Biochinuca et Biophysica Acta, 999 (1989) 233-238 233 Elsevier

BBAPRO 33503

Molecular characteristics of tke cholesterol oxidase and factors influencing its activity

F r a n c k Che i l l an t, H u g u e t t e L a f o n t ~, Efise T e r m i n e ~, F l o r e n c e F e r n a n d e z 2, Pau l Sauve 3 a n d G u y Lesga rds 2

i INSERM Unit~ 130, z Laboratoir¢ de Chimie des Produits Naturels, Facult~ des Sciences de St. Jerome, and J Centre de Biochimie et de biologic Moleculaire, Marseille (France)

(Received 18 May 1989)

Key words: Cholesterol oxidase; Bile salt; Electrochemistry; Enzyme activity

T h e purpose o f this study was to characterize Pseudomonas cholesterol oxidase (cholesterol:oxygen oxidoreductase, EC 1.1.3.6) and define factors influencing its activity with a view to perlorming electrochemical determinations of chelesterd. The enzyme wa; purified by gel filtration ~aul i~ molecular characteristics were determined. Its amino acid composition pcesents a high proportion of glutamic and aspartic acids. Its isoelectric point is between 5.0 and 5.5 u n i t / p H and its molecular mass is equal to 65.1 kDa, as determined by analytical ultracentrifugation. The influence of reaction predncts on enzyme activity was studied. The amount of hydrogen peroxide producr~ is not great enough to inhibit the enzyme. Adding exogenous fiavin adenine dinudeotide increases enzyme acti;,ity. Bile salts sUibilize cholesterol oxidase structure increasing its activity without changing its aff'mity for cholesterol.

lnmm'uction

Because cholesterol determination is of great impor- tance in medicine, numerous techniques have been de- scribed [1-3l. Currently the best method is enzymatic determination. All enzymatic n_ethods proposed until now are based on the specific oxidation of cholesterol into cholestenone by cholesterol oxidase (COD) [4]. One advantage of this technique is that it allows quantifica- tion not only of free cholesterol but also total cholesterol, i.e., free and esterified cholesterol, after addition of cholesterol esterase which hydrolyses cholesteryl esters [5]. Enzymatic methods are also more specific than purely chemical techniques and faster than cbdomato- graphic techniques [6].

The choice of the detection method to be used after enzymatic reaction depends on the nature of the bio- logical fluid tested. For example colorimetric methods cannot be used for cholesterol detection in bile which is a viscous and pigmented fluid [7]. Electrochemical de-

tection is more suitable because it takes advantage of the sp?cificity of the enzymatic reaction while avoiding the colorimetric interferences due to bile pigments. Among the different methods of electrochemical detec- tion, the oxygen electrode is particularly valuable [5,8], since it eliminates interferences from a number of dec_ troactive substances such as ascorbic acid [9].

Despite the key importance of COD in cholesterol determination, little information has been published about the characteristics of this enzyme and its mecha- nism of action. The aim of this report is to characterise COD and describe factors influencing its activity and kinetic parameters. The COD used in this study, which was obtain~J from Pseudomonas, was chosen because it gave good results in preliminary studies especially with respect to conservation. COD from different micro- organisms have been described by Noma and Nakayama [9].

Materials and Methods

COD, cholesterol oxydase; FAD, flavin adenine dinucleotide; SDS, sodium dodccyl sulfate; TDC, taurodcoxycholatc; C, cholate; TC, taurockolate; TDHC, ta&odehydrocholate.

Correspondence: F. Cl~llaa, INSERM Unit~ 130, 18 avenue Mozart, 13009 Marseille, France.

Reagents Cholesterol oxidase obtained from a species of Pseu-

dmnonas (Sigma Ch.mfical Co, St. Lom_~s, MO, U.S.A.) was used at the rate of 40 uuits/mg protein (1 unit converts 1.0 ~tmol of cholesterol into 4-cholestene-3-one per min at pH 7.5 and 250C). Aqueous cholesterol

0167-4838/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)

2M

calibrators obtained from serum (2.59, 5.]8 and 10.36 mM) and flavin adenine dinucleotide were also purchased from Sigma Chemical Co. Taurodeoxycholic acid, cholic acid, taurodehydrocholic acid and taurochoIic acid (sodium salt) were obtained from Calbiochem (La Jolla, CA, U.S.A.). All other chemicals and solvents were of at least reagent grade (ACS: Carlo Erba, Milan, Italy).

COD activity Oxygen consumption which is proportional to the

concentration of free cholesterol during enzymatic re, ac- tion was measured by means of an oxygen electrode (Beckman). The oxygen electrode is a polarographic electrode measuring the current limited by the diffvsion of oxygen through a teflon membrane between a rhodium cathode and a silver anode with a relative potential difference of -0.55 V. The cathode and the anode were connected by an electrolyte gel containing po.ta~sium chloride.

For cholesterol detection the electrode was immersed in a cell containing 5 nd of buffer solution (0.1 M phosphate buffer (pH 7.5), 0.05% THton X-100, 3 M NaCI) maintained at 37°C. The ceil was then placed on a magnetic shaker until current intensity stabilized, at which time 0.6 unit of cholesterol oxidase in 10 #l distilled water was added followed 1 min later by 50 tti of the test solution. For samples containing esterified cholesterol, 0.06 unit of cholesterol esterase in 10 pl distilled water was added to the cholesterol oxidase in order to convert cbolesteryi estt,~rs into free cholesterol. The reaction was completed within 5 rain and the response was read by means of a Cholesterol Analyser 11 (Beckman).

Gel filtration A 1.6 em x 45 em glass column was packed with

Ultrogel LKB Ac~-5, 5000-70000. A quantity of 500 pl of a solution contalnln$ 40 mg of COD (160 units) per ml of phosphate buffer 0.1 M (pH 7.5) and 1% of sodium azide was eluted with 0.1 M pho~'phate buffer (pH 7.5) Blue dextran and vitamin B-12 were used to determine the void volume (V0) and total volume (V,). The K^v of each fraction was calculated as follov~:

g x v " ( ~ - Vo)/(~ - Vo)

where ~ is the elution vo'ume i'O].

SDS polvacrylamide gel clectrop,,oresis Electrophoresis was performed in 15% polyacryl-

amide gel in the presence of 0.1~ sodium dodecyl sulfate (SDS) by the method of Laemmli [11]. The reference proteins were ribonuclease A (13.7 kDa), chymotrypsinogen (25 kDa), ovalbumin (45 kDa) and human albumin (69 kDa).

Eiectrofocalisation COD electrofocalisation was performed on LKB

Ampholine PAG plates at pH ranging from 3.5 to 9.5. The reference proteins were myoglobin (7.3), aldolase (6.0), albumin (4.7) and ferritin (4.4).

Amino acid analysis Amino acid analysis was performed after HCI hy-

drolysis of dry residues (5.6 M HCI, constant boiling, Pierce Chemical) with a 119 CL Beckman amino acid analyzer connected to a Spectra Physics Integrator. Performic oxidation was achieved by the method of Kimmel et al. [12].

Molecular mass determination Analytic ultracentrifugation was performed at 20°C

in a Model E Spinco-Beckman ultracentrifuge featuring speed and temperature controls. The sedimentatio:~ coefficient was measured using a double-sector capillary cell with Scl0ieren cptics. The Yphantis method [13] was used for the determination of molecular mass.

Results

Gel filtration SDS-polyacrylamide gel electroFhor,;,*is of th,- com-

mercial COD revealed a major band around 70.0 kDa and num,,rous :ninor bands of lower mt,'~u]at mass. COD activity corresponds to tk¢ major band. Proteins of low molecular masses were eliminated by eluting the commercial COD on an Ultrogel column. The elution profile is shown in Fig. 1. Three p~ t~ were observed with respective KAy values of 0.00, 0.21 and 0.57. The first peak eluted with the voiO ". o!ume corresponded to a high moleculaz mass protein not found on SDS poly- acrylamide gel. This hextion was able to oxidize cl',olesterol. The second peak was pure COD monomere with an estimated molecular mass of 70 kDa. This faction represents 30% of total proteins. The third peak contained proteins of lower molecular masses.

1.000~ 0.800~ 0.600~

o 2d 40 60 ml

Vo

I 1

8O

Fig. 1. Elution prof'de of PsetMomonas COD on Ultrogel AcA-5, 5000-70000.

TABLE 1

Amino acid composition of Pseudomonas COD

Values are calculat~'~d as residue number relative to the lowest amino acid value (HIS) which is fi~ed to 1.

Amino acid Residues

Asp 6 Thr 7 Set 3,7 Glu 6,S Pro 6 Gly 6 Ala 8 Cys 4 Vat 5,2 Met 3 lie 4 Leu 8.5 Tyr 2,8 Phe 3 His 1 Lys 1,3 Arg 6

?,,lole¢ular characteristics Assays were done at pH 7.5 with 2.5 mg ml- i of

purified COD in 0.1 M phosphate buffer. The symmetry of the peak obtained indicated substantial homogeneity of the preparation. Purified COD had a specific volume of 0.73 ml- mg -~ and a molecular mass of 65.1 ± 1.1 kDa. The sedimentation coefficient was 4.88 S. This coefficient fell to 4.73 when 0.5 mM taurodeoxycholate was added and to 5.52 S when 5.0 mM TDC was added.

T!,e isoelectric point of pure COD was between pH 5.0 and 5.5. Its amino acid composition is shown in Table I.

Detergents Bile salts which are structurally analoguous to

cholesterol are natural detergents. In order to study their effects on enzymatic activity, we determined the kinetic parameters of Pseudomonas COD. Results are shown in Table It.

The affi~Lity constant (Kin) and maximum velocity (Vm~) were determined in 0.1 M phosphate buffer and

235

,.,° F

0 0 0 / ~ ~ ' f I I I ! t t - 0 5 0.0 0,5 1.0 1.5 2.0 ~5

I / C ( M - I ) x 10 "'~

Fi 8. 2, Lincwcaver-Burk plot. Influence of the substrat¢ conocntration on COD activity. 0.1 M phosphate buffer, 3 M NaCi, 0.5% Triton X-100, 0.6 unit COl:), 50/~l 5.18 m M cholesterol standard, ra, without

bile salts; O, 2 raM TDC.

0.05% (v/v) Triton X-100 with 3.0 M NaCI (pH 7.5) and 40 ° C. Fig. 2 shows two Lineweaver-Burk plots: one drawn in the absence of bile salts (regression cozffi- cient = 0.997) and one drawn with ta,:rodeoxycholate (TDC) (regression coefficieitt = 0.979). The regy-~ion fires oi cholate (C) and taurocholate (TC) with respec- tive regression coefficients of 0.982 and 0.976 are suoer- imposable on that of taurodeoxycno, ate. Tauro- dehydrocholate (regression coefficient--0.971) gave a

regression line identical to the one obtained without bile salts. While enzyme affinity for the suL~trate was con- stant at 2.5.10 -5 M, Vm~ was 2-fok~ h~3c~ :vith than without bile salts reaching 10.5/xmol • rain -~ •mg - t in the presence of TDC. With increasing, co,.:entrations of TDC (0.5 mM to 5.0 raM), the regre~=3n curve was ahvays comparable with the one presented in Fig. 2. The degree of activation was )he same regardless of bile salt cor.centration. Only ta~,~odehvdrocholate failed to induce activation.

Time-dependent retention o f activity Whatever the cholesterol concentration used (2.59,

5.18 or 10.36 mM), COD activity was constant provided that the pH of the medium was maintained between 5.5 and 8.5.

Using immobilized COD involves that the enzyme remains for a long time in the reaction medium. With

TABLE Ii

Effect of different bile salts at dzfferent concentratons on COD kinetic parameters

Regression lines are calculated from Lineweaver-Burk p!ot 0.t M phosphate buffer 3 M NaCI, 0.5~o Triton X-100.

Straight line R-squared Km (M) Vm~ (It tool. rain- 1. rag- z )

Without bile salts 1 / V ffi 0 . 4 3 3 . 1 / C + 0.170 0.997 2 .5 ,10- s 5.88 Taurodehydrocholate (2 raM) 1 / V ~ 0 . 4 2 4 . 1 / C + 0.170 0.972 2 .5 ,10- s 5.88 Cholate (2 raM) 1 / V ffi 0 . 2 4 3 . 1 / C + 0.095 0.991 2.~. 10- ~ 10.52 Taurodeoxycholate (0,5 raM) 1 / V -- 0 . 3 0 1 . 1 / C + 0.081 0.984 2 .7 ,10- 5 12.30 Taurodeoxycholate (2 raM) 1 / V ffi 0.285. I / C + 0 .099 0.976 2 .5-10- 5 10.10 Taurodeoxycholate (5 mM) 1 / ( V ffi 0 . 3 0 8 . 1 / C + 0.082 0.985 2 .7-10- s 12.20

236

180

s .g

w

Incubation cloys

Fig. 3. pH stability of Pseadomomu COD. COD 0.6 unit, 50 /~i cholesterol sUmdard 5.18 mM. o, pH 5; O, pH 6; O, pH 7, O, pH

7.5; I pH ~, 0, pH 9.

this aim we studied the effect of this prolonged incuba- tion on the retention of activity.

To study the effects of pH on the conservation of COD activity, 0.8 unit of COD was incubated at 37 °C in solutions at pH 5.0, 6.0, 7.5, 8.0 and 9.0. At each pH value activity was measured three times at 24-h intervals on calibrated 50/~1 sample of cholesterol (5.18 mM). Results expressed relatively to the highest initial activity (100~) are shown in Fig. 3. At basic pH values, COD lost 80 to 100~ of its activity after 1 day of h~cubation. At acidic pH values, a loss of only 20 to 30~ was measured.

Temperature COD activity was measured under the conditions

previously described using 50 tAI of cholesterol standard 5.18 mM either without bile salts or with TDC 2 mM. These measurements were carried out at different tem- peratures between 25 and 50 ° C. The resulting curves are shown in Fig. 4 The regression coefficients ranged from 0.87 to 0.99, indicating that cholesterol oxidation obeys the Arrhenius equation.

The activation energies calculated from these data were 23.4 kJ. tool -~ without bile salts and 24.2 kJ.

'oL 5

~0 3.I 3.2 3.3 ~.,- I~ I/I' {K -I ). 10" j

Fig. 4..~rrhenius plot. Vnfluence of temperature on COD activity. 0.6 unit COD, 50/~I 5.18 mM cholesterol standard. ~ 2 mM TDC; iii,

without bile salts.

mol- ~ with TDC. Without bile salts a break appeared at about 37°C with a 20~ decrease of COD activity. When bile salts were used, this break did not occur.

Influence of the cofactor The spectrum of COD from Pseudomonas showed

two characteristic absorption bands of flavin adenine dinucleotide (FAD), i.e., 445 and 375 rim. Although FAD is covalently bound to the enzyme [14], addition of exogenous FAD to the buffer solution doubled COD activity. The same degree of stimulation was obtained regardless of the amount of FAD added (0.5, 1 and 10 rag- all-l).

Reaction products During cholesterol oxidation H202 and cholestenone

are ~deased in ~71::'~"~..olar ouantities. The effects of adding 30 vol. of H20, to the reacuon ~edium was tested: COD activity was 705 of the initial rate ¢,,hen the final concentration of H202 in the reaction medima was 5.89 mM, 37% for 11.8 mM H,O, and totally inhibited for 29.5 mM H20,.

Discussion

Commercial Pseudomonas cholesterol oxidase some- times contained minor proteins of low molecular mass. Depending on the application, for example in the case of protein immobilisation on support, this contamina- tion can be a disadvantage. Purification by gel filtration is a straightforward way of solving this problem.

In this study, elution profiles suggested that our commercial COD existed in two main forms. The first, a monomeric form, with a molecular mass of 65.0 kDa, was detected by both electrophoresis and gel filtration. This form accounted for 30~ of the total proteins in the preparation studied. The second, a polymeric form of the first, could only be detected by gel filuation, since it was destroyed by SDS for electrophoresis. This form represented 45~ of the total proteins.

With regard to the amino acid composition of COD, s~;veral findings are interesting. Among these is the imrot ~nce of the neutral amino acids (leucine, alanine, glycine). The presence of four free sulphated amino acids suggests that two disulfate bonds are involved in the tertiary protein structure. The higher number of acidic than basic amino acids, 12 vs. 8, is in good agreement with the isoelectric point of the COD which is 5.

Under our conditions the Michaelis constant is un- changed, regardless of whether bile salts are present in the reaction medium. The value of 2.5-10 -5 M is comparable to the one reported by Wollenberger et al. [15] and Blum et al. [16]. Bile salts do not modify the affinity of the enzyme for the substrate. Kinetic studies showing the effects of bile salts on COD activity explain

preliminary findings that the enzymatic reaction is fas- ter in bile (medium containing bile salts) than in serum. Indeed Vma ~ doubles in the presence of bile salts. Since this effect was observed with all but one of the bile salts tested, it can be concluded that degree of hydroxylation or conjugation type do not modify the activation. Only taurodehydrocholate which is not in micellar form did not induce stimulation, presumably because the critical micellar concentration (CMC) is too high [17]. Bile salts facilitate the incorporation of cholesterol into mixed micelles. Since cholesterol in micellar form is a better substrate for COD, bile salts stimulate the COD activ- ity.

The disappearance of the break noted around 37 °C on the Arrhenius plot with bile salts [18] shows that the micelles have a stabilizing effect on the COD structure. This observation reinforces the hypothesis of a COD-bi le salt micelle association. It is also important to note that only bile salts in the micellar form have this effect. This enzyme-bile salt association, while not modifying the tertiary structure of the protein, could give it a better access to its substrate and better stability in regard to temperature variations.

The fact that the degree of activation was the same no matter what concentration of bile salts was used (up to the CMC) suggests that this is a 'hit-or-miss' process in which as soon as the bile salts are added to the medium and cholesterol micelles are formed, COD ac- tion is enhanced and the reaction rate thus increases. This situation is different from the one described with cholesterol esterase [19] in which bile salts binding to the enzyme modify its affinity for the substrate with no change in enzyme activity.

~'he increase in the sedimentation coefficient ob- tained by adding bile salts to the medium indicates that this protein associates with one or more bile salt micelles. However, the coefficient value with 5 mM TDC is too low to assert that there is a protein-protein association and thus to conclude that COD polymerisation takes place. Nevertheless this ultracentrifugation study sug- gests that COD is associated with bile salt micelles and that this association increases enzyme activity without changing its affinity for the substrate. Fu~hcrmore, the fact that the activation energy of the enzymatic reaction is the same with our without bile salts implies that the catalytic constant of the enzyme is not modified. Ap- parently in the presence of bile selt micelles, COD presents a structure that facilitates the reaction with its substrate without changing the active site. Thus bile salt micelles do not act as a COD activator.

In contrast with Noma's observations [9] on COD from Nocordia, Brevibacterium, Sckizopkyl!um and Streptomyce$, Pseudomas COD is remarkably stable with respect to pH. However, pH is very important to the conservation of this protein. The most favorable pH is near the isoelectric point. This finding could be of value

237

especially in treatments aimed at immobilizing the en- zyme.

Like many oxidation enzymes, COD acts in associa- tion with a cofactor which ensures the transfer of elec- trons and hydrogen from cholesterol to oxygen which is the acceptor. As shown by Keney and Singer [14], the oxidoreduction cofactor of Pseudomonas COD is FAD. Although FAD is bound to enzyme, additional cofactor is needed to achieve maximum em.yme activity. How- ever, the increase in enzyme activity is not proportional to FAD concentration.

The specificity of COD for cholesterol is well docu- mented by Ikawa et al. [20] who drew up a list of cholesterol derivatives able to interfere in the enzymatic reaction. All these derivatives are characterized by a hydroxyl function in position/~ in relation to the cyclic nuclei. H202 which is produced during enzymatic reac- tion and serves as a basis for many colorimetric detec- tion techniques [4] and some electrochemical detection techniques [16], is an inhibitor of COD [21]. However, the amount of H202 required to produce strong inhibi- tion is much higher than the amount produced by the reaction under our conditions. As shown by Lee and Biellemann [21] zl4 cholestenone has no effect on the enzymatic reaction.

Hopefully, a better understanding of the structure and activity of Pseudomonas cholesterol oxidase will optimize the use of this protein in current cholesterol assay techniques, especially in complex biological media. We are attempting to use these results to develop a biomedical or industrial bioreactor based on the im- mobilisation of cholesterol oxidase.

References

1 Morin, RJ. (1976) Ciin. Chim. Acta 71, 75-80. 2 Roda, A., Festi, D., Sama, C., Mazzella, G., Aidini, R, Roda, E.

and Barbara, L. (1975) Clin. Chim. Aeta 64, 337-341. 3 Carlson, S.E. and Goldfarb, S. (1977) Clin. Chim. Acta 79,

575-582. 4 Fromm, H., Amin, P., Klein, H. and Kupk¢, I. (1980) J. Lipid Re,.

21, 259-261. 5 Noma, A. and Nakayama, K. (1976) Clin. Chem. 22, 336-340. o Haeckel, R., Somag, G.. Kiilpmann, W.R. and Feldmann, U.

(1979) J. Clin. Chem. Clia. Biochem. 17, 553-563. 7 Boltoth C.H., Nicholls, J.S. and He, aton, K.W. (1980) Ciin. Chim.

Acta 105, 225-230. 8 Clark, L.C., Duggan, C.A., Grooms, T.A., Hart, L.M. and Moore,

M.E. (1981) Clin. Chem. 27, 1978-1982. 9 Noma, A. and Nakayama, K. (1976) Clin. Chim. Aeta 73, 487-496.

10 Laurent, T. and Killander, J. (1964) J. Chromatog. 14, 317-330. 11 Laernmli, U.K. (1970) Nature 227, 680-685. 12 Kirnmel, J.IL, Kato, G.K., Paira, A.C.M. and Smith, E.L. (1962) J.

Biol. Chem. 237, 2525-2534. 13 Yphantis, D.A. (19t,4) B;~ochemistry 3, 297-317. 14 Kenney, W.C. and Singer, T.P. (1979) J. Biol. Chem. 254,

4689-4690. 15 Wollenberger, U., Khlin, M. and Sheller, F. (1983) Bio¢lec. Bio-

eng. 11, 307-317.

23~

16 Bt~m, L,2,. Bercmnd, C. j ~ l C~k t , p,9., (19~3) & ~ l . ~ l t , ]6. 1~ Lombar.~o. I). ~md Dennis, ~.A. I ] ~ ' ) J. J~o]. (~m~. 260. 5Z~-5~]. 16114-16121.

I? C~-),. F~.C. (19~$) $~:r#1~ ~nd ~1¢ Acid_~ (l[laniel~,,~r4 H. ~,d ~ I~:m~m, S,. ]-akili. 7vJ. mid ~,ura~ M, (1979) I , ~a~llcm. KS, S/~rall. 3., ~la.), ]Ebc'.~r ~ n ~ PubiJzJtcrs J~.Y. (BJcznneclJcal 1#.47-]4~2, Oisiek~a). ~tS-~]~. ~.1 Z,~#, K.M, a~d BicL]menn, J.F. (~i9~L,] I~;oo# 8. Chcn~ 14, 262-273.

~S I.ml~l~;~o, D., 13,y. CI. e~l Fi~;e,hr]l,~ C. (I978) ~ [ ~ ~ophy~. .~CL~ 5:Z?. 142-149.