Search

Recombinant, normal human medium-chain acyl-CoA dehydrogenase (MCADH) and the common, human disease-causing K304E mutant ([Glu304]MCADH) protein were expressed in Escherichia coli using an optimized system, and the enzymes were purified to apparent homogeneity. The crucial factor leading to the production of active [Glu304]MCADH protein is the expression in E. coli cells at reduced temperature (28 degrees C). Expression in the same system at 37 degrees C results in very low amounts of active mutant protein. Several catalytic and physicochemical parameters of these two proteins have been determined and were compared to those of purified pig kidney MCADH. Although [Glu304]MCADH has approximately the same rate of substrate reduction with dodecanoyl-CoA and the same V(max) as human MCADH with the best substrate for the latter, octanoyl-CoA, the K(m) in the mutant MCADH is fourfold higher, which generates a correspondingly lower catalytic efficiency. Importantly, V(max) obtained using the natural acceptor, electron transfer flavoprotein, is only a third that for human MCADH. The V(max)/K(m) versus chain-length profile of the mutant shows a maximum with dodecanoyl-CoA which differs markedly from that of human MCADH, which has maximal efficiency with octanoyl-CoA. The substrate specificity of the mutant is broader with a less pronounced activity peak resembling long-chain acyl-CoA dehydrogenase. The purified mutant enzyme exhibits a reduced thermal stability compared to human wild-type MCADH. The major difference between the two proteins expressed in E. coli is the more pronounced lability of the K304E mutant in crude extracts, which suggests a higher susceptibility to attack by endogenous proteases. Differences between tetrameric [Glu304]MCADH which survives the first step(s) of purification and corresponding MCADH are minor. The overall differences in properties of [Glu304]MCADH together with its impaired folding and tetramer assembly may contribute to the generation of the abnormalities observed in patients homozygous for the K304E mutation.

Long-chain-acyl-CoA dehydrogenase (LCADH) has been produced by recombinant techniques from the human cDNA and purified after expression in Escherichia coli. Pig kidney LCADH was purified using an optimized method which also produces apparently pure short-chain-acyl-CoA dehydrogenase (SCADH) and medium-chain-acyl-CoA dehydrogenase (MCADH) in good yields. LCADH from both sources has a maximal turnover rate (Vmax of 650-700 min(-1) at pH 7.6) with the best substrates, which is approximately fivefold higher than reported previously. The human enzyme has an approximately fivefold higher Km compared with the pig kidney enzyme with substrates of chain length from C10 to C18 and a significantly different dependence of Vmax on the chain length. Pig kidney LCADH has a similar Vmax/Km with C10 to C14 substrates as MCADH does with C6 to C10 substrates. Recombinant human LCADH, however, is significantly less efficient (approximately fourfold with C12) than purified pig kidney enzyme. We conclude that human LCADH is either quantitatively less important in beta-oxidation than in the pig, or that post-translational modifications, not present in the recombinant human enzyme, are required to optimize human LCADH activity. Our results demonstrate that LCADH is as important as the other acyl-CoA dehydrogenases in fatty acid oxidation at physiological, mitochondrial pH with optimal substrates of chain length C10-C14. The extent of the LCADH-flavin cofactor reduction observed with most substrates and the rate of the subsequent reoxidation with oxygen are markedly different from those found with human medium chain acyl-CoA dehydrogenase. Both LCADH are inactivated by the substrate analogue 2-octynoyl-CoA, possibly via covalent modification of Glu261, the active-site residue involved in deprotonation of the substrate (alpha)C-H.

The catalytically essential glutamate residue that initiates catalysis by abstracting the substrate alpha-hydrogen as H+ is located at position 376 (mature MCADH numbering) on loop JK in medium chain acyl-CoA dehydrogenase (MCADH). In long chain acyl-CoA dehydrogenase (LCADH) and isovaleryl-CoA dehydrogenase (IVDH), the corresponding Glu carrying out the same function is placed at position 255 on the adjacent helix G. These glutamates thus act on substrate approaching from two opposite regions at the active center. We have implemented the topology of LCADH in MCADH by carrying out the two mutations Glu376Gly and Thr255Glu. The resulting chimeric enzyme, "medium-/long" chain acyl-CoA dehydrogenase (MLCADH) has approximately 20% of the activity of MCADH and approximately 25% that of LCADH with its best substrates octanoyl-CoA and dodecanoyl-CoA, respectively. MLCADH exhibits an enhanced rate of reoxidation with oxygen, however, with a much narrower substrate chain length specificity that peaks with dodecanoyl-CoA. This is the same maximum as that of LCADH and is thus significantly shifted from that of native MCADH (hexanoyl/octanoyl-CoA). The putative, common ancestor of LCADH and IVDH has two Glu residues, one each at positions 255 and 376. The corresponding MCADH mutant, Thr255Glu (glu/glu-MCADH), is as active as MCADH with octanoyl-CoA; its activity/chain length profile is, however, much narrower. The topology of the Glu as H+ abstracting base seems an important factor in determining chain length specificity and reactivity in acyl-CoA dehydrogenases. The mechanisms underlying these effects are discussed in view of the three-dimensional structure of MLCADH, which is presented in the accompanying paper [Lee et al. (1996) Biochemistry 35, 12412-12420].

The influence of co-overexpression of the bacterial chaperonins GroEL and GroES on solubility, tetramer formation and enzyme activity of three variants of heterologously-expressed human medium-chain acyl-CoA dehydrogenase (MCAD) was analysed in order to investigate the molecular mechanism underlying MCAD deficiency caused by the prevalent K304E mutation. Depending on which of the three amino acids--lysine (wild-type), glutamic acid (K304E) or glutamine (K304Q) are present at position 304 of the mature polypeptide, three different patterns were observed in our assay system: (i) solubility, tetramer formation and yield of enzyme activity of wild-type MCAD is largely independent of GroESL co-overexpression; (ii) the larger part of the K304Q mutant is insoluble without and solubility is enhanced with GroESL co-overexpression; solubility correlates with the amount of tetramer detected and the enzyme activity measured as observed for the wild-type protein. (iii) Solubility of the K304E mutant is in a similar fashion GroESL responsive as the K304Q mutant, but the amount of tetramer observed and the enzyme activity measured do not correlate with the amount of soluble K304E MCAD protein detected in Western blotting. In a first attempt to estimate the specific activity, we show that tetrameric K304E and K304Q mutant MCAD display a specific activity in the range of the wild-type enzyme. Taken together, our results strongly suggest, that the K304E mutation primarily impairs the rate of folding and subunit assembly. Based on the data presented, we propose that lysine-304 is important for the folding pathway and that an exchange of this amino acid both to glutamine or glutamic acid leads to an increased tendency to misfold/aggregate. Furthermore, exchange of lysine-304 with an amino acid with negative charge at position 304 (glutamic acid) but not with a neutral charge (glutamine) negatively affects conversion to active tetramers. A possible explanation for this latter effect--charge repulsion upon subunit docking--is discussed.