Vol. 188, No. 2, 1992
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 898-904
October 30, 1992
PREFERENTIAL CYCLIZATION OF 2,3(S):22@),23-DIOXIDOSQUALENE BY MAMMALIAN 2,3-OXIDOSQUALENE-LANOSTEROL CYCLASE
Olivier BOUTAUD, Daniele DOLIS and Francis SCHUBER* Laboratoire de Chimie Bioorganique (CNRS URA-1386), FacultC de Pharmacie, 74 route du Rhin, 674~Illkirch, France Received
September
14,
1992
Kinetic studies on the cyclization of 2,3(S)-oxido and 2,3(S):22(S),23-dioxido[14C]squalene catalyzed by liver oxidosqualene-lanosterolcyclase revealed a specificity (in terms of V/Km) of the enzyme for the diepoxide. The specificity ratio was dependent on the enzyme preparation, i.e. purified or m icrosomal, and was highest (about 5) with the m icrosomal enzyme in the presenceof supernatant protein factors. These results explain why, in the presenceof cyclase inhibitors, the squalene epoxides can be channeled into a cholesterol biosynthesis regulatory pathway via 24(S),25-epoxylanosteroland 24(S),25epoxycholesterol. 0 1992 Academic Press, Inc.
In mammalian cells, inhibitors of 2,3-oxidosqualene-lanosterolcyclase (EC 5.4.99.7) affect very efficiently the biosynthesis of cholesterol. Inhibition of the cyclase results in the accumulation of 2,3(S)-oxidosqualene(OS) and also of 2,3(S):22(S),23-dioxidosqualene (DOS) 1l-31. Recently, it was shown that this diepoxide is formed by epoxidation of OS by purified squalene epoxidase 141.It was demonstrated that DOS is further converted into 24(S),25-epoxylanosterol and ultimately into 24(S),25-epoxycholesterol 151, which is a known repressor of HMG-CoA reductase [6,71,the main regulatory enzyme of cholesterol biosynthesis. Indeed, treatment of mammalian cells with cyclase inhibitors leads to a decrease of HMG-CoA reductase activity which was interpreted as resulting from the formation of regulatory oxysterols [3,6]; the repression is dependent, however, on the extent of inhibition. i.e. at high inhibitor concentrations - when presumably the cyclase is totally inhibited - the activity of HMG-CoA reductase seems unaffected 161. From a pharmacological point of view, the cyclaserepresents an attractive target for the design of specific hypocholesterolemicagents181;at a cellular level the cyclaseinhibitors m ight act synergistically: i) by direct reduction of the metabolic flow leading to cholesterol and ii) by repressing HMG-CoA reductase via the formation of 24(S),25-epoxycho-
*To whom correspondence
should
be addressed.
Abbreviurions: HMG, 3-hydroxy-3-methylglutaric acid; U18666A, 38(2-(diethylamino) ethoxy]androstJ-en-17-one; OS, 2,3-oxidosqualene; DOS, 2,3:22,23-dioxidosqualene, LN, lanosterol; ELN, 24,25-epox lanosterol; SPF, supematant protein factor; S,, supernatant fraction obtained at 1OJxg. 0006-291X/92 $4.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
898
Vol.
188,
No.
2, 1992
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH
COMMUNICATIONS
lesterol. Such a synergism could explain why cyclase inhibitors are more efficient at a cellular level than anticipated from their inhibition constants determined in vivo on the microsomal enzyme 131.The regulatory pathway uncovered by the cyclase inhibitors, i.e. formation of 24(S),25epoxycholesterol, will be relevant, however, only if DOS proves to be an excellent substrate of the cyclase. In this study we have compared the specificity of the cyclase towards OS and DOS considered as competing substrates;the results indicate that 2,3(S):22(S),23-dioxidosqualeneis the preferred substrate.
MATERIALS
AND
METHODS
Sodium [2-14C]acetate(48 mCi/mmol) was purchased from NEN-DuPont de Nemours (France). [3-3H12,3-oxidosqualeneand U18666A were kind gifts from respectively Drs. M. Cerutti and L. Cattel (Torino, Italy) and Dr. R.J. Cenedella (Kirksville, MO, USA). 2,3-Oxidosqualene and 2,3:22,23-dioxidosqualenewere synthesized according to literature [9,10]. 24,25Epoxylanosterol was prepared by oxidation of lanosterol with m-chloroperbenzoic acid 1111purified by TLC and characterized by mass spectrometry (see below). Analytical procedures- Synthetic compounds and nonsaponifiable lipid fractions were separated and purified by double migration TLC on silica gel plates (Merck 6O.F.,,,)with hexaneiethyl acetate (8515, v/v) as solvent (solvent A). Under these conditions OS, DOS, LN and OLN were separated, their respective Rf being: 0.86,0.65, 0.42 and 0.20. Radiochromatograms were obtained with a Berthold TLC linear analyzer LB 283. Preparation
of 2,3(S)-oxido[14C]squalene
and 2,369):22(S) 23-dioxidol14C]squafene-
Swiss 3T3 fibroblasts, grown as monolayers (3. lo6 cells/75 cm* dish) in the presence of 10% (v/v) h#id-depleted calf serum as described before 131.were treated on the second day with l@ M U18666A, an inhibitor of the oxidosqualene cyclase 121.in order to maximize the accumulation of the squalene epoxides. 114C1Acetate(30 @/dish) was simultaneously added to the culture medium. After 4 h incubation, the cells were rinsed 3 times with phosphate-buffered saline (pH 7.4) and subsequentlytreated with 2x6 ml of O.lN NaOH for 30 min. The nonsaponifiable lipid fraction, obtained as descr$ed before 131,was fractionated by TLC using solvent A. The zones correspgnding tot4 ClOS and I1 ClDOS were eluted with CH Cl, and stored in benzene at -20 C. The I Clsqualene epoxides were quantitated by G&, as described below, in order to determine their specific activity. Gas chromatography and mass spectrometry- Gas chromatography was performed on a 25m 0.25mm i.d. DB-1 coated fused silica capillary column using cholesterol as internal standard. A Carlo-Erba (Fractovap 4160) chromatograph was used equipped gith a Spectra Physics integrator. The sample was injected directly into the column at 60 C and the analysi? performed (H 2 ml/min) pith the following temperature program: 30 / min to 240 C followed by 3” lmin to 280 C. 24,25-Epoxylanosterol was characterized by gas chromatography coupled to mass spectrometry, as described before 131operated at 7Oev.The prominent peaks were at m/z: 442 (M), 427 (M-CH ; base peak), 409 (M-CH, -H,O), 315 (M-side chain), 313 (M-SC-2H), 273 (M-SC-42), 253 (M-SC-42-H,O). Preparation of the 2.3-oxidosquakne-lanosterol cyclase- Microsomal cyclase and supernatant fractions (S1 & were obtained from rat liver as described before 1121.The enzyme was also solubilize 9 from hog-liver microsomes with emulphogene BC-720, a non-ionic detergent, and partially purified (175-fold) as described previously 1131. Enzymatic assays and kinetics- Cyclase activity assays were described before 112,131. Pseudo-first-order rate constants (V/Km) were determined from progress curves using 0.5 -1.0 PM initial [14Cjlabeled substrate concentrations, where ISlo< < Km. For a typical assay, enzyme and substrates in O.lM potassium phosphate buffer (pH 7.4), containing 0.1% Tween-80, were incubated (final volume: 1.5 ml) at 37 C. At given times, 0.1 899
Vol. 188, No. 2, 1992
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
m l aliquots were wjthdrawn and the reaction stopped with 0.1 m l 10% KOH in methanol. After lh at 55 C, the m ixture was extracted with 0.5 m l cyclohexane and analyzed by TLC (solvent A). The radioactivity associatedwith the bands was then determined. About 0.24-0.36mg protein was used in the case of the purified cyclase whereas 5-12.5 mg protein were used for the m icrosomal preparation (+ 5-12.5 mg S,,). I, values (concentration of the inhibitor needed to decreaseby 50% the reaction rate) were determined under conditions where the cyclization rates were linear with time; in this case varying concentrations of U18666A were added to the reaction m ixture (final volume per assay:0.1 m l).
RESULTS To study the cyclization of mono- and dioxidosqualenescatalyzed by 2,3-oxidosqualenelanosterol cyclase we have first prepared [14C]labeled2,3(S)-oxidosqualeneand 2,3(S): 22(S),23-dioxidosqualene.These compounds were obtained biosynthetically in order to have accessto stereochemically pure substratesfor the enzyme which cyclizes preferentially the 2,3(S)-epoxystereoisomer.An accumulation of the labeled epoxideswas achieved by treating 3T3 fibroblasts, in lipid-depleted media, with U18666A, a powerful inhibitor of the cyclase 121,in the presence of [14C]acetate.Under our experimental conditions an average of 2.8 nmoles [14C]OSand 0.9 nmole ]14C]DOS were recovered from 106cells, with respectively 7.0 and 9.1 pCi lrmol specific activities. Cyclization of 2,3(S):22(S),23-dioxido114C]squalene by the purified cyclase- Using a solubilized and partially purified liver cyclase 1131,we have first demonstrated that this enzyme was able to catalyze the conversion of biosynthetic DOS into 24,25-epoxylanosterol. Under standard conditions, as analyzed by radio-TLC, [ 14C]DOSwas readily and fully transformed (Fig 1) into a single compound that m igrated identically to an authentic sample of 24,25-ELN. The identity of the reaction product was further confirmed by GC/MS (see “Materials and Methods”). Spec$city qf the cyclase - As analyzed by Fersht 1141for the specificity of an enzyme, the ratio of the rates of transformation of two substratesA and B, competing for a same active site, is given by Equation 1 as follows. VA
/Vg
=
(V/Km)A IA1 (1) (v
/Km)B
IBl
vA/vB =
(V/Km)A (2) (V/Km)B
With identical concentrations of A and B, this expression simplifies into Equation 2. Therefore, the specificity of 2,3-oxidosqualene-lanosterolcyclase which is biologically relevant, i.e. the ability to this enzyme to discriminate between competing mono- and dioxidosqualenes,will be governed solely by the ratios of the values of V/Km (specificity constant), the apparent first-order rate constants for the reaction of the two epoxides and the free enzyme (Eq. 1 or 2). The specificity constantscan be conveniently determined from kinetic analysis of cyclization progress curves obtained at low concentrations, i.e. with ]S]o< < Km. Under such experimental conditions, the integrated rate equation for irreversible single-substratereactions simplifies into a pseudo-first-order rate equa-
Vol.
188,
No.
2, 1992
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH
COMMUNICATIONS
100
. 75 Lizfz DOS
50
OS
.
25
.
0 0
IO
20
IO
20
I
Time (mm) I
I
40
50
30 Time (min)
30
40
50
/
Figure 1. Cyclization squalene-lanosterol
of 2,3(S):22(S),23-dioxido114C].~qualene by puriJed 2,3-oxidocyclase. The substrate (1 PM) was incubated in the presence of
enzyme (0.32 mg protein) at 37 C (final volume 1.5 ml) in 100 mM potassium phosphate buffer (pH 7.4) containing emulphogene (0.3 %) and Tween-80 (0.1 %). At times indicated, 0.1 ml aliquots were analyzed by TLC as described under “Materif! and Methods”. Inset: Comparative progress curves of the cyclization 2,3(S):22(S),23-dioxido[ 4C]squakne by microsomal
of 2,3(S)-oxido] Clsqualene and cyclase. The substrates (1 PM)
were incubated with rat liver microsomes (12.5 mg protein), S a0 (11 mg protein) as -. above. The solid lines represent the theoretical curves obtained h y ttttmg, with a nonlinear regression program, the data to Eq. 3.
tion which yields directly the ratio V/Km 1151.The progress curve can be analyzed according to Equation 3: P ( X) = lOO(1 - explkrl), where P represents reaction progress,
i.e. formation (in percentage) of (epoxy)lanosterol as a function of time (t), and k the observed pseudo-first-order rate constant for the cyclization. The specificity constant is obtained from the relation: V/Km = k/[E] where IEJ is the enzyme concentration. The cyclization of [14C]OSand [14C]DOSwas studied with different preparations of the cyclase: i) detergent-solubilized and partially purified enzyme (Fig. 1) and ii) microsoma1enzyme in the presence(Fig. 1; inset) or absenceof cytosolic fraction (S&. The results are given in Table 1. From the data one can conclude that 2,3-oxidosqualene
Table
1. Summary
Cyclase preparation
qf the
kinetic data .for the cyclization of OS and DOS by different sources of 2.3-oxidosqualene-lanosterol cyclase
Specificity ratio (DOS/OS)
DOS V/Km (~10~)
OS V/Km (~10~)
min-l Img protein /ml Purified enzyme Microsomes Microsomes + S,,,
2.3 2.8 5.0
182 1.18 6.48
80 0.42 1.3
The numbers given are means of two independent determinations. The specificity constants for cyclization (at pH 7.4 and 37°C) of the substrates, under pseudo-first-order kinetic conditions, were calculated from the progress curves as described in the text. 901
Vol.
188,
0
2
0
No.
2, 1992
IO
30 20 Time (min)
BIOCHEMICAL
40
F+ure 124.Competing cyclization dtoxtdo[ Cjsqualene by microsomal
AND BIOPHYSICAL
-b
50
03 qf 2,3(S)-oxido[‘%]squalene
RESEARCH
COMMUNICATIONS
-3 log 1 (Ml und 2,3@):22(S),23-4
2,3-oxidosqualene-lanosterol cyclase. The substrates were incubated at 1 pM concentration in the presence of rat liver microsomes (8 mg protein). S,,, (5 mg protein) at 37 C (final volume 1.5 ml). At times indicated, 0.1 ml aliquots were analyzed by TLC as described under “Material and Methods”. Finute 3. Inhibition of the cyclization qf 2,3(S/:22(S).23-dioxido114C]squalene by U18666A. The substrate (1 PM) was incubated in the presence of microsomal cyclase as
described in Fig. 1. in the presence of 10-6-10-2M U18666A (final volume 0.1 ml).
cyclase shows a specificity, in terms of V/Km, in favor of 2,3(S):22(S),23-dioxidosqualene. This specificity is however dependent on the type of preparation used, and appears highest when the cyclization is performed in the presence of S,, which has been described previously to contain proteins factors (SPF) facilitating the cyclization reaction [ 161. The importance of addition of supernatant factors is attested by the increase in V/Km values compared to those obtained with washed microsomes alone (Table 1). The preferential cyclization of [14C]DOS over ]14C]OSwas further illustrated by giving to the enzyme both substrates together. As shown in Fig. 2 for the microsomal enzyme (in the presence of S,,,), when starting with identical concentrations of both substrates, [14C]DOS is converted faster into [14C]ELN than 114C]OSinto [14C]LN. Moreover, we have determined the respective Iso for U 18666A, an inhibitor of the cyclase 121,for the cyclization of [14C]DOS and ]14C]OScatalyzed by the microsomal enzyme. When DOS was used as substrate, the observed value was 7.1~10~M (Fig. 3), i.e. about 7-fold higher than for the cyclization of OS.
DISCUSSION Our results indicate that when 2,3-oxidosqualene-lanosterol cyclase is faced with 2,3(S)-oxidosqualene and 2,3(S):22(S),23-dioxidosqualeneacting as competing substrates, the enzyme shows a specificity in favor of the diepoxide. One consequenceis that when both epoxides are present at identical initial concentrations, the enzyme will cyclize preferentially DOS. This contrasts with earlier work, based on rudimentary kinetic analysis using racemic epoxides and rat liver Sl, homogenates,which concluded the opposite 117,181.When considering the specificity ratio of = 2 in favor of DOS observed with the purified cyclase in a micellar environment it is reasonableto invoke an entropic factor, i.e. whereas OS could lead to non-productive binding, DOS can be cyclized by the 902
Vol.
188,
No.
2, 1992
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH
COMMUNICATIONS
enzyme from both ends of the molecule. A similar ratio was found in the case of the microsomal enzyme, but it increased to about 5 in the presence of S,,,, indicating the importance of supernatant factors on the observed specificity of the cyclase. According to Bloch and coworkers, the supernatant fraction contains a protein (SPF), known to activate both the squalene epoxidase and the cyclase. This SPF is believed to act by facilitating the translocation of exogenous and endogenoussubstrates between and within membranes [ 16,191.Since SFP shows no activation of the solubilized enzymes ] 191,i.e. in micellar environment, it seems logical to assume that the important specificity ratio found here might be due to a more efficient translocation of DOS by SFP. That this holds also true at a cellular level, would be consistent with our data (to be published) indicating that DOS is very actively transformed by cultured cells into oxysterols. As indicated in the Introduction, the specificity of the cyclase in favor of DOS will be of importance in channeling this metabolite into a regulatory pathway when the cyclase is inhibited. Two factors are relevant: i) the formation of ELN, and ultimately of 24(S), 25epoxycholestero1,depends on the concentration of DOS accumulated by cells treated with a cyclase inhibitor (see equation 1). It is generally observed that they are high enough (see above and ]1,2,11]) to, according to our data, efficiently compete with OS for the cyclization; ii) becauseof the specificity of the cyclase in favor of DOS, the cyclization of the diepoxide is less affected by an inhibitor of this enzyme. For example as found here, about 7-fold more U 18666A is required to inhibit by 50% the cyclization of DOS compared to OS, catalyzed by the microsomal cyclase. In conclusion, the present kinetic studies allow a better understanding of the occurrence of the regulatory pathway initiated with DOS, in the biosynthesis of cholesterol.
Acknowledgment: We thank Prof. P. Benveniste for the accessto GC facilities.
REFERENCES 1. Chang, T.-Y, Schiavoni, Jr., E.S., McCrae, K.R., Nelson, J.A. and Spencer, T.A. (1979) J. Biol. Chem. 254,11258-l 1263 2. Sexton, R.C., Panini, S.R., Azran, F. and Rudney, H. (1983) Biochemistry, 22,56875692 3. Gerst, N., Duriatti, A., Schuber, F., Taton, M., Benveniste, P. and Rahier, A. (1988) Biochem. Pharmacol. 37,1955-1964 4. Xiao, M.X. and Prestwich, G.D. (1992) Biochem. Biophys. Res. Commun. 185,323329 5. Nelson, J.A., Steckbeck, S.R. and Spencer,T.A. (1981) J. Biol. Chem. 256,1067-1068 6. Panini, S.R., Sexton, R.C. and Rudney, H. (1984) J. Biol. Chem. 259,7767-7771 7. Taylor, F.R., Kandutsch, A.A., Gayen, AK., Nelson, J.A., Steckbeck Nelson, S., Phirwa, S. and Spencer, T.A. (1986) J. Biol. Chem. 261,15039-15044 8. Cattel, L., Ceruti, M., Balliano, G., Viola, F., Grosa, G. and Schuber, F. (1989) Steroids 53,363-391 9. Willett, J.D., Sharpless,KB., Lord, K.E., van Tamelen, E.E. and Clayton, R.B. (1967) J. Biol. Chem. 242,4182-4191 10.Field, R.B. and Holmund, C.E. Arch. B&hem. Biophys. (1977) 180,465-471 11. Panini, S.R.,Sexton, R.C., Gupta, A.K., Parish, E.J., Chitrakom, S. and Rudney, H. (1986) J. Ltprd Res. 27, 1190-1204 12. Duriatti, A., Bouvier-Nave, P., Benveniste, P., Schuber, F., Delprino, L., Balliano, G. and Cattel, L. (1985) Biochem. Pharmacol. 34,2765-2777 903
Vol.
188,
No.
2, 1992
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH
COMMUNICATIONS
13. Duriatti, A. and Schuher, F. (1988) Biochem. Biophys. Res. Commun. 151,1378-1385 14.Fersht, A. (1977) in Enzyme Structure and Mechanism, Chapters 3 and 11,Freeman, San Francisco 15. Orsi, B.A and Tipton, K.F. (1979) Methods Enzymol. 63, 159-183 16.Caras, I.W. and Bloch, K. (1979) J. Biol. Chem. 254,11816-11821 17. Shishthori, T, Fukui, T. and Suga, T. (1973) Chem. Lett. 1289-1292 18.Corey, E.J. and Gross, S.K. (1967) J. Am. Chem. Sot. 89,4561-4562 19.Chin, and Bloch, K. (1984) J. Biol. Chem. 259, 11735-l1738
904
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 898-904
October 30, 1992
PREFERENTIAL CYCLIZATION OF 2,3(S):22@),23-DIOXIDOSQUALENE BY MAMMALIAN 2,3-OXIDOSQUALENE-LANOSTEROL CYCLASE
Olivier BOUTAUD, Daniele DOLIS and Francis SCHUBER* Laboratoire de Chimie Bioorganique (CNRS URA-1386), FacultC de Pharmacie, 74 route du Rhin, 674~Illkirch, France Received
September
14,
1992
Kinetic studies on the cyclization of 2,3(S)-oxido and 2,3(S):22(S),23-dioxido[14C]squalene catalyzed by liver oxidosqualene-lanosterolcyclase revealed a specificity (in terms of V/Km) of the enzyme for the diepoxide. The specificity ratio was dependent on the enzyme preparation, i.e. purified or m icrosomal, and was highest (about 5) with the m icrosomal enzyme in the presenceof supernatant protein factors. These results explain why, in the presenceof cyclase inhibitors, the squalene epoxides can be channeled into a cholesterol biosynthesis regulatory pathway via 24(S),25-epoxylanosteroland 24(S),25epoxycholesterol. 0 1992 Academic Press, Inc.
In mammalian cells, inhibitors of 2,3-oxidosqualene-lanosterolcyclase (EC 5.4.99.7) affect very efficiently the biosynthesis of cholesterol. Inhibition of the cyclase results in the accumulation of 2,3(S)-oxidosqualene(OS) and also of 2,3(S):22(S),23-dioxidosqualene (DOS) 1l-31. Recently, it was shown that this diepoxide is formed by epoxidation of OS by purified squalene epoxidase 141.It was demonstrated that DOS is further converted into 24(S),25-epoxylanosterol and ultimately into 24(S),25-epoxycholesterol 151, which is a known repressor of HMG-CoA reductase [6,71,the main regulatory enzyme of cholesterol biosynthesis. Indeed, treatment of mammalian cells with cyclase inhibitors leads to a decrease of HMG-CoA reductase activity which was interpreted as resulting from the formation of regulatory oxysterols [3,6]; the repression is dependent, however, on the extent of inhibition. i.e. at high inhibitor concentrations - when presumably the cyclase is totally inhibited - the activity of HMG-CoA reductase seems unaffected 161. From a pharmacological point of view, the cyclaserepresents an attractive target for the design of specific hypocholesterolemicagents181;at a cellular level the cyclaseinhibitors m ight act synergistically: i) by direct reduction of the metabolic flow leading to cholesterol and ii) by repressing HMG-CoA reductase via the formation of 24(S),25-epoxycho-
*To whom correspondence
should
be addressed.
Abbreviurions: HMG, 3-hydroxy-3-methylglutaric acid; U18666A, 38(2-(diethylamino) ethoxy]androstJ-en-17-one; OS, 2,3-oxidosqualene; DOS, 2,3:22,23-dioxidosqualene, LN, lanosterol; ELN, 24,25-epox lanosterol; SPF, supematant protein factor; S,, supernatant fraction obtained at 1OJxg. 0006-291X/92 $4.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
898
Vol.
188,
No.
2, 1992
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH
COMMUNICATIONS
lesterol. Such a synergism could explain why cyclase inhibitors are more efficient at a cellular level than anticipated from their inhibition constants determined in vivo on the microsomal enzyme 131.The regulatory pathway uncovered by the cyclase inhibitors, i.e. formation of 24(S),25epoxycholesterol, will be relevant, however, only if DOS proves to be an excellent substrate of the cyclase. In this study we have compared the specificity of the cyclase towards OS and DOS considered as competing substrates;the results indicate that 2,3(S):22(S),23-dioxidosqualeneis the preferred substrate.
MATERIALS
AND
METHODS
Sodium [2-14C]acetate(48 mCi/mmol) was purchased from NEN-DuPont de Nemours (France). [3-3H12,3-oxidosqualeneand U18666A were kind gifts from respectively Drs. M. Cerutti and L. Cattel (Torino, Italy) and Dr. R.J. Cenedella (Kirksville, MO, USA). 2,3-Oxidosqualene and 2,3:22,23-dioxidosqualenewere synthesized according to literature [9,10]. 24,25Epoxylanosterol was prepared by oxidation of lanosterol with m-chloroperbenzoic acid 1111purified by TLC and characterized by mass spectrometry (see below). Analytical procedures- Synthetic compounds and nonsaponifiable lipid fractions were separated and purified by double migration TLC on silica gel plates (Merck 6O.F.,,,)with hexaneiethyl acetate (8515, v/v) as solvent (solvent A). Under these conditions OS, DOS, LN and OLN were separated, their respective Rf being: 0.86,0.65, 0.42 and 0.20. Radiochromatograms were obtained with a Berthold TLC linear analyzer LB 283. Preparation
of 2,3(S)-oxido[14C]squalene
and 2,369):22(S) 23-dioxidol14C]squafene-
Swiss 3T3 fibroblasts, grown as monolayers (3. lo6 cells/75 cm* dish) in the presence of 10% (v/v) h#id-depleted calf serum as described before 131.were treated on the second day with l@ M U18666A, an inhibitor of the oxidosqualene cyclase 121.in order to maximize the accumulation of the squalene epoxides. 114C1Acetate(30 @/dish) was simultaneously added to the culture medium. After 4 h incubation, the cells were rinsed 3 times with phosphate-buffered saline (pH 7.4) and subsequentlytreated with 2x6 ml of O.lN NaOH for 30 min. The nonsaponifiable lipid fraction, obtained as descr$ed before 131,was fractionated by TLC using solvent A. The zones correspgnding tot4 ClOS and I1 ClDOS were eluted with CH Cl, and stored in benzene at -20 C. The I Clsqualene epoxides were quantitated by G&, as described below, in order to determine their specific activity. Gas chromatography and mass spectrometry- Gas chromatography was performed on a 25m 0.25mm i.d. DB-1 coated fused silica capillary column using cholesterol as internal standard. A Carlo-Erba (Fractovap 4160) chromatograph was used equipped gith a Spectra Physics integrator. The sample was injected directly into the column at 60 C and the analysi? performed (H 2 ml/min) pith the following temperature program: 30 / min to 240 C followed by 3” lmin to 280 C. 24,25-Epoxylanosterol was characterized by gas chromatography coupled to mass spectrometry, as described before 131operated at 7Oev.The prominent peaks were at m/z: 442 (M), 427 (M-CH ; base peak), 409 (M-CH, -H,O), 315 (M-side chain), 313 (M-SC-2H), 273 (M-SC-42), 253 (M-SC-42-H,O). Preparation of the 2.3-oxidosquakne-lanosterol cyclase- Microsomal cyclase and supernatant fractions (S1 & were obtained from rat liver as described before 1121.The enzyme was also solubilize 9 from hog-liver microsomes with emulphogene BC-720, a non-ionic detergent, and partially purified (175-fold) as described previously 1131. Enzymatic assays and kinetics- Cyclase activity assays were described before 112,131. Pseudo-first-order rate constants (V/Km) were determined from progress curves using 0.5 -1.0 PM initial [14Cjlabeled substrate concentrations, where ISlo< < Km. For a typical assay, enzyme and substrates in O.lM potassium phosphate buffer (pH 7.4), containing 0.1% Tween-80, were incubated (final volume: 1.5 ml) at 37 C. At given times, 0.1 899
Vol. 188, No. 2, 1992
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
m l aliquots were wjthdrawn and the reaction stopped with 0.1 m l 10% KOH in methanol. After lh at 55 C, the m ixture was extracted with 0.5 m l cyclohexane and analyzed by TLC (solvent A). The radioactivity associatedwith the bands was then determined. About 0.24-0.36mg protein was used in the case of the purified cyclase whereas 5-12.5 mg protein were used for the m icrosomal preparation (+ 5-12.5 mg S,,). I, values (concentration of the inhibitor needed to decreaseby 50% the reaction rate) were determined under conditions where the cyclization rates were linear with time; in this case varying concentrations of U18666A were added to the reaction m ixture (final volume per assay:0.1 m l).
RESULTS To study the cyclization of mono- and dioxidosqualenescatalyzed by 2,3-oxidosqualenelanosterol cyclase we have first prepared [14C]labeled2,3(S)-oxidosqualeneand 2,3(S): 22(S),23-dioxidosqualene.These compounds were obtained biosynthetically in order to have accessto stereochemically pure substratesfor the enzyme which cyclizes preferentially the 2,3(S)-epoxystereoisomer.An accumulation of the labeled epoxideswas achieved by treating 3T3 fibroblasts, in lipid-depleted media, with U18666A, a powerful inhibitor of the cyclase 121,in the presence of [14C]acetate.Under our experimental conditions an average of 2.8 nmoles [14C]OSand 0.9 nmole ]14C]DOS were recovered from 106cells, with respectively 7.0 and 9.1 pCi lrmol specific activities. Cyclization of 2,3(S):22(S),23-dioxido114C]squalene by the purified cyclase- Using a solubilized and partially purified liver cyclase 1131,we have first demonstrated that this enzyme was able to catalyze the conversion of biosynthetic DOS into 24,25-epoxylanosterol. Under standard conditions, as analyzed by radio-TLC, [ 14C]DOSwas readily and fully transformed (Fig 1) into a single compound that m igrated identically to an authentic sample of 24,25-ELN. The identity of the reaction product was further confirmed by GC/MS (see “Materials and Methods”). Spec$city qf the cyclase - As analyzed by Fersht 1141for the specificity of an enzyme, the ratio of the rates of transformation of two substratesA and B, competing for a same active site, is given by Equation 1 as follows. VA
/Vg
=
(V/Km)A IA1 (1) (v
/Km)B
IBl
vA/vB =
(V/Km)A (2) (V/Km)B
With identical concentrations of A and B, this expression simplifies into Equation 2. Therefore, the specificity of 2,3-oxidosqualene-lanosterolcyclase which is biologically relevant, i.e. the ability to this enzyme to discriminate between competing mono- and dioxidosqualenes,will be governed solely by the ratios of the values of V/Km (specificity constant), the apparent first-order rate constants for the reaction of the two epoxides and the free enzyme (Eq. 1 or 2). The specificity constantscan be conveniently determined from kinetic analysis of cyclization progress curves obtained at low concentrations, i.e. with ]S]o< < Km. Under such experimental conditions, the integrated rate equation for irreversible single-substratereactions simplifies into a pseudo-first-order rate equa-
Vol.
188,
No.
2, 1992
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH
COMMUNICATIONS
100
. 75 Lizfz DOS
50
OS
.
25
.
0 0
IO
20
IO
20
I
Time (mm) I
I
40
50
30 Time (min)
30
40
50
/
Figure 1. Cyclization squalene-lanosterol
of 2,3(S):22(S),23-dioxido114C].~qualene by puriJed 2,3-oxidocyclase. The substrate (1 PM) was incubated in the presence of
enzyme (0.32 mg protein) at 37 C (final volume 1.5 ml) in 100 mM potassium phosphate buffer (pH 7.4) containing emulphogene (0.3 %) and Tween-80 (0.1 %). At times indicated, 0.1 ml aliquots were analyzed by TLC as described under “Materif! and Methods”. Inset: Comparative progress curves of the cyclization 2,3(S):22(S),23-dioxido[ 4C]squakne by microsomal
of 2,3(S)-oxido] Clsqualene and cyclase. The substrates (1 PM)
were incubated with rat liver microsomes (12.5 mg protein), S a0 (11 mg protein) as -. above. The solid lines represent the theoretical curves obtained h y ttttmg, with a nonlinear regression program, the data to Eq. 3.
tion which yields directly the ratio V/Km 1151.The progress curve can be analyzed according to Equation 3: P ( X) = lOO(1 - explkrl), where P represents reaction progress,
i.e. formation (in percentage) of (epoxy)lanosterol as a function of time (t), and k the observed pseudo-first-order rate constant for the cyclization. The specificity constant is obtained from the relation: V/Km = k/[E] where IEJ is the enzyme concentration. The cyclization of [14C]OSand [14C]DOSwas studied with different preparations of the cyclase: i) detergent-solubilized and partially purified enzyme (Fig. 1) and ii) microsoma1enzyme in the presence(Fig. 1; inset) or absenceof cytosolic fraction (S&. The results are given in Table 1. From the data one can conclude that 2,3-oxidosqualene
Table
1. Summary
Cyclase preparation
qf the
kinetic data .for the cyclization of OS and DOS by different sources of 2.3-oxidosqualene-lanosterol cyclase
Specificity ratio (DOS/OS)
DOS V/Km (~10~)
OS V/Km (~10~)
min-l Img protein /ml Purified enzyme Microsomes Microsomes + S,,,
2.3 2.8 5.0
182 1.18 6.48
80 0.42 1.3
The numbers given are means of two independent determinations. The specificity constants for cyclization (at pH 7.4 and 37°C) of the substrates, under pseudo-first-order kinetic conditions, were calculated from the progress curves as described in the text. 901
Vol.
188,
0
2
0
No.
2, 1992
IO
30 20 Time (min)
BIOCHEMICAL
40
F+ure 124.Competing cyclization dtoxtdo[ Cjsqualene by microsomal
AND BIOPHYSICAL
-b
50
03 qf 2,3(S)-oxido[‘%]squalene
RESEARCH
COMMUNICATIONS
-3 log 1 (Ml und 2,3@):22(S),23-4
2,3-oxidosqualene-lanosterol cyclase. The substrates were incubated at 1 pM concentration in the presence of rat liver microsomes (8 mg protein). S,,, (5 mg protein) at 37 C (final volume 1.5 ml). At times indicated, 0.1 ml aliquots were analyzed by TLC as described under “Material and Methods”. Finute 3. Inhibition of the cyclization qf 2,3(S/:22(S).23-dioxido114C]squalene by U18666A. The substrate (1 PM) was incubated in the presence of microsomal cyclase as
described in Fig. 1. in the presence of 10-6-10-2M U18666A (final volume 0.1 ml).
cyclase shows a specificity, in terms of V/Km, in favor of 2,3(S):22(S),23-dioxidosqualene. This specificity is however dependent on the type of preparation used, and appears highest when the cyclization is performed in the presence of S,, which has been described previously to contain proteins factors (SPF) facilitating the cyclization reaction [ 161. The importance of addition of supernatant factors is attested by the increase in V/Km values compared to those obtained with washed microsomes alone (Table 1). The preferential cyclization of [14C]DOS over ]14C]OSwas further illustrated by giving to the enzyme both substrates together. As shown in Fig. 2 for the microsomal enzyme (in the presence of S,,,), when starting with identical concentrations of both substrates, [14C]DOS is converted faster into [14C]ELN than 114C]OSinto [14C]LN. Moreover, we have determined the respective Iso for U 18666A, an inhibitor of the cyclase 121,for the cyclization of [14C]DOS and ]14C]OScatalyzed by the microsomal enzyme. When DOS was used as substrate, the observed value was 7.1~10~M (Fig. 3), i.e. about 7-fold higher than for the cyclization of OS.
DISCUSSION Our results indicate that when 2,3-oxidosqualene-lanosterol cyclase is faced with 2,3(S)-oxidosqualene and 2,3(S):22(S),23-dioxidosqualeneacting as competing substrates, the enzyme shows a specificity in favor of the diepoxide. One consequenceis that when both epoxides are present at identical initial concentrations, the enzyme will cyclize preferentially DOS. This contrasts with earlier work, based on rudimentary kinetic analysis using racemic epoxides and rat liver Sl, homogenates,which concluded the opposite 117,181.When considering the specificity ratio of = 2 in favor of DOS observed with the purified cyclase in a micellar environment it is reasonableto invoke an entropic factor, i.e. whereas OS could lead to non-productive binding, DOS can be cyclized by the 902
Vol.
188,
No.
2, 1992
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH
COMMUNICATIONS
enzyme from both ends of the molecule. A similar ratio was found in the case of the microsomal enzyme, but it increased to about 5 in the presence of S,,,, indicating the importance of supernatant factors on the observed specificity of the cyclase. According to Bloch and coworkers, the supernatant fraction contains a protein (SPF), known to activate both the squalene epoxidase and the cyclase. This SPF is believed to act by facilitating the translocation of exogenous and endogenoussubstrates between and within membranes [ 16,191.Since SFP shows no activation of the solubilized enzymes ] 191,i.e. in micellar environment, it seems logical to assume that the important specificity ratio found here might be due to a more efficient translocation of DOS by SFP. That this holds also true at a cellular level, would be consistent with our data (to be published) indicating that DOS is very actively transformed by cultured cells into oxysterols. As indicated in the Introduction, the specificity of the cyclase in favor of DOS will be of importance in channeling this metabolite into a regulatory pathway when the cyclase is inhibited. Two factors are relevant: i) the formation of ELN, and ultimately of 24(S), 25epoxycholestero1,depends on the concentration of DOS accumulated by cells treated with a cyclase inhibitor (see equation 1). It is generally observed that they are high enough (see above and ]1,2,11]) to, according to our data, efficiently compete with OS for the cyclization; ii) becauseof the specificity of the cyclase in favor of DOS, the cyclization of the diepoxide is less affected by an inhibitor of this enzyme. For example as found here, about 7-fold more U 18666A is required to inhibit by 50% the cyclization of DOS compared to OS, catalyzed by the microsomal cyclase. In conclusion, the present kinetic studies allow a better understanding of the occurrence of the regulatory pathway initiated with DOS, in the biosynthesis of cholesterol.
Acknowledgment: We thank Prof. P. Benveniste for the accessto GC facilities.
REFERENCES 1. Chang, T.-Y, Schiavoni, Jr., E.S., McCrae, K.R., Nelson, J.A. and Spencer, T.A. (1979) J. Biol. Chem. 254,11258-l 1263 2. Sexton, R.C., Panini, S.R., Azran, F. and Rudney, H. (1983) Biochemistry, 22,56875692 3. Gerst, N., Duriatti, A., Schuber, F., Taton, M., Benveniste, P. and Rahier, A. (1988) Biochem. Pharmacol. 37,1955-1964 4. Xiao, M.X. and Prestwich, G.D. (1992) Biochem. Biophys. Res. Commun. 185,323329 5. Nelson, J.A., Steckbeck, S.R. and Spencer,T.A. (1981) J. Biol. Chem. 256,1067-1068 6. Panini, S.R., Sexton, R.C. and Rudney, H. (1984) J. Biol. Chem. 259,7767-7771 7. Taylor, F.R., Kandutsch, A.A., Gayen, AK., Nelson, J.A., Steckbeck Nelson, S., Phirwa, S. and Spencer, T.A. (1986) J. Biol. Chem. 261,15039-15044 8. Cattel, L., Ceruti, M., Balliano, G., Viola, F., Grosa, G. and Schuber, F. (1989) Steroids 53,363-391 9. Willett, J.D., Sharpless,KB., Lord, K.E., van Tamelen, E.E. and Clayton, R.B. (1967) J. Biol. Chem. 242,4182-4191 10.Field, R.B. and Holmund, C.E. Arch. B&hem. Biophys. (1977) 180,465-471 11. Panini, S.R.,Sexton, R.C., Gupta, A.K., Parish, E.J., Chitrakom, S. and Rudney, H. (1986) J. Ltprd Res. 27, 1190-1204 12. Duriatti, A., Bouvier-Nave, P., Benveniste, P., Schuber, F., Delprino, L., Balliano, G. and Cattel, L. (1985) Biochem. Pharmacol. 34,2765-2777 903
Vol.
188,
No.
2, 1992
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH
COMMUNICATIONS
13. Duriatti, A. and Schuher, F. (1988) Biochem. Biophys. Res. Commun. 151,1378-1385 14.Fersht, A. (1977) in Enzyme Structure and Mechanism, Chapters 3 and 11,Freeman, San Francisco 15. Orsi, B.A and Tipton, K.F. (1979) Methods Enzymol. 63, 159-183 16.Caras, I.W. and Bloch, K. (1979) J. Biol. Chem. 254,11816-11821 17. Shishthori, T, Fukui, T. and Suga, T. (1973) Chem. Lett. 1289-1292 18.Corey, E.J. and Gross, S.K. (1967) J. Am. Chem. Sot. 89,4561-4562 19.Chin, and Bloch, K. (1984) J. Biol. Chem. 259, 11735-l1738
904
