Partial Purification and Characterization of the Maize Mitochondrial Pyruvate Dehydrogenase Complex

Abstract

The pyruvate dehydrogenase complex was partially purified and characterized from etiolated maize (Zea mays L.) shoot mitochondria. Analysis by sodium dodecyl sulfate-polyacrylamide gel electrophoresis showed proteins of 40, 43, 52 to 53, and 62 to 63 kD. Immunoblot analyses identified these proteins as the E1β-, E1α-, E2-, and E3-subunits, respectively. The molecular mass of maize E2 is considerably smaller than that of other plant E2 subunits (76 kD). The activity of the maize mitochondrial complex has a pH optimum of 7.5 and a divalent cation requirement best satisfied by Mg²⁺. Michaelis constants for the substrates were 47, 3, 77, and 1 μm for pyruvate, coenzyme A (CoA), NAD⁺, and thiamine pyrophosphate, respectively. The products NADH and acetyl-CoA were competitive inhibitors with respect to NAD⁺ and CoA, and the inhibition constants were 15 and 47 μm, respectively. The complex was inactivated by phosphorylation and was reactivated after the removal of ATP and the addition of Mg²⁺.

The PDC catalyzes the oxidative decarboxylation of pyruvate to form acetyl-CoA and NADH. The PDC is composed of three fundamental enzymatic components: PDH (E1, EC 1.2.4.1), dihydrolipoyl transacetylase (E2, EC 2.3.1.12), and dihydrolipoamide dehydrogenase (E3, EC 1.8.1.4). The PDC from mammals (Gopalakrishnan et al., 1989), yeast (Behal et al., 1989), and perhaps plants (Taylor et al., 1992) contains an associated E3-binding protein. mtPDCs also contain two regulatory enzymes, PDH kinase and P-PDH phosphatase, which regulate PDC by reversible phosphorylation of the α-subunit of PDH (E1α). Mammalian and yeast mtPDC have a central pentagonal dodecahedryl core of E2-subunits to which the E1- and E3-subunits attach (Patel and Roche, 1990; Stoops et al., 1997). This E2 core comprises 20 trimers of a single polypeptide (Patel and Roche, 1990). Six to 12 E3 dimers, 6 to 12 E3-binding protein monomers, and 20 to 30 E1-α₂β₂ heterotetramers bind noncovalently to the E2 core (Patel and Roche, 1990).

The metabolic location of mtPDC and the irreversible nature of the reaction suggest that it is a site for regulation of mitochondrial carbon metabolism (Randall et al., 1996). All PDCs studied thus far are regulated by product inhibition (Patel and Roche, 1990; Luethy et al., 1996). In higher eukaryotes mtPDC activity is also regulated by reversible phosphorylation catalyzed by a PDH-specific protein kinase and a P-PDH-specific phosphatase (for review, see Patel and Roche, 1990; Randall et al., 1996).

The importance of mtPDC in controlling primary carbon metabolism is reflected by the many literature reports. However, there are a limited number of reports describing research on plant mtPDCs (for review, see Randall et al., 1996). Furthermore, our understanding of the regulation of plant mtPDC is derived from a limited number of C₃ species (e.g. pea, broccoli [Rubin and Randall, 1977], and castor bean [Rapp et al., 1987]).

In most C₃ species, leaf mtPDC is reversibly inactivated in the light in a photosynthesis- and photorespiration-dependent manner (Budde and Randall, 1990; Gemel and Randall, 1992). This is most likely the result of photorespiratory Gly metabolism that occurs in the leaf mitochondria during photosynthesis. Gly oxidation generates large amounts of NADH to support the necessary mitochondrial ATP production and NH₄⁺ to stimulate PDH kinase (Schuller et al., 1993). Consequently, mtPDC is negatively regulated as Gly oxidation increases, and all indications are that this light inactivation is caused by reversible phosphorylation of mtPDC. Pyruvate is the most effective inhibitor of mtPDC phosphorylation/inactivation (Schuller and Randall, 1990). In many C₄ species such as maize (Zea mays L.), pyruvate is a major metabolite in the photosynthetic CO₂-fixation process, and C₄ species lack significant photorespiration (Hatch, 1987). Therefore, light-dependent inactivation of mtPDC would not be expected. However, light-dependent inactivation of mtPDC was observed in maize leaves (Gemel and Randall, 1992), suggesting that the regulation of mtPDC may be different in maize and other C₄ plants.

To establish the properties and regulation of mtPDC in maize as a representative C₄ plant, we have undertaken a thorough examination of maize mtPDC beginning with nonphotosynthetic tissue to establish a baseline before proceeding to the characterization of the leaf mtPDCs, which will involve the two different cell types involved in C₄ photosynthesis. This report describes the partial purification and characterization of mtPDC from etiolated shoots of maize.

MATERIALS AND METHODS

Maize (Zea mays B73) seeds were obtained from the Illinois Seed Foundation (Urbana). Mitochondria were isolated from etiolated shoots as previously described (Hayes et al., 1991). Protein was quantified according to the method of Bradford (1976) using BSA as the standard. All other materials were from Sigma or Fisher Scientific.

Activity Assays

mtPDC activity was measured by monitoring NAD⁺ reduction at 340 nm (Randall et al., 1977) using a Response UV/visible spectrophotometer (Gilford, Oberlin, OH). The plastid marker TPI (EC 5.3.1.1) was assayed according to the method of Eisenthal and Danson (1992).

Electrophoresis and Immunoblot Analysis

SDS-PAGE, two-dimensional gel electrophoresis, and immunodetection of proteins bound to nitrocellulose membranes were performed as previously described (Luethy et al., 1995a). E1α monoclonal antibodies were raised against maize protein (Luethy et al., 1995a). E1β polyclonal antibodies were raised against recombinant Arabidopsis thaliana protein (M. Luethy, unpublished data). E3 polyclonal antibodies (generously provided by Dr. Steve Rawsthorne, John Innes Institute, Norwich, UK) were raised against the pea (Pisum sativum) L-protein of the GDC (Turner et al., 1992). E2-specific antibodies were affinity purified from total PDC antibodies raised against broccoli PDC (Randall et al., 1981) by incubating nitrocellulose-immobilized pea E2 with the antibodies and eluting as described by Smith and Fisher (1984).

Purification of Mitochondrial PDC

Purified mitochondria were resuspended in 30 mm Tes-KOH, pH 7.5, 2 mm DTT, lysed with a Polytron homogenizer (30 s at the 70% setting), and centrifuged for 15 min at 100,000g at 4°C in a rotor (model TL-100.3, Beckman) to remove membranes. The supernatant was subsequently centrifuged for 6 h at 400,000g. The resulting pellets were resuspended in a minimal volume of 30 mm Tes-KOH, pH 7.5, 2 mm DTT buffer, clarified by centrifugation at 13,000g for 15 min, and designated the 400K enzyme. The 400K enzyme (1 mL) was layered onto a 40-mL, 10 to 50% (v/v) linear glycerol gradient. The glycerol stock solutions contained 50 mm Tes-KOH, pH 7.5, 1.5 mm pyruvate, 1 mm MgCl₂, and 14 mm 2-mercaptoethanol. Gradients were centrifuged for 18 h at 25,000 rpm using an SW-28 rotor in an L8-55 ultracentrifuge (Beckman). Gradients were fractionated from the bottom, and the glycerol concentration was determined using a refractometer.

RESULTS

Plants are unique in that they contain a plastid PDC isoform (Williams and Randall, 1979; Camp and Randall, 1985) in addition to mtPDC. Therefore, to study the mitochondrial isoform, it was necessary to demonstrate that the purified mitochondria had low plastid contamination. Using TPI as the plastid marker enzyme, it was established that only 0.05% of the total TPI activity was present with the purified mitochondria.

Starting with 955 g fresh weight of etiolated maize shoots, 0.4 mg of highly enriched PDC was obtained, corresponding to 1.4% of the total mitochondrial protein (Table I). Almost 30% of total PDC activity was recovered, with a 21-fold enrichment. The specific activity of the partially purified maize mitochondrial PDC (0.81 μmol min⁻¹ mg⁻¹) was lower than values previously reported for purified cauliflower mtPDC (5.4 μmol min⁻¹ mg⁻¹, enriched approximately 100-fold; Randall et al., 1977) and purified broccoli mtPDC (6.3 μmol min⁻¹ mg⁻¹, enriched approximately 200-fold; Rubin and Randall, 1977). The peak of PDC activity consistently sedimented at 30% glycerol (Fig. 1A), which is similar to the peaks of other mtPDCs but larger than the peak of either plastid or Escherichia coli PDC (Camp and Randall, 1985). Compared with the sedimentation profiles of other mtPDCs, the molecular mass of the maize mtPDC was estimated at about 8000 to 9000 kD (Patel and Roche, 1990).

Table I.

Summarized purification of maize mtPDC

Fraction	Specific Activity	Total Activity	Enrichment	Yield	Protein
	μmol NADH formed min−1 mg−1 protein	μmol min−1	-fold	%	mg
Lysed mitochondria	0.039	1.2	1.0	100	30
100K Enzymea	0.12	1.0	3.1	83	8.5
400K Enzyme	0.16	0.72	4.1	60	4.4
Glycerol gradientb	0.81	0.35	21	29	0.43

^aSupernatant from 100,000g centrifugation. 

^bPooled mtPDC activity fractions from glycerol-gradient fractionation. 

Figure 1.

A, Fractionation profile for a typical rate-zonal glycerol gradient. Approximately 3 mg of 400K enzyme was loaded onto this gradient. Fractions 19 through 30 were pooled and concentrated for the SDS gel shown in Figure 2. B, Coomassie blue-stained SDS-PAGE of odd-numbered glycerol-gradient fractions 7 to 33. Positions of protein standards are indicated on the left (in kilodaltons). The position of the 110-kD protein is indicated on the right with an arrowhead. U, Units.

PDC Subunit Composition

SDS-PAGE analysis of the pooled mtPDC activity fractions from the glycerol gradient showed proteins at 40, 43, 52 to 53, 62 to 63, and 110 kD (Fig. 2, lane 4). Subunit-specific antibodies showed the enrichment of the putative mtPDC components through the purification (Fig. 2B), accounting for all of the major polypeptides observed in the gel except the 110-kD protein, which was probably a contaminant because it did not react with PDC antibodies and peaked at a higher point in the glycerol gradient (Fig. 1B).

Figure 2.

SDS-PAGE and corresponding immunoblots of lysed mitochondria (total mito), the supernatant from the 100,000g centrifugation (100K super), the 400K enzyme (400K pellet), and the pooled mtPDC activity fraction from the glycerol-gradient fractionation (glycerol). A, Coomassie blue-stained SDS-PAGE gel loaded with 5 μg of protein per lane. Positions of protein standards are indicated on the left and calculated molecular mass values of the predominant bands in the glycerol fraction are indicated on the right (in kilodaltons). B, Four replica protein blots of the fractions in A were probed with anti-subunit antibodies. One microgram of protein was loaded per lane. The molecular masses of the protein bands are indicated on the left (in kilodaltons).

Monoclonal antibodies to maize E1α recognized the 43-kD band in immunoblots (Fig. 2B). However, this single 43-kD band presented six isoelectric forms on immunoblots after two-dimensional IEF/SDS-PAGE separation (Fig. 3B). These multiple isoelectric forms may reflect the phosphorylation of multiple residues creating a gradient of phosphoproteins. Polyclonal antibodies to recombinant Arabidopsis E1β recognized a 40-kD polypeptide (Fig. 2B) highly enriched in the pooled glycerol-gradient fraction. In light of the strong evidence that PDC and GDC in pea mitochondria share identical E3 components (Bourguignon et al., 1996), antibodies raised against the pea L-protein of GDC were used to probe maize mtPDC. These antibodies recognized a 62- to 63-kD doublet that was enriched in the glycerol-gradient-purified fraction (Fig. 2B). Antibodies that recognize the 76-kD pea E2-subunit reacted with a 52- to 53-kD maize doublet that was enriched throughout the PDC purification (Fig. 2B).

Figure 3.

Two-dimensional gel electrophoresis of glycerol-gradient-enriched maize mtPDC and corresponding immunoblots. A, Ten micrograms of glycerol-gradient-enriched mtPDC was resolved by IEF in the first dimension followed by SDS-PAGE. B and C, One microgram of protein was resolved as in A, transferred to nitrocellulose, and probed with antibodies (Ab) to the E1α- and E1β-subunits. The molecular masses of the protein bands are indicated on the left (in kilodaltons).

The identity of the 52- to 53-kD doublet was further established by microsequencing the N terminus of these two proteins. The N-terminal amino acid sequence of the 52-kD protein (Fig. 4) had the highest similarity to a mammalian dihydrolipoamide transacetylase (E2), according to BLAST (Altschul et al., 1990), an amino acid alignment algorithm. N-terminal sequencing of the 53-kD protein revealed that it is related to the 52-kD protein (Fig. 4). Aligning the N-terminal sequence of the maize 52-kD protein with yeast (Niu et al., 1988), Arabidopsis (Guan et al., 1995), and human (Coppel et al., 1988) deduced E2 amino acid sequences showed the highest similarity within the lipoyl domains.

Figure 4.

Amino acid alignment of N-terminal amino acid sequences for maize 52- and 53-kD proteins and deduced amino acid sequences for yeast, Arabidopsis, and human E2-subunits. The number of amino acid residues (aa) before the homologous region is indicated to the left of the sequences. Shading indicates an identical amino acid. X indicates a cycle of Edman degradation for which no determination was made.

Reaction Requirements and Kinetic Properties

Maize mtPDC activity showed a sharp optimum at pH 7.5, similar to that for pea mtPDC (Miernyk and Randall, 1987), but lower than that for plastid PDC (pH 8.2; Williams and Randall, 1979). Maize mtPDC activity was sensitive to high ionic strength, similar to porcine PDC (Pawelczyk et al., 1992), with buffer concentrations higher than 75 mm and NaCl concentrations greater than 50 mm reducing mtPDC activity; NaCl, KCl, and NH₄Cl all gave similar patterns of inhibition.

Maize mtPDC required CoA, NAD⁺, thiamine pyrophosphate, and divalent cations for activity and did not use NADP⁺. The 400K enzyme specifically decarboxylated pyruvate and exhibited only minor activity with 2-oxobutyrate (15%), 3-hydroxypyruvate (6%), and 3-hydroxybutyrate (5%). No activity was seen with the branched-chain keto acids 2-oxoisovalerate and 2-oxoisocaproate, or with 2-oxoglutarate. The dialyzed maize 400K enzyme had an absolute requirement for divalent cations (Fig. 5), with Mg²⁺, Mn²⁺, and Ca²⁺ all able to restore activity.

Figure 5.

Divalent cation requirement for the 400K enzyme. The 400K enzyme was dialyzed for 2 h in 2 L of 2 mm EDTA and 2 mm EGTA to remove endogenous cations, and then dialyzed twice in 2 L of 20 mm Tes, pH 7.5, and 0.5 mm DTT to remove the chelators. Enzyme was added to an assay vessel that contained divalent cations and necessary components. Rates are expressed as relative percentages of the maximum rate (0.093 μmol min⁻¹ mg⁻¹).

K_m values were determined under optimal conditions of pH and saturating concentrations of nonvariable substrates. Most of the K_m values for maize mtPDC were in the range of those reported for other plant PDCs (Table II). The exception was TPP; its K_m value was 10-fold higher than that of pea. This may explain why the maize complex, unlike the pea complex (Miernyk and Randall, 1987), required exogenous TPP for activity. The K_i for NADH was 5-fold lower than the K_m for NAD⁺, whereas the K_i for acetyl-CoA was much higher than the K_m for CoA, suggesting that NADH could be a more potent product inhibitor (Table II). Both products were competitive inhibitors with respect to their substrates (data not shown).

Table II.

K_m for PDCs

Plant Tissue	Km				Ki
	Pyr	NAD	CoA	MgTPP	NADHa	AcCoAb
	μm
MtPDCs
Etiolated maize shoots	47 ± 3	77 ± 18	3 ± 2	1 ± 0.2	15 ± 6	47 ± 5
Pea leafc	57	122	4	0.08	18	10
Broccoli floretd	250	110	6	n.d.e	13	19
Plastid PDC
Pea leaff	120	36	10	n.d.	9	16

The maize 400K enzyme was used in this study. Values are the mean of at least three separate preparations ± sd.

^a[CoASH] was 20 μm; all other substrates and cofactors were saturating. 

^b[NAD] was 100 μm; all other substrates and cofactors were saturating. 

^cMiernyk and Randall (1987). 

^dRubin and Randall (1977). 

^en.d., Not determined. 

^fCamp et al. (1988). 

Intermediates of the Krebs cycle, amino acids, and polyamines had little effect on the activity of the 400K enzyme when tested at 2 mm. Hydroxypyruvate, previously shown to be a noncompetitive inhibitor of PDC (Randall et al., 1977), inhibited maize PDC by 36%. Ali et al. (1993) reported that the E1α-subunit of mammalian PDC contained an essential Cys (Cys-62 in humans). Mutation of this Cys to Ala or derivatization by sulfhydryl reagents completely inactivated the mammalian enzyme. Because this essential Cys is not present in prokaryotic or plastid (Johnston et al., 1997) forms of PDC, we determined the effect of sulfhydryl reagents on maize mtPDC. Mersalyl, p-hydroxymercuribenzoate, and N-ethylmaleimide rapidly inactivated the maize mtPDC.

ATP-Dependent Inactivation

The 400K enzyme was almost completely inactivated in 6 min with 200 μm MgATP, but after 15 min the complex began gradually reactivating (Fig. 6A). To determine if this reactivation was caused by the Mg²⁺-requiring P-PDH phosphatase activity, the 400K enzyme was inactivated with 200 μm ATP, the excess ATP was removed with hexokinase plus Glc, the sample was divided into four aliquots, and EDTA, MgCl₂, or buffer was added (Fig. 6B). EDTA prevented the gradual reactivation observed with the control, whereas 10 mm Mg²⁺ stimulated an 85% recovery of activity in 20 min, suggesting that the Mg²⁺-dependent reactivation is likely the result of a P-PDH phosphatase. Incubation of the 400K enzyme with [γ-³²P]ATP labeled a 43-kD protein that was recognized by monoclonal antibodies to the maize E1α-subunit (data not shown).

Figure 6.

A, ATP-dependent inactivation of the 400K enzyme. Equimolar amounts of Mg²⁺ and ATP were added to PDC preparations to the final concentrations indicated. The control did not have any MgATP added. One-hundred-microgram samples of enzyme were removed at various time intervals and assayed for activity. Activity is expressed as a percentage of the control (0.16 μmol NADH formed min⁻¹ mg⁻¹ protein) at time 0. B, The effect of Mg²⁺ on reactivation of P-PDC. The 400K enzyme was incubated with 200 μm MgATP until the inactivation of mtPDC ceased. Free ATP was then removed with 2.5 units of hexokinase (HK) and 2 mm Glc at 30 min. The 400K enzyme was then divided into four aliquots to which EDTA, MgCl₂, or buffer (control) was added to the final concentrations indicated.

DISCUSSION

Rate-zonal density-gradient centrifugation of maize mitochondrial matrix protein yielded a distinct peak of PDC activity with a specific activity ranging from 1.2 to 1.7 units mg⁻¹ protein (Fig. 1A), and the predominant polypeptides in this peak were identified as the E1β-, E1α-, E2-, and E3-subunits. Additionally, a 110-kD Coomassie blue-stained polypeptide was observed; however, this protein did not peak with PDC peak activity (Fig. 1B) and was not recognized by any PDC antibodies. Polypeptides of 50 and 55 kD were also observed in heavily loaded lanes; however, the 55-kD polypeptide also did not peak with PDC peak activity. The unidentified 50-kD protein was probably not PDH kinase or P-PDH phosphatase, since glycerol-gradient-enriched mtPDC lacked these activities. Alternatively, it could have been an E3-binding protein homolog, which is also a 50-kD protein in yeast (Behal et al., 1989).

A comparison of the apparent and calculated sizes of plant mtPDC subunits shows that only the maize E1α-subunit is identical in size to other plant E1α-subunits, whereas the E1β- and E3-subunits are 2 to 3 kD larger (Table III). In contrast, the maize E2-subunit (52 kD) is much smaller than other plant E2-subunits (76 kD), which can be explained by its variable multidomain structure.

Table III.

Estimated and deduced molecular mass values of mtPDC catalytic subunits

Molecular Mass
SDS-PAGE			Calculated/deduced
Subunit	Maize	Pea	Arabidopsis	Pea	Potato
	kD
E1α	43	43a	43.0b	43.5c	43.2d
E1β	40	38a	39.2e	38.7c	n.a.
E2	52 –53	76a	60.0f	n.a.	n.a.
E3	62 –63	57a,60g,h	n.a.i	50.4h	n.a.

Values are based on SDS-PAGE analysis of isolated mitochondria and therefore represent processed proteins. Molecular mass values deduced from sequenced cDNA clones represent precursor proteins (i.e. targeting peptide plus mature protein).

^aLuethy et al., 1995a. 

^bLuethy et al., 1995b. 

^cM.H. Luethy, unpublished data. 

^dGrof et al., 1995. 

^eLuethy, 1994. 

^fGuan et al. 

^gTurner et al., 1992. 

^hBourguignon et al., 1992. 

ⁱn.a., Data not available. 

The E2-subunit possesses a multidomain structure, with a lipoyl domain(s) connected by flexible linkers to the E1-/E3-binding domains followed by the catalytic domain (Reed and Hackert, 1990; Perham, 1991). The flexible lipoyl domains allow active-site coupling between the E1- and E3-subunits for the following series of reactions. The E1 reductively acylates the covalently bound lipoate within the E2 lipoyl domain. The E2-subunit catalyzes the acyl-transfer step to CoA, and E3 catalyzes the reoxidation of the dihydrolipoyl moiety using NAD⁺ as the electron acceptor. In addition to having a catalytic role, the inner (second) lipoyl domain of mammalian E2 binds the kinase and phosphatase regulatory components (Liu et al., 1995; Chen et al., 1996).

The number of E2 lipoyl domains found in nature is variable (Reed and Hackert, 1990). Multiple, tandemly repeated lipoyl domains have been observed in the E2-subunits described previously for all organisms except bacilli and yeast (Perham, 1991). Although the tandemly repeated lipoyl domains are functional (Allen et al., 1989), only one is required for E2 catalytic or complex function (Guest et al., 1985; Machado et al., 1992). The considerably smaller size of the maize E2 can be explained if only one lipoyl domain is present. A single lipoyl domain may be attributable to either novel gene structure or proteolysis. Proteolysis is unlikely, since we observed only the 52- to 53-kD band for E2 with purified mitochondria lysed directly into SDS-PAGE sample buffer followed by boiling (data not shown).

Support for the presence of a single lipoyl domain is found in the reduced size of the maize E2, the N-terminal amino acid sequence (which is most similar to the single E2 lipoyl domain from yeast), and its similarity to the inner lipoyl domain of Arabidopsis and human E2. The properties of the maize E2 are consistent with previous findings, i.e. a single lipoyl domain is sufficient for complex activity and appears to be sufficient for binding the kinase and phosphatase, although maybe not as tightly as with the mammalian complex, since both the kinase and the phosphatase can be stripped away during the glycerol-gradient purification. Molecular analysis shows that the E2 from Arabidopsis has multiple lipoyl domains (Guan et al., 1995), so it will be interesting to determine if this is true for other plant species or if single lipoyl domains are characteristic of maize alone.

The pyruvate and CoA K_ms for maize mtPDC are similar to those from pea (Miernyk and Randall 1987), but unlike the pea mtPDC the maize complex has a lower K_m for NAD⁺ and a higher K_m for MgTPP, which may reflect differences in the E3 and E1 components. The high K_i for acetyl-CoA in relation to other plant PDCs may reflect a different in vivo environment for the maize mtPDC or the various functions of the maize mitochondria.

The maize mtPDC requires divalent cations for catalytic activity. Of the divalent cations tested, Mg²⁺ best satisfied this requirement, as determined by the V_max/K_m ratio for the three cations Mg²⁺ (2.3), Mn²⁺ (1.3), and Ca²⁺ (1.0). The K_m for Mg²⁺ was approximately 40 μm, considerably lower than that of pea mtPDC (360 μm; Miernyk and Randall, 1987) and pea plastid PDC (1 mm; Camp and Randall, 1985), suggesting that the divalent cation requirement does not have regulatory significance in maize. All plant PDCs except mtPDC from cauliflower (Randall et al., 1977) will accept Mn²⁺ and Ca²⁺ as Mg²⁺ substitutes.

The maize mtPDC is capable of regulation by reversible phosphorylation. Increasing amounts of MgATP completely inactivated mtPDC, although reactivation immediately ensued, indicating that PDC activity reflects the relative activities of the regulatory kinase and phosphatase. MgATP concentrations below saturation will not entirely inactivate mtPDC, even after extended periods. This can be explained by contaminating ATPase activity, high P-PDH phosphatase activity, multiple phosphorylation sites that coordinate full inactivation, or all of the above.

In summary, we have partially purified PDC from maize mitochondria and identified the catalytic subunits by immunoblot analysis. The molecular masses of the maize PDC subunits are similar to those of other plant PDCs, the exception being the E2-subunit, which was 23 kD smaller than pea E2. Overall, the kinetic properties of maize mtPDC were similar to those of other plant mtPDCs, although slight differences were observed with regard to the divalent cation and TPP requirement, as well as the product inhibitor acetyl-CoA. The degree of similarity between maize mtPDC and C₃-plant mtPDCs was somewhat surprising considering the differences in pyruvate metabolism.

Abbreviations:

400K enzyme

the pellet from the 400,000g centrifugation

GDC

Gly decarboxylase complex

mtPDC

mitochondrial pyruvate dehydrogenase complex

PDC

pyruvate dehydrogenase complex

PDH

pyruvate dehydrogenase

P-PDH

phosphopyruvate dehydrogenase

TPI

triose phosphate isomerase

TPP

thiamine pyrophosphate