Molecular Psychiatry (2012) 17, 946 -- 954 & 2012 Macmillan Publishers Limited All rights reserved 1359-4184/12 www.nature.com/mp

ORIGINAL ARTICLE The ADHD-susceptibility gene lphn3.1 modulates dopaminergic neuron formation and locomotor activity during zebrafish development

M Lange1, W Norton1, M Coolen1, M Chaminade1, S Merker2, F Proft2, A Schmitt2, P Vernier3, K-P Lesch2 and L Bally-Cuif1

Attention-deficit/hyperactivity disorder (ADHD) is a neurodevelopmental disorder characterized by inattention, hyperactivity, increased impulsivity and emotion dysregulation. Linkage analysis followed by fine-mapping identified variation in the gene coding for Latrophilin 3 (LPHN3), a putative adhesion-G protein-coupled receptor, as a risk factor for ADHD. In order to validate the link between LPHN3 and ADHD, and to understand the function of LPHN3 in the etiology of the disease, we examined its ortholog lphn3.1 during zebrafish development. Loss of lphn3.1 function causes a reduction and misplacement of dopamine- positive neurons in the ventral diencephalon and a hyperactive/impulsive motor phenotype. The behavioral phenotype can be rescued by the ADHD treatment drugs methylphenidate and atomoxetine. Together, our results implicate decreased Lphn3 activity in eliciting ADHD-like behavior, and demonstrate its correlated contribution to the development of the brain dopaminergic circuitry.

Molecular Psychiatry (2012) 17, 946--954; doi:10.1038/mp.2012.29; published online 17 April 2012 Keywords: ADHD; dopamine; hyperactivity; impulsivity; Lphn3; zebrafish

INTRODUCTION European populations. Replication in a Spanish cohort of patients Attention-deficit/hyperactivity disorder (ADHD; OMIM no. 143465) with adult ADHD suggested a role for LPHN3 in the persistence of 22 is a clinically heterogeneous neurodevelopmental disorder that is the disease throughout life. Furthermore, LPHN3 was also characterized by inattention, hyperactivity and increased impulsiv- identified as one among 86 risk genes in patients with substance 23 ity/emotionality. ADHD represents the most common behavioral use disorders, a frequent comorbidity of ADHD. LPHN3 is a disorder in childhood and adolescence with prevalence rates of putative adhesion-G protein-coupled receptor, but its physiologi- 3--5% across different cultural settings.1--3 Although the substantial cal role is not well understood and an endogenous ligand has yet 24 heritability of ADHD is documented in numerous family, twin and to be identified. In recent years, the zebrafish has been adoption studies, with estimates of up to 80%, both genome-wide established as a valid animal model for probing the genetic and 25 -- 28 approaches and candidate gene studies have so far failed to reliably developmental bases of behavior. In this study, we have used identify and characterize ADHD-associated risk genes.4--8 larval zebrafish to assess the developmental and behavioral Insights from both pharmacological studies and animal models function of Lphn3.1 activity, one of the zebrafish orthologs of have consistently implicated altered neurotransmission in the LPHN3. We show that decreased Lphn3.1 activity selectively affects neurobiology of ADHD.9--14 The effectiveness of treating ADHD DA system development and triggers hyperactive/impulsive with both stimulants (for example, methylphenidate (MPH)) and locomotion. This behavioral phenotype can be rescued by MPH noradrenaline (NA) reuptake inhibitors (for example, atomoxetine and ATO, drugs that are effective in treating ADHD. These results (ATO)), that interact with the dopaminergic (DA) and NA systems, conclusively implicate LPHN3 as one of the factors involved in the and the ability of antidepressants (that target the 5-HT system) to locomotor phenotype of ADHD and highlight its critical role in DA moderate the ADHD-related emotion dysregulation have focused neuron development. research onto monoaminergic signaling.15 -- 19 A critical role of DA in the etiology of ADHD is also supported by genetic association studies.20 However, the changes during neural development that lead to the symptoms of ADHD are not well understood. MATERIALS AND METHODS Recently, Arcos-Burgos et al.21 reported evidence of a risk In situ hybridization and histochemical methods haplotype in the gene coding for Latrophilin 3 (LPHN3), by linkage In situ hybridization and antibody labeling were performed on whole- scan of a South American genetic isolate and subsequent fine- mount preparations (whole embryos or dissected brains) and sectioned mapping of the 4q13.2 locus in several North American and adult brains according to standard procedures.29 -- 31

1Zebrafish Neurogenetics Group, Laboratory of Neurobiology and Development (N&D), CNRS UPR 3294, Institute of Neurobiology Alfred Fessard, Avenue de la Terrasse, Gif-sur- Yvette ce´dex, France; 2Department of Psychiatry, Psychosomatics and Psychotherapy, Division of Molecular Psychiatry, ADHD Clinical Research Network, Laboratory of Translational Neuroscience, University of Wu¨rzburg, Fu¨chsleinstrasse 15, Wu¨rzburg, Germany and 3Research group ‘Development and Evolution of Neurotransmission’, Laboratory of Neurobiology and Development (N&D), CNRS UPR 3294, Institute of Neurobiology Alfred Fessard, Avenue de la Terrasse, Gif-sur-Yvette ce´dex, France. Correspondence: Dr L Bally-Cuif or Dr W Norton, Zebrafish Neurogenetics Group, Laboratory of Neurobiology and Development (N&D), CNRS UPR 3294, Institute of Neurobiology Alfred Fessard, Avenue de la Terrasse, 91198 Gif-sur-Yvette ce´dex, France. E-mail: [email protected] or [email protected] Received 21 July 2011; revised 31 January 2012; accepted 13 February 2012; published online 17 April 2012 Lphn3.1 controls locomotor activity and impulsivity M Lange et al 947 Morpholino injection of differentiated neurons (Supplementary Figure S2A, D), such as 32 Lphn3 splice morpholino oligonucleotides (MO) were purchased from those labeled by elavl3 expression. This pattern is maintained at GeneTools LLC. Lphn3-MO1 was designed to bind the splice-donor site of 48 h.p.f. (Supplementary Figure S2B, E), whereas by 6 days post exon 2, while Lphn3-MO2 blocks splicing at the exon6/intron6 boundary fertilization (d.p.f.) the expression of lphn3.1 becomes more (Supplementary Figure S4H). A control MO (Lphn3-CO) with 5 bp restricted, with strong expression in the ventral part of the mismatches was also used. Embryos were injected with 250 mM MO telencephalon and diencephalon and in the hindbrain as well as in (5.9 ng) at the single-cell stage and allowed to recover for 6 days with a the ventral spinal area (Supplementary Figure S2C, F). At 24 h.p.f., regular day--night cycle at 28 1C before being analyzed for behavior (below). lphn3.2 expression is seen in the ventral diencephalon (Supple- To confirm splicing defects following MO injection, 20 embryos were snap mentary Figure S2J, M). It expands at 48 h.p.f. to also include the frozen in liquid nitrogen, and RNA was extracted using the TRIzol reagent telencephalon and hindbrain (in stripes of expression that appear (Invitrogen, Carlsbad, CA, USA). We then used RT-PCR (carried out using the to be complementary to those of lphn3.1), as well as at the edge Superscriptase II kit, Invitrogen) to visualize aberrant lphn3.1 splicing in of the optic tectum and the otic vesicle (Supplementary Figure morphants, and real-time PCR (see Supplementary Methods) to quantify S2K, N). By 6 d.p.f., expression of lphn3.2 appears very similar to that lphn3.1 transcripts. See Supplementary Methods for sequences of MOs and of lphn3.1, and is seen in the ventral part of the telencephalon, the primers. diencephalon and the hindbrain (Supplementary Figure S2L, O). At 30 h.p.f. lphn3.1 is coexpressed with tyrosine hydroxylase Analysis of locomotor behavior in Lphn3 morphant larvae (a marker for DA neurons) in the area of the posterior tuberculum (PT)(Supplementary Figure S2X). In order to assess the conservation Locomotor activity was examined at 6 days post fertilization (d.p.f.) by of Lphn3 expression between species, we compared zebrafish recording the total distance swum over a 5-min period measured using the lphn3.1 and mouse Lphn3 in the adult brain. Expression of lphn3.1 ZebraLab software (Videotrack; ViewPoint Life Sciences, Lyon, France). is seen along the telencephalic midline as well as at lower levels in Injected larvae were placed into separate wells of a 12-well plate (BD) the telencephalic parenchyma, the anterior thalamus, periaque- inside a Zebrabox (ViewPoint Life Sciences) and were allowed to habituate ductal gray matter, the superior raphe´ nucleus, periventricular for 5 min before recording. The distance swum was calculated by tracking nucleus of the inferior hypothalamus, the cerebellum and the the larvae for 5 min using a high-speed infrared camera with the ZebraLab nucleus of the medial longitudinal fasciculus (Supplementary software set to a detection threshold of 11, inactive/small threshold of 5 Figure S3A--E). This wide-spread expression profile is shared by and small/large threshold of 10. To compare motor activities, the mean Lphn3, the murine homolog of lphn3.1 (Supplementary Figure total distance swum was calculated for control and morphant larvae. For S3F-- I). In the adult mouse brain, Lphn3 is expressed in all layers of measurements during the night, animals that had already been measured the cortex as well as in the dorsal and central raphe´ nuclei, the during the day were left in the 12-well plate and were recorded for 5 min hippocampus and cerebellum. 3 h after the light was switched off in the Zebrabox. The data obtained was This broad developmental expression provided little insight into exported to Excel (Microsoft) for analysis. To compare the distance swum Lphn3 function. We therefore chose to disrupt the function of between wild-type and morphant larvae, p-values were calculated using a lphn3.1, the LPHN3 homolog with strongest embryonic expression. one-way ANOVA followed by paired t-tests with unpooled SD and the lphn3.1 consists of 19 exons (Supplementary Figure S4B) and p-value adjusted according to HOLM. All data was analyzed using Excel codes for a protein with a similar predicted structure as human (Microsoft) and the R Foundation statistical software package. LPHN3 (Supplementary Figure S4A). We used two splice-blocking MO to knockdown lphn3.1 activity during development. Lphn3- Drug treatment MO1 targeted the exon2/intron2 boundary, and Lphn3-MO2 the All drug solutions were freshly prepared in embryo medium (containing exon6/intron6 boundary, regions where lphn3.1 and lphn3.2 0.01% dimethyl sulfoxide) on the day of the experiment. Before drug sequences differed significantly (Supplementary Figure S4B, H). treatment, larvae were injected with either Lphn3-CO or Lphn3-MO1 and Injection of either of these MOs resulted in the absence of wild- were allowed to recover for 6 days. Morphant larvae were then incubated type lphn3.1 transcripts in 2 d.p.f. embryos (Supplementary Figure in either a drug solution (1, 5, 10, 15 or 20 mM ATO (Sigma-Aldrich, St Louis, S4D, E, G). At later stages, (6 d.p.f.) the expression of lphn3.1 MO, USA) or 8, 10, 12, 15 or 20 mM MPH (Sigma-Aldrich)) or in 0.01% recovered (Supplementary Figure S4F, G), indicating that mor- dimethyl sulfoxide alone for a 1-h period. The distance swum during a pholino injection causes a transient downregulation of Lphn3.1 5-min time period was recorded for the different treatment groups as activity during development. The knockdown of lphn3.1 by described above. For real-time PCR analysis of the effect of MPH and ATO morpholino injection did not cause a developmental delay treatment on lphn3.1 expression levels, larvae were treated with either (Supplementary Figure S4C) nor an increase in apoptosis 10 mM MPH or 1 mM ATO for 1 h before RNA preparation. To observe the (Supplementary Figure S5A--D) compared with control animals. effect of MPH or ATO treatment on DA neuron formation, larvae were treated with either 10 mM MPH or 1 mM ATO from day 0 (immediately after fertilization), on day 3 or on day 6 (for 1 h) before being fixed at 6 d.p.f. Downregulation of Lphn3 function triggers locomotor hyperactivity and swimming bursts We first investigated the consequences of decreasing Lphn3 RESULTS activity on larval zebrafish behavior. Among the hallmarks of Isolation and knockdown of zebrafish LPHN3 orthologs ADHD-typical syndromal dimensions, we focused on motor In order to try and elucidate the endogenous function of LPHN3, activity, which can be most easily evaluated in this model system. we cloned and examined the corresponding gene in zebrafish. We We injected Lphn3-MO1 or Lphn3-MO2 into wild-type eggs and used BLAST analysis of the zebrafish genome to identify two partial measured locomotion 6 days later. We observed a significant transcripts with similarity to human LPHN3 and confirmed their increase in the distance swum by lphn3.1 morphants compared identity by phylogenetic analysis of Latrophilin family members with control embryos injected with a mismatch morpholino from several species (Supplementary Figure S1). Zebrafish Lphn3.1 (Lphn3-CO) (Figure 1a). Both MO1- and MO2-injected larvae dis- has 73.7% identity and 86.2% similarity to the duplicated Lphn3.2 played a similar increase in locomotion in multiple independent protein. Lphn3.1 has 73.3% identity and 85.7% similarity to human experiments (Supplementary Figure S6A and data not shown), LPHN3 and Lphn3.2 has 76.1% identity and 89.3% similarity to confirming the specificity of the phenotype. We consequently human LPHN3. Both LPHN3 orthologs show partially overlapping restricted our analysis to Lphn3-MO1 morphant larvae. expression profiles during development. At 24 h post fertilization The increased distance swum by lphn3.1 morphants (Figure 1a, (h.p.f.), lphn3.1 is broadly expressed in a pattern reminiscent of that Supplementary Figure S6A, B and Supplementary Video) could be

& 2012 Macmillan Publishers Limited Molecular Psychiatry (2012), 946 -- 954 Lphn3.1 controls locomotor activity and impulsivity M Lange et al 948

Figure 1. Downregulation of lphn3.1 induces hyperactivity in zebrafish larvae. (a) Mean distance swum in a 5-min time interval by 6 d.p.f. larvae injected with Lphn3-CO, Lphn3-MO1 or Lphn3-MO2. MO1- and MO2-injected larvae swam equally fast and significantly further than CO- injected animals. Error bars are ±s.e.m. Lphn3-CO n ¼ 43, Lphn3-MO1 n ¼ 42, Lphn3-MO2 n ¼ 49. **po0.025, NS ¼ non significant. (b) Average swimming speed of the Lphn3-CO- and Lphn3-MO1-injected populations during a 5-min experiment, excluding the time spent resting. For Lphn3-CO, n ¼ 47 animals analyzed; for Lphn3-MO1, n ¼ 48. Lphn3-CO-injected animals swim on average 3.93 mm s -- 1, whereas Lphn3-MO1 morphants swim faster, at an average speed of 6.44 mm s -- 1. Error bars are ±s.e.m., *po0.05. (c) Sum over time of the distance swum (in millimeters) by Lphn3-CO- and Lphn3-MO1-injected larvae. The total distance traveled is added every 3 s during a 120-second experiment; n ¼ 10 fish for each treatment group; grey lines represent Lphn3-MO1 whereas black lines represent Lphn3-CO larvae. Lphn3-MO1 morphant fish are constantly and stably hyperactive. Linear mixed t-test po0.025. (d) Mean distance swum during 5 min by 6 d.p.f. Lphn3-CO (n ¼ 12) and Lphn3-MO1 (n ¼ 12) injected larvae during day-time (*po0.05) and night-time conditions (**po0.025). For each treatment group the same animals are analyzed at both time points. Swimming activity during the night is decreased compared with day-time (***po0.001), but night-measured Lphn3-MO1 larvae are still significantly more active than night-measured Lphn3-CO controls (**po0.025).

caused by a decrease in the amount of time resting between at night.33 -- 35 To determine whether the activity of lphn3.1 swimming bouts, and/or an overall increase in swimming speed. morphants displayed comparable properties, we first plotted the We found no significant difference in the amount of time spent total distance swum by individual fish every 3 s during a 120-s resting between Lphn3-CO- and Lphn3-MO1-injected larvae experiment (Figure 1c). We represented the single curves obtained during a 5-min experiment (p ¼ 0.5 Supplementary Figure S6C). by the linear equation y ¼ Sx, with a slope S of 2.8 mm s -- 1 in Conversely, when excluding the resting periods, Lphn3-CO- Lphn3-CO- against 6.2 mm s -- 1 in Lphn3-MO1 animals. These injected animals swam in average at 3.93 mm s -- 1, against values are significantly different from each other (Linear mixed 6.44 mm s -- 1 for Lphn3-M01-injected larvae (Lphn3-CO vs Lphn3- t-test po0.025) and close to the mean speed of control and MO1 p ¼ 0.04; Figure 1b). Thus, the hyperactivity of lphn3.1 morphant fish (Figure 1b). Together, these results indicate that morphants reflected an increase in the overall speed of Lphn3.1 morphant fish display regular and stable hyperactivity. locomotion during periods of activity. Next, to determine whether the hyperactivity of Lphn3-MO1 The symptoms of hyperactivity manifested by ADHD patients morphants extended into the sleeping period, we recorded can be remarkably stable over time and may also be maintained the distance swum at night. As expected, both Lphn3-CO and

Molecular Psychiatry (2012), 946 -- 954 & 2012 Macmillan Publishers Limited Lphn3.1 controls locomotor activity and impulsivity M Lange et al 949 Lphn3-MO1 populations clearly swam less during the night. their behavioral phenotype (Supplementary Figure S8A). Group 1 However, Lphn3-MO1 morphant larvae still displayed marked comprised Lphn3-MO1-injected larvae that swam further than any hyperactivity compared with control injected animals (Figure 1d). control larva. Larvae in groups 2 and 3 swam more or less than the ADHD-associated increases in impulsivity, subdivided into both mean distance measured for Lphn3-CO larvae, respectively, within motor and cognitive components,36 is seen as a fundamental trait the range of activity levels displayed by control larvae. We then underlying various dimensions of the behavioral impairment performed anti-tyrosine hydroxylase antibody labeling and observed in patients.12 When plotting the distance swum every 3 s observed a gradient in the modification of PT DA neurons. during 90 s for individual animals, the curves for morphant fish Group 1 (severely hyperactive Lphn3-MO1) larvae had a strong displayed sharper peaks than those of Lphn3-CO larvae (Figure 2a, reduction and misplacement of PT DA neurons (Supplementary grey arrow). The individual curves for Lphn3-MO1 fish exhibit an Figure S8C). Group 2 Lphn3-MO1 larvae had a mild reduction of PT average of 6.2 peaks compared with 2.7 peaks per individual neurons compared with group 2 Lphn3-CO larvae (Supplementary Lphn3-CO fish (Lphn3-CO vs Lphn3-MO1 po0.001; Figure 2b). Figure S8D, E), whereas the PT of group 3 Lphn3-MO1 animals was Furthermore, the distance swum per peak was increased in largely similar to that of group 3 Lphn-CO animals (Supplementary morphants (Figure 2c), whereas the time spent to swim this Figure S8F, G). Group 2 and 3 Lphn3-CO larvae did not significantly distance was unchanged between both treatment groups (Lphn3- differ from each other (Supplementary Figure S8D, F). These CO vs Lphn3-MO1 p ¼ 0.2; Figure 2d). Therefore, lphn3.1 mor- observations were further validated by counting the number of phants display more activity bursts and also accelerate signifi- cells (Supplementary Figure S8B). Finally, DAergic projections cantly faster than controls (4.79 mm s -- 1 in Lphn3-MO1 compared appeared gradually but consistently reduced in morphant groups with 2.14 mm s -- 1 in Lphn3-CO; Figure 2e), suggesting increased compared with controls (Supplementary Figure S8C--G). We motor impulsivity in their swimming behavior. conclude that a severe impairment in the formation of the PT DA system does correlate with hyperactivity. However, the mild but significant reduction in DA cell numbers in group 2 Lphn3- Lphn3 morphants show impaired DA system development MO1 larvae, and the consistently decreased DAergic projections in We next aimed to determine whether any changes in neural all Lphn3-MO1 groups, suggests that a threshold may exist for the development and/or neurotransmitter signaling correlated with number of PT DA neurons and/or projections that need to be the behavior of morphants. Several lines of evidence connect DA reduced in order to trigger hyperactivity. (DAergic) signaling to the etiology of ADHD.15,37,38 We first used Other monoaminergic neurotransmitters, including NA and high-performance liquid chromatography to measure the con- serotonin (5-HT),18,44 have been linked to ADHD, and we next centrations of DA and its metabolite 3,4-dihydroxyphenylacetic analyzed hyperactive MO-injected larvae for the development of acid in whole larvae (Supplementary Figure S7A). Lphn3-MO1 these systems. At both 3 d.p.f. and 6 d.p.f., we found that the larvae had no significant changes in either DA level or turnover number (Supplementary Figure S9F) and position (Supplementary compared with Lphn3-CO (Supplementary Figure S7), suggesting Figure S9A--D) of NA neurons in the locus cœruleus45 were similar that DA signaling is not globally disrupted in Lphn3-MO1. in both Lphn3-CO- and Lphn3-MO1-injected larvae. We made an However, because this analysis would not detect subtle alterations analogous observation for 5-HT neurons. In zebrafish, 5-HT in the formation of some DA nuclei, we examined the formation of neurons are found in the raphe´ nucleus, the ventral hypothalamus DA neurons in the brain of 3 d.p.f. and 6 d.p.f. hyperactive lphn3.1 and a pretectal diencephalic cluster.31,46 -- 48 We counted the morphants using anti-tyrosine hydroxylase immunocytochemistry. number of these neurons in each cluster (Supplementary Figure We concentrated our analysis on the DAergic nuclei of the PT, S10M), but did not observe a significant variation in number some of which send projections to subpallial structures. Although or morphology in morphants compared with control fish the functional relationship of this zebrafish system with the (Supplementary Figure S10A--L). Therefore, among candidate human basal ganglia circuitry remains controversial,39 the zebra- neurotransmitter pathways, only DA neurons appeared disorga- fish PT has been implicated in the control of locomotion.40 -- 42 The nized in lphn3.1 morphant larvae. PT comprises 7 populations of neurons (Figure 3k).43 At 3 d.p.f., we detected a global decrease in the number of PT neurons in lphn3.1 morphants (Figures 3b and d, Supplementary Figure S7B). This The behavioral phenotype of Lphn3 morphants can be rescued by included a decrease in populations 1 and 2 and a complete ADHD treatment drugs absence of population 3 (Figures 3a--d). By 6 d.p.f., hyperactive One of the most effective drugs used to treat ADHD is the Lphn3-MO1-injected larvae displayed a severe disorganization of psychostimulant MPH,49 an amphetamine-like compound that all PT neurons (Figures 3e--j), coupled to a reduction in the total acts on the DA transporter.50 MPH is believed to improve attention number of both DA neurons (6 d.p.f. Lphn3-CO vs 6 d.p.f. Lphn3- and cognition, leading to a secondary decrease in psycho- MO1 po0.001, Supplementary Figure S7B) and projections. motor activity including impulsive actions in patients. To explore Populations 2, 3 and 7 were almost completely absent, and whether the hyperactivity of lphn3.1 morphants could also be populations 1 and 4/5 severely reduced (Figures 3f, h, j). To rescued by MPH, we treated larvae with the drug and measured confirm these observations, we analyzed a second DA marker, the locomotion. In a first dose-response study, we found that the gene coding for the DA transporter, slc6a3/dat. The number of concentration of MPH producing the largest behavioral alterations slc6a3-positive neurons was reduced in the PT of Lphn3-MO1 in morphants was 10 mM (Supplementary Figure S11A). We next compared with Lphn3-CO at both 3 d.p.f. (Supplementary Figure compared the locomotor activity of the same zebrafish larvae S7C, D, G, H) and 6 d.p.f. (Supplementary Figure S7E, F, I, J), con- before and after drug treatment at the same concentration. firming the modification of PT neurons upon loss of lphn3.1 function. Treatment of Lphn3-CO animals with 10 mM MPH decreased their Locomotion is a quantitative trait, and activity levels varied resting time (Supplementary Figure S11C) (CO vs CO þ MPH; within a single population of both Lphn3-CO and Lphn3-MO1 po0.01) but did not result in a significant change in their activity, larvae (Supplementary Figure S8A), although on average the whether measured as distance swum (Figure 4a, black bars) (CO vs population of Lphn3-MO1 animals were always hyperactive CO þ MPH; p ¼ 0.19) or speed (Figure 4c) (CO vs CO þ MPH: compared with Lphn3-CO. In order to directly examine the p ¼ 0.36). In striking contrast, MPH decreased the distance swum correlation between hyperactivity and the reduction of DA by Lphn3-MO larvae to a level similar to Lphn3-CO control animals neurons, we compared the PT of individual Lphn3-CO and (Figure 4a)(MO1 vs MO1 þ MPH: po0.001; CO vs MO1 þ MPH: Lphn3-MO1 larvae distributed across the spectrum of locomotion p ¼ 0.4). This rescue was achieved without modifying the resting levels. Larvae were first divided into three groups according to time of Lphn3-MO1 larvae (Supplementary Figure S11C) but

& 2012 Macmillan Publishers Limited Molecular Psychiatry (2012), 946 -- 954 Lphn3.1 controls locomotor activity and impulsivity M Lange et al 950

Molecular Psychiatry (2012), 946 -- 954 & 2012 Macmillan Publishers Limited Lphn3.1 controls locomotor activity and impulsivity M Lange et al 951 through decreasing their locomotion speed (Figure 4c; MO1 vs we re-analyzed the effect of ATO treatment in a single experiment. MO1 þ MPH; po0.05, CO vs MO1 þ MPH; p ¼ 0.3). Together, MPH- We used 1 mM ATO, the minimum dose found to affect morphant treated morphant animals showed a paradoxical reduction in animals (Supplementary Figure S11B). We observed that this locomotion via a rescue of their mean locomotion speed to the treatment had no significant effect on control-injected larvae level of untreated control fish, in a manner reminiscent of the (Figure 4d, Supplementary Figure S11D). However, it rescued the effects of MPH in ADHD patients. hyperactivity of Lphn3-MO1-injected larvae, reducing the distance The symptoms of ADHD in some patients can also be traveled to a level equivalent to that of control fish (Figure 4b; ameliorated by treatment with the selective NA reuptake inhibitor Lphn3-CO vs MO1 þ MPH: p ¼ 0.46), similar to the effect seen in ATO51 -- 54 that also indirectly modulates DAergic neurotransmis- human patients. This calming effect was associated with a sion.55 In a dose-response curve, we observed a general decrease reduction of the mean swimming speed to the level of untreated in the locomotion of both Lphn3-CO and Lphn3-MO1 injected or treated control fish (Figure 4d) and an increase in the larvae upon ATO treatment. However, ATO generally had a resting time of morphants compared with untreated Lphn3-CO stronger effect on morphant fish than controls (Supplementary and Lphn3-MO1 animals (CO vs MO þ ATO: po0.025; MO1 vs Figure S11B; Lphn3-CO vs Lphn3MO1; po0.001). Similar to MPH, MO1 þ ATO: po0.05; Supplementary Figure S11D). We next

Figure 3. The distribution of DA neurons is altered in the PT of Lphn3-MO1 larvae. (a-- j) Dorsal views of anti-tyrosine hydroxylase (TH) antibody staining at 3 d.p.f. (a-- d) and 6 d.p.f. (e-- j). High magnification views of the caudal (g, h) and rostral (i, j) PT area. DA neuron subgroups are labeled 1--7 according to Rink and Wulliman,43 and asterisks denote reduced, missing or possibly misplaced populations in Lphn3-MO1-injected larvae. (k, l) Cartoon dorsal view of TH-positive cell-groups in the posterior tuberculum. Lphn3-CO fish show seven distinct populations of diencephalic DA neurons similar to wild-type larvae (k). Lphn3-MO1-injected larvae have reduced (red asterisks) TH- positive neurons in the posterior tuberculum (l).

Figure 2. Individual morpholino-injected larvae exhibit motor impulsivity. (a) Lphn3-MO1 morphant fish display locomotion peaks. Distance swum (in millimeters) by Lphn3-CO- (left) and Lphn3-MO1-injected larvae (right) for five individual fish, plotted every 3 s during a 90-second experiment. On the first morphant graph the peak parameters are shown (including mean distance swum (horizontal black bar), acceleration time and height of peak (difference between the top and the base of the peak)). Peaks were defined as all acceleration events when the fish traveled 45mm in o12 s. (b) Number of activity peaks for the Lphn3-CO- and Lphn3-MO1-injected populations during a 120-second experiment. n ¼ 10 for each group. Peak episodes occurred on average 6.2 times in morphant larvae against 2.7 times in control fish. ***po0.01. (c) Hyperactive fish cover more distance in each peak. Average height of peaks (in millimeters; the distance values indicated in panel a) is summarized and reported in this graph for the Lphn3-MO1 (average 12.86 mm) and for the Lphn3-CO larvae (average 6.58 mm). ANOVA followed by t-test n ¼ 10 ***po0.001. (d) The duration of the acceleration phase of peaks (for the time values indicated in panel a) are plotted for the morphant and control larvae. Lphn3-MO1 injected (mean duration 3.53 s) and Lphn3-CO larvae (mean duration 3.90 s) exhibit no statistical difference. ANOVA followed by t-test n ¼ 10. NS ¼ non significant. (e) Acceleration values during the peaks episodes for Lphn3- CO and Lphn3-MO1 larvae. Peak acceleration (mm sÀ1) corresponding to the ratio between the distance and time swum during the different peaks for n ¼ 10 animals. The mean acceleration in one peak is 4.76 mm sÀ1 for an Lphn3-MO1- and 2.14 mm sÀ1 for Lphn3-CO-injected larvae. ***po0.01.

& 2012 Macmillan Publishers Limited Molecular Psychiatry (2012), 946 -- 954 Lphn3.1 controls locomotor activity and impulsivity M Lange et al 952

Figure 4. Application of ADHD treatment drugs rescues lphn3.1 morphant hyperactivity. (a) MPH treatment rescues morphant locomotion. Distance swum in a 5-min period by Lphn3-CO (n ¼ 11) and Lphn3-MO1 larvae (n ¼ 11) before and after a 1-hour treatment with 10 mM MPH. The same larvae are analyzed before and after treatment. (b) ATO treatment rescues of morphant locomotion. Distance swum in a 5-min period by Lphn3-CO (n ¼ 12) and Lphn3-MO1 larvae (n ¼ 12) before and after a 1-hour treatment with 1 mM ATO. The same larvae are analyzed before and after treatment. (c) MPH rescues the swimming speed of Lphn3-MO1 larvae to the level Lphn3-CO animals. Average swimming speed of the Lph3-CO- and Lphn3-MO1-injected populations before and after a 10 mM MPH treatment during a 5-min experiment, excluding the time spent resting. n ¼ 12 for each treatment group. (d) ATO reduces the swimming speed of Lphn3-MO1 larvae to the level Lphn3-CO animals. Average swimming speed of the Lph3-CO- and Lphn3-MO1-injected 6 d.p.f. larvae before and after a 1 mM ATO treatment during a 5-min experiment (excluding the time spent resting). n ¼ 12 for each treatment group. All error bars show ±s.e.m., *po0.05, **po0.025, ***po0.001, NS ¼ non-significant.

explored the possible mechanism by which MPH and ATO essential validation of genome-wide gene identification studies treatment could reduce the lphn3.1-linked hyperactivity. We and conclusively implicate LPHN3 in the etiology of ADHD.23 treated wild-type larvae with either 10 mm MPH or 1 mm ATO for The zebrafish genome contains two orthologs of LPHN3, lphn3.1 1 h and measured the expression level of lphn3.1 and lphn3.2. and lphn3.2, as a consequence of an extra whole-genome Application of either MPH or ATO had no effect on lphn3.1 duplication in teleost fish.56 The expression profiles of these two expression. Conversely, MPH mildly but significantly increased genes, appear to partially overlap during development. As a lphn3.2 expression but ATO treatment did not (Supplementary consequence, morpholino knockdown of lphn3.1 alone may cause Figure S12). We also analyzed the effect of ADHD-treatment drug a hypomorphic reduction of total Lphn3 protein (rather than a application on the formation of tyrosine hydroxylase-positive PT complete loss of function) in the brain areas where both genes are DA neurons during development. Treatment of wild-type larvae coexpressed. However, further studies of the effect of reducing with either 10 mm MPH or 1 mm ATO for 6 days (from 0 d.p.f.; both lphn3.1 and lphn3.2 will be required in order to confirm this Supplementary Figure S13B, C, F), 3 days (from 3 d.p.f.; Supple- hypothesis. Genes identified by genome-wide approaches contain mentary Figure S13D, G) or 1 h (on day 6; Supplementary Figure polymorphisms that most likely cause a modulation of gene activity S13E, H) had no global effect on the position or number of DA rather than a complete loss of function.57 Our data suggests that neurons in the PT (Supplementary Figure S13I). Together, these the phenotypic dimensions of ADHD patients carrying the LPHN3 results suggest that the rescue of lphn3.1 hyperactivity by MPH risk haplotype21 --23 mayinpartbecausedbyanincompletereduc- and ATO probably does not occur at the level of gene expression tion of gene activity. Nevertheless, future experiments involving (despite an increase in lphn3.2 expression upon ATO application), analyses of disease-associated LPHN3 variants identified in ADHD or by globally modifying the number of DA neurons in the PT. patients will be needed in order to directly address this issue. Our results have several important implications. Firstly, they highlight a primary and essential role of lphn3.1 in establishing DISCUSSION accurate DAergic signaling during development. How this specific In this study we, have analyzed the developmental and behavioral effect is achieved despite the rather broad expression of lphn3 function of lphn3.1, a zebrafish homolog of the human LPHN3 genes, which step(s) of DA system development are primarily gene that has been identified in ADHD patients by linkage and under Lphn3 control, and which signaling pathway(s) depend on fine-mapping analyses.21 The results presented here provide an Lphn3 action will be important issues to address in the future.

Molecular Psychiatry (2012), 946 -- 954 & 2012 Macmillan Publishers Limited Lphn3.1 controls locomotor activity and impulsivity M Lange et al 953 High-performance liquid chromatography analysis of the level and supported by the Neuroscience School of Paris (ENP), the ANR (Chair of Excellence turnover of DA in the brain reveal no significant differences ANR-08-CEXC-001-01), the FRM (program DPR 20081214424), the PIME program, the between Lphn3-MO1 and Lphn3-CO larvae. This may occur Schlumberger Association (grant DLS/GP/LB090305) and the EU projects NeuroXsys because the modification to DA neuron formation that we see (Grant agreement FP7 2007-2013, no 223262) and ZF-Health (grant agreement in Lphn3-MO1 is restricted to one discrete cluster of neurons, HEALTH-F4-2010-242048). Work in Klaus-Peter Lesch’s laboratory is supported by the Deutsche Forschungsgemeinschaft (DFG KFO 125, SFB 581 and SFB TRR 58) and the alteration of which may not be sufficient to globally change DA Bundesministerium fu¨r Bildung und Forschung (BMBF 01GV0605). levels. Thus, the connection between Lphn3 function and DAergic signaling needs to be studied at a higher resolution, with a focus on specific brain areas and/or projections. Lphn3 expression also appears broadly distributed in mouse but no models for Lphn3 REFERENCES loss-of-function have been examined so far. Latrophilin genes also 1 Geissler J, Lesch KP. A lifetime of attention-deficit/hyperactivity disorder: have an essential role in controlling the establishment of tissue diagnostic challenges, treatment and neurobiological mechanisms. Expert Rev 24 polarity during C. elegans development. Thus our present data Neurother 2011; 11: 1467 -- 1484. provides further evidence of the involvement of Latrophilin family 2 Polanczyk G, Rohde LA. Epidemiology of attention-deficit/hyperactivity disorder adhesion-G protein-coupled receptor proteins in early develop- across the lifespan. Curr Opin Psychiatry 2007; 20: 386 -- 392. ment. Furthermore, as MO only provide a transient knockdown of 3 Swanson JM, Sunohara GA, Kennedy JL, Regino R, Fineberg E, Wigal T et al. gene activity, our study also reinforces the conclusion that ADHD Association of the dopamine receptor D4 (DRD4) gene with a refined phenotype is predominantly a neurodevelopmental disorder.6,58 of attention deficit hyperactivity disorder (ADHD): a family-based approach. Mol Psychiatry 1998; 3:38--41. Our results also demonstrate that reduction of Lphn3 function 4 Faraone SV, Perlis RH, Doyle AE, Smoller JW, Goralnick JJ, Holmgren MA et al. induces robust behavioral changes in larval zebrafish, with Molecular genetics of attention-deficit/hyperactivity disorder. Biol Psychiatry 2005; characteristics reminiscent of the hyperactivity/impulsivity pheno- 57: 1313 -- 1323. type observed in ADHD patients. These include stably increased 5 Lesch KP, Selch S, Renner TJ, Jacob C, Nguyen TT, Hahn T et al. Genome-wide copy locomotor behavior, night-time hyperactivity and bursts of motor number variation analysis in attention-deficit/hyperactivity disorder: association impulsivity. The distinctive response of lphn3.1 morphants to with neuropeptide Y gene dosage in an extended pedigree. Mol Psychiatry 2011; ADHD-specific drugs also appears typical of ADHD behavior, with 16: 491 -- 503. both MPH and ATO restoring morphants to control activity levels. 6 Lesch KP, Timmesfeld N, Renner TJ, Halperin R, Roser C, Nguyen TT et al. Molecular The mechanism by which MPH and ATO rescue the locomotor genetics of adult ADHD: converging evidence from genome-wide association and extended pedigree linkage studies. J Neural Transm 2008; 115: 1573 -- 1585. behavior of lphn3.1 morphants, like in ADHD patients, remains 7 Neale BM, Medland SE, Ripke S, Asherson P, Franke B, Lesch KP et al. Meta-analysis unclear. Even though MPH causes an increased level of lphn3.2 of genome-wide association studies of attention-deficit/hyperactivity disorder. expression following acute treatment, ATO treatment did not J Am Acad Child Adolesc Psychiatry 2010; 49: 884 -- 897. affect the expression of either gene (Supplementary Figure S12). 8 Zhou K, Dempfle A, Arcos-Burgos M, Bakker SC, Banaschewski T, Biederman J et al. MPH and ATO treatment also had no obvious effect on the Meta-analysis of genome-wide linkage scans of attention deficit hyperactivity formation of PT DA neurons during development, even following disorder. Am J Med Genet B Neuropsychiatr Genet 2008; 147B: 1392 -- 1398. chronic drug treatment (Supplementary Figure S13). Thus, MPH 9 Banaschewski T, Becker K, Scherag S, Franke B, Coghill D. Molecular genetics of and ATO are unlikely to modify ADHD-linked behavior by globally attention-deficit/hyperactivity disorder: an overview. Eur Child Adolesc Psychiatry changing lphn3 gene expression levels or by normalizing the 2010; 19: 237 -- 257. 10 Kostrzewa RM, Kostrzewa JP, Kostrzewa RA, Nowak P, Brus R. Pharmacological development of DA neurons. Further experiments, including high- models of ADHD. J Neural Transm 2008; 115: 287 -- 298. performance liquid chromatography measures before and after 11 Russell VA, Sagvolden T, Johansen EB. Animal models of attention-deficit MPH and ATO treatment, will be required in order to better hyperactivity disorder. Behav Brain Funct 2005; 1:9. understand the mechanism by which MPH and ATO can rescue 12 Sagvolden T, Johansen EB, Aase H, Russell VA. A dynamic developmental theory the Lphn3-MO1 behavioral phenotype. of attention-deficit/hyperactivity disorder (ADHD) predominantly hyperactive/ ADHD patients may primarily suffer from deficits in attention impulsive and combined subtypes. Behav Brain Sci 2005; 28: 397 -- 419, discussion and concomitant failure of cognitive control, with hyperactivity 419-368. and increased impulsivity occurring as a consequence.59,60 13 Williams NM, Zaharieva I, Martin A, Langley K, Mantripragada K, Fossdal R et al. Inattention is complex to assess in animal models61 -- 63 and we Rare chromosomal deletions and duplications in attention-deficit hyperactivity disorder: a genome-wide analysis. Lancet 2010; 376: 1401 -- 1408. cannot currently measure it in zebrafish larvae. Nevertheless, the 14 Sagvolden T. Behavioral validation of the spontaneously hypertensive rat (SHR) as phenotypic expression of hyperactivity and increased impulsivity an animal model of attention-deficit/hyperactivity disorder (AD/HD). Neurosci 58,64 in most ADHD patients, and the data that we present here, Biobehav Rev 2000; 24:31--39. strongly suggest that lphn3.1 morphant larvae are a valid model 15 Cook Jr EH, Stein MA, Krasowski MD, Cox NJ, Olkon DM, Kieffer JE et al. for at least some ADHD-associated locomotor endophenotypes. Association of attention-deficit disorder and the dopamine transporter gene. Am J Finally, lphn3.1 morphants may provide a useful platform to screen Hum Genet 1995; 56: 993 -- 998. for novel ADHD therapeutic strategies. 16 Gizer IR, Ficks C, Waldman ID. Candidate gene studies of ADHD: a meta-analytic review. Hum Genet 2009; 126: 51 -- 90. 17 LaHoste GJ, Swanson JM, Wigal SB, Glabe C, Wigal T, King N et al. Dopamine D4 CONFLICT OF INTEREST receptor gene polymorphism is associated with attention deficit hyperactivity disorder. Mol Psychiatry 1996; 1: 121 -- 124. The authors declare no conflict of interest. 18 Oades RD. Role of the serotonin system in ADHD: treatment implications. Expert Rev Neurother 2007; 7: 1357 -- 1374. 19 Walitza S, Renner TJ, Dempfle A, Konrad K, Wewetzer C, Halbach A et al. ACKNOWLEDGEMENTS Transmission disequilibrium of polymorphic variants in the tryptophan hydro- We are grateful to all members of Laure Bally-Cuif’s lab for their support and to xylase-2 gene in attention-deficit/hyperactivity disorder. Mol Psychiatry 2005; 10: Se´bastien Bedu for expert fish care. We thank Drs Alessandro Alunni, Isabelle Foucher 1126 -- 1132. and Monika Krecsmarik for providing us with reagents. We are also indebted to Mario 20 Franke B, Vasquez AA, Johansson S, Hoogman M, Romanos J, Boreatti-Hummer A Kreutzfeldt and all members of Klaus-Peter Lesch’s lab for technical assistance. We et al. Multicenter analysis of the SLC6A3/DAT1 VNTR haplotype in persistent thank M Muenke and M Arcos-Burgos for helpful discussion and comments. We also ADHD suggests differential involvement of the gene in childhood and persistent thank Dr Sylvie Granon for her insightful comments on this project, Dr Theresa ADHD. Neuropsychopharmacology 2010; 35: 656 -- 664. Faus-Kessler for help with the statistical analyses presented in Supplementary Figure 21 Arcos-Burgos M, Jain M, Acosta MT, Shively S, Stanescu H, Wallis D et al. S6A, Dr Baptiste Mouginot for the mathematical analysis of Figure 1c, and Dr Manfred A common variant of the latrophilin 3 gene, LPHN3, confers susceptibility to Schartl for critical reading of the manuscript. Merlin Lange is supported by a ADHD and predicts effectiveness of stimulant medication. Mol Psychiatry 2010; 15: fellowship from the French Ministry of Education. Work in LBC’s laboratory is 1053 -- 1066.

& 2012 Macmillan Publishers Limited Molecular Psychiatry (2012), 946 -- 954 Lphn3.1 controls locomotor activity and impulsivity M Lange et al 954 22 Ribase´s M, Ramos-Quiroga JA, Sa´nchez-Mora C, Bosch R, Richarte V, Palomar G 44 Tripp G, Wickens JR. Neurobiology of ADHD. Neuropharmacology 2009; 57: et al. Contribution of LPHN3 to the genetic susceptibility to ADHD in adulthood: a 579 -- 589. replication study. Genes Brain Behav 2011; 10: 149 -- 157. 45 Ma PM. Catecholaminergic systems in the zebrafish. I. Number, morphology, and 23 Liu QR, Drgon T, Johnson C, Walther D, Hess J, Uhl GR. Addiction molecular histochemical characteristics of neurons in the locus coeruleus. J Comp Neurol genetics: 639 401 SNP whole genome association identifies many 00cell adhesion00 1994; 344: 242 -- 255. genes. Am J Med Genet B Neuropsychiatr Genet 2006; 141B: 918 -- 925. 46 Bellipanni G, Rink E, Bally-Cuif L. Cloning of two tryptophan hydroxylase genes 24 Langenhan T, Promel S, Mestek L, Esmaeili B, Waller-Evans H, Hennig C et al. expressed in the diencephalon of the developing zebrafish brain. Gene Expr Latrophilin signaling links anterior-posterior tissue polarity and oriented cell Patterns 2002; 2: 251 -- 256. divisions in the C. elegans embryo. Dev Cell 2009; 17: 494 -- 504. 47 Kaslin J, Panula P. Comparative anatomy of the histaminergic and other aminergic 25 Fero K, Yokogawa T, Burgess HA. The behavioral repertoire of larval zebrafish. In: systems in zebrafish (Danio rerio). J Comp Neurol 2001; 440: 342 -- 377. Kalueff AV, Cachat JM, eds. Zebrafish Models in Neurobehavioral Research 2011; 48 Teraoka H, Russell C, Regan J, Chandrasekhar A, Concha ML, Yokoyama R et al. 249 -- 291. Hedgehog and Fgf signaling pathways regulate the development of tphR- 26 Lieschke GJ, Currie PD. Animal models of human disease: zebrafish swim into expressing serotonergic raphe neurons in zebrafish embryos. J Neurobiol 2004; view. Nat Rev Genet 2007; 8: 353 -- 367. 60: 275 -- 288. 27 Norton W, Bally-Cuif L. Adult zebrafish as a model organism for behavioral 49 Buitelaar J, Medori R. Treating attention-deficit/hyperactivity disorder beyond genetics. BMC Neurosci 2010; 11:90. symptom control alone in children and adolescents: a review of the potential 28 Rihel J, Prober DA, Arvanites A, Lam K, Zimmerman S, Jang S et al. Zebrafish benefits of long-acting stimulants. Eur Child Adolesc Psychiatry 2010; 19: 325 -- 340. behavioral profiling links drugs to biological targets and rest/wake regulation. 50 Sulzer D, Sonders MS, Poulsen NW, Galli A. Mechanisms of neurotransmitter Science (New York, NY) 2010; 327: 348 -- 351. release by amphetamines: a review. Prog Neurobiol 2005; 75: 406 -- 433. 29 Guo S, Brush J, Teraoka H, Goddard A, Wilson SW, Mullins MC et al. Development 51 Bymaster FP, Katner JS, Nelson DL, Hemrick-Luecke SK, Threlkeld PG, Heiligenstein of noradrenergic neurons in the zebrafish hindbrain requires BMP, FGF8, and the JH et al. Atomoxetine increases extracellular levels of norepinephrine and homeodomain protein soulless/Phox2a. Neuron 1999; 24: 555 -- 566. dopamine in prefrontal cortex of rat: a potential mechanism for efficacy in 30 Lillesaar C, Tannhauser B, Stigloher C, Kremmer E, Bally-Cuif L. The serotonergic attention deficit/hyperactivity disorder. Neuropsychopharmacology 2002; 27: phenotype is acquired by converging genetic mechanisms within the zebrafish 699 -- 711. central nervous system. Dev Dyn 2007; 236: 1072 -- 1084. 52 Newcorn JH. New treatments and approaches for attention deficit hyperactivity 31 Norton WH, Folchert A, Bally-Cuif L. Comparative analysis of serotonin receptor disorder. Curr Psychiatry Rep 2001; 3: 87 -- 91. (HTR1A/HTR1B families) and transporter (slc6a4a/b) gene expression in the 53 Hammerness P, McCarthy K, Mancuso E, Gendron C, Geller D. Atomoxetine for the zebrafish brain. J Comp Neurol 2008; 511: 521 -- 542. treatment of attention-deficit/hyperactivity disorder in children and adolescents: 32 Thisse B, Heyer V, Lux A, Alunni V, Degrave A, Seiliez I et al. Spatial and temporal a review. Neuropsychiatr Dis Treat 2009; 5: 215 -- 226. expression of the zebrafish genome by large-scale in situ hybridization screening. 54 Prasad S, Arellano J, Steer C, Libretto SE. Assessing the value of atomoxetine in Methods Cell Biol 2004; 77: 505 -- 519. treating children and adolescents with ADHD in the UK. Int J Clin Pract 2009; 63: 33 Porrino LJ, Rapoport JL, Behar D, Sceery W, Ismond DR, Bunney Jr WE. 1031 -- 1040. A naturalistic assessment of the motor activity of hyperactive boys. I. Comparison 55 Benkert D, Krause KH, Wasem J, Aidelsburger P. Effectiveness of pharmaceutical with normal controls. Arch Gen Psychiatry 1983; 40: 681 -- 687. therapy of ADHD (Attention-Deficit/Hyperactivity Disorder) in adults - health 34 Taylor E. Clinical foundations of hyperactivity research. Behav Brain Res 1998; 94: technology assessment. GMS Health Technol Assess 2010; 6: Doc13. 11 -- 24. 56 Blomme T, Vandepoele K, De Bodt S, Simillion C, Maere S, Van de Peer Y. The gain 35 Gruber R. Sleep characteristics of children and adolescents with attention deficit- and loss of genes during 600 million years of vertebrate evolution. Genome Biol hyperactivity disorder. Child Adolesc Psychiatr Clin N Am 2009; 18: 863 -- 876. 2006; 7: R43. 36 Patton JH, Stanford MS, Barratt ES. Factor structure of the Barratt impulsiveness 57 Domene S, Stanescu H, Wallis D, Tinloy B, Pineda DE, Kleta R et al. Screening of scale. J Clin Psychol 1995; 51: 768 -- 774. human LPHN3 for variants with a potential impact on ADHD susceptibility. Am J 37 Bush G, Valera EM, Seidman LJ. Functional neuroimaging of attention-deficit/ Med Genet B Neuropsychiatr Genet 2011; 156B: 11 -- 18. hyperactivity disorder: a review and suggested future directions. Biol Psychiatry 58 Castellanos FX, Tannock R. Neuroscience of attention-deficit/hype- 2005; 57: 1273 -- 1284. ractivity disorder: the search for endophenotypes. Nat Rev Neurosci 2002; 3: 38 Zhuang X, Oosting RS, Jones SR, Gainetdinov RR, Miller GW, Caron MG et al. 617 -- 628. Hyperactivity and impaired response habituation in hyperdopaminergic mice. 59 Faraone SV, Biederman J, Mick E. The age-dependent decline of attention deficit Proc Natl Acad Sci U S A 2001; 98: 1982 -- 1987. hyperactivity disorder: a meta-analysis of follow-up studies. Psychol Med 2006; 36: 39 Tay TL, Ronneberger O, Ryu S, Nitschke R, Driever W. Comprehensive catecho- 159 -- 165. laminergic projectome analysis reveals single-neuron integration of zebrafish 60 Rapport MD, Bolden J, Kofler MJ, Sarver DE, Raiker JS, Alderson RM. Hyperactivity ascending and descending dopaminergic systems. Nat Commun 2011; 2: 171. in boys with attention-deficit/hyperactivity disorder (ADHD): a ubiquitous core 40 Blin M, Norton W, Bally-Cuif L, Vernier P. NR4A2 controls the differentiation of symptom or manifestation of working memory deficits? J Abnorm Child Psychol selective dopaminergic nuclei in the zebrafish brain. Mol Cell Neurosci 2008; 39: 2009; 37: 521 -- 534. 592 -- 604. 61 Granon S, Changeux JP. Attention-deficit/hyperactivity disorder: a plausible 41 Bretaud S, Lee S, Guo S. Sensitivity of zebrafish to environmental toxins implicated mouse model? Acta Paediatr 2006; 95: 645 -- 649. in Parkinson’s disease. Neurotoxicol Teratol 2004; 26: 857 -- 864. 62 Siesser WB, Zhao J, Miller LR, Cheng SY, McDonald MP. Transgenic mice 42 Sallinen V, Torkko V, Sundvik M, Reenila I, Khrustalyov D, Kaslin J et al. MPTP and expressing a human mutant beta1 thyroid receptor are hyperactive, impulsive, MPP+ target specific aminergic cell populations in larval zebrafish. J Neurochem and inattentive. Genes Brain Behav 2006; 5: 282 -- 297. 2009; 108: 719 -- 731. 63 Alkam T, Hiramatsu M, Mamiya T, Aoyama Y, Nitta A, Yamada K et al. Evaluation of 43 Rink E, Wullimann MF. Development of the catecholaminergic system in the early object-based attention in mice. Behav Brain Res 2011; 220: 185 -- 193. zebrafish brain: an immunohistochemical study. Brain Res Dev Brain Res 2002; 137: 64 Biederman J, Faraone SV. Attention-deficit hyperactivity disorder. Lancet 2005; 89 -- 100. 366: 237 -- 248.

Supplementary Information accompanies the paper on the Molecular Psychiatry website (http://www.nature.com/mp)

Molecular Psychiatry (2012), 946 -- 954 & 2012 Macmillan Publishers Limited