Gel-filtration, ultracentrifugation and sucrose density gradient centrifugation demonstrated differences in physico-chemical properties of holoenzyme and apoenzyme of lipoamide dehydrogenase. The native apoenzyme has a mol.wt. of approx. 52,000 which is half that of the native holoenzyme. The DCIP-active enzyme formed directly after reconstitution of apoenzyme with FAD, is still a monomer. Light-scattering experiments with holoenzyme at low protein concentrations (< 0.1 mg/ml) indicated a dissociation of the enzyme related to a decrease in lipoate activity. These results were not consistent with previously proposed models for the enzyme structure. The idea of two interchain disulfide bridges, which moreover take part in the catalysis, is replaced by a real protein association-dissociation model in which the two catalytically important disulfide bridges are intrachain ones. The dimerisation constants derived from light-scattering data and from the return of the lipoate activity in recombination studies at relatively high apoenzyme concentrations agree with each other (3-6x10 6
l mole -1
The influence of urea on the catalytic activities of different conformations of the enzyme has been studied, Recombined DCIP-active enzyme, and both the frozen holoenzyme and frozen Cu 2+
-modified are very urea-sensitive while the Cu 2+
-modified enzyme is sensitive at concentrations>4 M, The holoenzyme is very stable in 8 M urea except when the protein concentration is low (0.1 mg/ml). The apoenzyme is thermo-labile and also urea-sensitive; the stability is protein concentration-dependent. The apoenzyme population is inhomogeneous as indicated electrophoretically,
The FAD binding to the apoenzyme induces a series of protein conformational changes, before dimerisation occurs.
The K ass
(FAD) value for the protein (2-3x10 5
l mole -1
) is considerably lower than in the holoenzyme. The dimerisation is dependent on pH (7.2-7.5) and ionic strength (0.2 M) and is promoted by elevated temperatures. The return of the lipoate activity is due to dimerisation though even complete restoration of this activity does not prove that the native holoenzyme structure is regained. The formed dimer still has initially a high DCIP-activity and it is partially sensitive to ureaand FMN-treatment over a period of at least 24 hrs all in contrast to the holoenzyme.
Some of the flavin derivatives studied are able to restore the monomer- activity with DCIP viz
. F8-BrAD (K ass
l mole -1
), 3-methyl-FAD (K ass
l mole -1
) and 3-carboxymethylFAD (K ass
l mole -1
). Only 3-methylFAD partially restores the lipoate activity. The flavin binding is due to multiple binding forces as shown by the interference of related compounds with the binding site, FMN which is a competitive inhibitor of the FAD binding induces, like FAD, protein conformational changes as the K i
value increases with time. FMN-derivatives generally show a noncompetitive inhibition pattern with respect to the DCIP-activity. Parts other than the isoalloxazine moiety of the flavin molecule viz
, the adenine part and pyrophosphate, are also involved in the FAD binding: NADH, NAD +
, ADP, adenine, ATP and pyrophosphate all inhibit the flavin binding. All nucleotides which inhibit the flavin binding affect the dimerisation except NAD +
which promotes the return of the lipoate activity.
The "ping pong bi bi" mechanism previously proposed to be the reaction mechanism of lipoamide dehydrogenase is not the correct model, The involvement of a ternary complex has been proposed instead. As deviations from linearity are observed in the L-B plots with lip(SH) 2
and NAD +
asvariable substrates depending on the level of the second substrate a prefered order mechanism is postulated according to FERDINAND (1966). NAD +
is the substrate which is prefered as first one (K D
≈110 μM at 25°).
A second binding site with a higher affinity for NAD +
than the catalytical important one (K D
≈20-30 μM) prevents 4-electron reduction of the flavin by NADH. It is likely that this site is the NADH binding site in the transhydrogenase reaction. The prod-act inhibition by lipS 2
(NH) with respect to lip(SH) 2
is complicated and the type of inhibition is dependent on
the level of both substrates, Under conditions where the mechanism has become random, the inhibition of lipS 2
) is competitive, At lower NAD +
levels this inhibition is noncompetitive with respect to lip(SH) 2
Kinetical evidence is presented in the case of the native holoenzyme and the Cu 2+
-modified enzyme that the oxidised form exists in different temperature-dependent conformations. Analysis of the sets of rate constants revealed that this conformational. change is clearly reflected in the NAD +
/NADH binding properties.
Antisera against four different structures of lipoamide dehydrogenase have been succesfully prepared, viz
. against the native holoenzyme and the Cu 2+
-modified enzyme (both dimers) and against the apoenzyme and the reconstituted, DCIP-active enzyme (both monomers). The inhibition of the enzyme activity proved to be the most sensitive method to detect a reaction between antibodies and the different antigenic structures. The homologous combinations are the most specifically inhibited but always less than 100 %. The antiserum against the apoenzyme does almost not interfere with the activities of the different enzyme conformations. The antisera against the dimeric enzyme forms are closer to each other than to the one formed against the DCIP-active monomer. Inhibition patterns of the homologous Cu 2+
-enzyme/antiserum combination are noncompetitive.
The Cu 2+
-modified enzyme has many structural properties in common with the holoenzyme. NAD +
in low concentrations activates the DCIP-activity and promotes the linearity in the assay which is generally non-linear. The recombination phenomena are very similar and the K ass
value of FAD almost identical (2-5x10 5
l mole -1
) to the normal case. The non-linearity of the assay is probably a property of the dimer structure.
Freezing of diluted holoenzyme promotes the DCIP-activity while the lipoate activity drops. These changes are protein concentration-dependent. Li +
, Cs +
-ions are protecting in 0.1 M concentration against these changes but Na +
, K +
and Rb +
do not. However, a combination of K +
and Na +
ratio 7) is as protective as the Na +
combination of the same ratio. When the frozen enzyme is thawed and the decrease of the DCIP-activity and the return of the lipoate activity are followed, the structural changes induced by freezing pass at least through one intermediate as the decrease of the DCIP-activity follows a second-order reaction rate on ice. The order of the reaction decreases and approaches 1 at elevated temperatures. The freezing process is very complicated indeed. Results with polyacrylamide electrophoresis strongly favour the existence of two different lipoamide dehydrogenases isolated in a mixture from the pyruvate dehydrogenase complex and the α-ketoglutarate dehydrogenase complex. Freezing of the enzyme changes mainly the electrophoretic patterns of lipoamide dehydrogenase obtained from the pyruvate dehydrogenase complex.