Supporting Information

Blankman et al. 10.1073/pnas.1217121110 SI Materials and Methods RT-PCR. Total RNA was isolated from brain tissue of 6-wk-old −/− +/+ Materials. We purchased 2-arachidonoylglycerol (2-AG) from ABHD12 and ABHD12 mice using TRIzol (Invitrogen). Cayman Chemicals; 2-oleoylglycerol, pentadecanoic acid (PDA), First-strand cDNA was synthesized using the SuperScript III and dodecylmonoalkylglycerol ether (C12:0 MAGE) were pur- Reverse Transcriptase kit (Invitrogen) according to the manu- chased from Sigma-Aldrich. Monopentadecanoin (C15:0 MAG) facturer’s protocol. PCR amplification (25 cycles) of a 259 bp fragment of the ABHD12 cDNA was performed with primers 50- and monoheptadecanoin (C17:0 MAG) were purchased from Nu- 0 0 Chek-Prep. Phospholipids and lysophospholipids were purchased CTCAGTAGGAACACCATCGGAGC-3 and 5 -GCCAGGG- AAGTATCGGTATATCACTG-3. Amplification of a 245-bp from Avanti Polar . Fluorophosphonate (FP)-rhodamine 0 (1) and JZL184 (2) were synthesized as described previously. GAPDH product was performed as a control with primers 5 - GGTGAAGGTCGGTGTGAACGG-30 and 50-CCCATTTGA- 0 Generation of ABHD12−/− Mice. The α/β-hydrolase domain-con- TGTTAGTGGGGTCTCG-3 . Both sets of primers were de- taining (ABHD)12-targeting construct was generated by ampli- signed to span a >2-kb region of genomic DNA. fying 3.4- and 4.4-kb regions of the Abhd12 gene adjacent to the Untargeted liquid chromatography–mass spectrometry proteomic analysis. catalytic exon 8 from a BAC clone containing the Abhd12 locus Mouse brain membrane proteomes were prepared from 6-mo-old − − + + (clone ID RP23-193L22 from BACPAC Resources Center at ABHD12 / and ABHD12 / mice (n = 3 per genotype). Mem- Children’s Hospital Oakland Research Institute) and subcloning brane proteins (0.375 mg of total protein) were precipitated with these homologous arms into the SacII/BamHI and XhoI/HindIII 1:4 (vol/vol) CHCl3:MeOH and denatured with 25 mM ammo- sites of the pKO-NTKV vector. The targeting construct was nium bicarbonate in 6 M urea. Samples were reduced with designed to replace Abhd12 exons 8–10 with a neomycin selec- 10 mM DTT, alkylated with 40 mM iodoacetamide, and diluted tion cassette upon homologous recombination. Following elec- to 2 M urea with 25 mM ammonium bicarbonate. Digestion with troporation of the targeting construct in C57BL/6 Bruce 4-derived trypsin (0.5 μg/μL) was performed overnight at 37 °C in the murine embryonic stem (ES) cells, 300 integrated ES cell clones presence of 1 mM CaCl2. The tryptic peptide samples were were obtained and screened for homologous recombination by acidified with 5% (vol/vol) formic acid, and aliquots were frozen Southern blot analysis with external probes located 50 and 30 of the at −80 °C until use. targeted region. These probes gave differential band sizes in WT Multidimensional protein identification technology (MudPIT) (10.4 kb) or targeted (50 probe, 4.8 kb; 30 probe, 5.7 kb) genomic analysis was performed as described previously (3) on an LTQ DNA digested with EcoRI. One homologous recombinant, 96, was Orbitrap Velos mass spectrometer (ThermoFinnigan) coupled to identified, expanded and injected into albino C57BL/6 blastocysts. an Agilent 1200 series HPLC (50 μg of protein; 11-step gradi- Blastocyst implantation into pseudopregnant albino C57BL/6 fe- ent). MS spectra were acquired in profile mode, with a mass males generated nine chimeric males. Three of the chimeras pro- range of 400–1,800 in the Orbitrap analyzer with resolution set at duced germ-line transmission of the targeted mutation, which was 30,000, followed by 30 MS/MS scans in the ion trap. Dynamic confirmed by Southern analysis. PCR genotyping of genomic tail exclusion was enabled with a repeat count of 1, a repeat duration DNA was performed using the primers 50-CAGTGCTGGCCT- of 20 s, exclusion duration 20 s, and an exclusion list size of 300. GTCAGTCG-30, 5-GGTGCCCAGTGAATGGCC-30,and50-TA- All tandem mass spectra were collected using a normalized AAGCGCATGCTCCAGACTGCC-30, which amplify a 467-bp collision energy of 35%, an isolation window of 2 Da, and an product from the WT allele and an 295-bp product from the activation time of 10 ms. One microscan was applied for all ex- targeted allele. All mice used in this study were generated from periments. Spray voltage was set to 2.50 kV, and the flow rate + − breeding ABHD12 / mice and had ad libitum access to water through the column was 0.20 μL/min. and food. RAW files were generated from the mass spectra using XCalibur version 1.4, and MS/MS spectra data extracted using Biochemical Studies. Preparation of mouse brain proteomes. Mice were RAW Xtractor (Version 1.9.1), which is publicly available anesthetized with isoflurane and killed by decapitation. Brains (http://fields.scripps.edu/?q=content/download). The tandem were harvested, sectioned into hemispheres, immediately flash MS data were searched against the mouse International Protein frozen in liquid nitrogen, and frozen at −80 °C until use. One half Index (IPI) database using the ProLuCID search algorithm (4), brain was Dounce-homogenized in PBS (pH 7.5), sonicated, and allowing for modification of M with the crosslinking agent centrifuged at slow speed (1,400 × g for 10 min at 4 °C) to re- (15.9949), static modification of cysteine residues (57.02146 Da, move debris. The supernatant was centrifuged at high speed because of alkylation), half-tryptic specificity, and a mass (100,000 × g for 45 min at 4 °C), and this supernatant was saved tolerance set to 50 ppm for precursor mass and ±0.6 Da for as the soluble proteome. The pellet was washed and resuspended product ion masses. The resulting MS/MS spectra matches were in PBS, sonicated, and saved as the membrane proteome. The assembled and filtered using DTASelect2 (Version 2.0.27). The total protein concentration of each proteome was determined validity of peptide/spectrum matches was assessed using DTA- using the Bio-Rad Dc Protein Assay kit. Aliquots of the pro- Select2 (Version 2.0.27) and two SEQUEST-defined parame- teomes were stored at −80 °C until use. ters: the cross-correlation score (XCorr) and the normalized Activity-based protein profiling analysis. Brain membrane proteomes difference in cross-correlation scores (DeltaCN). Peptides with (50 μgin50μL of PBS buffer) were prepared from 6-wk-old XCorr scores greater than 1.8 (+1), 2.5 (+2), or 3.5 (+3), and − − + − + + ABHD12 / ,ABHD12/ ,andABHD12/ littermates and pre- DeltaCN scores greater than 0.08 were included in the spectral treated with 5 μM JZL184 or DMSO vehicle for 30 min at 25 °C, counting analysis. followed by incubation with 2 μM FP-rhodamine (1) for 1 h at 25 °C. substrate hydrolysis assays. The 2-AG hydrolysis activity of − − + + Reactions were quenched with 2× SDS/PAGE loading buffer (re- brain membrane homogenates from ABHD12 / and ABHD12 / ducing), separated by SDS/PAGE [10% (wt/vol) acrylamide] and mice (2 mo old; n = 4 per genotype) plus 1 μM JZL184 or vehicle visualized in-gel with a FMBio IIe flatbed fluorescence scanner (DMSO) was measured as described previously (5). Assessing the (Hitachi). Rhodamine fluorescence is shown in gray scale. activity of brain membrane homogenates against C18:1 mono-

Blankman et al. www.pnas.org/cgi/content/short/1217121110 1of19 acylglycerol (MAG), C18:1 lysophosphatidylserine (LPS), C18:1 The fatty acid or lysophospholipid hydrolysis products were lysophosphatidylinositol (LPI), C18:1 lysophosphatidylglycerol quantified by measuring the area under the peak in comparison (LPG), C18:1 lysophosphatidylcholine (LPC), C18:1 lysophos- with the internal standard and corrected for the background phatidic acid (LPA), and C18:1/18:1 bis(monoacylglcerol)phos- product present in reactions performed with heat inactivated (10 phate (BMP) was performed similarly. Mouse brain membrane min at 90 °C) protein homogenates. Data are presented as the − − proteomes were prepared from 6-mo-old ABHD12 / and mean of three replicates ± SEM. Statistical analysis was performed + + ABHD12 / littermates (n = 3 per genotype). Activity assays by Student’s t test. were performed with 20 μg of brain membrane homogenate di- luted in PBS (100 μL total volume) incubated with 100 μM lipid Behavioral Analysis. All behavioral testing, except for the tetrad substrate. After 30 min at 25 °C, the reactions were quenched tests for cannabimimetic activity, was performed in The Scripps with 2:1 (vol/vol) CHCl3:MeOH (350 μL), and 0.5 nmol PDA Research Institute mouse behavior assessment core facility. Mice was added as an internal standard. The reaction vials were vor- were group-housed in a temperature-controlled room in which texed to mix and centrifuged for 5 min at 1,400 × g to separate the lights were on a 12-h light/dark cycle (lights off at 0600 hours), phases. A portion of the lower organic phase (30 μL) was in- and behavior was assessed during the dark (active) phase. Food jected onto an Agilent 6520 series quadrupole–time-of-flight and water were available ad libitum. − − + − + + (QTOF) MS. Chromatography was performed on a 50 × 4.60 A cohort of ABHD12 / , ABHD12 / , and ABHD12 / lit- mm 5-μm Gemini C18 column (Phenomenex) coupled to a guard termates (n = 8–9 per genotype; mixed sex) were tested for lo- column (Gemini; C18; 4 × 3.0 mm; Phenomenex SecurityGuard comotor activity in an open-field, startle responses to auditory cartridge). The LC method consisted of 0.1 mL/min of 100% stimuli, rotarod performance, and hanging wire performance at – – – = – buffer A [95:5 (vol/vol) H2O:MeOH plus 0.1% (vol/vol) am- 5 6, 11 12, and 17 18 mo of age. A second cohort (n 8 10 per monium hydroxide] for 1.5 min, 0.5 mL/min linear gradient to genotype; mixed sex) was tested at 12 and 18 mo of age. The – – 100% buffer B [65:35:5 (vol/vol) iPrOH:MeOH:H2O plus 0.1% results obtained at 11 12 and 17 18 mo old from both cohorts (vol/vol) ammonium hydroxide] over 5 min, 0.5 mL/min 100% were combined. buffer B for 5.5 min, and equilibration with 0.5 mL/min 100% Locomotor activity. Locomotor activity was measured using an buffer A for 3 min (15 min total run time). MS analysis was open-field activity monitor in a square Plexiglas arena (software, performed in negative scanning mode with an electrospray ion- hardware, and protocols from Med Associates) (7). The total ization (ESI) source. The capillary voltage was 4.0 kV, the number of line crosses were recorded during a 10-min testing fragmentor voltage was 100 V, the drying gas temperature was session. Data are presented as average values ± SEM. Statistical 350 °C, the drying gas flow rate was 11 mL/min, and the nebu- analysis was performed by ANOVA and Fisher’s protected least- lizer pressure was 45 psi. Product (oleic acid for lysophospholi- significant difference (PLSD) test to determine significance pids and C18:1 LPG for BMP) release was quantified by among genotypes for each testing session. measuring the area under the peak in comparison with the PDA Auditory startle response. Startle testing was performed using SR- internal standard and corrected for the nonenzymatically formed Lab startle chambers (San Diego Instruments) (8). Acoustic product present in heat inactivated (10 min at 90 °C) control stimuli was produced by high-frequency speakers controlled by reactions. Data are presented as the average of three biological SR-Lab software. A 25-min test session was used, in which pulse replicates, error bars represent the SEM, and statistical signifi- values were 90, 95, 100, 105, 110, 115, and 120 dB on a 70-dB cance was calculated using the Student’s t test. background (B) level. Startle pulses were 40 ms in duration and Recombinant ABHD12 activity against a panel of lipid sub- the trial types were presented several times in a pseudorandom strates was assessed by a liquid chromatography–mass spec- order with intervening no stimulus trials to control for baseline trometry (LC-MS) method essentially as described previously movement. Data are presented as the average startle magnitude (6). HEK293T cells were grown to ∼70% confluence in 10 cm measured during each acoustic pulse ± the SEM. Statistical dishes in Dulbecco’s modified Eagle medium (DMEM) supple- analysis was performed by two-way ANOVA and Fisher’s PLSD mented with L-glutamine and 10% (vol/vol) FBS at 37 °C and 5% test to determine significance between genotypes for each testing (vol/vol) CO2. Cells were transiently transfected with 6 μg of the session. In the auditory startle test, comparison between geno- murine ABHD12 cDNA in pSPORT6-CMV (OpenBioSystems) types for each dB pulse tested per session was performed by or empty vector control (“mock”) using the FUGENE 6 (Roche Student’s t test. Applied Science) transfection reagent according to the manu- Rotarod test. Motor coordination was measured using the accel- facturer’s protocols. After 48 h, the cells were washed twice with erating rotarod test with a Roto-Rod Series 8 (IITC Life Sciences) PBS, collected by scraping, resuspended in 500 μL of PBS, and (9). Mice were placed on the stationary rod, which began to lysed by sonication. The lysates were centrifuged (100,000 × g for accelerate from 10 rpm, and the rotations per minute at fall were 45 min at 4 °C), and the pellet was washed, resuspended in PBS, recorded. Mice were subjected to six testing trials per day, which sonicated, and saved as the membrane homogenate. Protein consisted of two sets of three trials each, with 1 min between concentrations were determined using the Bio-Rad Dc protein each trial and 2 h between each set. Data are presented as av- assay, and aliquots of the homogenates were stored at −80 °C erage time to fall(s) ± SEM. Statistical analysis was performed until use. Successful overexpression of active ABHD12 protein by ANOVA and Fisher’s PLSD test to determine significance was confirmed by activity-based protein profiling (ABPP) anal- among genotypes. ysis as described above. ABHD12-expressing or mock cell Hanging wire test. Peripheral strength was assessed using the membrane homogenates (5 μg) in PBS (100 μL total volume) hanging wire test (9). Mice were held so that only their forelimbs were incubated at 25 °C with the lipid substrates listed in Table contacted an elevated metal bar (12 mm diameter; 37 cm above S7. After 30 min, the reactions were quenched by the addition of the floor). Latency to fall was measured up to a maximum of 350 μL of 2:1 vol/vol CHCl3:MeOH, doped with 0.5 nmol of in- 30 s; each mouse was tested in three trials separated by a 30-s ternal standards (Table S7), vortexed, then centrifuged (1,400 × g, rest period. Data are presented as the average latency to fall(s) ± 3 min) to separate the phases. A portion of the organic phase (20 SEM. Statistical analysis was performed by ANOVA and Fisher’s μL) was injected onto an Agilent 6520 QTOF MS. LC separation PLSD test to determine significance among genotypes. and MS parameters were identical to those described above, Optomotor test. Visual acuity was assessed in the optomotor test as except that the reaction products for dioleoylglycerol (DOG) and described previously (10) using a stationary elevated platform phosphatidylcholine (PC) were measured in positive-ion mode surrounded by a drum with black-and-white striped walls. The with mobile phase buffers containing 0.1% (vol/vol) formic acid. mouse was placed on the platform to habituate for 1 min and

Blankman et al. www.pnas.org/cgi/content/short/1217121110 2of19 then the drum was rotated at 2 rpm in one direction for 1 min, All procedures were conducted in accordance with the stopped for 30 s, and then rotated in the other direction for 1 guidelines established by the Department of Agriculture and the min. The total number of head tracks (15° movements at the National Institutes of Health in the Guide for the Care and Use of speed of the drum) was recorded. Laboratory Animals and approved by the Institutional Animal Auditory brainstem response. Brainstem auditory-evoked potentials Care and Use Committee of The Scripps Research Institute. + + (BAEPs) were measured in 9- to 10-mo-old ABHD12 / , Histological and immunohistochemical analyses. Mice were deeply + − − − ABHD12 / , and ABHD12 / mice (n = 5; mixed sex) after anesthetized using isoflurane and perfused with 0.1 M phosphate − − confirming that ABHD12 / mice of this age exhibit a reduced buffer (PB), followed by 4% (wt/vol) paraformaldehyde. Perfused auditory startle reflex as described above. Mice were anes- brains were postfixed in 4% paraformaldehyde overnight, cry- thetized using isoflurane and their heads were placed in a ste- oprotected in 30% (wt/vol) sucrose in 4% paraformaldehyde until reotaxic apparatus. Head hair was shaved off, and the incision they sank (∼3 d), and rapidly frozen on dry ice. Coronal 25-μm site was prepared with ethanol and betadine. An incision was sections were cut on a Leica CM1850 cryostat. Slide-mounted + + − − made, and the skull was exposed and cleaned. Three stainless- sections from 18- to 21-mo-old ABHD12 / and ABHD12 / steel screw electrodes were inserted into the skull: two over the mice (n = 4) were stained with hematoxylin and eosin (H&E), hippocampus (2.0 mm posterior and 2.0 mm lateral to bregma) Luxol Fast Blue (LFB) with a hematoxylin counterstain or cresyl and a third over the cerebellum as a control for signal artifacts. violet (Nissl). Stained sections were imaged using a Evos XL Insulated leads from these electrodes were then soldered to Core light microscope (Advanced Microscopy Group). a miniconnector that was cemented to the skull with dental Immunohistochemistry was performed on frozen, free-floating acrylic. Wounds were sutured closed, topical antibiotic ointment cryostat sections (25 μm) as described previously (12). Primary was applied, mice were injected s.c. with flunixin, and then the antibodies were rabbit anti-ionized calcium-binding adaptor mice were recovered in a clean warm cage. Mice were monitored molecule (Iba)1 (1/500 dilution; Wako), rabbit anti-human glial continuously during surgery and until fully recovered and then at fibrillary acid protein (GFAP) (1/500 dilution; Abcam), and least once per day until the experiment’s end. After surgery, mice rabbit anti-mouse calbindin (1/200 dilution; Abcam). Sections + + − − were housed in individual cages and allowed a 1–2 wk of re- from 18- to 21-mo-old ABHD12 / and ABHD12 / mice (n = 4) covery before recordings. were immunostained for GFAP and calbindin. Sections from 6-, To measure the BAEPs, mice were anesthetized using iso- 12-, and 18-mo-old mice (n = 3–4 per genotype per age) were flurane and then fitted with bilateral polyethylene ear tubes immunostained for Iba1. Sections were incubated with primary placed into the external auditory canal. A “Y” connector attached antibodies in 0.5% (wt/vol) BSA in 0.1 M PB for 48 h at 4 °C, the two tubes to the central sound source (Grass Instruments incubated with secondary antibody (anti–rabbit-biotin; Vector Audio Amplifier). A computer program written using National Laboratories; 1/300 dilution of 1.5 mg/mL stock) in 0.5% (wt/vol) Instruments software (“LabView”) for the Macintosh II micro- BSA in 0.1 M PB for 1 h at room temperature and detected with computer was used to generate the stimuli and to collect the ABC Elite Vectastain (Vector Laboratories) for 1 h. Dia- data. The sound stimulus consisted of a 65- to 80-dB sound- minobenzidine (DAB) (peroxidase substrate kit; Vector Labo- pressure level at a rate of 10 Hz. Raw signals were amplified by ratories) was used as the chromogen. Sections were washed in a Grass P-511 instrumentation amplifier filtered between 300 and 0.1 M PB after staining and mounted in VectaMount. Im- 3,000 Hz and averaged using custom software operating on a munostained sections were imaged using a Leica SCN400 whole Macintosh II series computer. Stimuli were delivered binaurally at slide scanner. a frequency of 10 per second for a total of 1,024 stimuli. Online Quantitation of Iba1-positive, enlarged microglia (>200 μm2) averaging of the signals allowed us to repeat trials within a re- was performed in the cerebellar region depicted by the black box cording session to determine the stability of the evoked responses. in Fig. S2G in five matching cerebellar sections per mouse using − − Tetrad tests for cannabimimetic activity. Two month old ABHD12 / ImageJ software (National Institutes of Health). + + and ABHD12 / littermates (n = 5 per genotype, mixed sex) were tested for catalepsy, locomotor activity and core tempera- LC-MS Metabolite Profiling. Untargeted metabolomics analysis. Dis- ture (11). Mice were housed on a 0600 hours/1800 hours light/ covery metabolite profiling was performed as described previously − − + + dark cycle, had ad libitum access to water and food, and were (13). ABHD12 / and ABHD12 / littermates between 2 and 6 housed individually overnight before testing. Catalepsy was as- mo of age (n = 3–5 per genotype) were anesthetized with iso- sessed on a 0.7-cm-diameter bar fixed 4.5 cm off of the ground. flurane and killed by decapitation. Brains were harvested, later- The mouse was placed with its front paws on the bar, and the ally sectioned, and immediately submerged in liquid N2. One length of time the mouse was immobile on the bar was recorded, frozen brain hemisphere per mouse was weighed and immedi- if applicable. If the mouse moved off the bar, it was placed back ately Dounce-homogenized in 8 mL of 2:1:1 (vol/vol/vol) CHCl3: on in the original position. The assay was stopped after 60 s or MeOH:PBS, with 10 nmol of PDA and C12:0 MAGE added as after the third time the mouse moved off the bar. Neither internal standards for negative- and positive-mode analysis, re- + + − − ABHD12 / nor ABHD12 / mice displayed any cataleptic be- spectively. Homogenates were centrifuged for 10 min at 1,400 × havior. Locomotor activity was assessed in a clear Plexiglas cage g. The organic (lower) phase was transferred to a clean vial and (18 × 10 × 8.5 inches) placed atop a 7 cm × 7 cm grid. The dried a stream of N2. The metabolomes were resolubilized in 2:1 number of grid-lines crossed by the hind limbs in 5 min was re- vol/vol CHCl3:MeOH (120 μL), and 30 μL was injected onto an corded. Rectal temperature was measured by a thermocouple Agilent 6520 QTOF instrument. LC separation was achieved probe inserted 1.2 cm into the mouse rectum, and temperature using the same solid and mobile phases described above for was determined using a telethermometer. Antinociceptive re- substrate assays. To assist in ion formation, 0.1% (vol/vol) am- − − sponses were assessed in 2- to 3-mo-old male ABHD12 / and monium hydroxide or 0.1% (vol/vol) formic acid was added to + + ABHD12 / littermates (n = 8–9 per genotype) by the thermal the buffers for negative or positive ionization mode, respectively. tail immersion test at four temperatures (52 °C, 54 °C, 56 °C, and The LC method consisted of 0.1 mL/min 0% buffer B for 5 min, 58 °C) over a period of 4 d, with one temperature tested each day. a 0.4 mL/min linear gradient over 40 min to 100% buffer B, 0.5 Mice were housed individually for the duration of the testing. mL/min 100% buffer B for 10 min, and 0.4 mL/min equilibration Each mouse was handheld and 1 cm of the tail was submerged with 0% buffer B for 5 min, for an overall run time of 60 min. MS into a heated water bath and the latency for the mouse to with- analysis was performed with an ESI source in scanning mode drawal its tail was timed and recorded. Data were compared by from m/z = 50–1,200. The capillary voltage was set to 4.0 kV, and Student’s t test are reported as average values ± SEM. the fragmentor voltage was set to 100 V. The drying gas tem-

Blankman et al. www.pnas.org/cgi/content/short/1217121110 3of19 perature was 350 °C, the drying gas flow rate was 11 L/min, and The m/z values corresponding to the altered endogenous me- − the nebulizer pressure was 45 psi. Analysis of the LC-MS data tabolites or the parent (M-H) ions of synthetic standards were were performed by XCMS (freely available at http://metlin. targeted for MS/MS analysis, and MS1 and MS2 spectra were scripps.edu/xcms) (14), which automatically identifies, matches, acquired at a rate of 1.02 spectra per second. The collision en- aligns, and integrates chromatographic peaks corresponding to ergies for LPS was 15, for PS was 25, and for PG was 15. The endogenous metabolites and identifies m/z values that are sig- capillary voltage was set to 4.0 kV, and the fragmentor voltage was nificantly altered in user-defined control vs. experimental data- set to 100 V. The drying gas temperature was 350 °C, the drying sets. The XCMS results from two independently performed gas flow rate was 11 L/min, and the nebulizer pressure was 45 psi. experiments were compared and filtered for m/z values that ap- Targeted metabolite measurements. Brain MAG levels in − − + − + + peared in both datasets with >twofold change between genotypes, ABHD12 / , ABHD12 / , and ABHD12 / littermates (6 mo >30,000 average peak-integration area, LC elution time during the old; n = 5) were measured by multiple reaction monitoring linear gradient, and P < 0.05. Peak integrations for the m/z values (MRM) methods as described previously (5). Tissue LPS, PS, − − + − that passed these criteria were checked manually and normalized for LPG, PG, LPI, and PI levels in ABHD12 / , ABHD12 / , and − − tissue weight and area of the internal standard. Results are pre- ABHD12 / littermates (6 mo old; n = 5) were quantified using sented as the ratio of the average normalized intensities measured in similar targeted MRM methods. Tissue metabolomes were − − + + the ABHD12 / vs. ABHD12 / (KO/WT) brain metabolomes. To prepared as described above for untargeted metabolite analysis, determine the molecular identity of the altered metabolites, the except that 17:1 LPI, LPG, and LPS and 17:0/20:4 PI, PG, and experimental m/z values were searched against known metabolite PS synthetic lipids were added as internal standards. A portion masses in three online databases: Metlin (hosted by the Scripps (10 μL) of the tissue metabolome was injected onto an Agilent Center for Metabolomics, http://metlin.scripps.edu/metabo_search_ G6410B QQQ instrument. LC separation was achieved using the alt2.php), the Human Metabolome Database (www.hmdb.ca), and solid and mobile phases described above for substrate assays and the Lipid Maps Structure Database (www.lipidmaps.org/data/ the following method: 0.1 mL/min 0% buffer B for 5 min, 0.4 structure/index.html). Putative assignments were confirmed by mL/min linear gradient from 60 to 100% buffer B over 15 min, coelution with synthetic standards and/or fragmentation analysis as 0.5 mL/min 100% buffer B for 8 min, and 0.5 mL 0% buffer B described below. for 3 min (31 min total run time per sample). + + The relative abundance of known lipid species was determined Relative brain BMP levels were measured in ABHD12 / and − − − by manually extracting the mass corresponding to the (M-H) or ABHD12 / mice (2–3 mo old; n = 3) using a method similar to + (M+H) parent ion (in negative- or positive-ionization mode, one described previously (15) that allows BMP species to be respectively), integrating the area under the peak and normal- distinguished from isobaric PG species. Brain metabolomes were izing this value for the tissue weight and internal standard in- prepared as described above, and 17:0/20:4 PG was added as an tensity. Statistical significance was determined by Student’s t test. internal standard. A 10-μL quantity of the solubilized brain lipids High mass-accuracy measurements. High mass-accuracy measure- was injected onto an Agilent G6410B QQQ instrument. LC was ments of m/z = 552.3, 578.4, 580.4, and 608.4 were achieved by performed using the solid phase described above for substrate via Fourier transform ion cyclotron resonance (FT-ICR)-MS assays and an 8-min isocratic gradient of 1 mM ammonium using an Apex II 7.0T FT-ICR mass spectrometer (Bruker formate in methanol. Daltonics) coupled to an Agilent 1100 LC. LC was performed MS analysis was performed by targeted MRM with an ESI as described above for untargeted metabolomics analysis in source in negative-ion mode for (L)PS, (L)PI, and (L)PG and negative-ionization mode. The FT-IRC-MS system is equipped positive-ion mode for BMP using the precursor to product ion with a custom electrospray source with two nebulizers for dual- transitions and collision energies (CEs) listed in Table S6. The spray ionization, enabling operation in “lock-mass” mode via dwell time for each lipid was set to 100 ms. The capillary was set to the constant infusion of three compounds [3-β-hydroxy-20-ox- 4 kV, the fragmentor was set to 100 V, and the delta electron opregn-5-en-17-α-yl sulfate (C21H32O6S), m/z = 411.1847; multiplier voltage (ΔEMV) was set to –200. The drying gas bradykinin fragment 1–8 (C44H61N11O10), m/z = 902.4530; temperature was 350 °C, the drying gas flow rate was 11 l/min, − hexatyrosine (C54H56N6O13), doubly charged ion [M-2H]2 and the nebulizer pressure was 35 psi. (L)PS, (L)PI and (L)PG m/z = 497.1880] for internal calibration. Acquisition time was species were quantified by measuring the area under the peak in 0.9 s per spectrum. Using these parameters, the 213,000 re- comparison with the appropriate unnatural internal standard solving power was achieved. and normalizing for wet tissue weight. The relative levels of BMP Metabolite fragmentation analysis. MS/MS analysis was performed species were determined by measuring the area under the peak on an Agilent 6520 QTOF instrument with an ESI source, using and normalizing to the area of an internal standard and wet the same LC separation and buffers as described above for tissue weight. Data are presented as the average values ± SEM. untargeted metabolomic analysis in negative ionization mode. The Student’s t test was used to assess significance.

1. Patricelli MP, Giang DK, Stamp LM, Burbaum JJ (2001) Direct visualization of 8. Risbrough VB, Hauger RL, Roberts AL, Vale WW, Geyer MA (2004) Corticotropin- hydrolase activities in complex proteomes using fluorescent active site-directed releasing factor receptors CRF1 and CRF2 exert both additive and opposing influences probes. Proteomics 1(9):1067–1071. on defensive startle behavior. J Neurosci 24(29):6545–6552. 2. Long JZ, et al. (2009) Selective blockade of 2-arachidonoylglycerol hydrolysis produces 9. Gokhin DS, et al. (2010) Tropomodulin isoforms regulate thin filament pointed-end behavioral effects. Nat Chem Biol 5(1):37–44. capping and skeletal muscle physiology. J Cell Biol 189(1):95–109. 3. Washburn MP, Wolters D, Yates JR, 3rd (2001) Large-scale analysis of the yeast 10. Amador-Arjona A, et al. (2011) Primary cilia regulate proliferation of amplifying proteome by multidimensional protein identification technology. Nat Biotechnol progenitors in adult hippocampus: Implications for learning and memory. J Neurosci 19(3):242–247. 31(27):9933–9944. 4. Xu T, et al. (2006) ProLuCID, a fast and sensitive tandem mass spectra-based protein 11. Long JZ, et al. (2009) Dual blockade of FAAH and MAGL identifies behavioral processes identification program. Mol Cell Proteomics 5:S174. regulated by endocannabinoid crosstalk in vivo. Proc Natl Acad Sci USA 106(48):20270–20275. 5. Schlosburg JE, et al. (2010) Chronic blockade causes 12. Nomura DK, et al. (2011) Endocannabinoid hydrolysis generates brain prostaglandins functional antagonism of the . Nat Neurosci 13(9): that promote neuroinflammation. Science 334(6057):809–813. 1113–1119. 13. Saghatelian A, et al. (2004) Assignment of endogenous substrates to by 6. Blankman JL, Simon GM, Cravatt BF (2007) A comprehensive profile of brain enzymes global metabolite profiling. 43(45):14332–14339. that hydrolyze the endocannabinoid 2-arachidonoylglycerol. Chem Biol 14(12): 14. Smith CA, Want EJ, O’Maille G, Abagyan R, Siuzdak G (2006) XCMS: Processing mass 1347–1356. spectrometry data for metabolite profiling using nonlinear peak alignment, 7. Barros CS, et al. (2009) Impaired maturation of dendritic spines without matching, and identification. Anal Chem 78(3):779–787. disorganization of cortical cell layers in mice lacking NRG1/ErbB signaling in the 15. Meikle PJ, et al. (2008) Effect of lysosomal storage on bis(monoacylglycero) central nervous system. Proc Natl Acad Sci USA 106(11):4507–4512. phosphate. Biochem J 411(1):71–78.

Blankman et al. www.pnas.org/cgi/content/short/1217121110 4of19 − − − − Fig. S1. Behavioral assessment of ABHD12 / mice. (A) ABHD12 / mice (black bars) exhibited modest hypermotility in the open field test at 5–6 mo old, but not 11–12 or 17–18 mo old, compared with ABHD12+/+ (white bars) and ABHD12+/− (gray bars) littermates (n = 8–9 per group). (B) ABHD12−/− (black triangles) mice at 9–10 mo old displayed a significantly reduced auditory startle reflex compared with ABHD12+/− (gray squares) and ABHD12+/+ (white diamonds) lit- + + + − − − termate controls (n = 5 per group). (C–F) Averaged BAEP waveforms of ABHD12 / (black line), ABHD12 / (teal line), and ABHD12 / (red line) mice measured in response to 65 dB (C), 68 dB (D), 70 dB (E), and 80 dB (F) auditory stimuli (n = 5 per group; wave peaks labeled P1 to P7). (G) Visual performance in the optomotor test was not significantly altered by genetic ABHD12 disruption. Data represent the averages ± SEM. Variance among genotypes was assessed by two-way ANOVA and Fisher’s PLSD test. The Student’s t test was used to determine significance for each individual dB pulse tested in B.*P < 0.05; **P < 0.01; + + ***P < 0.001 vs. ABHD12 / mice.

Blankman et al. www.pnas.org/cgi/content/short/1217121110 5of19 + + − − + + Fig. S2. Comparative histological analysis of 18-mo-old ABHD12 / and ABHD12 / mice. No obvious genotypic differences were observed between ABHD12 / − − and ABHD12 / mice in gross brain organization by H&E staining (A and B), myelin by LFB staining (C and D), neurons by Nissl staining (E and F), astrocytes by GFAP immunostaining (G and H), or Purkinje neurons by calbindin D28k immunostaining (I and J). Representative cerebellar images are shown. Black boxes in A, C, E, G, and I represent the magnified regions in B, D, F, H, and J. (Scale bars: A, C, E, G, and I, 1 mm; B, D, and F, 250 μm; H and J, 100 μm.]

Blankman et al. www.pnas.org/cgi/content/short/1217121110 6of19 Fig. S3. ABHD12 does not regulate bulk endocannabinoid metabolism and signaling in vivo. (A) Total 2-AG hydrolase activity was equivalent in brain + + − − − − membrane homogenates from ABHD12 / (white bars) and ABHD12 / (black bars) mice. In contrast, ABHD12 / brain membrane homogenates showed re- duced 2-AG hydrolysis activity following pretreatment with the selective MAGL inhibitor JZL184 (1 μM) (n = 4 per group). (B) Total brain levels of 2-AG and + + − − other nonendocannabinoid MAGs were equivalent in brain tissue from ABHD12 / (white bars) and ABHD12 / (black bars) mice (n = 5 per group). (C–E) + + − − ABHD12 / (white bars) and ABHD12 / (black bars) mice exhibited equivalent core body temperature (C), spontaneous activity in an open field (D), and antinociceptive responses to thermal pain (E)(n = 5–9 per group). Data are presented as average values ± SEM. Statistical analysis performed by Student’s t test. ***P < 0.001 vs. ABHD12+/+ homogenates treated with JZL184.

Blankman et al. www.pnas.org/cgi/content/short/1217121110 7of19 Fig. S4. Structural assignment of ABHD12-regulated metabolites. (A–C) MS/MS fragmentation analysis of endogenous m/z = 578.3 (A), 580.4 (B), and 608.4 (C) metabolites confirmed their structures as 22:1, 22:0, and 24:1 LPS species, respectively. (D–G) Comparison of retention time (D and F) and MS/MS fragmentation spectra (E and G) for endogenous metabolites (black traces) and synthetic lipid standards (red traces) confirmed the structures of m/z = 810.5 as 18:0/20:4 PS (D and E) and m/z = 797.5 as 18:0:20:4 PG (F and G). (H and I) MS/MS fragmentation analysis confirmed the structures of m/z = 808.5 as 18:1/20:4 PS (H)andm/z = 775.5 as 18:1/18:0 PG (I).

Blankman et al. www.pnas.org/cgi/content/short/1217121110 8of19 Fig. S5. ABHD12 hydrolyzes multiple LPS species. (A) ABHD12-transfected HEK293T cell membrane homogenates (gray bars) displayed increased hydrolytic − − activity compared with mock-transfected cell membrane homogenates (black bars) toward both 16:0 and 20:4 LPS. (B) ABHD12 / brain membranes (black bars) + + displayed significantly reduced hydrolytic activity toward 16:0 and 20:4 LPS compared with ABHD12 / brain membranes (white bars). Data are presented as average values ± SEM (n = 3 per group). *P < 0.05; **P < 0.01; ***P < 0.001 vs. mock or ABHD12+/+ homogenates (Student’s t test).

Blankman et al. www.pnas.org/cgi/content/short/1217121110 9of19 Table S1. BAEP amplitudes and latencies of waves I–VII and interpeak latency of waves I–IV for ABHD12+/+, ABHD12+/−, and ABHD12−/− littermates Auditory stimuli intensity

65 dB 68 dB 70 dB 80 dB

Average SEM P vs. +/+ Average SEM P vs. +/+ Average SEM P vs. +/+ Average SEM P vs. +/+

Amplitude (υV) Wave I + + ABHD12 / 0.09 0.01 0.11 0.00 0.09 0.01 0.13 0.02 ABHD12+/− 0.09 0.01 NS 0.09 0.01 NS 0.11 0.01 NS 0.15 0.01 NS − − ABHD12 / 0.09 0.01 NS 0.09 0.01 NS 0.09 0.01 NS 0.13 0.01 NS Wave II + + ABHD12 / 0.14 0.03 0.20 0.03 0.23 0.05 0.23 0.04 ABHD12+/− 0.10 0.01 NS 0.12 0.01 NS 0.17 0.02 NS 0.27 0.03 NS − − ABHD12 / 0.10 0.01 NS 0.19 0.05 NS 0.12 0.02 NS 0.21 0.05 NS Wave III ABHD12+/+ 0.17 0.05 0.20 0.03 0.24 0.03 0.29 0.03 ABHD12+/− 0.15 0.02 NS 0.14 0.02 NS 0.19 0.03 NS 0.28 0.03 NS − − ABHD12 / 0.10 0.01 NS 0.17 0.05 NS 0.18 0.03 NS 0.28 0.07 NS Wave IV ABHD12+/+ 0.21 0.02 0.27 0.04 0.25 0.02 0.25 0.05 ABHD12+/− 0.25 0.05 NS 0.23 0.03 NS 0.21 0.05 NS 0.25 0.02 NS − − ABHD12 / 0.27 0.03 NS 0.21 0.03 NS 0.31 0.03 NS 0.25 0.05 NS Wave V ABHD12+/+ 0.18 0.04 0.25 0.06 0.20 0.07 0.22 0.06 + − ABHD12 / 0.16 0.03 NS 0.36 0.06 NS 0.19 0.04 NS 0.31 0.04 NS − − ABHD12 / 0.19 0.02 NS 0.27 0.01 NS 0.26 0.01 NS 0.22 0.05 NS Wave VI ABHD12+/+ 0.22 0.05 0.18 0.02 0.16 0.04 0.31 0.04 + − ABHD12 / 0.16 0.04 NS 0.20 0.03 NS 0.20 0.04 NS 0.29 0.04 NS − − ABHD12 / 0.15 0.03 NS 0.20 0.06 NS 0.24 0.05 NS 0.34 0.03 NS Wave VII ABHD12+/+ 0.14 0.02 0.19 0.04 0.11 0.01 0.21 0.02 + − ABHD12 / 0.12 0.01 NS 0.11 0.01 NS 0.12 0.00 NS 0.19 0.03 NS − − ABHD12 / 0.14 0.03 NS 0.12 0.01 NS 0.10 0.01 NS 0.23 0.04 NS Latency (μs) Wave I + + ABHD12 / 4.34 0.02 4.36 0.02 4.35 0.02 4.35 0.02 + − ABHD12 / 4.33 0.02 NS 4.34 0.02 NS 4.33 0.01 NS 4.34 0.01 NS ABHD12−/− 4.34 0.02 NS 4.35 0.02 NS 4.34 0.02 NS 4.34 0.01 NS Wave II + + ABHD12 / 6.09 0.02 6.09 0.02 6.08 0.02 6.09 0.03 ABHD12+/− 6.10 0.02 NS 6.10 0.02 NS 6.09 0.03 NS 6.08 0.02 NS ABHD12−/− 6.09 0.02 NS 6.08 0.02 NS 6.09 0.03 NS 6.08 0.02 NS Wave III + + ABHD12 / 6.96 0.05 6.96 0.05 6.96 0.05 6.97 0.05 ABHD12+/− 6.97 0.05 NS 6.97 0.05 NS 6.98 0.04 NS 6.98 0.04 NS ABHD12−/− 6.98 0.04 NS 6.97 0.04 NS 6.98 0.04 NS 6.99 0.02 0.0347 Wave IV + + ABHD12 / 7.75 0.04 7.75 0.04 7.76 0.05 7.79 0.10 ABHD12+/− 7.94 0.05 0.0055 7.76 0.05 NS 7.84 0.02 NS 7.81 0.02 NS − − ABHD12 / 8.02 0.03 0.0004 8.00 0.03 0.0004 8.01 0.04 0.0007 8.01 0.05 0.0145 Wave V ABHD12+/+ 8.99 0.06 8.93 0.03 9.00 0.04 8.99 0.04 ABHD12+/− 9.23 0.02 0.002 8.95 0.04 NS 9.09 0.04 NS 9.03 0.02 NS − − ABHD12 / 9.29 0.04 0.0004 9.03 0.06 NS 9.17 0.03 0.0106 9.18 0.07 NS Wave VI ABHD12+/+ 10.13 0.15 10.13 0.15 10.16 0.05 10.16 0.06 ABHD12+/− 10.39 0.07 NS 10.33 0.09 NS 10.20 0.04 NS 10.23 0.06 NS − − ABHD12 / 10.42 0.11 NS 10.37 0.12 NS 10.36 0.11 NS 10.32 0.08 NS Wave VII ABHD12+/+ 11.07 0.04 11.09 0.03 11.09 0.04 11.10 0.03 + − ABHD12 / 11.25 0.02 0.0005 11.13 0.04 NS 11.14 0.04 NS 11.10 0.02 NS − − ABHD12 / 11.25 0.02 0.0005 11.21 0.02 0.0138 11.20 0.03 0.0469 11.10 0.05 NS

Blankman et al. www.pnas.org/cgi/content/short/1217121110 10 of 19 Table S1. Cont. Auditory stimuli intensity

65 dB 68 dB 70 dB 80 dB

Average SEM P vs. +/+ Average SEM P vs. +/+ Average SEM P vs. +/+ Average SEM P vs. +/+

Wave I-IV ABHD12+/+ 3.41 0.05 3.39 0.05 3.41 0.06 3.44 0.09 + − ABHD12 / 3.61 0.05 0.0096 3.42 0.06 NS 3.51 0.03 NS 3.47 0.02 NS − − ABHD12 / 3.68 0.04 0.0013 3.65 0.03 0.0022 3.67 0.04 0.0056 3.67 0.05 0.0167

Variance among genotypes was assessed by two-way ANOVA and Fisher’s PLSD test (n = 5 per group).

Blankman et al. www.pnas.org/cgi/content/short/1217121110 11 of 19 Table S2. Untargeted lipidomic profiling of ABHD12+/+ and ABHD12−/− brain tissue Lipid m/z Retention time (min) KO/WT P

MAG 16:0 331.3 40.9 0.8 0.3175 18:1 357.3 41.4 1.0 0.8027 20:4 379.3 40.1 1.3 0.1006 PC 34:0 762.4 52.5 0.9 0.5083 36:1 788.6 52.2 1.1 0.3244 LPC 16:0 496.3 38.8 1.1 0.2767 18:0 524.4 41.4 1.4 0.0961 18:1 522.4 39.4 1.2 0.1880 20:4 544.3 37.5 1.4 0.0775 PE 34:1 716.5 44.8 0.9 0.1831 36:4 738.5 44 1.1 0.4344 38:5 764.5 44.1 1.0 0.7376 38:4 766.5 46 1.4 0.0821 LPE 16:0 452.3 32.7 0.8 0.1222 18:1 478.3 33.1 1.0 0.9764 18:0 480.3 34.9 1.1 0.1418 20:4 500.3 32 1.0 0.7974 PA 32:0 647.5 36 0.8 0.2636 34:1 673.5 36 1.1 0.7274 36:4 695.5 35.5 0.9 0.2434 36:2 699.5 36.7 0.9 0.8312 38:5 721.5 35.7 1.2 0.5368 38:4 723.5 36.8 1.3 0.1578 LPA 16:0 409.2 24.9 0.8 0.2668 18:1 435.3 25.1 1.0 0.9734 18:0 437.3 27.3 1.0 0.9026 20:4 457.2 24.2 0.9 0.5147 Unassigned 718.6 41.2 3.4 0.0002 − 814.6 44.8 5.1 1.8 × e 6 PS 34:1 760.5 38.591 0.6 0.0094 36:2 786.5 38.7 0.6 0.0001 36:1 788.5 40 0.8 0.1156 38:5 808.5 38 2.9 0.0209 38:4 810.5 39.5 3.5 1.7 × e−06 LPS 16:0 496.3 29.2 1.5 0.0498 18:1 522.3 29.5 1.0 0.9370 18:0 524.3 30.8 3.1 0.0001 20:4 544.3 28.5 3.5 0.0014 20:0 552.3 31.3 7.7 0.0068 22:6 568.3 28.7 1.0 0.8429 22:1 578.3 31.3 12.0 0.0010 22:0 580.4 33 38.6 0.0004 24:0 608.4 34.7 5.9 0.0003 PG 32:0 721.5 42.3 0.5 0.0010 34:1 747.5 42.4 0.9 0.3335 34:0 749.5 43.2 1.0 0.5127 − 36:1 775.5 41.9 3.8 7.5 × e 7 − 36:4 797.5 41.6 3.0 7.8 × e 6 LPG 20:4 531.3 31.3 1.1 0.7069 LPI 16:0 571.3 31.8 0.6 0.0368

Blankman et al. www.pnas.org/cgi/content/short/1217121110 12 of 19 Table S2. Cont. Lipid m/z Retention time (min) KO/WT P

18:1 597.3 32 0.9 0.6734 18:0 599.3 33.7 1.1 0.6687 20:4 619.3 29.1 2.3 0.0123 22:6 643.3 31.3 0.9 0.7659 FFA 16:0 255.2 30 1.1 0.5423 18:1 281.2 30.4 1.0 0.9486 18:0 283.3 31.9 1.1 0.2305 20:4 303.2 29.5 1.1 0.4411 22:6 327.2 29.7 1.0 0.8323 22:0 339.3 35.5 1.2 0.5558 22:0 339.3 35.5 1.2 0.5558 24:0 367.4 37.3 1.4 0.4188 26:0 395.4 39.4 1.3 0.4665

LC-MS profiling of brain tissue from 2- to 6-mo-old ABHD12+/+ and ABHD12−/− mice revealed multiple lipid − − species that were altered in ABHD12 / brains. Bolded entries were identified as changing by the XCMS algo- rithm (14) and confirmed by manual integration. Other significantly changing metabolites were identified and quantified by manual extraction and integration, respectively. Note that XCMS identified two additional chang- ing metabolites that we could not structurally assign based on the acquired information (m/z = 718.6 and 814.6). MAG, PC, and LPC species were measured in positive ionization mode, and m/z values represent [M+H]+ ions. All other lipids were measured in negative ionization mode, and m/z values represent [M-H]− ions. Relative me- − − + + tabolite abundance in ABHD12 / vs. ABHD12 / brain metabolomes (KO/WT) is presented as a ratio of the average integrated peak intensities (n = 3–5 per genotype). Statistical analysis was performed by the Student’s t test. MAG, monoacylglycerol; PC, phosphatidylcholine; LPC, lysophosphatidylcholine; PE,; LPE, lysophosphatidylethanolamine; PA, phosphatidic acid; LPA, lysophosphatidic acid; PS, phosphatidylserine; LPS, lysophosphatidylserine; PG, phosphatidylglycerol; LPG, lysophosphatidylglycerol; LPI, lysophosphatidylinositol; FFA, free fatty acid.

Table S3. Structures and analytical data for representative VLC-LPS lipids elevated in brain tissue from ABHD12−/− mice

Observed Molecular Theoretical Mean error Molecular assignment and structure Fold elevation in m/z formula m/z (ppm) ABHD12−/− 552.3307 C26H51NO9P− 552.3307 −0.1 20:0 LPS 12.2

578.3471 C28H53NO9P− 578.3463 0.8 22:1 LPS 15.2

580.3619 C28H55NO9P− 580.362 −0.1 22:0 LPS 13.5

608.3941 C30H59NO9P− 608.3933 −1.1 24:0 LPS 15.2

Blankman et al. www.pnas.org/cgi/content/short/1217121110 13 of 19 Table S4. Targeted brain lipid measurements ABHD12+/+ (WT) ABHD12+/− (HET) ABHD12−/− (KO)

Average SEM Average SEM HET/WT P vs. WT Average SEM KO/WT P vs. WT

LPS 16:0 0.071 0.006 0.081 0.008 1.133 NS 0.154 0.010 2.165 0.0001 18:1 1.577 0.058 1.696 0.048 1.075 NS 1.894 0.057 1.201 0.0047 18:0 2.649 0.204 2.875 0.189 1.085 NS 10.034 1.212 3.788 0.0003 20:4 0.281 0.038 0.363 0.055 1.294 NS 0.550 0.054 1.958 0.0036 20:1 0.099 0.013 0.121 0.011 1.222 NS 0.309 0.067 3.113 0.0151 20:0 0.039 0.002 0.043 0.005 1.100 NS 0.477 0.058 12.210 0.0001 22:6 4.153 0.563 4.652 0.406 1.120 NS 4.402 1.272 1.060 NS − 22:4 0.223 0.022 0.253 0.019 1.139 NS 0.552 0.031 2.479 2.4 × e 5 22:1 0.033 0.004 0.037 0.006 1.138 NS 0.496 0.055 15.258 3.2 × e−5 22:0 0.025 0.004 0.030 0.008 1.199 NS 0.334 0.037 13.452 3.7 × e−5 − 24:1 0.018 0.004 0.021 0.002 1.146 NS 0.461 0.021 25.444 2.6 × e 8 − 24:0 0.038 0.008 0.082 0.029 2.149 NS 0.584 0.033 15.242 2.3 × e 7 PS 16:0/18:1 94.773 10.303 79.906 5.424 0.843 NS 31.328 1.792 0.331 0.0002 18:1/18:1 505.103 36.916 488.154 19.900 0.966 NS 271.902 5.082 0.538 0.0002 18:0/18:1 365.492 106.201 375.063 47.663 1.026 NS 314.216 48.558 0.860 NS 18:1/20:4 31.150 2.177 39.719 4.028 1.275 NS 52.483 4.192 1.685 0.0020 18:0/20:4 188.760 36.515 202.917 24.656 1.075 NS 493.122 69.991 2.612 0.0093 18:1/20:1 8.409 0.220 7.465 1.628 0.888 NS 3.766 0.567 0.448 0.0004 18:0/20:1 44.418 0.952 44.874 10.071 1.010 NS 20.386 3.254 0.459 0.0005 18:1/22:6 59.357 3.451 57.741 4.828 0.973 NS 42.689 2.782 0.719 0.0164 18:0/22:6 1,970.196 97.447 1,810.615 131.369 0.919 NS 1,656.428 135.181 0.841 NS 20:1/20:4 0.721 0.079 1.063 0.148 1.476 NS 3.225 0.387 4.475 0.0007 20:0/20:4 0.854 0.110 0.579 0.123 0.678 NS 2.542 0.341 2.977 0.0030 20:1/22:6 5.381 0.216 4.913 0.586 0.913 NS 4.608 0.469 0.856 NS 20:0/22:6 1.986 0.178 1.513 0.394 0.762 NS 1.530 0.142 0.770 NS 22:1/20:4 0.255 0.031 0.225 0.043 0.882 NS 2.315 0.281 9.078 0.0003 22:0/20:4 0.510 0.071 0.361 0.072 0.708 NS 2.562 0.295 5.022 0.0005 24:1/20:4 0.149 0.026 0.097 0.018 0.648 NS 1.284 0.143 8.611 0.0003 24:0/20:4 0.055 0.019 0.069 0.007 1.247 NS 0.299 0.076 5.432 0.0149 LPI 16:0 2.476 0.311 1.925 0.218 0.778 NS 1.503 0.178 0.607 0.0265 18:1 2.885 0.416 3.897 0.391 1.351 NS 2.886 0.390 1.001 NS 18:0 10.683 1.749 11.290 0.762 1.057 NS 11.612 2.426 1.087 NS 20:4 3.802 0.517 4.844 0.436 1.274 NS 7.144 1.141 1.879 0.0285 PI 18:1/20:4 35.863 3.193 38.733 3.370 1.080 NS 32.825 2.977 0.915 NS 18:0/20:4 228.008 26.338 266.172 28.842 1.167 NS 315.033 24.808 1.382 0.0428 LPG 16:0 0.499 0.036 0.520 0.040 1.044 NS 0.440 0.029 0.882 NS 18:1 1.987 0.152 2.230 0.124 1.122 NS 2.027 0.124 1.020 NS 18:0 0.269 0.033 0.273 0.047 1.017 NS 0.440 0.027 1.638 0.0040 20:4 0.571 0.047 0.649 0.051 1.137 NS 0.591 0.074 1.035 NS 22:6 0.348 0.033 0.337 0.034 0.968 NS 0.221 0.025 0.636 0.0166 PG 16:0/16:0 18.220 2.074 20.034 2.241 1.100 NS 12.211 1.519 0.670 0.0476 16:0/18:1 60.383 7.054 66.335 6.086 1.099 NS 64.905 8.338 1.075 NS 18:0/18:1 4.085 0.287 6.047 0.852 1.480 0.043 19.192 2.566 4.698 0.0004 18:0/20:4 7.604 0.893 10.831 1.140 1.424 0.045 23.034 3.730 3.029 0.0038 BMP 16:0/16:0 10.779 0.669 12.528 2.943 1.162 NS 16:0/18:1 12.329 1.129 13.781 3.349 1.118 NS 18:1/18:1 3.087 0.037 3.525 0.512 1.142 NS 20:4/22:6 2.503 0.349 2.182 0.421 0.872 NS 22:6/22:6 25.302 3.456 23.309 3.821 0.921 NS

MRM MS methods were used to quantify (L)PS, (L)PI, and (L)PG levels in ABHD12+/+ (WT), ABHD12+/− (HET), and ABHD12−/− + + − − (KO) brains and measure relative BMP levels in ABHD12 / and ABHD12 / brains. (L)PS, (L)PI, and (L)PG levels are presented as nanomoles per gram of tissue, and relative BMP levels are presented as arbitrary units normalized for tissue weight. Statistical analysis was performed by the Student’s t test (n = 3–5 per group). HET, heterozygous; NS, not significant.

Blankman et al. www.pnas.org/cgi/content/short/1217121110 14 of 19 Table S5. MRM transition list for targeted metabolite measurements Class Species Precursor Product CE (V)

PS 17:0/20:4 std 797 269 35 PS 16:0/18:1 761 255 35 PS 18:1/18:1 787 281 35 PS 18:0/18:1 789 283 35 PS 18:1/20:4 809 281 35 PS 18:0/20:4 811 283 35 PS 18:1/20:1 815 281 35 PS 18:0/20:1 817 283 35 PS 18:1/22:6 833 281 35 PS 18:0/22:6 835 283 35 PS 20:1/20:4 837 309 35 PS 20:0/20:4 839 311 35 PS 20:1/22:6 861 309 35 PS 20:0/22:6 863 311 35 PS 22:1/20:4 865 337 35 PS 22:0/20:4 867 339 35 PS 24:1/20:4 893 365 35 PS 24:0/20:4 895 367 35 LPS 17:1 std 508 421 15 LPS 16:0 496 409 15 LPS 18:1 522 435 15 LPS 18:0 524 437 15 LPS 20:4 544 457 15 LPS 20:1 550 463 15 LPS 20:0 552 465 15 LPS 22:4 572 485 15 LPS 22:6 568 481 15 LPS 22:1 578 491 15 LPS 22:0 580 493 15 LPS 24:1 606 519 15 LPS 24:0 608 521 15 PG 17:0/20:4 std 784 303 40 PG 16:0/16:0 722 255 40 PG 16:0/18:1 748 281 40 PG 18:0/18:1 776 281 40 PG 18:0/20:4 798 303 40 LPG 17:1 std 496 267 20 LPG 16:0 483 255 20 LPG 18:1 509 281 20 LPG 18:0 511 283 20 LPG 20:4 531 303 20 LPG 22:6 555 327 20 PI 17:0/20:4 std 872 303 40 PI 18:1/20:4 884 303 40 PI 18:0/20:4 886 303 40 LPI 17:1 std 583 267 40 LPI 16:0 571 255 40 LPI 18:1 597 281 40 LPI 18:0 599 283 40 LPI 20:4 619 303 40 PG 17:0/20:4 std 803 614 15 BMP 16:0/16:0 741 313 20 BMP 16:0/18:1 767 339 20 BMP 18:1/18:1 793 339 20 BMP 20:4/22:6 861 361 20 BMP 20:4/22:6 885 385 20

Listed are the precursor and product ion m/z values and collision energies (CEs) used to quantify lipid + + + − − − metabolite levels in ABHD12 / , ABHD12 / , and ABHD12 / tissues. std, standard.

Blankman et al. www.pnas.org/cgi/content/short/1217121110 15 of 19 Table S6. Targeted lipid measurements in dissected brain regions and peripheral tissues ABHD12+/+ (WT) ABHD12+/− (HET) ABHD12−/− (KO)

Average SEM Average SEM HET/WT P vs. WT Average SEM KO/WT P vs. WT

Cerebellum LPS 16:0 0.111 0.006 0.152 0.011 1.374 0.0189 0.196 0.012 1.766 0.0008 18:1 2.351 0.087 2.890 0.120 1.230 NS 2.730 0.273 1.161 NS − 18:0 11.246 0.677 9.083 1.180 0.808 NS 26.948 0.644 2.396 2.8 × e 6 − 20:4 0.177 0.016 0.197 0.017 1.116 NS 0.438 0.019 2.471 4.3 × e 5 20:0 0.064 0.010 0.085 0.003 1.330 NS 0.691 0.038 10.838 3.5 × e−6 22:6 2.626 0.419 2.757 0.587 1.050 NS 2.669 0.313 1.016 NS − 22:1 0.065 0.005 0.046 0.004 0.704 NS 1.091 0.054 16.734 1.4 × e 6 22:0 0.131 0.011 0.116 0.014 0.888 NS 2.089 0.107 16.003 1.8 × e−6 24:1 0.061 0.011 0.049 0.007 0.811 NS 1.500 0.076 24.650 1.4 × e−6 − 24:0 0.137 0.019 0.111 0.006 0.813 NS 3.318 0.105 24.285 9.4 × e 8 PS 16:0/18:1 61.244 3.915 74.716 13.522 1.220 NS 23.781 0.282 0.388 0.0001 16:0/18:0 71.590 13.549 64.999 15.304 0.908 NS 37.590 2.279 0.525 0.0482 18:1/18:1 517.582 10.009 607.106 78.046 1.173 NS 274.830 22.587 0.531 0.0001 18:0/18:1 336.359 75.806 377.764 55.387 1.123 NS 204.420 21.464 0.608 NS 18:1/20:4 28.916 1.262 43.833 5.902 1.516 0.0484 55.100 2.130 1.906 4.2 × e−5 18:0/20:4 88.293 7.101 124.897 17.039 1.415 NS 188.531 13.129 2.135 0.0005 18:1/22:6 43.105 1.575 51.536 6.458 1.196 NS 25.509 1.088 0.592 0.0001 18:0/22:6 897.481 76.118 1,123.395 160.488 1.252 NS 561.455 51.255 0.626 0.0106 20:1/20:4 0.620 0.118 0.891 0.120 1.437 NS 2.002 0.147 3.229 0.0003 Cortex LPS 16:0 0.146 0.025 0.196 0.033 1.343 NS 0.219 0.016 1.500 0.0498 18:1 1.102 0.120 1.318 0.172 1.195 NS 1.233 0.049 1.119 NS 18:0 12.166 3.685 14.396 5.394 1.183 NS 14.019 1.947 1.152 NS − 20:4 0.120 0.008 0.160 0.016 1.332 NS 0.344 0.010 2.863 2.4 × e 6 20:0 0.024 0.001 0.023 0.002 0.987 NS 0.177 0.013 7.460 2.6 × e−5 22:6 2.969 0.333 2.530 0.259 0.852 NS 2.814 0.196 0.948 NS − 22:1 0.021 0.004 0.017 0.002 0.815 NS 0.214 0.015 10.403 1.7 × e 5 22:0 0.027 0.004 0.040 0.002 1.485 0.0176 0.369 0.054 13.668 0.0007 24:1 0.010 0.000 0.017 0.001 1.778 0.0002 0.221 0.028 22.697 0.0003 24:0 0.044 0.002 0.039 0.008 0.900 NS 0.507 0.070 11.596 0.0006 PS 16:0/18:1 46.771 1.053 46.129 1.190 0.986 NS 21.486 1.221 0.459 4.2 × e−6 16:0/18:0 68.203 5.768 58.983 5.509 0.865 NS 41.079 5.928 0.602 0.0168 18:1/18:1 158.761 10.077 138.931 8.376 0.875 NS 86.792 6.314 0.547 0.0009 18:0/18:1 102.713 6.766 111.167 11.961 1.082 NS 61.146 5.068 0.595 0.0027 18:1/20:4 11.169 0.408 14.228 0.697 1.274 0.0091 18.559 0.994 1.662 0.0005 18:0/20:4 69.664 1.934 85.710 5.362 1.230 0.0305 138.129 9.346 1.983 0.0004 18:1/22:6 41.326 1.822 38.363 1.342 0.928 NS 26.065 1.418 0.631 0.0006 18:0/22:6 1,199.876 42.251 1,120.403 58.824 0.934 NS 772.844 50.032 0.644 0.0006 20:1/20:4 0.261 0.005 0.318 0.031 1.218 NS 0.855 0.055 3.279 3.8 × e−5 Hippocampus LPS 16:0 0.470 0.060 0.349 0.044 0.743 NS 0.358 0.052 0.762 NS 18:1 3.378 0.961 2.203 0.193 0.652 NS 2.749 0.259 0.814 NS 18:0 37.024 4.127 21.639 7.115 0.584 NS 34.106 6.171 0.921 NS 20:4 0.567 0.151 0.504 0.054 0.888 NS 1.051 0.165 1.854 NS 20:0 0.080 0.009 0.052 0.004 0.652 NS 0.257 0.064 3.209 0.0330 22:6 8.515 2.012 6.006 0.513 0.705 NS 6.287 0.935 0.738 NS 22:1 0.090 0.019 0.043 0.009 0.478 NS 0.336 0.090 3.747 0.0370 − 22:0 0.037 0.003 0.086 0.003 2.348 1.9 × e 5 0.272 0.083 7.443 0.0300 24:1 0.015 0.001 0.043 0.004 2.985 0.0003 0.205 0.061 14.090 0.0206 24:0 0.052 0.002 0.083 0.007 1.579 0.0055 0.514 0.136 9.816 0.0145 PS − 16:0/18:1 73.825 1.721 85.593 6.777 1.159 NS 33.670 1.409 0.456 1.9 × e 6 16:0/18:0 99.131 1.419 89.708 10.206 0.905 NS 82.133 10.680 0.829 NS 18:1/18:1 182.607 7.921 192.406 4.236 1.054 NS 128.316 17.495 0.703 0.0301 18:0/18:1 183.719 17.697 247.958 26.508 1.350 NS 200.222 47.229 1.090 NS

Blankman et al. www.pnas.org/cgi/content/short/1217121110 16 of 19 Table S6. Cont. ABHD12+/+ (WT) ABHD12+/− (HET) ABHD12−/− (KO)

Average SEM Average SEM HET/WT P vs. WT Average SEM KO/WT P vs. WT

18:1/20:4 19.185 0.540 27.689 1.266 1.443 0.0008 35.947 2.996 1.874 0.0015 18:0/20:4 127.128 5.757 182.942 5.424 1.439 0.0004 331.087 39.071 2.604 0.0021 18:1/22:6 49.644 0.622 58.645 3.654 1.181 NS 35.898 1.769 0.723 0.0003 18:0/22:6 1,586.937 48.575 1,741.337 48.002 1.097 NS 1,373.584 129.562 0.866 NS 20:1/20:4 0.422 0.030 0.636 0.038 1.507 0.0045 2.319 0.470 5.496 0.0069 Heart LPS 18:1 1.843 0.068 1.651 0.170 0.896 NS 1.924 0.252 1.044 NS 18:0 77.768 7.491 71.598 4.899 0.921 NS 65.145 5.913 0.838 NS 20:4 0.506 0.036 0.619 0.107 1.225 NS 0.807 0.150 1.596 NS 20:0 0.875 0.115 0.712 0.048 0.813 NS 0.664 0.047 0.758 NS 22:6 0.966 0.160 0.846 0.355 0.876 NS 1.674 0.230 1.732 0.036 22:1 0.142 0.009 0.131 0.010 0.922 NS 0.132 0.014 0.930 NS 22:0 0.121 0.017 0.099 0.009 0.821 NS 0.107 0.011 0.889 NS 24:1 0.033 0.005 0.031 0.003 0.934 NS 0.031 0.003 0.924 NS 24:0 0.057 0.011 0.040 0.004 0.699 NS 0.061 0.011 1.073 NS PS 16:0/18:1 1.426 0.231 0.938 0.095 0.658 NS 1.173 0.267 0.822 NS 18:1/18:1 3.334 0.116 2.388 0.145 0.716 NS 2.153 0.533 0.646 NS 18:0/18:1 50.190 4.172 35.505 4.089 0.707 NS 37.654 9.228 0.750 NS 18:1/20:4 5.453 0.732 3.939 0.579 0.722 NS 5.839 1.182 1.071 NS 18:0/20:4 77.589 4.351 70.076 5.054 0.903 NS 93.848 17.665 1.210 NS 18:1/22:6 7.199 0.460 6.443 0.564 0.895 NS 5.773 0.970 0.802 NS 18:0/22:6 260.491 27.944 293.383 25.615 1.126 NS 242.349 57.429 0.930 NS 20:1/20:4 0.965 0.056 0.844 0.073 0.875 NS 1.068 0.144 1.106 NS Kidney LPS 16:0 1.965 0.377 1.340 0.097 0.682 NS 1.961 0.765 0.998 NS 18:1 1.406 0.371 1.157 0.069 0.823 NS 0.998 0.228 0.710 NS 18:0 15.251 1.737 12.729 0.707 0.835 NS 16.781 4.831 1.100 NS 20:4 5.362 1.221 4.399 0.401 0.820 NS 4.095 0.398 0.764 NS 20:0 0.085 0.012 0.071 0.006 0.836 NS 0.090 0.014 1.059 NS 22:6 1.250 0.208 0.854 0.052 0.683 NS 0.818 0.091 0.654 NS 22:1 0.025 0.002 0.026 0.006 1.043 NS 0.038 0.009 1.493 NS 22:0 0.051 0.006 0.059 0.005 1.154 NS 0.090 0.012 1.746 0.024 24:1 0.019 0.003 0.018 0.002 0.946 NS 0.027 0.004 1.405 NS 24:0 0.057 0.006 0.058 0.008 1.005 NS 0.097 0.011 1.697 0.012 PS 16:0/18:1 3.958 0.331 4.133 0.282 1.044 NS 2.980 0.329 0.753 NS 18:1/18:1 1.517 0.094 1.676 0.119 1.105 NS 1.192 0.118 0.786 NS 18:0/18:1 14.476 2.034 16.704 2.029 1.154 NS 12.412 2.033 0.857 NS 18:1/20:4 2.749 0.194 3.048 0.123 1.109 NS 2.434 0.231 0.885 NS 18:0/20:4 118.126 14.986 135.112 13.935 1.144 NS 111.654 8.472 0.945 NS 18:1/22:6 1.173 0.096 1.055 0.047 0.900 NS 0.805 0.121 0.686 0.035 18:0/22:6 21.487 2.435 18.406 1.478 0.857 NS 14.373 1.855 0.669 0.042 20:1/20:4 0.181 0.021 0.222 0.020 1.229 NS 0.223 0.020 1.236 NS Liver LPS 16:0 3.840 0.886 2.624 0.340 0.683 NS 4.064 0.816 1.058 NS 18:1 1.169 0.231 0.696 0.032 0.595 NS 0.863 0.130 0.739 NS 18:0 18.958 2.622 18.402 2.639 0.971 NS 24.338 3.568 1.284 NS 20:4 2.939 0.333 2.863 0.261 0.974 NS 4.194 0.482 1.427 NS 20:0 0.150 0.041 0.085 0.024 0.565 NS 0.206 0.036 1.371 NS 22:6 2.003 0.203 1.596 0.206 0.796 NS 2.117 0.336 1.057 NS 22:1 0.030 0.003 0.034 0.010 1.130 NS 0.048 0.010 1.572 NS 22:0 ND ND ND 24:1 ND ND ND 24:0 ND ND ND PS 16:0/18:1 0.907 0.113 0.918 0.060 1.012 NS 0.751 0.023 0.827 NS 18:1/18:1 1.315 0.172 1.283 0.106 0.976 NS 1.023 0.039 0.778 NS

Blankman et al. www.pnas.org/cgi/content/short/1217121110 17 of 19 Table S6. Cont. ABHD12+/+ (WT) ABHD12+/− (HET) ABHD12−/− (KO)

Average SEM Average SEM HET/WT P vs. WT Average SEM KO/WT P vs. WT

18:0/18:1 5.904 0.716 6.172 1.030 1.045 NS 6.184 0.438 1.047 NS 18:1/20:4 8.001 1.149 8.339 1.372 1.042 NS 7.083 0.519 0.885 NS 18:0/20:4 157.452 17.089 171.544 14.002 1.090 NS 219.257 10.849 1.393 0.016 18:1/22:6 4.448 0.649 3.795 0.771 0.853 NS 3.190 0.251 0.717 NS 18:0/22:6 74.678 4.668 70.435 9.342 0.943 NS 73.659 6.121 0.986 NS 20:1/20:4 0.508 0.089 0.558 0.184 1.100 NS 0.930 0.146 1.832 0.038 Lung LPS 16:0 3.326 0.516 2.700 0.275 0.812 NS 3.609 1.504 1.085 NS 18:1 7.763 1.366 6.470 0.303 0.833 NS 7.473 2.517 0.963 NS 18:0 81.408 12.081 67.877 8.304 0.834 NS 77.148 32.753 0.948 NS 20:4 2.840 0.372 3.117 0.382 1.098 NS 2.963 0.455 1.043 NS 20:0 0.618 0.104 0.503 0.061 0.814 NS 0.700 0.272 1.134 NS 22:6 1.242 0.200 1.093 0.168 0.880 NS 1.036 0.176 0.834 NS 22:1 1.002 0.193 0.891 0.069 0.889 NS 0.740 0.171 0.738 NS 22:0 0.549 0.071 0.467 0.036 0.850 NS 0.500 0.081 0.910 NS 24:1 0.325 0.059 0.301 0.025 0.928 NS 0.287 0.054 0.883 NS 24:0 0.509 0.043 0.525 0.046 1.033 NS 0.629 0.079 1.237 NS PS 16:0/18:1 14.454 2.480 10.744 0.872 0.743 NS 10.380 3.253 0.718 NS 18:1/18:1 14.468 1.457 12.083 0.742 0.835 NS 10.888 3.216 0.753 NS 18:0/18:1 130.766 23.361 116.473 12.442 0.891 NS 112.326 36.066 0.859 NS 18:1/20:4 6.430 0.372 7.590 0.424 1.180 NS 7.240 0.654 1.126 NS 18:0/20:4 143.571 22.610 146.537 11.278 1.021 NS 143.625 26.044 1.000 NS 18:1/22:6 3.205 0.470 3.861 0.481 1.205 NS 2.800 0.520 0.874 NS 18:0/22:6 61.919 7.327 60.503 2.458 0.977 NS 37.238 4.754 0.601 0.022 20:1/20:4 0.957 0.152 1.029 0.081 1.076 NS 1.133 0.203 1.185 NS Eye LPS 16:0 0.655 0.281 0.550 0.166 0.840 NS 0.438 0.152 0.669 NS 18:1 3.034 0.373 3.797 0.832 1.251 NS 2.508 0.757 0.827 NS 18:0 48.901 7.355 42.534 11.010 0.870 NS 35.497 5.701 0.726 NS 20:4 1.175 0.371 1.534 0.466 1.306 NS 1.248 0.438 1.062 NS 20:0 0.188 0.045 0.159 0.043 0.848 NS 0.172 0.050 0.918 NS 22:6 8.545 2.106 9.648 4.958 1.129 NS 6.302 1.026 0.738 NS 22:1 0.179 0.044 0.152 0.030 0.850 NS 0.136 0.048 0.761 NS 22:0 0.133 0.030 0.112 0.020 0.845 NS 0.122 0.038 0.919 NS 24:1 0.122 0.015 0.095 0.013 0.782 NS 0.104 0.026 0.855 NS 24:0 ND ND ND PS 16:0/18:1 22.186 0.335 19.174 1.223 0.864 NS 14.136 0.493 0.637 1.5 × e−5 16:0/18:0 216.782 29.567 187.488 27.274 0.865 NS 253.100 17.644 1.168 NS 18:1/18:1 105.734 3.641 100.532 1.602 0.951 NS 97.691 3.926 0.924 NS 18:0/18:1 108.142 4.179 102.759 2.817 0.950 NS 99.437 6.479 0.920 NS 18:1/20:4 9.985 0.923 10.154 1.034 1.017 NS 10.499 1.239 1.051 NS 18:0/20:4 111.826 4.729 128.021 2.501 1.145 NS 137.912 9.969 1.233 NS 18:1/22:6 14.354 0.555 13.457 0.243 0.938 NS 13.195 0.744 0.919 NS 18:0/22:6 257.629 14.726 258.003 3.144 1.001 NS 232.762 25.471 0.903 NS 20:1/20:4 0.482 0.044 0.530 0.042 1.100 NS 0.636 0.060 1.319 NS

MRM MS methods were used to quantify LPS and PS levels in several brain regions (cerebellum, cortex, and hippocampus) and + + + − − − peripheral tissues (heart, kidney, liver, lung, and eye) from ABHD12 / (WT), ABHD12 / (HET), and ABHD12 / (KO) mice. Lipid levels are presented as nanomoles per gram of tissue. Statistical analysis was performed by the Student’s t test (n = 5 per group). HET, heterozygous; ND, not detected; NS, not significant.

Blankman et al. www.pnas.org/cgi/content/short/1217121110 18 of 19 Table S7. Recombinant ABHD12 substrate assay conditions Lipid substrate Substrate, μM Product Internal standard

18:1 MAG 100 18:1 FFA (OA) 15:0 FFA (PDA) 18:1 LPA 100 18:1 FFA (OA) 15:0 FFA (PDA) 18:1 LPC 100 18:1 FFA (OA) 15:0 FFA (PDA) 18:1 LPE 100 18:1 FFA (OA) 15:0 FFA (PDA) 18:1 LPG 100 18:1 FFA (OA) 15:0 FFA (PDA) 18:1 LPI 100 18:1 FFA (OA) 15:0 FFA (PDA) 18:1 LPS 100 18:1 FFA (OA) 15:0 FFA (PDA) 18:1/18:1 BMP 100 18:1 LPG 17:1 LPG 18:1/18:1 DOG 100 18:1 MAG 15:0 MAG 16:0/18:1 PA 100 18:1 PA 17:0 LPA 16:0/18:1 PC 100 16:0 PC 15:0 LPC 16:0/18:1 PE 100 18:1 PE 17:0 LPE 16:0/18:1 PG 100 18:1 PG 17:1 LPG 16:0/18:1 PI 50 18:1 PI 17:1 LPI 16:0/18:1 PS 100 18:1 PS 17:1 LPS

Listed are the substrate concentration, product detected, and internal standard for each lipid substrate tested. MAG, monoacylglycerol; FFA, free fatty acid; OA, oleic acid; PDA, pentadecanoic acid; LPA, lysophosphatidic acid; LPC, lysophosphatidylcholine; LPE, lysophosphatidylethanolamine; LPI, lysophosphatidylinositol; LPS, lysophosphatidylserine; BMP, bismonoacylgly- cerolphosphate; DOG, dioleoylglycerol; PA, phosphatidic acid; PC, phospha- tidylcholine; PE, phosphatidylethanolamine; PG, phosphatidylglycerol; PI, phosphatidylinositol; PS, phosphatidylserine.

Blankman et al. www.pnas.org/cgi/content/short/1217121110 19 of 19