Chemical xenobiotics and mitochondrial autoantigens in primary biliary cirrhosis: identification of antibodies against a common environmental, cosmetic, and food additive, 2-octynoic acid

Emerging evidence has suggested environmental factors as causative agents in the pathogenesis of primary biliary cirrhosis (PBC). We have hypothesized that in PBC the lipoyl domain of the immunodominant E2 component of pyruvate dehydrogenase (PDC-E2) is replaced by a chemical xenobiotic mimic, which is sufficient to break self-tolerance. To address this hypothesis, based upon our quantitative structure-activity relationship data, a total of 107 potential xenobiotic mimics were coupled to the lysine residue of the immunodominant 15 amino acid peptide of the PDC-E2 inner lipoyl domain and spotted on microarray slides. Sera from patients with PBC (n = 47), primary sclerosing cholangitis (n = 15), and healthy volunteers (n = 20) were assayed for Ig reactivity. PBC sera were subsequently absorbed with native lipoylated PDC-E2 peptide or a xenobiotically modified PDC-E2 peptide, and the remaining reactivity analyzed. Of the 107 xenobiotics, 33 had a significantly higher IgG reactivity against PBC sera compared with control sera. In addition, 9 of those 33 compounds were more reactive than the native lipoylated peptide. Following absorption, 8 of the 9 compounds demonstrated cross-reactivity with lipoic acid. One compound, 2-octynoic acid, was unique in both its quantitative structure-activity relationship analysis and reactivity. PBC patient sera demonstrated high Ig reactivity against 2-octynoic acid-PDC-E2 peptide. Not only does 2-octynoic acid have the potential to modify PDC-E2 in vivo but importantly it was/is widely used in the environment including perfumes, lipstick, and many common food flavorings.     

Immunization with a xenobiotic 6-bromohexanoate bovine serum albumin conjugate induces antimitochondrial antibodies

The E2 subunit of pyruvate dehydrogenase complex (PDC-E2) is the major autoantigen recognized by antimitochondrial Abs (AMA) in primary biliary cirrhosis (PBC). Recently, we replaced the lipoic acid moiety of PDC-E2 with a battery of synthetic structures designed to mimic a xenobiotically modified lipoyl hapten on a 12-aa peptide that was found within the immunodominant autoepitope of PDC-E2 and demonstrated that AMA in PBC reacted against several organic modified mimotopes as well as, or sometimes significantly better than, the native lipoyl domain. Based on this data, we immunized rabbits with one such xenobiotic organic compound, 6-bromohexanoate, coupled to BSA. One hundred percent of immunized rabbits developed AMA that have each and every characteristic of human AMAs with reactivity against PDC-E2, E2 subunit of branched chain 2-oxo-acid dehydrogenase, and E2 subunit of 2-oxoglutarate dehydrogenase complex. The rabbit AMA also inhibited enzymatic function of PDC-E2 and, importantly, binds to peptide sequences not present in the xenobiotic carrier immunogen. In contrast, BSA-immunized controls did not produce such activity. Our observation that animals immunized with a xenobiotic BSA complex produce autoantibodies that react not only with the xenobiotic, but also with mitochondrial autoantigens recognized by autoimmune PBC sera, suggests that environmental xenobiotic agents can be a risk factor for the induction of PBC.     

Comparative epitope mapping of murine monoclonal and human autoantibodies to human PDH-E2, the major mitochondrial autoantigen of primary biliary cirrhosis.

Immunization with recombinant human pyruvate dehydrogenase (PDH)-E2, the major autoantigen of primary biliary cirrhosis, readily induces a vigorous murine antibody response but does not generate hepatic disease. To determine the fine specificity of this response, 18 mAb were generated from three strains of mice and the reactive epitopes mapped. An initial examination of mAb suggested that they behaved similarly to the antimitochondrial autoantibodies in primary biliary cirrhosis (PBC) because i) all polyclonal antisera and 2 of 18 mAb reacted with all species of mammalian PDH-E2 examined including mouse PDH-E2, ii) 15 of 18 mAb inhibited PDH enzyme function, and iii) the reactivity of mAb toward rPDH-E2 were blocked by PBC sera. However, fine examination of the reactive sequences of the PDH-E2 complex revealed that antibodies identical to those in PBC patients were not produced by experimental immunization. In contrast to PBC, none of the mAb or murine polyclonal sera were able to react with protein X, a lipoic acid-containing component of the PDH complex previously shown to cross-react with PDH-E2 when probed with PBC sera. Although the epitopes for 12 mAb were localized within the inner lipoyl domain, none reacted with mouse PDH-E2 or cross-reacted with the outer lipoyl domain as observed in PBC. In addition, the epitopes of the two mAb which did react with all mammalian species of mitochondria were not localized within the PBC epitope. These findings indicate the highly immunogenic nature of the inner lipoyl domain of PDH-E2. The inability to elicit antibodies of the same specificity in mice, considered together with the highly localized autoantibody response in humans, suggests that antimitochondrial autoantibodies are most likely the result of specific breakdown of tolerance to a unique autoepitope.     

Phagotopes derived by antibody screening of phage-displayed random peptide libraries vary in immunoreactivity: studies using an exemplary monoclonal antibody, CII-C1, to type II collagen

Antibody screening of phage-displayed random peptide libraries to identify mimotopes of conformational epitopes is promising. However, because interpretations can be difficult, an exemplary system has been used in the present study to investigate whether variation in the peptide sequences of selected phagotopes corresponded with variation in immunoreactivity. The phagotopes, derived using a well-characterized monoclonal antibody, CII-C1, to a known conformational epitope on type II collagen, C1, were tested by direct and inhibition ELISA for reactivity with CII-C1. A multiple sequence alignment algorithm, PILEUP, was used to sort the peptides expressed by the phagotopes into clusters. A model was prepared of the C1 epitope on type II collagen. The 12 selected phagotopes reacted with CII-C1 by both direct ELISA (titres from < 100-11 200) and inhibition ELISA (20-100% inhibition); the reactivity varied according to the peptide sequence and assay format. The differences in reactivity between the phagotopes were mostly in accord with the alignment, by PILEUP, of the peptide sequences. The finding that the phagotopes functionally mimicked the C1 epitope on collagen was validated in that amino acids RRL at the amino terminal of many of the peptides were topographically demonstrable on the model of the C1 epitope. Notably, one phagotope that expressed the widely divergent peptide C-IAPKRHNSA-C also mimicked the C1 epitope, as judged by reactivity in each of the assays used: these included cross-inhibition of CII-C1 reactivity with each of the other phagotopes and inhibition by a synthetic peptide corresponding to that expressed by the most frequently selected phagotope, RRLPFGSQM. Thus, it has been demonstrated that multiple phage-displayed peptides can mimic the same epitope and that observed immunoreactivity of selected phagotopes with the selecting mAb can depend on the primary sequence of the expressed peptide and also on the assay format used.     

Structural requirement for autoreactivity on human pyruvate dehydrogenase-E2, the major autoantigen of primary biliary cirrhosis. Implication for a conformational autoepitope.

The E2 component (acetyltransferase) of the pyruvate dehydrogenase (PDH) complex is the major mitochondrial autoantigen recognized by autoantibodies in patients with primary biliary cirrhosis (PBC). Previous work, using only a partial length rat liver cDNA clone of PDH-E2, demonstrated that the immunodominant epitope was localized to the lipoic acid binding site. Human PDH-E2, in contrast to rat PDH-E2, has two lipoic acid binding sites. By using a full length human cDNA for PDH-E2, and by preparation of multiple overlapping recombinant fragments, we have determined that three autoreactive determinants are present on human PDH-E2: two cross-reactive lipoyl domains, and an area surrounding the E1/E3 binding region. The dominant epitope was localized to the inner lipoyl domain whereas the outer lipoyl domain only showed a weak cross-reactivity, and only 1/26 PBC sera reacted weakly to the E1/E3 binding region area. By probing recombinant fusion proteins expressed from small restriction fragments of the inner lipoyl domain, we have found that a minimum of 75 amino acids (residues 146-221) were required for detectable autoantibody binding, and that 93 amino acids (residues 128-221) were necessary for characteristically strong antimitochondrial autoantibody recognition. Such a requirement for a large region suggests the possibility that a conformational autoepitope may be recognized. In addition, we have found that absorption of PBC sera with the purified mammalian PDH complex does not remove reactivity against Escherichia coli Ag. The possible implications for such results are discussed.     

A lipoyl synthetic octadecapeptide of dihydrolipoamide acetyltransferase specifically recognized by anti-M2 autoantibodies in primary biliary cirrhosis.

Close to 95% of patients with established clinical, biochemical and histologic features of primary biliary cirrhosis (PBC) possess antimitochondrial M2 antibodies reacting with the E2 component, dihydrolipoamide acetyltransferase, of the pyruvate dehydrogenase complex. We examined the ability of synthetic peptides of E2 to be recognized in ELISA by sera from patients with PBC and autoimmune-related disorders. Sera from 14 PBC M2+ patients, 1 PBC M2- patient, 5 non-PBC M2+ patients, and 6 patients with chronic active hepatitis were studied. Among the seven E2 synthetic peptides tested (namely peptides 87-119, 167-184, 169-202, 267-302, 456-477, 498-513 and 530-543), only peptide 167-184 used as OVA conjugate and prepared with lipoic acid (LA) located on lysine 173 (natural inner lipoyl-binding site) was recognized in direct ELISA by PBC M2+ sera. The conjugated peptide 167-184 LA was not recognized in direct ELISA by non-PBC M2+ sera or by sera from patients with chronic active hepatitis. The free peptide 167-184 LA inhibited the ELISA reaction of PBC antibodies to PDH and totally abolished the typical immunofluorescence reaction of PBC sera on rat kidney, stomach and liver, or human HEp-2 cell substrates. No inhibition of ELISA or immunofluorescence reaction was found with the other E2 fragments including peptide 167-184 without LA. Our results show that the lipoyl moiety forms an integral part of a dominant conformational epitope recognized by PBC sera. Inasmuch as the peptide 167-184 LA was not recognized by non-PBC sera in direct ELISA, it could be used as a valuable probe for PBC diagnosis.     

Immunoreactivity of organic mimeotopes of the E2 component of pyruvate dehydrogenase: connecting xenobiotics with primary biliary cirrhosis

In primary biliary cirrhosis (PBC), the major autoepitope recognized by both T and B cells is the inner lipoyl domain of the E2 component of pyruvate dehydrogenase. To address the hypothesis that PBC is induced by xenobiotic exposure, we took advantage of ab initio quantum chemistry and synthesized the inner lipoyl domain of E2 component of pyruvate dehydrogenase, replacing the lipoic acid moiety with synthetic structures designed to mimic a xenobiotically modified lipoyl hapten, and we quantitated the reactivity of these structures with sera from PBC patients. Interestingly, antimitochondrial Abs from all seropositive patients with PBC, but no controls, reacted against 3 of the 18 organic modified autoepitopes significantly better than to the native domain. By structural analysis, the features that correlated with autoantibody binding included synthetic domain peptides with a halide or methyl halide in the meta or para position containing no strong hydrogen bond accepting groups on the phenyl ring of the lysine substituents, and synthetic domain peptides with a relatively low rotation barrier about the linkage bond. Many chemicals including pharmaceuticals and household detergents have the potential to form such halogenated derivatives as metabolites. These data reflect the first time that an organic compound has been shown to serve as a mimeotope for an autoantigen and further provide evidence for a potential mechanism by which environmental organic compounds may cause PBC.     

Rat liver mitochondria contain two immunologically distinct dihydrolipoamide dehydrogenases

We have raised antisera against dihydrolipoamide dehydrogenase. One antigen was isolated from purified bovine kidney pyruvate dehydrogenase complex (PDC). The other antigen was a commercial preparation of porcine heart dihydrolipoamide dehydrogenase (E3) which did not first involve purification of the alpha-keto acid dehydrogenase complex(es). Both antibody preparations cross-reacted with the E3 components of PDC, alpha-ketoglutarate dehydrogenase complex, and branched-chain keto acid dehydrogenase complex. This demonstrates the immunological identity of the E3 components. These sera totally precipitated E3 activity from the purified complexes, from purified preparations of E3, and from extracts of rat heart and kidney mitochondria. The two sera vary in their reaction with rat liver mitochondrial extracts: the anti PDC-E3 serum left residual E3 activity (approximately 50% of the original) that was precipitable by the anti-E3 anti-serum. This indicates that liver contains two immunologically distinct forms of E3. Metabolic assays measuring the differential effects of the two sera on the glycine decarboxylation reaction suggest that the form which is immunologically nonreactive with the anti-PDC-E3 serum could represent the E3 involved in the glycine cleavage system.     

Apoptosis-associated antigens recognized by autoantibodies in patients with the autoimmune liver disease primary biliary cirrhosis

There is growing evidence that the onset of autoimmune disorders can be linked to the inefficient removal of apoptotic cells. Since defects in the elimination of apoptotic cells lead to secondary necrosis and subsequent release of intracellular components, this might explain the generation of autoantibodies against intracellular antigens. Accordingly, we wanted to investigate, whether antibodies from patients with the autoimmune liver disease primary biliary cirrhosis (PBC) recognize self-proteins generated and released during apoptosis. Using Western blot analyses we could detect intracellular antigens with serum IgG from PBC patients but not with serum IgG from healthy donors in lysates of Jurkat T-leukemia, HepG2 hepatoma, and HT-29 colon-carcinoma cells. Interestingly, PBC serum IgG also recognized caspase substrates in cells undergoing apoptosis induced by staurosporine or TRAIL (TNF-related apoptosis inducing ligand). In addition to intracellular antigens, serum IgG from PBC patients detected caspase-dependent antigens in the supernatants of apoptotic (secondary necrotic) cells and antigens on the surface of apoptotic Jurkat cells. Among the caspase substrates recognized by PBC serum IgG we could identify the components PDC-E2 and -E1β of the known autoantigen PDC (pyruvate dehydrogenase complex). Thus, caspase-mediated processing of intracellular proteins might generate de novo autoantigens that upon release contribute to the generation of autoantibodies and autoimmune diseases as PBC. Christoph Peter Berg and Gerburg Maria Stein contributed equally to this paper and share first authorship. Sebastian Wesselberg and Kirsten Lauber share equal senior authorship.     

Substrate-induced structural changes of the pyruvate dehydrogenase multienzyme complex

The time course of the overall reaction catalyzed by the pyruvate dehydrogenase multienzyme complex produces an unexpectedly high lag (tau = 8 S) even in the presence of saturating concentrations of its substrates. The preincubation of the pyruvate dehydrogenase complex with one of the substrates alone decreases the duration of this lag, and all the substrates of the pyruvate dehydrogenase component (E1) and dihydrolipoyl transacetylase component (E2) together (pyruvate, thiamine pyrophosphate, and CoA) result in the complete disappearance of the lag. The reduction of the dihydrolipoyl dehydrogenase component (E3) of the pyruvate dehydrogenase complex with the substrates of the complex in the absence of NAD+ produces significantly different quenching in the FAD fluorescence, and then the reduction with the substrates of E3 as dihydrolipoic acid and dithioerythritol. (The formation of FADH2 was not observed in the system.) The higher fluorescence quenching in the presence of substrates of pyruvate dehydrogenase complex compared to the effect caused by the substrates of the E3 component (dihydrolipoic acid and DTE) indicates conformational changes additionally manifested in the fluorescence properties of the enzyme complex. The substrate-induced quenching of the enzyme-bound FAD fluorescence shows biphasic kinetics. The rate constant of the slow phase is comparable with the rate constant calculated from the time duration of the lag phase observed in the overall reaction. The kinetic analysis of both intensity and anisotropy decrease of the FAD fluorescence suggests a consecutive transmittance of an all substrate-coordinated, induced conformational changes directed from the pyruvate dehydrogenase-via the lipoyl transacetylase--to the lipoyl dehydrogenase. Two simultaneous conformational effects caused by binding of the substrates can be distinguished; one of them results the fluorescence of the bound FAD to be more quenched, while the other makes the FAD more mobile. The first-order rate constants of both these conformational changes were determined. The present observations suggest that the pyruvate dehydrogenase complex exists in a partially inactive state in the absence of its substrates, and it becomes active due to conformational changes caused by the binding of its substrates.     

Genetic reconstruction and functional analysis of the repeating lipoyl domains in the pyruvate dehydrogenase multienzyme complex of Escherichia coli

The dihydrolipoamide acetyltransferase component (E2p) of the pyruvate dehydrogenase complex of Escherichia coli contains three highly homologous sequences of about 100 residues that are tandemly repeated to form the N-terminal half of the polypeptide chain. All three sequences include a lysine residue that is a site for lipoylation and they appear to form independently folded functional domains. These lipoyl domains are in turn linked to a much larger (about 300 residues) subunit-binding domain of the E2p chain that aggregates to form the octahedral inner core of the complex and also contains the acetyltransferase active site. In order to investigate whether individual lipoyl domains play different parts in the enzymic mechanism, selective deletions were made in vitro in the dihydrolipoamide acetyltransferase gene (aceF) so as to excise one or two of the repeating sequences. This was facilitated by the high degree of homology in these sequences, which allowed the creation of hybrid lipoyl domains that closely resemble the originals. Pyruvate dehydrogenase complexes incorporating these genetically reconstructed E2p components were purified and their structures were confirmed. It was found that the overall catalytic activity, the system of active site coupling, and the ability to complement pyruvate dehydrogenase complex mutants, were not significantly affected by the loss of one or even two lipoyl domains per E2p chain. No special role can be attached thus far to individual lipoyl domains. On the other hand, certain genetic deletions affecting the acetyltransferase domain caused inactivation of the complex, highlighting particularly sensitive areas of that part of the E2p chain.     

Immunogenetic analysis reveals that epitope shifting occurs during B-cell affinity maturation in primary biliary cirrhosis

Primary biliary cirrhosis (PBC) is a liver disease characterized by serum autoantibodies against the pyruvate dehydrogenase complex (PDC) located in the inner mitochondrial membrane. The predominant target in PDC has previously been localized to the inner lipoyl domain (ILD) of the E2 subunit. The etiology of PBC is unknown, although molecular mimicry with bacterial PDC has been proposed. Here, we have investigated the etiology of PBC and nature of the autoimmune response by analyzing the structure of a human monoclonal antibody with ILD specificity. Mutants of the monoclonal antibody, which was originally isolated from a patient with PBC, were expressed as Fab by phage display, and tested for reactivity against recombinant domains of the E2 subunit. Fab in which the V(H)-encoded portions were reverted to germline lost reactivity against the ILD alone, but recognized a different epitope in a didomain construct encompassing the ILD, hinge region and E1/E3 binding domain. The complete V(H) and V(L )germline revertant was unreactive with the human ILD and didomain, the Escherichia coli didomain, and whole PDC. We hypothesize that the IgM on the surface of the na?ve B-cell first recognizes an as yet unidentified antigen, and that accumulation of somatic mutations results in an intermolecular epitope shift directed towards an epitope involving the E1/E3 binding domain. Further mutations result in the specificity being redirected to the ILD. These findings also suggest that bacterial molecular mimicry is not involved in initiating disease.     

A general strategy to identify mimotopes of pathological antigens using only random peptide libraries and human sera.

A strategy to identify disease-specific epitopes from phage-displayed random peptide libraries using human sera is described. Peptides on phage (phagotopes) that react with antibodies present in patient sera are purified from > 10(7) different sequences by affinity selection and immunological screening of plaques. Disease-specific phagotopes can be identified out of this pool through an 'antigen independent' procedure which avails itself only of patient and normal human sera. Using this strategy, we have selected antigenic mimics (mimotopes) of two different epitopes from the human hepatitis B virus envelope protein (HBsAg). We could show that a humoral response to these mimotopes is widespread in the immunized population, suggesting that the strategy identifies phagotopes that have a potential role as diagnostic reagents. Immunization of mice with the selected phagotopes elicited a strong specific response against the HBsAg. These results open new inroads into disease-related epitope discovery and provide the potential for vaccine development without a requirement for the use of, or even information about, the aetiological agent or its antigens.     

Structure function studies on the lipoate-acetyltransferase--component-X-core assembly of the ox heart pyruvate dehydrogenase complex

Component X, the recently recognised subunit of mammalian pyruvate dehydrogenase complex, was shown by immune blotting to be present in all of nine tissues dissected from rat. This finding indicated that component X was not an isoenzyme of the lipoate acetyltransferase (E2) associated with one or a limited number of tissues. Native pyruvate dehydrogenase complex was shown to bind IgG raised to isolated component X, indicating that there were at least some regions of the X subunit exposed at the periphery of the complex. Lipoyl groups of ox heart pyruvate dehydrogenase complex were specifically cross-linked by reaction with phenylene-o-bismaleimide in the presence of pyruvate and the subunits contributing to the products of cross-linking were identified by immune blotting. Species with very high Mr containing both E2 and component X, were formed in high yield, as well as apparent E2/E2 and E2/X dimers and trimers and an X/X dimer. These results showed that acetylated lipoyl groups of different E2 and X subunits were able to interact in all possible combinations. The types of cross-linked E2 products formed suggested that two thiols, reactible with phenylene-o-bismaleimide, were rapidly generated in the presence of pyruvate. The results were most easily explained by the presence of two acetylatable lipoyl groups on each E2 polypeptide.     

Isolation of human anti-branched chain alpha-oxo acid dehydrogenase-E2 recombinant antibodies by Ig repertoire cloning in idiopathic dilated cardiomyopathy

The generation of human monoclonal autoantibodies is critical for understanding humoral immune response in autoimmunity. In this study, Ig gene repertoire cloning was performed from a regional lymph node of a patient with idiopathic dilated cardiomyopathy (IDCM), and the resulting combinatorial IgG library was screened with bovine branched chain alpha-oxo acid dehydrogenase-E-2 (BCOADC-E2), one of the autoantigens in IDCM. After three rounds of affinity selection, we isolated three human recombinant IgG Fab molecules, named BC1, BC2 and BC3, that specifically react with BCOADC-E2 by ELISA. Interestingly, BC2 showed weak cross-reactivity to pyruvate dehydrogenase complex-E2 (PDC-E2), another mitochondrial autoantigen found in primary biliary cirrhosis (PBC), and their kappa light chain genes have 95% homology with a light chain of the human anti-DNA antibody. Although the exact pathogenic effect of anti-BCOADC-E2 autoantibodies is still unknown in IDCM, the potential binding specificity and limited light chain gene usage of our recombinant IgG molecules may shed light on the initial mechanism as to how autoantibodies start developing in IDCM.     

Kinetics and specificity of reductive acylation of lipoyl domains from 2-oxo acid dehydrogenase multienzyme complexes

Lipoamide and a peptide, Thr-Val-Glu-Gly-Asp-Lys-Ala-Ser-Met-Glu lipoylated on the N6-amino group of the lysine residue, were tested as substrates for reductive acetylation by the pyruvate decarboxylase (E1p) component of the pyruvate dehydrogenase multienzyme complex of Escherichia coli. The peptide has the same amino acid sequence as that surrounding the three lipoyllysine residues in the lipoate acetyltransferase (E2p) component of the native enzyme complex. Lipoamide was shown to be a very poor substrate, with a Km much higher than 4 mM and a value of kcat/Km of 1.5 M-1.s-1. Under similar conditions, the three E2p lipoyl domains, excised from the pyruvate dehydrogenase complex by treatment with Staphylococcus aureus V8 proteinase, could be reductively acetylated by E1p much more readily, with a typical Km of approximately 26 microM and a typical kcat of approximately 0.8 s-1. The value of kcat/Km for the lipoyl domains, approximately 3.0 x 10(4) M-1.s-1, is about 20,000 times higher than that for lipoamide as a substrate. This indicates the great improvement in the effectiveness of lipoic acid as a substrate for E1p that accompanies the attachment of the lipoyl group to a protein domain. The free E2o lipoyl domain was similarly found to be capable of being reductively succinylated by the 2-oxoglutarate decarboxylase (E1o) component of the 2-oxoglutarate dehydrogenase complex of E. coli. The 2-oxo acid dehydrogenase complexes are specific for their particular 2-oxo acid substrates. The specificity of the E1 components was found to extend also to the lipoyl domains.(ABSTRACT TRUNCATED AT 250 WORDS)     

Domain structure and 1H-n.m.r. spectroscopy of the pyruvate dehydrogenase complex of Bacillus stearothermophilus.

The pyruvate dehydrogenase complex of Bacillus stearothermophilus was treated with Staphylococcus aureus V8 proteinase, causing cleavage of the dihydrolipoamide acetyltransferase polypeptide chain (apparent Mr 57 000), inhibition of the enzymic activity and disassembly of the complex. Fragments of the dihydrolipoamide acetyltransferase chains with apparent Mr 28 000, which contained the acetyltransferase activity, remained assembled as a particle ascribed the role of an inner core of the complex. The lipoic acid residue of each dihydrolipoamide acetyltransferase chain was found as part of a small but stable domain that, unlike free lipoamide, was able still to function as a substrate for reductive acetylation by pyruvate in the presence of intact enzyme complex or isolated pyruvate dehydrogenase (lipoamide) component. The lipoyl domain was acidic and had an apparent Mr of 6500 (by sedimentation equilibrium), 7800 (by sodium dodecyl sulphate/polyacrylamide-gel electrophoresis) and 10 000 and 20 400 (by gel filtration in the presence and in the absence respectively of 6M-guanidinium chloride). 1H-n.m.r. spectroscopy of the dihydrolipoamide acetyltransferase inner core demonstrated that it did not contain the segments of highly mobile polypeptide chain found in the pyruvate dehydrogenase complex. 1H-n.m.r. spectroscopy of the lipoyl domain demonstrated that it had a stable and defined tertiary structure. From these and other experiments, a model of the dihydrolipoamide acetyltransferase chain is proposed in which the small, folded, lipoyl domain comprises the N-terminal region, and the large, folded, core-forming domain that contains the acetyltransferase active site comprises the C-terminal region. These two regions are separated by a third segment of the chain, which includes a substantial region of polypeptide chain that enjoys high conformational mobility and facilitates movement of the lipoyl domain between the various active sites in the enzyme complex.     

Distinct modes of recognition of the lipoyl domain as substrate by the E1 and E3 components of the pyruvate dehydrogenase multienzyme complex

Two-dimensional (15)N-heteronuclear single-quantum coherence (HSQC) NMR studies with a di-domain (lipoyl domain+ linker+ peripheral subunit-binding domain) of the dihydrolipoyl acetyltransferase (E2) component of the pyruvate dehydrogenase complex of Bacillus stearothermophilus allowed a molecular comparison of the need for lipoic acid to be covalently attached to the lipoyl domain in order to undergo reductive acetylation by the pyruvate decarboxylase (E1) component, in contrast with the ability of free lipoic acid to serve as substrate for the dihydrolipoyl dehydrogenase (E3) component. Tethering the lipoyl domain to the peripheral subunit-binding domain in a complex with E1 or E3 rendered the system more like the native enzyme complex, compared with the use of a free lipoyl domain, yet of a size still amenable to investigation by NMR spectroscopy. Recognition of the tethered lipoyl domain by E1 was found to be ensured by intensive interaction with the lipoyl-lysine-containing beta-turn and with residues in the protruding loop close to the beta-turn. The size and sequence of this loop varies significantly between species and dictates the lipoylated lipoyl domain as the true substrate for E1. In contrast, with E3 the main interaction sites on the tethered lipoyl domain were revealed as residues Asp41 and Ala43, which form a conserved sequence motif, DKA, around the lipoyl-lysine residue. No domain specificity is observed at this step and substrate channelling in the complex thus rests on the recognition of the lipoyl domain by the first enzyme, E1. The cofactor, thiamine diphosphate, and substrate, pyruvate, had distinct but contrasting effects on the E1/di-domain interaction, whereas NAD(+) and NADH had negligible effect on the E3/di-domain interaction. Tethering the lipoyl domain did not significantly change the nature of its interaction with E1 compared with a free lipoyl domain, indicative of the conformational freedom allowed by the linker in the movement of the lipoyl domain between active sites.     

Solution Structure of the Lipoyl Domain of the 2-Oxoglutarate Dehydrogenase Complex fromAzotobacter vinelandii

The three-dimensional solution structure of the lipoyl domain of the 2-oxoglutarate dehydrogenase complex fromAzotobacter vinelandiihas been determined from nuclear magnetic resonance data by using distance geometry and dynamical simulated annealing refinement. The structure determination is based on a total of 580 experimentally derived distance constraints and 65 dihedral angle constraints. The solution structure is represented by an ensemble of 25 structures with an average root-mean-square deviation between the individual structures of the ensemble and the mean coordinates of 0.71 Å for backbone atoms and 1.08 Å for all heavy atoms. The overall fold of the lipoyl domain is that of a β-barrel-sandwich hybrid. It consists of two almost parallel four-stranded anti-parallel β-sheets formed around a well-defined hydrophobic core, with a central position of the single tryptophan 21. The lipoylation site, lysine 42, is found in a β-turn at the far end of one of the sheets, and is close in space to a solvent-exposed loop comprising residues 7 to 15. The lipoyl domain displays a remarkable internal symmetry that projects one β-sheet onto the other β-sheet after rotation of approximately 180° about a 2-fold rotational symmetry axis. There is close structural similarity between the structure of this 2-oxoglutarate dehydrogenase complex lipoyl domain and the structures of the lipoyl domains of pyruvate dehydrogenase complexes fromBacillus stearothermophilusandEscherichia coli, and conformational differences occur primarily in a solvent-exposed loop close in space to the lipoylation site. The lipoyl domain structure is discussed in relation to the process of molecular recognition of lipoyl domains by their parent 2-oxo acid dehydrogenase.     

Crystal structure of pyruvate dehydrogenase kinase 3 bound to lipoyl domain 2 of human pyruvate dehydrogenase complex

The human pyruvate dehydrogenase complex (PDC) is regulated by reversible phosphorylation by four isoforms of pyruvate dehydrogenase kinase (PDK). PDKs phosphorylate serine residues in the dehydrogenase (E1p) component of PDC, but their amino-acid sequences are unrelated to eukaryotic Ser/Thr/Tyr protein kinases. PDK3 binds to the inner lipoyl domains (L2) from the 60-meric transacetylase (E2p) core of PDC, with concomitant stimulated kinase activity. Here, we present crystal structures of the PDK3-L2 complex with and without bound ADP or ATP. These structures disclose that the C-terminal tail from one subunit of PDK3 dimer constitutes an integral part of the lipoyl-binding pocket in the N-terminal domain of the opposing subunit. The two swapped C-terminal tails promote conformational changes in active-site clefts of both PDK3 subunits, resulting in largely disordered ATP lids in the ADP-bound form. Our structural and biochemical data suggest that L2 binding stimulates PDK3 activity by disrupting the ATP lid, which otherwise traps ADP, to remove product inhibition exerted by this nucleotide. We hypothesize that this allosteric mechanism accounts, in part, for E2p-augmented PDK3 activity.