Squalene Synthase Inhibitors_Todd L Capson

8/8/2019 Squalene Synthase Inhibitors_Todd L Capson

1/6

3734

90056-31-8;7 (isomer l ) , 90129-48-9;7 (isomer 2), 90056-32-9;7 al-coho1 (isomer l), 119637-88-6;7 alcohol (isomer 2), 119717-48-5;8(isomer I ) , 119619-14-6;8 (isomer 2), 119717-34-9;8 1-hydroxy deriv-ative, 119619-13-5;9 (isomer l), 119619-15-7;9 (isomer 2), 119619-16-8;10 , 119619-17-9;11, 119619-18-0;11 P,y-unsaturated derivative,119619-19-1;11 hydroxy enone derivative, 119619-20-4;11 epoxy alcoholderivative, 119619-21-5; 12 , 119619-22-6; 13 , 119717-35-0; 14, 64-7; Na+[PhSeB(O Et),]-, 117268-79-8.

J . A m . C h em .SOC. 989, 111 , 3734-3739

119717-36-1;15 (isomer l ) , 119619-23-7;15 (isomer 2) , 119619-24-8;15 7-alcohol derivative, 1 19619-25-9;16 , 119619-26-0;16 diol mesylatederivative, 119637-89-7;17 , 119619-27-1; 17 aldehyde derivative,119619-28-2; 18 , 119619-29-3; 19 , 119619-30-6; 19 diol derivative,119619-31-7;(-)-20, 20744-71-2;21 (isomer I), 119678-62-5;21 (isomer2), 119678-63-6;22 (isomer l), 119619-32-8;22 (isomer 2) , 119678-

Squa lene Synthetase. Inhibition by Am monium A nalogues ofCarb ocationic Interm edia tes in the Conversionof PresqualeneDiphosphate to Squalene

C. Dale Poulter,* Todd L. Capson, Michael D. Thompson, and Ronda S. Bard

Contribution fro,m the D epartments of Chem istry and Medicinal Ch emistry, Universityof Utah,Salt Lake City, Utah 84112. Received Septem ber 16 , 1988

Abstract: Squalene synthetase(EC 2.5.1.21) catalyzes the formation of squalene(3) from farnesyl diphosphate(1) via presqualenediphosphate (2 ) in two steps. The mechanism of the rearrangement of2 to 3 was studied with stable amm onium analogues6 an d 7 of primary and tertiary cyclopropylcarbinyl cations4 an d 5, respectively, proposed as intermediates. In non-pyro-phosphate-containing buffers,6 and 7 were not inhibitors. However, the combinationof 6 or 7 with PP, produced potent synergisticinhibition of squalene synthesis from1 an d 2. Amino acid 8, an analo gue in which a phosphonophosphate moiety was tetheredto the amino group in6 , was a potent inhibitorof squalene synthesis in pyrophosphate-free buffers. When synthesisof 2 an d3 from 1 was measured simultaneously in the presenceof 8, both rate s were depressed in an identical mann er. It was concludedthat squalene synthetase has a single active site which catalyzes1 - an d 2 - . The mechanism of the second reactionis discussed.

Squalene synthetase(farnesy1diphosphate:farnesyldiphosphatefarnesyl transferase, EC 2.5.1.21) cata lyzes th e format ion ofsqualene from farnesyl diphosphate in two distinct steps.' Inreaction 1, two molecules of farnesy l diphosp hate(1) ar e condensedto form presqualene diphosphate(Z), a prenyl transfer where theCl- C2 double bond of one farnesyl diphospha te serves as theprenyl acceptor for the farnesyl residue of the other? Presqualen e

diphosphateis then converted to squalene(3 ) in reaction 2 by arearran gem ent tha t cleaves the two newly formed cyclopropanebonds and joins theCI carbo ns of the two original farnesyl residuesto generate a 1 '-1-fused i~ op re no id .~

P O c1 9

1

PPi,H*

r e a c t i o n I

N A D P H

NADP! ppi

r e a c t i o n 2

3

( 1 ) Poulter, C . D.; Rilling, H. C. Biosynthesisof Isoprenoid Compounds;Porter, J . W .; Spurgeon,S. L. , Eds.; Wiley: New York, 1981; Vol. 1, pp41 3-441.

(2 ) Poulter, C. D.; illing,H . C. Biosynthesisof Isoprenoid Compoun ds;Porter, J . W.; Spurgeon,S. L., Eds.; Wiley: New York, 1981; Vol. 1, pp162-224.

0002-7863/89/1511-3734$01.50/0

Scheme I. A Mechanism for Conversion of Presqualene Diphosphateto Squalene

N A D P

2

" C H 3 C l l H 1 9

wc -1 19

NADP+

3

HI

N A D P

4

N A D P

5

In add ition to conversion to squalene, whose sole fateis sterolsynthesis, farnesyl diphosp hate also serves as a primer for th e

prenyl transfers which generate 2,3-dehydrodolicyl di p h~ sp ha te ,~al l - trans polyprenyl diphosphates for ubiquinone bio ~y nt he sis ,~and perhaps the hydrophobic prenyl units involved in modificationof nuclear proteins essential for cell division a nd main tenan ce ofcellular morphology.6 Th e 1'-1 condensation is th e first pat h-way-specific reaction in sterol metabo lism, and squalen e synthetase

(3 ) For a descriptionof non-head-to-tailat tachments in isoprenoids see:Poulter, C. D. ; Marsh,L . L.; Hughes, J. M. ; Argyle, J. C.; Satterwhite, D.M.; Goodfellow, R. J.; Moesinger, S . G . J . Am. Chem.SOC.1977, 99,3816-3823.

(4) Adair, W . L.; Cafmeyer,N .; Keller, R . K. J . Biol. Chem. 1984, 259,4441-4446.

(5 ) Faust, J. R.; Goldstein, J. L.; Brown, M . S. Arch. Biochem. Biophys.1979, 192, 86-99.

(6 ) Wolda, S. L.; Glomset,J . A. J . Biol. Chem. 1988, 263, 5997-6000.

0989 American Chemical Society

Todd L. Capson, application materials for Senior Program Officer, TRAFFIC North America


2/6

Squalene Synthetase

thus resides at an important branch point in the isoprenoidpathw ay. Rilling and co-workers recently solubilized and purifiedsqualene synthetase to h~mog eneity . ' ,~Th e intrinsic m icrosomalprotein had a molecular weight of 47000, requiredMg2+ andN A D H 9or NADPH as cofactors, and functioned independentlyof membrane-derived lipids. They also found that th e catalyticactivities for synthesis of presqualene diphosphate and squalen ecopurified a t all stages of their procedure.

Th e mechanism of the 1'-1 coupling has been the subject ofa considerable amoun t of speculation. With th e elucidationofthe s t ructure of 2 and the recognition that3 was synthesized in

two distinct steps, attention was directed to th e mechanisms ofthe individual transform ations. Although several were proposedfor reaction 1 , there is little work which bears directly on thisstep.lJO In contrast, several groups recognized that the reductiverearrangement of2 to 3 could be rationalized in terms of the bondreorganizations typically observed for cyclopropylcarbiny l cat-i o n ~ . " - ~ ~i ll in g an d c o - ~ o r k e r s ' ~riginally proposed the se-quenc e shown in Sch eme I. Subseq uently , Poulte r et a1.I6 dis-covered that t he rearrangement of primary cation4 to tertiarycation 5 was neither kinetically nor thermodynamically favoredin nonenzymatic model studies and concluded that squalenesynthetase must exert strict regiocontrol on the cationic inter-mediates if the high selectivity required for biosysthesis of squalenewas to be realized. They suggested that regiocontrol was achievedthrough favorable electrostatic interactions between P PI generatedas a consequenceof cleaving the C - O bond in2 and carbocationicspecies along the reaction path to squalene.I6-l8

I n s ev er al re ce nt s tu di es , a m m ~ n i u m ' ~ - ~ 'nd sulfonium2',22analogues of carbocations proposed as reactive intermediates forreactions in the isoprenoid pathway proved to be potent inhibitorsof the relevant enzymes. W e now presenta full account of ex-periments with ammo nium analogues6 an d 7.23 The compoundswere designed to mimic the electrostatic and topological propertiesof primary cation 4 and tertiary cation 5, respectively. Althoughneither analogue was effective in standard assays, both becamepotent synergistic inhibitors in combination with P PI, presumablyby duplicating features of the related cyclopropylsarbinyl cat-ion-PPI ion pairs. Synergism between PP I and sulfonium analoguesof putative carbocationic intermed iates in cyclizations of geranyldiphosphate to a-pinene and bornyl diphosphate was also recentlyreported by Cro teau an d co-workers.21 In addition, we found that

J . A m . Chem. SOC., ol. 111, N o. 10 , 1989 3735

(7) Kuswik-Rabiega, G .; Rilling, H. C. J . Biol. Chem. 1987, 262,

(8 ) Sasiak, K.; Rilling,H. C. Arch. Biochem. Biophys.1988,260,622627 .(9) Abbreviations: BHD A, endo-bicyclo[2.2.1]heptane-2,3-dicarboxylic

acid; DTT, dithiothreitol; EDTA,ethylenediaminetetraacetic acid; NADP ,nicotinamide adenine dinucleotide phosphate; N AD PH , reduced nicotinamideadenine dinucleotide phosphate; Pi , inorganic phosphate; P P,, inorganic py-roahosahate.

1505-1 509.

- r - - - r - - - -( I O ) Altman, L. J. ; Kowerski,R. C.; Laungani, D. R.J . Am . Chem.S O ~ .

( I 1) van Tamelen, E. E.; Schwartz, M. A.J . Am . Chem.SOC. 971, 93,1978, 100, 6174-6182.

1780-1 762.(12) Altman. L. J. ; Kowerski. R . C. ; Rillina, H . C. J . Am. Chem. SOC.

1971, 93 , 1782-1783.(13) Rilling, H.C.; Poulter, C. D.; Epstein, W . W. ; Larsen, B. R.J . A m .

Chem. SOC. 971. 93 . 1783-1785.(14) Coates, R. M.; Robinson, W. H . J . Am. Chem. SOC.1972, 94 ,

5920-5921.( 1 5 ) Qureshi, A. A.; Barnes, F. J . ; Semmler, E. J.; Porter, J. W . J . B i d .Chem. 1973, 248, 2755-2767.

(16) Poulter, C. D.; Marsh,L. L.; Hughes, J. M.; Argyle, J . C.; Satter-white, D. M.; Goodfellow, R.J. ; Moesinaer, S G. J . Am . Chem.SOC. 977.99 , 3816-3823.

(17) Poulter, C. D.; Muscio,0 . .I.; oodfellow, R. J .Biochemistry 1974.13 , 1530-1538.

(18) Poulter, C. D. J . Agric . Food Chem. 1974, 22, 167-173.(19) Narula, A. S.; Rahier, A.; Benveniste, P.; Schuber,F. J . Am . Chem.

(20) Sandifer, R . M. ; Thompson, M. D.; Gaughan, R.G.; Poulter, C. D.

(21) R ahier, A.; Gtno t, J.-C.; Schuber, F.; Benveniste,P.; Narula, A. J.

(22) C roteau, R.; Wheeler,C. J.; Aksela, R.; Oehlschlager, A. C.J . B i d .

(23) Preliminary results with7 were presented in ref 20.

SOC. 981, 103, 2408-2409.

J . A m . Chem. SOC. 982, 104, 7316-7378.

J . B i d . Chem. 1984, 259, 15215-15223.

Chem. 1986, 261 1251-1263.

6 7

8

phosphonophosphate 8, a tethered analogue of theb P P i on pair,was a potent inhibitor in the absence ofPP,. These results arediscussed in terms of an ion pair mechanism for th e rearrangementof presqualene diphosphate to squalene.

Experimental SectionMaterials and General Methods. Unless stated otherwise all chemicals

were purchased from Sigma or Aldrich. Cytochrom ec reductase andglucose-6-phosphate dehydrogenase were from Sigm a. Syntheses ofammonium analogues 6-8 were reporte d previously.24 endo-Bicyclo-[2.2.l]heptane-2,3-dicarboxylic a c i d a n h y d r i d e w a s s y n t h e s i ~ e d . ~ ~BHDA buffer stocks were prepared by heating the anhydride with 1equiv of 6 M KO H until the materi al dissolved. Th e solution was allowedto cool, and the p H was adjusted to 7.4. Th e solution was diluted to1M and stored until needed. [I-3 H]F arne syldiphosphate was preparedaccording to the p rocedure of Davisson et a1.26 an d dilute d with cold

material2' to the desired specific activit y. Fresh bakers' yea st( S a c c h a -romyces cereuisiae) was purchased in 1- lb blocks from West co Stand ardCo. (Salt Lake City, UT), s tored at -20 O C , and used within 1 month.

Radioactivity was measured on a Packard Tri -Ca rb 4530 liquid scin-tillation spectrometer using 4% (w/v) Omnifluor in toluene, 0.4% (w/v)Omnifluor (Amersham) in 2:1 toluene/Triton-X (Amersh am), Instagel(Packard), or Instafluor (Packard) cocktails. Protein concentrations weremeasured according to the procedure of Lowry.28 Solutions of amm o-nium salts 6 an d 7 for inhibition experiments were prepared by addingthe corresponding amines to 100 m M BHDA containing 1% (v/v)Tween-80, pH 7.4 (approximately 1 mg/ mL ). Th e materials were mixedat high speed on a vortex mixer for 5 min, and the resulting slightlyturbid suspensions were diluted to 5 mL with the above buffer to giveclear solutions that were stored until needed.

[1-'HIPresqualene Alcohol ([l-3H]-12). In a flame-dried 2-mL vialcontaining 116 mg (0.27 mmol) of presqualene aldehyde,24 solution ofsodium ['Hlborohydride (25 mCi, 500 mCi/mmol, Amersham) in0. 2

mL of basic methanol (one pellet of potassium hydroxide/30 mL) wasadded. Th e ampule that contained the sodium [3H]boroh ydride wasrinsed with two 0.1-m L portions of methanol, which were added to thereaction vial. Th e vial was capped , and the clear, light-yellow solutionwas stirred at room temperature for 18 h. Sodium borohydride ( I O mg,0.26 mmol) was added, and the cloudy solution was stirred for 2 h atroom tempera ture. Excess sodium borohydride was quenched by a slow,dropwise addition of 0.05 mL of a saturated solution of ammoniumchloride. The contentsof the vial were transferred to a culture tube, andthe vial was rinsed with 2 mL of pentane. Saturated sodium chloride (2mL ) was added, and the solution was carefully mixed. Th e pentane layerwas removed and passed through a plug of magnesium sulfate. Th eaqueous layer was extracted with two I-mL portions of pentane; eachI-mL portion was passed through ma gnesium sulfat e. Th e mater ial waspurified by flash chromatograph y (2.5X 13 cm column, 235-400-meshsilica gel, l.8:8.2 ethyl acetate/hexanes) to yield a clear, viscous oil (12.9mCi, 52% radiochemical yield) that com igrated with auth entic material.24The alcohol was stored at -20 O C in 4.5 mL of purified hexanes.

Determination of Specific Activity for [ I-'HIPresqualene Naphthoate([1-3H]-13). I n a flame-dried I-m L vial containing a solution ofI O mg(0.08 mmol) of 4-(dimethy1amino)pyridine and 34 mgof 2-naphthoicacid (0.2 mmol) in 0.5 mL of methylene chloride was added 56pCi of[I-'H]-12 in 20 p L of pentane. Th e solution was stirred for 2 min beforeaddition of 41 mg of 1,3-dicyclohexylcarbodiimide 0.20 mmol) dissolved

(24) Capson, T. L. ; Thompson, M . D.; Dixit, V . M .; Gaughan. R. G.;

(25) Durst, H. 0 .; Gokel, G. W. Experimental Organic Chemistry;

(26) Davisson,V . J. ; Zabriskie,T. M.; Poulter, C. D. Bioorg. Chem . 1986,

(27) Davisson, V . J .; Woodside, A. B.; Neal, T. R.; Stremler, K. E. ;

(28) Layne, E. Methods Enzymol. 1957, 3 , 441-454.

Poulter, C. D. J . Org. Chem. 1988, 5 3 , 5903-5908.

McG raw-H ill: New York, 1980; pp 230-231.

14 , 46-54.

Muehlbacher, M.; Poulter, C. D.J . Org. Chem. 1986, SI , 4768-4779.


3/6

3736 J . A m , C h e m .Soc., Vol. 111, N o. 10, 1989

i n 0.5 mL of methylene chloride. Th e solution was stirred for 4 h at roomtemp erat ure. Solvent was removed under a stream of nitrogen to afforda white residue that was suspended in 1 mL of hexanes. Th e solution wasfiltered, and the reaction vial was rinsed with four 1-m L portions ofhexanes. Th e hexane extracts were combined, and solvent was removedunder a stream of nitrogen to afford a white residue that was againtreated four times with 1-m L portions of hexanes as previously described.Solvent was removed from the combined filtrates under a stream ofnitrogen to afford a white residue that was suspended in 0.4 mL ofhexanes. A 0.2-mL portion was applied to a 20X 20 cm pla te of silicagel 60,0.25 mm thick, F 254 (EM reagents), and the plate was developedthree times with 0.3:9.7 ethyl acetate/hexanes. The UV-active material

that comigrated with auth entic cold13 prepared in an identical mannerwas dissolved in IO mL of methylene chloride. Solvent was removedunde r reduced pressure, and t he residue was dissolved in 1 .0 mL ofHP LC -gra de acetonitrile. Th e radiochemical recovery was 11.8 pCi(21%). Analysis by UV and liquid scintillation spectrometry26gave aspecific activity of 52.7 mCi/mmol.

In a previous experiment using the same procedure, a sample of cold13 (5.3 mg, 25%) was prepared from 15mg (0.036 mmol) of 12: ' HNM R (300 MHz, CDCIJ 8 .60 (brs, 1 H), 8.06(dd, J = 8.4 and 1.4 Hz,1 H), 7 .88 (m , 3 H) , 7 .54 (m , 2 H ) , 5.08 (m, 4 H, vinyl) , 4.97 (d, J =7.3 Hz, 1 H , vinyl coupled to C3'), 4.65 (m , 1 H, carbinyl), 4.19 (m, 1H, carbinyl) , 2.0 (m , 14 H, vinyl methylenes), 1.65(s, 2 H , vinyl me-thylenes), 1.64 (s, 6 H, methyls) , 1.56(s, 15 H, methyls), 1.5-1.0 (un-resolved cyclopropyl resonances), 1.20 ppm(s, 3 H, cyclopropyl methyl);

1130, 1090 cm-I; UV (A acetonitrile) 236 nm (f , 6.3 X IO4).[l-3H]Presqualene Diphosphate ([l-)H]-2). To a flame-dried two-neck

15-mL round-bottom flask was added 6.3 mCi of [l-3H]presqualenealcohol (0.12 mmol) in 2.2 mL of hexanes. Solvent was removed witha stream of nitrogen, and the flask was placed in an ice bath. Tri-chloroacetonitrile (163 mg, .13 mmol, freshly distilled) was addeddropwise, and the solution was stirred under nitrogen a troom tempera-ture for 45 min at 0 OC. The reaction m ixture was allowed to warm toroom temperatu re, and a solution of 112 mg (0.38 mmol) of bis(tri-ethylamm onium) hydrogen phosphate in 2.7 mL of acetonitrile wasadded dropwise over a 4-h period. The yellow-green solution was allowedto stir for 26 h at room temperature. Th e reaction mixture was trans-ferred to a 5 0-mL rou nd-bottom flask, and solvent was removed underreduced pressure. Th e residue was dissolved in 1:25 (v/v) 2-pro panol/25mM ammonium bicarbonate buffer and applied to a IO-mL (17-mequiv)column of AG 50W- X8 ion-exchange resin (ammon ium form). Th ecolumn was slowly eluted with 4 column volumes of the same buffer.Solvent was removed by lyophilization. Th e residual fluffy, yellow-whitesolid was dissolved in 8 mL of chloroform and transferred to a culturetube. Ammonium bicarbonate (8 mL of a 0.1 M solution) was added,and the suspension was vigorously mixed for 10s. The mixture wascentrifuged, and t he chloroform layer was removed. Th e aqueous layerwas extracted with two 8-m L portions of chloroform , and solvent wasremoved from the combined chloroform layers at reduced pressure. Theresidue was dissolved in a minimal am ount of 1:1:0.4 (v/v/v) butanol/tetrahydrofur an/O.l M ammon ium bicarbonat e and applied to a 2.5X13 cm column of CF11 cellulose (Whatm an). Th e column was elutedby gravity flow with the same solvent. Fractions tha t contained materialthat comigrated with an authentic sample of presqualene diphosphatewere combin ed, and solvent was removed to give 0.403 mCi (6.4% ra-diochemical yield) of material that comigrated with authenticpresqualene diphosphateon TLC [cellulose; 1:1:0.4 (v/v/v) 1-butanol/tetrahydrofuran/O.l M ammonium bicarbonate,R, = 0.31. The materialwas stored at -78 OC.

In a previous experiment using the same procedure, a sample of cold12 (5 9 mg, .14 mmol) was converted to2 (2 0 mg, 2%): 'H NM R (300MHz, CDCI,) 5.05 (br s , 4 H , vinyls), 4.87 (d, J = 8 Hz, vinyl adjace nt

to C3'), 4.0-2.6 (br , amm onium cou nterions, carbinyls), 1.96 (m, 16 H,vinyl methylenes), 1.64 (m, 9 H , vinyl methyls), 1.55(s , 12 H, vinylmethyls), 1.4-0.8 ppm (unresolved resonances for the cyclopropyl methyland C1' and C3' cyclopropyls); 31PN M R (121 MH z, CDCI,) -9.4 (brs, 1 P, P2), -1 1.3 (br s, 1 P, P l) ; negative ion FAB MS, m / z (re1 inten-sity) 585 (M - N3 HIo , 5) , 189 ( l l ) , 181 (12), 177 (14), 159 (30), 97(16).

Isolation of Yeast Microsomes. All procedures were a t 4 O C . Yeast(3 0 g) was suspended in 30 mL of 100 mM BHD A contain ing 100 pMDT T, pH 7.4, and the cells were disrupted at 15000-1800 0 psi in aFrench pressure cell. The suspension was spun at 23000g for 30 min.Th e supernatant was filtered through six layers of gauze and spun againat 23000g for 30 min. Th e resulting supernatan t was spun at 177000gfor 1.5 h, and the pellet was resuspended in 24 mL of100 m M B H D A ,10 pM DT T, 1 mM ED TA , and 10 mM MgC12, pH 7.4, by eight passeswith a Dounce homogenizer. Th e suspension was spun and resuspended

IR (CCI,) 296 0,29 20, 2860, 1718, 1450, 1375, 1350, 1280, 1225, 1190,

Poulter et a l.

again as described above. The m aterial was divided into 1-mL portionsand frozen at -20 OC.

Assays. (A ) Squalene Synthesis. Assays were in 500 pL of 50 m MBHDA, 11 mM KF, 5 mM M gCI2 ,50 p M D T T, 1 m M N A D P H , a n d0.2% Tween-80, pH 7.4, containing appropriate amounts of substrate (spact. 18-200 pCi/pmol) and yeast microsomes (2-7 fig of protein) in 12mm X 75 mm test tubes. Tubes containing all of the ingredients exceptfor microsomes were flushed with nitrogen, sealed with cork stoppers, andequilibrated at 30 OC for 10 min. Microsom es were added, and thecontents were mixed for 3s. Incubations were for I O min at 30 OC andwere stopped by addition of 250 pL of 1:l (v/v) 40% (w/v) aqueousKOH/ethanol, followed by mixing for3 s. Squalene (2 pL) and solid

Na Cl (sufficient to saturate the solution) were added. The mixture wasextracted with 1 mL of ligroin (b p 60-80 "C ), and a 0.8-m L portion ofthe ligroin layer was loaded onto a column containing2 mL of alumina(activity 11) in a IO-mL disposable syringe barrel. The assay mixture wasextracted with an additional I-m L portion of ligroin, and 0.8mL of theextract was added to the column. Th e combined extracts were eluted intoa scintillation vial with 9 mL of toluene, 1 mL of 4% Omnilluor cocktailwas added, and radioactivity was measured. All determinations were intriplicate agains t blanks. Calculations of specific activities (nmol min"mg-]) were corrected for the stoichiometry of the condensation withrespect to loss of a proton from C1 of one of the two farnesyl residues.The above assay was also used to measure squalene synthesis from [l-'H]presqualene diphosphate.

(B ) Presqualene Diphosphate Synthesis (Proton Release A ssay). T heproton release was modified for the simultaneou s evaluation ofpresqualene diphosphate and squalene synthesis. Assays were in 200 pLof 50 mM BHDA, 5 mM MgCI, , 50 p M D T T, 1 m M N A D P H , a n d0.2% Tween-80, pH 7.4, with 48 pg of protein. Incubations were con-ducted as described above. Th e reaction was quenched with80 pL of1:l (v/v) 40% (w/v) aqueous KOH/ethanol, and the resulting solutionwas saturated with NaC1. Squalene (2 pL) was added, and the solutionwas extracted three times with ligroin as before. The combined ligroinextracts were chromatogra phed, and radioactivity of the toluene eluatewas determined to measure synthesis of squalene. The remaining ligroinwas carefully removed from th e aqueous layer, 2 mL of methanol wasadded, and the mixture was mixed vigorously for1 min. T he mixturewas transferred to a 13 X 100 mM culture tube which was fitted witha glass tube elbow. Th e contents were distilled to near dryne ss at 78 OCwith a block heater into a test tube cooledon ice. A 0.75-mL portionof the distillate was removed, and its radioactivity was determined inInstagel (10 mL).

In order to compare specific activities for the proton release andsqualene assays, values determined by proton release were multiplied bya factor which reflected differences in proton stoichiometry and effi-ciencies of the two workups. With (f )-[ l-3 H]f arn esy ldiphosphate, 25%

of the total radioactivity was released in reaction 1, and the remaining75% was incorporated into squalene. Th e squalene workup was greaterthan 95% efficient, while the proton release workup only measured ap-proximately 33% of the released activity. When com bined, these featuresgive a correction factor of 8.6.

ResultsBuffer Inhibition. M o s t i n v e s t i g a t o r s h a v e u s e d a 55 mM

phospha te buffe r for s q u a l e n e s y n t h e t a s eassays. S i n c e i n o rg a ni cp y r o p h o s p h a t e (PPJ is a p r o d u c t of b o t h r e a c t i o n s c a t a l y z e d b yt h e e n z y m e , w e w e r e c o n c e r n e d t h a t t h e e n z y m e w a s s u s c e pt i bl eto inh ib i t ion in P i -con ta in ing buffe rs . An addi t iona l concern wast h a t h i g h c o n c e n t r a ti o n s of a m i n e b u f f e rs m i g h t a l s o i n h i b it t h eenzy me by b ind ing to th e reg ionof the ca ta ly t ic s i t e tha t normal lyaccommodates the carbocationic species proposedas in te rmedia tesdurin g catalysis . Figu re 1 shows the ac t iv ityof squa lene syn the tase

for convers ion of 1 t o 2 a s a f u n c t i o n o f b u f f e r c o n c e n t r a t i o n i nPi, Tr i s , a n d BHDA b u f fe r s a t p H 7.4 . T h e s l i g h t r is e in ac t iv i tybetween 10 a n d 25 mM in a l l th ree- sys tems was perhaps the resu l tof i n a d e q u a t e b u f f e r i n g a t l o w c o n c e n t r a t i o n s . A b o v e25 m M ,T r i s a n d Pi b o t h i nh i b it e d t h e en z y m e , wh i le B H D A d i d n o t . A sc o n c e n t r a t i o n s i n c r e a s e d , t h e d i f f e r e n c e s b e c a m e d r a m a t i c . In100 mM b u f f e r s, t h e a c t i v i t y o f s q u a l e n e s y n t h e t a s e i n Tr i s a n dPi dropped to 19% a n d 2%, respec t ive ly, of the BH D A va lues . A t55 m M Pi, h e c o n c e n t ra t i o n n o r m a l l y u s e d t oassay t h e e n z y m e ,s q u a l e n e s y n t h e t a s e h a d o n l y 19% o f t h e a c ti v i t y s e e n in B H D A .

A d i f f e r e n t p a t t e r n w a s s e e n f o r b u f f e r s c o n t a i n i n g Tr i s a n dPi or PPI. A s s h o w n in F i g u r e 2, PPI b e c a m e a p o t e n t i n h i b i t o r

(29) Rilling, H. C. J . Lipid Res. 1970, I Z , 480-485.


4/6

Squalene Synthetase J . A m . C h em .Soc., Vol. 111, N o. 10, 1989 3731

7 0. 4

n n I."I I I I

0 . 0 2 5 . 0 5 0 . 0 7 5 . 0 100.0

[Buffer] , mM

Figure 1. Rates for squalene synthesis from farnesyl diphosphate as afunction of buffer concentration. All buffers contained11 mM KF , 5mM MgCI2,50 pM DTT, 1 mM NADPH, and0.2% Tween-80. BHDA( 0 ) ; ris (m); phosphate (A). pH 7.4.

0.4,

m

L

i0 0 .0 0.5 1.0 1.5 2.0[PPi] , mM

Figure 3. Rates fo r squalene synthesis fromfarnesyl diphosphate as afunction of the concentrationsof PP, and 7: 0 ( O ) , 3 (B), 5 (A), 0 (v) ,and 20 pM ( e ) .

in 100 m M Tris. There wasa 45 % increase in activity at 0.1 m MPP, (vide infra), an d activity was essentially abo lished at2 m MPP,.

Th e individual effects of Tris andPPI in 2 5 m M B H D A w er esubstantially different. Whi le there wasno evidence for inhibitionby 0-20 m M Tris, low concentrationsof PP, elicited a markedstimulation of squalene synthetase. As shown in Figure 3, th e

Table I. Inhibition of Squalene Synthetase by AmmoniumAnalogues 6 , 7 , and 8

[substrate] [PPJ I,,analogue substrate (GM) (m M) (GM)

6 1 12 1 102 2 1 5

I 1 1 1 38 1 4 0 5

2 2 0 3

A

0 5 10 15 20

181. MM

25 -

k gw v

B

5 - 1

0 1 I I I 1

0 5 10 15 2 0 2 5

[81, 4 4

Figure 4. Inhibition of squalene synthetase by tethered ammoniumanalogue 8. (A) Rate of squalene synthesis from1 ( 0 ) nd from 2 (B).(B) rates for synthesis of2 (proton release,B) and 3 (squalene assay,0 )from 1 .

ra te of squalene synthesis almost doubled when 1 mMPPI wasadded to 50 m M BH DA. At higher concentrations,PP , slowedthe rate of squalene synthesis, as expected for a product inhibitor.Th e magn itude of stimulation/inhibition varied with differentpreparations of microsomes. When a fresh preparation was al-lowed to stand at 5 O C ,the activity of squalene synthetase inPPi-free buffer slowly increased until, aft er 3 0 days, no additional

stimulation wasseen between 0 and 1 mM PP,. As the microsomesaged further, the activity of the enzyme gradually decreased. Thestimulation in activity elicited byPP, in BHDA buffer was notseen for two other microsomal enzymes, cytochromec reductase30and g l u c o ~ e - 6 - p h o s p h a t a s e , ~ ~nd at this point we have no ex-planation to offer for th e simulation.

Inhibition by Ammonium Analogues 6 and 7. In BH DA buffercontaining chloride and fluoride as the only other anions, am -monium analogues 6 and 7 did not inhibit squalene synthetaseat concentrationsup to 17 0 bM . However, as illustrated in Figure3, potent inhibition of enzymatic activity was noted whenPP, was

(30) Aronson, N. N.; Touster, 0 . Methods Enzymol.1975 ,31A,90-102.(3 1 ) Sottocasa, G . L.; Kuylenstierna,B.; Ernster,L.; Bergstrand,A . J . Cell

Bioi. 1967, 32, 415-437.


5/6

3138

Scheme 11. Mechanisms for Synthesisof Squalene a nd Phytoene

J . A m . Chem. SOC., ol. 111, N o. 10, 1989 Poulter et a l .

IN A D P t N A D P

added to the buffer. A plot of the ra te of squalene synthesis versus[PPJ in the presence of 20pM 7 decreased sharply as the con-centrat ion of PPI increased from0 to 1 mM. At [PPI ]= 1 mM,7 was a potent inhibitor with I , , = 3 pM (the concentration of

inhibitor required for 50 % inhibition) in the presence of 1 m MNA DP H. Similar profiles were observed for inhibition of synthesisof 3 from 1 and from 2 by primary am monium analogue6 . Is0values for the inhibitors are listed in Table I.

Inhibition by Tethered Ammonium Analogue8. Figure 4A showsthe results for inhibition of react ion 1(1 2) a nd re ac ti on 2 ( 2-+ 3) by tethered ammonium analogue8. In contrast to exper-iments with analogues 6 and 7, 8 was a potent inhibitor of theenzyme in th e absence of PP I and had a similar effect on bothreactions. Figure 4B shows inhibition of reactions 1 and 2 in anexperiment where the rates of proton release from C1(1- )and accumulation of radiolabeled 3 (1 - ) were measured si-multaneously. Tethered ammonium analogue8 had similar effectson the rates of both reactions.

Discussion

Two major classes of isoprenoid metabolites, sterols! andcarotenoid^,^^ar e derived from simple 1-1 precurso rs. In sterolmetabolism, the condensation of farnesyl diphosphate proceedsin two steps. The first yields (lR,2R,3R)-presq uaIene di-phosphatelo [( R,2R,3R)-2 , R = Cl IH l9 ], cl-2-3-fused tri-terpene, and the second gives squalene(3 , R = C I I H 1 9 ) it h asaturated 1-1 linkage. Th e second reaction proceeds with in-version of configuration at th e C1 and C1 centers as outlined inScheme I. In carotenoid metabolism two molecules of geranyl-geranyl diphosphate are condensed to give (lR,2R,3R)-?re-phytoene diphosphateI0[( R,2R,3R)-2 , R = C16H27], hich inturn rearranges to phytoene(9, R = C&,,) with a 1-1 dou blebond. Since squalene synthetase requires NA D PH and phytoenesynthet ase does not, it is resonable to assume that the 1-1 con-densations have similar mechanisms until the final stage of reaction2 where reduction by NA DP H produces the saturated1-1 linkagein 3 and elimination generates the central double bond in9 (seeScheme 11).

Several groups proposed tha t the c1-2-3 to 1-1 rea rran ge-ments catalyzed by squalene and phytoene synthetases involvecyclopropylcarbinyl cations.-I8 Th e mechanisms suggested byP o u l t e r a n d c o - w ~ r k e r s ~ ~ , ~ *n Schemes I and I1 proceed byisomerization of primary cation 4 to its more stable tertiary isomer5 by a ring expansion/contraction sequence initiated by migrationof the C2-CI cyclopropane bond. In sterol biosynthesis thetertiary C30 species(5 , R = C 1 , H I 9 )s quenched by a hydride fromNA DP H to give3 ( R = Cl,H19), nd in carotenoid biosynthesisits Cd0( 5 , R = C16 H27 ) ounterpart loses a proton to give9 .However, the rearra ngem ents of cyclopropylcarbinyl cations ar enotoriously sensitive to substitution patterns on the cyclopropanering. Poulter and co-workersI6 examined th e solvolytic rear-

rangements of the chrysanthemyl cation (4, R= CH3 ), a s implemodel for presqualeneor prephytoene disphosphate (see Schem eIII), and found the rearrangement 4- required for biosynthesisof 1-I-fused isoprenoids was only a minor reaction channel (ca0.04%). Two competing rearrangements, 4-. 10 and 4 - f ,were observed, with the lat ter accounting fo r most of4. It becameevident that t he enzym es must exert strict regiocontrol if cyclo-propylcarbinyl cations were indeed intermediates in the en-zyme-catalyzed rearrangements.

A least-motion mechanism consistent with properties of cy-clopropylcarbinyl cations and the stereochemical studies of

(32) Spurgeon, S. L.; Porter, J. W . Biosynthesis of Isoprenoid Compounds:Porter, J . W.: Spurgeon, S. L. , Eds.; Wiley: Ne w York, 1981; Vol. 2, pp1-122.

Scheme 111. Rearrangements for Primary CyclopropylcarbinylCation 4

C l -2 e l - 2 -3 1 - 3

10 4 1 1

004% a1

c 1 - 1 - 2

5

C or nf or th a nd P ~ p j a k , ~ ~as proposed by Poulter and co-work-e r ~ . ~ , h e [1.21 sigmatr opic shifts of carbon-carbon bondsduring the ring expansion/contraction sequences of cyclo-propylcarbinyl rearrangements are suprafacial, and th e rotationalbarrier between the carbinyl carbon and the cyclopropane ringis too high to permit isomerization within the lifetimes of the~ a r b o c a t i o n s . ~ ~hus, for the rearrangement catalyzed by squalenesynthetase, inversion at C1 during 4- requires that 2 be boundto the active site in a conformation where the C1 -0 bond is synto the C2-C3 cyclopropane bond. The resulting conform ationof the 4-PPI tight ion pair generated upon heterolysis of the C - Obond is optimal for stabilizing 4-PP,-. P P , with positive chargemigrating from C1 (cation 4) across C2 (species which inter-connect 4 and 5 ) to C3 (cat ion 5 ) . Rearrangements 4-PP, -lO-PP, and 4-PP,- l-P P, are accompanied by large separationsof positive and negative centers in the ion pairs with a resultingloss of electrostatic stabilization . A selective stabilizat ion of 4- by approximately8 kcal/mol would be sufficientto accountfor the high regioselectivity of the enzym atic reaction. This isa reasonab le value to achieve in a tigh t ion pair given the distancesbetween centers bearing charge in the various isomeric ion pairs.35

Following 4 -+ 5 (R = C l l H 1 9 ) ,he reaction is completed byaddition of hydride from NADPH with concomitant opening ofthe cyclopropane ring. Stereoelectronic considerations dictate that

the addition occurs with inversion atC 1.36337 Con sequ entl y, th ecofactor must be bound in a conform ation to be properly positionedfor transfer of hydride following the rearrange ment. Pre matu retransfer of hydride from the cofactor to 4 is blocked becauseaddition would require retention at C1 and, thus, encounters astereoelectronic barrier . It is only after rearrangem ent of 4 to5 that the barrier vanishes.

Inhibition studies with ammonium analogues6 an d 7 constitutestrong support for the ion-pair mechanism. It was clear that theinhibitors did not bind tightly to the active site of squalenesynthetase in spite of their close resemblance to carbocations 4an d 5 . In fac t, we were initially disappointed when 170pMconcentrations of the ammonium analogues failed to inhibit theenzyme. These concentrations exceededSS0 substrate concen-trations required for half-maximal activity in the presenceof 1m M N A D P H ) f o r 1 (10 wM) and 2 (8 pM) by approximately15-20-fold.* However, 20 K M concentrations of the amm oniumanalogues, which failed to inhibit, and 1 mM PPI, which stimulatedthe enzyme, combined to produce potent ion-paired inhibitors.For many e nzymes which process isoprenoid d iphosph ates, thediphosphate moiety is a principal determinate forPresumably, complementary interactions between the hydrocarbon

(33) Edmond, J.; Popjak, G.; Wong, S . M.; Williams, V. P. J . Eiol. Chem.

(34) Wiberg, K. B .; Hess, B. A,; Ashe, A. J. Carbonium Ions; Olah, G .A.; Schleyer, P. v . R., Eds.; Wiley: New York, 972; Vol. 3, pp 1295-1345.

(35) Cantor, C. R. ; Schimmel, P. R. Biophysical Chemistry; W. H.Freeman: San Francisco, 1980; pp 290-291.

(36) Poulter, C. D.; Friederich, E. C.; Winstein,S . J . A m . Chem. SOC.1970, 92 , 4274-4281.

(37) Poulter, C. D.; Winstein, S . J . Am . Chem. SOC. 970, 92, 4282-4288.

1971, 246, 6254-6271.


6/6

Squalene Synthetase

moieties of 6 an d 7 and squalene synthetase w ere insufficient toovercome electrostatic repulsions between the unshielded am-monium moieties and the enzyme.Zsis fo r 6 an d 7 against 1 and2 in buffer containingPPi indicated that the diph osph ate unit wasparticularly effective in promoting binding of the ammoniumanalogues, as opposed to other anions such as chlorideor B H D A .Similar synergistic phenomena were observed, albeit a t 104-foldhigher concentrations, for mixtures of Tris andPPI,where thereis no apparent stru ctural similarity between the am mon ium speciesand th e hydrocarbon regions of carbocations4 an d 5. In additionto mechanistic insight, these studies emphasize the caution on emust exercise in developing assays for squalene synthe tase.

Analogue 8 was designed to increase the effective concentrationof PPi in the presence of 6 by tethering a phosphonophosphatemoiety to the amm oni um analogue . Exam ination of a space-fillingmodel suggested tha ta two-carbon spacer between the nitrogenof 6 and a nonbridging oxygen inPP, would permit the tetheredmolecule to fold into a relatively strain-free conformation witha nonbridging oxygen on the remote phosphorus near the positivelycharged amm onium center. At temp ts to synthesizea tethereddiphosphate analo gue of6 were unsuccessful, presumably becauseof the reactivity of the 2-aminoethyl diphosphate linkage, andphosphonophosphate 8 was prepared ins tead. Th e compoundproved to be a potent inhibitor of squalene synthesis.Z50)s of 3-5p M were found against1 an d 2, values similar to those measuredfo r 6 an d 7 in the presence of 1 m MPP,. As illustrated in Figu re3, the effectiveness of 6 and 7 decreased rapidly at lower con-centrations of PPi, and the analogues did not inhibit in bufferscontaining 3 p M PPi. Thus, the tether increased the effectiveconcentration of PP , at least 300-fold, notwithstanding (1 ) a nexpected decrease in potency when the dipho sphate moiety wasreplaced with a phosphonophosphate moiety and (2) steric andhydrophobic perturbations associated with introduction of a hy-drophobic ethyl mo iety into the diphosp hate binding region of thecatalytic site.

Rilling an d c o - ~ o r k e r s ~ ~ ~iscovered that squalen esynthetaseis a relatively small protein composed ofa single polypeptide chainwhich catalyzes synthesis of presqualene diphosphate and squalene.In microsomal preparations, 1 was converted to 3 with the ac-cumulation of relatively little2. However, the activities for thetwo steps became clearly delineated when the enzym e was solu-bilized. Un der these conditions the activity for reaction1 was

approximately Sfo ld that of react ion 2 .8Rillings observations raise the question of whether2 an d 3 aresynthesized a t the sameor different catalytic sites. In earlierinhibition studiesof squalene biosynthesis in rat liver microsomes,Corey and Volante3* eported th at a phosphonophosphate analogueof presqualene disphosphate blocked form ation of squalene frommevalonate and 2 under co nditions where some synthesis of2 frommev alonate was seen. Althou gh these results ar e consistent witha two-site hypothesis, such high levels of inhibitor(0.5-1 m M )were required to completely block squalene formation th at non-specific inhibition m a)/ have occurred. Mor e recently, Bertolinoet discovered that a variety of detergent-like analog ues were

J . A m . Chem. Soc., Vol. 111, N o. 10 , 1989 3139

(38) Corey, E. J .; Volante, R. P. J . Am . Chem. SOC. 976, 98, 291-1293.(39 ) Bertolino, A.; Altman, L. J. ; Vasak, J . ; Rilling, H . C. Biochim. Bio-

phys . Acta 1978, 530, 17-23.

nonspecific inhibitors of squalene synthetase.Agnew and Popjak40 noticed severe inhibition of t he second

reaction, but not the first, when th e rates of reactions 1 and 2 weremeasured simultaneously in microsomal prep arations of squalenesynthetase. The inhibition was transitory, and as1 was consumed,the two activities, equalized. They proposed that th e time courseof the recovery indicated that the enzyme had a distinct activesite for each reaction. W e suggest tha t their da ta are also con-sistent with a single-site mechan ism. Since the prenyl transferstep (reaction 1) utilizes two molecules of1, i ts rate depends on[112,and a n increase in the concentration of1 will only serve toaccelerate reaction 1 . However, at high concentrations,1 cancompete with 2 for binding and, thereby, inhibit reaction2 .

Our experiments with tethered analogue8 support a single-sitehypothesis. In the absence of the tethered in hibitor specificallydesigned to block reaction 2, the rates of the two reactions werealmost equal. As illustrated in Figure 4A,8 inhibited squalenesynthesis from 1 and 2 to similar extents. In addition when therates of both reactions were measu red simultaneously with1 asthe initial substrate (Figure4B), the extent to which reactions1 and 2 were inhibited was almost identical. The simplest ex-planation of these results is that both are catalyzed at a singleactive site, and when8 binds, both are blocked equally. However,more complicated scenarios which cannot be excluded at thepresent time include two spatially overlapping sitesor nonover-lapping but strongly interacting sites where binding of8 t o onecompletely blocks catalysis at the other.

From the inhibi t ion s tudies , we conclude that squalenesynthetase has a single active site tha t catalyzes both reactionsfor synthesis of squalene from farnesyl diphosphate and thatcyclopropylcarbinyl cations are reactive intermediates in the re-arrangement of2 to 3. In p articular, inhibitors whose topologicaland electrostatic properties mimic primary cation 4 and tertiarycation 5, originally proposed as intermediates by Poulter andc o - ~ o r k e r s ~ ~ ~ ~ ~ - ~ *ind tightly to the enzyme. However, bindingof the positively charg ed species depends upon the availabilityofPP,, and it is the cyclopropylcarbinyl cation (am monium ana -logue)-PP, ion pair that is tightly bound. W e further suggest thatPP , also provides a negatively charged template which controlsthe regioselectivity of the carbocation fragment during the re-arrangemen t .

A simple minimal mechanism can be proposed. Onc e squalene

synthetase has sequestered presqualene diphosphate and NADPHin the proper orientation, catalysisis achieved by activationofthe diphosphate moiety, presumably by neutralizing negativecharge in the leaving group. By maintaining the original relativeorientation of the positive and negative fragments during therearrangements, the regioselective rearrangement of 4-PP,-*5-PP, which ensues is driven by elec trostatic stabilization withinenzym e-boun d tight ion pairs. Similarities between the reactionscatalyzed by squalene synthetase and phytoene synthetase suggestthat the latter utilizes similar chemistry to convert prephytoenediphosphate to phytoene.

Acknowledgment. This work was supported byN I H Gran t sG M 2 1 32 8 a n d G M 25521.

(40 ) Popjak, G . : Agnew, W. S. J . Biot. Chem. 1978, 253, 4566-4573.

Documents

Squalene Synthase Inhibitors_Todd L Capson