Substrate specificity of squalene synthetase

218

Biochimica et Biophysics @I Elsevier/North-Holland

Acta, 617 (1980) 218-224 Biomedical Press

BBA 57499

SUBSTRATE SPECIFICITY OF SQUALENE SYNTHETASE *

TANETOSHI Chemical

KOYAMA,

Research

KYOZO

Institute

OGURA

of Non-Aqueous

and SHUICHI Solutions,

SET0 Tohoku

University,

Sendai

980

(Japan) (Received

July 5th, 1979)

Key words: microsome)

Squalene

synthetase;

Substrate

specificity;

Farnesyl

pyrophosphate;(Pig

liver

Summary 22 artificial homologues of farnesyl pyrophosphate were examined for the reactivity as substrate for squalene synthetase of pig liver microsomes. 16 of the homologues were found to be reactive to give corresponding squalene-like products. Extention of the w-terminal of the carbon chain of farnesyl pyrophosphate is acceptable to the enzyme at least by two carbon atoms in either tram or cis direction (2E,6E)-3,7,1l-Trimethyldodeca-2,6-dienyland (2E,6E)-3,7dimethyldodeca-2,6dienyl pyrophosphates are both good substrates, whereas (2E,6E)-3,7dimethylundeca-2,6dienyl-, (2E,6E)-3,7-dimethyltrideca-2,6dienyl-, (2E,6E)-3,7dimethyltetradeca-2,6dienyl-, (2E,6E)-3,7,10trimethylundeca-2,6dienyl-, and (2E,6E)-3,7,12-trimethyltrideca-2,6dienyl pyrophosphates are poor substrates. These results indicate that the carbon chain length rather than 10,lldouble bond is important for the reactivity as substrate. Replacement of 3-methyl of famesyl pyrophosphate by an ethyl group or introduction of a methyl group at C-4 results in a complete loss of activity. Introduction Squalene synthetase is a particulate enzyme which occurs widely in eucaryotic cells. This enzyme catalyzes a carbon to carbon bond formation between two

* Supplementary data to this article are deposited with, and can be obtained from. Elsevier/NorthHolland Biomedical Press B.V., BBA Data Deposition, P.O. Box 1345. 1000 BH Amsterdam. The Netherlands. Reference should be made to No. BBA/DD/122/57499/617 (1980) 218-224. The suPplementary information includes details of the preparation of the farnesyl pyrophosphate homologues.

219

molecules of farnesyl pyrophosphate in a very novel fashion to form the hydrocarbon, squalene. The substrate specificity with respect to the artificial substrate homologues of this enzyme has recently been studied in this and other laboratories [l--8]. The structural modification of farnesyl pyrophosphate with respect to the alkyl region near the pyrophosphate moiety greatly diminishes or loses the reactivity as substrate of squalene synthetase [4-71. However, some homologues in which only the alkyl region near the w-end was modified were found to be relatively good substrates for the enzyme to give the corresponding squalene-like products [l-3,7]. These results led us to investigate in more detail the effect of modification of the carbon chain on the reactivity. This paper describes the reactivities of various artificial homologues of famesyl pyrophosphate especially with respect to the terminal alkyl region. Part of the results with eight of the 22 substrate homologues has been the subject of preliminary communications [ 1,2]. Materials and Methods Famesyl pyrophosphate synthetase was purified from pig liver in the presence of 10 mM 2-mercaptoethanol by the method reported by us [9]. Doubly washed microsomes of pig liver were prepared according to the method of Goodman and Popjik [lo]. Glucose 6-phosphate, glucose-g-phosphate dehydrogenase, and NADP were purchased from Boehringer Mannheim GmbH. [ 1-‘4C]Isopentenyl pyrophosphate, dimethylallyl pyrophosphate, geranyl pyrophosphate, and their homologues used in the enzymatic synthesis of various famesyl pyrophosphate homologues were all the same preparations as used in the previous studies [ 11-161. Preparation of radiolabeled farnesyl pyrophosphate homologues. All the famesyl pyrophosphate homologues used in this study were synthesized enzymatically using the pig liver farnesyl pyrophosphate synthetase [ll-15, 17-191: For a general procedure, an incubation mixture containing, in a final volume of 30 ml, 600 ~01 of Tris-HCl buffer, pH 7.7, 150 /_mol of MgCl,, 300 pm01 of KF, 300 ~01 of 2-mercaptoethanol, 750 nmol of [l-‘4C]isopentenyl pyrophosphate (spec. act., 1.2 Ci/mol) or its 3H-labeled homologues, 1.5 ~01 of an appropriate homologue of dimethylallyl pyrophosphate or geranyl pyrophosphate, and 3 mg of farnesyl pyrophosphate synthetase was incubated at 37°C for 90 min. The reaction was terminated by the addition of 0.6 ml 6 N NaOH. (NH4@04 (approx. 20 g) was added and famesyl pyrophosphate homologue synthesized enzymatically was extracted with butan-l-01. After washing with water the butanol extracts were combined and concentrated on a rotary evaporator at 50°C. The residue was then applied to a preparative silica gel TLC plate, which was developed with a solvent system of propan-l-ol/NH40H/H20 (6 : 3 : 1, v/v). The radioactive pyrophosphate ester appearing around RF 0.31 was extracted with water. In this way 22 radiolabeled homologues of farnesyl pyrophosphate were prepared as per details deposited in the BBA data bank, and were subjected to the enzymatic reactions.

FPP : R, = R, = CH, 1 : R, = C,H,, R2 = CH, 2 : R, = n-&H,, R, = CH, 3 : R, = CH,, RZ = C,H, 4 : RI = CH,, R2 = n-C,H, 5 : R, = CH,, R, = n-C,H,

11 :n=Z 12 :n = 3 13:n=4

18:n -1 19:n=2 20 : n = 3

GPP : R = CH, 6 : R = n-C,H9 7 : R = n-C5H,, 8 : R * n-&H,, 9 : R = n-C7H,5 10 : R = n-C,H,,

14 15 16 17

: : : :

R, R, R, R,

= = = =

21 22

R2 = C,H,, R, = CH, R2 = CH,, R, = C,H, CH,, R, = R, = C,H, RE = R, = C,H,

: R, = H, : R, = R,

Rp = CH, = CH,

OPP = OP,0,H3 Scheme

I. Structures

of 22 homologues

of farnesvl

pyrophosphate

(FPP).

GPP, EeraW’l wrwhosphate.

Enzyme assay. The reaction mixture contained, in a final volume of 3 ml, 30 pmol of MgC12, 350 pmol of potassium phosphate buffer, pH 7.4, 5 ~01 of NADP, 5 ~01 of glucose 6-phosphate, 5 units of glucose-6-phosphate dehydrogenase, 20 nmol of radiolabeled famesyl pyrophosphate homologues, and 100 mg of doubly washed microsomes. The incubation was performed at 37°C for 60 min. After the addition of 2 ml of 15% (w/v) KOH/&H,OH solution, the mixture was heated at 65’C for 2 h. Radioactive non-saponifiable materials were extracted with light petroleum and the extracts were washed with water. The extracts were applied to a silica gel thin-layer plate, which was developed with hexane. Radioactivity peaks were located with an Aloka thinlayer radiochromatoscanner (Nippon Musen Co.). The radioactive regions were scraped and the products were extracted with diethyl ether. The radioactivity in the extracts was determined in a toluene scintillator with a Packard liquid scintillation counter Model 3390 and the total conversion into squalene-like products was calculated. Identification of the product. For gas chromatography-mass spectrometer-y (CC-MS), a large-scale incubation (20 times as above) was carried out and the non-polar products extracted with light petroleum were purified twice by preparative TLC (silica gel) with hexane. The purified products were submitted to GC-MS with a Shimadzu-LKB gas chromatography-mass spectrometry system type 9000.

221

Results The TLC radiochromatogram of the products derived from the reaction of the natural substrate, farnesyl pyrophosphate gave two radioactivity peaks at the origin and at the RF value of 0.32. The material at the origin was assigned to farnesol formed by the action of microsomal phosphatase. The latter peak corresponded to the hydrocarbon product, squalene. In a similar manner 16 out of 22 sets of incubation gave squalene-like hydrocarbon products at RF values of 0.3-0.6 on TLC. Radioactivity peak at the origin was also found in every set of reaction and it was assigned to the corresponding alcohol of the famesyl pyrophosphate homologue by CC-MS. In Table I are listed the RF values and the yield of the hydrocarbon products along with those of the alcohols at the origin. In order to confirm the structure of the squalene-like product, each hydrocarbon fraction was subjected to GC-MS. The hydrocarbon product obtained by the reaction of 1 gave a major peak in GLC at a retention volume of 1.77 relative to that of squalene which was usually found in the reaction extracts as an endogeneous component in microsomes. The mass spectrum (Fig. 1A) indicated that the product was 1,24dimethylsqualene (23). It showed a molecular ion at m/e 438 corresponding to C3*Hs4. The base peak at m/e 83 can be reasonably assigned to (C2H5C(CH3) =

TABLE I PERCENT CONVERSION OF FARNESYL PYROPHOSPHATE TION CATALYZED BY PIG LIVER MICROSOMES

(FPP) HOMOLOGUES

n.d., not detected. Substrate

FPP 1 z 3 4 5 s ? s 9 10 ii 12 13 iz 15 16 i-i 18 19 I;ii z 22 -

Alcohol (origin)

% hydrocarbon

(%)

(RF)

29.7 26.0 70.6 33.1 55.2 66.1 38.0 44.7 73.0 71.3 86.4 21.1 42.4 62.7 56.5 74.6 55.8 48.1 21.8 61.2 58.2 40.6 86.6

69.7 62.3 1.6 40.1 3.1 n.d. 3.4 32.0 1.4 0.2 0.2 3.0 14.1 0.8 3.8 n.d. n.d. n.d. 3.6 0.7 0.3 n.d. n.d.

(0.32) (0.36) (0.36) (0.34) (0.36) (0.63) (0.66) (0.67) (0.51) (0.47) (0.63) (0.60) (0.59) (0.34)

(0.33) (0.33) (0.30)

Relative activity

1 0.894 0.023 0.575 0.044 0.000 0.049 0.459 0.020 0.003 0.003 0.043 0.202 0.011 0.055 0.000 0.000 0.000 0.052 0.010 0.004 0.000 0.000

IN THE REAC-

Fig.

1.

column The

Mass

spectra

packed

accelerating

with

of 2%

voltage

the

hydrocarbon

Dexsil of MS

300 was

GC

products 80/100

derived

Chromosorb

from W-A%

1

(A). DtiC?

7

(B),

and

was

used

12 for

(CL GLC

A

1-m at

glass

22O’C.

70 eV.

CHCH’,) which is a homologue of (CH,C(CH,) = CHCH’,), m/e 69, observed as the base peak in the spectrum of squalene. Peaks were also observed at m/e 355 (M - 83), 219 (83 + 68 X 2), and 151(83 + 68). The squalene-like material obtained by the incubation of 3 emerged at a retention volume of 1.50 relative to that of squalene. The mass spectrum witc almost the same as that of 23, indicating that the product was a geometr;. isomer of 1,24dimethylsqual~e, 24. The hydrocarbon product derived from 7 had a retention volume of 0.50 relative to that of squalene in GLC, and its m-assspectrum was compatible with the structure of 2,23-dimethyl-2,3,22,23_tetrahydrosqualene (25). The molecular ion was found at m/e 386 corresponding to Cz8HSowith fragments at m/e 329 (M-57), 287 (M-99), 261 (M-57-68), 193 (57+68X2),125 (57 + 68), and 69 which was the base peak (Fig. 1B). The squalene-like substance obtained by the incubation of 12 had a retention volume of 0.77 in GLC relative to that of squalene. The rnz spectrum of

223

this material showed a molecular ion at m/e 414 corresponding to C3,,HS4. Peaks were also observed at m/e 343 (M - 71), 301 (M - 113), 275 (M - 7168), 207 (71+ 68 X 2), and the base peak at 83. These results support the structure of 2,3,22,23_tetrahydrosqualene (26) (Fig. 1C). The rest of the farnesyl pyrophosphate homologues showed too low reactivities to give the squalene-like product in an amount sufficient for the mass spectrometry. It is reasonable to assume, however, that 2, 4, 6, 8, 9, 10, 11, 13 -7 14 -._) 18 _, 19 and 20 are active as substrates since they gave !iy&oGbon -3 products showing sima characteristic movement on TLC as shown in Table I. Discussion In our preliminary communications we reported that the famesyl pyrophosphate homologues 2, 5, 5, and 8 were inactive as substrate for squalene synthetase in contrast with the high reactivity of 1, 3, 7, and 12 [1,2]. However, the formers were also found to be reactive, if no3 .% highly as the latters, w-hen assayed under improved conditions using an increased amount of NADPH and microsomes. Furthermore, eight new famesyl pyrophosphate homologues were found to act as substrates. These 16 homologues were compared in terms of the yield of hydrocarbon synthesis in the reaction with squalene synthetase. Six homologues, 5, 15, 16, l7, 2J, and -22 gave no detectable amount of hydrocarbon product as shown in Table I. It is reasonable that neither 16 nor 17 is reactive at all since even _._. 15 in which only 3-methyl group of farnesyl pyro--. phosphate is replaced by an ethyl group is not acceptable as substrate. Ortiz de Montellano et al. have also reported 15 to be inactive toward yeast or rat liver squalene synthetases [ 61. They have%0 reported that neither 4-methylthiofarnesyl pyrophosphate nor 4fluorofamesyl pyrophosphate is active as substrate for the enzymes. As 21 and 22 are also inert as substrate, it can be said that modification at the 4-positionby any substituent is hardly acceptable. It seems very strange that the modification of the 4-position or 3-methyl group is not acceptable, because even 2-methylfamesyl pyrophosphate and 3-demethylfamesyl pyrophosphate have been reported to be somewhat active as substrates [6,7]. There might be an essential binding site for the central alkyl region of famesyl pyrophosphate since saturation of the double bond at the 6,7-position also makes the substrate inert [ 6,7]. 1 and 3 which are both derived by Among the 16 reactive homologues, extending the w-end of famesyl pyrophosphate Gy one carbon atom are highly acceptable as substrates to give the corresponding squalene-like hydrocarbons. Polito et al. have also reported that 1 is enzymatically converted to 23, which has been shown not to be converted?nto the corresponding sterol [ 31. Extention of the terminal by two carbon atoms in either trans or cis direction, however, sharply diminishes the reactivity. A cyclic isomer (l8), which is formally derived from 2 or 4 by joining the extra carbon chain to lo-methyl to form a five-membered ring, shows a higher reactivity than 2, or 4. Even homologues with extra three or four carbons are active if the extra carbzns forms a ring, six or seven membered, at the terminal end. These results suggest that the steric perturbation due to free rotation of an open alkyl chain is likely to be the major factor for an unfavorable binding at the active site of the enzyme. This

224

conclusion is also compatible with the fact that 5 is inert as substrate. Although Qureshi et al. have reported that even ge%ylgeranyl pyrophosphate is converted to lycopersene by yeast squalene synthetase [ZO] , our preliminary experiments showed that pig liver microsomes could not accept the CzOpyrophosphate as substrate. In contrast with the fact that saturation of the central double bond makes the substrate inert [ 71, the terminal double bond is not essential for the reactivity because the homologues, 2, 7,8, z,_lO, ll,l2, and 13 all of which lack the double bond at 10, 11 position %gactive as substrates. Among these artificial substrates 7 and 12 both of which have the same carbon chain length as that of farnesyl peophozhate are good substrates. The carbon chain length is still rather important for the activity than the terminal double bond because as the chain length falls off from that of farnesyl pyrophosphate, the loss of the reactivity becomes serious. We have also observed that geranyl pyrophosphate can never be the substrate of squalene synthetase. It is of interest that 7 can be a better substrate than 12, though the latter resembles famesyl pyroph%sphate more closely than the former does. The gemdimethyl structure at the terminal position seems not to be so important because 11 and 13 are rather less reactive as compared with 6 and 8_,respectively. -

This work was supported by a Grant-in-Aid for Encouragement of Young Scientist (278059) from the Ministry of Education, Science, and Culture of Japan and by the Asahi Glass Foundation for the Contribution to Industrial Technology. References 1

Ogura,

2

Koyama,

3

Pobto,

4

Ortiz

K.,

Koyama, T..

A..

Ogura,

Popjak,

de

T. and

Seto.

S. (1972)

J. Am.

Chem.

K. and

Seto,

S. (1974)

Chem.

Lett.

G. and

Mont&arm,

Parker.

P.R.,

T. (1972)

Castillo,

R.,

J. Biol.

Sot.

Chem.

Vinson,

W.

94.307-309

529-530 247.3464-3470

and

Wei,

J.S.

(1976)

J.

Am.

Chem.

Chem.

Sot.

Sot..

98,

2018-2019 5

Ortiz

de

Montellano,

P.R.,

Castillo,

P.R.,

Wei,

R.,

Vinson,

W.

and Wei,

J.S.

(1976)

Castillo,

R. and

J. Am.

98.

3020-

3021 6

Ortiz 16,

de Montellano.

J.S.,

Vinson,

W.A.,

Boparai,

A.S.

(1977)

Biochemistry

2680-2685

7

Washburn,

8

Ortiz

9

Koyama,

W.N.

and

Kow,

de Montellano, T.,

Saito.

10

Goodman,

D.S.

11

Ogura,

Nishino,

12

Nishino,

13

Nishino.

14

Ggura,

15

Koyama,

K., T., T.. K..

and

P.R. Y.,

T.,

Ogura,

Popjdk, T.,

Tetrahedron

Boparai, K. and

A.9 Seto,

G. (1960)

Koyama.

Lett.

(1977) Res.

Seto,

S. (1970)

K. and

Seto.

S. (1971)

Biochim.

Ogura,

K. and

Seto,

S. (1972)

J. Am.

Chem.

S. (1972)

J. C&em.

Ogura,

K. and

Seto,

S. (1973)

Chem.

Ogura,

K. and

Seto,

S. (1973)

J. Biol.

Kouama, Popjak,

G.,

Hoilowa~,

18

Popidk,

G.,

Rabinowitz,

19

Koyama,

T.,

20

Qureshi,

A.A..

Ogura,

P.W. J.L.

K. and

Barnes.

F.J.,

and

Baron,

J.M.

and Baron, Seto,

J.M.

S. (1977)

Semmler.

(1969)

E.J.

and

Lett.

94.

Chem.

Commun.

76520-525

82,1585-1590

Sot.

235,

92,

60366041

322-325

6849-853 Commun.

881-882

401-404

Chem.

248,

Biochem. Chem. J.W.

8043-8051 J. 111.

Biochem.

Porter,

Chem. Aeta

Sot.

Sot.

(1969)

J. Am.

J. Am.

Biophys.

Seto,

16

Res.

(Tokyo)

1, 286-300

T. and

17

Biophys.

J. Biochem.

J. Lipid

T. and

1555-1558

Biochem.

S. (1977)

Ogura, Koyama,

T.,

R. (1577) and

Sot.

J. 113,

325-332 861-868

99.1999-2000

(1973)

J. Biol.

Chem.

248,

2755-2767