Jukić et al.Involvement of CYP2C19 in depressive phenotypes

SUPPLEMENTARY INFORMATION

ElevatedCYP2C19 expression is associated with depressive symptoms and hippocampal homeostasis impairment

Marin M Jukić1, Nils Opel2, Jennifer Ström1, Tania Carrillo-Roa3, Sharon Miksys4, Maria Novalen4, Anna Renblom1, Sarah C Sim1, Eva M Peñas-LLedó5, Philippe Courtet6, Adrián LLerena4, Bernhard T Baune7, Dominique J de Quervain8, Andreas Papassotiropoulos8,9, Rachel FTyndale4, Elisabeth B Binder3, Udo Dannlowski2,10, and Magnus Ingelman-Sundberg1,*

1 Section of Pharmacogenetics, Department of Physiology and Pharmacology, Karolinska Institutet, Stockholm, Sweden

2 Department of Psychiatry, University of Münster, Münster, Germany.

3 Department of Translational Research in Psychiatry, Max Planck Institute of Psychiatry, Munich, Germany

4 Campbell Family Mental Health Research Institute, Centre for Addiction and Mental Health & Department of Psychiatry & Department of Pharmacology and Toxicology, University of Toronto, Toronto, Ontario, Canada.

5 CICAB Clinical Research Center, Extremadura University Hospital & Medical School, Badajoz, SpainCIBERSAM, Madrid, Spain

6 CHU Montpellier, Hôpital Lapeyronie, Psychiatric Emergency & Post-Acute Care Department, Pole Urgence, Montpellier, France

7Discipline of Psychiatry, School of Medicine, University of Adelaide, Adelaide, Germany.

8 Transfaculty Research Platform, Department of Psychology & University Psychiatric Clinics, University of Basel, Basel, Switzerland

9 Life Sciences Training Facility, Department Biozentrum, University of Basel, Basel, Switzerland

10Department of Psychiatry, University of Marburg, Marburg, Germany

* Correspondence: Prof. Magnus Ingelman-Sundberg, Section of Pharmacogenetics, Department of Physiology and Pharmacology, Karolinska Insittutet, Stockholm 171 77, Sweden. E-mail:

Keywords:CYP2C19, animal model, hippocampus, depression, 5-HT1A receptor

Clinical Studies

CYP2C19 Genotyping and Enzymatic Capacity Determination

Since all participants were of European or African-American (AA) origin, only CYP2C19*2 and CYP2C19*17 alleles were taken into account, and not the Asian population-specific CYP2C19*3 defective allele1. The Münster subjects (Figure 1a-c)were genotyped using the Infinium PsychArray-24 (Illumina, CA, USA), Basel subjects (Figure 1d) were genotyped using theGenome-Wide Human SNP Array 6.0 (Affymetrix, CA, USA), and African-American subjects (Figure 2b-e) were genotyped using the Infinitum OmniExpress (Illumina) by analyzing tagging SNPs.CYP2C19*17 is rs12248560, which is perfectly tagged (r2=1.0) by the array SNP rs7918101 and CYP2C19*2 is rs4244285, which is perfectly tagged (r2=1.0) by the array SNP rs4641393. The CYP2C19*1 allele was scored with 100% enzymatic capacity, defective CYP2C19*2 allele was scored with 0%, while the CYP2C19*17 allele was scored with 150% enzymatic capacity according to the data from the previous pharmacokinetic reports2, 3. Enzymatic capacity was calculated according to these scores from the CYP2C19 genotype (see Fig. 2a)

Neuroimaging Studies

386 healthy subjects (191 women; age range 17-59 years, mean age ± standard deviation (SD) 37.3 ± 11.6) were recruited via newspaper advertisement. Clinical diagnosis were obtained using the DSM-IV Structured Clinical Interview (SCID-I)4, Hamilton Rating Scale for Depression (HAM-D)5, and Beck Depression Inventory (BDI)6 were administered to assess the current level of depression. Only the subjects showing scores of <10 in the BDI were included in the study andthey were also ensured to be free from any history of psychiatric disorders or any psychotropic medication. Further exclusion criteria were any history of severe neurological (eg, concussion, stroke, tumor, neuro-inflammatory diseases) and medical (eg, cancer, chronic inflammatory or autoimmune diseases, infections) conditions. The reported ethnicity was confirmed using multidimensional scaling (MDS) of genome-wide genotypes in PLINK7 to avoid possible confounding by population stratification; all subjects clustered together across MDS components. The Ethics Committee at the University of Münster approved the study, and written informed consent was obtained from all participants.

Image acquisition. T1-weighted high-resolution anatomical images of the head were acquired using a three-dimensional fast gradient echo sequence (turbo field echo), with a repetition time of 7.4 milliseconds, echo time = 3.4 milliseconds, flip angle = 9°, two signal averages, inversion prepulse every 814.5 milliseconds, acquired over a field of view of 256 (feet-head) x 204 (anterior-posterior, AP) x 160 (right-left, LR) mm, phase encoding in AP and RL direction, reconstructed to voxels of 0.5 mm x 0.5 mm x 0.5 mm. Images were then processed using theFSL-FIRST algorithms as we and others have previously described8, 9. Briefly, FIRST is a fully automated segmentation software included in the FMRIB Software Library (FSL, Version 5.0.0),which uses pre-defined anatomical landmarks to delineate the hippocampal boundaries, thus providing analyzing tools not only for overall volume information of the segmented structures but also for localized differences in shape( The segmentation process includes boundary correction and registration to a MNI152 template using 12 degrees of freedom as well as a subcortical mask to detect and eliminate voxels outside the subcortical structure. Quality of segmentation and registration were manually checked for each subject. Volumes of the bilateral hippocampi were computed and further analyzed using the SPSS software package (IBM SPSS Software Package Version 23).For visualization purposes, the localization of hippocampal shape differences was computed running surface based vertex analysis as implemented in FSL-FIRST, which is based on multivariate statistics yielding a 3D mesh of the structure of interest.

For the replication study, 1032 healthy young subjects (630 women; age range 18-34 years, mean age ± standard deviation (SD): 22.5 ± 3.3) were recruited. Participants filled in a self-rating questionnaire concerning their health status, medication, and drug consumption. All included subjects were free of any physical, neurological or psychiatric illness. SNP array-based correction for cryptic relatedness and for unbalanced genetic background was applied as described previously10, 11 and genetic outliers were excluded prior to the analysis. The ethics committee of the Canton of Basel approved the experiments (approval number: 190/08). Written informed consent was obtained from all subjects prior to participation.Imageacquisitionandextractionofhippocampalvolumes was done as reported in12. Briefly, we acquiredananatomicalsequencewitharadio-frequencypulses andrapidgradient-echo(MPRAGE)sequence.Segmentations ofcorticalandsubcorticalstructureswereretrieved fromFreeSurfer4.5andlabelingwasbasedontheDesikan–Killiany Atlas13. Leftandrighthippocampal volumeswerecorrectedseparatelyfor intracerebral volume,age,sexand differencesdue tosoftwareandgradientupdatesbyusingthez-transformed residualsofalinearregression.

Genetic Study of Depression Severity

The effect of CYP2C19 genotype onMDD was examined in the subjects derived from Grady Trauma Project (GTP), a large study conducted in Atlanta, Georgia, that investigates the role of genetic and environmental factors in shaping responses to stressful life events. The original GTP includes participants predominantly from the African-American urban population of low socioeconomic status14, 15. This population is characterized by high prevalence and severity of trauma over the lifetime.For the purpose of this study, the subset of 3849African-Americansubjects (2784 women, mean age ± SD: 40.2 ± 13.8) characterized by the BDI6 , CTQ16 , and PSS17 scales was used. The reported ethnicity was confirmed using multidimensional scaling (MDS) of genome-wide genotypes in PLINK7. To avoid possible confounding by population stratification, only subjects which clustered together across MDS components were included in the analyses. All participants provided written informed consent and all procedures were approved by the Institutional Review Boards of the Emory University School of Medicine and Grady Memorial Hospital (IRB00002114).

Genetic Study of Suicidality

A total of 209 depressed suicide attempters (181 women, mean age ± SD: 37.4 ± 12.4) were evaluated in a specialized unit of the Montpellier University Hospital and they represent a subset of the previously described larger cohort of participants18. A suicide attempt was defined as ‘a self-destructive behavior with intent to end one’s life independent of resulting damage’ and the patients were evaluated with the Beck Suicide Intent Scale (SIS)19within the initial hours of a failed suicide attempt. The objective circumstances section of SIS was used to address the severity of suicide intent. A high score in this subscale has shown predictive validity for eventual suicide19 and it assesses the severity of the characteristics related to the suicidal act or the active preparation or planning of the attempt such as: isolation, timing, suicide notes, and precautions taken against discovery and intervention. Participants were included in the analysis if they suffered from MDD as assessed by the Mini-International Neuropsychiatric Interview and they had no family history of suicide, previously shown to independently increase the suicide intent18. Since all subjects from this cohort had four biological grandparents originating from Western European countries, the chance for any noticeable population stratification is extremely low18.All participants gave written informed consent, as approved by the local Ethic Committee, after receiving information on the study.

Behavioral Studies

Restrain Procedure

Mice were taken from their home cage, placed inside a ventilated 50ml centrifuge tube, left there for a defined amount of time, and finally returned to the home cage. For the acute restrain stress (ARS) procedure, mice were restrained for 30 min. Mice subjected to ARS were removed from the tubes and decapitated either immediately or after additional 30 minutes of recovery in the home cage. For the chronic restrain stress (CRS) paradigm, mice were restrained for 4 hours every day over the course of 4 weeks. Animal weight was measured for the last three consecutive days of the week and then averaged; a relative decrease in weight calculated for each animal as the measurement of the stress-responsivity. On the last day of CRS, animals were exposed to the light-dark box before the restrain session for that day. On the next day, animals were exposed to the FST.

Light-Dark Box

The light-dark box was used for assessing anxious behavior and consisted of two compartments: one dark, closed compartment (10 lux, black Plexiglas) and one brightly illuminated, open compartment (500 lux, white Plexiglas). Both compartments measured 25 cm3 and were connected with an opening (10 cm wide and 5 cm tall) to allow animals to freely move between them. Each individual mouse was gently placed through the opening to the dark compartment and video-recorded for 10 minutes. One pilot animal was tested prior to the representative animals on the apparatus. The apparatus was wiped clean with 10% ethanol, carefully dried with paper towels, and left open for additional 5 min thereafter between the trials. Videos were analyzed using the top-view based behavior analysis software TopScan Lite (Clever Sys Inc, Reston, Virginia). Time spent in the dark compartment was recorded as the measurements of anxiety-like behavior.

Forced Swim Test

Mice were placed in a 2 l Pyrex beaker glass containing 1.6 l of water (diameter: 12 cm, depth: 20 cm) at 27.1 ± 0.1 °C. Water in the beaker was changed between two consecutive animals. All test sessions lasted 6 minutes and were recorded by a video camera positioned on the side. Immediately after the trial, each animal was carefully dried with a paper towel. The videotapes were later scored by an observer blind to group assignment. Immobility was only scored in case of complete immobility of the animal not shorter than 2 s. Small forepaw movements against the cylinder wall, gentle body sways without changing the body posture, and brief isolated kicks with a single hindpaw preluded and followed by long immobility time (>2 s) were exempted and immobility was scored when these behaviors occurred as well. Immobility was calculated for each minute of testing separatelyand total immobility time was analyzed as a measurement for despair-like behavior. The first two minutes of the habituation to the first exposure to the FST were omitted from the analysis. In repetitive exposure to the FST, the animals performed one trial every day in approximately same time (± 30 min) for the five consecutive days. 30 minutes prior to the fifth exposure to the FST, animals were intraperitoneally injected into the lower-left quadrant of the abdomen with the 10 ml per kg of animal weight injection containing either saline, citalopram, or ketamine solution. In the experiment with 8OH-DPAT and SB269970, animals were intraperitoneally injected bilaterally into lower quadrants of the abdomen with the 10 ml/kg injections containing either saline and saline, saline and 8OH-DPAT, or SB269970 and 8OH-DPAT. Injected solutions were made fresh on the day of experiment. Dosing was determined for optimal in vivoreceptor binding based on the previous reports for citalopram (10 mg/kg)20, ketamine (10 mg/kg)21, 8OH-DPAT (0.75 mg/kg)22, and SB269970 (7.5 mg/kg)23. Citalopram, 8OH-DPAT, and SB269970 solutionswere made by dissolving 10 mg of citalopram hydrobromide (Tocris #1427, USA), 0.75 mg of 8OH-DPAT hydrobromide (Tocris #0529, USA), or 7.5 mg of SB269970 hydrochloride(Tocris #1612, USA) into 10 ml of saline; for the 8OH-DPAT solution, gentle heating (37 °C) was applied to dissolve the compound quicker. Ketamine solution was prepared by mixing0.1ml of Ketaminol vet® 100 mg/mL (Intervet, Sweden) with9.9 ml of saline.

Biochemical Analyses

Western Blot

Flash-frozen mouse HC tissue was homogenized in lysis buffer (25 mM HEPES, 300 mM NaCl pH 7.4, 2% Triton-100) containing a cocktail of cOmplete proteinase inhibitor (Roche #04693116001, Switzerland) and PhosStop phosphatase inhibitor (Roche #04906845001, Switzerland). Tissue protein concentrations were determined using 595nm spectrophotometry with Bradford reagent (Sigma #B6916, Germany) and 10µg of the protein was loaded to a gel. All samples were run in duplicates, electrophoretically separated on an SDS-PAGE gel (12.5% polyacrylamide), and transferred to 0.45µm nitrocellulose blotting membranes (GE healthcare life sciences #1060016, Germany). The membranes were blocked with 5 % bovine serum albumin (BSA) or 5% nonfat dry milk (depending on the antibody specification)in Tris-buffered saline solution containing 0.1% Tween 20 (TBST) for 1 h at room temperature (RT). Membranes were then incubated overnight at 4 °C with pERK1/2 (1:1000, #9101),ERK1/2 (1:1000, #9102), pGSK3(1:1000, #9336), or GSK3β (1:1000, #9832) primary antibodies, diluted in the blocking buffer solution. After washing in TBST, the blots were incubated with corresponding horseradish peroxidase (HRP)-conjugated secondary antibody for 1h at RT (1:2000; #7074/#7076). All antibodies were purchased fromCell Signaling Technology (Beverly, MA, United states).Following the rinse in TBST, the blots were incubated with SuperSignal West Pico Chemiluminescent Substrate(ThermoFisher scientific #34079, USA) and developed using BioRad ChemiDoc MP imaging system (BioRad, USA). Densitometry analyses were performed by ImageLab 5.2.1 software (BioRad, USA). The average signal density value from duplicates was normalized by the one of the loading control from the same membrane.

Gene Expression Analysis by qPCR

RNA was extracted from whole hippocampal tissue using the RNeasy lipid tissue Minikit (#74804,Quiagen, Germany) according to the manufacturer recommendations. RNA was quantified using NanoDrop 2000C (Thermo Scientific, Wilmington, DE, USA). The volume carrying 2µg of RNA was used for the reverse transcription with qScript (Quanta, Italy), and cDNA was stored at -20 °Cuntil further use. cDNA was assayed using Taqman Mm99999915 (Gapdh) and Mm04230607 (Bdnf)primers (Life Technologies, Carlsbad, CA, USA). Samples were run in triplicates in 96-well plates using 7500HT Real-Time PCR Thermocycler(Applied Biosystems, CA, USA). Gene expression was calculated using the 2-ΔΔCt method24 normalized to Gapdh as a reference gene.

Serotonin and 5-HIAA Assays by HPLC-MS/MS

Before dissection, the brains of WT and 2C19TG mice were harvested and briefly washed in ice-cold saline. The dissected samples were weighedand homogenized in 0.009 ml/mgbrain tissue in extraction solution containing 7 parts water adjusted to pH 2.5 with acetic acid, 1.5 parts methanol and 1.5 parts acetonitrile. Internal standards 5-HT-d4 (Toronto Research Chemicals, Ontario, Canada) and 5HIAA-d2 (CDN Isotopes, Quebec, Canada) in the above solutionwere added at 1.0 ng/mg brain tissue. Homogenate was centrifuged at 11,000 g and 25 µL of the supernatant was injected into the HPLC-MS/MS for analysis. An Agilent 1260 LC system (Agilent, USA) consisting of Agilent 1260 Quaternary pump, Agilent 1260 Infinity Standard Autosampler and temperature-controlled column compartment was connected to Agilent 6430 QQQ. The analytes were resolved on a Gemini 5 µ C18 110 Å 150 x 2 mm column (Phenomenex, USA) kept at ambient temperature during analysis. The mobile phase consisting of Solvent A (aqueous solution ofacetic acid, pH 2.5) and Solvent B (methanol) was pumped at a flow rate of 0.200 ml/minute. The following gradient elution was used: 5% B 0 - 3 minutes, 3 - 5 minutes linear increase to 70% B, maintained 5-15 minutes, 15 - 16 minutes linear decrease to 5% B, maintained 16- 25 minutes. The QQQ was equipped with an electrospray ion source and operated in positive mode for measuring 5-HT and negative mode for 5HIAA. The following transitions were monitored:177  160 for 5HT, 181  164 for 5HT-d4, 190  146 for 5HIAA and 192  148 for 5HIAA-d2. The calibration standards were prepared in the extraction solution and ranged from 5 to 1000 ng/ml. The limits of quantification were 0.125 ng for 5-HT and 5-HIAA.

Supplementary figure 1. Humanized CYP2C19 mice (2C19TG) do not show different anxiety- and despair-like behavior after chronic restrain stress.Animals were exposed to the CRS for four hours every day over a period of four weeks and then tested in the light-dark box and forced swim test respectively. (a) Animals were recorded for ten minutes in the LDB and the percentage of the time spent in light compartment was measured thereafter. There was no difference in the time spent in the light chamber between WT and 2C19TG mice (t16=0.537, p=0.599). (b) Animals were tested in the FST for six minutes and immobility was scored for the last four minutes. 2C19TG mice spent more time immobile in the FST (t16=0.302, p=0.767)

Supplementary figure 2. 5-HT1A activation lowers immobility time in the forced swim test without changing the temporal profile of activity.Animals were exposed to the FST daily for five consecutive days and intraperitoneally injected with two injections containing either saline, saline and 8OH-DPAT, or 8OH-DPAT and SB269970 30 minutes before the fifth test. Acute 8OH-DPAT treatment reduced immobility time in 2C19TG more than in WT mice (Genotype*TreatmentF1,32=5.790, p=0.038; Fisher’s LSD post hoc test p<0.001); this reduction was not significantly altered by SB269970 in either of the genotypes. 8OH-DPAT and 8OH-DPAT/SB269970 treatment did not change temporal profile of immobility in WTs and 2C19TG mice (Treatment*Time(RM)F10,120=0.857, p=0.576; Treatment*Time(RM)F10,115=0.456, p=0.914)