Eur J Blochem. 202, 789 - 796 (1 991) V' FEBS 1991

Characterization and distribution of cis-prenyl transferase participating in liver microsomal polyisoprenoid biosynthesis Johan ERICSSON I , Anders THELIN ', Tadeusz CHOJNACKI' and Gustav DALLNER

' Department of Biochemistry, University of Stockholm, Sweden Institute of Biochemistry and Biophysics, Polish Academy of Sciences, Warsaw, Poland Clinical Research Center, F42, Division of Basic Research in Dementia, Huddinge Hospital, Sweden

(Received June 26/September 2, 1991) - EJB 91 0834

The properties of rat liver cis-prenyl transferase, mediating the synthesis of polyisoprenoid pyrophosphate from trans,truns-Farnesyl pyrophosphate and [3H]isopentenyl pyrophosphate were studied. The K, values for farnesyl pyrophosphate and isopentenyl pyrophosphate were found to be 25 pM and 4.4 pM, respectively. Appropriate conditions were established to measure the condensation reaction, which was linear during the first hour using 1 mg microsomal protein. Various detergents could solubilize the enzyme, but the presence of Triton X-100 was required during the incubation to obtain full activity. There was also an absolute requirement for Mg2+ and the pH maximum was 7.0. Inorganic phosphate, especially pyrophosphate, proved to be inhibitory. cis-Prenyl transferase is associated mainly with the cytoplasmic surface of rough microsomes and, to some extent, also with smooth I microsomes, but was almost absent from smooth I1 microsomes. At all localizations, the product is polyprenyl pyrophosphate and to some extent, also polyprenyl monophosphate. The isoprenoids formed contain 15 - 18 units in the presence of detergents and 16 - 20 units in the absence of detergents.

Dolichol biosynthesis, starting from farnesyl pyrophos- phate (FPP), involves complex and sequential enzymatic reac- tions assumed to be mediated by a single enzyme, cis-prenyl transferase [l]. The end-product of this enzyme, polyprenyl pyrophosphate, must be modified by the action of phospha- tases, saturases, acylating enzymes and kinases in order to obtain the final products [2,3] . Consequently, the production of polyisoprenoid pyrophosphate is the main event of the synthesis and of considerable interest in regulatory studies.

The enzyme catalyzing the addition of isopentenyl pyrophosphate (IPP) units using farnesyl pyrophosphate as primer has been isolated and characterized in bacterial sys- tems, but the enzyme characteristics are much less well known in animal tissues [4, 51. Using farnesyl pyrophosphate and isopentenyl pyrophosphate as substrates, synthesis of wun- saturated pyrophosphates and monophosphates could be obtained in hen oviduct, rat liver, Ehrlich ascites tumor cells and spermatogenic cells [6 - 1 I]. The activity was found to be either membrane bound or present in the cytosol. Obviously, the conditions used for measurement of the activity are of great importance, since the nature of the product obtained varies with the conditions used. In the case of the rat liver enzyme, very long incubation times were needed to obtain measurable activities [9 - 111. Solubilization and subsequent purification proved to be very difficult compared to the bac- terial enzyme, in spite of the fact that, in the initial phase of the isolation procedure, the maminalian enzyme appears to

C'orwspondunce t o J . Ericssoii, Department of Biochemistry, Uni- versity of Stockholm, Arrhenius Laboratories, S-106 91 Stockholm, Sweden

Ahhruviurions. IPP, isopentenyl pyrophosphatc; FPP, farnesyl pyrophosphatc.

~-

be resistant to relatively high detergent concentrations [7, 81. At present these difficulties prevent the isolation of the purified enzyme from animal cells for studying the character- istics of polyisoprenoid biosynthesis. In the present study, we have analyzed the properties of this enzyme in rat liver frac- tions in order to obtain an optimal system for kinetic studies using relatively short incubation times. The factors influencing the activity were found to be of importance in establishing the synthetic mechanism and also, for future purification studies.

MATERIALS AND METHODS

Chemicals

a-Unsaturated polyprenols were prepared from Sorhus suecicu as described previously [I 21. Polyprenyl phosphate and polyprenyl pyrophosphate were prepared according to Popjak et al. [13]. Labeled isopentenol was prepared from commercial isopentenol (Janssen Chimica, Belgium) by oxidation to iso- pentenal using ammonium trioxide/pyridine, followed by re- duction with [3H]sodium borohydride (Amersham, England, 16.5 Ci/mmol) according to Keenan and Kruczek [14]. The 3H-labeled isopentenol was subsequently phosphorylated with di(triethy1amine) phosphate in the presence of trichloro- acetonitrile in order to obtain isopentenyl pyrophosphate (15 Ci/mmol), according to Popjak et al. [13]. Unlabeled farnesyl pyrophosphate was synthesized by the same method from all trans-farnesol. trans-Farnesol was a gift from Dr Mizuno of the Kururay Co., Japan. Upon TLC in different solvent systems, all polyprenols and phosphorylated deriva- tives gave single spots.

790

a b

T

Y 0 ' I I I I I I I

0 10 20 30 40 50 60 70 Incubation time (min)

60

t 0 50 100 150 200 250

pM FPP

C d 150 7 100 7 1

0 10 20 30 0 1 2 3 4

mg protein FM IPP

Fig. 1. Properties of the microsomal ek-prenyl transferase. Activity as a function of (a) time course, (b) substrate concentration (FPP), (c) substrate concentration (IPP) and (d) protein concentration, in the presence of 0, 100 and 200 mM KF. Mean values +_ SEM (n = 7).

Isolation of liver subfractions at various concentrations. The solubilization medium also

Male Sprague-Dawley rats (1 80 - 200 g) were used with- out starvation. Isolation of microsomes and the particle-free supernatant fraction was performed as described earlier [ 151. The microsomal pellet was suspended in 0.15 M Tris/HCl, pH 8.0, and recentrifuged in order to remove adsorbed pro- teins [16]. Subfractionation of the microsomal fraction into rough, smooth I and smooth I1 microsomes was performed as described earlier [17]. The various subfractions were charac- terized using appropriate marker enzymes. Contamination of the total microsomal fraction by mitochondria, lvsosomes,

routinely contained 1 mM dithiothreitol and 0.4 M KC1. Without addition of monovalent ions, no solubilization could be achieved with any of the detergents tested. The solubiliza- tion mixture was incubated for 20 min at 4°C and was then centrifuged at 105000 x g for 1 h in order to remove remaining nonsolubilized proteins. The supernatant obtained from this centrifugation was used as an enzyme source. With detergent concentrations which solubilized all of the cis-prenyl transfer- ase activity, about half of the total microsomal protein was solubilized.

Golgi and plasma membranes was calculated to be i , 3 , 5 and 4%, respectively, of the total microsomal protein. lncubations

cis-Prenyl transferase activity was measured in a 30O-pl

Solubilization of the enzyme incubation mixture and the optimal system contained 25 mM imidazole/HCl, pH 7.0, 1 mM MgC12, 100 mM KF, 0.1 mM

In order to solubilize cis-prenyl transferase, the micro- somes were treated with the detergents P-octyl glycoside, taurodeoxycholate, Chaps, Triton X-I 14 and Triton X-100,

dithiothreitol, 1:0% Triton X-100, 50 pM FPP and 5.5 pM ['HI IPP (2.3 Ci/mmol). FPP was added in a 0.1% Triton X-100 solution and [3H]IPP in ethanol/0.15 M NH40H

791

(1 : 1 :0.5 pl per assay). When incubations were performed in the absence of detergents, FPP was also added in the latter solvent. The reaction was started by the addition of 1 mg microsomal protein and continued for 60 min at 37°C. The reaction was stopped by the addition of 300 pl 0.87 M KOH and hydrolysis was performed at I00 "C for 30 min. Finally, the samples were supplemented with 1.5 ml 2 M KCI and 300 ~1 1 M HC1.

Extraction and chromatography of lipids

The reaction mixture was extracted three times with 3 ml diethyl ether/petroleum ether (1 : 1). The pooled organic phases were then washed three times with 2 ml 0.9% NaCl. The organic phase was transferred to a scintillation vial for determination of radioactivity. The organic solvent was evap- orated and the residue dissolved in 10ml OptiScint (LKB Scintillation Products, Sweden) and the radioactivity deter- mined by scintillation counting. When the reaction products were to be analyzed by TLC, the washed organic phase was dried under N2 and the residue dissolved in chloroform/meth- anol (2:l). The lipids were applied to a silica-gel-60 plastic sheet (Merck). Development was performed using diisobutyl ketoneiacetic acidlwater (8 : 5 : 1). For autoradiography, the plate was sprayed with En3Hance spray (NEN Research Prod- ucts, Boston) and exposed to X-Omat AR film (Eastman Kodak, Rochester) at -80°C for 48 h. The radioactive spots comigrated with C g 5 polyprenyl pyrophosphate (Rf 0.4) and polyprenyl phosphate (Rf 0.6) added as internal standards and visualized with iodine vapor.

When the distribution of the chain lengths of the enzyme products was analyzed, the washed extracts were hydrolyzed enzymatically by the method of Fujii et al. [18] and the hexane- extractable products were analyzed by reverse-phase HPLC using a Hewlett-Packard Hypersil ODS 3-pm column. A con- vex gradient was employed from the initial 2-propanol/ methdnol/H20 (40:60: 5 ) in pump system A to 40% hexane/ 2-propanoI(70: 30) in pump system B, at a flow rate of 1.5 ml/ min with a program time of 30 min. The eluate was monitored at 210 nm and the radioactivity was monitored using a radio- active-flow detector (Radiomatic Instruments, Tampa, FI.) with a cell volume of 0.5 ml. To determine whether the polyisoprenoids synthesized in vitro were a-unsaturated or a- saturated, the individual radioactive peaks from the reverse- phase HPLC run were collected and rerun in a straight-phase HPLC system capable of separating polyprenols from dolichols [2].

Chemical measurement

Protein concentration was determined with the biuret method using bovine serum albumin as standard [19].

RESULTS

cis-Prenyl transferase activity in rat liver has, to date, been studied only to a limited extent, since the sensitivity of the assay methods employed is very low [7 - lo]. In order to increase this sensitivity, we have synthesized [3H]IPP with a high specific radioactivity. In addition to employing optimal incubation conditions, it was necessary to optimize extraction conditions after incubation. Complete extraction could be obtained only after alkaline hydrolysis prior to extraction. To achieve complete extraction, the hydrolysate was adjusted

lZO 1

[I 8

t 0 ' I I I I I

0 0.5 1 1.5 2 2.5

46 Triton X-100

Fig. 2. Effect of Triton X-100. Total microsomes were incubated in the presence of various detergent concentrations. Mean values & SEM (n = 6).

to an acid pH. At low pH, the phosphate groups of the polyisoprenoid are protonated, which prevents the loss of negatively charged reaction products to the water phase. The method employed for measuring enzyme activity could thus be used to study the enzyme properties in liver subfractions using relatively short incubation times.

Enzyme properties

The incubation procedure employed resulted in a linear increase in enzyme activity during the first 60 min (Fig. 1 a). In the presence of excess IPP, saturation of the enzyme was attained at a concentration of 70 pM FPP (Fig. 1 b). Measure- ments in the presence of excess FPP led to saturation at 5 pM IPP (Fig. lc). With up to 0.5 mg protein in the incubation medium, the activity increases linearly, while at higher protein concentrations the activity decreases, obviously because of the action of phosphatases hydrolyzing the substrate (Fig. 1 d). When KF, a phosphatase inhibitor, was included in the incu- bation mixture, an increase in activity was obtained. To obtain maximal levels, it was necessary to supplement the incubation medium with 100mM KF/mg protein. In this way, it was possible to increase the protein concentration and simulta- neously inhibit the increased phosphatase activity. Our incu- bation mixture routinely contained 100 mM KF.

Detergents

Triton X-100 increased the transferase activity 10-fold at a concentration of 0.2% and the activity remained at this level up to a detergent concentration of 2% (Fig. 2). Triton X-100 differs greatly from other detergents as an activator of thc enzyme when added to the incubation mixture (Table 1). The activation is considerably less with Triton X-I 14, even lower with Chaps and almost completely absent with taurodeoxy- cholate and /I-octylglucoside. However, when the solubilized fractions were supplemented with Triton X-100, all gave simi- larly high activity. Thus, all detergents are able to solubilize the enzyme, but only Triton X-100 is also able to activate the enzyme maximally.

792

Table 1. Effect of various detergents on the activity and solubilization of cis-prenyl transferase. Activity was measured by incubation of total microsomes in the presence of the various detergents. Microsomes were also treated with the various detergents and, after centrifugation, the solubilized material was assayed in the presence of 1% Triton X- 100. The values were means f SEM (n = 7).

~~

Detergent Total Solubilized activity activity

dpm . mg protein-'

None 6.13 f 0.5 0 Triton X-200, 1% 56.3 7.0 96.1 f 12 Triton X-114, 1% 32.4 f 4.3 104 + I 4 Taurodeoxycholate, 2% 8.81 f 0.6 84.4f 8.3 Chaps, 5% 18.4 f 1.9 96 f 7.6 ,B-Octyl glycoside, 2% 5.40 f 0.8 88.5f 8.1

Table 2. Effects of divalent cations on cis-prenyl transferase activity. The concentration of the various divalent cations was 1 mM. The values are means + SEM (n = 6).

Ion Activity

10-3 dpm I mg protein None 10.2 f 1.5 Mgz+ 55.3 f 6.9 Mn2+ 8.53 f 0.7 Zn2 + 8.34 0.8 co2+ 40.8 f 5.7 Ca2 + 19.1 f 2.3

120 1

80

8 40

/ + 1 mM EDTA

0 ' I I I

0 4 8 12

mM Mg2'

Fig. 3. Effect of MgCI2. Incubations were performed in the presence of various Mgz+ concentrations with or without 1 mM EDTA. Mean values & SEM (17 = 8).

Divalent cations

Divalent cations have great influence on the transferase activity (Table 2) . Mg2+ increased activity fivefold, while in the presence of Co2+ and Ca2+ the activity was elevated fourfold and twofold, respectively. Mn2+ and Zn" were slightly inhibitory. The concentration curve for Mg2+ is illus- trated in Fig. 3. Maximal activation is obtained at a concen-

Table 3. Effects of buffer and pH on cis-prenyl transferase activity. In experiment 1 the concentrations of the buffers were 50 mM. In experiment 2 the concentration of the imidazole buffer was 50 mM. The values are means +_ SEM (n = 7).

Experi- Buffer pI-1 Activity ment

1 Imidazole/HCl 7.0 Tris/HCl Mops Hepes

2 Imidazole/HCl 6.0 6.5 7.0 7.5 8.0 8.5

dpm . mg protein

70.5 f 8.5 51.3 f 4.9 52.3 f 4.6 27.8 f 2.0

30.7 f 2.9 45.2 f 4.8 71.5 f 5.9 43.4 * 3.5 32.5 f 3.6 21.7 f 2.8

T 120 7 NaCl

8 0 - - g u-

8

0 ' I I I

0 60

Fig. 4. Effects of inorganic monophosphates and pyrophosphate. The incubation mixtures contained iiicreasing concentrations of the vari- ous salts. Mean values * SEM (n = 8).

tration of 1 mM and at 10 mM there was a slight inhibition. Specificity of the Mg2' activation is indicated by the fact that the addition of 1 mM EDTA eliminated this activation at low concentrations.

Various conditions The type of buffer used during the incubation influences

enzyme activity (Table 3). Higher activity was obtained with imidazole/HCl than with Tris/HCl, Mops or Hepes. One ex- planation may be an effect on antagonistic enzymes, since imidazole is an inhibitor of several phosphatases. With imidazole/HCl, highest activity was obtained at pH 7.0 and this activity declined sharply at both acid and alkaline pH. A part of the radioactivity in the organic phase at pH 8.5 was not polyprenyl pyrophosphate or polyprenyl monophosphate, but isopentenol, as shown by chromatographic investigation (not shown in Table 3). Isopentenol was probably produced by hydrolysis of IPP by phosphatases.

Since inorganic monophosphates and pyrophosphates are produced during the condensation reactions as well as by phosphatases, their influence on the transferase activity was tested (Fig. 4). pyrophosphate at 50 mM concentration was

79 3

Table 4. Distribution of c&prenyl transferase activity in microsomal membranes. Subfractionation of the microsomal fraction was performed as described in Materials and Methods. Total microsomes were treated with 50 pg trypsin/mg protein and, after centrifugation, the pellet was washed and used for enzyme assay. The values are means f SEM (N = 6) .

Experi- Fraction ment

~-~

Activity

1 Total microsomes Rough microsomes Smooth I microsomes Smooth I1 microsomes Supernatant

2 Total microsomes Total microsomes, treated with trypsin

- - ~

dpm . mg protein

70.5 F 8.8 81.3 f 7.5 38.9 f 4.9 12.4k1.5 6.8 k 0.7

12.3 +_ 6.3

15.0 & 1.8

R SI T

Fig. 5. Autoradiography of the incubation products. R = rough, S, smooth I , SII smooth I1 and T total microsomes. The lipids were separated by TLC on silica. The plates were developed with diisobutyl kctone/acetic acid/water (8: 5 : 1). The radioactive products were ident- ified (using standards) as (A) polyprenyl phosphatc (R, 0.6) and (B) polyprenylpyrophosphdte (Rf 0.4). 0 = origin, F = front.

completely inhibitory, while monophosphate gave a 30% inhi- bition of the activity. These inhibitory effects were not the result of the increasing ionic strength, since monovalent cat- ions had no effect.

Other products of the transferase reaction, e.g. polyprenyl pyrophosphates and polyprenylmonophosphates, were also tested at increasing concentrations. No inhibition was .found up to a concentration of 100 pM.

Characterization of the product

Autoradiographic analysis of the products obtained upon incubation of microsomal subfractions and total microsomes is shown in Fig. 5. It is apparent that, in all three subfractions, the main product is polyprenyl pyrophosphate and, to some extent, polyprenyl monophosphate is also produced in amounts proportional to the level of the major product. Autoradiographic studies were performed in a number of experiments involving addition of various substances, differ- ent pH and detergents, and the ratio between the levels of

The distribution pattern of the polyisoprenoids after incu- bation is shown by C I 8 reverse-phase HPLC (Fig. 6). The pattern appearing after in vitro incubations in the presence of detergents consists of shorter polyisoprenoids than those found in tissues. The main product is that with 16 isoprene units, but substantial amounts of C75 and Cg5 polyisoprenoids are also recovered (Fig. 6a). In the absence of detergents, the main product of Cg0 polyisoprene and the additional major products are those with 17 and 19 isoprene residues (Fig. 6b). Thus, incubation in the absence of detergents gives products more similar to those in vivo, but the yields are strongly re- duced.

The individual polyisoprenes, isolated by reverse-phase HPLC, were also rechromatographed on silica HPLC in order to separate a-unsaturated and saturated species (not shown in Fig. 6). In all cases, only a-unsaturated lipids could be found, in agreement with previous findings that a-saturation requires the presence of NADH and a cytosolic factor [2]. Upon straight phase TLC in benzene/ethyl acetate (8: 3 ) , the individ- ual polyprenols comigrated with irans,trans-poly-cis-prenol standards (not shown in Fig. 6). This indicates that the microsomal enzyme adds the IPP units in the cisconfiguration.

Intracellular distribution Upon subcellular fractionation, the highest activity was

recovered in rough microsomes, exceeding that of the smooth I counterpart (Table 4). ‘The enzyme activity was much lower in smooth I1 microsomes. Some activity was also recovered in the membrane-free supernatant fraction. An increase in the specific activity was obtained when the microsomes were washed with 0.3 5 M TrisiHCI, pH 8.0 (not shown in Table 4). This may be explained by the removal of adsorbed cytosolic proteins.

The localization of the enzyme in the transverse plane of the membrane was studied using proteolytic treatment. Upon trypsin treatment of intact microsomal vesicles, 80% of the transferase activity was lost, indicating the presence of this enzyme on the outer, cytoplasmic surface of the microsomal membranes.

polyprenyl pyrophosphates and polyprenyl monophosphates were similar to that shown in Fig. 5. In contrast, in the absence of detergents an additional product, e.g. nonphosphorylated polyprenol, also appeared (Rf 0.9; not shown).

DlSCUSSlON The microsomal cis-prenyl transferase mediates a sequence

of condensation reactions in which an isopentenyl unit from

794

T A 15

t

10 20

Time (min)

30

17 18

10 20 30

Time (min)

Fig. 6. Distribution of polyisoprenes produced by cis-prenyl transferase. The radioactive products were enzymatically dephosphorylated and run on reverse-phase HPLC. (A) Incubation in the presence of 1% Triton X-100; (B) incubation in the absence of Triton X-100. The numbers denote the number of isoprene residues. The radioactivity was monitored using a radioactivity flow detector (Radiomatic Inc.) attached to the HPLC.

IPP is initially added to trans,truns-FPP followed by consecu- tive additions of additional isopentenyl units to obtain a trans- trans-poly-cis-isoprenoid chain. It is thought that one enzyme mediates the complete sequence of reactions, but the final product must be subjected to various types of modifications, e.g. dephosphorylation(s) and a-saturation. The two main polyisoprenoid products are dolichol and ubiquinone. This requires the presence of at least two condensation enzymes, since solanesyl pyrophosphate, utilized for ubiquinone bio- synthesis, is all trans. Interestingly, in our incubation system, only the cis-polyisoprenol series, with 13 - 17 isoprene units, are formed. Considering the relatively high amount and turn- over of cellular ubiquinone, one would expect to obtain a sizeable amount of solanesyl pyrophosphate.

It appears that the conditions we have used are optimal for cis-prenyl transferase activity, but are unfavourable for solanesyl pyrophosphate synthesis. Alternatively, some essen- tial factor is absent in our system, such as a specific substrate or an activating Factor. It is not clear why, in the system used, a specific series of polyprenols are synthesized which are shorter than those found in vivo. Previously, it was proposed that the terminating step is the addition of isopentenol instead of isopentenyl pyrophosphate and that the lack of pyrophos-

phate as a leaving group is the reason for the termination of the condensation reaction. Obviously the nature of the product depends on the incubation conditions such as substrate con- centration, Mg2 + concentration, type of detergent and mem- brane phospholipids [2, 8 - 111.

In previous experiments by others, conditions were used which gave low activities in spite of very long incubation times, and the kinetics proved to be both complex and difficult to interpret. In our system, short incubation times result in relatively high yields of product with high reproducibility. In addition, in order to attain optimal incubation conditions, we have synthesized the tritiated substrate with a high specific radioactivity, which gives much higher radioactivity in the product. Probably, it would also be advantageous to inhibit pyrophosphatases, which may be achieved by bacitracin [20]. Since this inhibitor binds to the substrates, it could not be used in our system.

An absolute requirement for the reaction is the presence of Triton X-100. Interestingly, different types of detergents were equally efficient in solubilization of the enzyme without inactivation. Triton X-100 has to be present during the incu- bation in order to reach full activity. The form of the enzyme

795

in Triton micelles probably resembles that in microsomal membranes, thereby facilitating the condensation reaction. The cation requirement is both great and specific, since it is restricted to Mg2+. In bacterial systems, the necessity for Mg2+ has also been documented, and it is believed that the Mg2+-substrate complex is utilized in the condensation reac- tion [21]. For optimal activity, imidazole buffer is required and the pH optimum is unusually sharp at pH 7.0.

One regulatory mechanism in most enzyme reactions is product inhibition, which is accentuated in our case. Inorganic phosphate, both monophosphate and particularly pyrophos- phate, are strong inhibitors. Inorganic pyrophosphate is con- tinuously produced during the condensation reactions and, obviously, mechanisms which counteract the accumulation of this inhibitory substance must exist. Neither polyprenols nor their phosphorylated derivatives exhibited any influence on the reaction.

cis-Prenyl transferase synthesizes several products with various chain lengths, both in vivo and in vitro, and the reason for this is not yet fully understood. In the case of the side chain of ubiquinone, only one main polyisoprenoid product is synthesized, with nine isoprene units in rat and ten isoprene units in human. In the case of dolichol, the pattern is complex, consisting of various quantities of between five and seven types of polyisoprenes, depending on the tissue and the species investigated [22]. The reason for this pattern has not yet been explained; in the case of dolichyl phosphate we know that no sugar specificity is associated with any individual dolichyl phosphate species and that glycoprotein synthesis is not influ- enced by the polyisoprenoid pattern [23]. However, membrane destabilization in a model system increases linearly with the incorporation of polyisoprenes with increasing chain length [24]. In the in vitro system used here, biosynthesis yields a family of polyisoprenes, but the individual species are shorter than in vivo. In the absence of detergents, longer isoprenoids are synthesized, but the biosynthetic activity is strongly re- duced. It would be desirable to devise an in vitro system in which long isoprenoids are synthesized with high efficiency. This aim may be attained in the future by identifying all the factors involved in this complex process. One of these factors is the substrate concentration, since it was found previously that both increasing amounts of mevalonate and IPP in the incubation medium resulted in the appearance of longer polyisoprenoids [2, 9 - 1 I].

The intracellular localization of an enzyme system is of importance when one is studying properties and interactions between enzyme systems. cis-Prenyl transferase is localized to the endoplasmatic reticulum and is enriched in the rough membranes. In the smooth I1 microsomal fraction, which is an as-yet unidentified segment of the Golgi system, no activity is present. Cholesterol biosynthesis has a similar distribution to that of cis-prenyl transferase, in contrast to ubiquinone, which is produced in the Golgi vesicles [25], at least the ter- minal reactions. It will be of great importance in the future to localize the intracellular site of solanesyl pyrophosphate synthesis.

Since these lipids are present in all intracellular mem- branes, transport from the site of synthesis to the final localiza- tion must take place. In this respect, the cellular organization plays an important role, since most lipids are subjected to modifications during translocation. This is also the case for dolichol, which is esterified in the Golgi system, while dolichyl ester hydrolysis occurs in lysosomes [26, 271.

It is also known that dolichol is synthesized in peroxisomes, but under normal conditions this synthesis rep-

resents only a small portion of the total and the enzymatic mechanism of the peroxisomal process has yet not been stud- ied in detail [28]. The low activity of dolichol synthesis in the cytosolic fraction most probably originates from peroxisomes, since a sizeable portion of this organelle is ruptured during homogenization. It is interesting to note that, in contrast to microsomes, these enzymes in peroxisomes are localized in the lumen in a soluble form [28]. It will be a future task to compare the mechanisms of condensation in these organelles, which are apparently organized in different manners. The incubation procedure presented in this study will probably also be suf- ficiently sensitive for peroxisomes, which is a minor organelle in the liver.

The properties of the cis-prenyl transferase found here, using appropriate incubation conditions, are similar to those described in bacterial systems [4, 51. This procedure and its future developments can be applied for further studies of the kinetic properties of the enzyme, of its subcellular and intramembraneous distributions, of the peroxisornal function and of the mode of regulation. Finally, the enzyme should be purified both from microsomes and peroxisomes for a detailed characterization.

This work was supported by the Swedish Cancer Society and the Swedish Medical Research Council. During part of this study T. Chojnacki was a visiting professor in Stockholm, supported by the Swedish Cancer Society. We thank P.-A. Ericsson for photography and technical assistance.

REFERENCES 1. Hemming, F. W. (1983) in Biosynthesis of isoprenoid compounds

(Porter, J . W. & Spurgeon, S. L., eds) vol. 2, pp. 305-354, John Wiley & Sons Inc., New York.

2. Ekstrom, T., Chojnacki, T. & Daliner, G. (1987) J . Biol. Chem.

3. Chojnacki, T. & Dallner, G. (1988) Biochem. J . 251, 1-9. 4. Baba, T. & Allen, C. M. (1980) Arch. Biochem. Biophys. 200,

5. Takahashi, I. & Ogura, K . (1982) J. Biuchem. (Tokyo) 92, 773-

6. Adair, W. L. & Cafmeyer, N. (1983) Biochim. Biophys. Acta 208,

7 . Adair, W. L., Cafmeyer, N. & Keller, R. K. (1984) J . Bid. Chem.

8. Chen, Z., Morris, C. & Allen, C. M. (1988) Arch. Biochim. Bio-

9. Sagami, H., Lennarz, W. J. & Ogura, K. (1989) Biochim. Biuphys.

10. Sagami, H., Matsuoka, S. & Ogura, K. (1991) J . Bid. Chem. 266,

11. Matsuoka, S., Sagami, H., Kurisaki, A. & Ogura, K. (1991) J .

12. Chojnacki, T. & Vogtman. T. (1984) Acta Biochim. Pol. 31,115-

13. Popjak, G., Cornforth, J. W., Cornforth, R. H., Ryhage, R. &

14. Keenan, R. W. & Kruczek, M. E. (1976) Biochemistry 15,1586-

15. Thelin, A,, Low, P., Chojnacki, T. & Dallner, G. (1991) Eur. J .

16. Glauman, kf. & Dallner. G. (1968) J . LipidRes. 9, 720-729. 17. Dallner. G. (1974) Methods Enzvmol. 31. 191 -201.

262,4090-4097.

414- 484.

777.

532- 534.

259,4441 -4446.

phys. 266,98 - 110.

Acta 1002,218-224.

3458 - 3463.

Biol. Chem. 266, 3464 - 3468.

126.

Goodman, D. S. (1962) J . Biol. Chem. 237,56-61.

1591.

Biochem. 195, 155 - 161.

~I

18. Fujii, H., Kogama, T. & Ogura,K. (1982) Biochim. Biophys. Acta 712.116-722.

19. Gornall, A. G., Bardawill. C. J. & David, M. M. (1949) J . Biol.

20. Wedgewood, J . F. & Strominger, J. L. (1980) J . Biol. Chem. 255, Chem. 177,751 -766.

1120- 11 23.

796

21. Gotoh, T., Koyama, T. & Ogura, K. (1988) Biochem. Biophys. Res. Commun. 156,396-402.

22. Tavares, I. A., Coolbear, T. & Hemming, F. W. (1981) Arch. Biochem. Biophys. 207,427 -436.

23. Low, P., Peterson, E., Mizuno, M., Takigawa, T., Chojnacki, T. & Dailner, G. (1986) Biosci. Rep. 6, 677-683.

24. van Duyn, G., Valtersson, C., Chojnacki, T., Verkleij, A. J., Dallner, G. & de Kruijff, B. (1986) Biochim. Biophys. Acta 861, 121 3 - 1223.

25. Kalen, A., Appelkvist, E.-L., Chojnacki, T. 8z Dallner, G. (2990)

26. Tollbom, O., Valtersson, C., Chojnacki, T. & Dallner, G. (1988)

27. Tollborn, O., Chojnacki, T. & Dallner, G. (1989) J . B id . Chem.

28. Appeikvist, E.-L. & Kalen, A. (1989) Eur. J . Biochem. 185,503-

J . Biol. Chem. 265, 1158- 1164.

J . Biol. Chem. 263,1347 - 1352.

264,9836-9841.

509.