schematic summary of the zebrafish larvae behaviour screen for AA phenotype is shown. Beside it is a summary of results obtained using this test to evaluate the effect of JNK1 pathway hub drugs on AA-like behaviour.\u003c/p\u003e","description":"","filename":"Ffig6.png","url":"https://assets-eu.researchsquare.com/files/rs-3830013/v1/9af6bed8f04dabdd960d5ca8.png"},{"id":56566300,"identity":"70386204-b5e1-4984-9cab-c5822480a062","added_by":"auto","created_at":"2024-05-16 00:38:49","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3390306,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3830013/v1/fcaa0e9e-ec81-45b4-8baf-86cd95955f24.pdf"},{"id":49384076,"identity":"9dc10a6c-cffb-4451-a41b-31fb57fa4de4","added_by":"auto","created_at":"2024-01-09 19:45:23","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":7748906,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarymaterialandlegends.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3830013/v1/d2f65118ee8a030b3a4b33dd.pdf"},{"id":49384069,"identity":"8890e8ca-9265-441c-b4d4-62d7a2ff0381","added_by":"auto","created_at":"2024-01-09 19:45:23","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":255386,"visible":true,"origin":"","legend":"","description":"","filename":"Table1.docx","url":"https://assets-eu.researchsquare.com/files/rs-3830013/v1/4ee5eec079e3d1a7d07b61bb.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"\u003cp\u003eJnk1 and Downstream Signalling Hubs Regulate Anxiety-like Behaviours in a Zebrafish Larvae Phenotypic Screen\u003c/p\u003e","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eAnxiety and depression are highly prevalent mental disorders that involve abnormal neural function and dysregulation of circuits associated with threat and fear responses \u003csup\u003e1,2\u003c/sup\u003e. Together they account for over half of the global burden of mental disorders, and are among the largest of disease burdens overall \u003csup\u003e3,4\u003c/sup\u003e. Current treatments are lacking as not all patients achieve remission from their symptoms \u003csup\u003e5\u003c/sup\u003e, while side effects are common \u003csup\u003e6\u003c/sup\u003e. These disorders develop as a result of a complex interplay between genes and the environment \u003csup\u003e7,8\u003c/sup\u003e, and involve a wide variety of mechanisms \u003csup\u003e5,9\u003c/sup\u003e. One of the most consistent changes associated with depression is increased plasma cortisol and deregulation of hypothalamic-pituitary-axis homeostasis associated with desensitised cortisol receptor feedback mechanisms \u003csup\u003e10,11\u003c/sup\u003e. Clinically useful antidepressant drugs such as serotonin reuptake inhibitors (SSRIs) and tricyclic antidepressants have been shown to recover cortisol homeostasis \u003csup\u003e12\u003c/sup\u003e. A screening assay based on behaviour can potentially accelerate mode of action analysis and mechanistic understanding.\u003c/p\u003e \u003cp\u003ePhenotypic screening represents a strategy that can identify molecules with the ability to alter behaviour \u003csup\u003e13\u003c/sup\u003e. While mouse models are used to model disease, the time and resources needed to screen compounds in mice would be prohibitive, thus simpler systems are needed when large numbers of compounds should be tested. The zebrafish larvae as a model organism is emerging as a viable approach for screening of phenotypic behaviours \u003csup\u003e14\u0026ndash;16\u003c/sup\u003e. For analysis of stress-induced behaviour, the zebrafish hypothalamic-pituitary-interrenal (HPI) axis is cortisol responsive already by four days post fertilization, and like the HP-adrenal (HPA) axis in mammals, its dysregulation is associated with dysfunctional coping behaviours \u003csup\u003e17,18\u003c/sup\u003e. Also, zebrafish larva habenula and amygdala are involved in affective behaviours at an early stage \u003csup\u003e17\u003c/sup\u003e, and zebrafish brain neurotransmitters and corresponding receptors (including NMDA receptors) are expressed and functional at the larval stage. Zebrafish express orthologues for about 70% of human genes with 47% of these genes bearing a 1:1 relationship \u003csup\u003e19,20\u003c/sup\u003e, compared to 80% in mice \u003csup\u003e21\u003c/sup\u003e. The larvae display stereotypical behavioural responses such as thigmotaxis which is an evolutionarily conserved anxiety behavioural response \u003csup\u003e22\u003c/sup\u003e, and hyper-locomotor activity in response to neuroactive drugs \u003csup\u003e15,16,23\u003c/sup\u003e. For these reasons it is used for central nervous system drug screens at an early developmental stage which is suitable for scaleup \u003csup\u003e18,24\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe JNK1 pathway represents a potential regulator of anxiety and depression. In brain there are three \u003cem\u003eJNK\u003c/em\u003e genes and 10 splice variants expressed. Among these, JNK1 is known to play important roles during morphogenesis and in axonal/dendritic architecture modelling during brain development \u003csup\u003e25\u0026ndash;28\u003c/sup\u003e. In nervous tissue, JNK has been shown to be activated by endocrine stress leading to retraction of dendritic spines, and JNK-inhibitor treated or \u003cem\u003eJnk1-/-\u003c/em\u003e mice show reduced anxiety- and depressive-like behaviours \u003csup\u003e29\u0026ndash;31\u003c/sup\u003e. Moreover, JNK signalling may contribute to glucocorticoid resistance and HPA-axis dysregulation underlying anxiety and depression \u003csup\u003e32\u003c/sup\u003e, as JNK phosphorylation of the glucocorticoid receptor blocks its nuclear translocation and subsequent gene regulation \u003csup\u003e33,34\u003c/sup\u003e. Also, JNK activity is reduced by the fast-acting anti-depressant ketamine when administered to hippocampal neurons in culture \u003csup\u003e31\u003c/sup\u003e. However, JNK1 is expected to phosphorylate a large number of downstream targets in brain \u003csup\u003e28\u003c/sup\u003e, and whether any of these downstream targets contribute to anxiety and depressive-like responses is not known.\u003c/p\u003e \u003cp\u003eIn this study we identify signalling hubs downstream of JNK1 by comparing the mouse brain phosphoproteome from wild-type and \u003cem\u003eJnk1-/-\u003c/em\u003e mice. We then utilize a zebrafish larvae behavioural assay to test whether any of these downstream hubs phenocopy known antidepressant and anxiolytic drugs, and JNK1 inhibition. This identifies downstream players from JNK1 that phenocopy AA-like stereotypic responses in a zebrafish larvae screen.\u003c/p\u003e"},{"header":"RESULTS","content":"\u003cp\u003e\u003cstrong\u003eEstablishing an acoustic/light stimulus battery to profile zebrafish larvae behavioural responses to anti-depressant and anxiolytic drugs \u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe developed a\u0026nbsp;multiparameter test battery\u0026nbsp;to assess the behaviour of zebrafish larvae\u0026nbsp;in response to light and acoustic stressors.\u0026nbsp;For this,\u0026nbsp;larvae were placed individually in\u0026nbsp;square\u0026nbsp;wells of a 96-well plate and exposed to blue (470 nm), red (635 nm), purple (a combination of red and blue lights), white light,\u0026nbsp;or\u0026nbsp;flickering light,\u0026nbsp;combined with\u0026nbsp;light or heavy tapping\u0026nbsp;as indicated (Fig.\u0026nbsp;1A\u0026nbsp;and\u0026nbsp;B).\u0026nbsp;Zebrafish larvae exhibited robust motility from 4\u0026nbsp;days post-fertilization (dpf)\u0026nbsp;onwards\u0026nbsp;for all stimuli, except\u0026nbsp;for\u0026nbsp;white light where older larvae\u0026nbsp;were less responsive\u0026nbsp;(Fig. 1C). We expanded the test to include five cycles, each followed by a 29-minute intermission\u0026nbsp;post\u0026nbsp;startle (Fig. 1D). This enabled\u0026nbsp;us to monitor stress, recovery, and habituation responses.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClassical and\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003efast acting anti-depressant\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eand anxiolytic\u0026nbsp;drugs\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;alter zebrafish larvae motility\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe next measured motility of zebrafish larvae following a 1-hour incubation with classical \u003cu\u003ea\u003c/u\u003enti-depressant or \u003cu\u003ea\u003c/u\u003enxiolytic (AA) drugs: fluoxetine (a SSRI), imipramine (an antagonist \u0026nbsp;against the serotonin transporter \u0026gt; noradrenaline transporter \u0026gt; dopamine transporter), LiCl (which targets inositol monophosphatase and GSK3 \u003csup\u003e35\u003c/sup\u003e), diazepam\u0026nbsp;(a Gamma Amino Butyric Acid receptor antagonist)\u0026nbsp;or low dose (1 \u0026ndash; 10 \u0026micro;M) ketamine;\u0026nbsp;a\u0026nbsp;fast-acting antidepressant,\u0026nbsp;or\u0026nbsp;high dose (100 \u0026micro;M) ketamine\u0026nbsp;or\u0026nbsp;MK801\u0026nbsp;(also known as dizocilpine)\u003csup\u003e36\u003c/sup\u003e; NMDA receptor antagonists with psychotic activities (Fig. 1E-K; Table 1). Additional data for each drug treatment (i.e. individual fish tracking plots and quantitative output from individual cycles) is found in supplementary figures 1 to 6. The responses to the anti-psychotic drug Haloperidol responses are shown for comparison with the AA drugs (Supplementary figure 7). Motility responses during the 1-minute startle period were quantified from six peak clusters (P1 to P6) (Fig. 1E).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eClinically relevant doses of the AA drugs reduced the distance moved during the startle period (Fig. 1F-K). Even lithium at 1 mM showed a response similar to other AA drugs, while at 10 mM (above the therapeutic range), LiCl increased distance moved, possibly due to engagement of additional targets. Diazepam, with its muscle relaxant properties, caused a more significant motility inhibition than other AA drugs. However, it induced an unexpected increase in motility in response to purple light plus tapping (P5) (Fig. 1H), resembling the paradoxical hyperactivity or agitation effect seen in some individuals in response to diazepam \u003csup\u003e16,37\u003c/sup\u003e. Overall the motility profiles from the AA drugs (at the clinically relevant dose) were similar.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eKetamine elicits an antidepressant effect in humans at steady state plasma levels of around 1 \u0026micro;M, and an anaesthetic effect at an average dose of 9.3 \u0026micro;M in humans \u003csup\u003e38,39\u003c/sup\u003e. We therefore treated larvae with doses in this range. During the acute startle phase, ketamine decreased motility at doses of 10 \u0026micro;M and above for peaks P1-P3 and increased motility for peaks P4-P6. This closely matched the effect of MK801 and is therefore likely to be mediated through an NMDA receptor antagonist effect as both drugs share the phencyclidine site on (Fig. 1J-K). The response at lower doses of MK801 is consistent with the higher affinity of MK801 for NMDA receptor binding (Ki = 30.5 nM for MK801 verses 417 nM for ketamine) \u003csup\u003e38\u003c/sup\u003e.\u0026nbsp;In summary,\u0026nbsp;treatment of zebrafish larvae with\u0026nbsp;classical\u0026nbsp;AA\u0026nbsp;drugs suppressed\u0026nbsp;their\u0026nbsp;motility\u0026nbsp;responses during the test battery,\u0026nbsp;at doses\u0026nbsp;that are clinically relevant\u0026nbsp;for\u0026nbsp;an anti-depressant/anxiolytic effect in mammals.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDistance, turning, pausing and spurting are altered by AA drugs during the startle period\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe next examined the behavioural syllables that make up the motility response. Thus, in addition to distance travelled, we also measured turning (sum of turning angles), pausing (total time spent pausing), spurt velocity, time thigmotaxis and distance thigmotaxis near the tank border (Fig. 2A). We characterised the effect of AA drugs on this expanded set of features using 411 zebrafish larvae behaviour tracks (Fig. 2B). Fluoxetine, imipramine, LiCl and diazepam elicited strikingly similar behavioural syllables during the startle period (Fig. 2B). All AA drugs increased turning behaviour and decreased thigmotaxis distance and time, and spurting behaviours at lower doses. Only diazepam (at 1 \u0026micro;M or higher) diverged somewhat in that thigmotaxis distance was increased. Ketamine mimicked MK801 at high doses (100 \u0026micro;M) in that both increased thigmotaxis, whereas turning increased with 10 \u0026micro;M ketamine or 1 \u0026micro;M MK801, consistent with the effect being mediated by the NMDA receptor for which MK801 has higher affinity. Notably, increased turning was a prominent feature of the AA-drug treated fish (Fig. 2B). MK801 induced a large increase in motility at 10 \u0026micro;M consistent with the known hyperlocomotor effect of NMDAR inhibition in rodents and fish \u003csup\u003e38,40-42\u003c/sup\u003e, however\u0026nbsp;when measuring the startle phase only, the overall effect of\u0026nbsp;ketamine\u0026nbsp;on motility was minor, even though for stimuli it increased motility (Supplementary figure 6A, B).\u0026nbsp;Interestingly, during the post-startle phase shown in figure 5, MK801 and ketamine responses were almost identical.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIdentification of JNK\u003c/strong\u003e\u003cstrong\u003e1\u0026nbsp;pathway signalling hubs\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHaving established phenotypic responses to clinically used AA drug classes in zebrafish larvae, we next turned our attention to the JNK1 pathway. As JNK1 regulates many physiological processes in brain, we were interested to identify downstream signalling hubs that could serve as alternative, possibly more specific AA drug targets than JNK1 itself. Initially we identified JNK1-regulated phosphoproteins by comparing the phosphoproteome from wild-type (WT) and \u003cem\u003eJnk1-/-\u003c/em\u003e mouse brains. Phosphoproteins that differed significantly by more than 1.5 fold in \u003cem\u003eJnk1\u003c/em\u003e knockout brain are presented as a circular array (Fig. 3A). We next predicted the most highly connected phosphoproteins to the JNK1-regulated ones using the GeneMANIA physical interaction network database. The predicted proteins that physically interact with the JNK1-regulated phosphoproteins are displayed at the centre of the circle (Fig. 3A). From the identified interacting proteins, we determined 29 candidate hubs from the JNK1-regulated phosphoproteome based on interaction counts and interaction confidence levels. To assess the significance of these interactions, we compared interaction counts for JNK1-regulated phosphoproteins to 1000 randomly generated datasets derived from the entire brain phosphoproteome (Fig. 3B). These hubs demonstrated significantly higher connectivity in \u003cem\u003eJnk1-/-\u003c/em\u003e mouse brains compared to WT brains. From among the hubs, we further selected those for which small molecule drugs were available for testing in the zebrafish behavioural screen. The selected hubs included YWHAZ/E (14-3-3s), AKT1, GSK3B, BRAF, and PKCE/G. \u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eJNK\u003c/strong\u003e\u003cstrong\u003e1\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;pathway hub compounds\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003einduce\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;AA-like\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003estereotypic behaviours during\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003ethe startle period\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe next screened JNK1 pathway drugs during the startle period. JNK inhibitors SP600125 at 1 \u0026micro;M, and JNK-IN-8 at 10 nM mimicked the AA drug profiles (Fig. 4A). Downstream of JNK1, we targeted signalling hubs 14-3-3 and AKT using R18, an inhibitor of 14-3-3 client binding \u003csup\u003e43\u003c/sup\u003e and SC79 (an AKT activator). This also induced a profile similar to that obtained with clinically used AA drugs. In contrast AKTi, an inhibitor of AKT showed an opposite effect on distance and pausing and less effect on turning (Fig. 4A). GSK-3 inhibitor (SB216763) at 1 \u0026micro;M and LiCl at 1 mM, both known to inhibit GSK-3, produced similar responses. We also screened PKC-targeted drugs. PKC-epsilon activator (FR236924)\u003csup\u003e44\u003c/sup\u003e at 0.1 \u0026micro;M increased turning and pausing while decreasing distance (Fig. 4A). Bryostatin-1, a mixed PKC agonist, also increased turning, but differed from FR236924 in that it reduced spurting. Also at 10 nM (a dose that activates PKCa and PKCd isoforms, but antagonizes PKCg), it increased thigmotaxis and decreased pausing. Conversely, the pan-PKC inhibitor 3,4-Bis(3-indoyl)maleimide (BIM) increased turning and decreased thigmotaxis at 100 nM, a dose that inhibits PKCa more effectively than PKCe \u003csup\u003e45\u003c/sup\u003e,\u003c/p\u003e\n\u003cp\u003eWe also evaluated whether JNK regulated HPI axis stress in this model. To this end, we used saline stress which has been shown to elicit cortisol-dependent motility in zebrafish larvae that is HPI axis dependent \u003csup\u003e46\u003c/sup\u003e. This treatment increased the distance moved and spurting, and decreased turning (Fig. 4B, C). Notably, pre-treatment with JNK inhibitor SP600125 (10 \u0026micro;M) for 1 hour significantly reduced these stress-induced behaviours (Fig. 4C). This identifies turning, pausing, spurting and distance as features that are JNK and HPI axis dependent from zebrafish motility tracking.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDrug behaviour\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;profiles during the post-startle period\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe post-startle period is also relevant for stress responses and habituation \u003csup\u003e47\u003c/sup\u003e, therefore we examined zebrafish behaviour during this period. Clinically validated drugs produced behaviours similar to those during the startle period but with a more pronounced reduction in distance moved, increased pausing, and milder effects on thigmotaxis (Fig. 5A). Turning behaviour during the post-startle response period increased more linearly with dose, and at 10 \u0026micro;M, ketamine\u0026apos;s behaviour profile resembled classical AA drugs in this period. Among the JNK1 pathway hub compounds, several induced anxiolytic-like profiles during the post-startle period (Fig. 5B). For example, the 14-3-3 inhibitor R18 (0.1 \u0026micro;M) mimicked classical AA drug profiles, reducing distance moved while increasing turning and pausing. The selective PKC\u0026epsilon; activator FR236924 (0.1 \u0026micro;M) showed a similar effect, decreasing distance and thigmotaxis while increasing turning and pausing. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eJNK inhibitor SP600125 (1 \u0026micro;M) reduced distance, decreased thigmotaxis, and increased pausing, while the less well characterised JNK-IN-8 at 0.01 \u0026micro;M, had no significant effect. The AKT activator SC79 (2 \u0026micro;M) produced a response similar to JNK inhibitor and AA drugs, while the AKT inhibitor AKTi (50 \u0026micro;M) induced an opposite response, possibly representing anxiogenic-like behaviours. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eGSK-3 inhibitor SB216763 mimicked the response of AKT activator SC79. The PKC activators FR236924 (0.1 \u0026micro;M) and PMA evoked similar responses, although Bryostatin-1, which activates some PKC isoforms, differed substantially in profile \u003csup\u003e45\u003c/sup\u003e. Notably, NMDA receptor antagonists MK801 and high-dose ketamine showed markedly different dose response profiles during and after the startle period (Fig.\u0026nbsp;5), with post-startle responses aligning more closely with AA drugs. AKTi, BIM, and Bryostatin-1 did not fit the AA phenotype, as expected, whereas haloperidol fit both \u0026ldquo;AA\u0026rdquo; and \u0026ldquo;other\u0026rdquo; classification.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMachine learning identifies compounds with anti-depressant/anxiolytic classification\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn addition to traditional statistical analysis, we employed machine learning to create a comprehensive AA phenotype classification. For this we used Receiver Operating Characteristic (ROC) curves to assess the \u0026quot;fit\u0026quot; of test drugs to this classification. This approach has the advantage of overcoming the limitations and biases of regular statistical analysis and enhances the interpretation of complex data. We tested three supervised learning models: Random Forest classification-regression, GLMNET linear-regression, and SVM non-linear-regression. The SVM non-linear regression model achieved the highest training accuracy (\u0026gt;0.99) for both the startle and post-startle response periods with AA drugs, making it the chosen model to validate the test compounds.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eUsing motility tracking from the startle response period, we foud that SP600125, JNK-IN-8, SC79, GSK-3 inhibitor (SB216763), PKC isoform activators (FR236924 and PMA), and the 14-3-3 inhibitor (R18) classified as \u0026quot;AA\u0026quot; drug types with an area under the curve (AUC) of \u0026gt;97% in the SVM ROC analysis. For JNK inhibitors, this aligned with mouse data where JNK1 inhibition exerts an anxiolytic effect (Fig. 6A, B) \u003csup\u003e29\u003c/sup\u003e. Conversely, inhibitors of the same molecules (BIM, AKTi, SB59085, and bryostatin-1) did not fit the \u0026quot;AA\u0026quot; category. Interestingly, Haloperidol, which is a D2 receptor antagonist used for its anti-psychotic drug effect, classified with both \u0026quot;AA\u0026quot; and \u0026quot;other\u0026quot; phenotypes.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWhen using post-startle response data, the accuracy of predictions for test drugs was generally higher. Thus, SC79, R18, SB216763, PMA, and FR236924 scored highly for an AA phenotype in the post-startle period, even surpassing the scores of JNK inhibitors, suggesting that greater functional specificity is achieved by targeting downstream of JNK1. A summary of the multiparametric zebrafish larvae behaviour test and the impact of signalling hubs downstream of JNK on the AA phenotype are outlined in Figure 6C.\u003c/p\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eIn this study, we establish a multi-parameter behavioural screen to classify a common AA phenotype in zebrafish larvae using clinically proven AA drugs. We use distinct groups of classical anti-depressant drugs; fluoxetine, representing a commonly used serotonin reuptake inhibitor; imipramine, a tricyclic anti-depressant, and ketamine, a fast-acting anti-depressant, as well as the mood stabilizing drug lithium chloride and diazepam, a commonly used anxiolytic of the benzodiazepine class. Despite having distinct pharmacological profiles, these drugs elicit similar effects on zebrafish larvae motility patterns. All AA drugs reduce distance moved, thigmotaxis and spurting, while they increase turning. This differs from the behavioural responses to psychosis-inducing drugs MK801 and high dose ketamine NMDA receptor antagonist drugs and anti-psychotic Haloperidol, and saline-induced HPI axis stress, all of which produce opposing behaviours in the same test battery. We employ machine learning to characterise an \u0026ldquo;AA phenotype\u0026rdquo; using all of these behavioural features, and go on to use the assay to evaluate the signalling hubs downstream of JNK1. The results uncover new regulators downstream of JNK1 that align with the classical AA-drug phenotype performing even better than JNK inhibitors, suggesting increased functional specificity.\u003c/p\u003e \u003cp\u003eDysregulation of the hypothalamic-pituitary-adrenal axis leading to hypercortisolemia is one of the main circuit irregularities contributing to anxiety disorders and major depression \u003csup\u003e48\u0026ndash;51\u003c/sup\u003e. This hormonal change can trigger a cascade of events including the activation of JNKs. Accordingly, JNK is activated by cortisol and induces pro-inflammatory cytokines as well as amygdaloid dendritic hypertrophy \u003csup\u003e31,52\u0026ndash;54\u003c/sup\u003e. Additionally, JNK plays a role in the homeostatic regulation of glucocorticoid receptor transcriptional activity \u003csup\u003e25,28\u003c/sup\u003e. In contrast, when JNK is inhibited in mouse brain, a reduction in anxiety- and depressive-like behaviours is observed \u003csup\u003e55\u003c/sup\u003e, as is the case with JNK inhibitor treatment or genetic deletion \u003csup\u003e29,56\u0026ndash;58\u003c/sup\u003e. In contrast, chimeric JNK activation induces depressive-like behaviour and impaired assessment of risk/reward benefit \u003csup\u003e58,59\u003c/sup\u003e. In zebrafish brain at the larval stage, JNK1 orthologs (\u003cem\u003emapk8a\u003c/em\u003e and \u003cem\u003emapk8b)\u003c/em\u003e are expressed \u003csup\u003e60\u003c/sup\u003e, and consistent with behavioural findings in rodents \u003csup\u003e54\u003c/sup\u003e, we find that treatment with structurally independent JNK1 inhibitors SP600125 and JNK-IN-8, induce an AA phenotype with A.U.C.s of 97.9% and 98.3% respectively in the startle period. Thus, JNK inhibitors elicit a behavioural response in zebrafish larvae that matches closely to that of clinically used AA drugs.\u003c/p\u003e \u003cp\u003eA main goal of this study was to identify signalling hubs downstream of JNK1 influencing anxiety- and depressive-like behaviours. We discovered 14-3-3ζ (YWAZ) and 14-3-3ε (YWHAE), members of the 14-3-3 client binding family \u003csup\u003e61\u003c/sup\u003e, as key hubs in this process. JNK is known to phosphorylate 14-3-3ζ leading to client protein release \u003csup\u003e62\u003c/sup\u003e. Consistent with this, the interaction count for ζ and ε 14-3-3 isoforms increases in \u003cem\u003eJnk1-/-\u003c/em\u003e mouse brain, indicating that JNK1 regulates 14-3-3 binding dynamics in the CNS. In zebrafish larvae, we show that R18 peptide which disrupts 14-3-3 client binding \u003csup\u003e63\u003c/sup\u003e, recapitulates the AA-drug phenotype with A.U.C. of 93.5% and 97.2% in the startle and post-startle period respectively, consistent with the involvement of ζ or ε isoforms \u003csup\u003e61,64\u003c/sup\u003e. 14-3-3ε has also been shown to mediate chronic stress-induced depression \u003csup\u003e65\u003c/sup\u003e. Interestingly, fluoxetine rescues the behavioural deficit in zebrafish lacking 14-3-3ζ suggesting a shared mechanism with fluoxetine \u003csup\u003e66\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eWe also identify AKT as a potential signalling hub downstream of JNK1 based on mouse brain phosphoproteome interaction network analysis. AKT and 14-3-3 are mechanistically linked, thus inhibition of 14-3-3 client binding activates AKT \u003csup\u003e67,68\u003c/sup\u003e, potentially placing AKT regulation by JNK1 downstream of 14-3-3. We found that the AKT activator SC79 evoked an AA phenotype with A.U.C.s of 99.4% and 98.3% for the startle and post-startle response periods in the zebrafish larva screen. Conversely, the allosteric inhibitor of AKT (AKTi) induced an opposite response (A.U.C. 60.3% and 60.4% respectively). That AKT activation induces an AA phenotype fits with the well-documented anti-depressant action of AKT in animal models and depressed subjects, and its regulation of BDNF \u003csup\u003e69,70\u003c/sup\u003e. Consistent with this, we also identify that GSK-3β is a hub downstream of JNK1 in mouse brain. This is not surprising as GSK-3α/β isoforms are phosphorylated directly by AKT on S21 and S9 (α and β isoforms respectively), leading to GSK-3 inhibition \u003csup\u003e71\u003c/sup\u003e. In the zebrafish larva assay, the GSK-3α/β inhibitor SB216763, elicits an AA phenotype with A.U.C.s of 97.6% and 99.0% during the startle and post-startle response phases.\u003c/p\u003e \u003cp\u003eB-Raf was another signalling hub identified downstream of JNK1. In the interaction screen, B-Raf displayed an increase in protein interactions (from \u0026ldquo;0\u0026rdquo; to \u0026ldquo;10\u0026rdquo;) in the phosphoproteome from \u003cem\u003eJnk1-/-\u003c/em\u003e mouse brain compared to wild-type. However, when we tested the B-Raf inhibitor SB590885, the resulting phenotype did not classify as \u0026ldquo;AA\u0026rdquo; either during the startle or post-startle period. Interestingly, B-Raf is also known to interact with 14-3-3, which functions to maintain its kinase domain in an inactive state \u003csup\u003e72\u003c/sup\u003e. The expectation therefore would be that activation of B-Raf upon disruption of 14-3-3 binding may allow it to contribute to an anxiolytic-like state, as observed with the 14-3-3 client binding inhibitor R18. However, it was not possible for us to test B-Raf further due to lack of activators for this molecule. The oncogenic nature of B-Raf activating mutations contraindicates the use of such molecules.\u003c/p\u003e \u003cp\u003eFinally, we tested PKC activators and inhibitors in the zebrafish screen to explore their impact as possible effector hubs downstream of JNK1. We used the PKCε-activator FR236924 and the pan-PKC activator PMA, both of which evoked a clear anxiolytic-like effect (A.U.C.s of 97.8% and 95.7% respectively). Consistent with this, the broad specificity PKC inhibitor (BIM) had the opposite effect. Notably however, in some contrast with our findings in zebrafish larvae, BIM treatment attenuates corticotrophin releasing factor (CRF) facilitation of the acoustic startle response in mice, although it had no effect in the absence of CRF \u003csup\u003e73\u003c/sup\u003e. Moreover, \u003cem\u003ePKCγ\u003c/em\u003e-/- mice exhibit low anxiety \u003csup\u003e74\u003c/sup\u003e. Interestingly also, that PKCε activator (FR236924) promotes an AA phenotype in zebrafish is supported by the response to byrostatin-1 at 100 nM, (a dose that activates PKCε \u003csup\u003e75\u003c/sup\u003e) during the startle response, even though when analysed at the full dose response in our study the overall phenotype is \u0026ldquo;other\u0026rdquo;. While these studies suggest isoform-specific roles for PKC variants in anxiety-like behaviours, we identify that PKCε represents a signalling hub downstream from JNK1 which controls an AA phenotype in zebrafish larvae, indicating that further analysis of this isoform in the context of stress-related behaviours is warranted.\u003c/p\u003e \u003cp\u003eThe observation of overlapping behavioural phenotypes elicited by drugs with distinct pharmacologies merits discussion. The primary target of fluoxetine is to block serotonin reuptake, and that of imipramine is to block serotonin and norepinephrine reuptake, and to additionally antagonise dopamine D2 and acetylcholine receptors \u003csup\u003e76\u003c/sup\u003e. These drugs take weeks to elicit a notable effect on neurotransmitter levels in both responders and non-responders and other effects such as neuroplasticity and neurogenesis changes may be additionally needed for the therapeutic effect \u003csup\u003e76,77\u003c/sup\u003e. A recent study identified that the Tropomyosin receptor kinase (TrkB) is a common binding site for a range of anti-depressant drugs including fluoxetine, imipramine and ketamine. The authors demonstrated that binding of these drugs to TrkB occurs within minutes and is required for the anti-depressant-like effect on behaviour \u003csup\u003e78\u003c/sup\u003e. The authors rationalized that the well studied neuroplasticity effects of BDNF, mediated via TrkB binding, could explain the anti-depressant effect and its timeline. TrkB could represent the common mediator of drug responses measured here as zebrafish larvae indeed express TrkB orthologs (\u003cem\u003etrkB1\u003c/em\u003e and \u003cem\u003etrkB2\u003c/em\u003e) from 6 days post fertilization onwards \u003csup\u003e79\u003c/sup\u003e. Moreover, activation of TrkB leads within minutes to the activation of AKT \u003csup\u003e69\u003c/sup\u003e. We uncover that JNK1 signalling hubs converge on the AKT pathway with 14-3-3ζ/ε and GSK-3 in addition to AKT itself identified. Moreover, treatment of zebrafish larvae with the AKT activator (SC79) evokes an AA phenotype with closes match to the clinically used AA drugs with an A.U.C. of 99.4%. AKT pathway activation is the most likely explanation for the temporal and phenotypic overlap observed in response to AA and JNK1 hub drugs.\u003c/p\u003e \u003cp\u003eAnimal models are limited when it comes to modeling complex disorders such as anxiety and depression. They do not account for subjective experience, or complex genetic and environmental background, and the models that do exist in mice require large numbers to attain statistical power \u003csup\u003e80,81\u003c/sup\u003e. Here we present an assay that is more straightforward and scalable than rodent models. It measures stereotypic responses to acoustic and light stress, based on larval behavioural sequalae (distance, turning, pausing, spurting, thigomataxis distance, thigmotaxis time, dose and size effect). This may lack face and construct validity, and it is not designed to model a specific disease mechanism. Nonetheless, it does measure HPI axis stress responses and the neuroendocrine and monoamine control systems underlying the stress axis is conserved from chordates to mammals \u003csup\u003e82\u003c/sup\u003e, moreover locomotor activity in zebrafish larvae is inherently cortisol dependent \u003csup\u003e12\u003c/sup\u003e. A similar approach has been used to identify novel anti-psychotic compounds in zebrafish larvae \u003csup\u003e15\u003c/sup\u003e, and to investigate the paradoxical excitatory effects of GABA and serotonin \u003csup\u003e16\u003c/sup\u003e, for example. Interestingly, here we also observe a similar hyperactivity effect with the GABAR agonist diazepam when fish are exposed to violet light and acoustic stimuli together (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eH). Our study in addition exploites machine learning to classify the desired phenotype, enabling use of complex combinations of features. The result is high accuracy prediction scores for test drugs that are known to be anxiolytic/anti-depressant in mice e.g. AKT activator and GSK-3 inhibitor \u003csup\u003e69,70,83\u003c/sup\u003e. This highlights the potential of the zebrafish larvae screen to potentially identify compounds that mimic AA drugs, for further testing in rodent models, which is rather the goal of a phenotypic screen.\u003c/p\u003e \u003cp\u003eA possible explanation for the high prediction score could be linked to the dependence of the zebrafish larvae locomotor response on the HPI-axis \u003csup\u003e46\u003c/sup\u003e, and the drugs used regulate the HPA-axis \u003csup\u003e84\u003c/sup\u003e. Although the stereotypic behaviours measured here may not directly relate to AA conditions, JNK is activated by cortisol \u003csup\u003e30,85\u003c/sup\u003e leading to synapse withdrawal \u003csup\u003e31\u003c/sup\u003e, which may influence circuit behaviour and long-term mental state. Notably also, the phenotype of AA-drugs differ from those of psychotic compounds MK801 and ketamine. Similarly, the anti-psychotic D2 antagonist Haloperidol shows dual classification which is consistent with its known paradoxical side effect, whereby some patients can respond with worsening of symptoms such as increased agitation, insomnia and hallucinations in patients \u003csup\u003e12\u003c/sup\u003e. Consistent with this in the zebrafish larvae, we find that Haloperidol shows a distinct hyper-motility phenotype, which possibly accounts for its classification as \u0026ldquo;other\u0026rdquo;; yet it also increases turning and diminishes spurting, consistent with the \u0026ldquo;AA\u0026rdquo; phenotype (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e; Supplementary Fig.\u0026nbsp;7). This paradoxical effect to Haloperidol was also previously reported in zebrafish larvae \u003csup\u003e15\u003c/sup\u003e. A long-term advantage of the screen is the potential reduction in the number of mice needed for drug discovery, while expanding the utility of zebrafish larvae in neuropsychopharmacology studies.\u003c/p\u003e"},{"header":"MATERIALS AND METHODS","content":"\u003cp\u003e\u003cem\u003eZebrafish husbandry -\u003c/em\u003e WT zebrafish (\u003cem\u003eD. Rerio\u003c/em\u003e) larvae were derived from a cross between the high-performance AB strain and the Tu short-fin wild type strain \u003csup\u003e20\u003c/sup\u003e. Fish were maintained in the Turku Bioscience Zebrafish Core facility at 28.5\u0026deg;C, under a 12 h day/night cycle. Embryos were cultured in E3 medium (5 mM NaCl, 0.17 mM KCl, 0.33 mM CaCl\u003csub\u003e2\u003c/sub\u003e, 0.33 mM MgSO\u003csub\u003e4\u003c/sub\u003e) with 30 eggs per 10 cm diameter dish. Larvae of indeterminate sex were used from 3 to 8 dpf as indicated. Experiments were carried out in the afternoon to maximise consistency of locomotor behaviour \u003csup\u003e86\u003c/sup\u003e. Fish were bred in two tanks. At least two plates of fish were tested for every drug. There were minor differences in baseline motilities between plates therefore data was normalised to internal control fish measured on the same plate. Afterward the experiment, zebrafish larvae are anesthetized with 200mg/L Tricaine for 5min, and then euthanized by transferring to 1% chloramine ice-water solution. Behavioural experiments were carried out following the guidelines that were approved by the National Animal Experiment Board (ELLA) and the Regional State Administrative Office of Southern Finland (ESAVI) under licenses ESAVI/16343/2019 and ESAVI/30686/2022, which is the authority in Finland that ensures that animal experimentation is carried out following the law and thus follows EU ethical guidelines for animal experimentation. All all methods are performed and reported in accordance with the ARRIVE guidelines (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://arriveguidelines.org\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eBehavioural battery test \u0026minus;\u003c/em\u003e\u0026thinsp;7 days post fertilization (dpf) zebrafish larvae were transferred to square well 96-well plates and drug dosages were administered following which the plate was directly transferred to the DanioVision\u0026trade; device (Noldus Information Technology, BV, Wageningen, Netherlands) and acclimatized for 1 h in the dark at 28.5\u0026deg;C. Larvae then underwent the test battery consisting of 5 cycles of a 60 s string of acoustic/vibratory (tapping) and light stimuli programmed using Ethovision13 XT software (Noldus Information Technology, BV, Wageningen, Netherlands). The 60 s stimuli were as follows; high (H) and low (L) intensity patterns (1.6 \u0026amp; 4.2 watt): 1 s pause, 6x L-tap (0.5 s break between individual taps), 4 s pause, 4x H-tap, 1 s pause, 8x L-tap, 8 s pause, 2x L- tap, 10x H- tap, 7 s pause, 2x H-tap, 13 s pause, 14x H-tap, 3 s pause. Visual stimuli evoked using red (635 nm, 192 Lm), blue (470 nm, 240 Lm), purple (red and blue together) and white LEDs, switched colour at 10 s intervals in the following order: darkness, red, blue, purple (red \u0026amp; blue), white, and sequential flickering light of all colours sequentially at 0.1 to 0.2 s intervals. After the stimuli period there was a 29-minute intermission period. This 30 min sequence was repeated 4 more times. Digital video tracking was performed using the Daniovision imaging unit using a pre-programmed protocol with the included Ethovision XT software. Acquisition was 60 fps and exported as csv file for automated analysis using customized software in R. The data analyst was agnostic to the drug properties.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eStress response assay \u0026minus;\u003c/em\u003e\u0026thinsp;5 or 7 dpf larvae (as indicated) were transferred to a 48-well plate with 1 fish/well and acclimatized for 1 h following which they were exposed to stress: either with white light illumination (10,000 Lux) for 50 s or addition of 100 mM NaCl. Larvae were tracked for a further 30 min following stressor in the dark using infrared light. NaCl experiments were performed using ambient level (500 Lux) white light.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eDrug treatments \u0026ndash;\u003c/em\u003e Larvae at the indicated ages were treated with increasing doses of drugs and compounds as follows: fluoxetine, diazepam, FR236924 and PMA (Phorbol 12-myristate 13-acetate) and ketamine were from Tocris Bioscience, Abingdon, UK; haloperidol, imipramine, LiCl, SP600125, SB590885, Bryostatin-1 and MK801 from MERCK Sigma Aldrich Solutions, Darmstadt, DE; R18 from Enzo Life Sciences distributed by AH Diagnostics, Helsinki, FI; SB216763, JNK-in-8 and AKTi from Selleck Chemicals LLC, Houston, Texas; bisindolylmaleimide I (BIM) from Santa Cruz Biotechnology Inc., Dallas, Texas and SC79 from R\u0026amp;D systems, McKinley Place NE, Minneapolis or NaCl in E3 medium. Treatments were for 1h at 28.5 \u003csup\u003eo\u003c/sup\u003eC before commencing the startle battery. Control fish were treated with equivalent volume of carrier.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eBehaviour data analysis \u0026ndash; Motility\u003c/em\u003e: Raw motility tracking data was exported in csv format, where each file contained data from 1 experimental cycle, and each worksheet contained time series data from 1 fish. Relative motility index (M.I.\u003csub\u003eR\u003c/sub\u003e) (plots of left side) represents the motility during a certain condition (M\u003cspan type=\"SmallCaps\" class=\"SmallCaps\" name=\"Emphasis\"\u003ec\u003c/span\u003e\u003csub\u003ex\u003c/sub\u003e) relative to the motility during the control condition (M\u003cspan type=\"SmallCaps\" class=\"SmallCaps\" name=\"Emphasis\"\u003ec\u003c/span\u003e\u003csub\u003e0\u003c/sub\u003e). The relative motility index was calculated as follows from all 5 cycles:\u003c/p\u003e\n\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\n \u003cdiv class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e$${\\text{M}.\\text{I}.}_{\\text{R}}\\left({\\text{c}}_{\\text{x}}\\right)\\left[\\text{%}\\right]=\\frac{\\left(M{\\text{c}}_{\\text{x}}\\right)-\\left(M{\\text{c}}_{0}\\right)}{\\left(M{\\text{c}}_{0}\\right)}\\text{*}100$$\u003c/div\u003e\u003c/div\u003e\u003cp\u003e\u0026ldquo;Relative Peak Analysis\u0026rdquo; (shown in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e right side panels) represents the motility during peaks P1, P2, P3, P4, P5, P6 of a test condition relative to the same peaks from control fish in the same plate. The data scientist handling the analysis was agnostic to the drug treatments. Analysis was done in R.\u003c/p\u003e\u003cp\u003e\u003cem\u003eAnalysis of other features\u003c/em\u003e: X and Y positions, heading, and distance features from the time series data were utilized to calculate behavioural features: distance, turning, pausing, spurt velocity, and thigmotaxis by distance and time. \u003cem\u003eDistance\u003c/em\u003e \u0026ndash; Distance was summed from all time points per fish and mean total distance travelled per treatment dose was calculated. \u003cem\u003eTurning\u003c/em\u003e \u0026ndash; Turning angles between consecutive time points were calculated from the heading parameter of the last and current time points, and converted to magnitude representing the smaller angle in 360 degrees. Relative turning angles were calculated by summing all time points for each fish, then normalizing to distance travelled, and presented as average values according to drug dose. \u003cem\u003ePausing\u003c/em\u003e - Time points with value 0 for distance were summed to give a value for total pause time for each fish. \u003cem\u003eSpurt velocity\u003c/em\u003e - Time bouts between pauses were identified and the velocity for each time point was calculated from distance feature. The maximum velocity from each time bout was recorded and the mean maximum velocities from all time bouts in a single fish were calculated. \u003cem\u003eThigmotaxis\u003c/em\u003e - The well border coordinates from the 96-well plate were recorded from Ethovision graphic user interface (GUI). An inner border was (virtually) created that was 40% of the width and length of the original border, and placed in the centre of each well. The coordinates of these inner borders served as thresholds for thigmotactic behaviour. Time points where a fish travelled outside the inner border were identified and the ratio of near outer border travel versus total travel for each fish was calculated for time and distance. Anaysis was done in R.\u003c/p\u003e\u003cp\u003e\u003cem\u003eMachine learning analysis -\u003c/em\u003e Machine learning algorithms were applied to zebrafish larvae behaviour data during the 1 min startle period and during the 10 min of the recovery period. Specifically, the quantitative variables total pause duration, total distance traveled, relative turning angle, average spurt speed, total thigmotaxis time and total thigmotaxis distance were normalized against control (or zero-dosage fish) data as percentages. These normalized metrics taking into account effect sizes, along with dosage level variable, were incorporated as features for the machine learning models. The models were tasked to differentiate between \u0026ldquo;antidepressant or anxiolytic\u0026rdquo; drugs and \u0026ldquo;other\u0026rdquo; drugs. For the training dataset, drug treatments including diazepam, fluoxetine, imipramine, and lithium chloride (LiCl) were categorized as \u0026quot;antidepressant/anxiolytic\u0026quot; class, while ketamine and MK801 treatments were categorized as \u0026quot;other\u0026quot;. Generalized Linear Model via penalized maximum likelihood (Glmnet), Support Vector Machine (SVM) with a polynomial kernel, and Random Forest algorithms were utilized to build 3 separate models each for startle response behaviour (SRB) data during the 1 min startle battery and post-startle response behaviour (PSRB) data during the 10 min recovery period. The constructed models were subsequently employed to predict the class category of other drug treatments. Machine learning was performed with \u0026ldquo;caret\u0026rdquo; package within the R environment.\u003c/p\u003e\u003cp\u003e\u003cem\u003eStatistics -\u003c/em\u003e All behavioural features were compared to control fish from the same experiment. For statistical testing, feature values were collected into a distribution, and this data according to drug dose was compared to respective control distributions using Wilcoxon signed-ranked test or one-way or two-way ANOVA with Tukey or Dunnett\u0026rsquo;s multiple comparison test as indicated. Calculations and plots were made using a customized pipeline in R programming language. Data analysis was automated and the data scientist was agnostic to the treatment drugs.\u003c/p\u003e\u003cp\u003e\u003cem\u003ePreparation of samples for mass spectrometry, SDS-PAGE and in-gel digestion \u0026ndash;\u003c/em\u003e All the chemicals for digestion and mass spectrometry analysis were from Sigma-Aldrich (Stockholm, Sweden). Whole brain was isolated from WT and \u003cem\u003eJnk1-/-\u003c/em\u003e C57B6J mice \u003csup\u003e87\u003c/sup\u003e at embryonic day 15 (E15), post-natal day 0 (P0), post-natal day 21 (P21), and 8 months (Adult). We extracted brains from three mice for each age and genotype and snap froze them in liquid N\u003csub\u003e2\u003c/sub\u003e and then pulverized them in a Micro-dismembrator II (Braun Biotech International, Melsungen, Germany). We weighed the powder and stored aliquots at -80\u0026deg;C before use. SDS (1%), supplemented with protease and phosphatase inhibitors (Sigma-Aldrich, St. Louis, USA), was added to the brain-powder. Mechanical disruption followed using a syringe twice. We centrifuged samples for 1 hour at 4\u0026deg;C and collected the supernatant. We quantified protein concentration using the Total Protein Kit, Micro Lowry, Peterson\u0026apos;s Modification (Sigma-Aldrich, St. Louis, USA).\u003c/p\u003e\u003cp\u003eA total protein amount of 100 \u0026micro;g for each sample was separated on 12% Criterion gels (Bio-Rad Laboratories, Hercules, Ca, USA), following the manufacturer\u0026rsquo;s instructions. We washed the gel in milliQ, stained for 30 min with GelCode (Bio-Rad Laboratories) and washed overnight in milliQ before manually slicing each lane into 5 equal slices. We de-stained gel slices by washing 3 times in 25 mM ammonium bicarbonate (AMBIC)/50% acetonitrile and dried in a vacuum centrifuge (Speedvac). Samples were reduced (10mM DTT/100 mM AMBIC, 1 h at 56\u0026deg;C), alkylated (55 mM iodoacetamide (IAA)/100 mM AMBIC, 45 min at room temperature in the dark), washed 2 times with 100 mM AMBIC and dehydrated with ACN before being dried in the Speedvac. We rehydrated slices with 12.5 \u0026micro;g/ml modified porcine trypsin (Promega, Madison, WI, USA) in 50 mM AMBIC and digested overnight at 37\u0026deg;C. Peptides were eluted 2 times in 75% ACN/5% FA, dried 1 h in the Speedvac and dissolved in 0.1% formic acid.\u003c/p\u003e\u003cp\u003e\u003cem\u003eTiO\u003c/em\u003e \u003csub\u003e\u0026nbsp;\u003cem\u003e2\u003c/em\u003e\u0026nbsp;\u003c/sub\u003e \u003cem\u003ephospho-peptide enrichment \u0026ndash;\u003c/em\u003e we incubated samples for 10 min with Buffer A (1 M GA, 80% ACN, 5% TFA) followed by incubation with the TiO2 Mag Sepharose (GE Healthcare Life Sciences) for 30 min in an Eppendorf Thermomixer at 800 rpm. Liquid was removed and washing steps were done using Buffer A and 2 times using Buffer B (80% ACN, 1% TFA). Phospho-peptides enriched were eluted 2 times using 100 mM ammonium hydroxide in water and dried 1h in the Speedvac.\u003c/p\u003e\u003cp\u003e\u003cem\u003eLC-MS/MS analysis-\u003c/em\u003e Dried peptides were resuspended in 0.1% formic acid and separated using an Eksigent nano-LC 2D system (Eksigent, Dublin, CA, USA) consisting of a solvent degasser, a nano-flow pump and a cooled auto-sampler. Peptide concentrations were measured by Nanodrop (ThermoFisher, Stockholm) and the concentration adjusted to the same amount in all samples. Eight \u0026micro;l of sample was loaded and washed for 15 min onto a pre-column (Acclaim PepMap 100, C18, 3 \u0026micro;m particle size, 50 \u0026micro;m diameter, Thermo Fisher Scientific, H\u0026auml;gersten, Sweden) at a constant flow of 5 \u0026micro;l/min solvent B (0.1% FA in ACN). We ran three biological replicates and two technical repeats for each of the two genotypes at the four time points. Phosphopeptides were loaded into a RP analytical column (10 \u0026micro;m fused silica emitter, 75 \u0026micro;m ID column, 16 cm Pico Tip, New Objective) packed in-house with C18 material ReproSil-Pur and separated using an eighty-minute gradient from 3\u0026ndash;40% solvent B at a flow rate of 300 nl/min. The gradient was followed by 20 min column washing with 90% ACN, 0.1% FA and 15 min re- equilibration with 3% solvent B. The HPLC system coupled to an LTQ-Orbitrap XL mass spectrometer (ThermoFisher Scientific, Bremen, Germany) operated in Data Dependent Acquisition mode. Spray voltage was set to 1.90 kV and the temperature of the heated capillary was set to 200\u0026deg;C. The ten most intense ions from the survey scan performed by the Orbitrap were fragmented by collision-induced dissociation (CID) in the LTQ (normalized collision energy 35, parent mass selection window 0.5 Da, 30 ms activation time and minimum signal threshold for MS/MS scans set to 100). We excluded unassigned charge states and singly charged ions from fragmentation. The dynamic exclusion list was limited to a maximum of 500 masses with a retention time window of 2 minutes and a relative mass window of 10 ppm. The X-calibur software version 2.0.7 (Thermo Scientific) controlled the HPLC, mass spectrometer and data acquisition.\u003c/p\u003e\u003cp\u003e\u003cem\u003eShotgun data analysis from LTQ Orbitrap \u0026ndash;\u003c/em\u003e The raw data was uploaded into Progenesis LC-MS (Waters) for analysis. After raw mass spectra were aligned and normalisation across all runs was carried out by total ion current in the analysis area manually defined for the reference HPLC for the time and mass ranges. We submitted the raw files from the mass spectrometer to Thermo Proteome Discoverer (version 1.0.43.0) for protein identification by Mascot. Cysteine carbamidomethylation was set as a fixed modification, methionine oxidation and phosphorylation of serine, threonine, tyrosine (STY) were set as variable modifications. Data were searched against Uniprot \u003cem\u003eMus musculus\u003c/em\u003e database containing 170.598 sequences (release date April 2014). The mass tolerance was set to 10 ppm for the precursor ion and 0.8 Da for the fragment ions. Trypsin was set as the protease and a maximum 2 of missed cleavages were allowed. We ran raw data using an FDR of 5% and a score cutoff of 20. We discarded proteins marked as contaminants, reverse hits and \u0026ldquo;non-unique peptides\u0026rdquo;. We selected phospho-peptides with a Mascot score higher than 20 and verified using the default Mascot Delta score of 0.1 for comparative analysis between groups.\u003c/p\u003e\u003cp\u003e\u003cem\u003eData merging and statistical analysis phosphoproteomics data \u0026ndash;\u003c/em\u003e Downstream analysis of MS intensity outputs from Progenesis LC-MS (Waters) utilized a customized R (v.3.4.0) software pipeline. Phosphopeptide data were refined as follows: (i) all phosphopeptides from 5 gel slices were merged, (ii) identical phosphopeptides, i.e. those with identical sequence, UniProt identifier and number and position of phosphorylated residues, were merged by summing the intensities. (iii) If a peptide entry had\u0026thinsp;\u0026gt;\u0026thinsp;1 missing value across 3 replicates, all entries were removed. For quality control of data, a distribution plot of phosphopeptide intensities for each unique genotype at given ages was visualized using boxplot and histogram analysis, using ggplot2 \u003csup\u003e88\u003c/sup\u003e (Supplementary F1). \u0026ldquo;Phosphopeptide-centric\u0026rdquo; and phosphosite-centric analysis was performed as appropriate and is defined in figure legends. We used the Bioconductor package RankProd \u003csup\u003e89,90\u003c/sup\u003e for statistical analysis. We calculated mean ratios of phosphopeptide intensities [\u003cem\u003eJnk1-/-/\u003c/em\u003eWT] per entry (for both phosphopeptide-centric and protein-centric changes). Only phosphorylation intensity changes that passed a threshold Ratio\u003csub\u003e\u003cem\u003eJnk1\u0026minus;/\u0026minus;\u003c/em\u003e/WT\u003c/sub\u003e \u0026gt; 2, with p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 were included.\u003c/p\u003e\u003cp\u003e\u003cem\u003eHub analysis \u0026ndash;\u003c/em\u003e 29 candidate hubs (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eH, I) that were significantly altered in \u003cem\u003eJnk1-/-\u003c/em\u003e mouse brain compared to wildtype brain and were part of the psychiatric disease associated genes according to the MetaCore\u0026trade; database schizophrenia gene list were selected to analyse as potential signalling hubs acting downstream of JNK1 in regulating depressive/anxiety-like behaviour. The selection criteria included a threshold for phosphoproteins that showed either \u0026ge;\u0026thinsp;7 physical interactions, or \u0026gt;\u0026thinsp;3 high weight (weight\u0026thinsp;\u0026gt;\u0026thinsp;0.04) physical interactions from the physical interaction network built with GeneMANIA from among the significantly altered phosphoproteins derived from the wildtype and \u003cem\u003eJnk1-/-\u003c/em\u003e mouse brain phosphoproteomes. 29 proteins fulfilled these criteria. To control for hub protein\u0026apos;s inherent functional plasticity variations, 1000 control phosphoprotein networks of the same size as the \u003cem\u003eJnk1-/-\u003c/em\u003e psychiatric disease phosphoprotein network were randomly generated from the entire brain phosphoproteome dataset using GeneMANIA (v.3.4.1). The number of interactions of the 29 proteins from the significantly altered phosphoproteome network in \u003cem\u003eJnk1-/-\u003c/em\u003e mouse brain was each compared to the distribution of interaction count from 1000 randomly generated control networks to determine the significance of the hubs. Statistical significance for the hubs was calculated using the hypergeometric test. P-values were adjusted using the Benjamini-Hochberg procedure.\u003c/p\u003e\u003cp\u003e\u003cem\u003eReceiver Operating Characteristic (ROC) analysis -\u003c/em\u003e ROC curves were constructed to visualise the predictive proficiency of the machine learning models. Individual curves were plotted for each test drug from the test dataset, paired with a control drug that was either classified as a \u0026quot;positive control\u0026quot; or a \u0026quot;negative control\u0026quot;. Each treatment drug is assigned with a \u0026quot;true\u0026quot; class label that reflected its known properties and molecular interactions. Drugs of the \u0026quot;antidepressant or anxiolytic\u0026quot; class were coupled with the \u0026quot;negative control\u0026quot; drug ketamine, which provided the \u0026quot;true\u0026quot; class label of \u0026quot;other\u0026quot;. Conversely, drugs classified under the \u0026quot;other\u0026quot; class were paired with the \u0026quot;positive control\u0026quot; drug fluoxetine for the \u0026quot;true\u0026quot; class label of \u0026quot;Antidepressant or anxiolytic\u0026quot; for the purposes of ROC curve calculation. The construction of the ROC curve was facilitated by the \u0026apos;pROC\u0026apos; package within the R statistical programming environment.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cem\u003eData availability-\u0026nbsp;\u003c/em\u003eThe phosphoproteomics data generated during the current study is available in the PRIDE repository, (www.ebi.ac.uk/pride/) and can be accessed with the following username and password [email protected]; Password: qNDO1jC9.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAUTHOR CONTRIBUTIONS\u003c/p\u003e\n\u003cp\u003eB.H., C.S. J.P. and J.J. carried out experimental work and data analysis. Y.H. carried out the phosphoproteomic and bioinformatics data analysis, machine learning. C.S., Y.H. and E.C. wrote the manuscript. All authors contributed to the manuscript and figure preparation. I.P. supervised the fish husbandry and contributed to data interpretation. E.C. conceptualized and supervised the project. \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFUNDING:\u003c/p\u003e\n\u003cp\u003eAcademy of Finland grant #135090 and 310583, Business Finland grant 1817/31/2015, \u0026Aring;bo Akademi University, ERASMUS Mundi, \u0026Aring;bo Akademi Foundation, Svenska Kulturfonden and K. Albin Johansson. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eACKNOWLEDGEMENTS\u003c/p\u003e\n\u003cp\u003eWe thank Zebrafish Core (Turku Bioscience Centre, University of Turku and \u0026Aring;bo Akademi University) supported by Biocenter Finland for assistance.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eCOMPETING INTERESTS\u003c/p\u003e\n\u003cp\u003eThe authors have nothing to disclose.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eGiacobbe, P. \u0026amp; Flint, A. Diagnosis and Management of Anxiety Disorders. \u003cem\u003eContinuum (Minneap Minn)\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, 893-919 (2018). https://doi.org/10.1212/CON.0000000000000607\u003c/li\u003e\n\u003cli\u003ePenninx, B. W., Pine, D. S., Holmes, E. A. \u0026amp; Reif, A. Anxiety disorders. \u003cem\u003eLancet\u003c/em\u003e \u003cstrong\u003e397\u003c/strong\u003e, 914-927 (2021). https://doi.org/10.1016/S0140-6736(21)00359-7\u003c/li\u003e\n\u003cli\u003eCollaborators, G. M. D. Global, regional, and national burden of 12 mental disorders in 204 countries and territories, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. \u003cem\u003eLancet Psychiatry\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 137-150 (2022). https://doi.org/10.1016/S2215-0366(21)00395-3\u003c/li\u003e\n\u003cli\u003eCollaborators, G. D. a. I. Global burden of 369 diseases and injuries in 204 countries and territories, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. \u003cem\u003eLancet\u003c/em\u003e \u003cstrong\u003e396\u003c/strong\u003e, 1204-1222 (2020). https://doi.org/10.1016/S0140-6736(20)30925-9\u003c/li\u003e\n\u003cli\u003eNemeroff, C. B. The State of Our Understanding of the Pathophysiology and Optimal Treatment of Depression: Glass Half Full or Half Empty? \u003cem\u003eAm J Psychiatry\u003c/em\u003e \u003cstrong\u003e177\u003c/strong\u003e, 671-685 (2020). https://doi.org/10.1176/appi.ajp.2020.20060845\u003c/li\u003e\n\u003cli\u003eGriebel, G. \u0026amp; Holmes, A. 50 years of hurdles and hope in anxiolytic drug discovery. \u003cem\u003eNat Rev Drug Discov\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 667-687 (2013). https://doi.org/10.1038/nrd4075\u003c/li\u003e\n\u003cli\u003eWang, T.\u003cem\u003e et al.\u003c/em\u003e Polygenic risk for five psychiatric disorders and cross-disorder and disorder-specific neural connectivity in two independent populations. \u003cem\u003eNeuroimage Clin\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 441-449 (2017). https://doi.org/10.1016/j.nicl.2017.02.011\u003c/li\u003e\n\u003cli\u003eMeier, S. M. \u0026amp; Deckert, J. Genetics of Anxiety Disorders. \u003cem\u003eCurr Psychiatry Rep\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 16 (2019). https://doi.org/10.1007/s11920-019-1002-7\u003c/li\u003e\n\u003cli\u003eGarakani, A.\u003cem\u003e et al.\u003c/em\u003e Pharmacotherapy of Anxiety Disorders: Current and Emerging Treatment Options. \u003cem\u003eFocus (Am Psychiatr Publ)\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, 222-242 (2021). https://doi.org/10.1176/appi.focus.19203\u003c/li\u003e\n\u003cli\u003eKeller, J.\u003cem\u003e et al.\u003c/em\u003e HPA axis in major depression: cortisol, clinical symptomatology and genetic variation predict cognition. \u003cem\u003eMol Psychiatry\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 527-536 (2017). https://doi.org/10.1038/mp.2016.120\u003c/li\u003e\n\u003cli\u003ePerrin, A. J., Horowitz, M. A., Roelofs, J., Zunszain, P. A. \u0026amp; Pariante, C. M. Glucocorticoid Resistance: Is It a Requisite for Increased Cytokine Production in Depression? A Systematic Review and Meta-Analysis. \u003cem\u003eFront Psychiatry\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 423 (2019). https://doi.org/10.3389/fpsyt.2019.00423\u003c/li\u003e\n\u003cli\u003eOta, K. T.\u003cem\u003e et al.\u003c/em\u003e REDD1 is essential for stress-induced synaptic loss and depressive behavior. \u003cem\u003eNat Med\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 531-535 (2014). https://doi.org/10.1038/nm.351310.1038/nm.3513. Epub 2014 Apr 13.\u003c/li\u003e\n\u003cli\u003eVincent, F.\u003cem\u003e et al.\u003c/em\u003e Phenotypic drug discovery: recent successes, lessons learned and new directions. \u003cem\u003eNat Rev Drug Discov\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 899-914 (2022). https://doi.org/10.1038/s41573-022-00472-w\u003c/li\u003e\n\u003cli\u003eColwill, R. M. \u0026amp; Creton, R. Imaging escape and avoidance behavior in zebrafish larvae. \u003cem\u003eRev Neurosci\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 63-73 (2011). https://doi.org/10.1515/RNS.2011.008\u003c/li\u003e\n\u003cli\u003eBruni, G.\u003cem\u003e et al.\u003c/em\u003e Zebrafish behavioral profiling identifies multitarget antipsychotic-like compounds. \u003cem\u003eNat Chem Biol\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 559-566 (2016). https://doi.org/10.1038/nchembio.2097\u003c/li\u003e\n\u003cli\u003eMcCarroll, M. N.\u003cem\u003e et al.\u003c/em\u003e Zebrafish behavioural profiling identifies GABA and serotonin receptor ligands related to sedation and paradoxical excitation. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 4078 (2019). https://doi.org/10.1038/s41467-019-11936-w\u003c/li\u003e\n\u003cli\u003ede Abreu, M. S.\u003cem\u003e et al.\u003c/em\u003e Unconventional anxiety pharmacology in zebrafish: Drugs beyond traditional anxiogenic and anxiolytic spectra. \u003cem\u003ePharmacol Biochem Behav\u003c/em\u003e \u003cstrong\u003e207\u003c/strong\u003e, 173205 (2021). https://doi.org/10.1016/j.pbb.2021.173205\u003c/li\u003e\n\u003cli\u003eMaximino, C.\u003cem\u003e et al.\u003c/em\u003e Fingerprinting of psychoactive drugs in zebrafish anxiety-like behaviors. \u003cem\u003ePLoS One\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, e103943 (2014). https://doi.org/10.1371/journal.pone.0103943\u003c/li\u003e\n\u003cli\u003eHowe, K.\u003cem\u003e et al.\u003c/em\u003e The zebrafish reference genome sequence and its relationship to the human genome. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e496\u003c/strong\u003e, 498-503 (2013). https://doi.org/10.1038/nature12111\u003c/li\u003e\n\u003cli\u003eKalueff, A. V., Stewart, A. M. \u0026amp; Gerlai, R. Zebrafish as an emerging model for studying complex brain disorders. \u003cem\u003eTrends Pharmacol Sci\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 63-75 (2014). https://doi.org/10.1016/j.tips.2013.12.002\u003c/li\u003e\n\u003cli\u003eWaterston, R. H.\u003cem\u003e et al.\u003c/em\u003e Initial sequencing and comparative analysis of the mouse genome. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e420\u003c/strong\u003e, 520-562 (2002). https://doi.org/10.1038/nature01262\u003c/li\u003e\n\u003cli\u003eSchn\u0026ouml;rr, S. J., Steenbergen, P. J., Richardson, M. K. \u0026amp; Champagne, D. L. Measuring thigmotaxis in larval zebrafish. \u003cem\u003eBehav Brain Res\u003c/em\u003e \u003cstrong\u003e228\u003c/strong\u003e, 367-374 (2012). https://doi.org/10.1016/j.bbr.2011.12.016\u003c/li\u003e\n\u003cli\u003eBasnet, R. M., Zizioli, D., Taweedet, S., Finazzi, D. \u0026amp; Memo, M. Zebrafish Larvae as a Behavioral Model in Neuropharmacology. \u003cem\u003eBiomedicines\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e (2019). https://doi.org/10.3390/biomedicines7010023\u003c/li\u003e\n\u003cli\u003eMenezes, F. P., Kist, L. W., Bogo, M. R., Bonan, C. D. \u0026amp; Da Silva, R. S. Evaluation of age-dependent response to NMDA receptor antagonism in zebrafish. \u003cem\u003eZebrafish\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 137-143 (2015). https://doi.org/10.1089/zeb.2014.1018\u003c/li\u003e\n\u003cli\u003eCoffey, E. T., Hongisto, V., Dickens, M., Davis, R. J. \u0026amp; Courtney, M. J. Dual roles for c-Jun N-terminal kinase in developmental and stress responses in cerebellar granule neurons. \u003cem\u003eJ Neurosci\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 7602-7613 (2000).\u003c/li\u003e\n\u003cli\u003eChang, L., Jones, Y., Ellisman, M. H., Goldstein, L. S. \u0026amp; Karin, M. JNK1 is required for maintenance of neuronal microtubules and controls phosphorylation of microtubule-associated proteins. \u003cem\u003eDev Cell\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 521-533 (2003). https://doi.org/S1534580703000947 [pii]\u003c/li\u003e\n\u003cli\u003eKomulainen, E.\u003cem\u003e et al.\u003c/em\u003e JNK1 controls dendritic field size in L2/3 and L5 of the motor cortex, constrains soma size, and influences fine motor coordination. \u003cem\u003eFront Cell Neurosci\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 272 (2014). https://doi.org/10.3389/fncel.2014.00272\u003c/li\u003e\n\u003cli\u003eCoffey, E. T. Nuclear and cytosolic JNK signalling in neurons. \u003cem\u003eNat Rev Neurosci\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 285-299 (2014). https://doi.org/10.1038/nrn3729\u003c/li\u003e\n\u003cli\u003eMohammad, H.\u003cem\u003e et al.\u003c/em\u003e JNK1 controls adult hippocampal neurogenesis and imposes cell-autonomous control of anxiety behaviour from the neurogenic niche. \u003cem\u003eMol Psychiatry\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 362-374 (2018). https://doi.org/10.1038/mp.2016.203\u003c/li\u003e\n\u003cli\u003eHollos, P., Marchisella, F. \u0026amp; Coffey, E. T. JNK Regulation of Depression and Anxiety. \u003cem\u003eBrain Plast\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 145-155 (2018). https://doi.org/10.3233/BPL-170062\u003c/li\u003e\n\u003cli\u003eHollos, P., John, J. M., Lehtonen, J. V. \u0026amp; Coffey, E. T. Optogenetic Control of Spine-Head JNK Reveals a Role in Dendritic Spine Regression. \u003cem\u003eeNeuro\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e (2020). https://doi.org/10.1523/ENEURO.0303-19.2019\u003c/li\u003e\n\u003cli\u003eHolsboer, F. The corticosteroid receptor hypothesis of depression. \u003cem\u003eNeuropsychopharmacology\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 477-501 (2000). https://doi.org/10.1016/S0893-133X(00)00159-7\u003c/li\u003e\n\u003cli\u003eJovicic, M. J., Lukic, I., Radojcic, M., Adzic, M. \u0026amp; Maric, N. P. Modulation of c-Jun N-terminal kinase signaling and specific glucocorticoid receptor phosphorylation in the treatment of major depression. \u003cem\u003eMed Hypotheses\u003c/em\u003e \u003cstrong\u003e85\u003c/strong\u003e, 291-294 (2015). https://doi.org/10.1016/j.mehy.2015.05.015\u003c/li\u003e\n\u003cli\u003eSevilla, L. M.\u003cem\u003e et al.\u003c/em\u003e Glucocorticoid Resistance: Interference between the Glucocorticoid Receptor and the MAPK Signalling Pathways. \u003cem\u003eInt J Mol Sci\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e (2021). https://doi.org/10.3390/ijms221810049\u003c/li\u003e\n\u003cli\u003eGould, T. D. \u0026amp; Manji, H. K. Glycogen synthase kinase-3: a putative molecular target for lithium mimetic drugs. \u003cem\u003eNeuropsychopharmacology\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 1223-1237 (2005). https://doi.org/10.1038/sj.npp.1300731\u003c/li\u003e\n\u003cli\u003eHuettner, J. E. \u0026amp; Bean, B. P. Block of N-methyl-D-aspartate-activated current by the anticonvulsant MK-801: selective binding to open channels. \u003cem\u003eProc Natl Acad Sci U S A\u003c/em\u003e \u003cstrong\u003e85\u003c/strong\u003e, 1307-1311 (1988). https://doi.org/10.1073/pnas.85.4.1307\u003c/li\u003e\n\u003cli\u003eMancuso, C. E., Tanzi, M. G. \u0026amp; Gabay, M. Paradoxical reactions to benzodiazepines: literature review and treatment options. \u003cem\u003ePharmacotherapy\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, 1177-1185 (2004). https://doi.org/10.1592/phco.24.13.1177.38089\u003c/li\u003e\n\u003cli\u003eZanos, P.\u003cem\u003e et al.\u003c/em\u003e Ketamine and Ketamine Metabolite Pharmacology: Insights into Therapeutic Mechanisms. \u003cem\u003ePharmacol Rev\u003c/em\u003e \u003cstrong\u003e70\u003c/strong\u003e, 621-660 (2018). https://doi.org/10.1124/pr.117.015198\u003c/li\u003e\n\u003cli\u003eKim, J. W. \u0026amp; Monteggia, L. M. Increasing doses of ketamine curtail antidepressant responses and suppress associated synaptic signaling pathways. \u003cem\u003eBehav Brain Res\u003c/em\u003e \u003cstrong\u003e380\u003c/strong\u003e, 112378 (2020). https://doi.org/10.1016/j.bbr.2019.112378\u003c/li\u003e\n\u003cli\u003eCarlsson, A., Waters, N. \u0026amp; Carlsson, M. L. Neurotransmitter interactions in schizophrenia-therapeutic implications. \u003cem\u003eEur Arch Psychiatry Clin Neurosci\u003c/em\u003e \u003cstrong\u003e249 Suppl 4\u003c/strong\u003e, 37-43 (1999). https://doi.org/10.1007/pl00014183\u003c/li\u003e\n\u003cli\u003eCarey, R. J., Dai, H. \u0026amp; Gui, J. Effects of dizocilpine (MK-801) on motor activity and memory. \u003cem\u003ePsychopharmacology (Berl)\u003c/em\u003e \u003cstrong\u003e137\u003c/strong\u003e, 241-246 (1998). https://doi.org/10.1007/s002130050616\u003c/li\u003e\n\u003cli\u003eMcDougall, S. A.\u003cem\u003e et al.\u003c/em\u003e MK801-induced locomotor activity in preweanling and adolescent male and female rats: role of the dopamine and serotonin systems. \u003cem\u003ePsychopharmacology (Berl)\u003c/em\u003e \u003cstrong\u003e237\u003c/strong\u003e, 2469-2483 (2020). https://doi.org/10.1007/s00213-020-05547-3\u003c/li\u003e\n\u003cli\u003eWang, B.\u003cem\u003e et al.\u003c/em\u003e Isolation of high-affinity peptide antagonists of 14-3-3 proteins by phage display. \u003cem\u003eBiochemistry\u003c/em\u003e \u003cstrong\u003e38\u003c/strong\u003e, 12499-12504 (1999). https://doi.org/10.1021/bi991353h\u003c/li\u003e\n\u003cli\u003eTanaka, A. \u0026amp; Nishizaki, T. The newly synthesized linoleic acid derivative FR236924 induces a long-lasting facilitation of hippocampal neurotransmission by targeting nicotinic acetylcholine receptors. \u003cem\u003eBioorg Med Chem Lett\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 1037-1040 (2003). https://doi.org/10.1016/s0960-894x(03)00089-1\u003c/li\u003e\n\u003cli\u003eKeenan, C., Goode, N. \u0026amp; Pears, C. Isoform specificity of activators and inhibitors of protein kinase C gamma and delta. \u003cem\u003eFEBS Lett\u003c/em\u003e \u003cstrong\u003e415\u003c/strong\u003e, 101-108 (1997). https://doi.org/10.1016/s0014-5793(97)01104-6\u003c/li\u003e\n\u003cli\u003eLee, H. B.\u003cem\u003e et al.\u003c/em\u003e Novel zebrafish behavioral assay to identify modifiers of the rapid, nongenomic stress response. \u003cem\u003eGenes Brain Behav\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e, e12549 (2019). https://doi.org/10.1111/gbb.12549\u003c/li\u003e\n\u003cli\u003eBeppi, C., Straumann, D. \u0026amp; B\u0026ouml;gli, S. Y. A model-based quantification of startle reflex habituation in larval zebrafish. \u003cem\u003eSci Rep\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 846 (2021). https://doi.org/10.1038/s41598-020-79923-6\u003c/li\u003e\n\u003cli\u003eBrown, E. S., Varghese, F. P. \u0026amp; McEwen, B. S. Association of depression with medical illness: does cortisol play a role? \u003cem\u003eBiol Psychiatry\u003c/em\u003e \u003cstrong\u003e55\u003c/strong\u003e, 1-9 (2004). https://doi.org/10.1016/s0006-3223(03)00473-6\u003c/li\u003e\n\u003cli\u003eBinder, E. B.\u003cem\u003e et al.\u003c/em\u003e HPA-axis regulation at in-patient admission is associated with antidepressant therapy outcome in male but not in female depressed patients. \u003cem\u003ePsychoneuroendocrinology\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, 99-109 (2009). https://doi.org/10.1016/j.psyneuen.2008.08.018\u003c/li\u003e\n\u003cli\u003eShin, L. M. \u0026amp; Liberzon, I. The neurocircuitry of fear, stress, and anxiety disorders. \u003cem\u003eNeuropsychopharmacology\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 169-191 (2010). https://doi.org/10.1038/npp.2009.83\u003c/li\u003e\n\u003cli\u003ePadilla-Coreano, N.\u003cem\u003e et al.\u003c/em\u003e Direct Ventral Hippocampal-Prefrontal Input Is Required for Anxiety-Related Neural Activity and Behavior. \u003cem\u003eNeuron\u003c/em\u003e \u003cstrong\u003e89\u003c/strong\u003e, 857-866 (2016). https://doi.org/10.1016/j.neuron.2016.01.011\u003c/li\u003e\n\u003cli\u003eQi, A. Q., Qiu, J., Xiao, L. \u0026amp; Chen, Y. Z. Rapid activation of JNK and p38 by glucocorticoids in primary cultured hippocampal cells. \u003cem\u003eJ Neurosci Res\u003c/em\u003e \u003cstrong\u003e80\u003c/strong\u003e, 510-517 (2005). https://doi.org/10.1002/jnr.20491\u003c/li\u003e\n\u003cli\u003eMitra, R. \u0026amp; Sapolsky, R. M. Acute corticosterone treatment is sufficient to induce anxiety and amygdaloid dendritic hypertrophy. \u003cem\u003eProc Natl Acad Sci U S A\u003c/em\u003e \u003cstrong\u003e105\u003c/strong\u003e, 5573-5578 (2008). https://doi.org/10.1073/pnas.0705615105 10.1073/pnas.0705615105. Epub 2008 Apr 7.\u003c/li\u003e\n\u003cli\u003eZhang, J.\u003cem\u003e et al.\u003c/em\u003e Inhibition of JNK ameliorates depressive-like behaviors and reduces the activation of pro-inflammatory cytokines and the phosphorylation of glucocorticoid receptors at serine 246 induced by neuroinflammation. \u003cem\u003ePsychoneuroendocrinology\u003c/em\u003e \u003cstrong\u003e113\u003c/strong\u003e, 104580 (2020). https://doi.org/10.1016/j.psyneuen.2019.104580\u003c/li\u003e\n\u003cli\u003eMohammad, H.\u003cem\u003e et al.\u003c/em\u003e JNK1 controls adult hippocampal neurogenesis and imposes cell-autonomous control of anxiety behaviour from the neurogenic niche. \u003cem\u003eMol Psychiatry\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 487 (2018). https://doi.org/10.1038/mp.2017.21\u003c/li\u003e\n\u003cli\u003eZhao, H. B.\u003cem\u003e et al.\u003c/em\u003e Xiao Yao San Improves the Anxiety-Like Behaviors of Rats Induced by Chronic Immobilization Stress: The Involvement of the JNK Signaling Pathway in the Hippocampus. \u003cem\u003eBiol Pharm Bull\u003c/em\u003e \u003cstrong\u003e40\u003c/strong\u003e, 187-194 (2017). https://doi.org/10.1248/bpb.b16-00694\u003c/li\u003e\n\u003cli\u003eStefanoska, K.\u003cem\u003e et al.\u003c/em\u003e Neuronal MAP kinase p38\u0026alpha; inhibits c-Jun N-terminal kinase to modulate anxiety-related behaviour. \u003cem\u003eSci Rep\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 14296 (2018). https://doi.org/10.1038/s41598-018-32592-y\u003c/li\u003e\n\u003cli\u003eZhou, X.\u003cem\u003e et al.\u003c/em\u003e Stress-Activated Protein Kinase JNK Modulates Depression-like Behaviors in Mice. \u003cem\u003eMol Neurobiol\u003c/em\u003e \u003cstrong\u003e60\u003c/strong\u003e, 2367-2378 (2023). https://doi.org/10.1007/s12035-023-03209-x\u003c/li\u003e\n\u003cli\u003eOpenshaw, R. L., Pratt, J. A. \u0026amp; Morris, B. J. The schizophrenia risk gene Map2k7 regulates responding in a novel contingency-shifting rodent touchscreen gambling task. \u003cem\u003eDis Model Mech\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e (2022). https://doi.org/10.1242/dmm.049310\u003c/li\u003e\n\u003cli\u003eSantos-Ledo, A.\u003cem\u003e et al.\u003c/em\u003e Alternative splicing of jnk1a in zebrafish determines first heart field ventricular cardiomyocyte numbers through modulation of hand2 expression. \u003cem\u003ePLoS Genet\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, e1008782 (2020). https://doi.org/10.1371/journal.pgen.1008782\u003c/li\u003e\n\u003cli\u003eFu, H., Subramanian, R. R. \u0026amp; Masters, S. C. 14-3-3 proteins: structure, function, and regulation. \u003cem\u003eAnnu Rev Pharmacol Toxicol\u003c/em\u003e \u003cstrong\u003e40\u003c/strong\u003e, 617-647 (2000). https://doi.org/10.1146/annurev.pharmtox.40.1.617\u003c/li\u003e\n\u003cli\u003eTsuruta, F.\u003cem\u003e et al.\u003c/em\u003e JNK promotes Bax translocation to mitochondria through phosphorylation of 14-3-3 proteins. \u003cem\u003eEMBO J\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 1889-1899 (2004). https://doi.org/10.1038/sj.emboj.7600194\u003c/li\u003e\n\u003cli\u003ePetosa, C.\u003cem\u003e et al.\u003c/em\u003e 14-3-3zeta binds a phosphorylated Raf peptide and an unphosphorylated peptide via its conserved amphipathic groove. \u003cem\u003eJ Biol Chem\u003c/em\u003e \u003cstrong\u003e273\u003c/strong\u003e, 16305-16310 (1998). https://doi.org/10.1074/jbc.273.26.16305\u003c/li\u003e\n\u003cli\u003eBesser, J.\u003cem\u003e et al.\u003c/em\u003e Expression analysis of the family of 14-3-3 proteins in zebrafish development. \u003cem\u003eGene Expr Patterns\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 511-520 (2007). https://doi.org/10.1016/j.modgep.2006.10.007\u003c/li\u003e\n\u003cli\u003eZhao, Y.\u003cem\u003e et al.\u003c/em\u003e Identification of 14-3-3 epsilon as a regulator of the neural apoptotic pathway for chronic-stress-induced depression. \u003cem\u003eiScience\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, 102043 (2021). https://doi.org/10.1016/j.isci.2021.102043\u003c/li\u003e\n\u003cli\u003eAnt\u0026oacute;n-Galindo, E.\u003cem\u003e et al.\u003c/em\u003e Deficiency of the ywhaz gene, involved in neurodevelopmental disorders, alters brain activity and behaviour in zebrafish. \u003cem\u003eMol Psychiatry\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 3739-3748 (2022). https://doi.org/10.1038/s41380-022-01577-9\u003c/li\u003e\n\u003cli\u003eCasta\u0026ntilde;eda, A.\u003cem\u003e et al.\u003c/em\u003e pVHL suppresses Akt/\u0026beta;-catenin-mediated cell proliferation by inhibiting 14-3-3\u0026zeta; expression. \u003cem\u003eBiochem J\u003c/em\u003e \u003cstrong\u003e474\u003c/strong\u003e, 2679-2689 (2017). https://doi.org/10.1042/BCJ20161097\u003c/li\u003e\n\u003cli\u003eG\u0026oacute;mez-Su\u0026aacute;rez, M.\u003cem\u003e et al.\u003c/em\u003e 14-3-3 Proteins regulate Akt Thr308 phosphorylation in intestinal epithelial cells. \u003cem\u003eCell Death Differ\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 1060-1072 (2016). https://doi.org/10.1038/cdd.2015.163\u003c/li\u003e\n\u003cli\u003eVaridaki, A., Mohammad, H., Coffey, E.T. \u003cem\u003eMolecular mechanisms of depression\u003c/em\u003e. 1st edn, (Elsevier, 2016).\u003c/li\u003e\n\u003cli\u003eCriado-Marrero, M.\u003cem\u003e et al.\u003c/em\u003e Early Life Stress and High FKBP5 Interact to Increase Anxiety-Like Symptoms through Altered AKT Signaling in the Dorsal Hippocampus. \u003cem\u003eInt J Mol Sci\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e (2019). https://doi.org/10.3390/ijms20112738\u003c/li\u003e\n\u003cli\u003eKing, M. K.\u003cem\u003e et al.\u003c/em\u003e Glycogen synthase kinase-3 inhibitors: Rescuers of cognitive impairments. \u003cem\u003ePharmacol Ther\u003c/em\u003e \u003cstrong\u003e141\u003c/strong\u003e, 1-12 (2014). https://doi.org/10.1016/j.pharmthera.2013.07.010\u003c/li\u003e\n\u003cli\u003eGunderwala, A., Cope, N. \u0026amp; Wang, Z. Mechanism and inhibition of BRAF kinase. \u003cem\u003eCurr Opin Chem Biol\u003c/em\u003e \u003cstrong\u003e71\u003c/strong\u003e, 102205 (2022). https://doi.org/10.1016/j.cbpa.2022.102205\u003c/li\u003e\n\u003cli\u003eToth, M., Gresack, J. E., Hauger, R. L., Halberstadt, A. L. \u0026amp; Risbrough, V. B. The role of PKC signaling in CRF-induced modulation of startle. \u003cem\u003ePsychopharmacology (Berl)\u003c/em\u003e \u003cstrong\u003e229\u003c/strong\u003e, 579-589 (2013). https://doi.org/10.1007/s00213-013-3114-9\u003c/li\u003e\n\u003cli\u003eBowers, B. J., Collins, A. C., Tritto, T. \u0026amp; Wehner, J. M. Mice lacking PKC gamma exhibit decreased anxiety. \u003cem\u003eBehav Genet\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 111-121 (2000). https://doi.org/10.1023/a:1001951104208\u003c/li\u003e\n\u003cli\u003eSun, M. K. \u0026amp; Alkon, D. L. Dual effects of bryostatin-1 on spatial memory and depression. \u003cem\u003eEur J Pharmacol\u003c/em\u003e \u003cstrong\u003e512\u003c/strong\u003e, 43-51 (2005). https://doi.org/10.1016/j.ejphar.2005.02.028\u003c/li\u003e\n\u003cli\u003eMalhi, G. S. \u0026amp; Mann, J. J. Depression. \u003cem\u003eLancet\u003c/em\u003e \u003cstrong\u003e392\u003c/strong\u003e, 2299-2312 (2018). https://doi.org/10.1016/S0140-6736(18)31948-2\u003c/li\u003e\n\u003cli\u003eAlvarez, J. C.\u003cem\u003e et al.\u003c/em\u003e Plasma serotonin level after 1 day of fluoxetine treatment: a biological predictor for antidepressant response? \u003cem\u003ePsychopharmacology (Berl)\u003c/em\u003e \u003cstrong\u003e143\u003c/strong\u003e, 97-101 (1999). https://doi.org/10.1007/s002130050924\u003c/li\u003e\n\u003cli\u003eCasarotto, P. C.\u003cem\u003e et al.\u003c/em\u003e Antidepressant drugs act by directly binding to TRKB neurotrophin receptors. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e184\u003c/strong\u003e, 1299-1313.e1219 (2021). https://doi.org/10.1016/j.cell.2021.01.034\u003c/li\u003e\n\u003cli\u003eMartin, S. C., Marazzi, G., Sandell, J. H. \u0026amp; Heinrich, G. Five Trk receptors in the zebrafish. \u003cem\u003eDev Biol\u003c/em\u003e \u003cstrong\u003e169\u003c/strong\u003e, 745-758 (1995). https://doi.org/10.1006/dbio.1995.1184\u003c/li\u003e\n\u003cli\u003eStrekalova, T.\u003cem\u003e et al.\u003c/em\u003e Chronic mild stress paradigm as a rat model of depression: facts, artifacts, and future perspectives. \u003cem\u003ePsychopharmacology (Berl)\u003c/em\u003e \u003cstrong\u003e239\u003c/strong\u003e, 663-693 (2022). https://doi.org/10.1007/s00213-021-05982-w\u003c/li\u003e\n\u003cli\u003eTaschereau-Dumouchel, V., Michel, M., Lau, H., Hofmann, S. G. \u0026amp; LeDoux, J. E. Putting the \u0026quot;mental\u0026quot; back in \u0026quot;mental disorders\u0026quot;: a perspective from research on fear and anxiety. \u003cem\u003eMol Psychiatry\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 1322-1330 (2022). https://doi.org/10.1038/s41380-021-01395-5\u003c/li\u003e\n\u003cli\u003eOrtega, V. A., Mercer, E. M., Giesbrecht, G. F. \u0026amp; Arrieta, M. C. Evolutionary Significance of the Neuroendocrine Stress Axis on Vertebrate Immunity and the Influence of the Microbiome on Early-Life Stress Regulation and Health Outcomes. \u003cem\u003eFront Microbiol\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 634539 (2021). https://doi.org/10.3389/fmicb.2021.634539\u003c/li\u003e\n\u003cli\u003ePan, J. Q.\u003cem\u003e et al.\u003c/em\u003e AKT kinase activity is required for lithium to modulate mood-related behaviors in mice. \u003cem\u003eNeuropsychopharmacology\u003c/em\u003e \u003cstrong\u003e36\u003c/strong\u003e, 1397-1411 (2011). https://doi.org/10.1038/npp.2011.24\u003c/li\u003e\n\u003cli\u003eTafet, G. E. \u0026amp; Nemeroff, C. B. Pharmacological Treatment of Anxiety Disorders: The Role of the HPA Axis. \u003cem\u003eFront Psychiatry\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 443 (2020). https://doi.org/10.3389/fpsyt.2020.00443\u003c/li\u003e\n\u003cli\u003eSolas, M., Gerenu, G., Gil-Bea, F. J. \u0026amp; Ram\u0026iacute;rez, M. J. Mineralocorticoid receptor activation induces insulin resistance through c-Jun N-terminal kinases in response to chronic corticosterone: cognitive implications. \u003cem\u003eJ Neuroendocrinol\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, 350-356 (2013). https://doi.org/10.1111/jne.12006\u003c/li\u003e\n\u003cli\u003eFitzgerald, J. A., Kirla, K. T., Zinner, C. P. \u0026amp; Vom Berg, C. M. Emergence of consistent intra-individual locomotor patterns during zebrafish development. \u003cem\u003eSci Rep\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 13647 (2019). https://doi.org/10.1038/s41598-019-49614-y\u003c/li\u003e\n\u003cli\u003eDong, C.\u003cem\u003e et al.\u003c/em\u003e Defective T cell differentiation in the absence of Jnk1. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e282\u003c/strong\u003e, 2092-2095 (1998).\u003c/li\u003e\n\u003cli\u003eH, W., (Springer-Verlag New York, 2009).\u003c/li\u003e\n\u003cli\u003eA language and environment for statistical computing. (R Foundation for Statistical Computing, Vienna, Austria., 2013).\u003c/li\u003e\n\u003cli\u003eDel Carratore, F.\u003cem\u003e et al.\u003c/em\u003e RankProd 2.0: a refactored bioconductor package for detecting differentially expressed features in molecular profiling datasets. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e33\u003c/strong\u003e, 2774-2775 (2017). https://doi.org/10.1093/bioinformatics/btx292\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Table","content":"\u003cp\u003eTable 1 is available in the Supplementary Files section.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-3830013/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3830013/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eCurrent treatments for anxiety and depression show limited efficacy in many patients indicating that research into new underlying mechanisms is needed. Inhibition of JNK1 has been shown to evoke an anxiolytic-and antidepressant-like phenotype in mice however the downstream effectors that elicit these behavioural effects are unknown. Here we employ a zebrafish (\u003cem\u003eD. Rerio\u003c/em\u003e) larvae behavioural assay to identify an antidepressant-/anxiolytic-like phenotype based on 2759 measured stereotypic responses to clinically proven antidepressant and anxiolytic (AA) drugs. Employing machine learning, we classify an AA phenotype from behavioural features measured during and after a startle battery in fish exposed to AA drugs (fluoxetine, imipramine, diazepam, lithium chloride, ketamine). We demonstrate that structurally independent JNK inhibitors replicate the AA classification with high accuracy, consistent with findings in mice. We go on to identify signalling hubs downstream from JNK1 by comparing phosphoproteome data from wildtype and \u003cem\u003eJnk1-/-\u003c/em\u003e mouse brains, and test these hubs as possible mediators of the AA phenotype in zebrafish larvae. Among these, we find that AKT, GSK-3, 14-3-3ζ/ε and PKCε, when pharmacologically targeted, phenocopy clinically proven AA drugs. This assay shows promise as an early phase screening for compounds with anti-stress-axis/anxiolytic-like properties, and for mode of action analysis.\u003c/p\u003e","manuscriptTitle":"Jnk1 and Downstream Signalling Hubs Regulate Anxiety-like Behaviours in a Zebrafish Larvae Phenotypic Screen","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-01-09 19:45:18","doi":"10.21203/rs.3.rs-3830013/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-02-26T15:35:02+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-02-12T13:34:28+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"86848c7f-3267-4f10-9d9f-ec760816c823","date":"2024-01-27T10:27:24+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-01-24T21:23:17+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"29b266c8-8189-44fa-b7da-ca2bbe5a12bb","date":"2024-01-18T14:44:50+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-01-08T05:36:44+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-01-08T04:45:49+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2024-01-07T17:30:32+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-01-07T17:29:08+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2024-01-02T17:07:25+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"f0121d7e-70ad-4e81-93bf-e3d72bc280c2","owner":[],"postedDate":"January 9th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":28030638,"name":"Biological sciences/Biochemistry"},{"id":28030639,"name":"Biological sciences/Biological techniques"},{"id":28030640,"name":"Biological sciences/Cell biology"},{"id":28030641,"name":"Biological sciences/Drug discovery"},{"id":28030642,"name":"Biological sciences/Neuroscience"},{"id":28030643,"name":"Biological sciences/Systems biology"},{"id":28030644,"name":"Health sciences/Diseases"},{"id":28030645,"name":"Health sciences/Medical research"}],"tags":[],"updatedAt":"2024-05-16T00:38:38+00:00","versionOfRecord":{"articleIdentity":"rs-3830013","link":"https://doi.org/10.1038/s41598-024-61337-3","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2024-05-15 00:38:38","publishedOnDateReadable":"May 15th, 2024"},"versionCreatedAt":"2024-01-09 19:45:18","video":"","vorDoi":"10.1038/s41598-024-61337-3","vorDoiUrl":"https://doi.org/10.1038/s41598-024-61337-3","workflowStages":[]},"version":"v1","identity":"rs-3830013","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3830013","identity":"rs-3830013","version":["v1"]},"buildId":"_2-kVJe1T_tPrBINL-cwx","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}