Tryptophan hydroxylase (TRH) loss of function mutations induce growth and behavioral defects in Daphnia magna

Tryptophan hydroxylase (TRH) is the rate limiting enzyme in the serotonin synthesis. CRISPR-Cas9 technology was used to generate seven indel TRH mutants in Daphnia magna. Mono-allelic indel TRH−/+ clones showed normal levels of serotonin, measured by both immunohistochemistry and mass spectrometry (LC-MS/MS), whereas bi-allelic indel TRH−/− clones showed no detectable levels of serotonin. Life history and behavioral responses of TRH−/− clones showed the anti-phenotype of those exposed to selective serotonin reuptake inhibitors (SSRI). Mutants lacking serotonin grew less and hence reproduced latter, produced smaller clutches of smaller offspring and responded to a greater extent to light than wild type individuals. Mono-allelic indel TRH−/+ individuals showed the intermediate phenotype. The SSRI fluoxetine enhanced offspring production in all clones and decreased the response to light only in those clones having serotonin, thus indication that behavioral effects of this drug in D. magna are associated to serotonin. Results obtained with the TRH mutants are in line with reported ones in TRH knockouts of Caenorhabditis elegans, Drosophila and mice, indicating that there is one gene encoding TRH, which is the serotonin limiting enzyme in both the central and the periphery nervous system in Daphnia and that deprivation of serotonin increases anxiety-like behavior.


Supplementary Tables
924 Chi square and Kruskal-Wallis results testing for effects of clone (C) within each food ratio on the % of individuals investing more than 5 juvenile instars to maturity (Maturation) and age at first reproduction, respectively. ANOVA results testing for: the effects of food (F) and clone on total offspring production (fecundity), offspring size and population growth rates ( r); for effects of photoperiod (Ph), food and clone on swimming distance (swimming); for effects of clone on detected levels of neurotransmitters, feeding responses, oxygen consumption rates and lipid droplets at high food ratio. df, 2, F, p are degrees of freedom, Chi-square, Fisher's coefficient and probability level, respectively. L, H are low and high food ratios, respectively. * Analyses were limited to TRHA-/+, TRHB-/-, TRHC-/-and Wild type clones ** only TRH-/+ and wild type clones were considered. Table S4 . Growth results in the food and SSRI studies showed reduced growth in bi-allelic indel mutated TRH clones lacking serotonin and negligible effects of fluoxetine. von Bertalanffy regression parameters (Mean ±SE) ,r 2 and sample size N and two factor repeated measure ANOVA testing for the effect of the repeated measure (adult instar, I), food ration (F) or fluoxetine (FX) and clone (C) on body length. All regressions curves and their parameters were significant p<0.001. Table S5. Life-history results for seven studied clones in the food study showed reduced fecundity at high food levels, reduced offspring size and population growth rates, and delayed reproduction in bi-allelic indel mutated TRH clones lacking serotonin.  518 Chi square (2) , Kruskal-Wallis(KW) and Mann-Whitney (MW)test results testing for effects of clone (C) and fluoxetine (FX) exposure on the % of individuals investing more than 5 juvenile instars to maturity (Maturation) and age at first reproduction, respectively. ANOVA results testing for: the effects of fluoxetine (FX) and clone on total offspring production (fecundity), offspring size and population growth rates ( r); for effects of photoperiod (Ph), fluoxetine and clone on swimming distance (swimming). df, F, p are degrees of freedom, Fisher's coefficient and probability level, respectively.

CRISPR-Cas9 mediated targeted mutagenesis
There is only one probable orthologue of mammalian tryptophan hydrolase in the D. magna genome (Dapma7bEVm006764t1 hereafter referred as TRH, scaffold00084:361112-363778), which contains 15 exons (Fig 1). Reverse transcription PCR of the TRH gene using a primer set encompassing the targeted mutagenic sites (Table S1) as well as sequencing of the PCR fragments revealed that this gene is transcribed in D. magna ( Figure   S1). Then CRISPR-Cas9 targeted mutagenesis was performed according to Nakanishi, et al. 1 . Briefly, for the synthesis of Cas 9 mRNAs, templates with T7 promoter were amplified by PCR using PrimeSTAR (Takara Bio, Shiga, Japan) from the pCS-Dmavas-Cas9. Amplified  Table S2) were annealed and then ligated into the linearized pDR274 vector (Addgene plasmid 42250) using a ligation mix (TaKaRa Bio, Shiga, Japan).
Oligonucleotides for knocking down the TRH gene were designed using ZiFiT targeter version 4.2 2 . The two genomic targeted mutagenesis sites and sequences of the oligonucleotides used in this study are listed in Figure 1. pDR274-TRH vectors were then digested by DraI and used as templates for in vitro transcription with the mMessage mMachine T7 kit, followed purification. By using a BLASTn search on the D. magna genome database (wfleabase.org), we looked for potential off-target sites and found that TRH-1 and TRH-2 target sites had no potential off-target sites.
Cas9 and gRNAs were co-injected into Daphnia eggs as performed in Kato, et al. 3 . Briefly, eggs were collected immediately after ovulation from 2 to 3 weeks old daphnids and kept in S15/S21 ice-chilled M4-sucrose 80 mM. Injection was performed through a glass needle under N 2 gas pressure. The injection volume was approximately 0.2 nL and microinjections were carried out within an hour after ovulation. Injected eggs were then kept in M4-sucrose for 60 hours and embryonal development was monitored. After first and second generation PCR amplification of target loci were performed on genomic DNA extracted from alive mutant clonal lines to characterize Cas9-induced mutations. Genomic DNA was extracted from single daphniids by homogenization in 90 mL of 50 mM NaOH. The lysate was heated at 95 o C for 10 min and then neutralized with 10 mL of 1 M Tris-HCl (pH 7.5). This crude DNA extract was centrifuged at 10,000 g for 5 min prior to being used as a template for genomic PCR. All PCRs were performed with ExTaq DNA polymerase, Hot Start version (Takara).
The PCR products were analyzed firstly by polyacrylamide gel electrophoresis (PAGE) and selected lines were DNA sequenced using the Sanger method. The primers used for PCR and DNA sequencing are also listed in Table S4 .

Feeding
Feeding assays were conducted with groups of five juveniles transferred into individual test vessels filled with 100 mL of media at high food ration level ( 5 x 10 5 cells/ml C. vulgaris).
Two replicate vessels filled with the same culture medium but with no animals were used as blank replicates. The mean initial cell concentration of the experimental vessels at the start of the experiment (time t0) was determined from three 5-ml samples obtained from the treatment medium before it was distributed into all the experimental vessels. After 24 h, the final cell concentration (time t24) was measured. Feeding was then measured using the celldifference method following Barata and Baird 4 and expressed as proportional responses relative to controls.

S16/S21
Oxygen consumption assays were performed without food using standard respirometry methods with 50 ml gas-tight syringes (Hamilton, USA) as described by Agra, et al. 5 , and expressed as expressed as proportional responses relative to the wild type clone.

Storage lipid accumulation
Quantification of storage lipids into lipid droplets follow previous methods 6 . Nile red stock solutions were prepared in acetone and store protected from light. Just before use, the working solution was obtained by dilution of stock solution to 1.5 µM in ASTM. Live individuals were then exposed to Nile red working solution in the dark for 1 h at 20 ºC.
After incubation, animals were place in 100 ml ASTM for 1 min to allow clearance of Nile red residuals. Following clearance animals were placed individually in 1.5 ml centrifuge tubes, the remaining water removed and sonicated in 300 µl of isopropanol. The homogenized extract was then centrifuged at 10 000 g. We used 200 µl of supernatant to measure Nile red fluorescence using an excitation/emission wavelength 530/590 nm and a microplate fluorescence reader (Synergy 2, BioTek, USA). For each quantification and treatment 10 blanks (non-exposed animals to Nile red) were used to account for background levels of fluorescence. Nile red fluorescence was expressed as proportional responses relative to the wild type clone.

Brain whole-mount immunofluorescence
Methods for immunofluorescence microscopy were performed as described previously 7,8 . In  isotope-labeled internal standards. Samples were then centrifuged to remove animal debris, before protein precipitation was allowed at -20C for 30 min. Finally, samples were centrifuged at 14.500 rpm for 10 min, evaporated to dryness under N2 current and resuspended in 50 µL ACN : H2O (1:1, v/v). A further step of 10 min centrifugation at 10 000 g was included to ensure sample purity before injection. All steps were carried out at 4C and in dark conditions. Analysis of neurotransmitters was performed by liquid chromatography-tandem mass spectrometry following Tufi, et al. 9 with some minor modifications. Briefly, analyte separation was obtained by using a TSK GelAmide 80 HILIC column (2 × 250 mm, 5 μm particle size, Sigma Aldrich) and the analyses was carried out  Table S8. Data was acquired and processed using the MassLynx v4.1 software package. Quantification was based on the most intense transition of each analyte. Stable isotope-labeled internal standards were used for calibration and quantification. For the method validation several parameters were determined and are depicted in Table S9: inter-and intra-day variation, linearity, limit of detection (LOD) and limit of quantification (LOQ). LOD and LOQ were defined as the minimum detectable amount of analyte with a signal to noise ratio (RMS) of 3:1 and 10:1, respectively. S19/S21