The Escherichia coli fmt Gene, Encoding Methionyl-tRNA Met Formyltransferase, Escapes Metabolic Control
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
JOURNAL OF BACTERIOLOGY, Feb. 1993, p. 993-1000 Vol. 175, No. 4
0021-9193/93/040993-08$02.00/0
Copyright C 1993, American Society for Microbiology
The Escherichia coli fmt Gene, Encoding Methionyl-tRNA Met
Formyltransferase, Escapes Metabolic Control
THIERRY MEINNEL, JEAN-MICHEL GUILLON, YVES MECHULAM,* AND SYLVAIN BLANQUET
Laboratoire de Biochimie, Unite de Recherche Associee no. 240 du Centre National de la Recherche
Scientifique, Ecole Polytechnique, F-91128 Palaiseau Cedex, France
Received 8 September 1992/Accepted 2 December 1992
The genetic organization near the recently cloned fint gene, encoding Escherichia coli methionyl-tRNAV et
formyltransferase (J. M. Guillon, Y. Mechulam, J. M. Schmitter, S. Blanquet, and G. Fayat, J. Bacteriol.
174:4294-4301, 1992), has been studied. The fint gene, which starts at a GUG codon, is cotranscribed with
Downloaded from http://jb.asm.org/ on March 30, 2021 by guest
another gene, fins, and the transcription start site of this operon has been precisely mapped. Moreover, the
nucleotide sequence of a 1,379-bp fragment upstream fromflint reveals two additional open reading frames, in
the opposite polarity. In the range of 0.3 to 2 doublings per h, the intracellular methionyl-tRNA;Iet
formyltransferase concentration remains constant, providing, to our knowledge, the first example of a gene
component of the protein synthesis apparatus escaping metabolic control. When the gene fusion technique was
used for probing, no effect onfint expression of the concentrations of methionyl-tRNA;tet formyltransferase or
tRNA;4eI could be found. The possibility that theJint gene, the product of which is present in excess to ensure
full N acylation of methionyl-tRNA;4et, could be expressed in a constitutive manner is discussed.
N formylation of initiator methionyl-tRNAMet (Met- MATERIALS AND METHODS
tRNAfMet) has been shown to strongly stimulate the rate of in Bacterial strains and plasmids. The E. coli K-12 strains,
vitro protein synthesis from an RNA template (17). The plasmids, and phages used in this study are listed in Table 1.
formylation reaction is catalyzed by 10-formyltetrahydrofo- General genetic methods were as described by Miller (24).
late:L-methionyl-tRNArfet N-formyltransferase (formylase Bacteria were grown either in MOPS minimal medium (25)
[FMT]). The specificity of the formylation reaction for supplemented with various carbon sources or in Luria-
Met-tRNAfet is governed by the sequence of the acceptor Bertani (LB) medium. When needed, culture media were
stem of tRNAfMet (11, 18). supplemented with Casamino Acids (final concentration,
The fint gene, encoding FMT in Escherichia coli, has been 0.2% [wt/vol]) or methionine (0.2 mM). Antibiotics were
characterized and disrupted (10). fnt mutant strains fail to used at concentrations of 50 pug/ml (ampicillin) or 25 jig/ml
grow at 420C and exhibit a strong reduction in the cellular (chloramphenicol).
growth rate at 370C (10). Actually, the growth parameters of Recombinant DNA techniques. General recombinant DNA
E. coli are particularly sensitive to the intracellular level of techniques were as described previously (22, 29, 32). The
FMT, as determined by the construction and study of a set of nucleotide sequence was determined by the dideoxy chain
strains expressing various reduced levels of FMT. Conse- termination method (30) with an ALF DNA sequencer
quently, FMT appears to play an essential role, most likely (Pharmacia). For primer extension analysis, the 21-mer
at the level of translation initiation (10). oligonucleotide 5'-CACTTCTTCTACCGGTFTAGC-3',
In our previous work, fint was proposed to be part of an corresponding to the sequence starting 117 bases down-
operon comprising another open reading frame (ORF), fins, stream from the XbaI restriction site within the fins ORF
located upstream fromfint (10). Moreover, the translation of (10), was synthesized. Two picomoles of this oligonucleotide
was 5' labeled with 10 _uCi of [y-32P]ATP in 10 RA (3,000
fint starts with the unusual codon GUG. Such a non-AUG Ci/mmol; NEN) and used as a template in primer extension
initiation codon might influence the efficiency of fint trans- analysis with 150 pug of total RNA extracted from strain
lation (3, 16, 27) and serve, in addition, as a regulatory signal JM1O1Tr(pBS936). The general procedure for this analysis
(4, 8). was as described previously (19). The resulting 32P-labeled
In this study, we further characterize the operon contain- extension products were analyzed on a 6% polyacrylamide
ing the fint gene and, in particular, its control region. The sequencing gel in parallel with DNA sequencing reaction
levels of FMT as a function of the cellular growth rate are performed with the same oligonucleotide and double-
measured. Moreover, the steady-state levels of expression stranded pBS936 DNA as a template.
of a reporter gene, lacZ, placed under the control of the Enzyme specific activity measurements. tRNA formylation
transcriptional and translational signals of flint, are mea- reactions were performed at 250C with 100-pd assay-mixtures
sured in cells containing altered concentrations of FMT or as described previously (1, 10). E. coli tRNA et (1,500
tRNAfMet. Finally, the influence of the GUG start codon on pmol/A260 unit), prepared from an overproducing strain (21),
the efficiency of fint expression is assayed. was used as a substrate. Valylation reactions were also
performed at 250C with 100-pI assay mixtures as described
previously (10). Unfractionated tRNA from E. coli (Boehr-
inger Mannheim) was used as a substrate.
One unit of FMT activity was defined as the amount of
enzyme capable of formylating 1 pmol of Met-tRNAfMet per
*
Corresponding author. s in 1 ml of the standard assay mixture. One unit of
993994 MEINNEL ET AL. J. BACTERIOL.
TABLE 1. E. coli strains, plasmids, and phages used in the present study
Strain, plasmid, Genotype or marker(s) Reference or source
or phage
K37 galK rpsL 23
JM1OMTr supE thi A(lac-pro) recA56 srl-300::TnlO F' (traD36 proAB+ lacl lacZAM15) 12
PAL13TrkFatg Derivative of JMOMTr; fntAl::kan; XFatg monolysogen 10
JMX14 Derivative of JM1OMTr; X14 monolysogen This work
JMX14AUG Derivative of JM1OMTr; X14AUG monolysogen This work
JMX15 Derivative of JM1O1Tr; X15 monolysogen This work
PAL14Tr Derivative of Pall3TrXFatg; X14 monolysogen This work
PAL15Tr Derivative of Pall3TrXFatg; X15 monolysogen This work
pBluescript SK+ bla Stratagene
pBSTNAV bla 20
pBStRNAr¶et bla; overexpresses tRNA$ et 21
pBStRNA:~et
pACYC184
bla; overexpresses tRNAmet 20
Cmr 5
Downloaded from http://jb.asm.org/ on March 30, 2021 by guest
pACform pACYC184 derivative; fint-i (ATG); overexpresses FMT 9
pEBNB Derivative of pBluescript SK+ having the BglII and NcoI sites between the This work
EcoRI and BamHI sites of the polylinker
pBS936 bla fmt (contains an 8-kb chromosomal insertion of theflmt region) 10
pEform bla fint'; derivative of pEBNB having the BamHI-NcoI fragment of pBS936 This work
inserted between the BglII and NcoI sites of pEBNB
pRS414 bla 'lacZYA 33
pRS415 bla lacZYA 33
pBN14 bla fmt::lacZ (in frame); derivative of pRS414 having the EcoRI-BamHI This work
fragment of pEform inserted between the EcoRI and BamHI sites
pBN14AUG Same as pBN14 but with an AUG codon as the translation start site of This work
flnt: :lacZ
pBN15 bla flnt::lacZ; derivative of pRS415 having the EcoRI-BamHI fragment of This work
pEform inserted between the EcoRI and BamHI sites
pE, bla flnt::lacZ; derivative of pBN15 with an EcoRI-E, deletion This work (see Fig. 2)
pE2 bla flnt::lacZ; derivative of pBN15 with an EcoRI-E2 deletion This work (see Fig. 2)
pE3 bla fmt::lacZ; derivative of pBN15 with an EcoRI-E3 deletion This work (see Fig. 2)
pE4 bla fmt::lacZ; derivative of pBN15 with an EcoRI-E4 deletion This work (see Fig. 2)
pAX bla fint::lacZ; derivative of pBN15 with an EcoRI-XbaI deletion This work (see Fig. 2)
pAS bla fint::lacZ; derivative of pBN15 with an EcoRI-SmaI deletion This work (see Fig. 2)
XRS45 i21 lacZYA 33
XFatg flmt-i (ATG) under the control of the lac operon promoter 10
X15 Derivative of XRS45 obtained by homologous recombination with pBN15 This work
X14 Derivative of XRS45 obtained by homologous recombination with pBN14 This work
X14AUG Derivative of XRS45 obtained by homologous recombination with pBN14AUG This work
valyl-tRNA synthetase activity was defined as the amount of paper has been assigned GenBank-EMBL accession number
enzyme capable of aminoacylating 1 pmol of tRNAVal per s X65946.
in 1 ml of the standard assay mixture.
For measurement of intracellular enzymatic activities in RESULTS AND DISCUSSION
the cultures, two 3-ml cell samples were withdrawn twice
during exponential growth, at optical densities at 650 nm of Nucleotide sequence of the chromosomal region upstream
-0.2 and -0.5. 13-Galactosidase activities in toluene-perme- from fins. The fint gene could be the 5'-distal cistron of an
abilized cells were determined as described by Miller (24). operon since (i) its putative ribosome binding site (RBS)
For FMT and valyl-tRNA synthetase activity measure- overlaps the translation stop codon of another ORF, of 167
ments, cells were centrifuged, resuspended in 1 ml of 20 mM amino acids, fms, which also is preceded by an RBS; (ii) no
Tris-HCl (pH 7.5)-0.1 mM EDTA-0.15 M KCl-10 mM putative transcription initiation site on the DNA sequence
2-mercaptoethanol, and disrupted by sonication. The values can be recognized immediately upstream from fint; and (iii)
reported represent the average for the two samples. fint is followed by a sequence that has dyad symmetry and a
Protein concentrations in the crude extracts were mea- stretch of T residues and resembles a rho-independent
sured by the technique of Bradford (2) by use of a Bio-Rad transcription terminator (28).
protein assay kit with bovine serum albumin as a standard. To search for the promoter(s) responsible for the tran-
tRNA;4et and N-acylmethionyl-tRNA concentrations in vivo.
For determination of the concentrations of tRNAret and
scription of both fins and fint, we determined the DNA
sequence of the 1,379-bp BamHI-XbaI DNA fragment lo-
N-acylmethionyl-tRNA in vivo, 250 ml of exponentially cated upstream from fins (10). Analysis of the DNA se-
growing cells in LB medium was collected by centrifugation quence (Fig. 1) revealed two additional ORFs, in the oppo-
and resuspended in 1 ml of 0.3 M sodium acetate (pH 4.8) site polarity compared with fms and flnt. The first ORF,
containing 10 mM EDTA. Thereafter, the procedure was as called smf, starts 130 bp upstream from the fms start codon,
described previously (10). on the opposite DNA strand. It encodes 311 amino acids and
Accession number. The DNA sequence described in this is preceded by a putative RBS. The second ORF, calledVOL. 175, 1993 EXPRESSION OF THE E. COLI FORMYLASE GENE 995
90
GCGAAGCCGCTCGTCCGGAATATGTAACACTTGCAAAACTGACATAAATCTCCAGAGATGTGTTCAGGAGTTAGAAAGATTATTTCTTCT
ArgLeuArgGluAspProIleHisLeuValGlnLeuValSerMet
fms
XbaI 180
ATTCTAGACAAATCCCCCTCTGATTGACAGCATCACTGACCAATCGCAAAGATTGCTAAGGCTGCTTATGGCAGGGAGATAAGGATGGTC
MetVal
I 270 I
GATACAGAAATTTGGCTGCGTTTAATGAGTATCAGCAGCTTGTACGGCGATGATATGGTCCGTATCGCTCACTGGGTGGCAAAACAGTCG
AspThrGluIleTrpLeuArgLeuMetSerIleSerSerLeuTyrGlyAspAspMetValArgIleAlaHisTrpValAlaLysGlnSer
I 360
CATATTGATGCGGTTGTATTGCAGCAAACAGGGCTTACATTGCGGCAGGCACAACGCTTTCTTTCATTTCCACGAAAGAGTATCGAAAGC
HisIleAspAlaValValLeuGlnGlnThrGlyLeuThrLeuArgGlnAlaGlnArgPheLeuSerPheProArgLysSerIleGluSer
I 450
TCACTTTGTTGGTTGGAGCAACCCAACCATCATTTAATTCCTGCGGACAGCGAATTTTATCCTCCTCAACTTCTGGCGACGACAGATTAC
SerLeuCysTrpLeuGluGlnProAsnHisHisLeuIleProAlaAspSerGluPheTyrProProGlnLeuLeuAlaThrThrAspTyr
Downloaded from http://jb.asm.org/ on March 30, 2021 by guest
I 540
CCCGGCGCACTGTTTGTTGAAGGAGAACTGCACGCGCTGCATTCATTTCAGCTTGCCGTAGTGGGGAGTCGGGCGCATTCATGGTATGGC
ProGlyAlaLeuPheValGluGlyGluLeuHisAlaLeuHisSerPheGlnLeuAlaValValGlySerArgAlaHisSerTrpTyrGly
ClaI 630
GAGCGATGGGGACGATTATTTTGCGAAACTCTGGCGACGCGTGGAGTGACAATTACGAGTGGACTGGCGCGTGGAATCGATGGTGTAGCG
GluArgTrpGlyArgLeuPheCysGluThrLeuAlaThrArgGlyValThrIleThrSerGlyLeuAlaArgGlyIleAspGlyValAla
720
CATAAAGCAGCCTTACAGGTAAATGGCGTCAGCATTGCTGTATTGGGGAATGGACTTAATACCATTCATCCCCGCCGTCATGCCCGACTG
HisLysAlaAlaLeuGlnValAsnGlyValSerIleAlaValLeuGlyAsnGlyLeuAsnThrIleHisProArgArgHisAlaArgLeu
810
GCTGCCAGTCTGCTTGAACAGGGGGGCGCTCTCGTCTCGGAATTTCCCCTCGATGTTCCACCCCTTGCTTACAATTTCCCACGAAGAAAT
AlaAlaSerLeuLeuGluGlnGlyGlyAlaLeuValSerGluPheProLeuAspValProProLeuAlaTyrAsnPheProArgArgAsn
900
CGCATTATCAGTGGTCTAAGTAAAGGTGTACTGGTGGTGGAAGCGGCTTTGCGTAGTGGTTCGCTGGTGACAGCACGTTGTGCGCTTGAG
ArgIleIleSerGlyLeuSerLysGlyValLeuValValGluAlaAlaLeuArgSerGlySerLeuValThrAlaArgCysAlaLeuGlu
990
CAGGGGCGAGAAGTTTTTGCCTTGCCAGGTCCAATAGGGAATCCGGGAAGCGAAGGGCCTCACTGGTTAATAAAACAAGGTGCGATTCTT
GlnGlyArgGluValPheAlaLeuProGlyProIleGlyAsnProGlySerGluGlyProHisTrpLeuIleLysGlnGlyAlaIleLeu
1080
GTGACGGAACCGGAAGAAATTCTGGAAAACTTGCAATTTGGATTGCACTGGTTGCCAGACGCCCCTGAAAATTCATTTTATTCACCAGAT
ValThrGluProGluGluIleLeuGluAsnLeuGlnPheGlyLeuHisTrpLeuProAspAlaProGluAsnSerPheTyrSerProAsp
SacI 1170
CAGCAAGATCGTGGCATTGCCATTTCCTGAGCTCCTGGCTAACGTAGGAGATGAGGTAACACCTGTTGACGTCGTCGCTGAACGTGCCGG
GlnGlnAspArgGlyIleAlalTeSerSTOP
1260
CCAACCTGTGCCAGAGGTAGTTACTCAACTACTCGAACTGGAGTTAGCAGGATGGATCGCAGCTGTACCCGGCGGCTATGTCCGATTGAG
I 1350
:AGGGCATGCCATGTTCGACGTACTAATGTATTTGTTTGAAACCTATATTCACACAGAAGCTGAGTTGCGTGTGGATCAAGACAAACTTG
MetPheAspValLeuMetTyrLeuPheGluThrTyrIleHisThrGluAlaGluLeuArgValAspGlnAspLysLeuGlu
smg -
I 1440
AACAGGATCTTACCGACGCAGGTTTTGAGCGAGAAGATATCTACAATGCCCTGCTATGGCTGGAAAAACTTGCTGATTATCAGGAAGGGT
GlnAspLeuThrAspAlaGlyPheGluArgGluAspIleTyrAsnAlaLeuLeuTrpLeuGluLysLeuAlaAspTyrGlnGluGlyLeu
BamHI 1469
TGGCAGAACCGATGCAACTGGCCTCGGATCC
AlaGluProMetGlnLeuAlaSerAspPro
FIG. 1. Nucleotide sequence of the 1,379-bp DNA fragment located upstream from fins. The nucleotide sequences of both strands of the
XbaI-BamHI DNA fragment located upstream fromfms (see Fig. 2) were determined by the dideoxy chain termination method (30). The DNA
sequence of one strand and the translated amino acid sequences deduced from smf and smg are shown. Putative RBSs are underlined.
Restriction sites used for cloning into M13mpl8 or M13mpl9 also are indicated.
smg, starts 164 bp downstream from the stop codon of smf idea, the BamHI-NcoI fragment of plasmid pBS936, which
and also is preceded by an RBS. It does not end within the contains the four ORFs, was cloned between the EcoRI and
cloned fragment. A comparison of the amino acid sequence BamHI sites of plasmid pRS415 to yield pBN15 (Table 1), in
deduced from smf or of the incomplete one deduced from which the transcription of the lacZ gene is under the control
smg with those of other proteins in the NBRF data bank of any promoter possibly occurring on the cloned fragment.
(release 31) did not reveal any significant resemblance. Six derivatives of plasmid pBN15 with deletions of various
Transcription of thefins-fint operon. The occurrence of the lengths originating from the EcoRI site were constructed
smf and smg cistrons, in the opposite polarity compared with thereafter. Four EcoRI-EcoRI deletions were obtained be-
fims andflnt, suggested that the transcription start site of the tween the EcoRI site of pBN15 (corresponding to the BamHI
putative fis-flnt dicistronic operon could be located in the site of smg; Fig. 2) and four EcoRI sites (E1, E2, E3, and E4)
intergenic region between fins and smf. For testing of this created in the inserted DNA fragment. The corresponding996 MEINNEL ET AL. J. kRIOL.
BACTE
deletion plasmids were called pE1, pE2, pE3, and pE4 (Fig.
2). The two other deletions were constructed by taking
. 1oo0 advantage of the single SmaI and XbaI sites on pBN15 to
C.) yield plasmids pAS and pAX, respectively (Fig. 2).
u 80.
The P-galactosidase activities expressed in lacZ strain
JM101Tr containing each of the above-described plasmids
X0
o1)
6 were then measured. As shown in Fig. 2, the 0-galactosidase
0) 60.
0
activity measured with pBN15 dropped by a factor of 7 upon
deletion of the EcoRI-XbaI fragment (plasmid pAX),
whereas it remained unmodified with the EcoRI-E1 deletion
p 20.
(plasmid pE1). Moreover, the 20-fold reduction in P-galacto-
sidase activity in JMlOlTr(pE2), compared with that in
a JMlOlTr(pBN15), was nearly identical to that obtained with
Q. 0 0.5 1 1.5 2 2.5 any of the three plasmids, pE3, pAS, and pE4, that carry
Deletion size (kb) larger deletions than pE2. These results demonstrated that
Downloaded from http://jb.asm.org/ on March 30, 2021 by guest
the inserted BamHI-NcoI DNA fragment contained one or
several promoters located between E1 and XbaI, upstream
CQ Glli
U from fins (Fig. 2), and responsible for the transcription of
F.K-- EI -fmt'-* both fins and fint.
I
For precise mapping of the transcription start site of the
I fins-fint operon, primer extension analysis of total RNA
from strain JM1O1Tr(pBS936) was undertaken by use of
reverse transcriptase (Fig. 3). Four bases, clustered within a
nrRNl .
10-base region, were identified as transcription start sites of
the fins-fint operon. In agreement with the results of this
pEl mapping experiment, sequences matching the consensus
pAX sequence for the E. coli transcription promoters were found
pE2 upstream from the above-described transcription initiation
pE3 region. The corresponding -10 and -35 sequences are
pAS boxed in Fig. 3. It is interesting to note that the -35 box is
pE4 located upstream from the XbaI site, whereas the -10 one
overlaps this restriction site. This localization is in agree-
FIG. 2. fint is cotranscribed with fins. Various deletions in ment with the observation that, in the above-described
plasmid pBN15, carrying the fint::lacZ operon fusion (Table 1), experiment, lacZ transcription was more strongly stimulated
were constructed (bottom). Four oligonucleotides, O1 (5'-CGCAG
GAATTCAATGATGGTT), 02 (5'-GTGAATGCAGGAATTCAGC when plasmid pAX rather than plasmid pE2 was used.
GT), 03 (5'-AACAACAACGAATTCGTCAGA), and 04 (5'-GTAC Finally, the promoter of the smf gene was mapped in a
ACCTGAATTCGCAGCGCGTC), complementary to the indicated similar way (data not shown). Transcription of this gene
positions (E1, E2, E3, and E4, respectively) in the fins-fint region starts at a single A located downstream from the transcrip-
(middle), were synthesized, as was another oligonucleotide, 05 tion start sites of thefins-fint operon (Fig. 3). Consequently,
(5'-TTCTTGTGGACGCAGGGAAACA), corresponding to a se- the mRNAs for smf and fins-fint overlap.
quence located downstream from the NcoI restriction site within the Influence of the GUG translation initiation start site on the
fint gene. 05 and either 01, 02, 03, or 04 were used to amplify by the
PCR the corresponding regions on plasmid pBS936. The polymerase expression offint. The effect on the translation of fint of the
chain reaction products were restricted with the following combina- occurrence of the unusual GUG translation initiation start
tions of restriction enzymes: EcoRI and XbaI (01-0° amplification), site was evaluated by use of the gene fusion technique.
EcoRI and SmaI (02-05 and O3-05 amplifications), and EcoRI and Advantage was taken of the NcoI restriction site to fuse in
NcoI (04-05 amplification). Each resulting DNA fragment was frame the fint gene to the lacZ coding sequence in plasmid
inserted between the corresponding sites in plasmid pBN15, yielding pRS414 to yield pBN14 (Table 1). The region upstream from
pEl, pE2, pE3, and pE4. The bottom part of the figure represents the theflnt::lacZ in-frame fusion in plasmid pBN14 extends up
portion of the BamHI-NcoI region remaining in the indicated to the BamHI site within the smg ORF. The expression of
plasmids. Plasmids pAS and pAX were obtained after restriction of
pBN15 with EcoRI-SmaI (pAS) or EcoRI-XbaI (pAX), filling in with 3-galactosidase from plasmid pBN14 depends therefore on
the Klenow fragment of DNA polymerase I, and recircularization. both the transcription and the translation signals offlnt. The
Note that the EcoRI site on plasmid pBN15 is located immediately fint::lacZ GUG translation initiation codon in pBN14 was
upstream from BamHI. The P-galactosidase activities in JM1O1Tr changed to an AUG one by site-directed mutagenesis (31),
cells harboring each of the studied plasmids were determined as yielding pBN14AUG. The wild-type and mutated flnt: :lacZ
described by Miller (24). Bacteria were grown in MOPS minimal fusions were then transferred into phage XRS45 by homolo-
medium supplemented with Casamino Acids and ampicillin. Sam- gous recombination in vivo to yield X14 and X14AUG,
ples of the culture were withdrawn during exponential growth. respectively. Each X phage was used to lysogenize lacZ
,-Galactosidase activity measurements were systematically ob- strain JM1O1Tr. The 3-galactosidase activities produced
tained from two samples and corrected for the average activity
found in JM101Tr(pE4) (328 + 40 U). The P-galactosidase activities from the corresponding monolysogens (strains JMX14 and
shown at the top of the figure are expressed as percentages of the JMX14AUG) were measured. From the results shown in
activity measured for JM101Tr(pE1) (9,304 200 U). Data are Table 2, it appeared that changing the GUG translation
plotted as a function of the sizes of the various deletions created in initiation codon into an AUG one resulted in a 2.2-fold
the studied plasmids. increase in the expression of the flnt::lacZ fusion. Such a
moderate effect is similar to those previously measured for
either the lacZ gene or the cya gene from E. coli, in whichVOL. 175, 1993
I TCGA
t
:\5'~ ~T
EXPRESSION OF THE E. COLI FORMYLASE GENE
3'T A 5'
T A
A
T
C
T
T
COG
T A
T A
TA
A T
T A
AT
G C
T
A
G
A
A
997
Downloaded from http://jb.asm.org/ on March 30, 2021 by guest
AT
S - -A A
T 3
--~ A T
GC
AG
TA
~~~5
~~~N 1T A 3'
m m y
TTTCTGTATCGACCATCCTTATCTCCCTGCCATAAGCAGCCTTAGCAATCTTTGCGATTGGTCAGT
GluThrAspValMet
XbaI NoL
GA¶E ~~TCAGAGGGGGATTTGTCEAGAAGA AATAATCTTTCTAACTCCTGAACACA
- 35 - 10
TCTCIGGATTTATGTCAGTTTTGCAAGTGTTACATATTCCGGACGAGCGGCTTCGCAAAGTTGCT
Glu~Jr
sp~alMetSerValLeuGlnValLeuHislleProAspGluArgLeuArgLysValAla
FIG. 3. Fine mapping of the fis-flnt promoter. The 21-mer oligonucleotide 5'-CACTTCTTCTACCGGTTAGC-3' was 5' labeled with
[_y-32P]ATP and used in primer extension analysis with total RNA extracted from strain JM1O1Tr(pBS936) or from strain JM1O1Tr as a control.
The labeled extension product (lane 1) was analyzed on a 6% sequencing gel in parallel with a DNA sequencing reaction performed with the
same oligonucleotide and double-stranded pBS936 plasmid DNA as a template (lanes T, C, G, and A). The autoradiogram is shown in the
upper left part of the figure. The corresponding DNA sequence is shown in the upper right part, together with arrows indicating the mapped
transcription start sites. Mapping of the smf promoter was performed in an analogous manner with 01 as an extension primer (data not
shown). The lower part of the figure indicates the locations of the various transcription start sites (vertical arrows) in the smf-fins intergenic
region and the directions of transcription (horizontal arrows). The indicated ORFs are those of smf (top) andfins (bottom). The RBS of fens
is underlined. Putative -10 and -35 consensus sequences for the transcription of fs-fint are boxed. Bases identical to the consensus
sequence proposed by Rosenberg and Court (28) are boldfaced.
initiation codons were changed into GUG or AUG and ribosomes and tRNAs (reviewed in reference 14) and also
GUG, respectively (16, 27). has been demonstrated to occur for several aminoacyl-tRNA
Control of the expression of thefint gene. Because of the synthetases as well as initiation factors IF1, IF2, and IF3
participation of FMT in protein synthesis, its cellular con- (reviewed in reference 7; see also references 6 and 26).
centration may be governed by metabolic regulation, i.e., a Variations in growth rate were obtained by use of various
general regulatory mechanism in E. coli, yet not fully under- culture media containing different carbon sources. Cells
stood at the molecular level, which couples the concentra- were harvested at the exponential growth stage, and crude
tions of components of the biosynthetic machinery with cell extracts were prepared for enzymatic activity measure-
growth rate. Such a coupling has been firmly established for ments. Surprisingly, the FMT specific activity remained998 MEINNEL ET AL. J. BACTERIOL.
TABLE 2. Influence of the fint initiation codon (GUG or AUG) 350
and of the cellular concentration of FMT or of tRNA et on ON
the expression of various fint::lacZ fusions in vivo 300
Bacterial Plasmidb 1-Galactosidase Overexpressed 250
straina activity macromolecule c)8 i
'Up
200
JMX14 27 ± 3 VI
JMX14AUG 60 ± 3 150
JMX14 pACYC184 25 ± 4
JMX14 pACform 28 ± 4 FMT 100
JMX14 pBSTNAV 26 ± 3
JMX14 pBStRNAfMet 24 ± 3 tRNAT~et 50
JMX14
PAL14Tr
pBStRNAm!et 22
25
±
±
3
3
tRNA.et
0
PAL14Tr pACYC184 28 + 3 0.25 0.50 0.75 1.00 2.00
PAL14Tr pACform 23 ± 4 FMT Frequency of cell division (hl)
PAL14Tr pBStRNAr~et 18 ±4 tRNAret
Downloaded from http://jb.asm.org/ on March 30, 2021 by guest
PAL14Tr
JMX15
pBStRNA:~et 23
226
+3
+ 10
tRNA:et FIG. 4. Cellular level of FMT as a function of E. coli growth rate.
K37 cells were grown at 370C in MOPS minimal medium (25)
PAL15Tr 230 ± 15 containing 0.4% glucose, 0.4% glycerol, or 0.4% acetate as tile
PAL1STr pACYC184 237 ± 25 carbon source and supplemented with either methionine (0.2 mM) or
PAL1STr pACform 200 ± 20 FMT Casamino Acids (0.2%). LB medium was also used to obtain the
PAL1STr pBStRNAmet 217 ± 25 tRNAfMet highest growth rate. One unit of FMT (or valyl-tRNA synthetase)
PAL15Tr pBStRNAe 241 + 20 tRNAMet activity was defined as the amount of enzyme capable of formylating
(or aminoacylating) 1 pmol of Met-tRNAfMet (or tRNAval) per s in
a
Monolysogenic derivatives of the X14 or X14AUG (fmt::lacZ in-frame the standard assay. Protein concentrations in the crude extracts
fusions) or X15 (fint::lacZ operon fusions) prophages integrated in strain were determined by the technique of Bradford (2). The FMT
JMlOMTr were isolated, yielding strains JMX14, JMX14AUG, and JMX15, as
were X14 or X15 lysogens of PAL13TrXFatg (expressing a low level of FMT concentrations in the crude extracts, given in units per milligram of
activity), yielding strains PAL14Tr and PAL15Tr (Table 1). ,B-Galactosidase total protein, are plotted as a function of the number of doubling
activities from several independent lysogens were shown to vary as multiples events per hour measured for exponentially growing cultures. Sym-
of the lowest measured value that was considered to come from a monolyso- bols: E, FMT specific activity; valyl-tRNA synthetase specific
[],
gen. activity.
b Derivatives of the various strains containing the indicated plasmids were
constructed. Depending on the carried plasmid, cells were grown in the
presence of ampicillin (pBStRNAj~et and pBStRNA:et) or chloramphenicol
(pACYC184 and pACform). 28%. In such strains, with a low FMT level, P-galactosidase
activity remained unmodified, compared with that in control
strains JMX14 and JMX15 (Table 2).
insensitive to the variations in growth rate (Fig. 4). In the Transformation of either PAL14Tr or PAL15Tr by an
same experiment, the valyl-tRNA synthetase specific activ- FMT-overproducing plasmid, pACform, resulted in a 5,000-
ity, which is subject to metabolic regulation through tran- fold increase in the, FMT content. As shown in Table 2, the
scriptional activation, was also assayed as a control. 13-galactosidase activities in strains PAL14Tr(pACform) and
The occurrence of a constant level of FMT in the cell PAL15Tr(pACform) were unmodified, compared with those
regardless of growth parameters may reflect the occurrence in PAL14Tr and PAL15Tr, respectively.
of a retroactive control of the expression of fint, in- The above-described experiments showed that the expres-
volving either FMT itself or formyl-Met-tRNAret, the prod- sion of the fint::lacZ protein and operon fusions was insen-
uct of the reaction catalyzed by FMT. For the purpose sitive to variations in the intracellular FMT concentration in
a 5,000-fold range (Table 2). Taken together, these results
of measuring whether the cellular concentration of FMT
or initiator tRNAret could influence the expression of strongly indicate that the expression of the lint gene is
fint::lacZ fusions, the monolysogenic strain JMX14, express-
ing an fint::lacZ in-frame fusion, was transformed with
various plasmids: pBStRNAret, pBStRNA!et, pBSTNAV, TABLE 3. Effect of the intracellular FMT concentration on the
pACYC184, and pACform (Table 2). Neither 50-fold over- extent of N acylation of tRNAfet
expression of FMT activity nor 5-fold overexpression of
tRNA vet modified the ,-galactosidase activity expressed Growth Total FMT Met-tRNAP'et
from the fusion, compared with that from the controls, Bacterial straina rate Ar concn N acylation
(min)b concn (%)d (%)e
pBStRNA:et and pBSTNAV (Table 2). Similar results were
obtained with a JM1O1Tr derivative carrying a single copy of
X15, which bears an fnt::lacZ operon fusion, i.e., for which JM1OMTr 38100 ± 8 100 5 81 5
0-galactosidase transcription is under the control of the fint JM101Tr(pBStRNAret) 35496 ± 10 93 ± 5 80 + 5
PAL13TrXFatg 155 113 ± 8 1.0 0.5 58 + 3
promoter (Table 2).
Next, the effect of a reduced intracellular FMT concen- a Cells were grown in LB medium. FMT specific activity, tRNAPlet
tration on the expression offlnt::lacZ fusions was measured. concentration, and Met-tRNAfMPt N acylation were determined as described
in reference 10.
For this purpose, strain PAL13TrAFatg, in which the intra- b Doubling times of the cultures.
cellular FMT concentration is reduced 100-fold (10), was c The concentration of tRNAvaI also was measured to standardize the
lysogenized with a single copy of either X14 (PAL14Tr) or tRNAflct content of these tRNA preparations. A value of 100% corresponded
to 0.48 pmol of tRNAfMet per pmol of tRNAvaI.
X15 (PAL15Tr). As shown in Table 3, in PAL13TrAFatg d A value of 100%
corresponded to 51 U/mg.
cells, because of the limiting intracellular FMT concentra- e The extent of Met-tRNAfMet N acylation is given as a percentage of the
tion, the extent of Met-tRNA; et N acylation was lowered by corresponding measured total tRNAf et concentration.VOL. 175, 1993 EXPRESSION OF THE E. COLI FORMYLASE GENE 999
insensitive to variations in either intracellular FMT or intra- 4. Butler, J. S., M. Springer, and M. Grunberg-Manago. 1987.
cellular tRNApMet concentrations or to the extent of N AUU-to-AUG mutation in the initiator codon of the translation
acylation of Met-tRNAret initiator factor IF3 abolishes translational autocontrol of its own
Conclusions. The present study shows first that fint is gene (infC) in vivo. Proc. Natl. Acad. Sci. USA 84:4022-4025.
cotranscribed with fins, a gene encoding a putative protein 5. Chang, A. C. Y., and S. N. Cohen. 1978. Construction and
characterization of amplifiable multicopy DNA cloning vehi-
sharing sequence identities with zinc metallopeptidases (10). cles derived from the P1SA cryptic miniplasmid. J. Bacteriol.
Substitution of an AUG for the GUG translation start codon 134:1141-1156.
of flint only slightly increases the expression of the fint 6. Cummings, H. S., J. F. Sands, P. C. Foreman, J. Fraser, and
product. J. W. B. Hershey. 1991. Structure and expression of the infA
Second, the intracellular expression of FMT is not sensi- operon encoding translation initiation factor IF1: transcriptional
tive to metabolic control. To our knowledge, this fact makes control by the growth rate. J. Biol. Chem. 266:16491-16498.
FMT the first example of a component of the cellular 7. Grunberg-Manago, M. 1987. Regulation of the expression of
machinery of protein biosynthesis that escapes control by aminoacyl-tRNA synthetases and translation factors, p. 1386-
cell growth rate. 1409. In F. C. Neidhardt, J. L. Ingraham, K. B. Low, B.
Magasanik, M. Schaechter, and H. E. Umbarger (ed.), Esch-
Finally, using the gene fusion technique, we failed to show erichia coli and Salmonella typhimurium: cellular and molecular
control of the expression of Jint mediated by either the biology. American Society for Microbiology, Washington, D.C.
Downloaded from http://jb.asm.org/ on March 30, 2021 by guest
extent of N acyjation of Met-tRNAret or the intracellular 8. Gualerzi, C. L., and C. 0. Pon. 1990. Initiation of mRNA
FMT or tRNAfMet concentration. Although one cannot ex- translation in prokaryotes. Biochemistry 29:5881-5889.
clude the possibility that an effector not tested in the present 9. Guillon, J. M. 1992. Approche enzymologique et gen6tique du
study has an effect on FMT expression, the expression of the r6le de la formylation N-terminale dans le d6marrage de la
fint gene appears to be constitutive. traduction chez les bacteries. Ph.D. thesis, Ecole Polytech-
The data in Table 3 show that, with a 100-fold decrease in nique, Palaiseau, France.
cellular FMT activity, the extent of N acylation of Met- 10. Guillon, J. M., Y. Mechulam, J. M. Schmitter, S. Blanquet, and
G. Fayat. 1992. Disruption of the gene for methionyl-tRNAPet
tRN et is reduced only by 30%. Moreover, when the formyltransferase severely impairs the growth of Escherichia
tRNAf et concentration is increased fivefold in wild-type E. coli. J. Bacteriol. 174:4294-4301.
coli cells, the proportion of formyl-Met-tRNAp~et to total 11. Guillon, J. M., T. Meinnel, Y. Mechulam, C. Lazennec, S.
tRNA;~et remains unchanged (Table 3). These data indicate Blanquet, and G. Fayat. 1992. Nucleotides of tRNA governing
that the number of FMT molecules normally present in an E. the specificity of Escherichia coli methionyl-tRNAMetf formyl-
coli cell is in large excess with respect to the number transferase. J. Mol. Biol. 224:359-367.
required to ensure full N formylation of Met-tRNA et. 12. Hirel, P. H., F. Leveque, P. Mellot, F. Dardel, M. Panvert, Y.
Notably, an extract obtained from 3 x 107 bacteria grown in Mechulam, and G. Fayat. 1988. Genetic engineering of methio-
LB medium allowed the N formylation of 0.18 pmol of nyl-tRNA synthetase: in vitro regeneration of an active syn-
thetase by proteolytic cleavage of a methionyl-tRNA syn-
Met-tRNAp~et per s in the standard assay. If one assumes a
turnover number of 20 s-1 in the reaction catalyzed by FMT thetase-p-galactosidase chimeric protein. Biochimie (Paris) 70:
773-782.
(15), the number of FMT molecules in one bacterial cell can 13. Jakubowski, H., and E. Goldman. 1984. Quantities of individual
be estimated to be on the order of 102. This number is aminoacyl-tRNA families and their turnover in Escherichia coli.
significantly smaller than those of the other proteins in- J. Bacteriol. 158:769-776.
volved in the initiation of translation, including methionyl- 14. Jinks-Robertson, S., and M. Nomura. 1987. Ribosomes and
tRNA synthetase (2103 molecules per cell [13]) or valyl- tRNA, p. 1358-1385. In F. C. Neidhardt, J. L. Ingraham, K. B.
tRNA synthetase (as deduced from the data in Fig. 4, with Low, B. Magasanik, M. Schaechter, and H. E. Umbarger (ed.),
the assumption of a turnover number of 1 so1). The synthesis Escherichia coli and Salmonella typhimurium: cellular and
in E. coli of the non-rate-limiting number of 102 FMT molecular biology. American Society for Microbiology, Wash-
ington, D.C.
molecules could therefore be at a negligible energetic cost, 15. Kahn, D., M. Fromant, G. Fayat, P. Dessen, and S. Blanquet.
compared with the energy expended for the synthesis of the 1980. Methionyl-transfer-RNA transformylase from Escherichia
other translational factors. This idea may explain why the coli. Purification and characterisation. Eur. J. Biochem. 105:
cell has no advantage in regulating fint expression, while 489-497.
tight control of the expression of other proteins of the 16. Khudyakov, Y. E., V. S. Neplyueva, T. I. Kalinina, and V. D.
translational apparatus appears beneficial (reviewed in ref- Smirnov. 1988. Effect of structure of the initiator codon on
erence 7). translation in E. coli. FEBS Lett. 232:369-371.
17. Kung, H. F., B. Eskin, B. Redfield, and H. Weissbach. 1979.
ACKNOWLEDGMENTS DNA-directed in vitro synthesis of 3-galactosidase: requirement
for formylation of methionyl-tRNAMetf. Arch. Biochem. Bio-
We thank P. Plateau for helpful advice and critical reading of the phys. 195:396-400.
manuscript. C. Lazennec is acknowledged for skillful technical help. 18. Lee, C. P., B. L. Seong, and U. L. RajBhandary. 1991. Structural
This work was supported by grants from the Fondation pour la and sequence elements important for recognition of Eschenichia
Recherche Medicale. coli formyl-methionine tRNA by methionyl-tRNA transformy-
lase are clustered in the acceptor stem. J. Biol. Chem. 266:
REFERENCES 18012-18017.
1. Blanquet, S., P. Dessen, and D. Kahn. 1984. Properties and 19. Levque, F., M. Gazeau, M. Fromant, S. Blanquet, and P.
specificity of methionyl-tRNAMetf formyltransferase from Esch- Plateau. 1991. Control of Escherichia coli lysyl-tRNA syn-
erichia coli. Methods Enzymol. 106:141-153. thetase expression by anaerobiosis. J. Bacteriol. 173:7903-7910.
2. Bradford, M. M. 1976. A rapid and sensitive method for the 20. Meinnel, T., Y. Mechulam, and G. Fayat. 1988. Fast purification
quantitation of microgram quantities of protein utilizing the of a functional elongation tRNAMet expressed from a synthetic
principle of protein-dye binding. Anal. Biochem. 72:248-254. gene in vivo. Nucleic Acids Res. 16:8095-8096.
3. Brombach, M., and C. L. Pon. 1987. The unusual translational 21. Meinnel, T., Y. Mechulam, D. Le Corre, M. Panvert, S. Blan-
initiation codon AUU limits the expression of the infC (initiation quet, and G. Fayat. 1991. Selection of suppressor methionyl-
factor IF3) gene of Escherichia coli. Mol. Gen. Genet. 208:94- tRNA synthetase: mapping the anticodon binding site. Proc.
100. Natl. Acad. Sci. USA 88:291-295.1000 MEINNEL ET AL. J. BACTERIOL.
22. Meinnel, T., E. Schmitt, Y. Mechulam, and S. Blanquet. 1992. 28. Rosenberg, M., and D. Court. 1979. Regulatory sequences
Structural and biochemical characterization of the Escherichia involved in the promotion and termination of RNA transcrip-
coli argE gene product. J. Bacteriol. 174:2323-2331. tion. Annu. Rev. Genet. 13:319-353.
23. Miller, H. I., and D. I. Friedman. 1980. An E. coli gene product 29. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular
required for X site-specific recombination. Cell 20:711-719. cloning: a laboratory manual, 2nd ed. Cold Spring Harbor
24. Miller, J. H. 1972. Experiments in molecular genetics. Cold Laboratory, Cold Spring Harbor, N.Y.
Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 30. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequenc-
25. Neidhardt, F. C., P. L. Bloch, and D. F. Smith. 1974. Culture ing with chain-terminating inhibitors. Proc. Natl. Acad. Sci.
medium for enterobacteria. J. Bacteriol. 119:736-747. USA 74:5463-5467.
26. Pedersen, S., P. L. Bloch, S. Reeh, and F. C. Neidhardt. 1978. 31. Sayers, J. R., W. Schmidt, and F. Eckstein. 1988. 5'-*3' exonu-
Patterns of protein synthesis in E. coli: a catalog of the amount cleases in phosphorothioate-based oligonucleotide-directed mu-
of 140 individual proteins at different growth rates. Cell 14:179- tagenesis. Nucleic Acids Res. 16:791-802.
190. 32. Schmitter, J. M., Y. Mechulam, G. Fayat, and M. Anselme.
27. Reddy, P., A. Peterkovsky, and K. McKenney. 1985. Transla- 1986. Rapid purification of DNA fragments by high-performance
tional efficiency of the Escherichia coli adenylate cyclase gene: size-exclusion chromatography. J. Chromatogr. 378:462-466.
mutating the UUG initiation codon to GUG or AUG results in 33. Simons, R W., F. Houman, and N. Kleckner. 1987. Improved
increased gene expression. Proc. Natl. Acad. Sci. USA 82: single and multicopy lac-based cloning vectors for protein and
5656-5660. operon fusions. Gene 53:85-96.
Downloaded from http://jb.asm.org/ on March 30, 2021 by guestYou can also read
