- Email: [email protected]
Cloning of the dihydroxyacid dehydratase-encoding gene (ILV3) from Saccharomyces cerevisiae
Gene, 137(1993)179-185 0 1993 Elsevier Science Publishers
GENE
B.V. All rights
reserved.
179
0378-l 119/93/$06.00
07536
Cloning of the dihydroxyacid
dehydratase-encoding
gene (ILV3) from
Saccharomyces cerevisiae (Yeasts; branched-chain evolutionary
amino
conservation;
acids; mitochondrial
enzymes;
nucleotide
sequence;
gene organization;
in vitro transcription/translation)
Juan A. Velasco, Jo& Cansado, and Vicente Notario
M. Carmen
Pefia, Toshiaki
Kawakami*,
Jorge Laborda”
Division oiExperimenta1 Carcinogenesis, Department ofRadiation Medicine, Georgetown University Medical Center, Washington, DC 20007, USA Received by J.A. Gorman:
18 May 1993; Revised/Accepted:
30 July 1993; Received at publishers:
23 August
1993
SUMMARY
The biosynthesis
of branched-chain
amino
acids (aa) involves
three
shared
pathways
through
which pyruvate
or
cr-ketobutyrate are converted into cl-keto acids, precursors of valine, leucine or isoleucine. In eukaryotes, few of these common enzymes have been purified to homogeneity, and the whole complement of biosynthetic genes has not been cloned from a single species. In yeasts, most of these genes (ZLV genes) have been cloned and sequenced, with the exception of that coding for dihydroxyacid dehydratase (DAD, EC 4.2.1.9), the third enzyme in the common pathways. We have isolated Saccharomyces cerevisiae genomic sequences by hybridization to an oligodeoxyribonucleotide (oligo) probe designed from a highly conserved domain among bacterial DAD-encoding genes. The cloned sequences have been located to S. cerevisiae chromosome X, mapped within 0.4 centiMorgans (CM) of the ilv3 locus, and found to complement the ilv3 mutations of various yeast strains. Nucleotide (nt) and aa sequence analyses of the longest open reading frame (ORF) located within the cloned sequences identified them as the ZLV3 gene, which codes for the yeast DAD. With our cloning of ILV3, yeast becomes the only eukaryotic system from which all ILL’ genes have been cloned, thus allowing direct molecular analyses of their regulation.
INTRODUCTION
Branched-chain
amino
acids
(aa)
are
synthesized
through metabolic pathways well conserved in bacteria, yeast, filamentous fungi and plants (Kanamori and Wixom, 1963; Umbarger, 1978). Pyruvate and c+ketobutyrate are the initial substrates for the synthesis Correspondence to: Dr. V. Notario, Department of Radiation Medicine, Georgetown University Medical Center, P-427 Lombardi, 3800 Reservoir Rd., N.W., Washington, DC 20007, USA. Tel. (l-202) 687-2102; Fax (l-202) 687-2221. *Present addresses: (J.L.) Center for Biologics Evaluation and Research, F.D.A, N.I.H., Bldg. 29, 8800 Rockville Pike, Bethesda, MD 20892, USA. Tel. (l-301) 496-4038; (T.K.) Division of Immunology, La Jolla Institute for Allergy and Immunology, 11149 North Torrey Pines Road, La Jolla, CA 92037, USA. Tel. (1-619) 558-3500.
of valine and isoleucine through parallel steps which share three bifunctional mitochondrial (mt) enzymes: (a) acetohydroxy acid (AH) synthetase (AHAS, EC 4.1.3.18) which catalyzes substrate conversion to AH, (b) AH isomeroreductase (AAI, EC 5.4.99.3), which carries out AH isomerization and reduction to yield a,gdihydroxy acids, and (c) DAD, which catalyzes their
Abbreviations: aa, amino acid(s): AAI, AH isomeroreductase; AH, acetohydroxy acid(s); AHAS, AH synthetase; bp, base pair(s); CM, centiMorgans; DAD, dihydroxyacid dehydratase; ds, double strand(ed); ILV, genes encoding isoleucine/valine biosynthetic enzymes; ILV3, gene encoding DAD; kb, kilobase or 1000 bp; L., Lactococcus; mt, mitochondrial; nt, nucleotide(s); oligo, oligodeoxyribonucleotide; ORF, open reading frame; PAGE, polyacrylamide-gel electrophoresis; S., Saccharomyces; TD, threonine
deaminase.
180 dehydration producing a-ketoisovalerate and a-ketoB-methylvalerate. Transamination of these a-keto acids yields valine or isoleucine, respectively. Alternatively, a-ketoisovalerate, the a-keto acid precursor of valine, undergoes a condensation reaction with acetyl-CoA, as the first step towards the synthesis of leucine. Carbon flow into the isoleucine biosynthetic pathway is regulated by threonine deaminase (TD) activity, the first enzyme in which deaminates threonine into the pathway a-ketobutyrate (Umbarger, 1978). The regulation of these pathways has been studied primarily in prokaryotes, with most genes cloned and sequenced (Lawther et al., 1987; Godon et al., 1992; Grandoni et al., 1992). Direct regulation studies of branched-chain aa synthesis in eukaryotes have been hampered by the lack of molecular clones and/or sequence information for all the genes involved from a given species. Genetic mapping of yeast mutants defective in the biosynthesis of isoleucine and valine (i/u mutants) allowed the assignment of the three common activities in branched-chain aa synthesis to three unlinked loci ilv2 (AHAS), ilv4/ilv5 (AAI) and ilv3 (DAD); another locus (ilvf ) has been assigned to TD (Kadar and Wagner, 1964; Jones and Fink, 1982). Genes encoding AHAS have been cloned from the filamentous fungus Neurospora crassa (Jarai et al., 1990) and the yeast S. cerevisiae (Falco and Dumas, 1985); while the yeast ILV2 has been sequenced (Falco et al., 198.5) no sequence information has been reported on the fungal gene. Also, AAI-encoding genes have been cloned and sequenced from N. crassa (Sista and Bowman, 1992), S. cerevisiae (Petersen and 1986) and from spinach chloroplasts Holmberg, (Spinacea oleracea, Dumas et al., 1991). The TD-encoding gene (ZLVI) has been cloned and sequenced from S. cerevisiae (Kielland-Brandt et al., 1984). However, the molecular cloning of an eukaryotic DAD has not been reported to date. In the present study we report for the first time the isolation, identification, nt sequence and deduced aa sequence of an eukaryotic DAD-encoding gene, the ILV3 gene from S. cerevisiae.
RESULTS
AND DISCUSSION
(a) Detection and cloning of yeast sequences homologous to bacterial DAD-encoding genes On the basis of comparisons of available DAD nt sequences from several bacterial strains (Lawther et al., 1987; Godon et al., 1992) we designed oligo probes that might recognize the yeast homolog sequences when hybridized to S. cerevisiae genomic DNA. The optimum probe (designated bDAD) corresponded to a 20 nt sequence showing the maximum homology (85%) between Luctococcus lactis sbsp. lactis and Escherichiu coli
(Fig. 3) starting at aa 269 and 264, respectively, of the DAD sequences. Hybridization of EcoRI cleaved yeast genomic DNA to radiolabeled bDAD probe under high stringency conditions allowed the detection of a single, well defined, band of about 9 kb (Fig. 1A). Only after extended autoradiographic exposures could other not so well defined hybridizing bands be visualized. This result strongly suggested that the yeast DAD-encoding gene was encompassed within that unique EcoRI fragment. Screening of a yeast genomic phage h library with the bDAD probe resulted in the isolation of 24 clones, all of which appeared to contain a single EcoRI insert of about 9 kb which hybridized very strongly to the bDAD probe under high stringency conditions. Further restriction endonuclease analyses demonstrated that all 24 clones contained identical inserts, and localized the bDAD hybridizing sequences to a BamHI internal fragment of about 3.6 kb (Fig. 1B). This fragment was cleaved from clone h-ILV-19, isolated through preparative agarose electrophoresis, purified by electro-elution and ligated into pBR322 and the yeast integration vector YIPS for further structural and functional studies. Fig. IB shows the restriction map of the BumHI insert into YIp5 (designated YIP-ZLV) and the localization of the bDAD hybridizing sequences within the XbaIIBamHI subfragment. (b) Identification of the cloned sequences as the ZLV3 gene In order to demonstrate that the cloned bDAD homologous sequences were indeed the yeast ILV3 gene, three experimental protocols were developed in parallel: (a) chromosomal assignment, (b) gene mapping, and (c) complementation studies of ilv3 mutants. Fig. 2 shows the pattern of hybridization obtained when the purified 3.6-kb insert of Yip-ILV was used as the probe on a blot containing electrophoretically resolved S. cerevisiue chromosomes. These data unequivocally demonstrated that the cloned sequences were located in chromosome X of S. cerevisiae, which correlated with results from our gene mapping experiments. For locus assignment, YIP-ZLV was linearized at its unique BgrII site located within the insert (Fig. lB), to direct integration into the homologous chromosomal region. Crosses between URA3 transformants of strain 1992 containing properly integrated YIP-ZLV sequences and strains carrying multiple centromere markers (Table I) located the bDAD homologous sequences very close to ilv3, previously mapped to the right arm of chromosome X (Kadar and Wagner, 1964). This tight linkage (< 0.4 CM) suggested a physical distance to ilv3 of less than 1 kb. To confirm that ilv3 was present in the 3.6-kb BumHI insert, it was ligated into the unique BamHI site of the centromere vector YCp50 (Kuo and Campbell, 1983).
181
1 kb
B
A kb
II II
23.5,
B
E
B
+
6.6, I
XII-
\
/
9.6,
E \
1’ /I 1’
BP
\\ \\ \
VII.XV-
. YIP-ILV
R-Y
2.2, 2.1,
I 0
.
I
I
I
1
2
3
RY I kb
4
IXlllVII-
Fig. 1
Fig. 2
Fig. 1. Detection from protoplast EcoRI
and
and cloning of S. cereuisiae sequences homologous to bacterial DAD-encoding genes. (A) Total genomic DNA was prepared lysates of exponentially growing X2180 diploid yeast cells as described by Winston et al. (1983). Aliquots were cleaved with bacterial DAD-encoding genes with an oligo probe highly conserved among analyzed by Southern hybridization
(5’-GCXTTTGAAAAXGCCATXAC-3’). This probe, designated bDAD, was synthesized including all possible degeneracies at the three (underlined) non-conserved nt positions, and was end-labeled with [Y-~‘P]ATP (222 TBq/mmol) using T4 polynucleotide kinase. Hybridization was carried out under high stringency conditions as described by Wahl et al. (1979). The mobilities of the major hybridizing fragment (arrow) and the size markers are indicated. (B) Location of the bDAD-hybridizing sequences within the insert of YIpplLI/, a BamHI subfragment of the genomic 9-kb EcoRI product cloned into A-ILF-19. A h-L47.1 genomic library was prepared with X2180 genomic DNA digested to completion with EcoRI. A-ILV-19 was isolated and purified from this library by hybridization to the bDAD probe, and found to contain a 9-kb insert; restriction analysis showed the presence within that insert of a 3.6-kb BamHI fragment containig the bDAD-hybridizing sequences. This BamHI fragment was subcloned into the unique BarnHI site of the yeast integrative plasmid YIp5 (New England Biolabs) generating YIP-ILI/. E = EcoRI; B = BarnHI; Bc = BclI; X = Xbul; Bg = Bg/II; Sp = SphI; K = KpnI; RV = EcoRV. The heavy arrow the bDAD homologous sequences. Fig. 2. Chromosomal location of the cloned characteristic of chromosome X was detected
indicates
the position
and translational
orientation
of the unique
ORF located
within
bDAD-related yeast sequences. A single hybridizing band migrating to the electrophoretic position when the 3583-bp Yip-ILV BamHI insert was used to probe a commercial S. cerevisiae ‘Chromoblot’
(Clontech Laboratories) containing 5 ug/lane of electrophoretically resolved chromosomes. The mobility for hybridization and washing of the nylon membrane were as recommended by the manufacturers.
of all chromosomes
of YIP-ZLV
DNA
BumHI
insert
ilv3, ura3 haploid
Analysis
of nt sequence
from this construct was transformed into several strains. Strains GU86-1 lb (MATa ade2 his1 ilv3 ura3), GU91-22c (MATa adel his3 ilv3 lysl ura3), GU150-12b (MATa arg4 his4 ilv3 ura3), 3914-9A (MATa ura3 ilv3 from Reed B. Wickner), and IVPXS-2B (ATCC 64448, MATa his- leu2 ilv3-12 ura3) were used for complementation assays (Sherman et al., 1984). ZLV3, URA3 transformants were isolated in all cases. These results demonstrated that the cloned bDAD-hybridizing fragment encompassed S. cerevisiae ILV3 sequences capable of functionally complementing the ilv3 mutations used. (c) Analysis of nt and aa sequences of the cloned IL V3 gene In order to precisely locate the ILV3 coding sequences within the cloned fragment, the nt sequence of the 3.6-kb
insert
as 3583 bp, and
ORF
(Fig. lB), starting
was
defined
is indicated.
Conditions
determined
the length
determined
(Fig. 3).
of the cloned
the existence
at nt position
of an
1601 and ending
at 3116, encoding a polypeptide of 504 aa with a predicted M, of 60480. The position of the bDAD probe was located
within
the coding
region
Fig. 3 shows the nt sequence
defined
of the ORF
by this ORF. and its align-
ment with the E. coli and L. lactis sbsp. luctis sequences used
to design
S. cerevisiae
the bDAD
sequence
probe.
The
corresponding
shows 90% homology
with each
of the bacterial bDAD oligos. An overall nt homology search of DNA databases showed that the sequence of the cloned yeast ORF was about 50% homologous to the E. coli DAD-encoding gene and nearly 60% homologous to that of L. lactis sbsp. lactis.
182 TABLE
I
Genetic
linkage
of integrated
Yip-ILV
Cross
Parental
3892 3914
19921LV x 1972 3905-1lD x 3892-28D
“Only
tetrads
strainsb
showing
2+:2-
ILC’: : URA.3 were used for compiling
(ILV:
: URA3) to ilr;3”
PD’
NPD’
TC
53
0 0
0
86 segregation
for
0 both
iit13 and
these results.
bStrains used were: 1992 (MA?& spol I uru3 cunl cgh2 ude2 his7 hom3; Klapholz and Easton-Esposito, 1982); 19921LV, 1992(ILV: : URA3); 1792 (MATa gall his4 trpl horn3 ura3 CUPR ilo ude3 rad52 mul; Mortimer et al. 1981); 3905-11D (MA’& adr- urrr3 met3 ILV: : URA3), and 3892-28D (MA% ~73 i/r3 hom3). “PD, parental ditype; NPD, non-parental ditype: T, tetratype. IMethods: Yip-ILV was linearized by cleaving within the insert &III, and transformed into strain 1992 using method (Ito et al., 1983). Five URA transformants
with
the lithium acetate were crossed with
ura3 haploids and shown by their non-linkage to horn3 to not have integrated by recombination at URA3. One of these strains was then used to map the site of integration by standard mapping techniques (Mortimer
and Hawthorne,
1975).
Alignment of the deduced aa sequence of the yeast ORF with those of the DAD-encoding genes of E. coli and L. lnctis sbsp. tactis indicated (Fig. 4) that the translational products of the three genes were highly homologous. The yeast sequence showed about 39% identity with the E. coli enzyme, and 58% with that of Lactococcus over their entire sequences. Considering conservative aa substitutions, the yeast protein showed about 57% and 74% similarity with the enzymes of E. co/i and LuctococcUs, respectively. The extent of homology is close to 90% within several subdomains of the yeast protein (Fig. 4). In particular, the peptide defined by the nt sequence in the cloning bDAD probe (aa 202.. 208) shows 100% identity between Su~~~u~o~~~ce~and Lactococcus, and 100% similarity with E. cob. These data unequivocally identify the cloned sequences as the DAD-encoding ILV3 gene of S. cereuisiue. Because the three common enzymes involved in the synthesis of branched-chaiil aa are encoded by nuclear genes and subsequently translocated to mitochondria
GAGGCmri;AAAACGCCA~ACTTATGT~GTTGCAACCGGTGGGTCCA~T~TGCTGT~TTGCATTTG~TGGCTGTTG~TCACTCTGC~GGTGTC~G~TGTCACCAG~TGATTTCC~ ........I..7 *....A.. (L.lactis)
720
AGAATCAGTEATACTACACCftfiGATCGGTGACTTC~~CTTCTGGT~TACGTCAT~GCC~TTTG~TT~~GTTG~TGGTACCC~TCTGT~TT~GTATCTAT~TG~C~~
840
. . G. . . . . . . . . . . . . . G.. (&.coli)
TACACAAGAi;GTACTCTATCCAAGTATGCjAAGTTGGiCTTGA Fig. 3. The nt sequence
of the yeast ILV3 ORF.
The coding
sequence
of the longest
ORF
identified
within
the YIP-ILV
insert is shown
from the
putative starting ATG. Other ATG codons with sequence contexts favorable for translation initiation are also underlined. The yeast sequence hybridizing to the bDAD cloning probe is underlined in bold, and shown in comparison with the corresponding sequences of E. co/i and L. luctis. Identical nt are indicated by dots (.). Only the yeast nt positions are numbered. This sequence Methods: Subfragments of the 358%bp BornHI insert were subcloned into pBluescriptI1 SK. vector were used to determine the nt sequence of both strands of each DNA subfragment, Biochemical), as recommended by the manufacturers. Additional sequencing primers required were obtained
from Bio-Synthesis.
has been deposited in GenBank (Accession No. L[3975). Primers for the T7 and T3 sequences in the p3luescript using a Sequenase sequencing system, version 2.0 (US for sequencing internal portions of the longest fragments
183 SC ::
SC :i
. . . . . . . . . . . . . . . .. .. . . . . . . . . . . . . . . . . . .. . . . .. . . . . . . . . . . . . . . .. .
NEFKYNGKVE SVELNKYSKT LTPRSTQPAT QAHYYGIGFK DEDFKKAQVG IVSMDYDGNP . ........ . ..NPKYRSAT TTHGRNNAGA RALWRATGNT DADFGKPIIA VVNSFTQFVP
60 48
QFN-ll GVSDGISNGT KGNRVSLQSR EIIADSFETI ::I:: :::::::::: L::::::V:: : ::::l::N I:::: I:D:::A::H 6::L:::P:: :I::::V:VI
47
...HFSIIEK RLK....... . . CNNHLGTLGS KI:SSVNQTD GL GHVNLRDLG. LVAEQIEAAG 6.
107 174 162
SC :i
SC Ll EC
SC Ll EC
167 234 222
SAFQSYGEYI SKQFTEEERE DVVE
AVSKEK.... LAECDNIGEY IKKTHELG.. .ILP ILTK EAFENAITVV VATGGSTNAV :::Q::.... QE:::D::LA ::NLL:KD.. .:K:!fT :THADRKQLF :NAGKR:V:L T:RYY:QNDE SA:::l: I: A:::::I:LD :H:::::l:
220 287 282 279 346 342
SC Ll EC
329 397 402
K KAPSLP.EGQ EIIKPLS...
SC Ll EC
:; EC
:z
SLAPGGAV.. GKITGKEGTY .... .. ... . ... ....... ....FKGRAR VFEEEGAFIE N::Q::S:.. A::S::::EF .... .. ... . .......... .. ... ..T.. ::DG:QH::D DR:N:VSARW NTP:A:TAAU RCSTVILRKR LHRETAGVDD SILK:T:P:K :Y:SQDDAV:
363 431 462
:5
423 489 520
SC LJ EC
481 547 580
SC
SC Ll EC
.......TRG TLSKVAKLVS NASNGCVLDA l .... . .. ........ V:A:F:::TR P::E:::T:L l . ... . PKNRERQVSF A:RA::S:AT S:DK:A:RHK SKLGG'
504 570 615
Fig 4. Alignment of the DAD aa sequences from L. lactis (LI) and E. (Ec) with that deduced from the ILV3 ORF (SC). Domains with conservation greater than 85% are boxed. Identical aa are indicated by coli
(:), conservative
substitutions
by (I). sequence
gaps introduced
for opti-
mum alignment by (.), and translation termination codons by (*). The yeast sequence shows a 74% similarity with the 15. lactis enzyme over the entire protein sequence, and a 57% with that of E. coli. Most DNA and protein analyses were performed with the software package PC/Gene. ORFs were found using the methods of Fickett (1982) and Shepherd (1990). Homology searches in DNA and protein databases (GenBank release 69.0, EMBL release 26.0, NBRF-protein release 26.0, and Swissprot release 18.0) were performed with the program FASTA (Pearson and Lipman, 1988). The method of Needleman and Wunsch (1970) was used to study the statistical significance of the homologies found.
(Ryan and Kohlhaw, 1974), we analyzed the deduced ZLV3 aa sequence for the presence of putative mt-targeting sequences. In general, these sequences are rich in positively charged and hydroxylated aa residues, contain few or no negatively charged residues, and can form short amphiphilic cl-helical structures (Hart1 et al., 1989). Fourteen of the 37 N-terminal aa residues of the ILV3 protein are either positively charged or hydroxylated, and only two negatively charged residues are encompassed within the same domain. In addition, computer analysis (Chou and Fasman, 1978) for predictions on protein structure suggested that these amino terminal residues could assume an amphiphilic cl-helical confor-
mation.
Thus, it seems very likely that these first 37 aa
are part cannot
of the ILV3 mt-targeting be demonstrated
yeast DAD is purified
until
signal,
although
the mature
form
this of the
to homogeneity.
(d) In vivo and in vitro expression In order to determine the level and pattern
of expres-
sion of ZLV3, RNA from S. cereuisiae cells harvested various purified
phases of growth from asynchronous (Elder
hybridization XbaIIKpnI recognized appear
et al., 1983) and analyzed
was
by Northern
with ZLV3-specific probes (fragments and KpnI-BarnHI, Fig. 1B). Both probes a single
transcript
to be highly
expressed
throughout
cultures
at
the different
phases
of about
2.3 kb which
in a constitutive of growth
fashion
(Fig. 5A,B).
The same transcript was also recognized by a probe corresponding to the upstream BarnHI-XbaI fragment. This, when considered in addition to the position of the ILV3 ORF within the YIP-ZLV insert and the ability of YIP-ZLV to complement ilu3 mutants carrying presumably different mutations, strongly suggests that the entire ZLV3 transcriptional unit is located within the cloned 3.6-kb BamHI fragment. These data are consistent with the genetic definition of ilv3 as a single yeast locus (Kadar and Wagner, 1964), and agree with our finding of a single genomic fragment hybridizing to the bDAD probe. The capacity of the cloned ORF to synthesize a protein of the expected 60.48 kDa, was studied by in vitro coupled transcription and translation assays, using rabbit reticulocyte lysate and pGEM-3Z and pSP6bpoly(A) constructs containing the ILV3 insert. As shown in Fig. 5C, several polypeptides were translated from the ZLV3 insert when placed in sense orientation relative to the SP6 promoter, whereas only background signals could be detected when antisense constructs were used. The largest of the most abundant polypeptides was about 60 kDa, indicative of translation initiation at the first AUG in the ZLV3 ORF. The other major species may result from translation initiation at internal AUG codons, which is known to occur in rabbit reticulocyte lysates (Kozak, 1990; Jackson, 1991). Indeed, the AUG codons at aa 49, 71, 82 and 149 (Fig. 3) lay within sequence contexts defined as favorable for translation initiation (Kozak, 1981). The only eukaryotic DAD purified to homogeneity to date has been that from spinach leaves (Flint and Emptage, 1988). The native enzyme is a dimer of 110 kDa which consists of two monomers of 63 kDa and a [2Fe-2S] cluster. The yeast ILV3 contains two Cys residues (aa 62 and 140) themselves conserved in E. co/i and Lactococcus and located within aa domains highly conserved in bacteria; these two Cys are optimum candidate
184
A
123456
c
12 __ *
kDa
that the yeast ILV3 protein may adopt this type of metalbound dimeric conformation. (e) Conclusions (1) A genomic fragment encompassing which codes for the branched-chain
the ILV3 gene, aa biosynthetic
enzyme DAD, has been cloned from the yeast S. cerevisiae by hybridization to an oligo probe designed on the basis
B
of the high level of conservation
among
bacterial
DAD
sequences. (2) Identification
of the cloned
gene has been established
Fig. 5. In viva (A) Hybridization
and in vitro expression of the ILV.3 gene. of total RNA extracted from S. cereuisiue X2180 cells
collected after 3 h (lane 1 ), 6 h (lane 2) 9 h (lane 3), 12 h (lane 4) 18 h (lane 5) and 24 h (lane 6) with the XhaI-KpnI fragment of the Yip-1LV insert detected a single 2.3-kb transcript under high stringency conditions. The ds probe was radiolabeled by random [E--“2P]dCTP (111 TBq/mmol) using the Megaprime
priming with DNA labeling
system (Amersham) and purified through Sephadex G-50 spin-columns. Specific activity was 4 x 10’ cpm/pg. Blots were washed three times at room temperature in 2 x SSC/O. 1% SDS, for 5 min each time, and then twice for 30 min each, at the hybridization temperature (42’C) in 0.I x SSC/O.l% SDS. SSC is 0.15 M NaCl/O.OlS M Naacitrate, pH 7.0. Dried blots were exposed at -70°C to Kodak XAR-2 film, with intensifying screens. The migration of the ribosomal RNA markers is indicated. Equal sample loading was checked by staining the same gel with ethidium bromide(B) prior to blotting and hybridization. In in vitro coupled assays, SP6-derived ‘sense’ ILV3 transcripts were efficiently translated by rabbit reticulocyte lysates (C, lane l), whereas transcripts of ‘antisense’ constructs failed to direct efficient polypeptide translation (C, lane 2). The 3583”bp BaniHl fragment, encompassing the longest ORF classified as coding by the two methods mentioned in the Fig. 4 legend. was ligated into the unique BanzHI site in the plasmid vectors pGEM-3Z and pSP64-poly(A) (Promega, Madison, WI, USA). DNA prepared from constructs containing the insert in sense or antisense orientation relative transcription and
to the SP6 promoter translation assays
was used for in vitro coupled using the ‘TNT Coupled
Reticulocyte Lysate System’ from the same company. This system uses a rabbit reticulocyte lysate for the translation of transcripts synthesized with recombinant SP6 or T7 RNA polymerase from the respective promoter. Reactions were carried out as recommended by the manufacturers in the presence of L-[“5S]methionine (Amersham, in vivo cell labeling grade, 37 TBq/mmol). Translation products were resolved by SDS-PAGE as described (Notario et al., 1990) and visualized by autoradiography. The arrow indicates the migration of the largest fLV3 derived 60-kDa polypeptide.
aa for the formation of putative [2Fe-2S] clusters. Rabbit reticulocyte lysates provide an iron-rich nonoxidizing environment where Fe-S linkages can be formed and preserved. The polypeptides of about 1 lo-115 kDa detected by in vitro translation of ILV3 sequences might be homo- or heterodimeric forms of the major monomeric polypeptides in which the stability of the [Fe-S] bridges has been preserved. Thus, it is likely
sequences
as the ILV3
on the basis of(i) localization
in chromosome X (lOR), (ii) genetic mapping within 0.4 CM of the ZLV3 locus, and (iii) ability to complement the i/u3 mutations
in several haploid
strains.
(3) Analysis of nt sequence confirmed the identity of the cloned sequences as the DAD-encoding gene and showed that the yeast sequence is highly homologous (up to about 60%) to those of bacterial counterparts, indicating a high level of evolutionary conservation. The yeast sequences contain an ORF with coding capacity for a protein of 504 aa (60.48 kDa), which is highly conserved (up to 74% similarity) relative to the bacterial enzymes. The size of the predicted ILV3 protein product matched that of the largest polypeptide species synthesized in in vitro transcription/translation assays. (4) A single ILV3 transcript of about 2.3 kb has been detected in total yeast RNA. The mRNA is transcribed in a constitutive fashion, being expressed at high levels in RNA prepared from yeast cells harvested at different stages of asynchronous growth. (5) With our cloning of the ILV3 gene, yeast becomes the only eukaryotic system from which all ILV genes have been cloned from, thus allowing direct molecular analyses of their regulation,
ACKNOWLEDGEMENTS
The authors would like to thank Reed B. Wickner for his help in genetic mapping experiments, Keith C. Robbins and Stuart A. Aaronson in whose laboratory this project was started, and Anatoly Dritschilo for his support and discussions during its development.
REFERENCES Chou, P.Y. and Fasman,
G.D.: Empirical
predictions
of protein
confor-
mation Annu. Rev. Biochem. 47 (1978) 251-276. Dumas, R., Lebrun, M. and Deuce, R.: isolation, characterization and sequence analysis of a full-length cDNA clone encoding acetohydroxy acid reductoisomerase from spinach chloroplasts. Biochem. J. 277
( 1991) 4699475.
185 AUG initiator
Elder, R.T., Loh, E.Y. and Davies, R.W.: RNA from yeast transposable element Tyl has both ends in the direct repeats, to retrovirus
a structure
Kozak,
RNA. Proc. Natl. Acad. Sci. USA 80 (1983) 2432-2436.
Falco, SC. and Dumas, K.S.: Genetic analysis of mutants of Saccharomyces cereuisiae resistant to the herbicide sulfometuron methyl. Genetics 109 (1985) 21-35. Falco, SC., Dumas, K.S. and Livak, K.J.: Nucleotide sequence of the yeast ILV2 gene which encodes acetolactate synthase. Nucleic Acids Res. 13 (1985) 401 l-4027. Fickett, J.W.: Recognition of protein
contains
a [2Fe-2S]
3558-3564. Godon, J.J., Chopin, biosynthesis
cluster.
J. Biol.
Chem.
l-45. lto, H., Fukuda,
from spin263
(1988)
genes in Lacrococcus lads
subsp.
lactis. J. Bacterial.
protein
of Bacillus subtilis. N.,
Nicholson,
import.
Y., Murata,
D.W.
Biochim.
regulation
J. Bacterial. and
Biophys.
K. and Kimura,
174 (1992)
Neupert,
W.:
Acta 988 (1989)
A.: Transformation
of
tion initiation in rabbit reticulocyte lysates: unique features of encephalomyocarditis virus RNA translation. Biochim. Biophys. Acta 1088 (1991) 3455358. Jarai, G., Yagmai, B., Fu, Y.-H. amino
acid
and
Marzluf,
biosynthesis
G.A.:
Regulation
in Neurospora
of
crassa.
Cloning and characterization of the leu-f and i/u-3 genes. Mol. Gen. Genet. 224 (1990) 3833388. Jones, E.W. and Fink, G.R.: Regulation of amino acid and nucleotide biosynthesis in yeast. In: Strathern, J.N., Jones, E.W. and Broach, J.R. (Eds.), The Molecular Biology of the Yeast Saccharomyces. Metabolism and Gene Expression, Coid Spring Harbor Laboratory, Cold Spring Harbor, NY, 1982, pp. 181-299. Kadar, S.N. and Wagner, R.P.: Genetic and biochemical isoleucine-valine mutants Kanamori, M. and Wixom,
analysis
of
of yeast. Genetics 49 (1964) 2133222. R.L.: Studies in valine biosynthesis.
V.
Characteristics of the purified dihydroxyacid dehydratase from spinach leaves. J. Biol. Chem. 238 (1963) 998-1005. Kielland-Brandt, M.C., Holmberg, S., Petersen, J.G.L. and NilssonTillgren, T.: Nucleotide sequence of the gene for threonine deaminase ilu-l of Saccharomyces cereuisiae. Carlsberg Res. Commun. 49 (1984) 5677575. Klapholz, S. and Easton-Esposito,
Nucleic
of the fidelity
of initiation
Acids Res.
lysates from commercial
sources.
of translation
Nucleic
in
Acids Res. 18
(1990) 2828. Kuo, C.-L., and Campbell, J.L.: Cloning of Saccharom~ces cereuisiue DNA replication genes. Isolation of the CDC8 gene and two genes that compensate the cd&l mutation. Mol. Cell. Biol. 3 (1983)
operon of Escherichia 2137-2155.
R.: A new mapping
method
R.K.
and
coli K-12.
Hawthorne,
Nucleic
D.C.:
Acids
Genetic
Res.
15 (1987)
mapping
in yeast.
Methods Cell Biol. 11 (1975) 221-233. Mortimer, R.K., Contopoulou, R. and Schild, D.: Mitotic
chromosome
loss in a radiation-sensitive strain of the yeast Saccharomyces cere1C.v iae. Proc. Natl. Acad. Sci. USA 78 (1981) 5778-5782.
intact yeast cells treated with alkali cations. J. Bacterial. 153 (1983) 1633168. Jackson, R.J.: Potassium salts influence the fidelity of mRNA transla-
branched-chain
M.: Evaluation
reticulocyte
Mortimer, amino acid
174 (1992) 6580-6589. Grandoni. J.A., Zahler, S.A. and Calvo, J.M.: Transcriptional
Mitochondrial
ribosomes.
Lawther, R.P., Wek, R.C., Lopes, J.M., Pereira, R., Taillon, B.E. and Hatfield, G.W.: The complete nucleotide sequence of the iluGMEDA
regions in DNA sequences.
M.C. and Erlich, S.D.: Branched-chain
of the ilu-lea operon 3212-3219. Hartl, F.-U., Pfanner,
by eukaryotic
1730&1737. coding
Nucleic Acids Res. 10 (1982) 5303-5318. Flint, D.H. and Emptage, M.H.: Dihydroxy acid dehydratase ach
codon
9 (1981) 5233-5252.
similar
employ-
ing a meiotic ret-mutant of yeast. Genetics 100 (1982) 387-412. Kozak, M.: Possible role of flanking nucleotides in recognition of the
Needleman, S.B. and Wunsch, CD.: A general method applicable to the search for similarities in the amino acid sequence of two proteins. J. Mol. Biol. 48 (1970) 443-453. Notario, V., Castro, R., Flessate, D.M., Doniger. Frequent activation of non-ras transforming Syrian hamster cells initiated with chemical 5 (1990) 1425-1430. Pearson, W.R. and Lipman, quence comparison. 2444-2448.
D.J.: Improved
Proc.
Natl.
J. and DiPaolo,
tools
Acad.
J.A.:
sequences in neoplastic carcinogens. Oncogene
Sci.
for biological USA
85
se-
(1988)
Petersen, J.G.L. and Holmberg, S.: The ILV5 gene of Saccharomyces cerevisiae is highly expressed. Nucleic Acids Res. 14 (1986) 9631-9651. Ryan, E.D. and Kohlhaw, G.B.: Subcellular valine biosynthetic enzymes in yeast. 631-637. Shepherd, J.C.W.: Ancient patterns Enzymol. 183 (1990) 180&192.
localization of isoleucineJ. Bacterial. 120 (1974)
in nucleic acid sequences.
Methods
Sherman, F., Fink, G.R. and Hicks, J.B.: Methods in Yeast Genetics. Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1984. Sista, H. and Bowman, B.: Characterization of the ilv-2 gene from Neurospora crassa encoding a-keto-B-hydroxylacyl reductoisomerase. Gene 120 (1992) 115~118. Umbarger, H.E.: Amino acid biosynthesis
and its regulation.
Annu.
Rev. Biochem. 47 (1978) 533-606. Wahl, G.M., Stern, M. and Stark, G.R.: Efficient transfer of large DNA fragments from agarose gels to diazobenzyloxymethyl-paper and rapid hybridization by using dextran sulfate. Proc. Natl. Acad. Sci. USA 76 (1979) 3683-3687. Winston, F., Chumley, F. and Fink, G.R.: Eviction and transplacement of mutant genes in yeast, Methods Enzymol. 101 (1983) 21 l-228.
GENE
B.V. All rights
reserved.
179
0378-l 119/93/$06.00
07536
Cloning of the dihydroxyacid
dehydratase-encoding
gene (ILV3) from
Saccharomyces cerevisiae (Yeasts; branched-chain evolutionary
amino
conservation;
acids; mitochondrial
enzymes;
nucleotide
sequence;
gene organization;
in vitro transcription/translation)
Juan A. Velasco, Jo& Cansado, and Vicente Notario
M. Carmen
Pefia, Toshiaki
Kawakami*,
Jorge Laborda”
Division oiExperimenta1 Carcinogenesis, Department ofRadiation Medicine, Georgetown University Medical Center, Washington, DC 20007, USA Received by J.A. Gorman:
18 May 1993; Revised/Accepted:
30 July 1993; Received at publishers:
23 August
1993
SUMMARY
The biosynthesis
of branched-chain
amino
acids (aa) involves
three
shared
pathways
through
which pyruvate
or
cr-ketobutyrate are converted into cl-keto acids, precursors of valine, leucine or isoleucine. In eukaryotes, few of these common enzymes have been purified to homogeneity, and the whole complement of biosynthetic genes has not been cloned from a single species. In yeasts, most of these genes (ZLV genes) have been cloned and sequenced, with the exception of that coding for dihydroxyacid dehydratase (DAD, EC 4.2.1.9), the third enzyme in the common pathways. We have isolated Saccharomyces cerevisiae genomic sequences by hybridization to an oligodeoxyribonucleotide (oligo) probe designed from a highly conserved domain among bacterial DAD-encoding genes. The cloned sequences have been located to S. cerevisiae chromosome X, mapped within 0.4 centiMorgans (CM) of the ilv3 locus, and found to complement the ilv3 mutations of various yeast strains. Nucleotide (nt) and aa sequence analyses of the longest open reading frame (ORF) located within the cloned sequences identified them as the ZLV3 gene, which codes for the yeast DAD. With our cloning of ILV3, yeast becomes the only eukaryotic system from which all ILL’ genes have been cloned, thus allowing direct molecular analyses of their regulation.
INTRODUCTION
Branched-chain
amino
acids
(aa)
are
synthesized
through metabolic pathways well conserved in bacteria, yeast, filamentous fungi and plants (Kanamori and Wixom, 1963; Umbarger, 1978). Pyruvate and c+ketobutyrate are the initial substrates for the synthesis Correspondence to: Dr. V. Notario, Department of Radiation Medicine, Georgetown University Medical Center, P-427 Lombardi, 3800 Reservoir Rd., N.W., Washington, DC 20007, USA. Tel. (l-202) 687-2102; Fax (l-202) 687-2221. *Present addresses: (J.L.) Center for Biologics Evaluation and Research, F.D.A, N.I.H., Bldg. 29, 8800 Rockville Pike, Bethesda, MD 20892, USA. Tel. (l-301) 496-4038; (T.K.) Division of Immunology, La Jolla Institute for Allergy and Immunology, 11149 North Torrey Pines Road, La Jolla, CA 92037, USA. Tel. (1-619) 558-3500.
of valine and isoleucine through parallel steps which share three bifunctional mitochondrial (mt) enzymes: (a) acetohydroxy acid (AH) synthetase (AHAS, EC 4.1.3.18) which catalyzes substrate conversion to AH, (b) AH isomeroreductase (AAI, EC 5.4.99.3), which carries out AH isomerization and reduction to yield a,gdihydroxy acids, and (c) DAD, which catalyzes their
Abbreviations: aa, amino acid(s): AAI, AH isomeroreductase; AH, acetohydroxy acid(s); AHAS, AH synthetase; bp, base pair(s); CM, centiMorgans; DAD, dihydroxyacid dehydratase; ds, double strand(ed); ILV, genes encoding isoleucine/valine biosynthetic enzymes; ILV3, gene encoding DAD; kb, kilobase or 1000 bp; L., Lactococcus; mt, mitochondrial; nt, nucleotide(s); oligo, oligodeoxyribonucleotide; ORF, open reading frame; PAGE, polyacrylamide-gel electrophoresis; S., Saccharomyces; TD, threonine
deaminase.
180 dehydration producing a-ketoisovalerate and a-ketoB-methylvalerate. Transamination of these a-keto acids yields valine or isoleucine, respectively. Alternatively, a-ketoisovalerate, the a-keto acid precursor of valine, undergoes a condensation reaction with acetyl-CoA, as the first step towards the synthesis of leucine. Carbon flow into the isoleucine biosynthetic pathway is regulated by threonine deaminase (TD) activity, the first enzyme in which deaminates threonine into the pathway a-ketobutyrate (Umbarger, 1978). The regulation of these pathways has been studied primarily in prokaryotes, with most genes cloned and sequenced (Lawther et al., 1987; Godon et al., 1992; Grandoni et al., 1992). Direct regulation studies of branched-chain aa synthesis in eukaryotes have been hampered by the lack of molecular clones and/or sequence information for all the genes involved from a given species. Genetic mapping of yeast mutants defective in the biosynthesis of isoleucine and valine (i/u mutants) allowed the assignment of the three common activities in branched-chain aa synthesis to three unlinked loci ilv2 (AHAS), ilv4/ilv5 (AAI) and ilv3 (DAD); another locus (ilvf ) has been assigned to TD (Kadar and Wagner, 1964; Jones and Fink, 1982). Genes encoding AHAS have been cloned from the filamentous fungus Neurospora crassa (Jarai et al., 1990) and the yeast S. cerevisiae (Falco and Dumas, 1985); while the yeast ILV2 has been sequenced (Falco et al., 198.5) no sequence information has been reported on the fungal gene. Also, AAI-encoding genes have been cloned and sequenced from N. crassa (Sista and Bowman, 1992), S. cerevisiae (Petersen and 1986) and from spinach chloroplasts Holmberg, (Spinacea oleracea, Dumas et al., 1991). The TD-encoding gene (ZLVI) has been cloned and sequenced from S. cerevisiae (Kielland-Brandt et al., 1984). However, the molecular cloning of an eukaryotic DAD has not been reported to date. In the present study we report for the first time the isolation, identification, nt sequence and deduced aa sequence of an eukaryotic DAD-encoding gene, the ILV3 gene from S. cerevisiae.
RESULTS
AND DISCUSSION
(a) Detection and cloning of yeast sequences homologous to bacterial DAD-encoding genes On the basis of comparisons of available DAD nt sequences from several bacterial strains (Lawther et al., 1987; Godon et al., 1992) we designed oligo probes that might recognize the yeast homolog sequences when hybridized to S. cerevisiae genomic DNA. The optimum probe (designated bDAD) corresponded to a 20 nt sequence showing the maximum homology (85%) between Luctococcus lactis sbsp. lactis and Escherichiu coli
(Fig. 3) starting at aa 269 and 264, respectively, of the DAD sequences. Hybridization of EcoRI cleaved yeast genomic DNA to radiolabeled bDAD probe under high stringency conditions allowed the detection of a single, well defined, band of about 9 kb (Fig. 1A). Only after extended autoradiographic exposures could other not so well defined hybridizing bands be visualized. This result strongly suggested that the yeast DAD-encoding gene was encompassed within that unique EcoRI fragment. Screening of a yeast genomic phage h library with the bDAD probe resulted in the isolation of 24 clones, all of which appeared to contain a single EcoRI insert of about 9 kb which hybridized very strongly to the bDAD probe under high stringency conditions. Further restriction endonuclease analyses demonstrated that all 24 clones contained identical inserts, and localized the bDAD hybridizing sequences to a BamHI internal fragment of about 3.6 kb (Fig. 1B). This fragment was cleaved from clone h-ILV-19, isolated through preparative agarose electrophoresis, purified by electro-elution and ligated into pBR322 and the yeast integration vector YIPS for further structural and functional studies. Fig. IB shows the restriction map of the BumHI insert into YIp5 (designated YIP-ZLV) and the localization of the bDAD hybridizing sequences within the XbaIIBamHI subfragment. (b) Identification of the cloned sequences as the ZLV3 gene In order to demonstrate that the cloned bDAD homologous sequences were indeed the yeast ILV3 gene, three experimental protocols were developed in parallel: (a) chromosomal assignment, (b) gene mapping, and (c) complementation studies of ilv3 mutants. Fig. 2 shows the pattern of hybridization obtained when the purified 3.6-kb insert of Yip-ILV was used as the probe on a blot containing electrophoretically resolved S. cerevisiue chromosomes. These data unequivocally demonstrated that the cloned sequences were located in chromosome X of S. cerevisiae, which correlated with results from our gene mapping experiments. For locus assignment, YIP-ZLV was linearized at its unique BgrII site located within the insert (Fig. lB), to direct integration into the homologous chromosomal region. Crosses between URA3 transformants of strain 1992 containing properly integrated YIP-ZLV sequences and strains carrying multiple centromere markers (Table I) located the bDAD homologous sequences very close to ilv3, previously mapped to the right arm of chromosome X (Kadar and Wagner, 1964). This tight linkage (< 0.4 CM) suggested a physical distance to ilv3 of less than 1 kb. To confirm that ilv3 was present in the 3.6-kb BumHI insert, it was ligated into the unique BamHI site of the centromere vector YCp50 (Kuo and Campbell, 1983).
181
1 kb
B
A kb
II II
23.5,
B
E
B
+
6.6, I
XII-
\
/
9.6,
E \
1’ /I 1’
BP
\\ \\ \
VII.XV-
. YIP-ILV
R-Y
2.2, 2.1,
I 0
.
I
I
I
1
2
3
RY I kb
4
IXlllVII-
Fig. 1
Fig. 2
Fig. 1. Detection from protoplast EcoRI
and
and cloning of S. cereuisiae sequences homologous to bacterial DAD-encoding genes. (A) Total genomic DNA was prepared lysates of exponentially growing X2180 diploid yeast cells as described by Winston et al. (1983). Aliquots were cleaved with bacterial DAD-encoding genes with an oligo probe highly conserved among analyzed by Southern hybridization
(5’-GCXTTTGAAAAXGCCATXAC-3’). This probe, designated bDAD, was synthesized including all possible degeneracies at the three (underlined) non-conserved nt positions, and was end-labeled with [Y-~‘P]ATP (222 TBq/mmol) using T4 polynucleotide kinase. Hybridization was carried out under high stringency conditions as described by Wahl et al. (1979). The mobilities of the major hybridizing fragment (arrow) and the size markers are indicated. (B) Location of the bDAD-hybridizing sequences within the insert of YIpplLI/, a BamHI subfragment of the genomic 9-kb EcoRI product cloned into A-ILF-19. A h-L47.1 genomic library was prepared with X2180 genomic DNA digested to completion with EcoRI. A-ILV-19 was isolated and purified from this library by hybridization to the bDAD probe, and found to contain a 9-kb insert; restriction analysis showed the presence within that insert of a 3.6-kb BamHI fragment containig the bDAD-hybridizing sequences. This BamHI fragment was subcloned into the unique BarnHI site of the yeast integrative plasmid YIp5 (New England Biolabs) generating YIP-ILI/. E = EcoRI; B = BarnHI; Bc = BclI; X = Xbul; Bg = Bg/II; Sp = SphI; K = KpnI; RV = EcoRV. The heavy arrow the bDAD homologous sequences. Fig. 2. Chromosomal location of the cloned characteristic of chromosome X was detected
indicates
the position
and translational
orientation
of the unique
ORF located
within
bDAD-related yeast sequences. A single hybridizing band migrating to the electrophoretic position when the 3583-bp Yip-ILV BamHI insert was used to probe a commercial S. cerevisiae ‘Chromoblot’
(Clontech Laboratories) containing 5 ug/lane of electrophoretically resolved chromosomes. The mobility for hybridization and washing of the nylon membrane were as recommended by the manufacturers.
of all chromosomes
of YIP-ZLV
DNA
BumHI
insert
ilv3, ura3 haploid
Analysis
of nt sequence
from this construct was transformed into several strains. Strains GU86-1 lb (MATa ade2 his1 ilv3 ura3), GU91-22c (MATa adel his3 ilv3 lysl ura3), GU150-12b (MATa arg4 his4 ilv3 ura3), 3914-9A (MATa ura3 ilv3 from Reed B. Wickner), and IVPXS-2B (ATCC 64448, MATa his- leu2 ilv3-12 ura3) were used for complementation assays (Sherman et al., 1984). ZLV3, URA3 transformants were isolated in all cases. These results demonstrated that the cloned bDAD-hybridizing fragment encompassed S. cerevisiae ILV3 sequences capable of functionally complementing the ilv3 mutations used. (c) Analysis of nt and aa sequences of the cloned IL V3 gene In order to precisely locate the ILV3 coding sequences within the cloned fragment, the nt sequence of the 3.6-kb
insert
as 3583 bp, and
ORF
(Fig. lB), starting
was
defined
is indicated.
Conditions
determined
the length
determined
(Fig. 3).
of the cloned
the existence
at nt position
of an
1601 and ending
at 3116, encoding a polypeptide of 504 aa with a predicted M, of 60480. The position of the bDAD probe was located
within
the coding
region
Fig. 3 shows the nt sequence
defined
of the ORF
by this ORF. and its align-
ment with the E. coli and L. lactis sbsp. luctis sequences used
to design
S. cerevisiae
the bDAD
sequence
probe.
The
corresponding
shows 90% homology
with each
of the bacterial bDAD oligos. An overall nt homology search of DNA databases showed that the sequence of the cloned yeast ORF was about 50% homologous to the E. coli DAD-encoding gene and nearly 60% homologous to that of L. lactis sbsp. lactis.
182 TABLE
I
Genetic
linkage
of integrated
Yip-ILV
Cross
Parental
3892 3914
19921LV x 1972 3905-1lD x 3892-28D
“Only
tetrads
strainsb
showing
2+:2-
ILC’: : URA.3 were used for compiling
(ILV:
: URA3) to ilr;3”
PD’
NPD’
TC
53
0 0
0
86 segregation
for
0 both
iit13 and
these results.
bStrains used were: 1992 (MA?& spol I uru3 cunl cgh2 ude2 his7 hom3; Klapholz and Easton-Esposito, 1982); 19921LV, 1992(ILV: : URA3); 1792 (MATa gall his4 trpl horn3 ura3 CUPR ilo ude3 rad52 mul; Mortimer et al. 1981); 3905-11D (MA’& adr- urrr3 met3 ILV: : URA3), and 3892-28D (MA% ~73 i/r3 hom3). “PD, parental ditype; NPD, non-parental ditype: T, tetratype. IMethods: Yip-ILV was linearized by cleaving within the insert &III, and transformed into strain 1992 using method (Ito et al., 1983). Five URA transformants
with
the lithium acetate were crossed with
ura3 haploids and shown by their non-linkage to horn3 to not have integrated by recombination at URA3. One of these strains was then used to map the site of integration by standard mapping techniques (Mortimer
and Hawthorne,
1975).
Alignment of the deduced aa sequence of the yeast ORF with those of the DAD-encoding genes of E. coli and L. lnctis sbsp. tactis indicated (Fig. 4) that the translational products of the three genes were highly homologous. The yeast sequence showed about 39% identity with the E. coli enzyme, and 58% with that of Lactococcus over their entire sequences. Considering conservative aa substitutions, the yeast protein showed about 57% and 74% similarity with the enzymes of E. co/i and LuctococcUs, respectively. The extent of homology is close to 90% within several subdomains of the yeast protein (Fig. 4). In particular, the peptide defined by the nt sequence in the cloning bDAD probe (aa 202.. 208) shows 100% identity between Su~~~u~o~~~ce~and Lactococcus, and 100% similarity with E. cob. These data unequivocally identify the cloned sequences as the DAD-encoding ILV3 gene of S. cereuisiue. Because the three common enzymes involved in the synthesis of branched-chaiil aa are encoded by nuclear genes and subsequently translocated to mitochondria
GAGGCmri;AAAACGCCA~ACTTATGT~GTTGCAACCGGTGGGTCCA~T~TGCTGT~TTGCATTTG~TGGCTGTTG~TCACTCTGC~GGTGTC~G~TGTCACCAG~TGATTTCC~ ........I..7 *....A.. (L.lactis)
720
AGAATCAGTEATACTACACCftfiGATCGGTGACTTC~~CTTCTGGT~TACGTCAT~GCC~TTTG~TT~~GTTG~TGGTACCC~TCTGT~TT~GTATCTAT~TG~C~~
840
. . G. . . . . . . . . . . . . . G.. (&.coli)
TACACAAGAi;GTACTCTATCCAAGTATGCjAAGTTGGiCTTGA Fig. 3. The nt sequence
of the yeast ILV3 ORF.
The coding
sequence
of the longest
ORF
identified
within
the YIP-ILV
insert is shown
from the
putative starting ATG. Other ATG codons with sequence contexts favorable for translation initiation are also underlined. The yeast sequence hybridizing to the bDAD cloning probe is underlined in bold, and shown in comparison with the corresponding sequences of E. co/i and L. luctis. Identical nt are indicated by dots (.). Only the yeast nt positions are numbered. This sequence Methods: Subfragments of the 358%bp BornHI insert were subcloned into pBluescriptI1 SK. vector were used to determine the nt sequence of both strands of each DNA subfragment, Biochemical), as recommended by the manufacturers. Additional sequencing primers required were obtained
from Bio-Synthesis.
has been deposited in GenBank (Accession No. L[3975). Primers for the T7 and T3 sequences in the p3luescript using a Sequenase sequencing system, version 2.0 (US for sequencing internal portions of the longest fragments
183 SC ::
SC :i
. . . . . . . . . . . . . . . .. .. . . . . . . . . . . . . . . . . . .. . . . .. . . . . . . . . . . . . . . .. .
NEFKYNGKVE SVELNKYSKT LTPRSTQPAT QAHYYGIGFK DEDFKKAQVG IVSMDYDGNP . ........ . ..NPKYRSAT TTHGRNNAGA RALWRATGNT DADFGKPIIA VVNSFTQFVP
60 48
QFN-ll GVSDGISNGT KGNRVSLQSR EIIADSFETI ::I:: :::::::::: L::::::V:: : ::::l::N I:::: I:D:::A::H 6::L:::P:: :I::::V:VI
47
...HFSIIEK RLK....... . . CNNHLGTLGS KI:SSVNQTD GL GHVNLRDLG. LVAEQIEAAG 6.
107 174 162
SC :i
SC Ll EC
SC Ll EC
167 234 222
SAFQSYGEYI SKQFTEEERE DVVE
AVSKEK.... LAECDNIGEY IKKTHELG.. .ILP ILTK EAFENAITVV VATGGSTNAV :::Q::.... QE:::D::LA ::NLL:KD.. .:K:!fT :THADRKQLF :NAGKR:V:L T:RYY:QNDE SA:::l: I: A:::::I:LD :H:::::l:
220 287 282 279 346 342
SC Ll EC
329 397 402
K KAPSLP.EGQ EIIKPLS...
SC Ll EC
:; EC
:z
SLAPGGAV.. GKITGKEGTY .... .. ... . ... ....... ....FKGRAR VFEEEGAFIE N::Q::S:.. A::S::::EF .... .. ... . .......... .. ... ..T.. ::DG:QH::D DR:N:VSARW NTP:A:TAAU RCSTVILRKR LHRETAGVDD SILK:T:P:K :Y:SQDDAV:
363 431 462
:5
423 489 520
SC LJ EC
481 547 580
SC
SC Ll EC
.......TRG TLSKVAKLVS NASNGCVLDA l .... . .. ........ V:A:F:::TR P::E:::T:L l . ... . PKNRERQVSF A:RA::S:AT S:DK:A:RHK SKLGG'
504 570 615
Fig 4. Alignment of the DAD aa sequences from L. lactis (LI) and E. (Ec) with that deduced from the ILV3 ORF (SC). Domains with conservation greater than 85% are boxed. Identical aa are indicated by coli
(:), conservative
substitutions
by (I). sequence
gaps introduced
for opti-
mum alignment by (.), and translation termination codons by (*). The yeast sequence shows a 74% similarity with the 15. lactis enzyme over the entire protein sequence, and a 57% with that of E. coli. Most DNA and protein analyses were performed with the software package PC/Gene. ORFs were found using the methods of Fickett (1982) and Shepherd (1990). Homology searches in DNA and protein databases (GenBank release 69.0, EMBL release 26.0, NBRF-protein release 26.0, and Swissprot release 18.0) were performed with the program FASTA (Pearson and Lipman, 1988). The method of Needleman and Wunsch (1970) was used to study the statistical significance of the homologies found.
(Ryan and Kohlhaw, 1974), we analyzed the deduced ZLV3 aa sequence for the presence of putative mt-targeting sequences. In general, these sequences are rich in positively charged and hydroxylated aa residues, contain few or no negatively charged residues, and can form short amphiphilic cl-helical structures (Hart1 et al., 1989). Fourteen of the 37 N-terminal aa residues of the ILV3 protein are either positively charged or hydroxylated, and only two negatively charged residues are encompassed within the same domain. In addition, computer analysis (Chou and Fasman, 1978) for predictions on protein structure suggested that these amino terminal residues could assume an amphiphilic cl-helical confor-
mation.
Thus, it seems very likely that these first 37 aa
are part cannot
of the ILV3 mt-targeting be demonstrated
yeast DAD is purified
until
signal,
although
the mature
form
this of the
to homogeneity.
(d) In vivo and in vitro expression In order to determine the level and pattern
of expres-
sion of ZLV3, RNA from S. cereuisiae cells harvested various purified
phases of growth from asynchronous (Elder
hybridization XbaIIKpnI recognized appear
et al., 1983) and analyzed
was
by Northern
with ZLV3-specific probes (fragments and KpnI-BarnHI, Fig. 1B). Both probes a single
transcript
to be highly
expressed
throughout
cultures
at
the different
phases
of about
2.3 kb which
in a constitutive of growth
fashion
(Fig. 5A,B).
The same transcript was also recognized by a probe corresponding to the upstream BarnHI-XbaI fragment. This, when considered in addition to the position of the ILV3 ORF within the YIP-ZLV insert and the ability of YIP-ZLV to complement ilu3 mutants carrying presumably different mutations, strongly suggests that the entire ZLV3 transcriptional unit is located within the cloned 3.6-kb BamHI fragment. These data are consistent with the genetic definition of ilv3 as a single yeast locus (Kadar and Wagner, 1964), and agree with our finding of a single genomic fragment hybridizing to the bDAD probe. The capacity of the cloned ORF to synthesize a protein of the expected 60.48 kDa, was studied by in vitro coupled transcription and translation assays, using rabbit reticulocyte lysate and pGEM-3Z and pSP6bpoly(A) constructs containing the ILV3 insert. As shown in Fig. 5C, several polypeptides were translated from the ZLV3 insert when placed in sense orientation relative to the SP6 promoter, whereas only background signals could be detected when antisense constructs were used. The largest of the most abundant polypeptides was about 60 kDa, indicative of translation initiation at the first AUG in the ZLV3 ORF. The other major species may result from translation initiation at internal AUG codons, which is known to occur in rabbit reticulocyte lysates (Kozak, 1990; Jackson, 1991). Indeed, the AUG codons at aa 49, 71, 82 and 149 (Fig. 3) lay within sequence contexts defined as favorable for translation initiation (Kozak, 1981). The only eukaryotic DAD purified to homogeneity to date has been that from spinach leaves (Flint and Emptage, 1988). The native enzyme is a dimer of 110 kDa which consists of two monomers of 63 kDa and a [2Fe-2S] cluster. The yeast ILV3 contains two Cys residues (aa 62 and 140) themselves conserved in E. co/i and Lactococcus and located within aa domains highly conserved in bacteria; these two Cys are optimum candidate
184
A
123456
c
12 __ *
kDa
that the yeast ILV3 protein may adopt this type of metalbound dimeric conformation. (e) Conclusions (1) A genomic fragment encompassing which codes for the branched-chain
the ILV3 gene, aa biosynthetic
enzyme DAD, has been cloned from the yeast S. cerevisiae by hybridization to an oligo probe designed on the basis
B
of the high level of conservation
among
bacterial
DAD
sequences. (2) Identification
of the cloned
gene has been established
Fig. 5. In viva (A) Hybridization
and in vitro expression of the ILV.3 gene. of total RNA extracted from S. cereuisiue X2180 cells
collected after 3 h (lane 1 ), 6 h (lane 2) 9 h (lane 3), 12 h (lane 4) 18 h (lane 5) and 24 h (lane 6) with the XhaI-KpnI fragment of the Yip-1LV insert detected a single 2.3-kb transcript under high stringency conditions. The ds probe was radiolabeled by random [E--“2P]dCTP (111 TBq/mmol) using the Megaprime
priming with DNA labeling
system (Amersham) and purified through Sephadex G-50 spin-columns. Specific activity was 4 x 10’ cpm/pg. Blots were washed three times at room temperature in 2 x SSC/O. 1% SDS, for 5 min each time, and then twice for 30 min each, at the hybridization temperature (42’C) in 0.I x SSC/O.l% SDS. SSC is 0.15 M NaCl/O.OlS M Naacitrate, pH 7.0. Dried blots were exposed at -70°C to Kodak XAR-2 film, with intensifying screens. The migration of the ribosomal RNA markers is indicated. Equal sample loading was checked by staining the same gel with ethidium bromide(B) prior to blotting and hybridization. In in vitro coupled assays, SP6-derived ‘sense’ ILV3 transcripts were efficiently translated by rabbit reticulocyte lysates (C, lane l), whereas transcripts of ‘antisense’ constructs failed to direct efficient polypeptide translation (C, lane 2). The 3583”bp BaniHl fragment, encompassing the longest ORF classified as coding by the two methods mentioned in the Fig. 4 legend. was ligated into the unique BanzHI site in the plasmid vectors pGEM-3Z and pSP64-poly(A) (Promega, Madison, WI, USA). DNA prepared from constructs containing the insert in sense or antisense orientation relative transcription and
to the SP6 promoter translation assays
was used for in vitro coupled using the ‘TNT Coupled
Reticulocyte Lysate System’ from the same company. This system uses a rabbit reticulocyte lysate for the translation of transcripts synthesized with recombinant SP6 or T7 RNA polymerase from the respective promoter. Reactions were carried out as recommended by the manufacturers in the presence of L-[“5S]methionine (Amersham, in vivo cell labeling grade, 37 TBq/mmol). Translation products were resolved by SDS-PAGE as described (Notario et al., 1990) and visualized by autoradiography. The arrow indicates the migration of the largest fLV3 derived 60-kDa polypeptide.
aa for the formation of putative [2Fe-2S] clusters. Rabbit reticulocyte lysates provide an iron-rich nonoxidizing environment where Fe-S linkages can be formed and preserved. The polypeptides of about 1 lo-115 kDa detected by in vitro translation of ILV3 sequences might be homo- or heterodimeric forms of the major monomeric polypeptides in which the stability of the [Fe-S] bridges has been preserved. Thus, it is likely
sequences
as the ILV3
on the basis of(i) localization
in chromosome X (lOR), (ii) genetic mapping within 0.4 CM of the ZLV3 locus, and (iii) ability to complement the i/u3 mutations
in several haploid
strains.
(3) Analysis of nt sequence confirmed the identity of the cloned sequences as the DAD-encoding gene and showed that the yeast sequence is highly homologous (up to about 60%) to those of bacterial counterparts, indicating a high level of evolutionary conservation. The yeast sequences contain an ORF with coding capacity for a protein of 504 aa (60.48 kDa), which is highly conserved (up to 74% similarity) relative to the bacterial enzymes. The size of the predicted ILV3 protein product matched that of the largest polypeptide species synthesized in in vitro transcription/translation assays. (4) A single ILV3 transcript of about 2.3 kb has been detected in total yeast RNA. The mRNA is transcribed in a constitutive fashion, being expressed at high levels in RNA prepared from yeast cells harvested at different stages of asynchronous growth. (5) With our cloning of the ILV3 gene, yeast becomes the only eukaryotic system from which all ILV genes have been cloned from, thus allowing direct molecular analyses of their regulation,
ACKNOWLEDGEMENTS
The authors would like to thank Reed B. Wickner for his help in genetic mapping experiments, Keith C. Robbins and Stuart A. Aaronson in whose laboratory this project was started, and Anatoly Dritschilo for his support and discussions during its development.
REFERENCES Chou, P.Y. and Fasman,
G.D.: Empirical
predictions
of protein
confor-
mation Annu. Rev. Biochem. 47 (1978) 251-276. Dumas, R., Lebrun, M. and Deuce, R.: isolation, characterization and sequence analysis of a full-length cDNA clone encoding acetohydroxy acid reductoisomerase from spinach chloroplasts. Biochem. J. 277
( 1991) 4699475.
185 AUG initiator
Elder, R.T., Loh, E.Y. and Davies, R.W.: RNA from yeast transposable element Tyl has both ends in the direct repeats, to retrovirus
a structure
Kozak,
RNA. Proc. Natl. Acad. Sci. USA 80 (1983) 2432-2436.
Falco, SC. and Dumas, K.S.: Genetic analysis of mutants of Saccharomyces cereuisiae resistant to the herbicide sulfometuron methyl. Genetics 109 (1985) 21-35. Falco, SC., Dumas, K.S. and Livak, K.J.: Nucleotide sequence of the yeast ILV2 gene which encodes acetolactate synthase. Nucleic Acids Res. 13 (1985) 401 l-4027. Fickett, J.W.: Recognition of protein
contains
a [2Fe-2S]
3558-3564. Godon, J.J., Chopin, biosynthesis
cluster.
J. Biol.
Chem.
l-45. lto, H., Fukuda,
from spin263
(1988)
genes in Lacrococcus lads
subsp.
lactis. J. Bacterial.
protein
of Bacillus subtilis. N.,
Nicholson,
import.
Y., Murata,
D.W.
Biochim.
regulation
J. Bacterial. and
Biophys.
K. and Kimura,
174 (1992)
Neupert,
W.:
Acta 988 (1989)
A.: Transformation
of
tion initiation in rabbit reticulocyte lysates: unique features of encephalomyocarditis virus RNA translation. Biochim. Biophys. Acta 1088 (1991) 3455358. Jarai, G., Yagmai, B., Fu, Y.-H. amino
acid
and
Marzluf,
biosynthesis
G.A.:
Regulation
in Neurospora
of
crassa.
Cloning and characterization of the leu-f and i/u-3 genes. Mol. Gen. Genet. 224 (1990) 3833388. Jones, E.W. and Fink, G.R.: Regulation of amino acid and nucleotide biosynthesis in yeast. In: Strathern, J.N., Jones, E.W. and Broach, J.R. (Eds.), The Molecular Biology of the Yeast Saccharomyces. Metabolism and Gene Expression, Coid Spring Harbor Laboratory, Cold Spring Harbor, NY, 1982, pp. 181-299. Kadar, S.N. and Wagner, R.P.: Genetic and biochemical isoleucine-valine mutants Kanamori, M. and Wixom,
analysis
of
of yeast. Genetics 49 (1964) 2133222. R.L.: Studies in valine biosynthesis.
V.
Characteristics of the purified dihydroxyacid dehydratase from spinach leaves. J. Biol. Chem. 238 (1963) 998-1005. Kielland-Brandt, M.C., Holmberg, S., Petersen, J.G.L. and NilssonTillgren, T.: Nucleotide sequence of the gene for threonine deaminase ilu-l of Saccharomyces cereuisiae. Carlsberg Res. Commun. 49 (1984) 5677575. Klapholz, S. and Easton-Esposito,
Nucleic
of the fidelity
of initiation
Acids Res.
lysates from commercial
sources.
of translation
Nucleic
in
Acids Res. 18
(1990) 2828. Kuo, C.-L., and Campbell, J.L.: Cloning of Saccharom~ces cereuisiue DNA replication genes. Isolation of the CDC8 gene and two genes that compensate the cd&l mutation. Mol. Cell. Biol. 3 (1983)
operon of Escherichia 2137-2155.
R.: A new mapping
method
R.K.
and
coli K-12.
Hawthorne,
Nucleic
D.C.:
Acids
Genetic
Res.
15 (1987)
mapping
in yeast.
Methods Cell Biol. 11 (1975) 221-233. Mortimer, R.K., Contopoulou, R. and Schild, D.: Mitotic
chromosome
loss in a radiation-sensitive strain of the yeast Saccharomyces cere1C.v iae. Proc. Natl. Acad. Sci. USA 78 (1981) 5778-5782.
intact yeast cells treated with alkali cations. J. Bacterial. 153 (1983) 1633168. Jackson, R.J.: Potassium salts influence the fidelity of mRNA transla-
branched-chain
M.: Evaluation
reticulocyte
Mortimer, amino acid
174 (1992) 6580-6589. Grandoni. J.A., Zahler, S.A. and Calvo, J.M.: Transcriptional
Mitochondrial
ribosomes.
Lawther, R.P., Wek, R.C., Lopes, J.M., Pereira, R., Taillon, B.E. and Hatfield, G.W.: The complete nucleotide sequence of the iluGMEDA
regions in DNA sequences.
M.C. and Erlich, S.D.: Branched-chain
of the ilu-lea operon 3212-3219. Hartl, F.-U., Pfanner,
by eukaryotic
1730&1737. coding
Nucleic Acids Res. 10 (1982) 5303-5318. Flint, D.H. and Emptage, M.H.: Dihydroxy acid dehydratase ach
codon
9 (1981) 5233-5252.
similar
employ-
ing a meiotic ret-mutant of yeast. Genetics 100 (1982) 387-412. Kozak, M.: Possible role of flanking nucleotides in recognition of the
Needleman, S.B. and Wunsch, CD.: A general method applicable to the search for similarities in the amino acid sequence of two proteins. J. Mol. Biol. 48 (1970) 443-453. Notario, V., Castro, R., Flessate, D.M., Doniger. Frequent activation of non-ras transforming Syrian hamster cells initiated with chemical 5 (1990) 1425-1430. Pearson, W.R. and Lipman, quence comparison. 2444-2448.
D.J.: Improved
Proc.
Natl.
J. and DiPaolo,
tools
Acad.
J.A.:
sequences in neoplastic carcinogens. Oncogene
Sci.
for biological USA
85
se-
(1988)
Petersen, J.G.L. and Holmberg, S.: The ILV5 gene of Saccharomyces cerevisiae is highly expressed. Nucleic Acids Res. 14 (1986) 9631-9651. Ryan, E.D. and Kohlhaw, G.B.: Subcellular valine biosynthetic enzymes in yeast. 631-637. Shepherd, J.C.W.: Ancient patterns Enzymol. 183 (1990) 180&192.
localization of isoleucineJ. Bacterial. 120 (1974)
in nucleic acid sequences.
Methods
Sherman, F., Fink, G.R. and Hicks, J.B.: Methods in Yeast Genetics. Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1984. Sista, H. and Bowman, B.: Characterization of the ilv-2 gene from Neurospora crassa encoding a-keto-B-hydroxylacyl reductoisomerase. Gene 120 (1992) 115~118. Umbarger, H.E.: Amino acid biosynthesis
and its regulation.
Annu.
Rev. Biochem. 47 (1978) 533-606. Wahl, G.M., Stern, M. and Stark, G.R.: Efficient transfer of large DNA fragments from agarose gels to diazobenzyloxymethyl-paper and rapid hybridization by using dextran sulfate. Proc. Natl. Acad. Sci. USA 76 (1979) 3683-3687. Winston, F., Chumley, F. and Fink, G.R.: Eviction and transplacement of mutant genes in yeast, Methods Enzymol. 101 (1983) 21 l-228.