Main

In mammalian tissues, CPT in conjunction with acyl-CoA synthetase and carnitine acylcarnitine translocase provides the mechanism whereby LCFA are transferred from the cytosolic compartment to the mitochondrial matrix to undergo β-oxidation. According to the current working model, the CPT system consists in two enzymes located in the outer (CPT I) and in the inner(CPT II) mitochondrial membrane(1). The existence of distinct proteins supporting both activities had been disputed for a long time, from documented differences between CPT I and CPT II activities in terms of kinetic and regulatory properties(2, 3), or differential sensitivity to solubilizing agents(4, 5). Recent data, obtained in rat and human, established that CPT I and CPT II are distinct proteins(6–9).

In human pathology, CPT deficiency realizes schematically three distinct clinical presentations, characterized by muscular(10), hepatic(11), and hepatocardiomuscular(12) symptoms. CPT I activity was shown to be deficient in fibroblasts from patients with the hepatic form of the disease, but CPT II activity was decreased in fibroblasts from patients with the muscular and the hepatocardiomuscular forms identified in adults and infants, respectively(12, 13). The cloning of human CPT II cDNA(7) has allowed for the determination of the fact that CPT II defects resulted from mutations in the CPT II gene(14, 15). Molecular analysis of the CPT I defect has not been performed until now.

We show here, by complementation experiments performed between CPT I- and CPT II-deficient fibroblast cell lines, that CPT I and CPT II defects result from mutations in distinct genes.

METHODS

Twenty-four well microtiter plates were from Costar (Cambridge, MA).[9,10(n)-3H]Palmitic acid was obtained from Amersham International (Buckinghamshire, UK). Most other chemicals including PEG, fatty acid-free fraction V BSA, and Dowex-1 anion exchange resin were from Sigma Chemical Co. (St. Louis, MO). Ficoll 400 was from Pharmacia (Uppsala, Sweden). Ham's F-10 culture medium and FCS were from Life Technologies, Inc.(France).

Cells. Skin fibroblasts were grown in Ham's F-10 medium, supplemented with antibiotics and 12% FCS (complete Ham's F-10).

Cells were issued from controls and patients with fatty acid oxidation defects: CPT I-deficient cells, CPT IA, CPT IB; CPT II-deficient cells, infantile form CPT IIA, CPT IIB(12), CPT IIC, CPT IID, adult form CPT IIE, all identified by Dr. Demaugre, Paris; MCAD- and VLCAD-deficient cell lines were identified by Dr. Vianney Lyon.

Complementation analysis. The fusion procedure was similar to that described in Brivet et al.(16). Briefly, two cell lines were plated in 25-cm2 culture flasks at a density of 0.75 × 106 cells/line or 1.5 × 106 cells of the same line and grown in complete Ham's F-10 at 37 °C. Cells were exposed twice to 50% (vol/vol) PEG in Ham's F-10, for 60 s at 25 °C, 24 h and 48 h after plating. Twenty-four hours after the second PEG exposure, enrichment in multinucleated cells was obtained with a Ficoll density gradient consisting of 5-mL steps of 2.5, 5, 7.5, 10, 12.5, and 15% (wt/vol) Ficoll in Ham's F-10 buffered with 25 mM HEPES, pH 7.35, containing 20% FCS. Cells were layered on a 2.5% fraction, and after a 3-h sedimentation at room temperature, fractions 1 (2.5%) and 6 (15%), containing very few cells, were discarded; fractions corresponding to the 5 and 12.5% sedimentation steps were collected separately. The cells from the 5% fraction were used as cocultivation control within each fusion experiment, and the cells from the 12.5% fraction were used as fused cells. After washing, unfused cells and fused cells were plated in triplicate in 24-well microtiter plates and grown in complete Ham's F-10 for 48 h at 37 °C. Monolayers were then washed twice with PBS supplemented with 0.9 mM CaCl2 and 0.5 mM MgCl2, pH 7.35. 3H2O production from 100 μM [9,10(n)-3H]palmitate bound to 0.5 mg/mL albumin (sp act: 75 μCi/μmol) was measured at 37 °C in 200μL of the same buffer, and 3H2O release was recovered by ion-exchange treatment on Dowex-1 columns, as described elsewhere(17, 18). 3H2O production from unfused or fused cells was linear between 15 and 55 μg of protein/well. These conditions were obtained when 60,000 mononuclear cells or 20,000 polynuclear cells were platted in each well. 3H2O production was linear with time for up to 4 h of incubation. All experiments were conducted for a 2-h period. Cell proteins were measured as described previously(16). Results were expressed as nanomoles of3 H2O released/mg of protein/h.

Cocultures. [9,10(n)-3H]Palmitate detritiation was measured in cocultures including two cell lines. Cells were platted in triplicate in 24-well microtiter plates at a density of 30,000 cells/line or 60,000 cells of the same line, and grown for 48 h in complete Ham's F-10 at 37°C. Monolayers were washed twice with PBS, then 3H2O release from [9,10(n)-3H]palmitate was measured. At the end of the culture, the level of proteins per well was comprised between 20 and 30 μg, both in single cultures and in cocultures.

RESULTS

Complementation experiments were performed between two infantile CPT II(CPT IIA, CPT IIB)-deficient cell lines from patients with the infantile form of the defect and two CPT I (CPT IA, CPT IB)-deficient cell lines. LCFA oxidation estimated either by 14CO2 release from 1-14C-fatty acids or by 3H2O release from 9,10(n)-3H-fatty acids was shown strongly decreased in CPT I(about 70%)- and infantile CPT II (about 90%)-deficient fibroblasts(13, 19–22) (M. Brivet, unpublished data). 3H2O production from[9,10(n)-3H]palmitate is an assay requiring low numbers of cells. Thus restoration of 3H2O production from[9,10(n)-3H]palmitate was chosen as the criterion for complementation in fusion experiments between CPT I- and CPT II-deficient cell lines.

Fibroblasts were fused in the presence of PEG. Unfractionated cells generally contain between 10 and 30% multinucleated cells. Such fusion frequencies are not always sufficient for the detection of complementation(23). Thus fused and unfused cells were sorted by sedimentation on a six-step gradient of Ficoll. The 5% Ficoll fraction containing less than 10% of multinucleated cells was considered as the cocultivated control (unfused cells), whereas the 12.5% Ficoll fraction containing 90% of multinucleated cells was considered as the fused cell fraction. MCAD(24)- and VLCAD(25)-deficient cell lines were included in this study as controls for positive complementation.

The results of the different fusions carried out are summarized inTable 1. 3H2O production by homopolykaryons was 64-98% of that measured for unfused cells, suggesting some depressed metabolic activity of multinucleated cells obtained by PEG fusion as previously reported(16, 17). Fusion between the two distinct genetic defects, MCAD and VLCAD, resulted in an increased3 H2O production by heterokaryons relatively to that of unfused cells (49%) or to the mean 3H2O production by the homopolykaryons (61%).

Table 1 Complementation analysis of CPT I × CPT II-deficient fibroblasts
Full size table

3H2O production measured in the 5% Ficoll fraction was decreased by about 70% in CPT I-deficient cells and by more than 90% in CPT II-deficient cells (Table 1) when compared with control cells. In experiments including CPT IIA (CPT IIA × CPT IA and CPT IIA× CPT IB), 3H2O production measured in unfused cells was close to the mean 3H2O production by both cell lines. In heteropolykaryons, 3H2O production was increased as compared either with that of unfused cells (60%) or with the mean production of3 H2O by homopolykaryons (90%). The magnitude of this increase is similar to that observed in the positive complementation resulting from fusion of MCAD and VLCAD cell lines. In experiments including CPT IIB cells (CPT IIB× CPT IA and CPT IIB × CPT IB), 3H2O production by unfused cells was low (about 30% of the mean 3H2O production of both cell lines). In heteropolykaryons, 3H2O production was increased by 480 and 290% compared with unfused combinations but only by 30 and 25% when compared with the mean production of 3H2O by the homopolykaryons.

The possible inhibitory effect of CPT II-deficient cell lines upon LCFA oxidation by cocultured cells was tested in combinations including either CPT IA or control cell lines (see Table 2). Five CPT II-deficient cell lines were included in this study: four of them were from patients with the infantile form of the defect (CPT IIA, CPT IIB, CPT IIC, and CPT IID), and the last one was from a patient with the adult form of the defect (CPT IIE). 3H2O production was lower than the predictible values (mean 3H2O production by cell lines cultivated alone) in cocultures including infantile CPT II- and CPT I-deficient cell lines. Similar data were obtained when infantile CPT II-deficient cells were cocultured with a control cell line. No impaired LCFA oxidation was observed in cocultures including either adult CPT II (CPT IIE)- or CPT IA-deficient cell line.

Table 2 3 H 2 O production from [9,10(n)- 3 H]palmitate in CPT IA- and CPT IIA-E-deficient cell lines cultivated alone or in cocultures with CPT IA or control cells
Full size table

To determine whether the inhibitory effect of infantile CPT II-deficient cell lines upon LCFA oxidation by cocultured cells was L-carnitine-dependent, we measured 3H2O production by cocultivated control and CPT IIB-deficient cell lines, in the presence of L-carnitine. As shown inFigure 1, L-carnitine was without effect upon palmitate detritiation by control cells. Conversely, it increased palmitate oxidation by CPT IIB cells and by the coculture, including both cell lines. Complete restoration of LCFA oxidation by coculture required a high concentration of L-carnitine (2 mM). 3H2O production from[9,10(n)-3H]palmitate measured without and with 2 mM L-carnitine was 9.37 ± 0.55 versus 9.54 ± 0.55 in control cells, 0.31 ± 0.03 versus 0.98 ± 0.05 in CPT IIB, and 1.26 ± 0.09 versus 5.73 ± 0.12 nmol/h/mg of protein, in cocultures (mean ± SD of three distinct experiments). The effect of 2 mM L-carnitine was then tested in fusion experiments(Table 1) between the CPT IIB cell line and the CPT I-deficient cell lines. Compared with the experiments without L-carnitine,3 H2O production was increased by about 300 and 100% in unfused and fused cells, respectively. The percentage change with fusion was about 150% relative to unfused cells. 3H2O production by heterokaryons was now similar to that of heterokaryons resulting from CPT IIA × CPT IA or CPT IIA × CPT IB.

Figure 1
figure 1

Effect of L-carnitine on 3H2O production from [9,10(n)-3H]palmitate in cocultures associating control and CPT IIB-deficient cells. Palmitate detritiation by control (□) and CPT IIB (•) cell lines, and by coculture (▴) including both cell lines, was studied as described in “Methods,” in the presence of different concentrations of L-carnitine. Results of a typical experiment are shown.

Full size image

DISCUSSION

CPT consists of two activities located in the outer (CPT I) and the inner(CPT II) mitochondrial membranes. Both enzymes display very different kinetic and regulatory properties(2, 3), and exhibit different stability in the presence of detergents(4, 5). Until recently, the basis of their differences was not clear. One hypothesis was that CPT I and CPT II were similar polypeptides. Their different behavior, although membrane-associated, could be ascribed to differences in their physical environment. The second hypothesis was that CPT I and CPT II were different proteins. The debate was recently resolved: cDNA cloning of CPT I and CPT II from rat and human liver confirms that CPT I and CPT II are distinct proteins(6–9). Data from human pathology pleaded in favor of such conclusions: CPT I and CPT II activities were never simultaneously depressed in CPT defects(12). Antibodies raised against human hepatic CPT II recognized a protein in normal amounts in fibroblasts from patients with a CPT I defect, whereas this protein was strongly decreased in fibroblasts from CPT II-deficient patients(26).

In this work, we have performed a complementation analysis of CPT I and CPT II defects. LCFA oxidation, estimated by 3H2O production from[9,10(n)-3H]palmitate, was always depressed in CPT I-deficient fibroblasts(19–21). The CPT II defect results in two distinct clinical forms, one adult(10), the other infantile or neonatal(13, 27). Mutations in the CPT II cDNA were reported in both forms(14, 15, 28). However, LCFA oxidation was strongly reduced only in fibroblasts from patients with the infantile form of the disease(22). As a consequence, the complementation studies were performed with fibroblasts from patients with the infantile form of the CPT II defect. The increase of 3H2O production by fused cells was used as the criterion of positive complementation.

In heterokaryons resulting from the fusion of CPT II-deficient cells (CPT IIA × CPT IIB), no restoration of palmitate detritiation was observed. This suggests that the enzymatic defects of these cell lines resulted from mutations in the same gene. These data were predictible, because the enzymatic defect of these cell lines was shown to be associated with a defective CPT II protein(13, 22) resulting from the substitution of a tyrosine by a serine in CPT II protein for CPT IIA(29). Similarly, no increased palmitate detritiation was observed in CPT IA × CPT IB heteropolykaryons, indicating that, in these cell lines, the CPT I defect results from mutations in the same gene which might be the CPT I gene. CPT I activity was reported to be modulated by membrane components(30). Thus one cannot exclude that the CPT I defect could result from mutations in genes encoding such components. Tools of molecular biology will clarify this point. In combinations including CPT I (CPT IA or CPT IB)- and CPT IIA-deficient cells,3 H2O production from [9,10(n)-3H]palmitate by unfused cells was close to the mean value of 3H2O production by both cell lines. Fusion of CPT IIA with either CPT IA or CPT IB resulted in a positive complementation, as evidenced by the similar increase of3 H2O production by heteropolykaryons when compared with that of unfused cells or to the mean 3H2O production by homopolykaryons(about 60 and 90%, respectively). These findings confirm, as suggested previously(12, 26), that CPT I and CPT II defects result from mutations in different genes. Similar results were expected in experiments including CPT I and CPT IIB cell lines. Surprisingly, palmitate oxidation by unfused cells was low (near to that of the CPT IIB cell line), suggesting that CPT IIB cells impaired palmitate oxidation by CPT I-deficient cells. The percent change of 3H2O production by fused cells was high (480 and 290%) relative to the unfused cells, but slight (30 and 25%) when compared with mean 3H2O production by homopolykaryons. The possible inhibitory effect of CPT IIB upon palmitate oxidation by cocultured cells did not allow an unequivocal analysis of data from experiments including those with the CPT IIB cell line.

The effect of CPT II-deficient cells upon palmitate oxidation was further investigated directly in cocultures, including either CPT I-deficient or control cell lines. Four infantile and one adult CPT II-deficient cell lines were included in this study. LCFA oxidation was decreased in cocultures including CPT I- and infantile CPT II-deficient cells, but normal in experiments with adult CPT II-deficient cells. Palmitate oxidation was decreased by about 70 and 30% in cocultures, including CPT IIB-D and CPT IIA cell lines, respectively. The inhibitory effect of the CPT IIA cell line was not observed in complementation experiments, suggesting that PEG and/or Ficoll treatment of cells could modulate slightly this effect. Similar results were obtained when control cells were cocultivated with CPT II-deficient cells. Residual enzymatic activity of infantile CPT IIA-D and adult CPT IIE was of similar magnitude (11, 5, 8, 7, and 15% of control values, respectively). However, LCFA oxidation was strongly decreased (>90%) in infantile CPT II-deficient cells but moderately impaired (50%) in CPT IIE cells. Whether the magnitude of the inhibitory effect of CPT II-deficient cells is correlated with the severity of their metabolic block is questionable.

Medium chain fatty acid oxidation was normal in cocultures including CPT II-deficient cells (data not shown). This indicates that the inhibitory effect of these cell lines is restricted to LCFA oxidation. One hypothesis is that infantile CPT II cell lines impair carnitine metabolism in cocultured cells. Thus we tested the effect of L-carnitine upon palmitate oxidation by coculture including CPT IIB and control cells. L-Carnitine has no metabolic effect in control cells, but it increased, by a dose-dependent effect, palmitate oxidation by coculture as well as by CPT II cells. These data suggest that control cells (and probably CPT I-deficient cells) when cocultivated with infantile CPT II-deficient fibroblasts are unable to maintain an adequate carnitine level for LCFA oxidation. Carnitine uptake in fibroblasts is mediated by a high affinity carrier (Km = 5 μM) which is inhibited(31) by palmitoylcarnitine(Ki = 0.37 μM). Fatty acid (15 μM) and L-carnitine (2.5 μM) are supplied in culture medium by FCS. Thus CPT II-deficient cells, unable to further metabolize palmitoylcarnitine, would release in the medium long chain acylcarnitines (probably palmitoylcarnitine), which could impair carnitine uptake by cocultured cells. Thus, only L-carnitine from passive diffusion would be available for LCFA oxidation. This may explain why a high carnitine concentration was needed to suppress CPT IIB-induced inhibition upon palmitate oxidation by cocultured cells. Further studies are required to determine more accurately the mechanism of this inhibition. A metabolic cooperation between cocultivated cells might be also suggested to explain the metabolic effect of L-carnitine. High carnitine concentrations could increase palmitoylcarnitine formation in CPT II-deficient cells, so that palmitoylcarnitine exported to the medium could be taken up by CPT I-deficient or control cells to be oxidized more effectively. Such an effect is probably of minor importance. Indeed, metabolic effects of L-carnitine do not require its presence in the assay medium. They were similarly observed when carnitine was provided in culture medium during the 48-h period of culture and then discarded by washing before palmitate detritiation experiments (data not shown).

Analysis of complementation experiments including CPT IIB cell line was difficult owing to the inhibitory effect of CPT IIB cells upon palmitate oxidation by CPT I-deficient cells. When 2 mM L-carnitine was provided in a such experiments, palmitate oxidation was increased by 300 and 100% in unfused and fused cells, respectively. In heteropolykaryons, 3H2O production (increased by 150% compared with that of unfused cells) was similar to that measured in heteropolykaryons resulting from MCAD and VLCAD or CPT IIA- and CPT I-deficient cells. These data show that 2 mM L-carnitine exposes a positive complementation between CPT IIB- and CPT I-deficient cells by overcoming the inhibitory effect of CPT IIB cells upon LCFA oxidation.

Distinct clinical presentations were ascribed to CPT II deficiency. Most cases are characterized by episodes of muscle necrosis with paroxysmal myoglobinuria in young adults(32). Another presentation has been more recently described in infants. Symptomatology consists of attacks of hypoketotic hypoglycemia with a high level of plasma creatine kinase and heart injury, leading frequently to sudden death(12, 27). The pathophysiologic basis of these two presentations is not clear, because the magnitude of the enzymatic defect is not very different in the two presentations. We previously showed that LCFA oxidation was strongly decreased in fibroblasts from patients with the infantile form of the defect, whereas it was not depressed or only moderately impaired in adult patients(22). In this work, we identified another metabolic difference between these two clinical forms of the CPT II defect.