Báo cáo khoa học: Protein engineering of pyruvate carboxylase Investigation on the function of acetyl-CoA and the quaternary structure doc

The reaction of PC is believed to proceed in two steps, just likethose of other biotin-dependent carboxylases such as acetyl-CoA carboxylase: ATPþ HCO3 þ enz-biotin Ð enz-biotin-CO2 þ AD

Protein engineering of pyruvate carboxylase

Investigation on the function of acetyl-CoA and the quaternary structure

Shinji Sueda, Md Nurul Islam and Hiroki Kondo

Department of Biochemical Engineering and Science, Kyushu Institute of Technology, Japan

Pyruvate carboxylase (PC) from Bacillus thermodenitrificans

was engineered in such a way that the polypeptide chain was

divided into two, between the biotin carboxylase (BC) and

carboxyl transferase (CT) domains The two proteins thus

formed, PC-(BC) and PC-(CT+BCCP), retained their

catalytic activity as assayed by biotin-dependent ATPase

and oxamate-dependent oxalacetate decarboxylation, for

the former and the latter, respectively Neither activity was

dependent on acetyl-CoA, in sharp contrast to the complete

reaction of intact PC When assessed by gel filtration

chromatography, PC-(BC) was found to exist either in

dimers or monomers, depending on the protein concentra-tion, while PC-(CT +BCCP) occurred in dimers for the most part The two proteins do not associate spontaneously

or in the presence of acetyl-CoA Based on these observa-tions, this paper discusses how the tetrameric structure of

PC is built up and how acetyl-CoA modulates the protein structure

Keywords: acetyl-CoA; biotin; biotin-dependent carboxy-lase; protein engineering; pyruvate carboxylase

Pyruvate carboxylase (PC) is a biotin-dependent enzyme

and is involved in gluconeogenesis by converting pyruvate

to oxalacetate [1–3] There are two forms of PC, single

polypeptide chain type and subunit type, but a large

majority belongs to the former class [1,4–7] This form of

PC is made of about 1200 amino acids and is distributed

widely in both eukaryotes and some prokaryotes The

reaction of PC is believed to proceed in two steps, just

likethose of other biotin-dependent carboxylases such as

acetyl-CoA carboxylase:

ATPþ HCO3 þ enz-biotin Ð enz-biotin-CO2 þ ADP þ Pi

Scheme 1 enz-biotin-CO2 þ pyruvate Ð enz-biotin þ oxalacetate

Scheme 2

In the first step (Scheme 1), the biotin moiety covalently

attached to the enzyme is carboxylated by bicarbonate

and ATP In the second step (Scheme 2), the carboxyl

group is transferred from carboxybiotin to pyruvate

Thus, PC carries at least three functional domains: a

biotin carboxyl carrier protein (BCCP) domain, a biotin

carboxylase (BC) domain which mediates the first partial reaction and a carboxyl transferase (CT) domain which catalyzes the second partial reaction The BC domain is located in the amino terminus of the single polypeptide chain PC, followed by CT with the BCCP domain in the carboxyl terminus [Fig 1] The activity of PC is activa-ted by acetyl-CoA and inhibiactiva-ted by aspartate [2,8–10] Because of the lack of a three-dimensional structure, the detailed mechanism of carboxylation and regulation of

PC remains obscure Obviously, elucidation of the three-dimensional structure of PC will unveil much of this uncertainty and in fact such an undertaking is under way

in this laboratory Additionally, a protein engineering approach would be useful to examine the two partial reactions individually In this study, PC from Bacillus thermodenitrificans (previously Bacillus stearothermophilus) was engineered in such a way as to divide the protein into two at the boundary of the BC and CT domains (Fig 1) The properties of the resulting two proteins, PC-(BC) and PC-(CT +BCCP), were examined and compared with those of the intact PC in order to gain insight into the domain organization, the function of acetyl-CoA and the reaction mechanism of PC

Experimental procedures

Materials Inorganic salts and common organic chemicals were obtained from commercial sources Acetyl-coenzyme A was from Wako Pure Chemical (Osaka, Japan) and avidin was from ProZyme (San Leandro, CA, USA) Reagents for genetic engineering, such as restriction enzymes, were purchased from Takara (Kyoto, Japan) Oligonucleotides were custom synthesized by Hokkaido Science (Sapporo, Japan) The TOPO TA cloning kit was the product of Invitrogen

Correspondence to S Sueda, Department of Biochemical Engineering

and Science, Kyushu Institute of Technology, Kawazu 680-4, Iizuka

820-8502, Japan Fax: +81 948 29 7801, Tel.: +81 948 29 7834,

E-mail: sueda@bse.kyutech.ac.jp

Abbreviations: BC, biotin carboxylase; BCCP, biotin carboxyl

carrier protein; CT, carboxyl transferase; DTT, dithiothreitol;

KP i , potassium phosphate; PC, pyruvate carboxylase.

Enzymes: pyruvate carboxylase from Bacillus thermodenitrificans

(P94448) (EC 6.4.1.1); biotin carboxylase subunit of acetyl-CoA

carboxylase from Escherichia coli (P24182) (EC 6.4.1.2).

(Received 15 January 2004, revised 16 February 2004,

accepted 24 February 2004)

Construction of an overexpression plasmid

for intact PC

Previously, the B thermodenitrificans PC gene was cloned

into pBluescript vector [11] The resulting recombinant

plasmid (pPC) allowed Escherichia coli to express PC, albeit

at a relatively low level (data not shown) To enhance

expression, the promoter region was replaced with the high

expression promoter trc of pTrc99A vector Thus, the

 800 bp downstream region from the NcoI site containing

the trc promoter was amplified with pTrc99A as template

and using the following primers: Trc 1, 5¢-TTAGCGG

GCCCATTAAGTTCTGTC-3¢ and Trc 2, 5¢-TTGCGA

ATTCGTCTTGTCTCCATGGTCTGTTTCCTGTGTG

AAAT-3¢ (restriction enzyme sites are underlined) The

EcoRI site, present at about 10 bp downstream from the

initiating ATG codon of the PC gene, and the ApaI site,

present on pBluescript and pTrc99A, were exploited for

gene manipulation A 19 bp segment of the PC gene (shown

above in italics) containing an EcoRI site was incorporated

into the reverse primer, Trc 2 The PCR reaction was

conducted under the following conditions: The reaction

mixture contained 5 units of Ex TaqTM(Takara), 1· Ex Taq

buffer, 200 lMeach of the four dNTPs, 1 lMeach of the

primers and 10 ng of pTrc99A in a final volume of 100 lL

After denaturation at 94C for 5 min, the samples were

subjected to 30 cycles of denaturation (94C, 1 min),

annealing (58C, 1 min) and extension (72 C, 1 min),

and subsequently subjected to additional extension (72C,

10 min) The PCR products were TA cloned and sequenced

The plasmid thus prepared was digested with ApaI and

EcoRI, and the resulting fragment was ligated into the ApaI/

EcoRI sites of pPC The second amino acid of native PC

is converted to glutamic acid from lysine because of the

introduction of the NcoI site into the start codon region of

this recombinant This plasmid allowed E coli to express

PCat a much higher level and the enzyme produced was

as active as native PC Hence, this PC is called intact PC

despite the mutation of the second amino acid residue

Construction of over-expression plasmids for PC-(BC)

and PC-(CT + BCCP)

The boundary of the BC and CT domains of B

thermo-denitrificansPC was estimated to reside at residue 462 on the

basis of the reasoning described in the Results section The

polypeptide chain was divided into two at this point by

placing a stop codon or an initiation codon for the

expres-sion of BC and CT plus BCCP, respectively Expresexpres-sion

plasmids for PC-(BC) and PC-(CT +BCCP) were

constructed as follows: For the former, 440 bp fragment

was amplified with pPC as the template using the following primers: BC1, 5¢-ATTGATATCGTCCAGTCG CAAATTTTAATTGCT-3¢ and BC2, 5¢-ATAGGATCC TTAGAACACGAATAGTTCCGGCGTCGTATCGAT-3¢ (restriction enzyme sites are underlined) The forward primer, BC1, harbored the EcoRV site present on the PC gene, and the reverse primer, BC2, harbored a stop codon (denoted in bold) A BamHI site was introduced for subsequent manipulation PCR conditions were the same

as those for the amplification of the trc promoter, and the PCR product was TA cloned and sequenced The resulting plasmid was digested with EcoRV and BamHI, and the fragment formed was ligated into the EcoRV/BamHI sites

of pPC The promoter of this plasmid was replaced with the high expression promoter trc in exactly the same way as that

of the intact PC This plasmid, pPC-(BC), allowed E coli

to express the BC domain of PC at a high level

The PC-(CT +BCCP) expression plasmid was con-structed as follows: an 400 bp fragment was amplified with pPC as template using the following primers: CT1, 5¢-ATATCCATGGCACGCCGGAAAGACGGAACGA AAATG-3¢ and CT2, 5¢-CCGATCCCACGGATCCTCT TTTAAAAAGCG-3¢ (restriction enzyme sites are under-lined) The forward primer, CT1, harbored an NcoI site introduced for placing the start codon and cloning, and the reverse primer, CT2, harbored a BamHI site present on the

PCgene As a result of the engineering, the second amino acid residue is converted from proline to alanine PCR conditions were the same as those described above, and the PCR product was TA cloned and sequenced Likewise,

a fragment representing the downstream region from the BamHI site to the end of the open reading frame was prepared (S Sueda, unpublished observation) These two fragments were cloned into pTrc99A through multiple steps

to yield a recombinant plasmid, pPC-(CT +BCCP), which allowed E coli to express the desired CT plus BCCP domain of PC to a high level

Purification of proteins

E coli JM109 transformed with either one of the over-expression plasmids prepared above was grown in Luria-Bertani medium containing 50 lgÆmL)1 ampicillin and

1 lgÆmL)1 D-biotin, where a biotin-binding domain was present Cells were harvested by centrifugation, suspended

in 0.12M potassium phosphate (KPi) buffer, pH 7.0, containing 1 mM EDTA, 1 mM dithiothreitol (DTT) and

1 mM phenylmethanesulfonyl fluoride, disrupted by soni-cation and then centrifuged The precipitate that formed was removed by centrifugation, and ammonium sulfate was added to the supernatant to 40–50% saturation for intact

PC and PC-(CT +BCCP), and 30–40% saturation for PC-(BC) Again, the precipitate formed was collected by centrifugation, dissolved in buffer A (20 mM KPi buffer,

pH 7.0, containing 0.1 mMEDTA and 0.1 mMDTT), and dialyzed against the same buffer The samples were subjec-ted to anion exchange chromatography on diethylamino-ethyl (DEAE)-cellulose (Whatman) Proteins were eluted by

a salt gradient from buffer A to buffer B (buffer A +0.5M NaCl) The desired fractions, inspected by SDS/PAGE, were collected and dialyzed against buffer A The samples were applied to gel filtration chromatography on

Super-Fig 1 Schematic representation of the domain structures of intact PC

and engineered proteins, PC-(BC) and PC-(CT +BCCP).

dexTM 200 (Amersham), eluted with 50 mM KPi buffer,

pH 7.0, containing 0.1MNaCl, 0.1 mMEDTA and 0.1 mM

DTT, and the desired fractions collected Intact PC and

PC-(CT+BCCP) were further purified by monomeric

avidin-Sepharose affinity chromatography [12–14] as reported

previously [15] The samples were applied at a flow rate of

0.5 mLÆmin)1onto the monomeric avidin column

equili-brated with running buffer (50 mM KPi buffer, pH 7.0,

0.2M KCl, 1 mM EDTA, 5 mM2-mercaptoethanol) The

column was washed with several column volumes of running

buffer to remove unbound material Proteins were eluted

with 1 mgÆmL)1biotin in running buffer at a flow rate of

0.2 mLÆmin)1 The eluted intact PC and PC-(CT+BCCP)

were dialyzed against 5 mMKPibuffer, pH 7.0, containing

0.1 mMEDTA and 0.1 mMDTT, and stored at 4C In the

meantime, PC-(BC) was further purified by anion exchange

chromatography on Mono QTM HR 5/5 (Amersham)

Protein was eluted by a salt gradient from buffer C (20 mM

Tris/HCl, pH 7.5) to buffer D (buffer C+0.35M NaCl)

The desired fractions were collected and dialyzed against

5 mMKPibuffer, pH 7.0, with 0.1 mMEDTA and 0.1 mM

DTT, and stored at 4C The specific activity of PC,

determined below, was 9.5 UÆmg)1, where 1 U is defined as

the amount of enzyme to produce 1 lmol of oxalacetate per

min, and the protein concentration was determined from the

amino acid composition

Pyruvate carboxylase assays

Pyruvate carboxylase activity was measured by monitoring

the oxalacetate formation using the coupled reaction with

malate dehydrogenase according to the methods described

previously [16–18] Oxidation of NADH in the malate

dehydrogenase reaction was followed

spectrophotometri-cally at 340 nm All assays were carried out at 30C, and

the reaction mixture contained the following components,

unless otherwise stated: 100 mM Tris/HCl (pH 8.0), 2 mM

ATP, 5 mMMgCl2, 100 mMKCl, 5 mMpyruvate, 50 mM

NaHCO3, 0.1 mMacetyl-CoA, 0.15 mMNADH and 5 units

of malate dehydrogenase

The Km(Michaelis constant) values for ATP, bicarbonate

and pyruvate were determined as follows: the Kmfor ATP

was obtained by varying its concentration from 0–5 mMat

fixed concentrations of bicarbonate (100 mM) and pyruvate

(5 mM), where the enzyme was 77% and 92% saturated with

them, respectively In addition, free Mg2+ concentration

was kept constant, with MgCl2, at 3 mMin excess of ATP;

free Mg2+concentration was approximated to be its

ana-lytical concentration minus that of ATP, as the true

concen-tration of free Mg2+calculated based on the dissociation

constant for MgATP of 0.0143 mM [19] was only 1%

different from the approximate value At high ATP

concen-trations, substrate inhibition was evident, and thus two kinds

of analysis were applied for the data on ATP First, the

simple Michaelis–Menten equation was fitted to the kinetic

data in the low concentration range (0–1 mM) using the

nonlinear regression analysis program,ENZFITTER(Biosoft,

Cambridge, UK) Then, the entire data (from 0–5 mM) were

analyzed by Eqn (1), which takes into account substrate

inhibition, where v, Vmax, [S] and KI represent observed

reaction rate, maximum rate, substrate concentration and

the substrate inhibition constant, respectively:

v¼ Vmax ½S

½SþK m þ½S2

½K I 

The Kmvalue for bicarbonate was determined by varying its concentration from 0.5–100 mMat fixed concentrations

of ATP (2 mM) and pyruvate (5 mM) As the endogenous level of bicarbonate is known to be 0.5 mM at pH 8.0 [20], the concentration of bicarbonate was corrected for this value Likewise, the Km value for pyruvate was determined by varying its concentration from 0–5 mMat fixed concentrations of ATP (2 mM) and bicarbonate (100 mM) The simple Michaelis–Menten equation was used for the analysis of the data for bicarbonate and pyruvate

ATP cleavage assays ATP cleavage activity of intact PC and PC-(BC) was assayed according to the previously reported procedure [21] The progress of the reaction was followed by monitoring the formation of ADP in the presence of phosphoenolpyru-vate and pyruphosphoenolpyru-vate kinase The pyruphosphoenolpyru-vate formed was then reduced to lactate by lactate dehydrogenase with the concomitant oxidation of NADH, and this was measured from a decrease in absorbance at 340 nm All assays were conducted at 30C, and the reaction mixture contained the following components, unless otherwise stated: 100 mM Tris/HCl (pH 8.0), 2 mM ATP, 5 mM MgCl2, 100 mM KCl, 50 mMNaHCO3, 0.1 mMacetyl-CoA, 0.5 mM phos-phoenol pyruvate, 0.15 mM NADH, 5 units of lactate dehydrogenase and 5 units of pyruvate kinase In the case of the PC-(BC) assay, 50 mMfreeD-biotin was added to the above reaction mixture

The kinetic parameters, Km and Vmax, for the ATPase reaction of PC-(BC) were determined as follows: the kinetic parameters for ATP were obtained by varying its concentration from 0–5 mM at fixed concentrations of biotin (100 mM) and bicarbonate (100 mM), where the enzyme is 68% and 62% saturated with them, respect-ively Obviously, this situation is not ideal for the accurate estimation of kinetic parameters, but concentrations higher than this will deviate too much from those of physiological conditions Accordingly, experiments were carried out under these subsaturating conditions with respect to biotin and bicarbonate In the kinetics for ATP, substrate inhibition was manifest at high concentrations just like for intact PC, and thus two kinds of data analysis were also made in this case The kinetic parameters for bicarbonate were determined by varying its concentration from 0.5–100 mMat fixed concentrations

of ATP (2 mM) and biotin (100 mM); again, the concen-tration of bicarbonate was corrected for the endogenous bicarbonate at pH 8.0, 0.5 mM The simple Michaelis– Menten equation was used for the analysis of the data obtained The kinetic data for biotin were obtained by varying its concentration from 0–100 mM at fixed con-centrations of ATP (2 mM) and bicarbonate (100 mM) In the ATPase reaction of PC-(BC), a weak activity (2% of maximum) was observed in the absence of biotin, and thus the data were analyzed by Eqn (2), which takes into account this basal activity (v0):

v¼ Vmax ½S

Kmþ ½S þ v0

Eqnð2Þ

Oxalacetate decarboxylase assays

Oxalacetate decarboxylase activity of intact PC and

PC-(CT +BCCP) was measured with oxamate as the

stimu-lant, according to the procedures previously reported [22]

The reactions were monitored by measuring the formation

of pyruvate which was then reduced to lactate by lactate

dehydrogenase, and the concomitant oxidation of NADH

was monitored at 340 nm All assays were performed at

30C, and the reaction mixture contained the following

components, unless otherwise stated: 100 mM Tris/HCl

(pH 8.0), 5 mMMgCl2, 100 mMKCl, 0.1 mMoxalacetate,

0.1 mMacetyl-CoA, 1 mMoxamate, 0.15 mMNADH, and

5 units of lactate dehydrogenase The reactions were started

by the addition of intact PC or PC-(CT +BCCP), but

prior to the addition, a background rate of oxalacetate

decarboxylation was established, and this (2.4% of the

maximum) was subtracted from the rate in the presence

of enzyme

Avidin-blot analysis and determination

of the N-terminal amino acid sequence

For avidin-blot analysis, electrophoresed samples were

electroblotted onto a nylon membrane (Pall Biosupport,

Portsmouth, UK) according to the conventional procedure

[23] The membrane with blotted proteins was blocked with

skimmed milk in NaCl/Tris-Tween [20 mM Tris/HCl,

pH 7.6, 136 mM NaCl, 0.1% (v/v) Tween] for one hour

The blocked membrane was washed three times with NaCl/

Pi-Tween, and then immersed in NaCl/Tris buffer

contain-ing 0.4 UÆmL)1alkaline phosphatase-conjugated

streptavi-din (Boehringer Mannheim) for 20 min The membrane

was then washed with NaCl/Tris-Tween three times, before

being developed by 0.78 mM4-nitroblue tetrazolium

chlor-ide and 0.40 mM 5-bromo-4-chloro-3-indolylphosphate in

20 mM Tris/HCl, pH 9.5, containing 100 mM NaCl and

50 mMMgCl2

For determining the amino-terminal sequence of the

proteins, electrophoresed samples were electroblotted onto

a poly(vinylidene difluoride) membrane (Atto, Tokyo,

Japan) according to the conventional procedure Pieces

of the membrane containing the desired bands, as visualized

by ponceau S, were used for sequencing by Edman

degradation on a protein sequencer Model 491 (Applied

Biosystems)

Molecular size determination by HPLC gel-filtration

chromatography

High performance gel filtration chromatography was

carried out on a TSKgel G3000SWXL column

(7.8 mm· 30 cm) with TSK guard column SWXL

(6.0 mm· 4.0 cm) (Tosoh, Tokyo, Japan) using an HPLC

system (Hitachi, Tokyo, Japan) The samples were eluted at

a flow rate of 0.5 mLÆmin)1using a mobile phase of 100 mM

KPibuffer (pH 7.0) containing 100 mMNa2SO4, and the

eluted samples were monitored at 280 nm The gel filtration

column was calibrated using a set of proteins (Amersham): ribonuclease A (13.7 kDa), chymotrypsinogen A (23 kDa), ovalbumin (43 kDa), albumin (67 kDa), aldolase (158 kDa) and thyroglobulin (669 kDa) The apparent molecular masses of the samples were estimated from the calibration curve obtained The samples were analyzed at concentrations ranging from 1–100 lM, and 20 lL each was applied to the column

Molecular mass determination by mass spectrometry The molecular masses of PC-(BC) and PC-(CT +BCCP) were determined by MALDI TOF mass spectrometry with

a Voyager DE-STR mass spectrometer (PerSeptive Bio-systems, Framingham, MA, USA)

4-Hydroxyazobenzene-2¢-carboxylic acid (10 mgÆmL)1in 0.1% (v/v) trifluoroacetic acid in 70 : 30 water/acetonitrile) was used as the MALDI matrix Samples were prepared by mixing the protein solution with the matrix solution One microliter of this mixture was deposited on the sample plate, dried at ambient temperature and analyzed

Results

Construction and purification of the engineered proteins of PC

The boundary of the BC and CT domains of PC was estimated as follows: the BC subunit of E coli acetyl-CoA carboxylase is catalytically active and the three-dimensional structure is known [24,25] Its C-terminus appeared to correspond to residue 460 of B thermodenitrificans PC by sequence alignment [26] Likewise, the amino acid sequences

of PCs from various sources, including those of subunit-type PCs, were aligned to reveal that the N-terminus of CT seemed to reside at residue 470 of B thermodenitrificans PC Although there still remains some ambiguity concerning the exact location of the boundary because the C- and N-terminal regions of BC and CT domains, respectively, are barely conserved, it seemed safe to divide the two domains

at residue 462 without impairing the two activities (Fig 1) Based on this assumption, over-expression plasmids for PC-(BC) and PC-(CT +BCCP) which produce BC and the rest of the molecule, respectively, were constructed as detailed in Experimental procedures

The engineered proteins of PC as well as intact PC were purified by methods described under Experimental pro-cedures Monomeric avidin-Sepharose affinity chromato-graphy was used for the purification of intact PC and PC-(CT +BCCP) carrying the biotin prosthetic group within their structures Each purified protein was nearly homogeneous as judged by visual inspection of SDS/PAGE (Fig 2A) The yields were typically 10, 15 and 20 mg, for intact PC, PC-(BC) and PC-(CT +BCCP), respectively, from a 2 L culture Western blot analysis with alkaline phosphatase-conjugated streptavidin for intact PC and PC-(CT +BCCP) revealed that bands were observed at the positions corresponding to those of SDS/PAGE (Fig 2B) The amino-terminal sequence of each protein was analyzed by Edman degradation The amino acid sequences of intact PC and PC-(BC) were determined to be

METRRIRKVL, which was consistent with that deduced

from the DNA sequences The correct mutation of the

second amino acid residue, arising from the introduction

of an NcoI site in the start codon region, to glutamate from

the original lysine, was confirmed Likewise, the amino

acid sequence of PC-(CT +BCCP) was determined to be

ARRKDRGTKM, and this sequence was consistent with

that deduced from its DNA sequence except for the absence

of the first amino acid methionine It was also confirmed that

the second residue was properly converted to alanine from

original proline, because of the design of the expression

plasmid

In SDS/PAGE, the bands of intact PC and PC-(BC) were

observed at the positions corresponding to the molecular

masses deduced from their sequences, 128.5 and 51.4 kDa,

respectively, while that of PC-(CT +BCCP) was observed

at a position (65 kDa) considerably smaller than that

expected (77.1 kDa) To confirm the integrity of

PC-(CT +BCCP), this protein was analyzed by MALDI

TOF mass spectrometry together with PC-(BC) The mass

(m/z value) obtained was 77 047 ± 76 for PC-(CT +

BCCP) and 51 428 ± 53 for PC-(BC) (mean ± SD from

three determinations) These values are identical, within

experimental error, to the molecular masses deduced from

their sequences, 77 082 and 51 438 Da, respectively,

prov-ing that the two engineered proteins have the correct

structure

Molecular properties of the engineered proteins of PC

Association states of the proteins were investigated by high

performance gel filtration chromatography Apparent

molecular masses of the samples were estimated on the

basis of the calibration curve obtained by using a set of

standard proteins (Fig 3) Typical elution profiles of intact

PC, PC-(BC) and PC-(CT +BCCP) are shown in Fig 4

For intact PC, two peaks were observed at 13.76 min and

17.02 min (Fig 4A) and the apparent molecular masses

estimated from their retention times were 501.3 ± 11.5 kDa and 137.0 ± 4.8 kDa (mean ± SE from three separate experiments), which were considered to be the tetramer and monomer, respectively The intensity of the tetramer peak was about 10 times greater than that of the monomer and this ratio did not change with protein concentration over the range adopted (1–100 lM), verifying that intact PC exists mainly as a tetramer, which is typical for single polypeptide type PCs [4,27] Also for PC-(CT +BCCP), two peaks, major and minor, were observed at 17.17 min and 18.82 min (Fig 4B) and the apparent molecular masses estimated from them were 128.2 ± 1.8 kDa and 66.2 ± 1.1 kDa, which appeared

to represent a dimer and monomer, respectively Again, the ratio of the intensity of the two peaks (10 : 1), did not change with the protein concentration

By contrast, the behavior of PC-(BC) on gel filtration chromatography was different from those of the above two proteins Although two peaks were also observed for PC-(BC), the ratio of the intensity of the peaks changed markedly with the protein concentration At a high concentration (100 lM), a major peak was observed at 18.28 min (Fig 4C) and the molecular mass estimated from this peak was 81.9 ± 2.9 kDa, which appeared to represent

a dimer On the other hand, at a low concentration (5 lM), a major peak was observed at 19.41 min (Fig 4D) and the molecular mass estimated from it was 66.2 ± 1.1 kDa, which appeared to represent the monomer

Moreover, the mixtures of PC-(BC) and PC-(CT + BCCP) at various ratios were analyzed to study their interaction, but no new peak was observed other than those derived from the constituent proteins, suggesting that

Fig 3 Estimation of the molecular masses of intact and engineered PC

by gel filtration chromatography on the TSK G3000SWXL column The molecular masses of the proteins used for construction of the calib-ration curve (s) were ribonuclease A (13.7 kDa), chymotrypsino-gen A (23 kDa), ovalbumin (43 kDa), albumin (67 kDa), aldolase (158 kDa) and thyroglobulin (669 kDa) Elution times of intact PC and engineered proteins are represented by d, as labelled.

Fig 2 SDS/PAGE (A) and avidin-blot analysis (B) of purified intact

PC and engineered proteins (A) SDS/PAGE was run with 12.5%

polyacrylamide and 0.5 lg of proteins: M, marker; lane 1, intact PC;

lane 2, PC-(BC); lane 3, PC-(CT +BCCP) (B) SDS/PAGE was run

with 0.1 lg of proteins and electroblotted onto the membrane The

proteins carrying biotin were detected by the reaction with alkaline

phosphatase-conjugated avidin: lane 1, intact PC; lane 2,

PC-(CT +BCCP).

PC-(BC) and PC-(CT +BCCP) do not interact signifi-cantly under the experimental conditions employed Acetyl-CoA had almost no effect on the association of the two proteins, as the elution profile was hardly affected by preincubation of the samples with 0.1 mM acetyl-CoA followed by elution with buffer containing the same concentration of acetyl-CoA

Enzymic activity of intact PC Pyruvate carboxylase activity of intact PC was assayed by measuring the oxalacetate production in the presence of malate dehydrogenase and NADH as described under Experimental procedures The Km values determined for bicarbonate and pyruvate were 29.9 ± 1.4 mM and 0.31 ± 0.03 mM, respectively (estimate ± standard error from the nonlinear regression analysis), which were virtually identical to those of the literature, 28.6 mMfor bicarbonate and 0.33 mM for pyruvate [28] In the kinetic analysis of ATP, substrate inhibition was evident at high ATP concentration, and thus two kinds of data analysis were conducted as described in Experimental procedures The Km obtained from the data where substrate inhibition is insignificant was 0.46 ± 0.06 mM, while the value obtained from the whole data based on Eqn (1) that takes into account substrate inhibition was 0.87 ± 0.11 mM The Km for ATP of PC from the same source, obtained by simple Michaelis–Menten analysis on the data where substrate inhibition is not evident, was reported as 0.38 mM[28], close

to the corresponding value of the present work Also, the effect of acetyl-CoA on the pyruvate carboxylase reaction was nearly the same among the present work and literature; the activity was greatly increased upon addition of acetyl-CoA and the activity in the absence of acetyl-acetyl-CoA was approximately 0.3% of the maximum (Table 1)

PC is known to catalyze the cleavage of ATP in the absence of pyruvate [21] It is hence possible to study the reaction of BC (Scheme 1) independently of the CT reaction (Scheme 2) by measuring this activity This activity of PC was determined under essentially the same conditions as the complete reaction except for the omission of pyruvate (Table 1) The rate of the ATP cleavage reaction is about 0.2% of that of the complete reaction of PC, which almost coincides with that reported for chicken liver enzyme [21]

Fig 4 Typical elution profiles for intact PC (A), PC-(CT +BCCP) (B) and PC-(BC) (C and D) on the TSK G3000SWXL gel filtration column Proteins were chromatographed over a concentration range of 1–100 l M under the conditions described in Experimental procedures.

M, D and T denote monomer, dimer and tetramer, respectively (A) Two peaks were observed for intact PC at 13.76 and 17.02 min, which correspond to the tetramer and monomer, respectively, and the elution profile did not change over the concentration range examined (B) Two peaks were also observed for PC-(CT +BCCP) at 17.17 and 18.82 min, which correspond to the dimer and monomer, respectively, and the elution profile was not dependent on the protein concentra-tion (C) and (D) The elution profile of PC-(BC) was markedly dependent on the protein concentration: at the high concentration [100 l M , (C)], the dimer predominated, while at the low concentration [5 l M , (D)], the monomer predominated.

It was found that activity increased about 10-fold by the

addition of acetyl-CoA to 0.1 mM, thus this ATP cleavage

reaction is also dependent on acetyl-CoA Similar

depend-ence on acetyl-CoA was also observed previously [21]

Enzymic activity of PC-(BC)

It was found that the truncated enzyme, PC-(BC), is as

capable of mediating ATP cleavage in the presence of free

D-biotin as intact PC, suggesting that its three-dimensional

structure remains intact even in the absence of other

domains The enzymic activity of PC-(BC) in the presence of

various concentrations of biotin, bicarbonate and ATP are

depicted in Fig 5 As expected from the reaction

mechan-ism proposed for BC [2,29,30], this enzymic reaction was

completely dependent on three substrates; it is worthy of

noting that biotin is necessary for this reaction to proceed

(the activity in the absence of biotin is about 2% of

maximum in its presence) This subject is discussed in more

detail below Kinetic parameters for the three substrates

were determined from the data shown in Fig 5 The Kmfor

bicarbonate was 62.2 ± 5.3 mM, which was comparable

to that for the complete reaction of intact PC (29.9 ± 1.4 mM) In the kinetics for ATP, substrate inhi-bition was observed just like in intact PC, and the Kmvalues determined based on the simple Michaelis–Menton equa-tion and Eqn (1), were 0.54 ± 0.04 mM and 1.03 ± 0.15 mM, which were close to those of intact PC (0.46 ± 0.06 mMand 0.87 ± 0.11 mM) The Kmvalue for biotin of 50.9 ± 5.4 mMis considerably smaller than that

of the BC subunit of acetyl-CoA carboxylase from E coli (135 mM) [31] To investigate the effect of acetyl-CoA on this reaction, the assay was carried out under the standard conditions but omitting acetyl-CoA, and the data obtained are shown in Table 1 Unexpectedly, the activity of PC-(BC)

in the absence of acetyl-CoA was virtually unchanged from that in its presence In other words, the ATP cleavage activity of PC-(BC) is not dependent on acetyl-CoA, in sharp contrast to that of intact PC

Enzymic activity of PC-(CT + BCCP)

It was reported that oxamate stimulated the decarboxyla-tion of oxalacetate by PC [22] It is hence possible to study the CT reaction (Scheme 2) of PC separately from the BC reaction (Scheme 1) with this assay [22] The enzymic activity of PC-(CT +BCCP), investigated by measuring the oxalacetate decarboxylase activity in the presence of oxamate, increased with an increase in oxamate concentra-tion, as expected (Fig 6A) The activity in the presence of a saturating concentration of oxamate was about 40 times higher than that in its absence (Table 2) The effect of oxalacetate concentration on the decarboxylation reaction

at a fixed concentration of oxamate is shown in Fig 6B In this case, substrate inhibition at high concentration of oxalacetate is evident from the profile Such a phenomenon

Fig 5 Kinetic analysis for the ATP cleavage activity of PC-(BC) Activity of PC-(BC) (0.22 mg in 1 mL) was assayed with free biotin as the substrate in 100 m M Tris/HCl (pH 8.0) containing 5 m M MgCl 2 , 100 m M KCl, 0.1 m M acetyl-CoA, 0.5 m M phosphoenol pyruvate, 0.15 m M NADH, 5 units of lactate dehydrogenase, 5 units of pyruvate kinase and variable concentrations of ATP, bicarbonate and biotin at 30 C (A) Biotin was the variable substrate with 2 m M ATP and 100 m M bicarbonate; the K m for biotin was 50.9 ± 5.4 m M and the V max

2.39 ± 0.12 UÆlmol)1 Kinetic parameters were determined by fitting Eqn (2) to the data (B) Bicarbonate was the variable substrate with 2 m M ATP and 100 m M biotin; the K m for bicarbonate was 62.2 ± 5.3 m M and the V max 2.58 ± 0.12 UÆlmol)1 Kinetic parameters were determined by fitting the simple Michaelis–Menten equation to the data (C) ATP was the variable substrate with 100 m M bicarbonate and 100 m M biotin In the kinetics for ATP, substrate inhibition was evident, and thus two different kinds of analysis were made for the obtained data Kinetic parameters determined with the data from 0–1.0 m M on the basis of simple Michaelis–Menten were as follows: K m 0.54 ± 0.04 m M and V max

2.39 ± 0.10 UÆlmol)1, while those determined with the data from 0–5.0 m M on the basis of Eqn (1) were as follows: K m 1.03 ± 0.15 m M and V max

3.91 ± 0.40 UÆlmol)1 The theoretical curve shown in this figure was drawn on the basis of Eqn (1) In each case, the standard errors in V max and

K were determined from the nonlinear regression analysis.

Table 1 Effect of 0.1 m M acetyl-CoA on the pyruvate carboxylation of

intact PC and on the ATP cleavage reactions of intact PC and PC-(BC).

One unit of enzyme activity was defined as the amount of enzyme

required to catalyze the formation of 1 lmol of each product per min.

Values are the means ± SD from three separate experiments.

Protein Activity

Enzymic activity (UÆlmol)1) With acetyl-CoA Without acetyl-CoA Intact PC Overall 1220 ± 50 4.52 ± 0.36

Intact PC ATP cleavage 2.84 ± 0.21 0.31 ± 0.04

PC-(BC) ATP cleavage 1.02 ± 0.06 0.99 ± 0.07

was observed also for PC from chicken liver and was

accounted for by competitive substrate inhibition [22] The

decarboxylation activity of PC-(CT +BCCP) was found

to be of similar magnitude to that of PC (Table 2), and thus

PC-(CT +BCCP) retains the enzymic activity present in

the native structure despite lacking the BC domain It

is noted that the decarboxylation activity of PC-(CT +BCCP) and intact PC was virtually the same in the presence and absence of acetyl-CoA under the standard conditions used in this study Therefore, the catalytic reaction of the CT domain appears to be independent of acetyl-CoA, just like the ATP-cleavage reaction of PC-(BC)

Discussion

In general, single polypeptide-type PCs exist in the tetra-meric and subunit-type PCs in octatetra-meric form [1,2], but little is known as to how these oligomeric structures are formed PC consists of three domains, BC, CT and BCCP, but again little is known about how these domains are organized three-dimensionally to generate active enzymes These are the subjects addressed in this article It was found that the separated BC and CT +BCCP domains of the former type of PC from B thermodenitrificans retain their own catalytic activity, demonstrating that these two domains are independent entities as a protein Moreover, from the elution profiles of gel filtration HPLC, PC-(CT +BCCP) was found to exist mainly as a dimer, while PC-(BC) was found to exist as a monomer or a dimer depending on its concentration In other words, both engineered proteins associate with themselves to form homodimers, and the association of PC-(CT +BCCP) seems to be stronger than that of PC-(BC) In addition, the association between PC-(BC) and PC-(CT +BCCP) was not observed under the experimental conditions examined, demonstrating that they do not possess strong affinity for each other Given that the same applies to intact PC as well,

it is deduced that the tetrameric form of PC is built up in the following way: first, a dimer of PC is formed through the association of each (CT +BCCP) domain of two proto-mers of PC, and subsequently, individual BC domains of the resulting two dimers associate to form a tetramer In other words, the tetrameric structure of PC appears to be constructed through the interaction of the same domains, namely BC with BC and (CT +BCCP) with (CT + BCCP) This hypothesis awaits verification by X-ray crystallographic analysis, which is under way in this laboratory

As for the reaction of BC, the following mechanism involving the formation of an enzyme–carboxylphosphate complex seems to be the most plausible one [2,29,30]: ATPþ HCO3þ enz-biotin Ð ð2O3POCO2enz-biotinÞ

ð2O3POCO2enz-biotinÞ Ð enz-biotin-CO2 þ Pi

Scheme 4 Apparently biotin is not required in the reaction of bicarbonate with ATP (Scheme 3), but it is essential for the putative carboxylphosphate intermediate to form Biotin appears to participate indirectly in this step by inducing a conformational change so as to dispose the active site residues in correct orientations

to undergo nucleophilic attack on the c-phosphate of ATP In the present work, ATP cleavage activity (Scheme 3) of PC-(BC) was investigated with free biotin

Fig 6 Oxalacetate decarboxylation reaction of PC-(CT +BCCP).

Activity of PC-(CT +BCCP) (0.39 mg in 1 mL) was assayed in

100 m M Tris/HCl (pH 8.0) containing 5 m M MgCl 2 , 100 m M KCl,

0.1 m M acetyl-CoA, 0.15 m M NADH, 5 units of lactate

dehydro-genase and variable concentrations of oxalacetate and oxamate at

30 C (A) Oxamate was the varied substrate with 0.1 m M oxalacetate.

(B) Oxalacetate was the varied substrate with 1.0 m M oxamate Error

bars represent the standard deviations from the mean of three

deter-minations.

Table 2 Effect of 0.1 m M acetyl-CoA on the oxalacetate

decarboxy-lase activity (UÆlmol)1) of PC-(CT +BCCP) and intact PC One unit

of enzyme activity was defined as the amount of enzyme required to

catalyze the formation of 1 lmol of pyruvate per min Values are the

means ± SD from three separate experiments.

Acetyl-CoA

PC-(CT +BCCP) Intact PC

+Oxamate – Oxamate +Oxamate – Oxamate

Present 6.62 ± 0.48 0.18 ± 0.03 3.88 ± 0.32 0.43 ± 0.05

Absent 6.42 ± 0.54 0.19 ± 0.04 3.78 ± 0.37 0.41 ± 0.06

as substrate As the activity of PC-(BC) was completely

dependent not only on bicarbonate but also on biotin, it

was confirmed that biotin is essential in the reaction of

bicarbonate with ATP

Although a large number of studies has been devoted

to clarifying the role of acetyl-CoA in the PC reaction

[1,2,32–34], little is known about its activation mechanism

It was found that acetyl-CoA did not affect the ATP

cleavage activity of PC-(BC), although it is essential in the

same reaction of the BC domain of intact PC Likewise,

oxalacetate decarboxylation reactions of PC-(CT +

BCCP) and intact PC were not dependent on

acetyl-CoA Taken together, it seems that acetyl-CoA

partici-pates in the reaction of BC but not of CT, and judging

from the disappearance of acetyl-CoA dependence in the

reaction of PC-(BC) with free biotin, acetyl-CoA may act

as a regulator in the interaction between the active site of

the BC domain and the biotin moiety of the BCCP

domain

Based on these arguments, it is tempting to propose the

following hypothesis: in the absence of acetyl-CoA, the

active site of the BC domain cannot interact with biotin of

the BCCP domain due to the spatial separation between the

active site and biotin; however, upon binding of acetyl-CoA,

a conformational change is induced, so that biotin can reach

the active site to carry out the catalytic reaction In the

reaction of PC-(BC) with free biotin, such a steric constraint

is absent; as a result, its acetyl-CoA dependence may be lost

Conformation changes of PC induced by acetyl-CoA have

been observed by various means such as electron

micros-copy [35,36], ultracentrifugation [37] and others [38] In

order to verify the above hypothesis, further investigation is

needed and studies using other engineered proteins as well

as X-ray crystallographic analysis are under way in this

laboratory

Acknowledgements

The authors are grateful to Ms Tomoko Ishiguro and Ms Masayo

Nonaka for their assistance with construction of the recombinant

plasmids.

References

1 Jitrapakdee, S & Wallace, J.C (1999) Structure, function and

regulation of pyruvate carboxylase Biochem J 340, 1–16.

2 Attwood, P.V (1995) The structure and the mechanism of action

of pyruvate carboxylase Int J Biochem Cell Biol 27, 231–249.

3 Wallace, J.C., Jitrapakdee, S & Chapman-Smith, S (1998)

Pyruvate carboxylase Int J Biochem Cell Biol 30, 1–5.

4 Barden, R.E., Taylor, B.L., Isohashi, F., Frey, W.H., Zander, G.,

Lee, J.C & Utter, M.F (1975) Structural properties of pyruvate

carboxylases from chicken liver and other sources Proc Natl

Acad Sci USA 72, 4308–4312.

5 Lim, F., Morris, P., Occhiodoro, F & Wallace, J.C (1988)

Sequence and domain structure of yeast pyruvate carboxylase.

J Biol Chem 263, 11493–11497.

6 Cohen, N.D., Duc, J.A., Beegen, H & Utter, M.F (1979)

Quaternary structure of pyruvate carboxylase from Pseudomonas

citronellolis J Biol Chem 254, 9262–9269.

7 Goss, J.A., Cohen, N.D & Utter, M.F (1981) Characterization of

the subunit structure of pyruvate carboxylase J Biol Chem 256,

11819–11825.

8 Cazzulo, J.J & Stoppani, A.O.M (1968) The regulation of yeast pyruvate carboxylase by acetyl-coenzyme A and L -aspartate Arch Biochem Biophys 127, 563–567.

9 Libor, S.M., Sundaram, T.K & Scrutton, M.C (1978) Pyruvate carboxylase from a thermophilic Bacillus Studies on the specificity

of activation by acyl derivatives of coenzyme A and on the properties of catalysis in the absence of activator Biochem J 169, 543–558.

10 Ashman, L.K., Keech, D.B., Wallace, J.C & Nielsen, J (1972) Sheep kidney pyruvate carboxylase Studies on its activation by acetyl coenzyme A and characterization of its acetyl coenzyme A independent reaction J Biol Chem 247, 5818–5824.

11 Kondo, H., Kazuta, Y., Saito, A & Fuji, K (1997) Cloning and nucleotide sequence of Bacillus stearothermophilus pyruvate carboxylase Gene 191, 47–50.

12 Henrikson, K.P., Allen, S.H.G & Maloy, W.L (1979) An avidin monomer affinity column for the purification of biotin-containing enzymes Anal Biochem 94, 366–370.

13 Kohanski, R.A & Lane, M.D (1990) Monovalent avidin affinity column Methods Enzymol 184, 194–200.

14 Buckley, J.J., Libor, S & Sundaram, T.K (1979) Biotin subunits

of acetyl CoA carboxylase and pyruvate carboxylase from a thermophilic Bacillus Arch Biochem Biophys 192, 396–404.

15 Jitrapakdee, S., Walker, M.E & Wallace, J.C (1999) Functional expression, purification, and characterization of recombinant human pyruvate carboxylase Biochem Biophys Res Commun.

266, 512–517.

16 Modak, H.V & Kelly, D.J (1995) Acetyl-CoA-dependent pyru-vate carboxylase from the photosynthetic bacterium Rhodobacter capsulatus: rapid and efficient purification using dye-ligand affinity chromatography Microbiol 141, 2619–2628.

17 Mukhopadhyay, B., Stoddard, S.F & Wolfe, R.S (1998) Purification, regulation, and molecular and biochemical charac-terization of pyruvate carboxylase from Methanobacterium ther-moautotrophicum strain DH J Biol Chem 273, 5155–5166.

18 Mukhopadhyay, B & Purwantini, E (2000) Pyruvate carboxylase from Mycobacterium smegmatis: stabilization, rapid purification, molecular and biochemical characterization and regulation of the cellular level Biochim Biophys Acta 1475, 191–206.

19 Morrison, J.F (1979) Approaches to kinetic studies on metal-activated enzymes Methods Enzymol 63, 257–294.

20 Levert, K.L., Lloyd, R.B & Waldrop, G.L (2000) Do cystein 230 and lysine 238 of biotin carboxylase play a role in the activation of biotin? Biochemistry 39, 4122–4128.

21 Attwood, P.V & Graneri, B.D.L.A (1992) Bicarbonate-dependent ATP cleavage catalysed by pyruvate carboxylase in the absence of pyruvate Biochem J 287, 1011–1017.

22 Attwood, P.V & Cleland, W.W (1986) Decarboxylation of oxalacetate by pyruvate carboxylase Biochemistry 25, 8181–8196.

23 Sambrook, J., Fritsch, E F & Maniatis, T (1989) Transfer of proteins from SDS-polyacrylamide gels to solid supports: immunological detection of immobilized proteins (western blot-ting) In Molecular Cloning: a Laboratory Manual, 2nd edn, pp 18.60–18.75 Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

24 Waldrop, G.L., Rayment, I & Holden, H.M (1994) Three-dimensional structure of the biotin carboxylase subunit of acetyl-CoA carboxylase Biochemistry 33, 10249–10256.

25 Thoden, J.B., Blanchard, C.Z., Holden, H.M & Waldrop, G.L (2000) Movement of the biotin carboxylase B-domain as a result

of ATP binding J Biol Chem 275, 16183–16190.

26 Kondo, H., Shiratsuchi, K., Yoshimoto, T., Masuda, T., Kitaz-ono, A., Tsuru, D., Anai, M., Sekiguchi, M & Tanabe, T (1991) Acetyl-CoA carboxylase from Escherichia coli: Gene organization and nucleotide sequence of the biotin carboxylase subunit Proc Natl Acad Sci USA 88, 9730–9733.

27 Libor, S., Sundaram, T.K., Warwick, R., Chapman, J.A &

Grundy, S.M.W (1979) Pyruvate carboxylase from a

thermo-philic Bacillus: some molecular characteristics Biochemistry 18,

3647–3653.

28 Cazzulo, J.J., Sundaram, T.K & Kornberg, H.L (1970)

Proper-ties and regulation of pyruvate carboxylase from Bacillus

stearo-thermophilus Proc R Soc Lond., B, Biol Sci 176, 1–19.

29 Knowles, J.R (1989) The mechanism of biotin-dependent

enzymes Annu Rev Biochem 58, 195–221.

30 Lynen, F., Knappe, J., Lorch, E., Jutting, G & Ringelmann, E.

(1959) Die biochemische Funktion des Biotins Angew Chem 71,

481–486.

31 Polakis, S.E., Guchhait, R.B., Zwergel, E.E & Lane, M.D (1974)

Acetyl coenzyme A carboxylase system of Escherichia coli J Biol.

Chem 249, 6657–6667.

32 Attwood, P.V (1993) Locus of action of acetyl CoA in the

biotin-carboxylation reaction of pyruvate carboxylase Biochemistry 32,

12736–12742.

33 Phillips, N.F.B., Sonswell, M.A., Chapman-Smith, A., Keech,

D.B & Wallace, J.C (1992) Isolation of a carboxylphosphate

intermediate and the locus of acetyl-CoA action in the pyruvate carboxylase reaction Biochemistry 31, 9445–9450.

34 Attwood, P.V & Wallace, J.C (1986) The carboxybiotin complex

of chicken liver pyruvate carboxylase A kinetic analysis of the effects of acetyl-CoA, Mg 2+ ions and temperature on its stability and on its reaction with 2-oxobutyrate Biochem J 235, 359–364.

35 Rohde, M., Lim, F & Wallace, J.C (1986) Pyruvate carboxylase from Saccharomyces cerevisiae Quaternary structure, effects of allosteric ligands and binding of avidin Eur J Biochem 156, 15–22.

36 Attwood, P.V., Mayer, F & Wallace, J.C (1986) Avidin as a probe of the conformational changes induced in pyruvate car-boxylase by acetyl-CoA and pyruvate FEBS Lett 203, 191–196.

37 Taylor, B.L., Frey, W.H., Barden, R.E., Scrutton, M.C & Utter, M.F (1978) The use of the ultracentrifuge to determine the cata-lytically competent forms of enzymes with more than one oligo-meric structure J Biol Chem 253, 3062–3069.

38 Frey, W.H & Utter, M.F (1977) Binding of acetyl-CoA to chicken liver pyruvate carboxylase J Biol Chem 252, 51–56.