Suplementary Methods

Jukic et al.

SUPPLEMANTARY

SUPLEMENTARY METHODS

En1+/Otx2 mutants

The generation and genotyping of En1+/Otx2 mice has beenreported earlier (Broccoli, et al 1999; Brodski, et al 2003). Mutants were kept on a CD-1 genetic background. For lithium and carbamazepine repetitive OFTexperiments mutants were backcrossed for 6 generations to a Black Swiss genetic background. Mice were kept in a temperature-controlled (21–22 °C) roomunderreversed 12 h light/dark cycle (lights were turned off at 10:00 a.m.),with free access to food and water.After weaning, animals were kept as groups of 4-6 in a cage. All experiments were performed on 3-6 months old males, which were single cagedone week prior to any behavioral test. OFT, EPM, LDB, and social interaction tests were all conducted between 12:00 and 17:00 inthe separate behavior testing room. Mice were transferred to the behavior testing room 30 min prior to the beginning of the above mentioned behavioral tests in order to habituate to the new environment. In case that the same animals were used repetitively in more than one behavioral test (Supplementary figure 2a,d,e), tests wereconducted from the least invasive to the most invasive test: OFT, LDB, and EPM with at least 72 h break between the tests (Flaisher-Grinberg and Einat, 2010).

Home cage locomotor activity

Prior to the testing, animals were anesthetized with an intraperitoneal injection of a ketamine-xylazine mixture (100 mg/kg and 10 mg/kg, respectively) and chronically implanted with DSI TA-F10 telemetry probes. The transmitter was able to freely move among the peritoneal organs because it was not attached to the peritoneum. After ten days of recovery, locomotor activity of animals was recorded and analyzed with DSI DataQuest ART 4.3 software. Locomotor activity (counts) was obtained by counting the number of impulses, detected by changes in signal strength, per time unit. The signal was received by an antenna under each animal’s cage and transferred to a peripheral processor connected to a personal computer. All transmitters were calibrated before surgery and at the completion of experimentation to ensure the validity of biotelemetry measurements. Locomotor activity was calculated as an hourly or daily average of the raw data, collected at 1 s intervals by counting the number of impulses, detected by changes signal strength.(Tarasiuk, et al2014).

Open Field Test

Mice were placed in the open fields made of Plexiglas and measuring 40 x 40 x 30 cm under red light. The box was cleaned with a dilute ethanol solution (10%) between trials and carefully dried. For locomotor activity, animals were videotaped and recordings were analyzed off-line using Noldus EthoVision® 9 program. In order to define the center/periphery exploration, we defined the center zone in Noldus Ethovision® 9 program (20 x 20 cm). In order to study intra-session habituation of En1+/Otx2 mice in more detail, we quantified the level of activity change ratios (ACR). Values derived with the formula ACR=b/(a+b), were computed for each trial of each individual mice as described previously (Bothe, et al 2004; Leussis and Bolivar 2006). In this formula a is the activity during the first 5 minutes, and b is the averaged value of the 5 min intervals after 15 minutes of habituation. Values below 0.5 indicate a decrease, values above 0.5 indicate an increase and values closely around 0.5 indicate no changes in activity over time.

Light-Dark Box

The experiments were conducted in a well-lit behavioral testing room.The Plexiglas apparatus was constructed from a 13 x 40 x 30 cm dark compartment closed with a lid and a 27 x 40 x 30 cm light compartment. Those two were joined by a5 x 5 cm opening in the wall between them. The apparatus was cleaned with a dilute ethanol solution (10%) between trials and carefully dried. Mice were placed in the light compartment, videotaped for 10 min, and recordings were analyzed off-line using the Noldus EthoVision® 9 program.

Elevated Plus Maze

The experiments were conducted in a well-lit behavioral testing room.The apparatus was constructed out of transparent Plexiglas and consisted of two opposing open arms and two perpendicular opposing closed arms. The arms were 50 cm in length and 5 cm wide. The closed arms had walls that were 40 cm high, whereas the open arms had a half cm ledge to prevent falling. The EPM was wiped clean with a dilute ethanol solution (10%) between trials and carefully dried. At the start of the trial, the mouse was placed in the center of the maze, facing a closed arm and videotaped for 5 min. The number of entries and time spent in each of the arms were analyzed off line using the Noldus Ethovision® 9 program. We measured the number of entries and the time spent in the open arms as a percentage of the total number of entries and the time spent in all arms.

Chronic sucrose preference

Prior to the experiment animals were exposed for two weeks to two bottles of tap water. During the following 6 weeks animals received a choice between water and a 2% or 5% sucrose solution and sucrose preference was calculated for the last four weeks according to the weekly protocol. Every week, mice were given access to two drinking bottles positioned one next to the other. Twenty-four hour fluid consumption was measured by daily weighing the bottles during the last hour before the onset of the dark phase (9:00-10:00 AM). All interaction with the animals was conducted during this period. Sucrose solution (2% and5% prepared in tap water) was placed in one bottle and the tap water in the other. In order to study fluctuations in sucrose preference over time, we designed a one week protocol according to a previously described method (Taliaz, et al 2011). For day one and two of the week, the sucrose solution was placed on the left side of the cage and water on the right. On day three of the week, only tap water was administered on both sides and all animals were weighed. On day four and five the sucrose solution and water were placed in reverse to control for side preference. On day six and seven of the week animals received two tap water bottles. Sucrose preference for each mouse was defined as the average percentage of sucrose consumption of the total liquid consumption. Sucrose and total liquid intake were calculated by normalizing the volume of consumed sucrose solution or total liquid intake with the animal weight for that week. Only day two and day five were included in the analysis. This procedure was repeated for six weeks, out of which the first two weeks represent the adjustment period and were not taken into the analysis.

Social interaction test

Two mice of the same age, weight and genotype were placed inthe opposite corners of the open-fieldunder the red light as previously described (Shaltiel, et al 2008). Their activities were videotaped for 5 min, and thetime spent in social activity was manually scored using the Noldus EthoVision® program.The open field was wiped cleanbetween trials with a 10% alcohol solution.

Acute (IP) injections

Quipazine maleate (5mg/kg, purchased from Sigma-Aldrich Israel) and CP-809101 (5mg/kg, purchased from Tocris Bioscience) were both dissolved in saline and injected IP into test animals 15 minutes prior to behavioral tests. These doses were chosen to robustly increase serotonin receptor activation in the brain as previously described(Hutchinson, et al 2012; Kaur and Kulkarni 2002; Siuciak, et al 2007). Olanzapine (purchased from Sigma-Aldrich Israel) was dissolved in 0.1 M HCl, diluted with water, and titrated with 0.1 M NaOH towards neutrality (pH higher than 5) without any precipitation. Vehicle and olanzapine solutions (1 mg/kg) were injected IP into test animals one hour before behavioral tests.This protocol was chosen according to the previously reported anti-manic effect in mice (Engel, et al, 2010).Animals were tested in the OFT, LDB, and EPM in this orderfor the animals treated with olanzapine (figures S4, 3a, b) and quipazine (figure 4a, b), with a distinct cohort used for each drug. Established wash-out periods of 14 days were used between repetitive acute IP injections followed by behavioral tests (Mattiuz et al, 1997; Valencia-Flores et al, 1990). The sequence of the tests was conducted from the least invasive to the most invasive test (Flaisher-Grinberg and Einat, 2010).

Chronic lithium and carbamazepine treatment

Lithium carbonate and carbamazepine (purchased from Sigma-Adrich Israel) were mixed into powdered food. For the lithium experiments, an additional bottle of saline was added to each cage in order to compensate for the electrolyte loss in lithium treated animals. After one week of food containing 0.12% lithium, lithium concentrations were raised to 0.24%, for four weeks before the start and throughout the testing (Einat, et al 2003). After the experiment lithium serum concentrations were measured and only animals having lithium levels in the therapeutic range (0.5-1.2 mEq/l) were included in the analysis. Six OFTs were conducted during a period of three weeks. For carbamazepine treatment mice received for one week powdered food containing carbamazepine at a concentration of 0.25%. Two weeks before the start and throughout the testing, carbamazepine concentrations in the food were 0.5% as previously suggested (Kara, et al 2014). Four OFTs were conducted during a period of two weeks.

Statistical analysis:

All results are expressed as mean ± SE. IBM® SPSS® Statistics 21 software was used for the statistical analysis. A two-tailed unpaired Student's t-test was used for the comparative analyses. For the experiments that included pharmacological treatment, in addition to genotype, treatment effect was analyzed as a factor for two-way multifactorial analysis of variance (ANOVA test).Where significant effects were detected, Fisher’s LSD post hocanalysis indicated significant differences between individual groups.

INRICH Pathway analysis

1. Compilation of gene lists:Lists of genes specifying DA and 5HT neurons were compiled in a two-step process. In the first step, a search was performed using the search terms “dopamine and development”, “dopamine and embryo”, as well as “serotonin and development” and “serotonin and embryo”, including studies published up untilMay 2014. In a second step, the published studies obtained from this search were screened according to the following criteria: 1. Studies performed in vivo 2. Studies performed in mammals 3. Studies related to midbrain dopaminergic or rostral serotonergic neurons. 4. Studies related to the specification of monoaminergic neurons. We choose these criteria according to the following rationales. 1: In vitro studies are not necessarily relevant in vivo (Hegarthy et al., 2013). 2: Only experiments performed in mammals were considered, since pathways for monoaminergic neuronal specification in non-mammals differ substantially from mammalian pathways (Hegarthy et al., 2013). 3. Only data relevant for midbrain DA or rostral 5HT sub-populations were considered, since they are affected in En1+/Otx2 mutants and relevant to psychiatric disorders. In contrast, other DA and 5HT sub-populations, not affected in En1+/Otx2 mutants, e.g. olfactory DA or caudal hindbrain 5HT neurons projecting to the spinal cord, are less likely to be involved in these disorders. 4. Since En1+/Otx2 mutants exhibit alterations in the specification of monoaminergic neurons rather than in their migration or projection (Brodski, et al 2003), only studies that report this aspect of developmentwere taken into consideration. Based on these publications a gene list was compiled for DA as well as for 5-HT neurons.

2. INRICH AnalysisIn order to test if the sequence variants of selected gene sets are correlated with psychiatric disorders we used INRICH, an interval-based enrichment analysis tool for GWAS (Lee, et al 2012). INRICH evaluates each target set by considering the number of intervals that contain at least one target gene sequence. The data used for the statistical analysis was obtained from the meta-analysis study of bipolar disorders conducted by the Psychiatric Genomics Consortium (PGC) (Psychiatric GWAS Consortium Bipolar Disorder Working Group 2011). Associated LD-independent genomic intervals were based on the LD pruned SNP dataset downloaded from PGC website ( on 10.11.2013 for each disease (Supplementary files: pgc.bip.clump for BPD, pgc.scz.clump for SCZ, and pgc.mdd.clump for MDD). Intervals were spanning 250kbp up/downstream on variants attaining statistical significance for relationships at P≤0.001. We defined gene regions as 100 kb up/downstream of the transcription starting/ending sites for all genes on autosomal chromosomes based on the ENTREZ gene map hg 18 (Maglott, et al 2005) downloaded from the INRICH website (atgu.mgh.harvard.edu/inrich/downloads.html) on 10.11.2013 (Supplementary file entrez.gene.map). Significance was assessed through 100,000 permutations in which randomly selected genomic regions matched for gene and SNP density were compared with the target interval set. In order to test whether there is a degree of correspondence between the test intervals and the target sets, we calculated global enrichment. It indicates whether more targets than expected by chance are within one of the associated intervals and within at least one set that is nominally significant at P≤0.01. Finally, resampling-based 10000 second-step permutation was conducted in order to adjust the empirical P-values for testing multiple candidate target sets.

Since we repeatedly tested the hypothesis that sequence variants of genes directing the development of DA neurons are associated with BPD using different data sets, we corrected for multiple testing. We used the p-values obtained from the INRICH analysis using the lists of genes directing the specification of dopaminergic neurons compiled by us and by Hegarthy et al, for the FDR correction, according to SanGiovanni and Lee, 2013.

TABLE 1 REFERENCES

Alberi L, Sgado P, Simon HH (2004). Engrailed genes are cell-autonomously required to prevent apoptosis in mesencephalic dopaminergic neurons. Development 131: 3229-3236.

Alves dos Santos MT and Smidt MP (2011). En1 and Wnt signaling in midbrain dopaminergic neuronal development. Neural Dev 6: 23-8104-6-23.

Andersson E, Jensen JB, Parmar M, Guillemot F, Bjorklund A (2006a). Development of the mesencephalic dopaminergic neuron system is compromised in the absence of neurogenin 2. Development 133: 507-516.

Andersson E, Tryggvason U, Deng Q, Friling S, Alekseenko Z, Robert B, Perlmann T, Ericson J (2006b). Identification of intrinsic determinants of midbrain dopamine neurons. Cell 124: 393-405.

Andersson ER, Prakash N, Cajanek L, Minina E, Bryja V, Bryjova L, Yamaguchi TP, Hall AC, Wurst W, Arenas E (2008). Wnt5a regulates ventral midbrain morphogenesis and the development of A9-A10 dopaminergic cells in vivo. PLoS One 3: e3517.

Baquet ZC, Bickford PC, Jones KR (2005). Brain-derived neurotrophic factor is required for the establishment of the proper number of dopaminergic neurons in the substantia nigra pars compacta. J Neurosci 25: 6251-6259.

Blum M (1998). A null mutation in TGF-alpha leads to a reduction in midbrain dopaminergic neurons in the substantia nigra. Nat Neurosci 1: 374-377.

Briscoe J, Sussel L, Serup P, Hartigan-O'Connor D, Jessell TM, Rubenstein JL, Ericson J (1999). Homeobox gene Nkx2.2 and specification of neuronal identity by graded Sonic hedgehog signalling. Nature 398: 622-627.

Brodski C, Weisenhorn DM, Signore M, Sillaber I, Oesterheld M, Broccoli V, Acampora D, Simeone A, Wurst W (2003). Location and size of dopaminergic and serotonergic cell populations are controlled by the position of the midbrain-hindbrain organizer. J Neurosci 23: 4199-4207.

Chakrabarty K, Von Oerthel L, Hellemons A, Clotman F, Espana A, Groot Koerkamp M, Holstege FC, Pasterkamp RJ, Smidt MP (2012). Genome wide expression profiling of the mesodiencephalic region identifies novel factors involved in early and late dopaminergic development. Biol Open 1: 693-704.

Craven SE, Lim KC, Ye W, Engel JD, de Sauvage F, Rosenthal A (2004). Gata2 specifies serotonergic neurons downstream of sonic hedgehog. Development 131: 1165-1173.

Deng Q, Andersson E, Hedlund E, Alekseenko Z, Coppola E, Panman L, Millonig JH, Brunet JF, Ericson J, Perlmann T (2011). Specific and integrated roles of Lmx1a, Lmx1b and Phox2a in ventral midbrain development. Development 138: 3399-3408.

Ding YQ, Marklund U, Yuan W, Yin J, Wegman L, Ericson J, Deneris E, Johnson RL, Chen ZF (2003). Lmx1b is essential for the development of serotonergic neurons. Nat Neurosci 6: 933-938.

Dudok JJ, Groffen AJ, Toonen RF, Verhage M (2011).Deletion of Munc18-1 in 5-HT neurons results in rapid degeneration of the 5-HT system and early postnatal lethality.PLoS One6:e28137

Ellisor D, Rieser C, Voelcker B, Machan JT, Zervas M (2012). Genetic dissection of midbrain dopamine neuron development in vivo. Dev Biol 372: 249-262.

Ferri AL, Lin W, Mavromatakis YE, Wang JC, Sasaki H, Whitsett JA, Ang SL (2007). Foxa1 and Foxa2 regulate multiple phases of midbrain dopaminergic neuron development in a dosage-dependent manner. Development 134: 2761-2769.

Fox SR and Deneris ES (2012). Engrailed is required in maturing serotonin neurons to regulate the cytoarchitecture and survival of the dorsal raphe nucleus. J Neurosci 32: 7832-7842.

Hendricks TJ, Fyodorov DV, Wegman LJ, Lelutiu NB, Pehek EA, Yamamoto B, Silver J, Weeber EJ, Sweatt JD, Deneris ES (2003). Pet-1 ETS gene plays a critical role in 5-HT neuron development and is required for normal anxiety-like and aggressive behavior. Neuron 37: 233-247.

Hoekstra EJ, von Oerthel L, van der Heide LP, Kouwenhoven WM, Veenvliet JV, Wever I, Jin YR, Yoon JK, van der Linden AJ, Holstege FC, Groot Koerkamp MJ, Smidt MP (2013). Lmx1a encodes a rostral set of mesodiencephalic dopaminergic neurons marked by the Wnt/B-catenin signaling activator R-spondin 2. PLoS One 8: e74049.

Hoekstra EJ, von Oerthel L, van der Linden AJ, Smidt MP (2012). Phox2b influences the development of a caudal dopaminergic subset. PLoS One 7: e52118.

Hynes M, Poulsen K, Tessier-Lavigne M, Rosenthal A (1995). Control of neuronal diversity by the floor plate: contact-mediated induction of midbrain dopaminergic neurons. Cell 80: 95-101.

Hynes M, Stone DM, Dowd M, Pitts-Meek S, Goddard A, Gurney A, Rosenthal A (1997). Control of cell pattern in the neural tube by the zinc finger transcription factor and oncogene Gli-1. Neuron 19: 15-26.

Hynes M, Ye W, Wang K, Stone D, Murone M, Sauvage F, Rosenthal A (2000a). The seven-transmembrane receptor smoothened cell-autonomously induces multiple ventral cell types. Nat Neurosci 3: 41-46.

Hynes M, Ye W, Wang K, Stone D, Murone M, Sauvage F, Rosenthal A (2000b). The seven-transmembrane receptor smoothened cell-autonomously induces multiple ventral cell types. Nat Neurosci 3: 41-46.

Iwakura Y, Wang R, Abe Y, Piao YS, Shishido Y, Higashiyama S, Takei N, Nawa H (2011). Dopamine-dependent ectodomain shedding and release of epidermal growth factor in developing striatum: target-derived neurotrophic signaling (Part 2). J Neurochem 118: 57-68.

Jacob J, Ferri AL, Milton C, Prin F, Pla P, Lin W, Gavalas A, Ang SL, Briscoe J (2007). Transcriptional repression coordinates the temporal switch from motor to serotonergic neurogenesis. Nat Neurosci 10: 1433-1439.