A microtiter-plate assay of nitric oxide synthase activity

Microtiter-Plate Assays 275

MOLECULAR BIOTECHNOLOGY Volume 12, 1999

PROTOCOL

275

A Microtiter-Plate Assay of Nitric Oxide Synthase Activity

John Dawson and Richard G. Knowles

AbstractWe describe here a microtiter-plate assay for measuring nitric oxide synthase (NOS) activity by utilizing

the spectral shift in optical absorbence between the wavelengths 405 and 420 nm on conversion of oxyhe-moglobin to methemoglobin by nitric oxide (NO). This is a high-throughput assay permitting 96 or 384simultaneous kinetic measurements and is ideal for the study of NOS inhibitors and their time dependence.It is also possible to measure enzyme rates under different conditions simultaneously for the study of thecofactor and substrate dependence of NOS preparations. The assay requires approximately l0 pmol/min ofNOS activity to achieve a 1moD/min rate.

Index Entries: Nitric oxide synthases; oxyhemoglobin; microtiter-plate; kinetic.

Molecular Biotechnology 1999 Humana Press Inc. All rights of any nature whatsoever reserved. 1073–6085/1999/12:3/275–279/$11.25

*Author to whom all correspondence and reprint requests should be addressed. Glaxo-Wellcome Research and Development, 5G120, Gun-nels Wood Road, Stevenage, Herts SG1 2NY. E-mail jd43048@ Glaxowellcome.Co.UK

1. IntroductionNitric oxide synthase (NOS) catalyzes the con-

version of L-arginine, molecular oxygen, andnicotinamide adenine dinucleotide phosphate(NADPH) to NO, citrulline, and NADP+

(reviewed in ref. 1). The neuronal (n) and endot-helial (e) NOS isozymes are highly regulated byCa2+ and calmodulin (CaM), whereas the iNOShas CaM tightly bound. NOS are heme proteinsthat also contain tightly bound flavin adeninedinucleotide (FAD) and flavin mononucleotide(FMN) and require tetrahydrobiopterin (BH4) foractivity, whereas NADPH is used as a substrate.

This article describes a microtiter-plate assayfor measuring NOS activity by utilizing the spec-tral shift in optical absorbence on conversion ofoxyhemoglobin to methemoglobin by nitric oxide(NO). The oxyhemoglobin assay was firstdescribed by Feelish and Noack to quantify NOrelease by chemical NO donors (2), was modifiedby Salter and Knowles (3) to apply it to measure

NOS activity and increase the sensitivity, and hasnow been further modified to permit it to be runin microtiter-plate format. The original assay (2)used the absorbence difference between 401 nm(maximum point) and 411 nm (isosbestic point);however, greater sensitivity is achieved by moni-toring the absorption difference between thewavelengths 401 and 421 nm (minimum point).This microtiter-plate assay uses the closest readilyavailable filters to do this, 405 nm and 420 nm(giving close to the maximal sensitivity), and usesa relatively full well (250 µl total volume) tomaximize the optical path-length and thereby thesensitivity.

With the appropriate microtiter-plate reader,this assay permits 96 or 384 simultaneous kineticmeasurements of NOS activity. It is run at 37°Cand is ideal for the study of NOS inhibitors andtheir time dependence, which can also be tem-perature-dependent. Using a similar method, it ispossible to measure enzyme rates under differentconditions simultaneously. This is useful, for

276 Dawson and Knowles


example, for the study of the cofactor and sub-strate dependence of NOS preparations.

Other microtiter-plate assays of NOS activityhave been described. The oxymyoglobin assay(4,5) should on theoretical grounds be very simi-lar to that described here, kinetically measuringNO formation. Assays of NOS by measuringNADPH oxidation (6) are much less sensitive andmay not accurately reflect NO synthesis, sinceNADPH consumption is under some circum-stances uncoupled from NO synthesis. Assays ofNOS using the Griess reaction to measurenitrite formed (7) are end-point rather thankinetic, are somewhat less sensitive, and aresubject to the concern that only one of the twomajor products of NO breakdown (nitrate andnitrite) is being measured: the ratio of nitrite tonitrate formed can vary depending on what ispresent in the assay. Microtiter-plate assays basedon oxyhemoglobin or oxymyoglobin are thereforethe most sensitive and reliable assays for NOSactivity.

This version of the assay is somewhat less sensi-tive than the optimized spectrophotometric (3) orradiometric (8) assays, requiring approx 10 pmol/minof NOS activity to achieve a 1 mODU/min rate. Thissensitivity is quite adequate for many studies, e.g.,assays of brain-cytosol NOS, of recombinant-ex-pressed NOS, or of purified NOS.

2. Materials

1. Extraction buffer (EB): The extraction buffershould be prepared in advance. The basicbuffer is first prepared by dissolving sucrose(250 mM), Tris base (50 mM), andethylenediaminetetraacetic acid (EDTA) (1mM) in double-distilled or milliQ-grade waterand bringing its pH to 7.4 at room tempera-ture by the addition of HCl. The followingconstituents are then added to the final con-centrations indicated: 0.1 mM D/L-dithiothreitol (DTT), 0.5 µM leupeptin, 0.5µM pepstatin A, and 10 µM Antipain; thebuffer is then made up to its final volume withwater. This EB is then distributed intoaliquots (typically 50 mL per aliquot) andstored at –20°C until required.

2. 10 mg/mL Phenylmethylsulphonyl fluoride(PMSF): Because it is unstable in aqueoussolution, PMSF is not included in the buffer atthis stage, but prepared as a solution in abso-lute ethanol, stored at –20°C, and added to theEB during the extraction procedure (seebelow). The composition of this EB is designedto permit extraction of NOS from cells or tis-sues without breaking intracellular organellesand minimizing proteolysis by chelating diva-lent cations with EDTA and the inclusion ofprotease inhibitors.

3. Assay buffer: 100 mM HEPES is dissolved indouble-distilled or milliQ-grade water andbrought to pH 7.4 by the addition of NaOH. Thiscan be stored for several weeks at 4°C. Prior tothe assay of NOS, DTT is added to the buffer togive 100 µM. CaCl2, MgCl2, and hemoglobinare added as required and the buffer mixturewarmed to the required temperature. The cofac-tor and substrate stocks are kept in the dark onice until use (see Notes 1–3).

4. 1 M stock solutions of CaCl2 and MgCl2 can bestored for several weeks at 4°C.

5. Cofactor stocks: 7.5 mM L-arginine (HCl), 250mM NADPH, 250 µM FMN, 250 µM FAD, and25 µM CaM (Sigma P2277, from Bovine Brain,assumed M.W. 17,000) are dissolved in HEPESbuffer and stock solutions kept at -20°C. A“cocktail” of these can be made and stored inthe same way to reduce the number of additionsto the assay.

6. (6R)-5,6,7,8-Tetrahydro-L-biopterin hydro-chloride (BH4): On the day of assay, DTT isadded to 10 mM HCl (stock stored at roomtemperature) to give 500 µM. BH4 is dissolvedin this to give 500 µM and the solution kept onice in the dark until use. It can be used forapprox 4 h.

7. NOS inhibitors: Inhibitors are dissolved in25% (v/v) DMSO/100 mM HEPES buffer pH7.4 at 12.5 times the final concentrationrequired. A stock solution of 12.5 mM NG-me-thyl-L-arginine (L-NMMA) for blanks,dissolved in this way, can be stored at –20°C.Up to 2% DMSO can be tolerated in the assay.For IC50 determination, we routinely use 10concentrations with twofold dilution.



8. Ion-exchange resin: Dowex 50W Na+ form,200–400 mesh, 8% cross-linked. This is pre-pared by converting the H+ form (Sigma50X8–400) by washing the resin in a beakerwith 1M NaOH. Two washes with 1 vol ofNaOH are carried out by adding NaOH to theresin, mixing several times, allowing the resinto settle under gravity (~5 min) and aspiratingthe supernatant with a water-jet vacuum pump.The resin is then washed with water until thepH is <8 and finally brought to 50% resin inwater (approximately by the height of thesettled resin). The resin can be prepared in bulkand stored at 4°C.

9. 3mM oxyhemoglobin (monomer). We use oxy-hemoglobin prepared from human blood (9).Commercially available hemoglobin consistslargely of the met form and must be convertedto its reduced oxy-form by treating with so-dium dithionite blowing with oxygen and de-salting by running the solution through aSephadex G-25 column (4).

10. Enzyme: Baculovirus-expressed human-re-combinant iNOS, nNOS, and eNOS (6). 1U =1 nmol/min per plate at 37°C under the condi-tions of the assay.

11. Microtiter plates: Costar 3598 flat bottom.12. Microtiter-plate reader: A dual-wavelength

microplate reader with 405 and420 nm filters, temperature control, and kineticsoftware is required. We have used theMolecular Devices THERMOmax, DynatechMR7000, and Dynatech DIAS readers withSOFTmax, Biolinx, and Revelation software,respectively.

13. A microtiter-plate incubator and incubated or-bital-plate shaker, both set at 37°C, are alsorequired.

3. Methods3.1. Enzyme Preparation

We routinely use crude extracts of the humanrecombinant NOS isoforms expressed inSpodoptera frugiperda Clone 21 (Sf 21) cells usingthe baculovirus expression system (10).

1. Scrape Sf 21 cells grown as monolayers into asmall volume of EB using a rubber policemanor a purpose-made cell scraper with the plate

or flask standing on ice. Centrifuge cells grownin suspension to collect them as a cell pelletand resuspend the cell pellet in a small volumeof ice cold EB.

2. Disrupt the cell extracts by sonication. As withthe procedures with tissue extraction, it isimportant to keep the temperature at 0–4°C.Most probe sonicators should be suitable forthis purpose; we have routinely used an MSESoniprep 150 with a pre-cooled 5 mm tipprobe, sonicating three times for 10 s at 10 µmamplitude with 30 s cooling in ice water inbetween. PMSF is added after the first sonica-tion to give 100 µg/mL.

3. Centrifuge the cell extract at 100,000g for 30min at 4°C to remove insoluble material andstore at –70°C.

3.2. Removal of Endogenous ArginineThe arginine concentration in the standard

assay is 30 µM. However, high concentrations ofarginine can be present in the cell extracts. Weroutinely remove endogenous arginine by a pre-treatment with Dowex 50W (Na+ form) ion-ex-change resin, prepared as described, and thenwashed in EB.

1. Add 2 vol of enzyme extract to 1 vol of ice-cold, packed resin in microfuge tubes and mix.

2. Briefly centrifuge the mixture (e.g., 10,000gfor 1 min at 0–4°C) and collect the supernatantfor assay. It is crucial that no resin is carriedforward into the assay: it will adsorb the sub-strate!

3.3. The Assay Procedure1. Add 20 µL of NOS inhibitor, 12.5 mM L-

NMMA (for blanks, to produce 100% inhibi-tion) or 25% (v/v) DMSO/100 mM HEPESbuffer pH 7.4 (controls) to the 96-well plate.

2. Warm the plate in the 37°C incubator forapproximately 30 min (see Note 4).

3. Dilute the enzyme stock to the required work-ing strength, 12.5 times the final concentration,in HEPES buffer (2.5 mL of working strengthper plate, 0.5 U per mL) (see Note 5).

4. Add 0.6 mL of cofactor mix and 0.6 mL of BH4to 23.3 mL of the pre-warmed oxyhemoglobin,CaCl2, MgCl2, and buffer mixture. Final con-

278 Dawson and Knowles


centrations of the effectors used with the humanrecombinant enzyme are 5 µM oxyhemoglobin;1 mM MgCl2; 30 µM L-arginine (or as required);100 µM NADPH; 1 µM FMN;1 µM FAD; 100nM CaM; and 10 µM BH4 (see Note 6).

5. Add 210 µL of this complete mixture to each wellof the plate, keeping it at 37°C.

6. Rapidly add 20 µL of working strength enzymeper well and mix the plate with an incubated(37°C) orbital shaker for 15 s or a multipipet, tak-ing care not to cross contaminate wells.

7. Place the plate in the reader (pre-warmed to37°C) and start reading.

8. Follow the reaction by reading the difference inabsorbance between 420 and 405 nm, using adual-wavelength kinetic-plate reader at 37°C.Rates are determined as mOD/min by linearregression. [See Fig. 1 for representative resultswith L-NMMA and N -nitro-L-arginine(NitroArg).]

3.4. Calculation of ResultsThe L-NMMA-blank rate is subtracted from all

the control and test rates and the percentage inhibi-

tion calculated if required. Absolute NOS activity(enzyme units per weight of tissue or protein) can becalculated using the apparent-extinction coefficient(see Note 7). The path length will depend on theassay volume and type of plate.

4. Notes1. Phosphate buffer can also be used for the assay.2. We have found that concentrations of DTT

above 500 µM inhibit the enzyme, when usingthe crude human-recombinant isozymes(unpublished data). This is in contradiction withthe literature reporting stabilization of purifiedNOS by higher concentrations of DTT (11,12).

3. It is not necessary to add calcium under theseconditions, since the calcium concentrationpresent in the buffers as a contaminant (prob-ably low micromolar) is sufficient for expres-sion of full activity. Addition of highconcentrations (e.g., 200 µM ) can be detrimen-tal, decreasing the linearity of the enzymewhen using unpurified human-recombinantpreparations.

Fig. 1. (Top) Time-courses of NO synthesis by nNOS and its inhibition by L-NMMA and NitroArg. (Bottom) 96-well plate layout. Concentration of inhibitor(µM), Human Recombinant nNOS, 37°C, 30 min. Scales: x axis time 0–30 min, yaxis absorbence 0–100 mOD.



4. Plates at room temperature cool rapidly. Thetemperature settings of incubators and platereaders should be checked. We routinely moni-tor the well temperature with a miniature ther-mocouple (Physitemp BAT-12 with IT-23thermocouple microprobe).

5. It is crucial to ensure that under the particularconditions of the assay, temperature andenzyme source, rates are linear both with timeand concentration of enzyme extract added.

6. The “complete-assay mix” can only be used forapproximately 1 h at 37°C as the cofactors/sub-strates are unstable at this temperature. Do notleave exposed to light for long periods.

7. Calculation of the rates as pmol per min permass of sample or protein requires the deter-mination of the apparent-extinction coefficientfor the NO-mediated spectral shift, becausethis depends on the precise wavelengths andbandwidths of the filters used, the volume ofthe assay, and the shape of the wells used. Thisis determined by measuring the change inabsorbence at 405–420 nm following oxidationof a known concentration of hemoglobin byNO (as the dissolved gas or from NO donors).The stoichiometry of the reaction is 1 mol ofoxyhemoglobin monomer oxidized per mol ofNO. Typically, the apparent-extinction coeffi-cient is in the range 30,000–50,000 M–1.

8. Where not specified, the reagents can beobtained from Sigma.

References1. Knowles, R. G. and Moncada, S. (1994) Nitric oxide

synthases in mammals. Biochem. J. 298, 249–258.2. Feelisch, M. and Noack, E.A. (1987) Correlation

between nitric oxide formation during degradation oforganic nitrates and activation of guanylate cyclase.Eur. J. Pharmacol. 139, 9–30.

3. Salter, M. and Knowles, R.G. (1989) Assay of NOSactivity by the measurement of conversion of oxyhe-

moglobin to methemoglobin by NO, in Nitric OxideProtocols (Titheridge, M.A., ed.), Humana Press,New Jersey, pp. 61–65.

4. Feelisch, M., Kubitzek, D., and Werringloer, J.(1996) The oxyhaemoglobin assay, in Methods inNitric Oxide Research (Feelisch, M. and Stamler, J.S., eds.),Wiley, New York, pp. 472, 473.

5. Gross, S.S., Jaffe, E.A., Levi, R., and Kilbourn, R.G.(1991) Cytokine-activated endothelial cells expressan isotype of nitric oxide synthase which istetrahydrobiopterin-dependent, calmodulin-depen-dent and inhibited by arginine analogs with a rank-order of potency characteristic of activatedmacrophages. Biochem. Biophys. Res. Comm. 178(3), 823–829.

6. Stuehr, D. J. and Griffith, O. W. (1996) Purification,assay and properties of mammalian nitric oxidesynthases, in Methods in Nitric Oxide Research(Feelisch, M. and Stamler, J. S., eds.), Wiley, NewYork, p. 185.

7. Stuehr, D. J. and Griffith, O. W. (1996) Purification,assay and properties of mammalian nitric oxidesynthases, in Methods in Nitric Oxide Research(Feelisch, M. and Stamler, J. S., eds.), Wiley, NewYork, pp. 183–184.

8. Knowles, R.G. and Salter, M. (1989) Measurementof NOS activity by conversion of radiolabelled argin-ine to citrulline using ion-exchange separation, inNitric Oxide Protocols (Titheridge, M.A., ed.),Humana Press, New Jersey, pp. 67–73.

9. Paterson, R.A., Eagles, P.A.M., Young, D.A.B., andBeddell, C.R. (1976) Rapid purification of large quanti-ties of human hemoglobin with low phosphate contentby counter-flow dialysis. Int. J. Biochem. 7 117, 118.

10. Charles, I. C., Scorer, C. A., Angeles Moro, M.,Fernandez, C., Chubb, A., Dawson, J., Foxwell, N.,Knowles, R. G., and Baylis, S. A. (1996) Expressionof the three human NO synthase isozymes. MethodsEnzymol. 268, 449–460

11. Komori, Y., Hyun, J., Chiang, K., and Fukuto, J. M.(1995). The role of thiols in the apparent activationof rat brain nitric oxide synthase (NOS). J. Biochem117, 923–927

12. Hofmann, H. and Schmidt, H. H. H. W. (1995). Thioldependence of nitric oxide synthase. Biochemistry 34(41), 13,443–13,452.

Documents

A microtiter-plate assay of nitric oxide synthase activity