Main

In some peroxisomal inherited diseases, such as ZWS and ALD, the peroxisomes are absent or reduced in number(1). Because these organelles contain several important enzymic systems participating in lipid synthesis and breakdown, the peroxisome-deficient conditions lead to serious functional disturbances.

Peroxisomes participate in the synthesis of two lipids deriving from the mevalonate pathway, i.e. cholesterol and dolichol(2–4). Several enzymes participating in the common initial part, between acetyl-CoA and FPP(5,6), and in the terminal pathways of both cholesterol and dolichol synthesis are described as peroxisomal components(7,8). Because both of these lipids are also synthesized in the ER, the question arises to what extent these two lipids are produced in the two individual compartments.

The enzymes participating in the mevalonate pathway are not yet isolated from the two cellular locations, and, consequently, it is not known if they are isoenzymes with different properties. Dietary and drug treatments have established that independent up- and down-regulation of the same enzyme occurs at the two locations(9), thereby showing the individual regulation of the synthetic pathways in peroxisomes and microsomes.

The cellular content of peroxisomes or microperoxisomes is low in comparison with the extensive ER system. Because the specific activities of branch-point and probably also rate-limiting enzymes are similar in these organelles(10), it seems that the major part of the lipids are synthesized in the ER system. It is plausible that the lipids produced in the ER and peroxisomes serve different purposes, i.e. cholesterol is the lipoprotein precursor in the ER and a substrate for bile acids in peroxisomes.

The fibroblasts from patients who have peroxisomal disorders have previously been studied to some extent. By use of radioactive acetate, decreased cholesterol synthesis was found(11), but, in a separate study, the synthetic rate was unchanged or increased(12). Some activities of the mevalonate pathway enzymes, among them mevalonate kinase, seemed to be unchanged in patients with ZWS(13), whereas the kinase was found to be decreased in another study(14). In peroxisome-defective fibroblasts, in the presence of exogenous LDL, the activity of acyl-CoA:cholesterol acyltransferase was significantly decreased(15).

In the present investigation, the extent of cholesterol and dolichol formation was studied in fibroblasts of patients with peroxisomal deficiency by using mevalonate as precursor and determining the activity of some branch-point enzymes. The diseased fibroblasts were found to be appropriate to analyze the quantitative contribution of peroxisomes to the total cellular synthesis of the mevalonate pathway lipids.

METHODS

Chemicals. (R, S)-[5-3H]-mevalonolactone (specific activity, 11.6 Ci/mmol) was prepared as described by Keller(16). Mevinolin was a generous gift from Dr. A. W. Alberts (Merck). Before use, the mevalonolactone and mevinolin were converted from the metabolically inactive lactone form to the active dihydroxy acid form. This was achieved by evaporation of the solvent in which the compound was stored under a nitrogen stream, addition of 0.05 M NaOH to increase the pH to >11, and incubation at 30°C for 40 min, followed by evaporation again to dryness. Mevalonic acid was then dissolved in 95% ethanol (25 mCi/mL) and mevinolin in 10 mM Tris-HCl, pH 8.0 (10 mM, final concentration). The labeling of isopentanol and farnesol with tritium was performed using [3H]-sodium borohydride (Amersham Corp., 16.5 Ci/mmol) according to Keenan and Kruczek(17). Labeled isopentenyl-PP and FPP were prepared from the corresponding labeled alcohols according to Popják et al.(18). Upon thin-layer chromatographic analysis in different solvent systems, the radioactive products ran as single spots when detected. All cell culture media and antibiotics were purchased from GIBCO BRL (Life Technologies, Paisley, Scotland, UK) and the solvents used in this study from Merck (Darmstadt, Germany) and were of HPLC grade. The other chemicals were of pro analysis grade (Merck).

Patients. Fibroblast cultures were obtained from two patients with ZWS. These patients were both diagnosed as neonates based on a typical clinical presentation, highly elevated levels of very long-chain fatty acids, almost total lack of plasmalogen synthesis, and accumulation of trihydrocoprostanoic acid in serum. Both patients died within 3 mo. Five patients with X-linked ALD were boys with typical clinical presentation and elevated levels of very long-chain fatty acids in serum. Patient 8 was heterozygous for X-linked ALD. Patients 3, 6, and 7 have been treated with bone marrow transplantation, and clinical data for these have been published elsewhere(19). Control fibroblast lines were obtained from healthy subjects of the same age. The study was approved by the Ethics Committee of the Karolinska Institutet.

Cell culture. Fibroblasts were cultured from skin biopsies using Eagle's minimum essential medium (MEM) containing FCS (10%, vol/vol), penicillin (125U/mL), streptomycin (125 µg/mL), and Tylosine (6 mg/mL). Stock cultures of individual fibroblast stains were cultured in 75-cm2 Costar plastic tissue culture flasks in a humidified atmosphere of 5% CO2 in air at 37°C. Fibroblast lines were used between the 6th and 14th passage. The cells were frequently checked for mycoplasma and bacterial contamination.

[3H]-Mevalonate labeling. Confluent cells were harvested, resuspended in complete MEM medium, and plated into 75-cm2 Costar plastic tissue culture flasks. After 48 h in culture, the cells were incubated in 10 mL of MEM containing 1% (vol/vol) Ultroser G (an FCS substitute), penicillin (125 U/mL), streptomycin (125 µg/mL), and Tylosine (6 mg/mL), and without addition of FCS. After the cells were grown to confluence (for 2-3 d), sodium salt of mevinolin was dissolved in 10 mM Tris-HCl, pH 8.0, and added to the cells at the final concentration of 10 µM. The cells cultured in the absence of mevinolin were supplied with an equivalent volume of the buffer. Twenty-four hours later, the medium was replaced by 5 mL of fresh medium containing 0.5 mCi (R,S)-[5-3H]mevalonic acid (11.6 Ci/mmol), and the cells were incubated for a further 2 h. After removal from the flasks by trypsinization, the cells were sedimented by centrifugation and washed three times with Dulbecco's PBS, pH 7.4.

Lipid extraction and HPLC-radioflow analysis. The fibroblast pellets from two identically treated cell culture flasks were suspended in 0.4 mL of distilled water, followed by the addition of 3.6 mL of methanol and 2.4 mL of petroleum ether. As internal standards, 2.5 µmol of ergosterol, 1 nmol of ubiquinone-6, and 0.25 nmol of dolichol-23 were added to quantitate the recoveries after HPLC-radioflow analysis. The samples were vortexed vigorously. The upper solvent phase was then transferred into another tube and evaporated under nitrogen. The lipids were redissolved in 50 µL of chloroform:methanol (2:1, vol/vol) and injected for HPLC analysis by use of a reversed phase C18 column (Hewlett-Packard Hypersil ODS, 3 µm)(20). The lipids were separated by using a binary convex gradient at a flow rate of 1.5 mL/min from an initial 90% methanol:water (9:1, vol/vol) in pump A to 100% methanol:2-propanol:hexane (2:1:1, vol/vol) in pump B, and monitored by a UV detector at 210 nm. The radioactivity in cholesterol, ubiquinone, and dolichol was determined by a radioflow detector (Radiomatic Instruments, Tampa, FL).

Enzyme assays. The fibroblasts from controls and patients were harvested by trypsinization, homogenized by short sonication (20 s), and centrifuged at 100 000 × g for 60 min to separate the membranes from the supernatant. The enzyme activities were assayed in 300 µL of incubation medium containing 25 mM imidazole chloride, pH 7.0, 1 mM MgCl2, 0.1 mM DTT, and 5 mM KF(21,22). When squalene synthase activity was analyzed, 100 µM [3H]-FPP (44 Ci/mol) was added as a substrate. In assays to determine FPP synthase activity, the substrates were 20 µM [3H]-isopentenyl-PP and 50 µM geranyl-PP. When analyzing the activities in membrane fractions, the reactions were started by the addition of 300 µg of protein and allowed to proceed for 60 min at 37°C. In supernatant fractions, FPP synthase activity was assayed in the presence of 75 µg of protein for 15 min. All reactions were terminated by the addition of 2.0 mL of water-saturated n-butanol. The remaining butanol phase was evaporated under nitrogen stream and the residue was redissolved in 200 µL of 3% n-octylpyranoside and subsequently dephosphorylated according to Wong et al.(23).

Product analysis by HPLC. After dephosphorylation, the products were extracted with ethyl ether:petroleum ether (1:1, vol/vol) and this extract was then dried. The residue was redissolved in ethanol and analyzed by reversed-phase HPLC by use of a Hewlett-Packard Hypersil ODS C18 column at a flow rate of 1.5 mL/min. The products from the FPP synthase assay were analyzed by using a linear gradient from methanol:water (7:3, vol/vol) in pump system A to methanol:water (9:1, vol/vol) in pump system B for 20 min. To estimate the squalene synthase activity, the products were separated by using a gradient from the initial methanol:water (9:1, vol/vol) in pump system A to methanol:2-propanol:hexane (2:1:1, vol/vol) in pump system B, with a program time of 30 min. The absorbance at 210 nm and radioactivity of the effluent were monitored using a UV detector and a radioactivity flow detector, respectively. The individual radioactive products were identified using unlabeled internal standards added to each sample.

Protein assay. Protein levels were determined by the Lowry procedure with BSA as the standard(24).

RESULTS

The fibroblasts used in the experiments were taken as frozen cells and cultivated for 6-14 cycles before they were used. Three types of cells were used: fibroblasts derived from patients with 1) fully developed ZWS (ZWS group), 2) X-linked ALD characterized by a specific deficiency of fatty acid β-oxidation (ALD group), and 3) from nondiseased subjects (control group). Because we had only two cases of ZWS with complete peroxisomal deficiency, the cells from these patients were cultured at four separate occasions and analyzed in separate experiments. Because the cells from ALD patients have well-defined deficiencies only in the β-oxidation, we consider this group as equivalent with the control group. The cells were grown to a point just before confluence, and all experiments were adjusted to this same time point of growth. Measurements of lipid contents and radioactive labeling were performed on whole cells, and the cell suspensions were subjected to a short sonication to eliminate the permeability barrier when being used for enzyme analyses.

The cells obtained after trypsinization and centrifugation contained approximately 1 mg of protein in each sample. The cholesterol content of the fibroblasts varied between 26 and 39 nmol/mg protein. The mean value of the cholesterol level was 12% higher in the control group in comparison with the ALD group and patients having ZWS, but this value was not statistically significant.

The incorporation of [3H]-mevalonate into cholesterol, ubiquinone, and dolichol was determined by analyzing lipid extracts with a radioflow detector attached to the HPLC equipment. The fibroblasts exhibited extensive labelling of cholesterol after incubation with [3H]-mevalonate, whereas incorporation into ubiquinone was much lower and into dolichol low or undetectable (Fig. 1A). This is similar to the situation found in most other cells and tissues. After treatment of the cells with the HMG-CoA reductase inhibitor mevinolin, the radioactive labeling of the individual lipids was greatly altered; the corresponding chromatogram is shown in Figure 1B. A decrease in cholesterol labeling and an increase in both ubiquinone and dolichol labeling were apparent. The labeling of the cholesterol intermediate squalene can be observed in the chromatogram and was also elevated after mevinolin treatment.

Figure 1
figure 1

Separation of mevalonate pathway lipids from control fibroblasts by HPLC. The radioactivity elution profiles shown were obtained from cells after pretreatment in the absence (A) or in the presence (B) of mevinolin. 1, cholesterol; 2, squalene; 3, ubiquinone; 4, dolichol.

Full size image

After [3H]-mevalonate labeling, the specific radioactivity of total cellular cholesterol varied in both nondiseased subjects and patients with peroxisomal diseases (Fig. 2A). Mevinolin treatment decreased this cholesterol labeling extensively, and, after a 24-h treatment of the cells with the inhibitor, only one third or less of the activity remained in comparison with the untreated cells. Ubiquinone labeling was approximately 35-fold less than that of cholesterol, but the ratio of specific activities (ubiquinone per cholesterol) in fibroblasts was still considerably higher than in most other tissues or cells(25) (Fig. 2B). The level of the radioactivity incorporated into ubiquinone also varied greatly between study subjects. However, after mevinolin pretreatment of cells, the labeling of ubiquinone increased approximately 2.5-fold in all experiments performed. The dolichol content was found to be very low in fibroblasts, and, in some cases, both the amount and the [3H]-mevalonate labeling of this lipid were below the detection limits (data not shown).

Figure 2
figure 2

Labeling of fibroblasts lipids with [3H]-mevalonate. Incorporation into cholesterol (A) and ubiquinone (B). Light bars, no inhibitor; dark bars, in the presence of 10 µM mevinolin. Patient 1 and 2, ZWS; 3-7, X-linked ALD (homozygote); 8, X-linked ALD (heterozygote); 9-11, controls. The fibroblasts from the ZWS patients were cultured on four different occasions and analyzed separately.

Full size image

The mean value of [3H]-mevalonate incorporation into cholesterol was not significantly different between the ZWS, ALD, and control groups whether or not a pretreatment with mevinolin was performed (Table 1). In addition, the labeling level of ubiquinone in the three groups was similar. Mevinolin treatment increased the specific activity of ubiquinone labeling 2.5-fold in all groups.

Table 1 Lipid labeling of patient and control fibroblasts
Full size table

Squalene synthase is a membrane-bound enzyme, and only a small amount of the enzyme protein is found in the cytosol if homogenization is appropriately performed and proteolysis is absent. In our experiments, <2% of the total activity was recovered in the supernatant, and, therefore, the activity of this enzyme was analyzed in the total membrane fraction (Fig. 3A). Squalene synthase activity was found in all samples tested, and, taking into account the variations observed, the enzyme activity of control fibroblasts was in the same range as that measured in the patient groups. In our fibroblasts preparations, only low levels of FPP synthase activity were recovered in the membrane fraction and almost all the activity was found in the cytosol. Comparing the samples from ZWS patients with those of the control group, the FPP synthase activity in the cytosol was approximately equal or higher in the ZWS group (Fig. 3B). The mean values of the enzyme activities are given in Table 2. No differences were identified in the activity of squalene synthase when the fibroblasts from the ZWS group were compared with those of the other groups. On the other hand, the mean value of the cytosolic FPP synthase activity was higher in both ZWS and ALD patients compared with the controls.

Figure 3
figure 3

Branch-point enzyme activities in fibroblasts from patients with peroxisome deficiency. (A) Squalene synthase (total membranes); (B) FPP synthase (supernatant); (C) FPP synthase (total membranes). The patients studied are presented in Figure 2.

Full size image
Table 2 Branch-point enzyme activities in patient and control fibroblasts
Full size table

DISCUSSION

In genetic diseases of peroxisomes, such as ZWS, peroxisomes are absent from cells. Consequently, the enzymes located in peroxisomes are also absent in these patients. X-linked ALD is a milder peroxisomal disease in which very long-chain fatty acid oxidation is deficient. Because the enzymes, which catalyze the terminal reactions involved in cholesterol and dolichol biosynthesis, are present not only in the ER but also in the peroxisomes, the fibroblasts of ZWS patients seemed to be an interesting model to study the contribution of peroxisomes to the cellular synthesis of these two lipids.

[3H]-Mevalonate incorporation into cholesterol in the fibroblasts of the patients with ZWS, in comparison with control individuals, was decreased by 21% and was also lower than in the ALD group, but this decrease was not statistically significant. One explanation for the low decrease could be that the contribution of peroxisomes to the total cellular cholesterol synthesis is small. The extent of peroxisomal participation in the synthesis of mevalonate pathway lipids has not been studied in detail but approximately 10% of cholesterol and 25% of dolichol have been estimated to be synthesized by this organelle(10). In fact, the amount of cholesterol in the fibroblasts from ZWS patients was found to be decreased to a limited extent (approximately 12%). Another explanation could be that the absence or decreased content of peroxisomal enzymes in patient fibroblasts results in a compensatory increase of the microsomal enzyme activities and the absolute amount of cholesterol is not affected on a cellular level. The extent of [3H]-mevalonate labeling of ubiquinone in these patients was not significantly different from that of the controls. This was to be expected, because ubiquinone is probably not synthesized in peroxisomes and trans-prenyltransferases as well as the terminal enzymes modifying the ubiquinone ring have been shown to be localized mainly in the ER-Golgi system(26–28).

Squalene synthase activity in the fibroblasts of all three groups was similar, whereas the mean value of FPP synthase activity in both ZWS and ALD fibroblasts was higher in comparison with the control cells but not significantly different. Obviously, the branch-point enzymes are not deficient in these conditions, and the biosynthetic pattern of cholesterol as well as other related lipids was not significantly altered.

The level of incorporation of [3H]-mevalonate into cholesterol was only slightly modified, indicating that extensive peroxisomal deficiency does not interfere with its biosynthetic pathway. Previously, it has been suggested that cholesterol synthesis is decreased in fibroblasts from patients with ZWS, because [3H]-acetate incorporation has been found to be diminished in the fibroblasts of the diseased patients(11,29). The differences obtained between acetate and mevalonate incorporation may be explained by an enzyme deficiency localized to the initial part of the mevalonate pathway. However, it is also possible that acetate is not an optimal precursor to be used in studies of cholesterol biosynthesis. Because of the extensive use of this intermediate in a number of metabolic and synthetic reactions, the actual level of this compound seems to be relatively low in cellular pools. Accordingly, saturation of enzymes of the mevalonate pathway, at least under certain conditions, may not occur.

Upon pretreatment of the cells with mevinolin, an inhibitor of the HMG-CoA reductase, the incorporation rate of [3H]-mevalonate into cholesterol was reduced whereas incorporation into ubiquinone and dolichol was increased. According to the flow diversion hypothesis(30), a decrease of the mevalonate pool mainly affects the synthesis of cholesterol because of the low affinity of squalene synthase for its substrate. On the other hand, the high affinities of the other branch-point enzymes, trans- and cis-prenyltransferases as well as farnesyl:protein transferase, for their substrates result in an unaffected synthesis of ubiquinone and dolichol and, also, unchanged rates of protein farnesylation. One possibility is that in our tissue culture system, no saturation of the branch-point enzymes of the latter lipids was reached and an increased mevalonate pool size resulted in elevated lipid synthesis. Alternatively, mevinolin treatment results in a decreased size of the mevalonate pool that affects the dilution of the externally added [3H]-mevalonate. Therefore, the smaller cellular pool and the absence of an increased synthetic rate result in higher labeling of ubiquinone and dolichol. Regardless of these alternatives, the labeling patterns of cholesterol, ubiquinone, and dolichol were similarly affected by mevinolin treatment in all types of fibroblasts investigated.

A number of proposals have been presented detailing the quantitative and qualitative localization of both the mevalonate pathway enzymes and those participating in the terminal synthetic reactions of cholesterol and dolichol biosynthesis(4,31). The experiments described here show that the role of fibroblast peroxisomes in cholesterol synthesis is of minor importance. This finding, however, does not necessarily mean that the peroxisomal contribution is quantitatively unimportant for cellular functions. The peroxisomal β-oxidation of fatty acids is also very limited in comparison with mitochondrial oxidation, but several substrates, i.e. polyunsaturated very long-chain fatty acids or branched fatty acids, are exclusively oxidized in peroxisomes(32). At present, we do not know the fate of cholesterol and dolichol synthesized in this organelle, but these lipids are most probably used for different purposes than those synthesized in the ER. Additionally, during various drug treatments, peroxisomal lipids and lipid-synthesizing enzymes are produced at significantly increased levels(9,33,34). Such an increase may reflect a requirement of cellular adaptation to certain conditions. This possibility for adaptation has been lost in peroxisomal disorders with direct functional consequences.