|Home | About | Journals | Submit | Contact Us | Français|
Endocannabinoids are lipid-signaling molecules that bind cannabinoid receptors. These G-protein-coupled receptors were discovered as binding sites for Δ9-tetrahydrocannabinol (THC), the psychoactive component of marijuana (Cannabis sativa) (1). Two types of cannabinoid receptors are known: cannabinoid receptor 1 (CB1) and cannabinoid receptor 2 (CB2) (2,3). CB1 is expressed in the central and peripheral nervous system. CB2 is expressed in immune cells such as macrophages and B lymphocytes (4). 2-arachidonoylglycerol (2-AG) is the most abundant endocannabinoid found in the brain. 2-AG plays an important role in a variety of physiological functions including neuroprotection, regulation of food intake, anti-inflammation, and anti-nociception (4,5). It is believed that 2-AG is produced on demand in post-synaptic neurons, secreted, and acts at CB1 receptors expressed on pre-synaptic neurons (4). Monoacylglycerol lipase (MAGL) is co-expressed with CB1 receptors and is responsible for ~85% of 2-AG hydrolysis in the brain (6–8). MAGL is a 33-kDa serine hydrolase containing a catalytic triad (Ser-122, Asp-239, and His-269 in the mouse homolog) characteristic of members of this superfamily of proteins (9).
2-AG hydrolysis may be quantified by mass spectrometry or assays with radioactive substrate. Mass spectrometry is impractical for high-throughput assays, while radiolabeled 2-AG is very costly. Radiolabeled 2-oleoylglycerol (2-OG), which does not bind to cannabinoid receptors, is often used as a substitute for 2-AG since MAGL hydrolyzes both at similar rates (10,11). Nevertheless, radioactive 2-OG is still relatively expensive. In this study, we explored an alternative method of measuring 2-AG hydrolysis based upon reagents originally developed by Cayman Chemical Company to measure the potency of MAGL inhibitors. The assay employs a thioester-containing analog of 2-AG, arachidonoyl-1-thio-glycerol (A-1-TG), which MAGL can hydrolyze. The thioglycerol released upon hydrolysis by MAGL reacts with 5,5′-dithiobis(2-dinitrobenzoic acid) (DTNB) resulting in the release of a yellow thiolate ion (TNB) (Fig. 1) (12). This assay was utilized to measure MAGL kinetics, activity in cytosolic and membrane fractions, and loss of activity in a catalytically inactive mutant. Saario et al. developed the inhibitor N-arachidonyl maleimide (NAM) that irreversibly binds to sulfhydryl groups of cysteine residues (Cys208 and Cys242) in the active site of MAGL (13–15). The potency of this inhibitor was measured using the spectrophotometric assay and a conventional radioactive assay using 3H-2-OG for comparison. These experiments serve to measure the assay’s applicability and limitations.
N-arachidonyl maleimide, methylarachidonyl fluorophosphonate, and arachidonoyl-1-thio-glycerol were purchased from Cayman Chemical Company (Ann Arbor, MI, USA). 5,5′-dithiobis(2-dinitrobenzoic acid) and 2-oleoylglycerol were purchased from Sigma-Aldrich (St. Louis, MO, USA). Mono-oleoyl glycerol racemic 2-[glycerol-1 2 3-3H] (1 mCi/mL) was purchased from American Radiolabeled Chemicals (St. Louis, MO, USA).
Total mRNA was extracted from mouse brain using the RNAeasy Mini Kit (Qiagen). Mouse MAGL cDNA was generated from whole mouse brain mRNA using SuperScriptIII First-Strand Synthesis for RT-PCR (Invitrogen) and cloned into a pCDNA4/TO/myc-HISA vector (Invitrogen). The Ser122Ala mutation was obtained using the PCR overlap extension method as described by Heckman and Pease (16). Dual PCR reactions containing (1) a forward primer annealing to the 5′ end of MAGL (5′-GATATAGGTACCGCCACCATGCCTGAGGCAAGTTCACC-3′) and a reverse mutation primer (5′-GAAGGAGGACCCGGTGCGGTACCCGCCACGGT-3′) and (2) a forward mutation primer (5′-TTCCTCCTGGGCCACGCCATGGGCGGTGCCATC-3′) and a reverse primer annealing to the 3′ end of MAGL (5′-GATATATCTCGAGTCAGGGTGGACACCCAGC-3′) initially inserted nucleotide changes into fragments of the MAGL sequence. Amplification products from both of these reactions were used as templates for a second PCR reaction using MAGL 5′-and 3′-end primers. PCR reactions were carried out using Platinum Pfx DNA Polymerase (Invitrogen) using the following cycling conditions: denaturing at 95°C for 30 s, annealing for 1 min, and elongation at 68°C for 30–60 s (35 cycles). The amplification product from this reaction was gel extracted using QIAquick Gel Extraction Kit (Qiagen), subjected to restriction digest with KpnI and XhoI (New England Biolabs), and re-ligated to the pCDNA4 vector. The vector with insert was grown in competent DH5α E. coli (Invitrogen) and isolated using the QIAPrep Spin Miniprep Kit (Qiagen). Insertion of mutated nucleotides was confirmed by Big Dye Terminator Sequencing (Applied Biosystems, Stony Brook HSC Sequencing Facility) over the entire open reading frame. The nucleotides in italics are nucleotide changes while underlined nucleotides are restriction enzyme recognition sites in the above oligomer sequences.
COS-7 cells were maintained in 10 cm culture dishes in DMEM (Gibco) containing 10% calf serum (Gibco), 5% penicillin/streptomycin (Gibco), and 5% sodium pyruvate (Gibco). Cells were transfected at ~70% confluence using 6 μg cDNA with Lipofectamine 2000 (Invitrogen) as prescribed by the manufacturer.
Cells were harvested 24 h after transfection and mechanically homogenized in 20 mM Tris/1 mM EDTA pH 8.0 (with 320 mM sucrose for fractionation) using a 26-gauge needle. Whole cell lysate was obtained from the soluble fraction after spinning the homogenate at 1,000×g for 15 min. Membrane and cystosolic fractions were obtained by spinning whole cell lysate at 100,000×g for 30 min. Protein content was quantified using the Pierce BCA Assay Kit.
Reaction mixtures containing 10 mM Tris/1 mM EDTA pH 7.2, MAGL-transfected lysate, membrane, or cytosolic fraction, and inhibitor or vehicle were pre-incubated at 4°C (or room temperature for NAM) for 15 min. They were then incubated at 37°C for 3–5 min after addition of A-1-TG. One millimolar DTNB was added to the reactions at room temperature. Absorbances were read at 412 nm in 1-cm cuvettes using a Pharmacia Biotech Ultraspec 2000.
A cuvette containing only buffer was used as a blank. The concentration of thiolate ion per reaction was calculated using the Beer–Lambert equation, A=εcl, where A is the absorbance, ε is the molar extinction coefficient (14,150 M−1 cm−1), c is the concentration, and l is the path length. The thiolate concentration is proportional to the thioglycerol concentration and thus a measure of substrate hydrolysis efficiency.
Reaction mixtures containing 10 mM Tris–HCl (pH 8.0), MAGL-transfected lysate, 0.5 mg/mL fatty acid free BSA, and NAM or vehicle were pre-incubated at room temperature for 15 min. They were then incubated at 37°C for 15 min after addition of 0.8 nCi [3H]-2-OG + 70 μM 2-OG. Reactions were stopped by the addition of 2 volumes of chloroform:methanol (1:1, v/v) and spun at 1,000×g for 10 min. The aqueous phase was removed for sampling with 3 mL of ScintiVerse (Fisher). Samples were counted using a Beckman LS 6500 scintillation counter.
Goodness of fit, kinetic constants, and statistical significance were determined by using GraphPad Prism 4. Statistical significance was evaluated by using two-tailed unpaired t tests.
Experiments were conducted to determine if results obtained with the spectrophotometric assay were comparable to those obtained in previous radiometric studies. Increasing amounts of MAGL-transfected COS-7 cell lysate were incubated with 200 µM A-1-TG. MAGL activity increased linearly with up to 11 µg cell lysate protein (R2=0.96, Fig. 2a). An apparent Km=67.9±3.0 µM and Vmax=659.5±81.8 nmol/min/mg were obtained when increasing amounts of A-1-TG were incubated with 5 μg MAGL-transfected cell lysate protein (Fig. 2b). A comparison of MAGL activity in membrane versus cytosolic fractions revealed contributions of 49.2±2.0% and 50.8±2.0% to total activity, respectively (Fig. 3a). Mutation of catalytic serine 122 to an alanine (MAGLS122A) abolished hydrolytic activity in both membrane and cytosolic fractions (Fig. 3b).
In this set of experiments, the rates of hydrolysis of the substrates for the spectrophotometric and radiometric assays were 605.3 and 74.2 nmol/min/mg protein (n=6) with 3.2% and 5.8% relative standard errors of the mean, respectively. The spectrophotometric assay was also compared to the radiometric assay to test its sensitivity. Shown in Fig. 4 is a typical experiment where the effects of increasing amounts of the inhibitor NAM on MAGL activity was measured using both assays. An IC50 of 45.7 nM and 94.3 nM was obtained for the radiometric and spectrophotometric assay, respectively. At the lowest NAM concentration (10−9M), the spectrophotometric assay was not sensitive enough to detect inhibition. The spectrophotometric assay and the radiometric assay between 10−8M NAM and 10−6M NAM were comparable. Very high concentrations of NAM (10−6 and 10−5M) gave nearly complete inhibition of MAGL activity for both assays.
The study of 2-AG metabolism is crucial to understanding its physiological function as its synthesis and hydrolysis are tightly coupled to cannabinoid receptor activation. MAGL, as the major enzyme for 2-AG hydrolysis, is an attractive target for inhibitor design. Although cloned more than a decade ago, few selective inhibitors for MAGL have been synthesized. This may be owing to, in part, the lack of available fast, cost-effective assays. The experiments performed in this study used a spectrophotometric assay to look at MAGL hydrolytic activity and the effect of an inhibitor upon the enzyme.
The proportional increase of MAGL activity with increasing protein lysate demonstrated linearity with the spectrophotometric assay. Furthermore, the assay detected at least 2 μg changes in protein quantity with short incubation times. The saturation curve for A-1-TG hydrolysis demonstrated that this reaction follows Michaelis–Menten kinetics. Hydrolytic activity was equally divided between membrane and cytosolic fractions, which is characteristic of peripheral membrane proteins. It showed a total loss of enzyme activity when the catalytic serine of MAGL was mutated to an alanine.
These results are comparable to those obtained with radiometric assays in previous studies. The Km of 67.9±26.1 calculated with the spectrophotometric assay was nearly identical to the Km of 67.8±4.0 µM determined for rat cerebellar membranes using a radiometric assay (17). Muccioli and colleagues also determined that MAGL activity was not significantly different between membrane and cytosolic fractions when compared to total cell lysate activity (18). Initial studies showed MAGL possessed a catalytic triad characteristic of serine hydrolases. Subsequent mutation of serine 122 to an alanine abolished enzyme activity showing that this residue is essential for catalysis (9). This result was duplicated with the spectrophotometric assay.
The sensitivity and precision of the spectrophotometric assay was compared to a radiometric assay through inhibition studies. The colorimetric assay could not detect inhibition at the lowest concentration of NAM used and generally showed higher percent remaining activity with all NAM concentrations examined when compared with the radiometric assay. However, the spectrophotometric assay was found to be as precise as the radiometric assay.
Other alternative assays have been previously developed. One colorimetric assay used 4-nitrophenylacetate to quantify MAGL activity with Vmax=52.2 µmol/min/mg (19). A second assay utilized the fluorogenic substrate 7-hydroxycoumarinyl-arachidonate with Vmax=1,700 µmol/min/mg (20). Both these methods obtained results comparable to radiometric assays. The fluorescent assay was intended for use with purified MAGL while the assay using 4-nitrophenylacetate as well as the assay described herein have the advantage of working with homogenates and do not require a spectrofluorimeter or scintillation counter. The rates of hydrolysis for the various substrates vary widely. This is owing to their specific structure and the resulting susceptibility to hydrolysis by MAGL and the purity of the MAGL preparations. Conveniently, the substrate (A-1-TG) used in the current study is hydrolyzed at approximately a tenfold rate relative to the radioactive substrate (3H-2-OG).
The spectrophotometric assay described here was useful for measuring linear increases in activity with an increase in MAGL-transfected cell lysate protein, MAGL kinetics, and distribution of enzyme activity in cellular fractions. Although not as sensitive as a radiometric assay, NAM inhibition of MAGL was detectable except at the lowest concentration of inhibitor. Loss of total activity in a catalytic mutant was also detectable. This assay can be used with both purified MAGL or cultured cells.
We acknowledge the National Institute on Drug Abuse for support (DA016419, DA027103, and DA026953).