Comparison of the amino acid sequences of the transacylase components of branched chain oxoacid dehydrogenase of Pseudomonas putida, and the pyruvate and 2-oxoglutarate dehydrogenases of Escherichia coli

Eur. J. Biochem. 176, 165-169 (1988)

(0FEBS 1988

Comparison of the amino acid sequences of the transacylase components of branched chain oxoacid dehydrogenase of Pseudomonas putida, and the pyruvate and 2-oxoglutarate dehydrogenases of Escherichia coli Gayle BURNS, Tracy BROWN, Kenneth HATTER and John R. SOKATCH Department of Biochemistry and Molecular Biology, University of Oklahoma Health Sciences Center (Received March 1/May 4, 1988)

-

EJB 88 0264

The nucleotide sequence of bkdB, the structural gene for E2b, the transacylase component of branched-chainoxoacid dehydrogenase of Pseudornonas putida has been determined and translated into its amino acid sequence. The start of bkdB was identified from the N-terminal sequence of E2b isolated from branched-chain-oxoacid dehydrogenase of the closely related species, P. aeruginosa. The reading frame was composed of 65.5% G C with 82.3% of the codons ending in G or C. There was no intergenic space between bkdA2 and bkdB. No codons requiring minor tRNAs were utilized and the codon bias index indicated a preferential codon usage. The hkdB gene encoded 423 amino acids although the N-terminal methionine was absent from E2b prepared from P. ueruginosa. The relative molecular mass of the encoded protein was 45134 (45003 minus methionine) vs 47000 obtained by SDS/polyacrylamide gel electrophoresis. There was a single lipoyl domain in E2b compared to three lipoyl domains in E2p, and one domain in E20, the transacylases of pyruvate and 2-oxoglutarate dehydrogenases of Escherichiu coli respectively. There was significant similarity between the lipoyl domain of E2b and of E2p and E20 as well as between the EI-E2 binding domains of E2b, E2p and E20. There was no similarity between the E3 binding domain of E2b to E2p and E20 which may reflect the uniqueness of the E3 component of branchedchain-oxoacid dehydrogenase of P. putida. The conclusions drawn from these comparisons are that the transacylases of prokaryotic pyruvate, 2-oxoglutarate and branched-chain-oxoacid dehydrogenases descended from a common ancestral protein probably at about the same time.

+

Branched-chain-oxoacid dehydrogenase is a multienzyme complex which is common to the metabolism of the branched chain oxoacids which are formed from valine, isoleucine and leucine by transamination. Branched-chain-oxoacid dehydrogenase complex performs three catalytic activities. The E l component is the dehydrogenase and catalyzes the decarboxylation of the oxoacid substrate and transfer of the branched-chain fatty acyl residue to E2 reducing the lipoyl prosthetic group of the latter subunit. E2 catalyzes transacylation of the fatty acyl group from the lipoyl residue of E2 to coenzyme A. E2 is the core of all of the oxoacid multienzyme complexes and binds the E l and E3 components. The E3 component is a lipoamide dehydrogenase which catalyzes the oxidation of the dihydrolipoyl residue of the E2 subunit and the reduction of NAD+. The overall reaction catalyzed by Correspondence to J. R. Sokatch, Department of Biochemistry and Molecular Biology, University of Oklahoma, Health Sciences Center, P.O. Box 26901, Oklahoma City, Oklahoma, USA-73190 Abhreviutions. E2b, E2p and E20 are the transacylase components of branched-chain-oxoacid, pyruvate and 2-oxoglu'tarate dehydrogenase respectively. The structural genes for the components of branched-chain-oxoacid dehydrogenase from Pseudomonas putida are hkdA1, E l i , hkdA2, EIP, hkdB, E2 and lpdV, E3 or LPD-Val. The structural genes for the transacylase (E2) components of pyruvate and 2-oxoglutarate dehydrogenases of Escherichia coli are aceF and sucB. The structural gene for lipoamide dehydrogenase of E. coli is lpd. Enzyme. Branched-chain-oxoacid dehydrogenase (EC 1.2.1.25).

branched-chain-oxoacid valerate is:

dehydrogenase

+

2-Oxoisovalerate + CoA NAD+ + Isobutyryl-CoA

with

2-oxoiso-

+ C 0 2 + NADH.

Branched-chain-oxoacid dehydrogenase has been purified from Pseudornonas putida and Pseudornonas aeruginosa [l, 21, from bovine kidney [3] and rabbit liver [4]. In all cases, the complete complex consists of four subunits, Elcl and E1B being responsible for the El catalytic activity and single proteins catalyzing E2 and E3 reactions. Pseudornonus utilizes a specific E3 component, LPD-Val [2, 51 for branched-chainoxoacid dehydrogenase in contrast to Escherichia coli which makes a single lipoamide dehydrogenase for both pyruvate and 2-oxoglutarate dehydrogenases [6]. The structural genes for the P. putida complex have been cloned [7] and the nucleotide sequence of the E l a and EIB structural genes determined [S]. There is a striking similarity between the amino acid sequence of Ela subunit and the corresponding subunit of branched-chain-oxoacid dehydrogenase of rat liver, including conservation of the phosphorylated region. There is also substantial similarity between the Pseudornonas EIP subunit and the EIB subunit of human pyruvate dehydrogenase. There is however, no similarity between either of these Elcr subunits and the E l subunits of pyruvate [9] and 2-oxoglutarate [lo] dehydrogenases of E. coli. The amino acid sequences of the E2 components of pyruvate and 2-oxoglutarate dehydrogenases of E. coli, E2p

166 and E20 respectively, have been determined from the nucleotide sequences of their structural genes [II, 121. The objective of the present research was to compare the deduced amino acid sequence of the transacylase, E2b of branchedchain-oxoacid dehydrogenase of P. putida to the sequences of E2p and E20 to determine if any evolutionary relationships could be identified.

I

I

I

1

I

0

1

2

3

4

I 5

I

kb

MATERIALS AND METHODS Bacterial strains Recombinant plasmids and growth conditions are described in [7]. P. putida PpG2 was obtained from I.C. Gunsalus. E. coli JM109 was used as a host for MI3 phage containing recombinant DNA [13]. E. coli TBI was the host for pUC-derived plasmids and was obtained from Bethesda Research Laboratories (Gaithersburg, Maryland). E. coli TB1 is similar to strain JM83 except that the former is hsdR hsdM. Determination of nucleotide sequence The Sanger chain-termination method [14] was used for the determination of nucleotide sequence with subcloning into M13mpl8 and M13mp19 [I31 to determine the sequence of both strands. All clones overlapped so that sequences could be matched from adjacent restriction fragments. The clones for this purpose were pJRS10, which contains bkdAI and bkdA2, the structural genes for E l a and EIP subunits, and pJRS23, which contains IpdV, the structural gene for the E3 component [7]. The SphI -@hI restriction fragment which includes bases 283 - 2963 (Fig. I), the SphI - SphI fragment which includes bases 2963 - 3610 and the KpnI - KpnI fragment which includes bases 3735 - 4908 were also isolated and ligated into M13mp18 and M13mp19. These clones were also shortened by treatment with Ex0111 and S1 nucleases [15] to provide a series of subclones 2-300 bp shorther than the preceding clone. Analysis of nucleotide sequence Nucleotide sequence was analyzed using BIONET programs [16]. IFIND is a program which compares a short sequence with sequences in a data base and is useful for identifying domains. GENALIGN is a program which aligns entire sequences and is frequently used for evolutionary studies. GENALIGN searches for similar internal sequences which are in corresponding parts of the whole. The program provides a score indicating degree of relationship, and a plot showing the alignments. The deduced amino acid sequences of E2p [9] and E20 [lo] were obtained by translation of the nucleotide sequences of the structural genes from the GenBank and EMBL databases using SEQ. The latter included a correction to the data originally published. The methods for the isolation of the E2b component from P. aeruginosa, including purification by HPLC and N-terminal analysis of the isolated protein, will be described in [8]. Purified E2b eluted at 39.75 min in thesystem used. RESULTS AND DISCUSSION Nucleotide sequence of bkdB The relationship of the structural gene for E2b, bkdB, to the genes for the components of branched-chain-oxoacid

Fig. 1 . Strategy used to determine the nucleotide sequence of bkdB, the structural gene for the E2 component of branched-chain-oxoacid dehydrogenase q f P . putida PpGZ. The open and diagonally-filled bars represent the structural genes for the indicated components. The arrows represent the length of subclones in M13 and direction of reading of nucleotide sequence. The abbreviations used for the restriction enzymes are E, EcoRI; K, KpnI; P, PstI; S, SstI; Sa, S a n ; Sp, SphI

dehydrogenase of P. putida is shown in Fig. 1. The direction of transcription is from left to right in Fig. 1 [7]. All of bkdB has been cloned into pUC19 producing pJRS24 [7] and is expressed in E. coli TBI. E2b activity can be detected by complementation of extracts lacking E2b, but containing El and E3. The start of bkdB was identified with the aid of the Nterminal sequence of E2b purified by HPLC. The source of E2b was P. aeruginosa since we have not yet been able to isolate E2b from P. putida [2]. The N-terminal sequence of E2b from P. aeruginosa closely matched the translated sequence beginning at base 2333 with the exception of the absence of Nterminal methionine and glutamic acid in place of glutamine at residue 17 of the translated sequence (Fig. 2). There is no intergenic space since the stop codon for bkdA2 is the immediate preceding triplet. For comparison, the intergenic distance between the aceE and aceF genes and sucA and sucB in E. coli is 14 bases in each case [9, 101. There is a potential ribosome binding site, GGAGG, with five intervening bases before the start of the coding region. The open reading frame which begins at base number 2333 and continues through to the stop codon ending at 3604, encodes 423 amino acids. The relative molecular mass of the encoded protein was 45134 (45003 minus methionine) vs 47000 obtained using SDS/ polyacrylamide gel electrophoresis [I]. The G C content of bkdB (Table 2) was 65.5% with 82.3% of the codons ending in G or C both of which are typical of Pseudomonas genes including bkdAl and bkdA2 (our unpublished results). The hkdB gene was examined for sequences with dyad symmetry using the BIONET program SEQ which identified a total of 20 such sequences, 17 of which had free energies of formation of greater than - 84 kJ. The greatest free energy of formation was -163.9 kJ for a hairpin which extended from 2684 to 2761 with a loop of 28 bases. This is a much larger number of hairpins than the five found in the genes for Elm and ElP (our unpublished results). It is not clear at present if any of these stem and loop structures have regulatory functions.

+

167 2310

2320

2330

2347

2362

GTGCGGCATI GAAAAAGGTC TGGGGTCT GA ATG GGC ACG CAC GTC ATC AAG ATG CCG GAC SD M G T H V l R M P 2 2377

2392

2407

ATT GGC GAA GGC ATC GCG CAG GTC GAA TIY: GTG GAA TGG TTC GTC AAG GTG GGC GAC

LGECIAQVE'ixEWFVKVGD 2422

2437

2452

2467

ATC ATC Gcc GAG GAC CAA G1y: GTA GCC GAC Gl'C ATG ACC GAC AAG GCC ACC GTG GAA I I A E D Q V V A D V M T D K A T V E 2482

2497

2512

2557

2572

2587

ATG GCG GTC GGC AGT GAG Cl'G ATC CGC ATC GAA GTG GAA GGC AGC GGC AAC CAT GTG M A V G S E L I R I E V E G S G N H V 2602

TlT-Phe

0

TCT-Ser

0

TAT-Tyr

3

TGT-Cys

0

TIC-Phe

8

TCC-Ser

2

TAC-TJ r

2

TGc-Cys

1

TTA-Leu

0

TCA-Ser

2

TAA-.

0

TGA-

TTGLeu

2

TCG-Ser

4

TAG-.

0

TGG-Trp

2

CiT-Leu

0

CCT-Pro

2

CAT-His

3

CGT-Arg

3

CTC-Leu

5

CCC-Pro

5

CAC-His

8

CGC-Arg

21

CI'A-Leu

0

CCA-Pro

2

CAA-Gln

6

CGA-Arg

0

CTGLeu

25

CCG-Pro

18

CAG-Gln

14

CGGArg

2

.

1

2527

A'E CCG Ta: CCG GTC AGC GGC AAG GTG CTG GQ: CTG GGT GGC CAG CCA GGT GAA GTG I P S P V S G K V L A L G G Q P G E V 2542

Table 1 . Codon utilization by bkdB

2617

2632

2647

GAT GTG CCG CAA GCC AAG CCG GCC GAA GTG CCI Gffi GCA CCG GTA GCC GCX AAA CCT D V P Q A K P A E V P A A P V A A K P 2662

2677

2692

GAA CCA CAG AAA GAC GIT AAA CCG GCG GCG TAC CAG GCG TCA CCC AGC CAC GAG GCA E P Q K D V K P A A Y Q A S A S H E A 2707

2722

2737

2752

GCG CCC ATC GTG CCG CGC CAG Cffi GGC GAC AAG Cffi Cl'G GCC TCG CCG GCG GTG CGC A P I V P R Q P G D K P L A S P A V R 2767 AAA

K

2782

2027

R

2842

2857

2872

A'E Cn: CAC GAA GAC C'E GAC GCG TIC ATG AGC AAA Cffi CAA AGC GCI GCC I L H E D L D A F M S K P Q S A A

2887 2902 2917 2932 GGG CAA ACC-& AAT GGC TAT GCC AGG OGC ACC GAC AGC GAG CAG GTG CCG GTG ATC

G

Q

T

P

N

G

Y

A

R

R

T

D

3

ACT-Thr

0

AAT-Asn

2

AGT-Ser

1

21

ACC-Thr

12

AAC-Asn

10

AGC-Ser

15

ATA-Ile

1

ACA-Thr

1

AAA-Lys

5

AGA-Arg

0

ATGMet

11

ACG-Thr

2

AAG-Lys

12

AGGArg

1

GTT-Val

2

UT-Ala

3

GAT-Asp

4

GGT-Gly

6

GTC-Val

15

GCC-Ala

31

GAC-Asp

19

GGC-Gly

27

GTA-Val

3

GCA-Ala

3

GAA-Glu

19

GGA-Gly

0

GTGVal

29

GCG-Ala

15

GAG-Glu

8

GGGGly

2

2812

CGC GCC CTC GAT GCC GGC ATC GAA TIY: CGT TAT GTG CAC GGC AGC GGC CCG GCC R A L D A G I E L R Y V H G S G P A

GGG CGC

G

2797

ATT-Ile ATC-Ile

S

E

Q

V

P

V

I

2947

2962 2977 GGC Cn: CGC CGC AAG ATC GCC CAG CGC ATG CAG GAC GCC AAG CGC CGG GTC GCG CAC

G

L

R

R

K

2992

I

A

Q

R

M

Q

D

A

K

R

R

V

A

H

3037 TIC AGC TAT GTG GAA GAA ATC GAC GTC ACC GCC CTG GAA GCC CTG CGC CAG CAG CTC

F

3007

S

Y

V

E

E

3052

3022

I

D

V

T

A

3067

L

E

A

L

R

3082

Q

Q

L

3097

AAC AGC AAG CAC GGC GAC AGC CGC GGC AAG CTG ACA Cn: CTG CCG TIC Cn: GTG CGC N S K H G D S R G K L T L L P F L V R 3112 3127 GCC CTG GTC GTG CCA Cn: CGT GAC TTC CCG

A

L

V

V

A

L

R

D

3172

F

P

3142

3157

CAG ATA AAC GCC ACC TAC GAT GAC GAA Q I N A T Y D D E

3187

3202

3217

GCG CAG ATC ATC ACC CGC CAT GGC GCG GTG CAT GTG GGC A X GCC ACC CAA GGT GAC A Q I I T R H G A V H V G I A T Q G D 3232

3247

3262

AAC CXC CTG ATG GTA (lcc GTG CTG CGC CAC GCC GAA GCG GGC AGC (SIT: TGG GCC AAT N G L M V P V ' L R H A E A G S L W A N 3277

3292

GCC GGT

A

G

3307

3322

GAG ATT TCA CGC CTG Gcc AAC GCT Gffi CGC AAC AAC AAG Gcc AGC CGC GAA E I S R L A N A A R N N K A S R E

3337

3352

3367

3382

GAG CXG TCC GGT TCG ACC A'IT ACC CTG ACC AGC CTC CGC GCC (SIT: GGC GGC ATC GTC E L S G S T I T L T S L G A L G G I V 3397

3412

3427

3442

AGC Affi CCG GTG GTC AAC ACC Cffi GAA GTG Gffi ATC GTC GGT G'E AAC CGC ATG G m S T P V V N T P E V A I V G V N R M V 3457

3472

3487

3502

GAG O X CCC GTG GTG ATC GAC GGC CAG ATC GTC GTG CGC AAG ATG ATG AAC CTG TCC E R P V V I D G Q I V V R K M M N L S 3517

3532

3547

AGC Ta: TIC GAC CAC CGC GTG GTC GAT GGC ATG GAC Gcc GCC Cn: TIC ATC CAG GCC S S F D H R V V D G M D A A L F I Q A 3562 3577 3592 3614 GTG CGT GGC CTG CTC GAA CAA CCC GCC TGC Cl'G TIC GTG GAG TGA GCATGCAACA

V

R

G

L

L

E

Q

P

A

C

L

F

V

E

*.

Fig. 2. Nucleotide sequence of bkdB and the trunsluted amino acid sequence. The underlined amino acids were identical to the corresponding N-terminal amino acids of E2b isolated from P. aeruginosu. The numbering begins with the SstI site immediately preceding the start of the coding region (Fig. 1). SD is the Shine-Dalgarno ribosome binding site

Codon bias Codon usage by bkdB was biased indicating a strongly expressed gene (Table 1) and the bias was quantified by calculation of the codon bias index [17]. The extreme examples of codon bias index are completely random codon usage, in which case the index is zero, and completely biased, i.e. only one of the codons is used, in which case the index is 1.0. The index for bkdB was 0.544. The index for bkdAI was 0.648 and for bkdA2, 0.627 [8]. The indices for aceF and sucB were calculated from data in [9, 101 and were 0.4335 and 0.407 respectively. The index calculated from pooled codon usage data for four RNA polymerases, which are considered to be strongly expressed genes, was 0.607 [18]. The conclusion is that E2b, E2p and E20 are moderately to strongly expressed genes. Codons recognized by minor tRNAs of E. coli were not used by bkdB. Comparison of amino acid sequences of transacylases of oxoacid dehydrogenases E2p and E20 are organized into three functional regions, the lipoyl domain, the E3 binding domain and the El-E2 binding domain [19,20] (Fig. 3). The lipoyl domains are located at the N-terminal region and are thought to be mobile, allowing them to react sequentially with E l and E3 components of oxoacid dehydrogenases. The hinge region, which connects the lipoyl domains to the inner core of E2 is rich in alanine and proline providing the necessary flexibility (Fig. 3). E2p of E. coli contains three lipoyl domains of approximately 100 amino acids each [9] while E20 of E. coli contains a single lipoyl domain [lo]. It is not obvious if possession of three lipoyl domains is an advantage to E2p since deletion of two of the domains as well as the hinge region connecting the lipoyl domains to the rest of the protein does not appear to diminish activity of the pyruvate dehydrogenase

168

*

--r-,-

E2p2 E2pl E2p3 E2b E20

105

kdvnVPDIG sDEVEvTEILVKVGDKVEAESLITVEGDKASMEVPaPfAGtVKEIKVnVGDKvs~sL

I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I II1111 IIII I I I I IIIII I I I I II IIIIIII IIIIIIIIIIIIIIIIII I I I I ll II IIII I1 I 205 vkEvnVPDIG gDEVEvTEvmVKVGDKVAA~SLITVEGDKASMEVPaPfAGv~ElKVnVGDKvkTGsL I IIII II I IIIII I I I I l l I I 1 gthvIkmPDIGEgiAqVElvEWfVKVGDiiAeDQVvadVmTDKAtvEiP spvsgkVLalggqpge I I I I I I I I I I I Ill I I I 1 aiEikVPDIG aDEVEiTEILVKVGDKVEAELITVEGDgftSMEVPsPqAGiVKEI~sVGDKtq~aL

1 ssvdIlvPDlpEsvAdatvatWhkKpGDavvrDeVlveieTDKvvlEvP

asadgiLdavledeg

*

E2p2

173 IMvFevageagA AAPA

E2pl

70 IMIFd sadgAAdAAPA

II I IIII

aRdEAAPAAA;ipAAg.

I IIII IIIII

1

1 IIIIIIIII I I I I IIIIIII I

II

I

qAeeKKEAAPAAAPAAAAa

E2p3