Interchain acetyl transfer in the E2 component of bacterial pyruvate dehydrogenase suggests a model with different roles for each chain in a trimer of the homooligomeric component

Article abstract of DOI:10.1021/bi201614n

The bacterial pyruvate dehydrogenase complex carries out conversion of pyruvate to acetyl-coenzyme A with the assistance of thiamin diphosphate (ThDP), several other cofactors, and three principal protein components, E1-E3, each present in multiple copies. The E2 component forms the core of the complexes, each copy consisting of variable numbers of lipoyl domains (LDs, lipoic acid covalently amidated at a lysine residue), peripheral subunit binding domains (PSBDs), and catalytic (or core) domains (CDs). The reaction starts with a ThDP-dependent decarboxylation on E1 to an enamine/C2α carbanion, followed by oxidation and acetyl transfer to form S-acetyldihydrolipoamide E2, and then transfer of this acetyl group from the LD to coenzyme A on the CD. The dihydrolipoamide E2 is finally reoxidized by the E3 component. This report investigates whether the acetyl group is passed from the LD to the CD in an intra- or interchain reaction. Using an Escherichia coli E2 component having a single LD, two types of constructs were prepared: one with a Lys to Ala substitution in the LD at the Lys carrying the lipoic acid, making E2 incompetent toward post-translational ligation of lipoic acid and, hence, toward reductive acetylation, and the other in which the His believed to catalyze the transthiolacetylation in the CD is substituted with A or C, the absence of His rendering it incompetent toward acetyl-CoA formation. Both kinetic evidence and mass spectrometric evidence support interchain transfer of the acetyl groups, providing a novel model for the presence of multiples of three chains in all E2 components, and their assembly in bacterial enzymes.

Full text of DOI:10.1021/bi201614n

Article
pubs.acs.org/biochemistry
Interchain Acetyl Transfer in the E2 Component of Bacterial Pyruvate
Dehydrogenase Suggests a Model with Different Roles for Each
Chain in a Trimer of the Homooligomeric Component
Jaeyoung Song and Frank Jordan*
Department of Chemistry, Rutgers University, Newark, New Jersey 07102, United States
S
* Supporting Information
ABSTRACT: The bacterial pyruvate dehydrogenase complex carries out conversion of
pyruvate to acetyl-coenzyme A with the assistance of thiamin diphosphate (ThDP),
several other cofactors, and three principal protein components, E1−E3, each present in
multiple copies. The E2 component forms the core of the complexes, each copy
consisting of variable numbers of lipoyl domains (LDs, lipoic acid covalently amidated
at a lysine residue), peripheral subunit binding domains (PSBDs), and catalytic (or
core) domains (CDs). The reaction starts with a ThDP-dependent decarboxylation on
E1 to an enamine/C2α carbanion, followed by oxidation and acetyl transfer to form
̃
S-acetyldihydrolipoamide E2, and then transfer of this acetyl group from the LD to
coenzyme A on the CD. The dihydrolipoamide E2 is finally reoxidized by the E3
component. This report investigates whether the acetyl group is passed from the LD to the CD in an intra- or interchain reaction.
Using an Escherichia coli E2 component having a single LD, two types of constructs were prepared: one with a Lys to Ala
substitution in the LD at the Lys carrying the lipoic acid, making E2 incompetent toward post-translational ligation of lipoic acid
and, hence, toward reductive acetylation, and the other in which the His believed to catalyze the transthiolacetylation in the CD
is substituted with A or C, the absence of His rendering it incompetent toward acetyl-CoA formation. Both kinetic evidence and
mass spectrometric evidence support interchain transfer of the acetyl groups, providing a novel model for the presence of
multiples of three chains in all E2 components, and their assembly in bacterial enzymes.
he bacterial pyruvate dehydrogenase complex (PDHc)
We here address the question of whether the acetyl group is
Tcarries out conversion of pyruvate to acetyl-coenzyme A passed from the LD to the CD in an intra- or interchain
reaction by using an E. coli E2 component designed with only a
single LD (1-lip E2ec), rather than the three LDs in the wild-
type enzyme (3-lip E2ec). Earlier, it was shown that the activi-
ties of 1-lip E2ec and 3-lip E2ec are virtually the same, while
1-lip E2ec offers advantages for mechanistic studies.17−19
Two types of constructs of 1-lip E2ec were prepared to test
our hypothesis: one in which the lysine on the LD ordinarily
carrying the lipoic acid is changed to alanine (K41A) and two
others in which the histidine believed to catalyze the trans-
thiolacetylation in the CD is substituted with A or C (H399C
or H399A, respectively). The first variant is incompetent
toward post-translational ligation of the lipoic acid and, hence,
toward reductive acetylation. The other two variants are
potentially incompetent toward acetyl-CoA formation by virtue
of the absence of the catalytic histidine residue. These con-
structs then allow us to conduct a complementation experi-
ment: should the acetyl transfer reaction proceed in an intra-
chain fashion, the two types of constructs should each be
inactive either together or individually; however, should acetyl
transfer proceed in an interchain fashion, addition of the two
types of constructs should produce measurable activity. As a
with the assistance of three principal protein components, E1−
E31 (Scheme 1). In the mammalian enzyme,2 there are three
additional components mostly responsible for regulation: a
kinase, a phosphatase, and an E3-binding protein (E3BP).3,4
Each component exists in multiple copies with total molecular
masses in the range of 4.5−10 MDa. The multidomain E2
component forms the core of the complexes [see the domain
structure for E2 from Escherichia coli (E2ec) in Scheme 2]. It
consists of variable numbers1−3 of LDs to which the cofactor
lipoic acid is covalently attached in a post-translational reaction,
a PSBD, and a CD, where acetyl-CoA is produced.1 Although
the number of copies of icosahedral mammalian E2 is still
controversial, there is agreement that the sum (E2 + E3BP) is
60, a multiple of three,5−7 while there are 24 copies of the
octahedral E. coli E2 (E2ec), again a multiple of three.1 For the
E. coli PDHc, a stoichiometry of 12 E1ec dimers, 8 E2ec
trimers, and 6 E3ec dimers was proposed.8−12 All 24 copies of
E2ec have the LD and the CD, and it is plausible that interchain
acyl transfer would take place between an LD of one chain and
the CD of a different one, with the mobility of the LD afforded
by flexible Ala- and Pro-rich linkers.8−11,13−16 The comment in
ref 12 that, “A morphological unit consisting of three subunits
appears to be important in the assembly of both types of
polyhedral forms” (icosahedral or octahedral), prompted us to
explore potential mechanistic consequences of this observation.
Received: October 24, 2011
Revised: February 13, 2012
Published: March 13, 2012
© 2012 American Chemical Society
2795
dx.doi.org/10.1021/bi201614n | Biochemistry 2012, 51, 2795−2803

Biochemistry
Article
Scheme 1. Mechanism of the Pyruvate Dehydrogenase Complex
Scheme 2. Domain Architecture of Wild-Type E2ec
of E2ec variants is illustrated in Figure 1. We confirmed the
results of the kinetics experiments by ascertaining acetyl-CoA
formation from CoA using mass spectrometric analysis. The
results suggest a plausible model for utilization of multiples of
three chains present in E2 components, and for their assembly
in the bacterial class of such complexes.
EXPERIMENTAL PROCEDURES
■
Materials. Thiamin diphosphate (ThDP), NAD+, dithio-
threitol (DTT), isopropyl β-D-1-thiogalactopyranoside (IPTG),
micrococcal nuclease, deoxyribonuclease I, acetylcoenzyme A
(acetyl-CoA), phenylmethanesulfonyl fluoride (PMSF), and
control, the K41A/H399C doubly substituted variant was also
constructed, carrying neither a lipoylation site nor the active
center His. The experimental design for this complementation
Figure 1. Complementation experiments designed to test for interchain acetyl transfer on E2ec. (a) Scheme for monitoring intersubunit acetyl
transfer between K41A E2ec and H399C E2ec. (b) Control experiment with the doubly substituted variant (K41A/H399C E2ec). Both the
lipoylation site and the transthiolesterification site were modified simultaneously.
2796
dx.doi.org/10.1021/bi201614n | Biochemistry 2012, 51, 2795−2803

Biochemistry
Article
coenzyme A (CoA) were from USB (Cleveland, OH). E. coli
BL21(DE3) cells were from Novagen (EMD Chemicals,
Gibbstown, NJ). Lysozyme, formic acid, and methanol as the solvent
of mass spectrometry were from Sigma-Aldrich (St. Louis, MO).
Protease inhibitor cocktail tablets supplied in glass vials were from
Roche Applied Science (Indianapolis, IN)
Plasmid Purification and Site-Directed Mutagenesis.
The Wizard Plus Minipreps DNA purification system was used
for purification of DNA (Promega, Madison, WI). The Quik-
Change site-directed mutagenesis kit was used for single-site
substitution (Stratagene, La Jolla, CA). Mutagenic primers were
from IDTdna (Coralville, IA). The following primers were used
to create the K41A, H399A, and H399C variants of E2ec
(substituted bases underlined) according to methods published
by this laboratory:18,19 K41A, 5′ GATCACCGTAGAAGGC-
GACGCAGCTTCTATGGAAGTTCCG 3′; H399A, 5′
CTCTCTCCTTCGACGCCCGCGTGATCGACG 3′;
H399C, 5′ TCTCTCTCCTTCGACTGCCGCGTGATC-
GACGG 3′.
Creation of the K41A/H339C Doubly Substituted
Variant. For a control experiment, the K41A/H399C doubly
substituted variant was created. The plasmid of H399C was
purified and used for the substitution of K41 with alanine
(using the primer from above) to create K41A/H399C, which
carries neither the lipoylation site nor the catalytic histidine.
Protein Expression and Purification. The procedure for
expression and purification of 1-lip E2ec was modified from our
previous publication.18,19 A single colony from a freshly plated
cell bank was used to grow the inoculums at 37 °C overnight
(five or six culture tubes each with 10 mL of medium and
50 μg/mL ampicillin). Subsequently, it was inoculated in five or
six shaker flasks containing 500 mL of LB and 50 μg/mL am-
picillin fortified with 0.1 mM DL-α-lipoic acid and incubated
at 37 °C and 250 rpm until the A650 reached 0.6−0.8. To the
culture was then added 1.0 mM IPTG, and the mixture was
incubated for an additional 4 h. The cells were harvested by
centrifugation (2200g for 5 min) and washed with 20 mM
KH2PO4 (pH 7.5) containing 0.10 M NaCl and 0.25 mM
EDTA. The washed pellet could be stored below −20 °C until
further use. Cells were thawed and resuspended in 50 mL of
20 mM KH2PO4 buffer (pH 7.5) containing one protease
inhibitor cocktail tablet, which also supplies 2 mM EDTA and
0.1 M NaCl. Lysozyme (0.60 mg/mL) and 1 mM PMSF were
added, and the cell suspension was incubated for 20 min on ice
with intermittent stirring. Cells were then sonicated to disrupt
the membrane (total time of 4 min, 10 s on and 10 s off); the
lysate was fractionated with (NH4)SO4 (0−25% and 25−50%),
and the 25−50% pellet fraction was collected. After
fractionation with (NH4)SO4, the pellet was dissolved in
20 mM KH2PO4 dialysis buffer (pH 7.5) containing 1 mM
EDTA, 1 mM benzamidine hydrochloride, 1 mM DTT, and
0.2 NaCl and dialyzed against 2 L of dialysis buffer. Dialyzed
protein was centrifuged at 28978g for 20 min and treated with
1000 units of DNase, 500 units of nuclease, and 5 mM MgCl2
and incubated for 1 h on ice followed by centrifugation at
29000g for 20 min. Protein was applied to HiPrep 26 × 60
Sephacryl S-300 high-resolution gel-filtration column pre-
equilibrated with 3 column volumes of 20 mM KH2PO4
column buffer (pH 7.2) containing 0.5 mM EDTA, 1 mM
benzamidine hydrochloride, 1 mM DTT, and 0.2 M NaCl at a
flow rate of 1.5 mL/min, and the column was run overnight in
the cold room at a rate of 0.2 mL/min. Protein was monitored
at 280 nm and by sodium dodecyl sulfate−polyacrylamide gel
electrophoresis (SDS−PAGE). Pure fractions were combined
and centrifuged for 4 h at 121000g, and then the pellet was
resuspended in 1.0 mL of column buffer containing an addi-
tional 0.30 M NaCl (total of 0.50 M) and kept on ice in the
cold room for 15 h. Undissolved protein was removed by cen-
trifugation (29000g for 20 min), and the final purity was judged
by SDS−PAGE.
Measurement of PDHc Activity. The overall activity of
the PDHc consisting of 1-lip E2ec and its variants was mea-
sured after reconstitution to form a complex with indepen-
dently expressed E1ec and E3ec components at 25 °C, and the
wild-type E1ec:E2ec:E3ec complex mass ratio was 1:1:1. First,
the E2−E3 subcomplex was assembled by preincubating E2ec
(100 μg) and E3ec (100 μg) in 1 mL of 20 mM KH2PO4
(pH 7.5) for 1 h. Next, E1 (10 μg) was added to the mixture of
20 μg of the E2ec−E3ec subcomplex (from the preincubation)
in 200 μL of 20 mM KH2PO4 (pH 7.5), and the mixture was
incubated for 10 min. The ovarall activity was measured in
980 μL of reaction medium containing 100 mM Tris-HCl
(pH 8.0), 1 mM MgCl2, 2 mM sodium pyruvate, 2.5 mM
NAD+, 0.2 mM ThDP, and 2.6 mM DTT, initiating the reac-
tion by addition of 20 μL of CoA (final concentration of
100 μM) and 3 μg of the E1−E2−E3 complex. The overall
activities of K41A E2ec, H399A E2ec, and H399C E2ec were
monitored with the same mass ratio that was used for the wild-
type complex. The pyruvate-dependent reduction of NAD+ was
monitored at 340 nm on a Varian DMS 300 spectropho-
tometer.
To test for interchain acetyl transfer, K41A E2ec and H399C
E2ec were incubated at various mass ratios for 30 min prior
to the reconstitution with E1ec, E3ec, and pyruvate. Because
K41A E2ec carries a catalytic His, which may produce overall
activity by the transfer of an acetyl group from lipoyllysine on
H399C E2ec, the concentration of K41A was fixed. The K41A
E2ec:H399C E2ec mass ratios for incubation were 1:1, 1:1.5,
and 1:2. After incubation for 30 min in 20 mM KH2PO4 (pH
7.0), E1ec and E3ec were introduced into this mixture with a
1:1:1 E1ec:K41A E2ec, H399C E2ec:E3ec mass ratio. In a
control experiment, the doubly substituted variant, K41A/
H399C E2ec, was used, and the overall activity was again mea-
sured upon reconstitution of the complex with E1ec, E3ec, and
pyruvate. Then, K41A E2ec and K41A/H399C E2ec were mix-
ed and incubated at the same mass ratio used for K41A and
H399C above. The concentration of K41A was fixed again to
K41A:K41A/H399C ratios of 1:1, 1:1.5, and 1:2.
To test whether interchain acetyl transfer takes place via E2
chain communication within one (holo)PDHc molecule or
communication between E2 chains located on different PDHc
molecules, two PDH complexes were prepared, one from K41A
E2ec and the other from H399C E2ec. These two types of
complexes were then mixed in a 1:1 molar ratio, and NADH
production was monitored for 30 min.
Mass Spectrometric Analysis of Acetyl-CoA Formation.
As an independent measure of the complementation ability of
the K41A and H399C E2ec variants, the formation of acetyl-
CoA from CoA, starting the reaction with pyruvate, was
assessed using an Apex-ultra 7.0 T 70 hybrid FTMS from
Bruker Daltonics (Billerica, MA) with an ESI source. Both
positive and negative ion modes were tested, and the positive
ion mode provided better separation and resolution for acetyl-
CoA. For detecting acetyl-CoA formation in the overall PDHc
reaction, we used the mass ratios of wild-type and variant E2
2797
dx.doi.org/10.1021/bi201614n | Biochemistry 2012, 51, 2795−2803

Biochemistry
Article
Table 1. Overall PDHc Activities of E2ec Variants with
a
Various Mass Ratios
overall
E1ec:E2ec:E3ec
mass ratio in
overall reaction
PDHc
activity
b
type
single
variant
E2ec variant
Lys41Ala
(%)
1:1:1
2.5 0.2
His399Cys
His399Ala
1:1:1
1:1:1
1:1:1
2.8 0.2
5.6 0.3
0.9 0.1
Lys41Ala/His399Cys
(doubly substituted)
mixture of Lys41Ala:His399Cys
1:(1:1):1
10.3 0.2
two
(separate variants)
variants
1:(1:1.5):1
1:(1:2):1
1:(1:1):1
18.3 0.3
22.7 0.3
2.1 0.1
Lys41Ala:Lys41Ala/
His399Cys (singly and
doubly substituted)
1:(1:1.5):1
1:(1:2):1
3.0 0.1
3.0 0.1
a
The production of NADH was monitored at 340 nm. All experiments
were performed at 25 °C; each experiment performed more than 20
b
times. The full activity of PDHc assembled from the E1, E2, and E3
components is in the range of 14−17 μmol of NADH produced min−1
(mg of complex)−1.
components as used in the activity measurements described
above.
Sample Preparation for Fourier Transfer Mass
Spectrometry Experiments. All biochemical experiments
were conducted at 25 °C. For the experiment with 1-lip E2ec, a
mixture of 100 μg each of E2ec and E3ec was first incubated in
1 mL of 20 mM KH2PO4 (pH 7.0) for 1 h. Subsequently, 1 μg
of E1ec was added to the mixture of 2 μg of the E2ec−E3ec
subcomplex (from the previous preincubation) in 980 μL of
0.1 M Tris-HCl (pH 8.0) containing 1 mM MgCl2, 2 mM sodium
pyruvate, 2.5 mM NAD+, 0.2 mM ThDP, and 2.6 mM DTT,
and the mixture was incubated for 10 min. The reaction was
initiated by adding CoA (100 μM), and the mixture was
incubated for 10 min. After 10 min, the reaction was quenched
by adding a solution of 12.5% trichloroacetic acid in 1 M HCl.
[It is important to note that this is not a kinetic experiment;
rather, we are measuring the amount of acetyl-CoA formed
under the same conditions by different E2 constructs or their
mixtures during a fixed time. The reaction in fact is over in
1 min according to the steady state assay, after which the
absorbance no longer increases.] Next, protein was removed by
centrifugation at 16000g for 20 min, and then 50 μL of the
clarified sample was mixed with 50 μL of MS running solvent
containing 0.1% (v/v) formic acid and 50% MeOH (v/v) in
deionized water. Samples were then injected into the ESI-
FTMS. Twenty scans were acquired.
Quantitative Analysis of Acetyl-CoA Production Using
the FTMS. To quantify acetyl-CoA production by the variants,
a standard curve was created at a variety of acetyl-CoA concen-
trations. Samples were prepared as described above.
Standard Curve for Acetyl-CoA Determination using
the ESI-Detected FTMS. A sample (50 μL) was mixed with
50 μL of MS running solvent containing 0.1% (v/v) formic acid
and 50% MeOH (v/v) in deionized water for the FTMS,
yielding final concentrations of 1, 5, 10, 20, 30, 40, 50, 100,
and 200 μM. These samples were then injected into the FTMS
with a syringe pump (rate of 2 μL/min). Twenty scans were
Figure 2. Activity assay for E2ec variants reconstituted with E1ec and
E3ec. The top panel shows data for mixtures of K41A E2ec and
H399C E2ec incubated at various mass ratios; then E1ec and E3ec
were added, allowing use of the NADH assay for the overall activity of
the complex. In the bottom panel, a fixed concentration of K41A E2ec
was incubated with increasing concentrations of K41A/H399C E2ec.
Next, E1ec and E3ec were added to assess PDHc activity.
acquired. After data had been acquired, Sigmaplot was used to
calibrate and plot the data.
RESULTS AND DISCUSSION
■
Substitutions of lipoyl-bearing lysine and the putative catalytic
histidine affected the overall PDHc activity. The overall PDHc
activity of K41A E2ec was diminished to baseline, while
surprisingly, that of H399A E2ec was not completely abolished
(Table 1). In contrast to that of H399A E2ec, the overall activ-
ity of H399C E2ec was reduced to baseline. The activity was
never reduced to zero even with impairment of the lipoylation
site and the catalytic His. This could be the result of some
endogenous wild-type E2 component present in the E. coli cells
(the gene for E2 was not knocked out); however, recombinant
E2ec variants were all overproduced, and the endogenous
2798
dx.doi.org/10.1021/bi201614n | Biochemistry 2012, 51, 2795−2803

Biochemistry
Article
Figure 3. Acetyl-CoA produced by E2ec variants detected by the FTMS (left). Acetyl-CoA was observed at m/z 810.13 in different reactions from
top to bottom; 1-lip E2ec, H399A E2ec, H399C E2ec, and K41A E2ec. Acetyl-CoA produced by complementation of K41A E2ec and H399C E2ec
(right). The amount of acetyl-CoA detected at m/z 810.13 is clearly enhanced by complementation of a fixed K41A E2ec concentration with an
increasing concentration of H399C E2ec. From top to bottom: K41A, H399C, and a mixture of K41A and H399C at the indicated mass ratios of 1:1,
1:1.5, 1:2, 1:2.5, 1:3, 1:3.5, 1:4, 1:4.5, and 1:5. The Y axis is arbitrary intensity units, on the same scale on each side.
activity is unlikely to account for a significant fraction of the
activity observed. It is also possible that during expression and
purification of enzymes, endogenous substances are introduced
from the expression host.20 Nevertheless, the controls below
clearly eliminated this issue from consideration.
followed by reconstitution with E1ec and E3ec (Figure S1 of
the Supporting Information) at the same ratio that was used for
K41A and H399C; i.e., at K41A:K41A/H399C mass ratios of
1:1, 1:1.5, and 1:2, the overall activity (Table 1 and Figure 2)
yielded very low activity once more according to ΔA340/time
(Figure 2, bottom; notice significant noise for these measure-
ments). This indicates that there is little if any interchain acetyl
transfer from the doubly substituted variant (impaired in both
lipoylation and catalytic sites) to the catalytic site of K41A E2ec
and ultimately to CoA. Most importantly, the sparingly small
activity of this doubly substituted variant provides excellent
support for the conclusions drawn regarding enhanced activity
observed on complementation of the two singly substituted
variants.
Given that each E2ec variant probably exists as a 24-mer,
their ability to mix chains had to be established. The overall
activity (NADH production) of the mixture of two PDH com-
plexes, one reconstituted from K41A E2ec and the other from
H399C E2ec, was nearly zero (2.6%), which is nearly the same
as the overall PDHc activity for K41A E2ec by itself (Table 1)
after incubation for 30 min. In contrast, preincubation of K41A
Upon reconstitution with E1ec and E3ec, K41A E2ec and
H399C E2ec displayed diminished activity compared to that of
parental 1-lip E2ec (2.5 and 2.8%, respectively), indicating they
were nearly totally impaired in terms of overall activity. Else-
where, extensive mutagenesis studies on E2ec20,21 also indi-
cated that 2−3% activity is the minimum we can achieve with
the strain used for single substitutions, with more activity reduc-
tion with multiple substitutions, but when K41A E2ec and
H399C E2ec were mixed and incubated (Figure S1 of the
Supporting Information) at K41A:H399C mass ratios of 1:1,
1:1.5, and 1:2, reconstituted PDHc activities increased with an
increasing concentration of H399C E2ec to a maximum of
22.7%, clearly well above the baseline activity measured with
the individual variants (Table 1 and Figure 2, top).
As an important control, we used the K41A/H399C E2ec
doubly substituted variant: its preincubation with K41A E2ec
2799
dx.doi.org/10.1021/bi201614n | Biochemistry 2012, 51, 2795−2803

Biochemistry
Article
E2ec with H399C E2ec in a 1:1 molar ratio and then recon-
stitution with E1ec and E3ec produced 10% activity from the
inactive state almost immediately upon mixing (Figure S2 of
the Supporting Information), and the activity reached a maxi-
mum in 25−30 min (data not shown). This result indicated
that exchange of E2 chains between preformed PDH com-
plexes, one formed exclusively from K41A E2ec and the other
from H399C E2ec, is much slower than exchange of chains
between the two E2 oligomers, one formed from K41A E2ec
and the other from H399C. Presumably, these 24-mers are pro-
duced as soon as the protein is released from the ribosome.
Given the central position of the E2 chains in PDHc, the result
is perhaps not surprising but needed to be experimentally
demonstrated. This experiment is important in the interpreta-
tion of the results: it appears that the E2 chains (eight trimers
for a total of 24 at the vertexes of a cube according to the octa-
hedral symmetry accepted for this PDHc) are in essence
“sequestered” in each complex molecule once assembled from
the component enzymes.
Mass spectrometric data provided dramatic independent sup-
port for the production of acetyl-CoA in the PDHc reaction by
interchain acetyl transfer, in accord with the kinetic activity
assay, showing that considerably more NADH was produced on
complementation of K41A E2ec and H399C E2ec than with
either variant alone. The MS experiments report the amount of
acetyl-CoA produced by each variant or a mixture thereof un-
der the same conditions, but not the kinetics of acetyl-CoA
formation (see Experimental Procedures). Standard solutions
of CoA and acetyl-CoA were prepared and observed at m/z
768.123 and 810.133, respectively (Figure S2 of the Supporting
Information). Isotopic patterns of these standards also matched
theoretical data. To quantify the concentration of acetyl-CoA
produced by E1ec, E2ec, and pyruvate in the PDHc reaction by
the FTMS, a standard curve was constructed for acetyl-CoA
(1, 5, 10, 20, and 30 μM), generating a linear plot of intensity
versus acetyl-CoA concentration (Figure S3 of the Supporting
Information). Excellent goodness of fit of the linear regression
assured that it could be used to quantify the acetyl-CoA be-
ing produced in the PDHc reaction of all single-site variants
(K41A, H399A, and H399C) and mixtures of two variants,
K41A and H399C.
1-lip E2ec clearly produced acetyl-CoA at m/z 810.13
(Figure 3, left), while the amount of acetyl-CoA produced by
H399A E2ec, H399C E2ec, and K41A E2ec was much smaller.
The scales of the X and Y axes are the same in these figures to
allow comparison of the acetyl-CoA being produced. Consis-
tent with the NADH assay for PDHc, the H399A E2ec
produced much more acetyl-CoA than H399C E2ec or K41A
E2ec. Using the standard plot in Figure S3 of the Supporting
Information, the concentration of acetyl-CoA produced by
E1ec, E2ec (and its variants), E3ec, and pyruvate was quanti-
fied. The reaction was started with 100 μM CoA, yielding
the concentrations of acetyl-CoA in parentheses: 1-lip E2ec
(26.0 μM), H399A (12.3 μM), H399C (0.47 μM), and K41A
(0.03 μM). Again, the H399C, but not H399A, substitution
abolished activity.
Interchain acetyl transfer in the E2 component was next
tested by the FTMS. The production of acetyl-CoA with a fixed
concentration of K41A E2ec complemented with increasing
concentrations of H399C was quantified (Figure 3 right, X and
Y axes of all spectra are the same) with K41A E2ec:H399C
E2ec mass ratios of 1:1, 1:1.5, 1:2, 1:2.5, 1:3, 1:3.5, 1:4, 1:4.5,
and 1:5. Up to 14.8% conversion to acetyl-CoA is in evidence,
exceeding the amount produced by the respective variants by
themselves by numbers well in excess of the experimental error
(Table 2). Under the experimental conditions, at a 1:1 K41A
Table 2. Estimated Quantities of Acetyl-CoA Obtained by
a
Single and Complemented Variants by the FTMS
acetyl-CoA
E1ec:E2ec:E3ec
mass ratio
produced
type
substitutions of E2ec
none
(μM)
1-lip E2ec
1:1:1
26.0 0.3
0.03 0.01
0.47 0.02
12.3 0.2
11.3 0.03
single variant Lys41Ala
His399Cys
1:1:1
1:1:1
His399Ala
1:1:1
b
mixture of
two
variants
Lys41Ala:His399Cys
(separate variants)
1:(1:1):1
1:(1:1.5):1
1:(1:2):1
1:(1:2.5):1
1:(1:3):1
1:(1:3.5):1
1:(1:4):1
1:(1:4.5):1
1:(1:5):1
11.8 0.05
12.1 0.03
12.7 0.04
13.1 0.02
13.2 0.01
13.3 0.03
13.4 0.02
14.8 0.08
a
In all experiments, 100 μM coenzyme A was used, and all reaction
mixtures (after reconstitution) were incubated for 10 min.
Quantitative Fourier transform mass spectrometry data have been
b
averaged using 10 independent measurements. Two variants of E2ec
were mixed and incubated to monitor the production of acetyl-CoA.
E2ec:H399C E2ec mass ratio, 11.3 μM acetyl-CoA was pro-
duced from 100 μM CoA, while even a 1:5 mass ratio only
produced 14.8 μM, indicating that the 1:1 mass ratio was
sufficient to confirm interchain acetyl transfer between the two
E2ec variants. H399A E2ec provides a very good control and
also raises a new mechanistic question. On one hand, a com-
parison of the activities and acetyl-CoA production by H399A
E2ec and H399C E2ec supports our use of the latter to demo-
nstrate interchain acetyl transfer within the E2ec component.
On the other hand, the same comparison also points out our
lack of understanding of the transthiolacetylation mechanism,
transfer of the acetyl group from the acetyldihydrolipoyl E2 to
CoA at the E2 catalytic center. Should H399 be a crucial cata-
lytic residue, say a general acid−base catalyst, its substitution
with alanine should abolish the activity, which is clearly not the
case. The H399C E2ec substitution suggests that this cysteine
diminishes the activity for acetyl-CoA production to baseline by
a hitherto uncharacterized mechanism.
These results collectively strongly support our hypothesis
that interchain communication between the lipoyllysine site on
H399C and the catalytic site on K41A gave rise to the pro-
duction of acetyl-CoA upon reconstitution with E1ec, E3ec,
and pyruvate.
CONCLUSION
■
A complementation experiment was designed to test the
hypothesis that acetyl transfer between acetyldihydrolipoyl-E2
on the LD and coenzyme A, presumably at the CD of E2, takes
place by an interchain rather than an intrachain mechanism.
Both an activity assay and the direct mass spectral measurement
of the acetyl-CoA support the interchain acetyl transfer
hypothesis. A comparison of the behavior of H399A E2ec
with that of H399C E2ec showed that upon complementation
2800
dx.doi.org/10.1021/bi201614n | Biochemistry 2012, 51, 2795−2803

Biochemistry
Article
Figure 4. Model for utilization and assembly of multiples of three E2ec chains in reductive acetylation and acetyl-CoA formation.
with K41A E2ec, H399A E2ec produced a significantly higher
activity and level of acetyl-CoA than did complementation with
H399C E2ec. This is consistent with the report that the activity
of H602A in 3-lip E2ec (corresponding to H399 in 1-lip E2ec)
remained detectable,16,17 also suggesting the need for further
work to elucidate this mechanism.
While the notion of “active site coupling” in such systems has
been well substantiated and accepted,1,13−17 it typically refers to
interchain transfer of acetyl groups and reducing equivalents
between LDs, whereas in this study, we demonstrate interchain
acetyl transfer from a LD to the CD and to coenzyme A. The
results prompt us to propose the following tentative
2801
dx.doi.org/10.1021/bi201614n | Biochemistry 2012, 51, 2795−2803

Biochemistry
Article
explanation for the fact that bacterial E2 components exist as
multiples of three chains. Let us assume that of the three chains
in E2, say chains A, B, and C, chains A and C are in parallel
while chain B has an altered orientation such that its active cen-
ter can be reached by the acetyldihydrolipoyl group of chain A
so transthiolacetylation to CoA can take place (Figure 4). In a
single turnover according to this model, chain A would be re-
ductively acetylated (Figure 4 step 1) and chain B would accept
and then transfer the acetyl group to CoA, not using its lipoyl
group at all, while chain A would not use its catalytic center
(Figure 4, step 2). Chain C, with an orientation similar to that
of A, would not use its catalytic center either; however, it would
communicate reducing equivalents between chain A and E3ec
(Figure 4, step 3), and chain C would be reoxidized by the E3
component leading to NADH as a final product of the catalytic
cycle (Figure 4, step 4). This model also provides an expla-
nation for the finding that E1ec and E3ec compete for
overlapping,13−16 in our experience nonidentical,20 sites on the
PSBD.
It is useful to consider some of the many important contri-
butions prior to this work pointing to the possibility or proba-
bility of interchain group transfer on the basis of structural and
mechanistic information. (1) The trimeric nature of the core
domains has been established since the publication of X-ray
structures some years ago (e.g., refs 22 and 23). (2) In the
1-lipoyl E2 construct used here, the importance of the length of
the linker connecting the lipoyl domain to the PSBD for
ASSOCIATED CONTENT
* Supporting Information
Three figures. This material is available free of charge via the
Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
*E-mail: frjordan@rutgers.edu. Phone: (973) 353-5470. Fax:
(973) 353-1264.
■
Funding
Supported by National Institutes of Health Grant GM050380.
Notes
The authors declare no competing financial interest.
ABBREVIATIONS
■
ThDP, thiamin diphosphate; PDHc, pyruvate dehydrogenase
complex; E1, first pyruvate dehydrogenase component of
PDHc; E2, second dihydrolipoylacetyltransferase component of
PDHc; 1-lip E2, single-lipoyl construct of E2 from E. coli; 3-lip
E2, wild-type three-lipoyl E2 from E. coli; E3, third
dihydrolipoamide dehydrogenase component of PDHc; LD,
lipoyl domain of E2; PSBD, peripheral subunit binding domain
of E2; CD, core or catalytic domain of E2; FTMS, Fourier
transform mass spectrometer; CoA, coenzyme A.
REFERENCES
■
̀
efficient active center coupling (hence flexibility vis-a-vis other
(1) Reed, L. J. (1974) Multienzyme complexes. Acc. Chem. Res. 7,
40−46.
(2) Patel, M. S., Roche, T. E., and Harris, R. A. (1996) α-Keto acid
chains) was shown by Guest and Perham and co-workers.24
The stiffness of the inner linker regions was implied by NMR
studies of peptides corresponding to the linker sequence, sug-
gesting an “extended yet flexible” E2 structure,25 while for hu-
man E2, cryo-electron microscopy suggested flexible N-terminal
but less flexible core domains,26 both papers concluding
there are significant distances among E2 chains but also a sig-
nificant span for reaching the E1 and E3 components. Active site
coupling was shown in the E. coli and other complexes:1,13−17,24,27
for example, PDHc activity was retained after release of as
much as 50% of either lipoyl domains (by trypsin) or lipoate
(by lipoamidase) in the E. coli complex,27 while release of as
much as 50% of the lipoyl domains in the human complex had
little or no effect on the activity.28 Different functions for dif-
ferent chains of E2 (one chain engaged with E1, one with E3,
and the third with undefined function) were explicitly sug-
gested in Figure 14 of ref 15.
dehydrogenase complexes, Birkhauser, Basel, Switzerland.
̈
(3) Gopalakrishnan, S., Rahmatullah, M., Radke, G. A., Powers-
Greenwood, S., and Roche, T. E. (1989) Role of protein X in the
function of the mammalian pyruvate dehydrogenase complex. Biochem.
Biophys. Res. Commun. 160, 715−721.
(4) Neagle, J., De Marcucci, O., Dunbar, B., and Lindsay, J. G. (1989)
Component X of mammalian pyruvate dehydrogenase complex:
Structural and functional relationship to the lipoate acetyltransferase
(E2) component. FEBS Lett. 253, 11−15.
(5) Brautigam, C. A., Wynn, R. M., Chuang, J. L., and Chuang, D. T.
(2009) Subunit and catalytic component stoichiometries of an in vitro
reconstituted human pyruvate dehydrogenase complex. J. Biol. Chem.
284, 13086−13098.
(6) Hiromasa, Y., Fujisawa, T., Aso, Y., and Roche, T. E. (2004)
Organization of the cores of the mammalian pyruvate dehydrogenase
complex formed by E2 and E2 plus the E3-binding protein and their
capacities to bind the E1 and E3 components. J. Biol. Chem. 279,
6921−6933.
(7) Vijayakrishnan, S., Kelly, S. M., Gilbert, R. J., Callow, P., Bhella,
D., Forsyth, T., Lindsay, J. G., and Byron, O. (2010) Solution structure
and characterisation of the human pyruvate dehydrogenase complex
core assembly. J. Mol. Biol. 399, 71−93.
(8) Eley, M. H., Namihira, G., Hamilton, L., Munk, P., and Reed, L. J.
(1972) α-Keto acid dehydrogenase complexes. 18. Subunit composi-
tion of the Escherichia coli pyruvate dehydrogenase complex. Arch.
Biochem. Biophys. 152, 655−669.
(9) Bleile, D. M., Munk, P., Oliver, R. M., and Reed, L. J. (1979)
Subunit structure of dihydrolipoyl transacetylase component of
pyruvate dehydrogenase complex from Escherichia coli. Proc. Natl.
Acad. Sci. U.S.A. 76, 4385−4389.
(10) Stephens, P. E., Darlison, M. G., Lewis, H. M., and Guest, J. R.
(1983) The pyruvate dehydrogenase complex of Escherichia coli K12.
Nucleotide sequence encoding the dihydrolipoamide acetyltransferase
component. Eur. J. Biochem. 133, 481−489.
(11) Angelides, K. J., Akiyama, S. K., and Hammes, G. G. (1979)
Subunit stoichiometry and molecular weight of the pyruvate
Our work has shown explicitly such interchain acetyl transfer
for the E. coli E2 component. The transfer among chains of
acetyl groups and reducing equivalents had been proposed
many years ago, but their rates need further examination.29
What our results could not do is to clarify whether the acetyl
transfer from a dihydrolipoyl group of one chain produces
acetyl-CoA at a core domain of a different chain according to
an “intratrimer” or “intertrimer” path. However, the experi-
ments did rule out transfer to an adjacent complex. Previously,
it was shown by others that E1 chains in this complex exchange
at very slow rates (t1/2 values of many hours).29,30 For this com-
plex, for the purposes of such experiments, each complex mol-
ecule is sequestered once it is assembled. Our work addresses
and provides a tool for solving an issue common in enzymes
and proteins: when there are multiple copies of a component,
they may serve mechanistically distinct functions in each turn-
over. We present testable models for further work.
2802
dx.doi.org/10.1021/bi201614n | Biochemistry 2012, 51, 2795−2803

Biochemistry
Article
dehydrogenase multienzyme complex from Escherichia coli. Proc. Natl.
Acad. Sci. U.S.A. 76, 3279−3283.
Transacetylase Component But Not from the Protein X Component.
J. Biol. Chem. 265, 14512−14517.
(29) Collins, J. H., and Reed, L. J. (1977) Acyl group and electron pair
relay system: A network of interacting lipoyl moieties in the pyruvate and
α-ketoglutarate dehydrogenase complexes from Escherichia coli. Proc. Natl.
Acad. Sci. U.S.A. 74, 4223−4227.
(30) Hale, G., Bates, D. J., and sand Perham, R. N. (1979) Subunit
exchange in the pyruvate dehydrogenase complex of Escherichia coli.
FEBS Lett. 104, 343−346.
(12) CaJacob, C. A., Frey, P. A., Hainfeld, J. F., Wall, J. S., and Yang,
H. (1985) Escherichia coli pyruvate dehydrogenase complex: Particle
masses of the complex and component enzymes measured by scanning
transmission electron microscopy. Biochemistry 24, 2425−2431.
(13) Reed, L. J., and Hackert, M. L. (1990) Structure-function
relationships in dihydrolipoamide acyltransferases. J. Biol. Chem. 265,
8971−8974.
(14) Perham, R. N. (1991) Domains, motifs, and linkers in 2-oxo
acid dehydrogenase multienzyme complexes: A paradigm in the design
of a multifunctional protein. Biochemistry 30, 8501−8512.
(15) Perham, R. N. (2000) Swinging arms and swinging domains in
multifunctional enzymes: Catalytic machines for multistep reactions.
Annu. Rev. Biochem. 69, 961−1004.
(16) de Kok, A., Hengeveld, A. F., Martin, A., and Westphal, A. H.
(1998) The pyruvate dehydrogenase multi-enzyme complex from
Gram-negative bacteria. Biochim. Biophys. Acta 1385, 353−366.
(17) Russell, G. C., Machado, R. S., and Guest, J. R. (1992)
Overproduction of the pyruvate dehydrogenase multienzyme complex
of Escherichia coli and site-directed substitutions in the E1p and E2p
subunits. Biochem. J. 287 (Part 2), 611−619.
(18) Yi, J., Nemeria, N., McNally, A., Jordan, F., Machado, R. S., and
Guest, J. R. (1996) Effect of substitutions in the thiamin diphosphate-
magnesium fold on the activation of the pyruvate dehydrogenase
complex from Escherichia coli by cofactors and substrate. J. Biol. Chem.
271, 33192−33200.
(19) Wei, W., Li, H., Nemeria, N., and Jordan, F. (2003) Expression
and purification of the dihydrolipoamide acetyltransferase and
dihydrolipoamide dehydrogenase subunits of the Escherichia coli
pyruvate dehydrogenase multienzyme complex: A mass spectrometric
assay for reductive acetylation of dihydrolipoamide acetyltransferase.
Protein Expression Purif. 28, 140−150.
(20) Park, Y. H. (2008) Ph.D. Dissertation, Rutgers University
Graduate Faculty at Newark, Newark, NJ.
(21) Park, Y. H., Wei, W., Zhou, L., Nemeria, N., and Jordan, F.
(2004) Amino-terminal residues 1−45 of the Escherichia coli pyruvate
dehydrogenase complex E1 subunit interact with the E2 subunit and
are required for activity of the complex but not for reductive
acetylation of the E2 subunit. Biochemistry 43, 14037−14046.
(22) Mattevi, A., Obmolova, G., Schulze, E., Kalk, K. H., Westphal,
A. H., de Kok, A., and Hol, W. G. (1992) Atomic structure of the cubic
core of the pyruvate dehydrogenase multienzyme complex. Science 255,
1544−1550.
(23) Knapp, J. E., Carroll, D., Lawson, J. E., Ernst, S. R., Reed, L. J.,
and Hackert, M. L. (2000) Expression, Purification, and Structural
Analysis of the Trimeric Form of the Catalytic Domain of the
Escherichia coli Dihydrolipoamide Succinyltransferase. Protein Sci. 9,
37−48.
(24) Miles, J. S., Guest, J. R., Radford, S. E., and Perham, R. N.
(1988) Investgation of the Mechanism of Active Site Coupling in the
Pyruvate Dehydrogenase Complex of Escherichia coli by Protein
Engineering. J. Mol. Biol. 202, 97−106.
(25) Lengyel, J. S., Stott, K. M., Wu, X., Brooks, B. R., Balbo, A.,
Schuck, P., Perham, R. N., Subramaniam, S., and Milne, J. L. (2008)
Extended polypeptide linkers establish the spatial architecture of a
pyruvate dehydrogenase multienzyme complex. Structure 16, 93−103.
(26) Yu, X., Hiromasa, Y., Tsen, H., Chiu, W., Stoops, J. K., Roche,
T. E., and Zhou, Z. H. (2008) Structures of the human pyruvate
dehydrogenase complex cores: A highly conserved catalytic center with
flexible N-terminal domains. Structure 16, 104−114.
(27) Stepp, L. R., Bleile, D. M., McRorie, D. K., Pettit, F. H., and
Reed, L. J. (1981) Use of trypsin and lipoamidase to study the role of
lipoic acid moieties in the pyruvate and α-ketoglutarate dehydrogenase
complexes of Escherichia coli. Biochemistry 20, 4555−4560.
(28) Rahmatullah, M., Radke, G. A., Andrews, P. C., and Roche, T. E.
(1990) Changes in the Core of the Mammalian-Pyruvate Dehydrogen-
ase Complex upon Selective Removal of the Lipoyl Domain from the
2803
dx.doi.org/10.1021/bi201614n | Biochemistry 2012, 51, 2795−2803