Main

MKD is an inherited disorder of cholesterol and non-sterol isoprenoid biosynthesis in man resulting in severe pre- and postnatal pathology(1). The mevalonate pathway produces isoprenoids, which are critical for multiple cellular functions, from cholesterol biosynthesis to growth control(2–4) (Fig. 1). To ensure constant production of different isoprenoid compounds at all stages of growth while maintaining cholesterol levels, cells must carefully regulate mevalonate production and avoid overaccumulation of potentially toxic products. The results of in vitro studies on the affinities of the branch-point enzymes for farnesyl pyrophosphate(5), analyses of cholesterol and non-sterol isoprenoid biosynthesis in response to HMG-CoA reductase inhibitors(6–8), and in vivo experiments focused on dietary manipulation and treatment with HMG-CoA reductase inhibitors(9) cumulatively suggest that incomplete inhibition of cholesterol biosynthesis does not significantly reduce non-sterol isoprenoid levels. This observation is apparently the result of the high affinity of the branch-point enzymes for farnesyl and geranyl pyrophosphates (Fig. 1). However, the results of other in vivo studies using HMG-CoA reductase inhibitors in rats are at variance with this concept(10, 11). HMG-CoA reductase inhibitors were found to act nonexclusively on HMG-CoA reductase, influencing later steps in the pathway, including the branch-points leading to increases or decreases of non-sterol isoprenoids as a function of the experimental conditions, as well as effects on the LDL receptor(12).

Figure 1
figure 1

The branched pathway of cholesterol and non-sterol isoprenoid biosynthesis. Mevalonate kinase deficiency is identified by the solid square. Cholesterol-lowering drugs inhibit the preceding enzyme, HMG-CoA reductase, the rate-limiting enzyme of the pathway. Arrows indicate points of feedback inhibition of HMG-CoA reductase and mevalonate kinase by cholesterol and pyrophosphorylated intermediates.

Full size image

MKD is a multisystem pathology with substantial phenotypic variation. In the approximately 15 known patients, the phenotype includes mild to moderate psychomotor retardation, hypotonia, ataxia, myopathy, dysmorphism, cataracts, hepatosplenomegaly, lymphadenopathy, and anemia. About one-half of patients have died, whereas others have survived into late adolescence(1). Because the cholesterol pathway is involved in production of sterol in addition to non-sterol isoprenoids, including ubiquinone, dolichol, heme A among others, we sought to ascertain if impaired synthesis or quantity of non-sterol isoprenes might be a causative factor in the pathology of MKD.

In earlier work, we observed a substantial level of unsuppressible HMG-CoA reductase activity in MKD fibroblasts, grown in medium containing cholesterol supplied by fetal bovine serum(13). Recently we demonstrated a deficiency of ubiquinone 10 in MKD in vivo and in vitro(14). We have now extended these studies in MKD cell lines by assessing the regulatory adaptation of intracellular cholesterol biosynthesis and the LDL receptor pathway in response to different concentrations of LDL and non-lipoprotein-bound cholesterol. In addition, we estimated rates of N-linked protein glycosylation in MKD cells to gain insight into intracellular levels of dolichol and dolichol phosphate. Our results provide further evidence that the pathophysiology of MKD involves alterations in sterol as well as in non-sterol isoprenoid biosynthesis.

METHODS

Cell culture. Lymphoblasts were grown in RPMI 1640 medium and fibroblasts in α-medium (Biochrom, Berlin, FRG). All media contained additional glutamine (2 mM), streptomycin (100 mg/L), penicillin 100 U/L), and 100 g/L fetal bovine serum, or fetal bovine serum from which lipoproteins had been removed as described previously(13). Incubation conditions were 37 °C and 5% CO2. The fibroblasts were harvested by trypsinization(12).

Metabolite and pathway flux estimations. Cholesterol biosynthesis was estimated by monitoring the incorporation of[14C]acetate into cholesterol(13). Acetate is metabolized through other pathways, and its intracellular concentration is perhaps insufficient for saturation of the cholesterol synthetic pathway. Conversely, methods for the determination of absolute rates of sterol biosynthesis require milliCurie quantities of 3H2O, which can present a considerable biohazard. For these reasons, we chose[14C]acetate as a precursor. The dolichol and cholesterol contents in cultured cells were determined using HPLC(15). Cholesteryl ester formation, an indicator of LDL receptor pathway activity, was assessed by monitoring the formation of cholesterol esters with[14C]oleate as a substrate(16).

[14C]Galactose incorporation in cultured fibroblasts. N-Linked protein glycosylation was estimated by monitoring the incorporation of [14C]galactose into macromolecules. Fibroblasts derived from control individuals, patients with MKD, and patients suffering from galactose-1-phosphate uridyl diphosphate transferase deficiency (classical galactosemia) were grown to confluency in 60× 15-mm Falcon Petri dishes using Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. Cultures were labeled with 5 mL of medium containing 10% fetal bovine serum, 0.55 mmol/L D-galactose, and 0.1 mCi/mL D-[1-14C]galactose for 24 h. At the end of the incubation period, the medium was removed and frozen at -20 °C until further use. After washing the cultures with 0.9% NaCl, cells were harvested by trypsinization and pelleted by centrifugation at 500 × g for 3 min. The cell pellets were washed with cold (4 °C) saline and centrifuged again. Pellets were stored in tubes at -20 °C before extraction. For extraction of [14C]galactose-labeled macromolecules, 2 mL of 80% ethanol were added to the frozen pellets, and tubes were heated in an 85°C waterbath without boiling. After cooling the ethanol-precipitated cells, the tubes were centrifuged at 800 × g for 5 min, and the ethanolic supernatant was aspirated. Fresh 80% ethanol was added, and the entire procedure was repeated twice(17). Finally the pellets were dried by decanting the tubes. Pellets were then sonicated in 1 mL of distilled water for 20 s at 50 W (Branson Sonifier, Danbury, CT). Triplicate aliquots of the sonicate were counted by β-liquid scintillation spectrophotometry. Protein was determined by the Lowry method.

Secreted [14C]galactose-labeled macromolecules were estimated after dialysis of the incubation medium. To 1.5 mL of medium, 2.7 mg of D-galactose(nonlabeled, final concentration 10 mmol/L) were added. Then 1 mL of the medium was dialyzed against 10 mM cold galactose in 3 L of serum-free medium for 6 h, followed by dialysis against distilled water overnight. An aliquot of the dialysate was counted by β-liquid scintillation spectrophotometry. Total disintegrations/min were estimated by correcting the radioactivity of the sample with respect to total volume and related to cellular protein determined by the Lowry method.

RESULTS

Despite mevalonate kinase deficiency in extracts of fibroblasts derived from the patients (≅ 1% of control mean)(1, 13), there was still substantial cholesterol biosynthesis, determined using[14C]acetate incorporation into cholesterol in intact fibroblast monolayers (Fig. 2). Time-dependent experiments showed a linear incorporation of [14C]acetate into cholesterol from 12 to 48 h in control as well as MKD fibroblasts (data not shown).

Figure 2
figure 2

Effect of LDL cholesterol on [14C]acetate incorporation into cholesterol in control and MKD fibroblasts. Cells were incubated in the presence of the indicated amount of LDL and 0.1 mCi/mL[14C]acetate. After a 24-h incubation, radiolabeled cholesterol was estimated as described in “Methods.”

Full size image

The data of Fig. 2 may actually provide insight into some aspects of the clinical phenotype of MKD. Whereas patient 1 increased cholesterol biosynthesis to within the control range in these studies, patient 2 responded only minimally. This may in part reflect the nature of the defect in these patients, as patient 1 is one of the more mildly affected patients, presently attending the third grade of a school for physically handicapped children at the age of 9 y, whereas patient 2 was one of the very severely affected, having died at 19 mo(1). After withdrawal of cholesterol, MKD fibroblasts increased cholesterol biosynthesis. However, in the presence of only small amounts of LDL cholesterol, cholesterol biosynthesis in MKD and control cells was suppressed. In MKD cells the incorporation fell to approximately 30% of that of control cell lines.

We previously demonstrated that the enzymatic defect in MKD was partially counteracted by increased and partially unsuppressible HMG-CoA reductase activity in comparison with that of control cells(13). In those studies, the greatest percentage difference between patient and control fibroblasts was observed in the presence of exogenous cholesterol. Mean HMG-CoA reductase activity for MKD fibroblasts maintained in medium containing 10% fetal bovine serum was 63 ± 44 pmol/min/mg of protein(±1 SD, range 38-146)(13). This was 6-fold higher than the control mean of 11 ± 4 (range 8-15). HMG-CoA reductase activity was regulated in MKD fibroblasts comparable to control cells.

The activity of the LDL receptor pathway was estimated in control and MKD fibroblasts via determination of cholesterol ester formation. Upon addition of unbound cholesterol to culture medium, rates of cholesterol ester formation were indistinguishable between control and MKD fibroblasts(Fig. 3, left). However, cholesterol ester formation, an indicator of the activity of the LDL receptor pathway, was roughly 2-fold increased in MKD fibroblasts when cells were incubated in the presence of LDL-cholesterol (Fig. 3, right), suggesting an approximate 2-fold increased activity of the LDL receptor pathway.

Figure 3
figure 3

Effect of LDL cholesterol on cholesteryl ester formation from [14C]oleate as a function of cholesterol(left) and LDL-cholesterol (right). Incubation period was 5 h.

Full size image

To estimate the influence of the isoprenoid pathway upon cellular growth, the growth rates of cultured fibroblasts were determined under different conditions (Table 1). All cell lines grew at comparable rates in media supplemented with 10% fetal bovine serum. In lipid-depleted medium, supplemented with insulin and fibroblast growth factor, control fibroblasts grew equally well or better than in the preceding medium, but the growth rate of MKD fibroblasts decreased. When the medium was then supplemented with exogenous cholesterol (50 mg/mL) to suppress HMG-CoA reductase, the growth rate of control fibroblasts increased further. However, growth of MKD fibroblasts was further retarded. Apparently, down-regulation of HMG-CoA reductase in MKD fibroblasts by cholesterol is detrimental, perhaps leading to a shortage of non-sterol isoprenoids. We obtained evidence to support this concept by estimating quantities of dolichol and cholesterol in cultured lymphoblasts derived from MKD patients.

Table 1 Slopes of semilogarithmic growth curves of fibroblasts
Full size table

The availability of cultured lymphoblasts derived from MKD patients offered the opportunity to quantitate sterol and non-sterol isoprenoids. For these studies, large numbers of cells were needed, and cultured lymphoblasts from patients fulfilled this requirement. In lipid-free media, control lymphoblasts increased their cholesterol content 157% in comparison with growth in medium containing untreated fetal bovine serum (Table 2). MKD lymphoblasts kept their cholesterol levels roughly constant, perhaps suggesting that the cells were functioning at their maximum level of cholesterol biosynthesis. The concentration of dolichol in control lymphoblasts maintained in lipid-free media appeared lower, whereas dolichol levels remained about constant in MKD lymphoblasts. Notably, the mean dolichol concentration from MKD lymphoblasts was one-half the value detected in control lymphoblasts under both growth conditions.

Table 2 Dolichol and cholesterol concentrations in Epstein-Barr virus-transformed lymphoblasts
Full size table

N-Linked protein glycosylation in fibroblasts was estimated by determining the incorporation of labeled [14C]galactose into protein, and the secretion of [14C]galactose-labeled macromolecules into cell culture medium (Table 3). These studies were performed to assess the possibility that intracellular dolichol (and possibly dolichol phosphate) levels might be altered in cultured MKD lymphoblasts. In these studies, two fibroblast cell lines derived from patients with galactose-1-phosphate uridyl transferase deficiency (classical galactosemia) were investigated in parallel. The incorporation of [14C]galactose into intracellular macromolecules in MKD fibroblasts was not significantly different from that of control cells after 24 h. Conversely, the incorporation of [14C]galactose into macromolecules that were secreted into the cell culture medium was significantly decreased in MKD fibroblasts in comparison with control cells (p < 0.0001, Mann-Whitney nonparametric analysis). The same activity was also significantly decreased in fibroblasts derived from the parents of one patient (p < 0.01)(Table 3). As expected, both incorporation activities(radiolabeled intracellular and secreted macromolecules) were severely reduced in fibroblasts derived from patients with classical galactosemia(Table 3).

Table 3 Incorporation of [14C]galactose into intracellular macromolecules and secretion of[14C]galactose-labeled macromolecules into cell culture medium using fibroblasts derived from controls, patients with MKD, parents of MKD patients, and patients with galactose-1-phosphate uridyltransferase deficiency (galactosemia) fibroblasts
Full size table

DISCUSSION

The clinical and biochemical investigation of patients with MKD may provide new insight into the diversity of cellular functions of isoprenoids and the regulation of their biosynthesis. The location of the defective enzyme in the isoprenoid pathway makes MKD an interesting model in which to study the(patho)biochemical effects of inhibiting the preceding enzyme, HMG-CoA reductase, using drugs which are widely prescribed clinically (Fig. 1). We have demonstrated regulatory mechanisms in MKD fibroblasts that appear to mimic those observed in normal cultured cells treated with HMG-CoA reductase inhibitory drugs(3, 4). It is of interest that in both situations HMG-CoA reductase activity and LDL receptor activity are increased, although the mechanisms leading to these parallel situations may be quite different. These intracellular adaptations likely account in at least a partial restoration of intracellular cholesterol biosynthesis in MKD cells. The main mechanisms for restoring intracellular cholesterol are increased activities of HMG-CoA reductase and the LDL receptor pathway. The latter mechanism is likely responsible for the observed reduction of LDL cholesterol by cholesterol-lowering agents in hypercholesterolemic patients(3, 4).

Combining our earlier results on HMG-CoA reductase regulation and the present data on cholesterol uptake via the LDL receptor pathway, it would appear that MKD cells attempt to counteract their defect in two ways. First, there is an upregulation of the HMG-CoA reductase to provide the cell with precursors, and additionally, there is an increased uptake of LDL-cholesterol displayed by a larger number of LDL receptors. This process might be considered as two counteractive mechanisms, with increased reductase activity to ensure sufficient non-sterol isoprenoid intermediates at the expense of cholesterol, and increased expression of LDL receptors to ensure sufficient uptake of LDL cholesterol, thus enabling production of non-sterol isoprenoids from mevalonate.

Growth factors are located in the non-sterol region of isoprenoid biosynthesis. For example, isoprenyl derivatives are responsible for prenylation of Ras proteins and dolichol derivatives(18). In addition, cholesterol is also needed for the synthesis of cellular membranes during successive cell divisions. Along these lines, the growth rates in the lymphoblasts used to estimate cholesterol and dolichol levels (Table 2) were higher than the same rates in fibroblasts used for the growth curves in Table 1, which could suggest enhanced cholesterol utilization in the lymphoblastoid lines. However, to estimate the small amount of dolichol which we measured, substantial numbers of cells were needed which could not be reasonably achieved with cultured fibroblasts (Table 2). A shortage of non-sterol isoprenoids due to alterations in dolichol or dolichol phosphate pools might explain the decreased growth of MKD fibroblasts in cholesterol-free media (Table 1). After addition of exogenous cholesterol, the further inhibition of HMG-CoA reductase could lower the pool of non-sterol isoprenoids further, perhaps leading to reduced cell division (Table 1).

Altered dolichol or dolichol phosphate pools in MKD lymphoblasts, which was inferred from our results with galactose, is consistent with earlier results demonstrating reduced concentrations of ubiquinone 10 in MKD in vivo and in vitro(14). Reduced intracellular dolichol pools would argue against the hypothesis that the synthesis of non-sterol isoprenoids in comparison with cholesterol is always assured by the high affinities of the branch-point enzymes for farnesyl pyrophosphate(3, 4, 6–8). It is likely that Km values for farnesyl pyrophosphate for the enzymes of the protein prenylation pathway are lower than those for incorporation of farnesyl pyrophosphate into sterol(8). On the other hand, other investigators have demonstrated that Km values for intermediates at the branch points between cholesterol formation and the dolichol phosphate synthetic pathway are not significantly different(19). It is more likely that constant synthesis of non-sterol isoprenoids is ensured through coordinate regulation of HMG-CoA reductase with branch-point enzymes, thereby maintaining the levels of isopentenyl pyrophosphate and farnesyl pyrophosphate, to serve both sterol and non-sterol isoprenoid biosynthesis.

Overall, our data suggest altered dolichol (and possibly dolichol phosphate) pools in MKD fibroblasts. First, dolichol levels were lower in cultured lymphoblasts derived from MKD patients in comparison with control cells. Second, N-linked protein glycosylation, which is likely controlled by concentrations of dolichol phosphate(11), was impaired in MKD fibroblasts when the level of secreted[14C]galactose-labeled macromolecules was determined. This quantitative reduction in [14C]galactose-labeled secreted macromolecules in MKD fibroblasts is even more significant when considering that galactose is incorporated into O-linked in addition to collagen saccharides, and cultured fibroblasts will produce collagen. Therefore, the residual level of[14C]galactose secreted macromolecules into cell culture medium for MKD fibroblasts may reflect levels of collagen synthesis, which could suggest an even greater defect in N-linked protein glycosylation in MKD cells. However, one should be aware that the procedure used is somewhat unspecific to follow the function of dolichyl phosphate-mediated glycosylation.

Intracellular accumulation of [14C]galactose-labeled macromolecules results from the balance between synthesis of N-glycosylated galactose-containing glycoproteins and the rate of their lysosomal degradation. This process follows a nonlinear time curve with an equilibrium after 24-48 h(17). The secretion of galactose-labeled macromolecules reflects the cumulative actions of synthesis, transport, and secretion of glycoprotein that accumulates in the extracellular medium in a linear fashion. Based upon these time courses, it is perhaps not surprising that a difference in glycosylation can be demonstrated for MKD fibroblasts in the secreted macromolecules only. Presumed alterations of dolichol levels and a known deficiency of ubiquinone in MKD suggests that a deficiency of non-sterol isoprenoids in MKD patients (probable impairment of dolichol phosphate-dependent glycosylation and ubiqui-none-related oxidative stress) may be a pathogenetic factor in the course of the disease, and suggests new therapeutic approaches in MKD.