T

ftiha

dp

8

Analytical Biochemistry 291, 84–88 (2001)doi:10.1006/abio.2001.5005, available online at http://www.idealibrary.com on

Assays for Angiotensin ConvertingEnzyme Inhibitory Activity

Bin-Wha Chang,*,† Richie L. C. Chen,‡,1 I-Jen Huang,§ and Hsien-Chang Chang†*Department of Healthcare Administration, Hungkuang Institute of Technology, Taichung, Taiwan, Republic of China;‡Department of Bio-Industrial Mechatronics Engineering, College of Agriculture, National Taiwan University, Taipei,Taiwan, Republic of China; §Department of Products & Process Development, Taiwan Sugar Research Institute,

ainan, Taiwan, Republic of China; and †Institute of Biomedical Engineering, College of Engineering,National Cheng-Kung University, Tainan, Taiwan, Republic of China

Received August 21, 2000; published online March 7, 2001

tmtoasttssp

A colorimetric method and a capillary electrophore-sis procedure were developed for quantifying histidyl-leucine and hippurate, respectively. The colorimetricmethod is sensitive (extinction coefficient 5 7.5 mM21

cm21) and reproducible (CV 5 1.7%, n 5 5), which isbased on a selective chromogenic reaction for histidyl-leucine (lmax 5 390 nm) using o-phthalaldehyde. Forsamples containing unusually high levels of histidineand/or histidyl peptides, the separation-based ap-proach is preferable. The capillary electrophoresismethod makes use of an in-capillary microextractiontechnique; complicated samples can be measured inless than 4 min without pretreatment. Protocols usingboth methods to measure angiotensin converting en-zyme inhibitory activity were proposed. © 2001 Academic

Press

Key Words: angiotensin converting enzyme; histidyl-leucine; hippurate; o-phthalaldehyde; capillary elec-trophoresis; microextraction.

Angiotensin converting enzyme (ACE)2 is importantor the control of blood pressure and sodium ion excre-ion (1); its synthetic inhibitors are the drug of choice tonitiate a stepped-care therapy for mild-to-moderateypertension (2). The enzyme is relatively nonspecific,nd several foods (3, 4) and traditional medicines (5)

1 To whom correspondence should be addressed at Department ofBio-Industrial Mechatronics Engineering, College of Agriculture, Na-tional Taiwan University, 136 Chou-San Road, Taipei, Taiwan, Repub-lic of China. Fax: 1886-2-23627620. E-mail: rlcchen@ccms.ntu.edu.tw.

2 Abbreviations used: ACE, angiotensin converting enzyme; HHL,hippuryl-L-histidyl-L-leucine; HL, L-histidyl-L-leucine; OPA, o-phthala-ehyde; TNBS 2,4,6-trinitrobenzene sulfonate; CE, capillary electro-

horesis; MEKC, micellar electrokinetic chromatography.

4

have recently been noted to have ACE-inhibitory activ-ity.

Several methods for quantifying ACE-catalyzed re-action were developed for screening both synthetic andnaturally occurring ACE inhibitors. Most studies useda synthetic peptide, hippuryl-L-histidyl-L-leucine(HHL), to simulate the reaction (6); either the hippuricacid or the L-histidyl-L-leucine (HL) released by theaction of ACE was used to quantify the following sim-ulated reaction:

hippuryl-histidyl-leucine 3

histidyl-leucine 1 hippurate. [1]

Since determination of hippurate involves severaledious steps including extraction and evaporation (7),ethods based on measuring the other parameter, his-

idyl-leucine, were considered. The method utilizing an-phthalaldehyde (OPA) fluorogenic reaction (8) gener-ted precipitation that must be removed before mea-urement. The colorimetric methods using 2,4,6-trini-robenzene sulfonate (TNBS) are simpler (9, 10), buthe specificity for HL is not enough for a complicatedample (11). The methods proposed here are simple,elective, and reliable, which will ease the screeningrocess for ACE inhibitors.

MATERIALS AND METHODS

Chemicals and Tested Materials

Angiotensin I converting enzyme (EC 3.4.15.1.; fromrabbit lung; 4 units per mg protein), hippuryl-L-histi-dyl-L-leucine, L-histidyl-L-leucine, and o-phthalalde-hyde were purchased from Sigma Chemical Co. So-

dium hippurate, hippuric acid, benzoic acid, sodium

0003-2697/01 $35.00Copyright © 2001 by Academic Press

All rights of reproduction in any form reserved.

fiu

1trw1

A

w

d2vwmpfot[w

85ANGIOTENSIN CONVERTING ENZYME INHIBITORY ASSAYS

benzoate, 2-mercaptoethanol, and sodium borate deca-hydrate (borax) were from Nacalai Tesque Co. Allchemicals were of analytical grade and used as re-ceived.

Spray-dried powders of the mycelia of Cordyceps si-nensis were obtained from Taiwan Sugar Research In-stitute (16, 17). The fungus was submerged-cultivatedin a 6-kL stirred-tank fermenter. Typically 2 g of thepowders was extracted with 100 ml of hot water(65°C 3 1 h). The extracts of several batches of themycelia cultivated in different fermentation broth weretested for their ACE-inhibitory activities.

Reagents

Buffer A (pH 8.3 with HCl) contained 20 mM sodiumborate and 0.3 M NaCl. Buffer B (pH 12.0 with NaOH)was composed of 0.1 M sodium borate and 0.2 MNaOH. OPA reagent was prepared at least 1 h beforethe experiment by mixing 1.5 ml of OPA solution (10mg/ml in ethanol) and 1.5 ml of 2-mercaptoethanolsolution (5 ml/ml in ethanol) in 100 ml of Buffer B. Thefinal concentrations of OPA and 2-mercaptoethanol inthe OPA reagent were both about 1 mM. ACE solutionand HHL solutions were freshly prepared, respec-tively, by dissolving ACE (40 mU/ml) and HHL (15mM) with Buffer A.

Capillary electrophoresis (CE) running buffer was a30 or 20 mM sodium borate solution (pH 9.3). Benzoateinternal standard is an 80 mg/L (0.556 mM) solution ofsodium benzoate in the CE running buffer. All solu-tions were prepared with deionized water.

CE Instrumentation and Separation Conditions

Capillary electrophoresis was performed with a com-puter-interfaced CE machine (Beckman P/ACE 5000)equipped with an UV detector (214 6 10-nm bandpass

lter) and a thermal controlling cartridge (25°C). Anncoated fused-silica capillary (75 mm i.d. 3 27 cm

length) was used; the length from the detecting win-dow (100 mm width 3 200 mm length) to the outlet was7 cm. Prior to the experiment, the capillary was rinsedwith the alkaline CE running buffer (20 psi 3 1.5 min)to charge the capillary surface. Immediately before theseparation, a segment of sample solution (2.5 mm 3 75mm, ca. 11 nl) was introduced into the inlet end of thecapillary by pneumatic injection (typically 0.5 psi 3

s). Separation was started by imposing 10 kV acrosshe capillary (inlet to outlet). At the end of each sepa-ation, the capillary was rinsed (20 psi) subsequentlyith 0.1 N NaOH for 0.5 min and deionized water for.5 min.

CE-Inhibitory Assay Procedures

Sample solutions were diluted to different extents

ith Buffer A for the ACE-inhibitory assays. The ACE-

catalyzed reactions (37°C 3 2 h) were performed in testtubes (10 mm i.d. 3 5 cm) containing 100 ml of samplesolution, 100 ml of ACE solution, and 100 ml of HHLsolution (Mixture 1). Another mixture containing 100ml of sample solution and 200 ml of Buffer A (Mixture 2)was used to obtain the background absorbance of thesample solutions for the colorimetric method. The thirdmixture containing 100 ml of Buffer A, 100 ml of ACEsolution, and 100 ml of HHL solution (Mixture 3) wasused to obtain the data for 100% reaction. The fourthmixture containing 300 ml of Buffer A (Mixture 4) wasused to obtain the background absorbance of the OPAreagent. The enzymatic reactions (working pH 5–10)were terminated by adding either 3 ml of the alkaline(pH 12.0) OPA reagent (for the OPA method) or 1.0 mlof 1 N HCl (for the CE method).

For the OPA method, A 390s of Mixture 1 (A1), Mix-ture 2 (A2), Mixture 3 (A3), and Mixture 4 (A4) weremeasured after 20 min of 25°C incubation. The inhib-itory ratios were calculated by the following equation.

I~%! 5 @1 2 ~A1 2 A2!/~A3 2 A4!# 3 100. [2]

For the CE method, 1 ml of benzoate internal stan-ard was added before the extraction. Extraction withml of ethyl acetate was performed by the aid of a

ortexing mixer in the same test tube, and the mixtureas then transferred into a sample vial for the CEachine. The upper organic phase served as the sam-

le solution for the CE measurement. The peak areasor hippurate (P1 for Mixture 1 and P3 for Mixture 3)r their ratios to the internal standard were used ashe parameters to calculate the inhibitory ratios (Eq.3]). Typically, 20 mM sodium borate solution (pH 9.3)as used as the running buffer.

I~%! 5 ~1 2 P1/P3! 3 100. [3]

RESULTS AND DISCUSSION

OPA-Chromogenic Reaction for Histidyl-Leucine

Figure 1a shows the pH dependency of the OPA(o-phthalaldehyde; ca. 1 mM) chromogenic reactionwith 0.1 mM of HL in the presence of 2-mercaptoetha-nol (ca. 1 mM) as the reducing agent. The pHs werecontrolled by 0.1 M borate buffer. At the usual pH (pH,10.0) for OPA reaction (12), reaction of HL with theOPA reagent was slow, and the absorption spectrum ofthe OPA-HL adduct was similar to those obtained withother primary amines (lmax 5 340 nm). The reaction istherefore nonspecific for HL. As pH . 11.0, HL reactedwith the OPA reagent in a different manner and theresulting reaction mixture turned yellowish. At pH12.0 (Buffer B), the spectrum of OPA-HL shifted to the

visible region (lmax 5 390 nm), while the spectrum of

f

OwtoaFbb

86 CHANG ET AL.

OPA-glutamate remained unchanged (Fig. 1b). TheOPA-HL adduct was light-stable, which permittedreal-time monitoring of the chromogenic reaction (un-published data). The reaction was finished after 20 minof incubation (25°C in Buffer B), and the apparentmolar extinction coefficient (l 5 390 nm) of theOPA-HL adduct was 7.5 mM21 cm21. The detectionlimit of HL was 8 mM.

The background absorbance of a freshly preparedOPA reagent was low ( A 390 , 0.01) and its reactionwith HL was linear (r 5 0.997) up to 0.3 mM. The CVof A 390 obtained by five successive experiments with 0.1mM HL was 1.7%. The activity of the OPA reagent (pH12.0) was stable at room temperature and normal lab-oratory-lighted conditions for at least 24 h. The back-ground absorbance (l 5 390 nm) of an OPA reagentprepared 1 day earlier was 0.015, and the apparentmolar extinction coefficient of the OPA-HL adduct wasreduced to ca. 5.7 mM21 cm21. The A 390s originatedrom ACE and HHL were negligible.

2-Mercaptoethanol is essential for the statedPA-HL chromogenic reaction. As the reducing agentas omitted in the OPA reagent, the reaction condi-

ions were similar to the reported fluorometric meth-ds for histidyl peptide (12) and histamine (13). Thebsorption maximum was at 450 nm (the side peak inig. 1b). However, the 2-mercaptoethanol-omitteduffered reagent (pH 12.0) was not stable and should

FIG. 1. Spectrometric properties of OPA/2-mercaptoethanol-medi-ated reactions. (a) pH dependency of the chromogenic reaction withHL. (b) Absorption spectra of OPA-HL (solid curve) and OPA-gluta-mate adducts (dash curve) at pH 12.0. The reaction mixture (15min 3 25°C) contained 1 mM o-phthalaldehyde (OPA), 1 mM 2-mer-captoethanol, 0.1 mM L-histidyl-L-leucine (HL) or glutamate, and 0.1M sodium borate for pH control.

e used within 30 min. Actually, when a freshly pre-

pared 2-mercaptoethanol-containing OPA reagent (pH12.0) was used, the side peak was higher than thoseobtained with the OPA reagent prepared a few hoursearlier. Therefore, it is recommended that the sidereaction be reduced by setting aside the reagent forhours. A 450 can also be used as a parameter to measureHL, but the sensitivity, linearity, and CV (4.3%, n 5 5)were worse than A 390.

Electrophoretic Separation of ACE-CatalyzedReaction Mixture

To verify the analytical results obtained by the OPA-mediated chromogenic method, a CE-based separationmethod was developed for analyzing HHL, HL, andhippurate. Figure 2a shows the electropherogram of0.25 mM HHL, HL, and hippurate; the separation wasbetter and faster than a CE approach reported recently(14). The migration time for hippurate was 2.85 min(CV 5 0.13%, n 5 7); the peakwidth was 3.4 s. How-ever, the method still cannot be used to screen ACE-inhibitory compounds from biological samples such as

FIG. 2. Electropherograms ( A 214) of (a) a standard solution con-taining 0.25 mM hippuryl-histidyl-leucine (HHL), histidyl-leucine(HL), and hippurate (H) in the CE running buffer (b and c) anACE-catalyzed reaction mixture extract. The capillary electrophero-grams were obtained by injecting (0.5 psi 3 1 s) a segment (ca. 11 nl)of the above solutions into an uncoated fused-silica capillary (27cm 3 75 mm, 10 kV inlet to outlet) prefilled with (a and b) 30 mMsodium borate or (c) 20 mM sodium borate as the running buffer (pH9.3). ACE-catalyzed enzymatic reaction conditions and the extrac-tion procedures were detailed in the text. EA, ethyl acetate; B,

benzoate.

87ANGIOTENSIN CONVERTING ENZYME INHIBITORY ASSAYS

fermentation broth and traditional medicines. Biologi-cal samples often contain various zwitterionic com-pounds such as amino acids and peptides, and thecompounds will complicate the electropherogram. Thefollowing approach was developed for removing theproblematic compounds from biological sample solu-tions.

Typically 1 ml of sample solution was first acidifiedby adding 1 ml of 1 N HCl(aq) and then extracted with 2ml of ethyl acetate under vortexing. After minutes of aspontaneous phase-separation process, over 85% ofprotonated hippurate was extracted into the upperethyl acetate layer (Eq. [4]), while the interfering ma-terials (along with HHL and HL) were retained in thelower aqueous layer.

Hippurate 1 H1~Cl2! 3

hippuric acid in ethyl acetate. [4]

The resulting mixture was transferred directly into asample vial for the CE machine. Since the samplingprocesses of most CE machines involve only the mate-rial in the upper region of their sample vials, the sub-sequent manual phase-separation process was unnec-essary. A segment (2.5 mm 3 75 mm) of the ethylacetate solution of hippuric acid (11 nl) was pneumat-ically injected (0.5 psi 3 1 s) into the separation capil-lary, and the protonated hippuric acid was extractedback into the alkaline water phase, the CE buffer, anddeprotonated so that

hippuric acid in ethyl acetate 1 OH2 3

hippurate 1 H2O. [5]

The in-capillary microextraction procedure was per-formed online and almost simultaneously with theelectrophoretic separation; the dynamic process ren-dered the unusually high extraction efficiency. Over93% of hippuric acid was extracted from the sampleplug, which was calculated from the ratio of peak areasobtained by injecting an equal amount (0.1 mM each) ofhippuric acid in ethyl acetate and in the CE buffer (asthe marker for 100% extracted). Since the solubility ofethyl acetate in water (7.7% w/w at 25°C) is the highestamong regular hydrophobic solvents, the partition ofsolutes between water and ethyl acetate phases is thefastest. Ethyl acetate was tested as the best solvent forthe online in-capillary microextraction approach.

Compared with the electropherogram (Figs. 2a and2b) obtained by injection of aqueous hippurate solu-tion, the migration time was slightly longer (2.93 minfor hippurate, CV, 0.15%, n 5 7) and the resolutionwas better (peakwidth 5 2.1 s). Although the injected

organic sample plug may disturb the electroosmotic

flow, no significant change of the electrophoretic cur-rent (88 mA for both Figs. 2a and 2b) was observed andthe migration times were similar. Recently, Zhan et al.documented a similar microextraction approach (18)based on injecting an additional NaOH (0.1 M) seg-ment adjacent to the ethyl acetate sample segment,and analytes were separated in MEKC (micellar elec-trokinetic chromatography) mode. Zhan et al. statedthat saponification of ethyl acetate under alkaline con-ditions may help to establish the current at the earlystage of electrophoresis. The exact mechanism is cur-rently under our investigation.

With respect to the peakwidths of the electrophero-grams, sample dispersion restricted by the organic/aqueous interface may partially explain the higher res-olution obtained by the microextraction approach. TheCVs for the peak area of hippurate and its ratio to theinternal standard were 4.6% (n 5 7) and 2.1% (n 5 7),respectively. Internal standard strategies will be effec-tive for a CE machine with poor reproducibility. For alonger injection (0.5 psi 3 2 s), the extraction efficiencywas worse (about 70%), but the precision of the pneu-matic injection and thus the CV (3.1%, n 5 7) for thepeak area were better. An injection interval longerthan 2.5 s was not able to obtain a stable electroosmoticflow and the corresponding electropherogram, whichwas similar to the reported observation (18). Samplethroughput can be further increased by reducing theionic strength of the CE running buffer (15). By using20 mM sodium borate as the CE running buffer, themigration time for hippurate was reduced to 2.44 min(Fig. 2C). The electrophoretic current was 50 mA. Thedynamic range for quantifying hippurate by the CE-microextraction method was 10 mM to 0.5 mM (linear;r 5 0.992).

ACE-Inhibitory Assays

Several reported methods for quantifying ACE-cata-lyzed reactions were tried by the authors but failed toinvestigate the ACE-inhibitory activities of hot waterextracts of cultivated Cordyceps (16, 17). The photo-metric methods are based on nonselective chromogenicreactions with primary amines using OPA (10) andTNBS (12); the resulting high background absorbanceof the complicated samples restricted the dynamicrange of the measurement. The proposed colorimetricmethod is selective for histidyl-leucine, which will bemore preferable for an untreated biological sample.The lowest measurable Vmax of the enzyme was 0.08mM/min.

With respect to the methods for quantifying hippu-rates, extraction-based methods (7) were laborious andpoor in reproducibility. CE-based methods are gener-ally more rapid and economic than LC-based methods.

Unfortunately, ACE-catalyzed reaction is highly chlo-

2

1

1

1

1

1

1

1

1

1

88 CHANG ET AL.

ride-dependent (7) and therefore the ionic strength ofthe enzymatic reaction mixture is usually too high fora normal CE separation. Before a CE measurement,the reaction mixtures should be desalted by chromato-graphic methods or by using an electrodialyzer. Thereported CE method (14) used a running buffer con-taining 0.2 M sodium borate resulted in a high electro-phoretic current and thus an unwanted peak broaden-ing via joule-heating effect (15). The proposed CE-microextraction method used 20 mM borate buffer; thein-capillary microextraction procedure desalted thesample solutions and removed the interfering materi-als in the same time.

A parallel experiment was conducted to measure theACE-inhibitory activities of hot water extracts ofspray-dried powders of cultivated mycelia of C. sinen-sis; the correlation coefficient of the inhibitory ratios(I%) obtained by the two novel methods was 0.985 (n 5

5).

CONCLUSIONS

An automatic, high-throughput bioanalytical methodis needed for screening naturally occurring physiolog-ically active materials and for a combinatorial ap-proach for new drug development. The OPA-chromo-genic method was specific to the reaction product ofangiotensin converting enzyme, histidyl peptide; thereproducibility, the stability of the reagent and thesimplicity of the one-stroke reaction operation furtherexpand its practical use. The CE-microextractionmethod was rapid and selective; the in-capillary micro-extraction approach provided a useful online methodfor removing interfering peaks from the electrophero-grams. We currently attempted to use the microextrac-tion approach to determine the contents of benzoate-type preservatives in soy sauce and to measure liverbiotransformation function (conversion of benzoate tohippurate), the “H test.”

By virtue of the selectivity, simplicity, and rapidity,the authors strongly recommend researchers to auto-mate the proposed methods as a routine clinical diag-nostic assay for humoral ACE activity and a high-throughput method for screening effective ACEinhibitors.

ACKNOWLEDGMENT

The authors thank the National Science Council, Republic ofChina, for the financial support under Grant NSC 89-2313-B-002-

247.

REFERENCES

1. Ulm, E. H., and Greenlee, W. J. (1989) Inhibitors of the renin-angiotensin system enzymes. In Design of Enzyme Inhibitorsas Drugs (Sandler, M., and Smith, H. J., Eds.), pp. 146–177,Oxford, UK.

2. Hollenberg, N. K. (1992) Treatment of hypertension: The place ofangiotensin-converting enzyme inhibitors in the nineties. J. Car-diovasc. Pharmacol. 20(Suppl. 10), S29–S32.

3. Arai, S. (1996) Studies on functional foods in Japan—State of theart. Biosci. Biotechnol. Biochem. 60, 9–15.

4. Okamoto, A., Hanagata, H., Matsumoto, E., Kawamura, Y., Koi-zumi, Y., and Yanagita, F. (1995) Angiotensin I converting en-zyme inhibitory activities of various fermented foods. Biosci.Biotechnol. Biochem. 59, 1147–1149.

5. Morigiwa, A., Kitabatake, K., Fujimoto, Y., and Ikekawa, N.(1986) Angiotensin converting enzyme-inhibitory triterpenesfrom Ganoderma lucidum. Chem. Pharm. Bull. 34, 3025–3028.

6. Piquilloud, Y., Reinharz, A., and Roth, M. (1970) Studies on theangiotensin converting enzyme with different substrates. Bio-chim. Biophys. Acta 206, 136–142.

7. Cushman, D. W., and Cheung, H. S. (1971) Spectrophotometricassay and properties of the angiotensin-converting enzyme ofrabbit lung. Biochem. Pharmacol. 20, 1637–1648.

8. Friedland, J., and Silverstein, E. (1976) A sensitive fluorimetricassay for serum angiotensin-converting enzyme. Am. J. Clin.Pathol. 66, 416–424.

9. Fields, R. (1972) The rapid determination of amino groups withTNBS. Methods Enzymol. 25, 464–468.

0. Matsui, T., Matsufuji, H., and Osajima, Y. (1992) Colorimetricmeasurement of angiotensin I-converting enzyme inhibitory ac-tivity with trinitrobenzene sulfonate. Biosci. Biotechnol. Bio-chem. 56, 517–518.

1. Hazra, A. K., Chock, S. P., and Albers, R. W. (1984) Proteindetermination with Trinitrobenzene sulfonate: A method rela-tively independent of amino acid composition. Anal. Biochem.137, 437–443.

2. Roth, M. (1971) Fluorescence reaction for amino acids. Anal.Chem. 43, 880–882.

3. Shore, P. A., Burkhalter, A., and Cohn, V. H., Jr. (1959) Amethod for the fluorometric assay of histamine in tissues.J. Pharmacol. Exp. Ther. 127, 182–186.

4. Zhang, R.-Z., Xu, X.-H., Chen, T.-B., Li, L., and Rao, P.-F. (2000)An assay for angiotensin-converting enzyme using capillary zoneelectrophoresis. Anal. Biochem. 280, 286–290, doi:10.1006/abio.2000.4535.

5. Kuhn, R., and Hoffstetter-Kuhn, S. (1993) in Capillary Electro-phoresis: Principles and Practice, pp. 37–101, Springer-Verlag,Berlin.

6. Chen, R. L. C., Wang, Z. S., and Yang, J. P. (1998) Determinationof adenosine and cordycepin contents in cultivated Cordyceps bymicellar electrokinetic capillary chromatography. Rep. TaiwanSugar Res. Inst. 161, 49–56.

7. Liu, J.-F., Wu, Y.-J., Yang, J. P.-W., Chen, R. L. C., and Shieh,M.-J. (1999) Evaluation of food safety and physiological effect ofcultured mycelium of Cordyceps sinensis. New Taipei J. Med. 2,107–112.

8. Zhan, W., Wang, T., and Li, S. F. Y. (2000) Coupling of solventsemimicroextraction with capillary electrophoresis using ethyl

acetate as sample matrix. Electrophoresis 21, 573–578.