May 5, 2006 ... pSLl = 1.6-kb cDNA in pUCl9 EcoRI-EcoRI. pSL2 = 1.6-kb cDNA in pGEM 4
EeoRI-EcoRI. pKB3 = 2.9-kb cDNA from 3.0-kb cDNA clone in.
THEJOURNAL OF BIOLOGICAL CHEMISTRY
Vol. 264, No. 13, Issue of May 5,pp. 7742-7746,1989 Printed in U.S.A.
0 1989 by The American Society for Biochemistry and Molecular Biology, Inc.
Construction and Nucleotide Sequence ofa cDNA Encoding the Fulllength Preproteinfor Human Branched Chain Acyltransferase” (Received for publication, September 26, 1988)
Dean J. Danner, Stuart Litwer, W. Joseph Herring$, and Janet Pruckler From the Division of Medical Genetics,Department of Pediatrics, Emory University School of Medicine, Atlanta, Georgia 30322
A cDNA (1.6 kilobases) for branched chain acyltransferase (E2b) isolated from a human liver library encoded only the amino-terminal half of the protein (Hummel, K. B., Litwer, S., Bradford, A. P., Aitken, A., Danner, D. J., and Yeaman, S . J. (1988) J. Biol. Chem. 263,6165-6168). Here we report the isolation of other cDNAs which encode the carboxyl-terminal half of E2b and the construction of a cDNA which encodes the entire pre-E2b. cDNA from the original clone encoding the leader sequence, lipoate binding domain, and E3 binding domain was ligated to the cDNA from a clone which by restriction maps contained an additional 3’ sequence. Both cDNAs used in the construct made a fusion protein in their original phage isolate recognized by antibodies to E2b. The nucleotide sequence of the constructed cDNA was determined, and the 1431 base pairs in the open reading frame encoded a protein of 477 amino acids. In vitro transcription and translation of this cDNA produced a 57-kDa protein recognized by E2b-specific antibodies. Mouse liver mitochondria imported and processed the 57-kDa protein to a 52-kDa antigenic protein which co-migrated with E2b isolated from tissue. Comparing the protein structure of this human pre-E2b protein with that for other acyltransferase proteins showed a similarity in structure throughout all the proteins suggesting evolutionary conservation. Branched chain acyltransferase from Pseudomonas putida showed the most similarity to human E2b.
Branched chain acyltransferase (E2b) forms the core of the branched chain a-ketoacid dehydrogenase complex, a component of mitochondria in eukaryotes. Analysis of a 1.6kilobase (kb)’ cDNA for E2b is previously reported (1, 2). This 1.6-kb cDNA encodes an in uitro made protein with an M , value of 39,000, and the protein isimported and processed by mitochondria to a 36-kDa protein (3). However, mature E2b isolated from tissue as a component of the branched chain a-ketoacid dehydrogenase complex migrates with an M , value of 52,000. Analysis of the protein structure shows the human E2b encoded by the 1.6-kb cDNA contains the leader
* This work was supported by Grant DK38320 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solelyto indicate this fact. The nucleotide sequence(s)reported in this paper has been submitted to the GenBankTTM/EMBL Data Bank withaccessionnumber($ J04 723. $Supported by Predoctoral Grant 18-88-20 from the March of Dimes. I The abbreviations used are: kb, kilobase(s); SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; bp, base pair(s).
sequence, lipoate binding domain, and theE3 binding domain but lacks a major portion of the carboxyl-terminal end of the protein, tprmed the catalytic domain. The catalytic domain is thought to contain the binding sites for the other subunits of the complex and a highly conserved amino acid sequence reported to be the CoASH binding site (4, 5). Here we report the isolation of other longer cDNAs for E2b which hybridize to the 1.6-kbcDNA.We used the 1.6-kb cDNA and a3.0-kb cDNA to construct acDNA whichencodes full-length pre-E2b. The protein from this constructedcDNA, made in uitro, is imported and processed by mitochondria to mature E2b with an M, of 52,000. Analysis of the deduced amino acid sequence and protein structurewhen compared to other acyltransferase proteins confirmed the identity of this pre-E2b protein. MATERIALSANDMETHODS
Isolation of cDNA-The 1.6-kb cDNA was isolated by antibody selection from a human liver cDNA library in Xgtll (6). This cDNA was subcloned into pUC19 (pSL1) for production of large quantities of the 1.6-kb fragment for use as a probe in DNA-DNA hybridization for selection of other clones. The fragment was labeled with [LY-~*P] CTP by the oligomer primer extension reaction (7). Additional cDNAs were isolated from a hepG2 cDNA library in Xgtll (8) (see Fig. 1). The following is a summary of the plasmids and the cDNA inserts of the pre-E2b sequence used in these studies. Since a fulllength pre-E2b was constructed from the 1.6- and 3.0-kb cDNAs, the other cDNA clones were not used. pSLl = 1.6-kb cDNA in pUCl9 EcoRI-EcoRI pSL2 = 1.6-kb cDNA in pGEM 4 EeoRI-EcoRI pKB3 = 2.9-kb cDNA from 3.0-kb cDNA clone in pUC19 XbI-EcoRI pSL5 = 1.9-kb cDNA construct in pGEM 4 (see Fig. 2) pKB5 = 1.9-kb cDNA from pSL5 in pGEM 7Zf(+) p J P l = 1.9-kb cDNA from pSL5 in pGEM 3Zf(-) pJP2 = 1.6-kb cDNA (KpnI(1400)-EcoRI(3000))from pKB3 in pGEM 3Zf(+) pJP3 = 1.6-kb cDNA from pJP2 in pGEM 3Zf(-) pJHl = 608 bp of pSLl (EcoRI(O)-EcoRV(608))in pGEM 4 Sequence Analysis-Plasmids p J P l and pJP3 were used with an Erase-a-base System (Promega) for size reduction according to instructions of the manufacturer. Nucleotide sequence in both directions was determined using the Sequenase kit (USB).The determined sequence was analyzed and translated using the Genepro software programs, Riverside Scientific, on an Emory XT-PC. Mitochondria Isolation-Mouse liver mitochondria were isolated by standard differential centrifugation as previously described (3). In Vitro Transcription and Translation-Transcripts of linearized pSL5 were produced from the SP6 promoter according to directions with the Promega Gemini System I1 kit. Translation of these transcripts was done with arabbit reticulocyte lysate kit (Promega) according to the instructions using [35S]methionineas radiolabel. Immunoprecipitation-In uitro translation products were subjected to immunoprecipitation and resolved by SDS-PAGE for analysis by autoradiography as previously described (3). RNA Analysis-RNAwas isolated from human liver or human
7742
Entire cDNA Human forBranched Chain
Acyltransferase
7743
-.
lymphoblasts by standard methods (9). Poly(A+) RNA was isolated by chromatography on oligo(dT)-cellulose. After resolution of RNA by denaturing gel electrophoresis (7), thenucleic acid was transferred to Zeta-Probe (Bio-Rad) nylon membranes and probed with [a-"PI UTP-labeled antisense RNA from linearized p J H l using the T7 promoter to produce transcripts. RESULTS
Three cDNAs ranging in size from 1.9 to 3.0 kb were isolated from human cDNA libraries in Xgtll and purified. Restriction enzyme analysis demonstrated that these three cDNAs were longer than the original 1.6-kb cDNA by an extended 3' sequence (Fig. 1). The @-galactosidase fusion protein from phage containing the 3.0-kb cDNA fragment was recognized by antibodies to E2b (data notshown). Since we know the 1.6-kb cDNA encodes the leader sequence to direct this protein to themitochondria (1, 3), we constructed a cDNA using the first 608 bp from the 5' end of the 1.6-kb cDNA and 1300 bp from the 3' portion of the 3.0-kb cDNA -1.9 kb of (Fig. 2). The finalconstruct(pSL5)contained cDNA in the pGEM 4 vector. Transcripts made from the SP6 promoter of linearized pSL5 were used for in vitro translation with [35S]methionine as label. Protein products were immunoprecipitated with antibodies against E2b and analyzed by autoradiography after resolution of the proteins on SDS-PAGE. The major protein detected by this analysis had an M , value of 57,000 with a 0
400
800
1200
600
2000
2400
2800
3200
3, HXHPPV
P
E HXMPPY
p
If
F
EXHPPV
f'
K
K
V
EXMPPV
P
K
K
Y
f 1.6 €1.9 '
L
L L
b
1
2
3
4
FIG. 3. Autoradiograph of immunoprecipitated in vitro made pre-E2b and its mitochondrially processed products. Lane I shows translation products produced from transcripts from pSL5. Lane 2 shows the products after incubation with mouse liver mitochondria. Lane 3 is as in lane 2 except the mitochondria were treated with trypsin (1.25 mg/ml) after import of proteins. Lane 4 is purified bovine liver branchedchaina-ketoacid dehydrogenase stained with Coomassie Blue, the four bands being E3, E2b, Elba, and ElbP asindicated by marks in the right margin.
3600 bp
5. E
PreE2b-,. E2b-
F*.3
5: S O
FIG. 1. Restriction enzyme maps of cDNAs for human branched chain acyltransferase. The 1.6- and 2.3-kb cDNAs were isolated from the Woo (6) human liver cDNA library in X g t l l and the 1.9- and 3.0-kb cDNAs from the Mueckler HepG2 cDNA library in Xgtll (8).E, EcoRI; H , HindIII; X, XbaI; P, PstI;V , EcoRV; K, KpnI.
FIG. 2. Construction of composite cDNA in pGEM 4 which encodes full-length pre-E2b. Plasmid pKB3 contains the 3.0-kb cDNA, and pSL2 contains the 1.6-kb cDNA (see Fig. 1).
minor nonspecific low molecular weight protein being present in most translation mixtures (Fig. 3, lune 1). When these proteins were incubated with freshly isolated mouse liver mitochondria, a third protein was immunoprecipitated with an M , value of 52,000 (Fig. 3, lune 2 ) . The 52-kDa protein was protected from trypsin digestion, suggesting the 52-kDa protein was sequestered inside the mitochondria (Fig. 3, lune 3 ) . This protectedprotein co-migrated with mature E2b. Compare Fig. 3, lune 3, with the resolved branched chain aketoacid dehydrogenase proteins from bovine liver in lune 4 (10). Nucleotide sequence analysis of the cDNA insert in pSL5 revealed an open reading frame of 1431bp encoding 477 amino acids with a summed molecular weight of 52,958(Fig. 4). Neither the constructed cDNA nor the parent 3.0-kb cDNA contains a polyadenosine sequence. The long stretch of 3'untranslated region giving rise to a cDNA in excess of 3000 bp is consistent with the size of mRNA seen in Northern blot analysis of human liver or lymphoblast poly(A+) RNA. Fig. 5 demonstrates that two poly(A+)transcripts of 3.4 and 2.5 kb hybridize to radiolabeled antisense RNA from pJH1. We do not know if both messages produce authentic pre-E2b. No attempt was made to quantitate the amount of mRNA for pre-E2b in these two tissues. Deduced amino acid sequence was used to predict the protein structure of pre-E2b for comparison to the structure of other acyltransferase proteins. As seen inFig. 6,the protein size and defined domain regions of these proteins align. Amino acid sequence of matureE2b from human was compared with the amino acid sequence of E2b from Psuedornonas putida. Identical matches were found in 31% of the sequence, and when conservative substitutions were included this number increased to 45% (Fig. 7). The regions most highly conserved included the E3 binding domain and the seven-amino acid sequence near the carboxyl-terminal end containing the histidine residue involved in CoASH binding. A conserved region unique to theE2b proteins inthe catalytic domain was found at residues 270-290of human E2b and
cDNA for Entire Human Branched Chain Acyltransferase
7744
-10 VrCIGcAGTCCGT
M
1
60
90 ATITGCTW6UICCT~GUGT~~GCTWTTTGTGTTC~TATTTTTUUUTGTffiTMTG~UTG~6UI~TUUT M
L
R
T
Y
S
R
Y
A
G
K
L
I
C
V
R
Y
F
E
T
C
G
Y
V
H
V
L
K
P
1 2
Y
120
180
T A T G T G T G ~ E m C U 7 A T C C ~ T T ~ G T A T A G ~ U T C U U T U C T T C C T ~ C M C T G C T ~ C T C ~ T ~ U f f i ~ ~
Y V C F F G Y P S F K Y S H P H H F L K T T A A L R G E V V 210 240 270 UG~CMGCTCTUWU~TWGUffiWTTAW~GTMCTGTT~6UIT~TATGT~6UI~WTAUGT~CTU~T E
F
K
L
S
O
I
G
E
G
I
R
E
V
T
V
K
E
Y
Y
MO
V
K
E
G
D
T
V
S
E
F
360
330
WTAWTCTGT6UIG~TUUGTWTlMGCTTCTGTTlCUTUCTAGTCGTTATWTGWGTUTT~CTCTATTATMTCTA O
S
I
C
E
V
E
S
O
K
A
S
V
T
I
T
S
R
Y
O G 420
390
V
I
K
K
L
Y
Y
Y
L 450
~ C W T A T T ~ T A T G T V ~ U T T A G T A W U T A ~ C G ~ G C T T T ~ W T T U 6 U I 6 U I W T G ~ G ~ ~ C T C C T ~ D
D
I
A
Y
V
G
K
P
L
V
D
I
E
T
E
A
L
K
D
S
E
E
D
V
V
E
T
P
A
4KB
“2
~ K B
540
510
480
-3
GTGTCTUTWTGMUTAUUCCMWWT~GffiCC~UCTGGCM~CCT~GTTCGCCGTN~T~CMT
V
S
H
D
E
H
T
H
E
E
I
K
G
R
K
T
L
A
T
P
570
A
V
R
R
L
A
M
E
Y
Y
600
6M
ATTMGCTWGTWIGTTGTTffiCTUG~WTGGU6UITACTT~6UIWTATCCT~CTATTT~GUWUGWGCT
I
K
L
S
E
V
V
G
S
G
K
D
G
R
I
L
K
E
D
I
660
L
Y
Y
L
E
K
E
T
G
A
no
690
r\TATTCCCTCtTTU~UUGTT~TTATGCUCCTCUCTUUGCCUMWUTWCTGTTCCTATACTA~ATUMIICCTCCG
I L P P S P K V E I M P P P P K P K D M T V P I L V S K P P 750
G
T
A
~
U
U
V
V
F
T
G
K
D
K
.
U
T
U
E
P
W I
C K
U G
M F
U E
~ K
C
A
C
~
840
780 U T
V
K
T
810
~
G
G
C
T
l
T
M
S
A
A
L
K
I
G P
G
U
T
H
900
870
TTTffiTTATTGTWTWWTTWCCTlACT~CTGG~MGCTCCW6UI~TT~CCU~GUTTTGCTffiT~TT~CTC
F G Y C O E I D L T E L V K L R E E L K P I A F A R G I K L 930
ppo
960
T~ATGCCmCn~T~GGCT~TTCC~TTACTAUG~TCCTATC~MCGCTTCTGTGWT~CT~U6UIT
S F M P F F L K A A S L G L L E F P I L Y A S V D E M C E Y
FIG.5. Autoradiograph of mRNA transcripts in wild type human tissue. Radiolabeled antisense 5’-E2b RNA was hybridized to mRNA bound on nylon filters. Lane 1 contains 20 pg of poly(A+) I T Y K A S H Y I G 1 A M D T E E G L I V P ~ V K Y V O l C 1110 RNA from human lymphoblasts, and lane 2 contains 20 pgof poly(A+) T C T A T A T T T W U T C G C U C T ~ C T G U C C G C C T C U G U i U RNA from human liver. Size of hybridizing species was determined S I F D I A T E L Y R L E K L G S V C E L S T T O L T G G T by comparison to RNA laddermarkers from Bethesda Research 1200 1260 Laboratories. ~TACTCTTTCCUUTTTWTCMTTffiTGGTAC~TT~TUUCUGTWTMT~UCCT6UIGTAGCU~~CCCCTT~TU
0
1080
1020
ATMUTATMV.CTTCTUTMUTTffiWTA~TGWTA~WGU~~WTTGTCCCTMTGT~T~TUWTCT~
40
30
-
477AA
PE2b
.
423 AA
WTffiTGCTACMTGTUCCCTT~~T~GTG~TCCTATTTA~CCUGCT~AT~TA~AWTCT~T6UIW ECET2G0
3-
405AA
S
613 AA
F T L S Y I G S I G G T F A K P V I M P P E V A I G A L G S 125-3
1350
1320
HEPb
A~MGtCU~CCC~T~MCU~GW6UIGTATATMGGUUWTMT6UITGTWGCTGGTUGCTWTUUWGTTATT
I K A I P R F Y E K G E V Y K A P I M Y V S U S A D H R V I l y u l
1440 D
wI
G
A
T
M
S
P
F
S
Y
L
1410 Y
K
S
Y
L
E
Y
P
A
F
M
L
L
D
L
K
O
HE2p
1470 ATMWUTTCTTUlCTTTTTUGCTTC~WGTATGT~CCCTAGCTGT~UGUUTGTTUTCTTTACM~ATATTGTM
154.0
15W
1620
ACWTTTGTATCGTAT~~MGWTCTMG~CMTATTTGTUCTGTTCTATTA~CTTTTTACT~T6UITMTffiTGTMTG
0
1650 CTTCTCCTCU%CICTUUTTTTATAffiTUWGTGT~CTTCTTMTATGGTGCTWTGTTTTTGTGT~T~T~CTffiCM 1740
1Tm
1800
WTTMUMIITTAU;CCGTffiT~CTUC~CTGTMTCUGUCTTTffiWffiCCUGGTGffiGCWTU~TWffiTTA~G
1860
1830 TTTMWCUGCCTGGCCMUTGGT~CCTGGCCTCTACCT~TACUMTTWCCGGGTGTGGTffiTffiGTACC
FIG.4. Nucleotide sequence and deduced amino acid sequence for the open reading frame of the cDNA insert in pSL5. Numbers above the lines indicate base pairs beginning with the ATG start codon. The nucleotide sequence was determined in both directions using plasmid constructionsas described under “Materials and Methods.”
EE2p
P
629AA
N C FIG.6. Comparison of human pre-E2b protein structure to the protein structure of other acyltransferases. HE26, branched chain acyltransferase from human (thisstudy); PE2b, branched chain acyltransferase from P.putida (13); EE20, succinyltransferase from E. coli (26); HE2p, acetyltransferase (17); EE2p, acetyltransferase from E. coli (19). Open box, leader sequence; K, lipoatebinding domain; hatched box, E3 binding domain; open circle, histidine in the concerned seven-amino acid motif implicated in CoASH binding. N AA, total number of amino acids in sequence.
kb cDNA make an antigenic fusion protein in their original Xgtll phage, suggesting that bothcontaininformation to 219-239 of P. putida (lined unmarkedregion in Fig. 7). Despite encode the E2b protein. Previously we showed the 1.6-kb the similarity in amino acid content and protein structure, cDNA to encode the leader sequence the lipoate binding antisera to mammalian E2b did not recognize the E2b protein domain, and the E3binding domain of pre-E2b (1).Comparof P. putida. cDNA probes for each E2b did not cross-hybrid- ison of the restriction enzyme sites within the two cDNAs ize (data not shown). showed that the 3.0-kb cDNA was longer than the 1.6-kb cDNA in the 3‘ end of the molecules (Fig. 1).Therefore, we DISCUSSION ligated the first 608 bp from the 1.6-kb cDNA (EcoRITwo issues will be addressed 1)Does the constructed cDNA EcoRV(598)) to 1250 bp from the 3.0-kb cDNA (EcoRV(150)(pSL5) encode the full preE2b protein? and 2) How does the 2nd KpnI(1400)) to make a construct with enough coding original 1.6-kb cDNA differ from the constructed 1.9-kb sequence to direct the synthesis of a full-length pre-E2b. cDNA and why? Although in uitro synthesized pre-E2b migrated with an M , Evidence that the 1.9-kb cDNA construct (pSL5) encoded value of 57,000, the summed amino acid weights gave a value a full-length pre-E2b protein comes from antibody recogni- of52,958. The mitochondrially processed immunoreactive tion, co-migration of i n uitro made and processed protein with protein had an M , value of 52,000. A 56-amino acid aminotissue-prepared E2b in the branched chain a-ketoacid dehy- terminal leader sequence (2) should be removed by mitochondrogenase complex, and analysis of the deduced amino acid drial proteaseactivity leaving 421 amino acids with a summed sequence and implied protein structure. Boththe 1.6- and 3.0- weight 46,369 in the mature E2b. Acyltransferase proteins are
cDNA for Entire Human
Branched Chain Acyltransferase
7745
genase. Experiments designed to test binding and function are now possible using the cloned cDNAs and in vitro made HE2b LRGQWQFXLSDIGEGIREVTVKEWYVKEGDTVSQFDSICEVQSDKASVTITSR ...... . . ....... . . . ..... . ........... . . . . . . . 108 proteins. . . . PEZb MGTHVI--KXPDIGEGIAQVELVEWFVKVGDIIAEDQWADVMTDKATVEI?S? 52 A difference in E3 binding tothe complexes has been 150 HE2b YDG7iXKLYYNLDCIAYVGXPLVEIETEALKDSEEDVVETPA-----------observed. The branched chain a-ketoacid dehydrogenase com. . ... . . . .. .. .. .. .. .. . .. plex in mammals showslower affinity for E3 than either PE25 VSGKVLALGGQPGEVMA7GSELIRIEVZGSGNBPDVPQAXPAEVPAA?VAAXPE 106 E3 pyruvate dehydrogenase complex or a-ketoglutarate dehydroVSHDE-HTHQEIKGRKTLATPAVRRLAWENNIKLSEWGSG 190 HE2b .. . . . . . . . . . . ....... . . ... . . . . . . genase complex. When the branched chain a-ketoacid dehy160 PE2b PQKDVKPAAYQASASHEAAPIVPRQPGDK?LASPAVRKRALDAGIELRYVHGSG drogenase complex is purified from tissue, the E3 component HE2b KDGRILKEDILNYLEKQTGAILPPSPKVEIMPPPPKPKDMTVPILVSKPPVFTG 244 is usually lost (10, 21-24). There is no obvious difference in ....... ........ . . 192 PEZb PAGRILHED-LD---------------------AFMSKPQSMGQTPNGYARRT the general structure of E2b when compared with the other acyltransferases. A more extensive analysis and additional HE2b XDKTEPIKGFQKAWVKTMSAALX-IPHFGYCDEIDLTELVXLREELKPIAFARG 297 . ... . . . . . . . .. .. ......... .. . . . experimental designs are needed to determine the properties PE2b DSEQVPVIGLRRKIAQRMQDAKRRVAHFSYVEEIDVTALEALRQQLNSKHGDSR 246 of E2b which produce the altered affinity for E3 in the HE2b IKLSFMPFFLKAASLGLLQFPILNASVDENCQNITYKASHNIGIAMDTEQGLIV 351 . ... . . . ... . ... ... . ... . ... ... . . . . . ..... . . . . . . branched chain a-ketoacid dehydrogenase complex. Subunit PEZb GKLTLLPFLVRALWALRDFPQINATYDDEAQIITRHGAVlWGIATQGDNGLMV 300 interactions canbe analyzed after engineering an altered E2b HE2b PNVKNVQICSIFDIATELNRLQKLGSVGQLSTTDLTGGTFTLSNIGSIGGTFAK 405 protein structure by changing the base sequence in thecDNA . . . . . . . . . . . . . . ... .... .. .. .. .. .. PE2b PVLRHAEAGSLWANAGEISRLANAARNNKASREELSGSTITLTSLGALGGIVST 354 clone for pre-E2b. CoASH The catalytic domain in these proteins is similar in size and RE2b PVIMPPEVAIGALGSIKAIPR-FNQXGEWXAQIMSWSADHRVIDGATMSRF 458 . . . . . . . . . . . . . ............ .......... . amino acid sequence. Yet, within this domain must reside 406 PE2b PVVNTPEVAI--VGVNRRPWIDGQIV?RK"NLSSSFDHRWDGMDAALF specificity for binding of the complex-specific subunits. This HE2b SNLWKSYLENPAFMLLDLK* 477 specificity is known from mixing experiments with subunits . ..... . . . . . 423 PE2b IQAVRGLLEQPACLFVE* from mammalian pyruvate dehydrogenase complex and FIG. 7. Comparison of amino acid composition for human branched chain a-ketoacid dehydrogenase (27). HeterogeE2b (HE2b) and P.putida E2b (PE2b).Alignment begins with neous complexes would not decarboxylate either pyruvate or the glycine amino-terminal of mature HE2b. The E3 binding domain branched chain a-ketoacids efficiently. We suggest that difand CoASH binding domains are indicated. The other lined region ferences in amino acid sequence in the catalytic domain denotes similarity between the E2b proteins which does not exist impart the specificity of complex formation. with other E2 proteins. Human andP. putida E2b amino acid sequence and general protein structure were more comparable to each other than known to show retarded migration in SDS-PAGE; conse- to other E2 proteins. One region (lined in Fig.7) in the quently the size differences were anticipated (11-15). A simcatalytic domain of the two E2b proteins appeared unique. ilar comparison can be made with pre-E2p and E2p which Amino acid sequence similarity in this region of the otherE2 migrates in SDS-PAGE with M , values of 78,000 and 70,000proteins is low relative to E2b. Thus, this region becomes a 74,000, respectively (16). Amino acid composition deduced candidate site for E l b and E2b specific binding. Additional from the nucleotide sequence gave values of 65,715 for prestudies are necessary to test thishypothesis. E2p and 59,613 for mature E2p (16,17).P. putida E2b consists Since 1431 bp define the coding region of full-length preof 423 amino acids with a molecular weight of 45,128 while E2b, it was important to examine the 1.6-kb cDNA we origithe protein migrates with an M , value of 47,000 in SDSnally isolated and consider why this cDNA did not contain PAGE (13). Acyltransferase proteins from different sources showed a the entire coding sequence. As reported, the nucleotide seremarkable similarity in structure. Of the five proteins com- quence for the 1.6-kb cDNA has 844 bp in the open reading pared, all those with a single lipoate binding domain contain frame. The 3"untranslated sequence contains two polyadebetween 405 and 423 amino acids as mature E2. The two E2p nylation signals and a poly(A) tail (1).What possible explaproteins are longer due to additional lipoate binding domains nations can be given for a seemingly full-length cDNA which (Fig. 6). All the proteins compared contain a histidine residue fails to encode critical carboxyl-terminal regions of the E2b in ahighly conserved 7-residue motif 28-31 amino acids prior protein? The nucleotide sequence of the construct throughbp to the carboxyl-terminal end of the proteins. It is postulated 924 remains the same as reported for the original 1.6-kb that histidine in this position is involved in CoASH binding cDNA. Bases 925-1013 of the construct are identical to bases for catalyzing the covalent linkage of the thiol ester to the 1291-1369 of the 1.6-kb cDNA (1);thus the 1.6-kb cDNA essentially contains 353 nucleotides inserted between bases acyl group (4, 5 ) . The E3 binding domain has a similar amino acid sequence 924 and 925 of the construct. It is tempting to speculate that for all acyltransferases (1)(Figs. 6 and 7). It is suggested that this 353 bp in the 1.6-kb cDNA results from an unspliced a single gene encodes the E3 protein in humans (28) and in intron. Bases 922-924, AAG, fit the consensus sequence upEscherichia coli (18, 19), while two genes are known in P. stream of 5' splice sites. Likewise, bases 1272-1291 of the 1.6putida, one used with the pyruvate dehydrogenase complex kb cDNA match the consensus (a pyrimidine-rich region followed by C or T, AG/G) for a 3' splice junction (25). To and a-ketoglutarate dehydrogenase complex and a separate gene for the induced branched chain a-ketoacid dehydrogen- investigate this possibility we are isolating genomic clones for ase complex (20). Thissuggests an evolutionary conservation the E2b gene and will compare the nucleotide sequence with of the E3binding site in these E2 proteins. However, binding that of the cDNAs we have already sequenced. Further comof E3 toE2 may be independent of catalytic function. The E3 parison of the 3"untranslated regions of both cDNAs usedin protein from bacteria used with pyruvate dehydrogenase com- the construct (pSL5) was uninformative. plex anda-ketoglutarate dehydrogenase complexwill not Since two mRNAs hybridize to probes for pre-E2b it refunction catalytically with the mammalian complexes. It is mains a possibility that the 1.6-kb cDNA represents a natunot known whether the branched chain a-ketoacid dehydro- rally occurring species in the cDNA libraries we screened. genase-specific E3 from induced P. putida will work catalyti- Another possible explanation would be that the1.6-kb cDNA cally with mammalian branched chain a-ketoacid dehydro- resulted from a rearrangement during amplification of the HE25
MLRTWSRNAGXLiCVRYFQTCGNVEVLKPNYVCFFGYPSFXYSHPHHFLXTTAA
_"""
54
7746
cDNA for Entire Human
Branched Chain Acyltransferase
possibilities with RNA protection experiments designed to study the naturally occurring mRNA species directly. REFERENCES 1. Hummel, K. B., Litwer, S., Bradford, A. P., Aitken, A., Danner, D. J., and Yeaman, S. J. (1988) J. Biol. Chem. 263,6165-6168 2. Lau, K. S., Griffin, T. A., Hu, C-W. C., and Chuang, D. T.(1988) Biochemistry 27,1972-1981 3. Litwer, S., and Danner, D. J. (1988) Am. J. Hum.Genet. 4 2 , 764-769 4. Fussey, S. P. M., Guest, J. R., James, 0. F. W., Bassendine, M. F., and Yeaman, S. 3. (1988) Proc. Natl. Acad. Sci.U. S. A . 8 5 , 8654-8658 5. Leslie, A. G. W., Moody, P. C. E., and Shaw, W. V. (1988) Proc. Natl. Acad. Sci. U. S. A . 85,4133-4137 6. Litwer, S., and Danner, D. J. (1985) Bwchem.Biophys. Res. Commun. 131,961-967 7. Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Smith, J. A., Seidman, J. G., and Struhl, K.(1987) Current Protocols in Molecular Biology, pp. 3.5.8-3.5.9, Wiley-Interscience, New York 8. Mueckler, M., Caruso, C., Baldwin, S. A., Panico, M., Blench, I., Morris, H. R., Allard, W. J., Leinhard, G. E., and Lodish, H. F. (1985) Science 229,941-945 9. Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J., and Rutter, W. J. (1979) Biochemistry 1 8 , 5294-5299 10. Heffelfinger, S. C., Sewell, E. T., and Danner, D. J. (1983) Biochemistry 22,5519-5522 11. Guest, J. R., Lewis, H. M., Graham, L. D., Packman, L. C., and Perham, R. N. (1985) J. Mol. Bwl. 185,743-754
198 13. Burns, G., Brown, T., Hatter, K., and Sokatch, J. R. (1988) Eur. J. Biochem. 1 7 6 , 165-169 14. Perham, R. N., and Thomas, J. 0.(1971) FEBS Lett. 15,8-12 15. Kreszw, G.-B., Dietl, B., and Ronft, H. (1980) FEBS Lett. 1 1 2 , 48-50 16. De Marcucci, 0.G. L., Gibb, G. M., Dick, J., and Lindsay, J. G. (1988) Biochem. J. 251,817-823 17. Coppel, R. L., McNeilage, L. J., Surh, C. D., Van de Water, J., Spithill, T. W., Whittingham, S., and Gershwin, M. E. (1988) Proc. Natl. Acad. Sci. U. S. A . 8 5 , 7317-7321 18. Yeaman, S. J. (1986) Trends Bwchem. Sci. 11, 293-296 19. Stephens, P. E., Lewis, H. M., Darlison, M. G., and Guest, J. R. (1983) Eur. J. Biochem. 136,519-527 20. Sokatch, J. R., McCully, V., Gerbrosky, J., and Sokatch, D. J. (1981) J. Bacteriol. 148,639-646 21. Lawson, R., Cook, K. G., and Yeaman, S. J. (1983) FEBS Lett. 157,54-58 22. Pettit, F. H., Yeaman, S. J., and Reed, L. J. (1978) Proc. Natl. Acad. Sci. U. S. A . 75,4881-4886 23. Paxton, R., and Harris, R.A. (1982) J . Biol. Chem. 2 5 7 , 1443314439 24. Ono, K., Hakozaki, M., Nishimaki, H., and Kochi, H. (1987) Bwchem. Med. Metab. Biol.37,133-141 25. Watson, J. D., Hopkins, N. H., Roberts, J. W., Steitz, J. A., and Weiner, A. M. (1987) Molecular Biology of the Gene, pp. 639640, The Benjamin/Cummings Publishing Co. Inc., Menlo
Park, CA 26. Spencer, M. E., Darlison, M. G., Stephens, P. E., Duckenfield, I. K.. and Guest, J. R. (1984) Eur. J. Biochem. 141.361-374 27. Cook, K.G., Bradford,'A. P.', and Yeaman, S. J. (1985) Biochem. J. 225,731-735 28. Otulakowski. G.. Robinson. B.H.. and Willard. H. F. (1988) Somatic~CellMol. Genet. 1 4 , 411-414 ~
~~~
