Science & Technology

A visible wavelength spectrophotometric assay suitable for high-throughput screening of 3-hydroxy-3-methylglutaryl-CoA synthase

A visible wavelength spectrophotometric assay suitable for high-throughput screening of 3-hydroxy-3-methylglutaryl-CoA synthase
of 7
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  A visible wavelength spectrophotometric assay suitable for high-throughputscreening of 3-hydroxy-3-methylglutaryl-CoA synthase D. Andrew Skaff, Henry M. Miziorko * Division of Molecular Biology and Biochemistry, School of Biological Sciences, University of Missouri–Kansas City, Kansas City, MO 64110, USA a r t i c l e i n f o  Article history: Received 17 July 2009Available online 23 August 2009 Keywords: HMG-CoA synthaseVisible wavelength assayHigh-throughput screeningDithiobisnitrobenzoic acid (DTNB) a b s t r a c t 3-Hydroxy-3-methylglutaryl-CoA (HMG-CoA) synthase catalyzes the first physiologically irreversiblestep in biosynthesis of isoprenoids and sterols from acetyl-CoA. Inhibition of enzyme activity by  b -lac-tone-containing natural products correlates with substantial diminution of sterol synthesis, identifyingHMG-CoA synthase as a potential drug target and suggesting that identification of effective inhibitorswould be valuable. A visible wavelength spectrophotometric assay for HMG-CoA synthase has beendeveloped. The assay uses dithiobisnitrobenzoic acid (DTNB) to detect coenzyme A (CoASH) release onacetylation of enzyme by the substrate acetyl-CoA, which precedes condensation with acetoacetyl-CoAto form the HMG-CoA product. The assay method takes advantage of the stability of recombinant enzymein the absence of a reducing agent. It can be scaled down to a 60  l l volume to allow the use of 384-wellmicroplates, facilitating high-throughput screening of compound libraries. Enzyme activity measured inthe microplate assay is comparable to values measured by using conventional scale spectrophotometricassays with the DTNB method (412 nm) for CoASH production or by monitoring the use of a second sub-strate, acetoacetyl-CoA (300 nm). The high-throughput assay method has been successfully used toscreen a library of more than 100,000 drug-like compounds and has identified both reversible and irre-versible inhibitors of the human enzyme.   2009 Elsevier Inc. All rights reserved. Introduction 3-Hydroxy-3-methylglutaryl-CoA (HMG-CoA) 1 synthase (EC2.3.3.10) catalyzes the condensation of acetyl-CoA (Ac-CoA) withacetoacetyl-CoA to form HMG-CoA in a reaction that involves twocovalent reaction intermediates: acetyl- S  -enzyme and enzyme- S  -HMG-CoA (Scheme 1). The reaction is the first physiologically irre-versible reaction and, thus, represents a key early step in the path-way for biosynthesis of isoprenoids and sterols from Ac-CoA(Scheme 2), and a deficiency in HMG-CoA production has been dem-onstrated to block growth of prokaryotic ( Streptococcus pneumoniae )[1] and eukaryotic (yeast) [2] organisms. Inactivation of the purified enzyme by modification of the cys-teine involved in forming covalent reaction intermediates has beenaccomplished using the mechanism-based inhibitor 3-chloropropi-onyl-CoA [3]. The same cysteine is covalently modified (thioesteri-fied) by the use of   b -lactone-containing fungal metabolites such ashymeglusin (also referred to as 1233A, F-244, or L659,699) [4,5].Treatment with  b -lactone inhibitors can dramatically diminish(by >85%) cholesterol synthesis in animals [6]. However, the en-zyme has a low intrinsic thioesterase activity [7]; this may explainwhy, under in vivo conditions, enzyme activity quickly reboundswhen hymeglusin treatment is discontinued [8].Such observations validate HMG-CoA synthase as a target fordevelopment of therapeutic agents that diminish isoprenoid or ste-rol synthesis. It seems likely, however, that identification of atightly binding reversible inhibitor with drug-like properties ratherthan covalent inactivators of the types described above would bedesirable. For this reason, efforts have been made to screen a pro-prietary library of more than 100,000 compounds for HMG-CoAsynthase inhibitors. To pursue this goal, it was important to devel-op an enzyme assay that would be appropriate for use with typicalmicroplate multiwell formats. This requires volumes of   6 100 l land also requires reaction conditions under which substrates andenzyme are stable for the extended times needed for screeningthe multiple microplates containing the library of drug-like com-pounds. A variety of strategies were considered, and initial exper-imental tests were performed. Preliminary work led to thedevelopment of an assay that required no coupling enzymes and 0003-2697/$ - see front matter    2009 Elsevier Inc. All rights reserved.doi:10.1016/j.ab.2009.08.030 *  Corresponding author. Address: Division of Molecular Biology and Biochemis-try, School of Biological Sciences, University of Missouri–Kansas City, 5007 RockhillRoad, Kansas City, MO 64110, USA. Fax: +1 816 235 5595. E-mail address: (H.M. Miziorko). 1  Abbreviations used:  HMG-CoA, 3-hydroxy-3-methylglutaryl-CoA; Ac-CoA, acetyl-CoA; CoASH, coenzyme A; UV, ultraviolet; EDTA, ethylenediaminetetraacetic acid;DTNB, dithiobisnitrobenzoic acid; SD, standard deviation; CPM, 7-diethylamino-3-(4 0 -maleimidylphenyl)-4-methylcoumarin; GSH, glutathione; DTT, dithiothreitol; DMSO,dimethyl sulfoxide. Analytical Biochemistry 396 (2010) 96–102 Contents lists available at ScienceDirect Analytical Biochemistry journal homepage:  relied on spectrophotometric measurements of absorbance in avisible wavelength range that detected the production of the reac-tion product, coenzyme A (CoASH). As described in this article, theassay method could be scaled down to a volume compatible with amicroplate format, allowing it to be used productively for com-pound library screening. Materials and methods Ac-CoA and acetoacetyl-CoA were produced by reaction of CoASH with acetic anhydride and diketene, respectively, accordingto the method of Simon and Shemin [9]. Recombinant humanHMG-CoA synthase was produced in  Escherichia coli  using theexpression plasmid pET-HSyn described by Rokosz and coworkers[5], and the protein was isolated from bacterial lysates as describedfor the avian enzyme by Misra and coworkers [10]. Enzyme assays The standard ultraviolet (UV) wavelength spectrophotometricsemi-micro cuvette (1.0 ml) assay [11] contained 100 mM Tris–Cl(pH 8.0), 0.1 mM ethylenediaminetetraacetic acid (EDTA), 7 l Macetoacetyl-CoA, and Ac-CoA ranging from 10 to 1000 l M (as re-quired for  K  M  or  V  max  measurements). Reaction progress was mea-sured by the disappearance of substrate acetoacetyl-CoAmonitored at 300 nm ( e 300  = 3.6 mM  1 ) at 30   C. The dithiobisni-trobenzoic acid (DTNB) semi-micro cuvette (1.0 ml) assay con-tained 0.067 M Tris–Cl (pH 8.0), 130 l M Ac-CoA, 7 l Macetoacetyl-CoA, and 130 l M DTNB. The assay mix was incubatedat 24   C, and reaction progress was measured using a PerkinElmer k 35 spectrophotometer at 412 nm ( e 412  = 13.6 mM  1 ). Microplateassays were performed at 412 nm using a SpectraMax 96 or 384plate reader; the latter is capable of a fast reading time (30 s for ADPOS-CoA - O OO HO CH 3 PO O- Ac-CoA - O OO HO CH 3 P OOO - P O - OO - CoASHO OS-CoAADPAc-CoA CoASHO P OO O- P O - O O- R S  - OO HO CH 3  OOHO HO CH 3 - O2 NADPHS-CoAAcetyl-CoA Acetoacetyl-CoA 3-Hydroxy-3-Methylglutaryl-CoAMevalonic AcidMevalonate-5-PhosphateMevalonate-5-diphosphate Isopentenyl 5-DiphosphateATP+ Mg 2+ ATP+ Mg 2+ ATP + Mg 2+ ADP+P i  +CO 2 Thiolase HMG-CoA SynthaseHMG-CoA ReductaseMevalonate KinasePhosphomevalonate KinaseMevalonate DiphosphateDecarboxylase O- 2 NADPCoASH Scheme 2.  Intial reactions in biosynthesis of isoprenoids and sterols from Ac-CoA.   S-CoAOH 3 CH 3 C S-EnzOS-EnzOO + S-CoAOEnz-S S-CoA OO  CH 3 HO S  HO S-CoA OO  CH 3 HO S  H 2 C H 2 CS-EnzO EnzymeH + H 2 OEnzyme+CoASH Scheme 1.  HMG-CoA synthase reaction mechanism.  Assay for high-throughput screening of HMG-CoA synthase/D. Andrew Skaff, H.M. Miziorko/Anal. Biochem. 396 (2010) 96–102  97  a 384-well plate). During microplate assay development and vali-dation (96-well plate format), assay mixtures (120 l l) contained67 mM Tris–Cl (pH 8.0), 130 l M DTNB, 130 l M Ac-CoA, and7 l M acetoacetyl-CoA.Minor adjustments were implemented for the large number of 384-well microplate assays involved in compound library screen-ing. Liquid handling for microplate assays used a Thermo Multi-Drop 384 dispenser, which has eight channels for no-contactliquid dispensing of bulk reagent and can dispense liquids into a384-well plate in less than 20 s. A multichannel micropipettewas used for making serial dilutions in hit confirmation experi-ments. Well assay volume was reduced to 60 l l, but assay compo-nent concentrations were maintained at levels used in the twolarger format DTNB assays: (i) 67 mM Tris–Cl (pH 8.0), 130 l MDTNB, and 130 l M Ac-CoA and (ii) 7 l M acetoacetyl-CoA (the ace-toacetyl-CoA level used in the assay of Rokosz et al. [5] for the hu-man enzyme). For initial validation of the assay using the 384-wellmicroplate format, absorbance was measured at 30 s intervals. All384-well plate screening measurements were performed at theUniversity of Kansas High Throughput Screening Laboratory. Dataquality and reproducibility from replicate microplate assays wereestimated using the  Z  0 factor [12] calculated from the followingequation:  Z  0 ¼  1   ½ 3   ð SD positive  þ  SD negative Þ = ð Average Rate positive   Average Rate negative Þ : Thus,  Z  0 factors were calculated using the average rate values andcorresponding standard deviation (SD) values measured for repli-cate positive and negative controls included on each microwellplate. Typical  Z  0 factors observed during our screening of compoundlibrary plates ranged from 0.7 to 0.8. Kinetic parameters were calculated from data generated withstandard spectrophotometric or multiwell plate assays using theSigmaPlot Enzyme Kinetics Module. Results Strategies for microplate format assays of enzyme activity Standard assays for measurement of HMG-CoA synthase activ-ity rely on a sensitive radiometric method that follows conversionof [ 14 C]Ac-CoA into the acid-stable radioactivity in product[ 14 C]HMG-CoA, whereas on acidification and heating (95   C) todryness unreacted substrate is hydrolyzed to form volatile[ 14 C]acetic acid. Alternatively, a spectrophotometric method fol-lows depletion of substrate acetoacetyl-CoA at a wavelength(300 nm) in the UV range [11]. The radiometric method was pre-cluded by policy at the core screening laboratory associated withthe proprietary compound library. Although a strategy involvingrescaling of volumes used in the assay requiring 300-nm detectionmight have been pursued productively, detection at a higher wave-length also seemed worthy of consideration. Assays involvingdetection at visible wavelengths may offer some advantage overUV wavelength detection given that background absorbance con-tributions from library compounds, light scattering (microprecipi-tates), and polystyrene microplates typically used for screeningcompound libraries (  A 300 nm  0.3) should be decreased. Therefore,two alternative approaches were investigated, both involvingdetection of the reaction product CoASH. One method involveduse of the fluorescent sulfhydryl modification reagent CPM (7-diethylamino-3-(4 0 -maleimidylphenyl)-4-methylcoumarin). Thisapproach was previously reported to work for detection of CoASH[13]. It involved quenching aliquots of assay mixtures at fixedelapsed reaction times, adding fluorescent sulfhydryl reagent,and incubating for an appropriate time prior to fluorescence mea-surements in a microplate reader. The sensitivity of the fluorimet-ric method seemed to be attractive, but a continuous assay(directly monitoring the spectroscopic signal of unquenched sam-ple at multiple time points) seemed to be preferable; backgroundabsorbance contribution (measured at  t  o ) is constant and can besubtracted from assay measurements at subsequent times to cal-culate a corrected reaction rate. When the fluorescence approachwas tried using continuous monitoring of HMG-CoA synthase pro-duction of CoASH with CPM in the reaction mixture, the time per-iod that elapsed prior to observation of a linear reaction rate wasnot consistent. In an attempt to circumvent this complication, aspectrophotometric method for CoASH detection was developed.As indicated below, this alternative method proved to representa more reliable assay. Development of a DTNB assay for HMG-CoA synthase Although sulfhydryl detection using DTNB [14] is not as sensi-tive as a fluorimetric method, the production of thionitrobenzoate( e 412  = 13.6 mM  1 ) is monitored with reasonable sensitivity at412 nm, a visible wavelength that eliminates much of the back-ground absorbance contribution from a cuvette or polystyrenemicroplate (  A 412 nm 6 0.05) that can complicate detection at UVwavelengths. In addition, DTNB-based microplate assays for otherenzymes, such as glutathione (GSH) reductase [15] and trypanothi-one reductase [16], have been reported, suggesting the feasibilityof this approach. A spectrophotometric (  A 412 nm ) assay was de-signed and performed using standard semi-micro cuvettes(1.0 ml volume) and contained 67 mM Tris–Cl (pH 8.0), 130 l MDTNB, 130 l M Ac-CoA, and microgram amounts of enzyme. Thisassay approach was used to generate estimates of enzyme  V  max (0.67 ± 0.06 U/mg) and  K  m Ac-CoA  (73 ± 8 l M) that are in reasonableagreement with values for corresponding kinetic parameters( V  max  = 1.06 ± 0.04 U/mg and  K  m Ac-CoA  = 76 ± 10 l M) from thestandard spectrophotometric (  A 300 nm ) assay (Table 1). A  K  i  value(1.5 ± 0.2 l M) for the product HMG-CoA, which is a competitiveinhibitor with respect to substrate Ac-CoA, was determined usingthe DTNB assay. This estimate agrees well with the  K  i  value(1.1 ± 0.2 l M) measured using the standard spectrophotometric(  A 300 nm ) assay. An additional benefit of using the 412-nm methodis the abilityto determinean accurate K  m  value for acetoacetyl-CoA(5 ± 1 l M) without relying on use of the radiometric assay. In con-trast, when the 300-nm assay method is used, signal-to-noise lim-itations develop at low, variable acetoacetyl-CoA levels,complicating accurate measurement of   K  m  for this substrate. Design of a spectrophotometric microplate assay for HMG-CoAsynthase Two basic issues needed to be favorably resolved if the DTNBdetection method was to be successfully applied to a microplateassay of HMG-CoA synthase activity that could be useful for com-pound library screening. In comparison with single cuvette assays,for microplate assays there can be a longer time delay between thetime of reagent addition and the start of rate data acquisition dueto the robotic methods used for microplate loading and screening.This raises questions concerning the stability not only of reagentsbut also of enzyme, especially in an assay mixture that containsno sulfhydryl protective reagent (e.g., dithiothreitol [DTT]) becausesuch a reagent would be incompatible with CoASH detection byDTNB. A second related issue involves the possible DTNB inactiva-tion of enzyme, which is known to contain a reactive cysteine inthe active site (Scheme 1). Although human HMG-CoA synthase re-tains substantial activity in the presence of DTNB for a few minutesin the absence of a protective substrate, there is significantdiminu-tion in activity on incubation for longer time periods 98  Assay for high-throughput screening of HMG-CoA synthase/D. Andrew Skaff, H.M. Miziorko/Anal. Biochem. 396 (2010) 96–102  ( t  1/2  12 min). Microplate assay component solutions that elimi-nate any complications related to the concerns listed above weredeveloped. The first solution contained 0.2 M Tris–Cl (pH 8.0),400 l M Ac-CoA,and 400 l M DTNB. Solution 1 components are sta-ble for more than 6 h at 23   C; any substrate hydrolysis to formCoASH would be detectable by elevation of   A 412 nm  due to reactionwith DTNB. The second solution contained 21 l M acetoacetyl-CoAand enzyme (in dilute phosphate buffer, pH 7.0). Enzyme in solu-tion 2 is stable for more than 5 days; the tight-binding acetoace-tyl-CoA substrate is well established to form a binary complexthat protects the active site cysteine from modification [3] and,thus, maintains enzyme activity.The DTNB-based assay was scaled and tested for a 96-wellmicroplate format. During development and validation of the ap-proach, assay plate wells contained 40 l l of aqueous dimethyl sulf-oxide (DMSO, 2.5%, v/v), a volume and solvent level that couldcontain a library compound at 25 l g/ml. Aliquots (40 l l) of assaysolution 1 (containing Tris–Cl [pH 8.0], DTNB, and Ac-CoA) wereadded to each well, and the mixture was equilibrated at ambienttemperature (  22   C). The reaction was subsequently initiated bythe addition of 40- l l aliquots of assay solution 2 (containing ace-toacetyl-CoA and enzyme). Final concentrations in each microplatewell (120 l l) were 67 mM Tris–Cl (pH 8.0), 130 l M DTNB, 130 l MAc-CoA, 7 l M acetoacetyl-CoA, and enzyme levels ranging from 0.1to 0.3 l g. Next, 3 min after the addition of assay solution 2, reac-tion rates were determined by the measurement of   A 412 nm  (whichreflects production of the product CoASH). Fig. 1 depicts results of aquadruplicate assay performed over 4 min. Rates were linear withrespect to enzyme concentration (main panel,  R 2 = 0.99). For therepresentative progress curves shown in the inset, calculated cor-relation coefficients ( R 2 values) were 0.92, 0.98, and 0.95. Tracescorrespond to reactions containing 0.15-, 0.25-, and 0.30- l gamounts of protein, respectively. Linearity with time was observeduntil depletion of substrate acetoacetyl-CoA was approached. Goodreproducibility is indicated by calculation of an average  Z  0 factor of 0.7 [12] for data from the replicate assays. To further validate themicroplate assay, this method was used at variable concentrationsof Ac-CoA (10–1000 l M) (Fig. 2) so as to generate estimates of   V  max (1.04 ± 0.03 U/mg) and  K  m Ac-CoA  (84 ± 9 l M) (Table 1). This micro-plate assay method was also employed to determine a  K  i  value forHMG-CoA (3.1 ± 0.2 l M). These kinetic parameters (Table 1) are inreasonable agreement with values measured using standard spec-trophotometric assays performed at 300 or 412 nm. Implementation of a high-throughput library screen using the HMG-CoA synthase microplate assay In applying this methodology for the screening of library com-pounds in 384-well microplates, minor assay modifications weremade to accommodaterobotic plate filling and plate reader loadingas well as the scale of the screen (>100,000 compounds). Reaction  Table 1 Comparison of kinetic parameters measured in spectrophotometric cuvette and microplate format assays of HMG-CoA synthase. Kinetic parameter  D  A 300 nm  assay a (acetoacetyl-CoA use)  D  A 412 nm  assay a (DTNB measurement of CoASH formation)  D  A 412 nm  assay b (microplate format) V  max  (U/mg) c,d 1.06 ± 0.04 0.67 ± 0.06 1.04 ± 0.03 K  m Ac-CoA  ( l M) d 76 ± 10 73 ± 8 84 ± 9 K  i HMG-CoA  ( l M) e 1.1 ± 0.2 1.5 ± 0.2 3.1 ± 0.2 Note.  Components of assay mixtures are documented in Materials and methods. Calculation of kinetic parameters was performed using the SigmaPlot Enzyme KineticsModule. a Standard spectrophotometric assays are performed at 30   C (  A 300 nm  assay) or 24   C (  A 412 nm  assay) in semi-micro cuvettes using a PerkinElmer  k 35 UV–visible doublebeam spectrophotometer. b Microplate format assays are performed at ambient temperature (22–24   C) in 96- or 384-well plates using a SpectraMax plate reader. c One unit of HMG-CoA synthase activity corresponds to conversion of 1 l mol of substrate to product per minute. d For  V  max  and  K  m  determinations, Ac-CoA concentrations ranged from 10 to 1000 l M. e For  K  i  determinations, inhibitor (HMG-CoA) concentrations ranged from 2 to 16 l M. 00.250.50 0.1 0.2 0.3 0.4 HMG-CoA synthase (   g)    R  a   t  e   (     m  o   l   /  m   i  n   ) -0.0020.0020.006 0 1 2 3 4 Time (min) Fig. 1.  Dependence of HMG-CoA synthase enzymatic reaction rate on proteinconcentration. Averaged rates of quadruplicate measurements are shown versusprotein amounts rangingfrom 0.1 to 0.3 l g; rates are fit to a linear model with an  R 2 value of 0.99. Progress curves of the HMG-CoA synthase reaction were measured byDTNB detection (  A 412 nm ) of product CoASH formation. The assay was tested forreproducibility and time dependence by performing quadruplicate measurementsover a 4-min interval with various concentrations of enzyme in a microplate formatassay mix; time points were taken every 5 s. Ac-CoA and acetoacetyl-CoAconcentrations were 130 and 7 l M, respectively. The inset depicts progress curvesrepresentative of the data set used for determination of rates. Enzyme levels usedfor the progress curves depicted in the inset were as follows: 0.15 l g ( h ); 0.25 l g( D ); 0.3 l g ( e ). For clarity, only alternating time point measurements (i.e., 10 sintervals) are shown in the inset traces. All progress curves were fit to a linearmodel.  R 2 values of 0.92, 0.98, and 0.95 correspond to the inset trace data measuredat 0.15-, 0.25-, and 0.3- l g amounts of protein, respectively. 0 250 500 750 1000 [Ac-CoA] (   M)    R  a   t  e   (   U   /  m  g   ) 02468 0 0.025 0.05 0.075 0.1 1/[Ac-CoA] (   M -1 )    1   /   R  a   t  e   (   U   /  m  g   )   -   1 Fig. 2.  Ac-CoA concentration dependence of HMG-CoA synthase activity, asmeasured using the DTNB assay (412 nm) in a microplate format. Each pointrepresents an average of four measurements at each concentration of substrate.Concentrations of Ac-CoA were varied from 10 to 1000 l M in the 120  l l assayvolume. The acetoacetyl-CoA concentration was held constant at 7 l M, equivalentto levels used for compound library screening. The protein concentration was1.7 l g/ml. Nonlinear regression fit of the data to the Michaelis–Menten equation( R 2 = 0.96) yielded values of   V  max  = 1.04 ± 0.03 U/mg and  K  m  = 84 ± 9 l M. The insetdepicts a double reciprocal plot of the data.  Assay for high-throughput screening of HMG-CoA synthase/D. Andrew Skaff, H.M. Miziorko/Anal. Biochem. 396 (2010) 96–102  99  rates were determined by 412-nm absorbance measurements overa 3-min period taken at 30-s intervals. Final assay component con-centrations in each 60  l l well were 67 mM Tris–Cl (pH 8.0),130 l M DTNB, 130 l M Ac-CoA, 7 l M acetoacetyl-CoA, and librarycompounds (8 l g/ml). Background rate (negative control) wells(1–16) (Fig. 3) contained no enzyme and provide a benchmarkfor rates corresponding to 100% inhibition. Positive controls (wells17–32) contained enzyme as well as a 20- l l aliquot of DMSO solu-tion without a library compound; these indicate the rates expectedfor no inhibition. The use of a subsaturating substrate level allowssensitive detection of reaction inhibition in microplate wells con-taining library compounds.The compound library consisted of 104,000 compounds (molec-ular weight range = 150–480 Da) selected from the ChemDiv,ChemBridge, Prestwick, and MicroSource collections. A completeprimary screen required 11 half-day experiments. The top com-pounds (>200), which produced apparent inhibition of enzymeactivity by at least 50% at levels of 8 l g/ml, were selected for test-ing in a confirmation screen (Table 2). The resulting raw data setswere subjected to individual evaluation to rule out screening arti-facts and to validate the automated computer calculations of rateeffects. Of the compounds exhibiting the most efficacy in enzymeinhibition, 88 were selected for the design of serial dilution exper-iments to further discriminate among relative potencies of thesecompounds. Again, inspection of individual raw data sets was per-formed to eliminate screening artifacts and to confirm calculatedpotencies. Samples of solutions of the 30 most potent inhibitorswere acquired and retested for inhibition in buffer supplementedwith 0.1% Triton X-100 to rule out nonspecific effects of compoundaggregation [17] on measured enzyme activity. Compounds withinhibitory potency that is unaffected by detergent in the micro-plate assay (experiments performed in the home laboratory using120 l l reactionmixes in half-area96-well microplates with a Spec-traMax plate reader) were tested, and inhibition was confirmed, bykinetic assays using standard semi-micro cuvettes in a doublebeam spectrophotometer. Results of these tests indicate that sev-eral compounds (e.g., tetramethylthiuram disulfide, tetraethylthiu-ram disulfide (disulfiram), phenyl mercuric acetate) resulted intime-dependent irreversibleinhibition (Fig. 4). Other more promis-ing library compounds are reversible inhibitors exhibiting  K  i  valuesof less than 10 l M. These compounds, which include a variety of polyheterocyclic compounds (e.g., 2-benzyl-9-(3,6-dioxocyclo-hexa-1,4-dienyl)-2,3,7,8-tetrahydrol-6 H  -pyrido[1,2- a ]-pyrazine-1,4-dione, 2-amino-4-(4-bromothiophen-2-yl)-5-oxo-1- m -tolyl-1,4,5,6,7,8-hexahydroquinoline-3-carbonitrile, (  Z  )-5-(5-bromo-2-methoxy-benzylidene)-2-(4-methoxyphenyl)thiazolo[3,2- b ]-[1,2,4]triazol-6(5 H  )-one), are currently under further investigation aimed at elu-cidation of the molecular basis for inhibition and optimization of inhibitory potency. Examples of some types of these inhibitorsare depicted in Scheme 3. Discussion Previous reports [15,16] of the utility of DTNB in monitoringformation of thiol-containing products in microplate format assaysof enzyme activity encouraged pursuit of a similar strategy for -500501001500 64 128 192 256 320 384 Compound    %    I  n   h   i   b   i   t   i  o  n Fig. 3.  Results observed with a representative microplate from the HMG-CoAsynthase high-throughput screen. Wells 1–16 are the negative controls (opendiamonds) containing no enzyme. Rate estimates for the negative controls wereaveraged and assigned to 100% inhibition. Wells 17–32 are the positive controls(open diamonds) containing enzyme but no library compounds. Rate estimates forthe positive control wells were averaged and assigned to 0% inhibition. Wells 33–384 contain separate compounds (closed diamonds) from the library. Slopesrepresenting enzyme activity (reaction rates) are calculated using the softwareassociated with the SpectraMax Plus 384 plate reader. Baseline rate data for thecompound wells were normalized to the positive control wells. One microplatecompound with apparent inhibitory properties is identified (observed inhibi-tion > 50%, which is the cutoff value used to qualify compounds for furtherevaluation). A 60 l l reaction mixture contained 67 mM Tris (pH 8.0), 130 l M DTNB,130 l M Ac-CoA, 7 l M acetoacetyl-CoA, and 0.04 mU HMG-CoA synthase in 0.8%DMSO. Library compound concentrations were 8 l g/ml. The calculated  Z  0 factor forthis microplate was 0.80, a value typical of estimates for other microplates in thelibrary screen.  Table 2 Summary of results from a confirmation screen of selected library compounds. Residual HMG-CoA synthase activity(% of uninhibited control)Number of compounds0–9 1010–19 920–29 1030–39 1340–49 650–59 1060–69 170–79 780–89 1390–99 20 P 100 (no inhibition) 138 Note.  Compounds (>200) exhibiting the most effective apparent inhibition of HMG-CoA synthase activity in a primary screen were reassayed in a 384-well microplateformat to confirm inhibition. The same assay conditions employed in the primaryscreen were used for the confirmation plate; the 60- l l reaction mixture contained67 mM Tris (pH 8.0), 130 l M DTNB, 130 l M Ac-CoA, 7 l M acetoacetyl-CoA, and0.04 mU HMG-CoA synthase in 0.8% DMSO. The library compound concentrationwas 8 l g/ml. Kinetic data were collected at 412 nm using a SpectraMax 384microplate reader. The level of activity is calculated by comparison of reaction ratesfor the compound-containing wells with the average of the reaction rates in thepositive controls (without compound) on the microplate. The calculated  Z  0 factor fordata from the confirmation microplate was 0.72. Fig. 4.  Time-dependent inactivation of HMG-CoA synthase by phenyl mercuricacetate. Enzyme (1 mg/ml) and inhibitor (64 l M) were incubated in 100 mMpotassium phosphate buffer (pH 7.0) at 0   C. Then 5  l l aliquots of this inactivationmix were withdrawn at time intervals ranging from 15 s to 4 min. Residual enzymeactivity was measured using the standard  A 300 nm  assay (1.0 ml/cuvette), whichincluded Ac-CoA and acetoacetyl-CoA at concentrations of 100 and 10 l M,respectively. The inhibitor concentration in the assay was 3.2 l M. A linear fit of asemi-log plot of the data ( R 2 = 0.99) indicated a  t  1/2  of 1.2 min, which corresponds toan inactivation rate constant of 0.58 min  1 under these conditions.100  Assay for high-throughput screening of HMG-CoA synthase/D. Andrew Skaff, H.M. Miziorko/Anal. Biochem. 396 (2010) 96–102
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks