ABSTRACT
The codon UGA located 5
'
adjacent to an mRNA hairpin within
fdhF
mRNA promotes the incorporation of the amino acid selenocysteine into formate dehydrogenase H of
Escherichia coli
. The loop region of this mRNA hairpin has been shown to bind to the special
elongation factor SELB, which also forms a complex with selenocysteinyl-tRNA
Sec
and GTP. We designed seven different mRNA constructs derived from the
fdhF
mRNA which contain a translation initiation region including an AUG initiation
codon followed by no, one, two, three, four, five or six UUC phenylalanine codon(s) and the UGA selenocysteine codon 5
'
adjacent to the
fdhF
mRNA hairpin. By binding these different mRNA constructs to 30S ribosomal
subunits
in vitro
we attempted to mimic intermediate steps of elongation of a structured mRNA approaching the ribosome by one codon at a time. Toeprint analysis of the mRNA-ribosome complexes showed that the presence of the
fdhF
mRNA hairpin strongly interferes with binding of the
fdhF
mRNA to 30S ribosomal subunits as soon as the hairpin is placed closer than 16
bases to the ribosomal P-site. Binding is reduced up to 25-fold compared with mRNA constructs where the hairpin is located outside the ribosomal mRNA track. Surprisingly, no toeprint signals were observed in any of our mRNA constructs
when tRNA
Sec
was used instead of tRNA
fMet
. Lack of binding of selenocysteinyl-tRNA
Sec
to the UGA codon was attributed to steric hindrance by the
fdhF
mRNA hairpin. By chemical probing of the shortest mRNA construct (AUG-UGA-fdhF hairpin) bound to 30S ribosomal subunits we demonstrate that
the hairpin structure is not unfolded in the presence of ribosomes
in vitro
;
also, this mRNA is not translated
in vivo
when fused in-frame 5
'
of the
lacZ
gene. Therefore, our data indicate that
the
fdhF
mRNA hairpin has to be unfolded during elongation prior to entering the
ribosomal mRNA track and we propose that the SELB binding domain within the
fdhF
mRNA is located outside the ribosomal mRNA track during decoding of the UGA
selenocysteine codon by the SELB-selenocysteinyl-tRNA
Sec
-GTP complex.
There is ample evidence that specific mRNA secondary structures are required for so-called recoding events during translation of certain mRNAs (
1
,
2
). However, very little is known about the direct interaction of these mRNA
secondary structures with the ribosome (
3
-
5
). Several questions as to when and where these mRNA structures interact with
the ribosome remain unanswered so far and it is still unclear when and by which
mechanism(s) these mRNA structures become unfolded during translation (
6
). While the interaction of structured mRNAs with the ribosome during translation initiation in
Escherichia coli
has been intensively studied (for reviews see
7
-
10
) their interaction with the ribosome during elongation is obscured by the dynamic process of this event.
In this respect, our study represents an initial approach to elucidate the
interaction of an mRNA secondary structure with the ribosome during elongation. We chose the RNA hairpin within
fdhF
mRNA, which promotes the incorporation of selenocysteine into formate
dehydrogenase H in
E.coli
. This hairpin, located 3' adjacent to the UGA selenocysteine codon (
11
,
12
), is a prerequisite for selenocysteine incorporation (
13
). Besides its presence, the incorporation pathway requires a specific tRNA (selenocysteinyl-tRNA
Sec
) containing a UCA anticodon complementary to the UGA codon as well as a specific elongation factor, designated SELB. Elongation factor SELB, exhibiting extensive sequence similarity to EF-Tu (
14
), was shown to bind to selenocysteinyl-tRNA
Sec
(
14
) as well as to the loop region of the mRNA hairpin (
12
,
15
). It is assumed that by this mechanism selenocysteinyl-tRNA
Sec
is tethered to the UGA selenocysteine codon.
We performed toeprinting assays of mRNA constructs containing one, two, three, four, five or six codons between an AUG initiation codon
(preceded by a Shine-Dalgarno sequence) and the UGA selenocysteine codon adjacent to the mRNA
hairpin. By using these constructs we attempted to mimic intermediate steps of
elongation of the
fdhF
mRNA on the ribosome by moving the hairpin closer to the decoding site by one
codon at a time. The objective of this study was to gain a first insight into
the interaction of the
fdhF
mRNA stem-loop structure with the translation apparatus.
All reagents were obtained from Sigma (Germany) unless indicated otherwise. T7 polymerase was a generous gift from Thomas Maier (München), AMV reverse transcriptase was purchased from Appligene (France)
and kethoxal was supplied by Upjohn (UK).
tRNA
Phe
and tRNA
fMet
were obtained from Sigma (Germany) and tRNA
Sec
was a generous gift from Christian Baron (München).
mRNAs were transcribed from DNA templates containing a T7 promotor (
16
) followed by a translational initiation region including an AUG initiation
codon and the UGA selenocysteine codon 5' adjacent to the
fdhF
hairpin. Spacing between the AUG and UGA codon was designed containing no (AH75
[UUC]
0
), one (AH78 [UUC]
1
), two (AH81 [UUC]
2
), three (AH84 [UUC]
3
), four (AH87 [UUC]
4
), five (AH90 [UUC]
5
) or six (AH93 [UCC]
6
) UUC phenylalanine codons. DNA templates were generated by PCR amplification of
plasmid DNA pAF1 (
12
) essentially as described by Saiki
et al
. (
17
), using the following primers.
5
'
-Primers
AH75, 5'-GGC ACA TGT TAA TAC GAC TCA CTA TAG GGC TAA ATT TTG GAG GCA TTA
ATG
TGA
CAC GGC CCA TCG GTT GCA GGT-3';
AH78, 5'-GGC ACA TGT TAA TAC GAC TCA CTA TAG GGC TAA ATT TTG GAG GCA TTA
ATG
TTC
TGA
CAC GGC CCA TCG GTT GCA GGT-3';
AH81, 5'-GGC ACA TGT TAA TAC GAC TCA CTA TAG GGC TAA ATT TTG GAG GCA TTA
ATG
TTC TTC
TGA
CAC GGC CCA TCG GTT GCA GGT-3';
AH84, 5'-GGC ACA TGT TAA TAC GAC TCA CTA TAG GGC TAA ATT TTG GAG GCA TTA
ATG
TTC TTC TTC
TGA
CAC GGC CCA TCG GTT GCA GGT-3';
AH87, 5'-GGC ACA TGT TAA TAC GAC TCA CTA TAG GGC TAA ATT TTG GAG GCA TTA
ATG
TTC TTC TTC TTC
TGA
CAC GGC CCA TCG GTT GCA GGT-3';
AH90, 5'-GGC ACA TGT TAA TAC GAC TCA CTA TAG GGC TAA ATT TTG GAG GCA TTA
ATG
TTC TTC TTC TTC TTC
TGA
CAC GGC CCA TCG GTT GCA GGT-3';
AH93, 5'-GGC ACA TGT TAA TAC GAC TCA CTA TAG GGC TAA ATT TTG GAG GCA TTA
ATG
TTC TTC TTC TTC TTC TTC
TGA
CAC GGC CCA TCG GTT GCA GGT-3'.
ATG and TGA sequences are indicated in bold. 5'-Primers contained an
Afl
III site, the T7 promotor sequence, the translation initiation region including
the AUG codon as well as no (AH75), one (AH78), two (AH81), three (AH84), four (AH87), five (AH90) or six (AH93) TTC phenylalanine codons followed by a region complementary to
the
fdhF
mRNA hairpin (positions 418-441); numbering according to Zinoni
et al
. (
18
)
3
'
-Primer
AM2/T, 5'-GGC GGA TCC TCG GTA TTA TCA ATT TCG TTA ATA GC-3'.
The 3'-primer contained a
Bam
HI site followed by a sequence complementary to a region downstream of the
fdhF
hairpin (positions 478-503); numbering according to Zinoni
et al
. (
18
). Purification of DNA oligonucleotides and T7 transcription were performed as
described by Hüttenhofer and Noller (
5
).
Escherichia coli
600 MRE 0.5 M salt washed 70S ribosomes were prepared as described by Moazed and Noller (
19
) and 30S subunits were obtained as described by Moazed
et al
. (
20
). 30S subunits were activated by heating in reaction buffer A (10 mM MgCl
2
, 140 mM NH
4
Cl, 80 mM potassium cacodylate, pH 7.2) at 42oC for 20 min before being used for mRNA and tRNA binding (
21
).
Binding of mRNAs to 30S ribosomal subunits was performed by incubating 1 pmol
mRNA, 7 (or 14) pmol 30S ribosomal subunits and 40 pmol tRNA in the absence or
presence of 45 pmol SELB in reaction buffer B (10 mM Tris-acetate, pH 7.4, 60 mM NH
4
Cl, 10 mM Mg-acetate, 0.5 mM GTP, 2 mM DTT). The
32
P-end- labeled AM2/T primer (see above), which is complementary to the 3'-end of the various RNA transcripts, was annealed as
described in Hartz
et al
. (
22
). Toeprinting assays were performed according to Ringquist
et al
. (
23
). Quantification of toeprints was performed on a Molecular Dynamics Personal Densitometer. Binding affinity of mRNAs was assessed as the intensity of the +16 toeprint signal divided by the sum of the full-length signal plus the +16 toeprint (%).
Chemical probing of mRNA-ribosome complexes with KE was performed as described by Hüttenhofer and Noller (
5
). Primer extension reactions for analysis of modified bases were performed as described by Stern
et al
. (
24
), using 5'-
32
P-end-labeled primer AM2/T (see above). Samples were loaded onto 6% (w/v)
polyacrylamide-7 M urea gels. Electrophoresis was performed at 2000 V, 22 mA for 1.5 h.
A translational fusion was constructed which consisted of a translational
initiation codon directly fused to a UGA codon followed by the
fdhF
hairpin and the
lacZ
gene. To this end, we amplified a 140 bp PCR product from plasmid pWT which contains a translational fusion of a selenium insertion cartridge into the
lacZ
gene (
11
). Amplification was achieved with a 5'-primer containing the desired mutations and a 3'-primer complementary to a region of the
lacZ
gene (positions 6291-6306).
5'-primer, 5'-GGA AGC TTA AGG AGG AAA TTA TTA TGT GAC ACG GCC CAT GC-3';
3'-primer, 5'-GTA AAA CGA CGG CCA GT-3'.
The amplification product was cloned into plasmid pSKS106 utilising
Hin
dIII and
Bam
HI sites introduced at the insert borders (
25
). The
Hin
dIII site was filled in with the Klenow enzyme. The resulting plasmid contained
the
lac
promotor followed by an artificial reading frame of seven codons that was
terminated by a UAG codon and the AUG-UGA-
fdhF
hairpin-
lacZ
gene fusion preceded by an optimised Shine-Dalgarno sequence. Selenocysteine insertion into fusion proteins was assessed by measuring [beta]-galactosidase activity obtained with this construct and
plasmid pWT (
11
).
Escherichia coli
strains FM434 and FM464 (
13
) were transformed with the plasmids and analysed for synthesis of [beta]-galactosidase as previously described (
13
).
We performed toeprinting studies of ribosome-mRNA complexes using mRNA constructs containing no, one, two, three, four or five UUC phenylalanine codon(s) between the UGA selenocysteine codon adjacent to the
fdhF
hairpin and the AUG initiation codon (AH75 [UUC]
0
-AH90 [UUC]
5
, Figs
1
and
2
). The mRNA constructs were bound to 30S ribosomal subunits in the presence of
equal amounts of tRNA
fMet
(see Materials and Methods) and toeprints assays were performed in the presence
or absence of special elongation factor SELB. Elongation factor SELB has been
shown previously to bind to the loop region of the
fdhF
hairpin (
12
,
15
). In the absence of SELB the intensities of the +16 toeprint signals strongly decrease when the
fdhF
hairpin approaches the ribosomal decoding site; this is achieved by reducing the distance between
the AUG and UGA codon gradually by one codon at a time. Reduction in
intensities of toeprint signals indicates a reduced binding of these mRNA
constructs to the 30S ribosomal subunits. Thereby, binding was assessed as the
ratio of the +16 toeprint with respect to the full-length cDNA (see Materials and Methods). Densitometric evaluation of
toeprint signals showed that 45% of AH90 [UUC]
5
mRNA bound to 30S ribosomal subunits and an up to 25-fold reduction in binding of the remaining mRNA constructs (Figs
2
and
3
). About 47% of an mRNA construct containing a spacer region of six codons, AH93
[UUC]
6
, was bound to 30S ribosomal subunits, comparable with binding of the AH90 [UUC]
5
mRNA (Fig.
1
and data not shown).
Next, we wanted to determine whether the extended toeprint signal was solely due
to the presence of tRNA
fMet
positioning the mRNA with respect to the ribosome. To enhance the intensity of
the toeprint signals a 14-fold excess of ribosomes over mRNAs was used, instead of the 7-fold excess used for the previous experiment (see Materials and
Methods). We incubated the AH78 [UUC]
1
mRNA (the AUG-UUC-UGA construct) with 30S ribosomal subunits in the presence of tRNA
fMet
, tRNA
Phe
or tRNA
Sec
and compared the resulting toeprint signals to those obtained with the AH75
[UUC]
0
mRNA (the AUG-UGA construct) in the presence of tRNA
fMet
and tRNA
Sec
. Figure
4
shows binding of the AUG-UUC-UGA mRNA to 30S ribosomal subunits in the presence of tRNA
fMet
, which results in a toeprint at position +16, however, no extended toeprint is
visible. Binding of tRNA
Phe
to the UUC codon results in a toeprint shifted by three bases, as expected. In
addition, an extended toeprint is visible at position +71/72 (Fig.
4
), which is also observed within the AH75 [UUC]
0
mRNA (the AUG-UGA construct) when tRNA
fMet
is bound to the ribosome; note again the presence of two toeprint signals at
positions +16 and +18 due to binding of tRNA
fMet
to AUG or GUG. Surprisingly, with tRNA
Sec
as an initiator tRNA neither the +16 nor extended toeprint is visible when the
AH78 [UUC]
1
or AH75 [UUC]
0
mRNAs are bound to the ribosome, despite the fact that the UGA selenocysteine codon is located within a proper toeprint distance from the Shine-Dalgarno sequence.
Figure
At position 72, in the presence of any tRNA (fMet, Phe or Sec) and with both
constructs used, there is a weak background toeprint signal observed in some of
our experiments; the position of the signal is not shifted by three bases, as
is the case for the +16 toeprints in the presence of the different tRNAs. We
therefore attribute this signal, as well as the extended toeprint signal, to a
conformational change within the mRNA upon binding to 30S ribosomal subunits
(see Discussion).
One of the difficulties in interpreting toeprint signals is their lack of
information on the structural changes an mRNA might undergo upon binding to the
ribosome. We therefore performed a footprint analysis of the AH75 [UUC]
0
mRNA-ribosome complex to investigate whether the mRNA secondary structure
becomes unfolded upon binding to 30S ribosomal subunits. As a chemical probe to
investigate the accessiblity of G bases, KE was used. Modified bases were
analysed by primer extension analysis (see Materials and Methods). Unfolding of
the stem-loop structure by the ribosome can be monitored by the increased
accessibility of G bases in the stem structure of the hairpin. By chemical
probing these G bases have been shown not to be accessible (G39, G40 and G43)
or weakly accessible (G44) to chemical modification by KE in the free mRNA (
12
; Fig.
5
, lane 1). Binding of the mRNA was performed in the presence or absence of
special elongation factor SELB.
Figure
Quantitative binding of the AH75 [UUC]
0
mRNA is demonstrated by the complete protection of G bases of the Shine-Dalgarno sequence from chemical modification in the presence of 30S ribosomal
subunits (Fig.
5
, lane 2). However, no increased accessibilty of G bases in the stem structure
of the RNA hairpin can be observed, indicating that no unfolding of the mRNA
hairpin occurs (Fig.
5
). In the presence of SELB (Fig.
5
, lane 3), G+26 in the loop of the mRNA hairpin becomes protected from
modification by KE due to interaction with SELB, as shown previously (
12
). Surprisingly, despite the presence of tRNA
fMet
, G+3 of the AUG initiation codon is not protected from chemical modification.
Instead, a slight increase in the reactivity of G+3 and G+5 towards KE is
observed in the presence of SELB. A moderate increase in reactivitiy towards KE
is also visible for base G+52 (Fig.
5
, lanes 2 and 3; note the different numbering of bases compared with Fig.
1
).
To investigate
in vivo
expression of the AH75 [UUC]
0
mRNA (the AUG-UGA mRNA construct), we cloned the
fdhF
hairpin sequence, including the Shine-Dalgarno sequence and ATG initiation codon, in-frame 5' of the
lacZ
gene (see Materials and Methods) and measured [beta]-galactosidase synthesis of the resulting plasmid in bacterial
strains FM434 and FM464 (
13
); spacing between the Shine-Dalgarno sequence and the ATG start codon was optimized to be 7 instead
of 5 nt, to avoid out-of-frame decoding by tRNA
fMet
at the GUG sequence (see Discussion). Expression of [beta]-galactosidase was compared with a construct, pATG-(NNN)
7
-TGA, where a spacer region of seven codons was introduced between the ATG
and TGA codons (
11
). As can be seen in Table
1
, reduction of the distance between the ATG and TGA codons from seven codons to
none resulted in a dramatic decrease in [beta]-galactosidase synthesis. Expression of the AUG-UGA construct in strain FM464 (
13
), lacking selenocysteine-tRNA (by deletion of the tRNA
Sec
, encoding
selC
gene), resulted in the same decrease in [beta]-galactosidase synthesis (Table
1
). This indicates that UGA read-through and [beta]-galactosidase expression is reduced to background levels in the AUG-UGA mRNA construct.
Table 1
In this study we show that the presence of the
fdhF
mRNA hairpin, promoting selenocysteine incorporation into formate dehydrogenase
H in
E.coli
, results in a strongly reduced binding of
fdhF
mRNA to the ribosome when the hairpin is placed within the ribosomal mRNA
binding track. By toeprint or footprint analysis the `entry site' of the
ribosomal mRNA track has been shown to be located between +16 and +19 bases
away from the first base of the P-site codon (
5
,
26
). An up to 25-fold reduction in binding is observed of those mRNA constructs which place
the mRNA hairpin closer than ~16 bases to the ribosomal P-site (Fig.
6
). Thereby, the AH84 [UUC]
3
and AH81 [UUC]
2
mRNA constructs resulted in the strongest decrease in binding to 30S subunits,
while, in comparison, binding of AH75 [UUC]
0
and AH78 [UUC]
1
mRNAs was slightly higher. A possible explanation would be that, assuming an A-helical conformation of the
fdhF
mRNA within the ribosomal mRNA track, the
fdhF
hairpin might inhibit binding to the ribosome differently dependent on which
side of the mRNA helix the hairpin is located.
Figure
Within the AH75 [UUC]
0
mRNA-ribosome complex a second toeprint signal at position +18 in addition to
the one at position +16 is observed. The two toeprint signals can be
rationalised by postulating that tRNA
fMet
binds to the AH75 [UUC]
0
mRNA within two different reading frames; the +16 toeprint being due to
decoding of tRNA
fMet
at
Figure
To test for translation
in vivo
, a DNA fragment resembling AH75 [UUC]
0
mRNA was fused in-frame 5' of the
lacZ
gene lacking its translation initiation region.
In vivo
expression of the resulting plasmid, pATG-TGA, showed a dramatic decrease in [beta]-galactosidase synthesis. However, introducing a spacer region
of seven codons between the AUG and UGA codon restored read-through of the UGA codon to wild-type levels. This is consistent with the
fdhF
hairpin preventing translation of the mRNA when placed adjacent to the AUG
initiation codon.
In the presence of tRNA
Sec
we could not observe any ribosome-dependent toeprint signals in our mRNA constructs, indicative of the mRNA
hairpin interfering with decoding of the UGA codon by tRNA
Sec
. This could be due to either steric hindrance by the
fdhF
hairpin structure or the fact that the UGA codon might be embedded within a
secondary structure and thereby unable to base pair with the anticodon of tRNA
Sec
. Chemical probing data demonstrate (Fig.
5
and data not shown) that the selenocysteine codon is accessible to base-specific probes. Therefore, steric hindrance is the most likely
explanation for the absence of any toeprint signals in the presence of tRNA
Sec
. In fact, deletion of the hairpin structure 3' of the UGA codon within AH75 [UUC]
0
mRNA results in a +16 toeprint signal in the presence of tRNA
Sec
(
23
).
Since toeprinting studies only show a low resolution picture of how mRNAs
interact with the ribosome, we tried to elucidate the mechanism of interaction
of the AH75
[UUC]
0
mRNA with the ribosome by footprint analysis with chemical probes. By this
approach we show that the AH75
[UUC]
0
mRNA is not unfolded by 30S ribosomal subunits (Fig.
5
). Despite the presence of tRNA
fMet
, we did not observe protection of the G base of the AUG initiation codon, as
was shown previously for gene 32 mRNA (
5
). Although we have no direct evidence for binding of tRNA
fMet
to 30S subunits, lack of interaction with the AUG codon might be indicative of the presence of a `pre-ternary complex', where the tRNA is bound to the ribosome but not engaged in a codon-anticodon interaction (
28
-
30
). Lack of codon-anticodon interaction at the P-site codon preceeding the UGA selenocysteine codon might be
indicative of the
fdhF
mRNA hairpin interfering with decoding. Therefore, this might be consistent
with a model where partial unfolding of the hairpin structure has to occur
prior to the UGA codon entering the ribosomal A-site.
Our results are at variance with a model suggested by Ringquist
et al
. (
23
), which implies that the ribosomal mRNA binding track is able to accomodate
even large mRNA structures. This model is based on the presence of an
`extended' toeprint signal 36 bases downstream of the +16 signal within an AUG-UGA-fdhF hairpin mRNA. We also observed this extended toeprint, but only
when we used the shortest construct (AH75 [UUC]
0
mRNA), containing the stem-loop structure immediately next to the AUG initiation codon. However,
within all other mRNA constructs (e.g. AH78 [UUC]
1
-AH90 [UUC]
5
mRNAs), which extended the spacing between the initiation codon and the stem-loop structure by one codon each, did not result in an extended toeprint
(Figs
2
and
6
). One possible explanation for the presence of the extended toeprint within AH75
[UUC]
0
mRNA only is a conformational change within the mRNA upon binding to the
ribosome. This conformational change could stall the reverse transcriptase at the extended toeprint position.
In fact, an increase in the reactivity of base G+52, located adjacent to the
extended toeprint signal, towards KE was observed upon binding of the mRNA to
30S ribosomal subunits, indicative of some structural rearangement occuring
within AH75 [UUC]
0
mRNA (Fig.
5
).
In conclusion, our study shows that: (i) the
fdhF
mRNA secondary structure interferes with binding of the mRNA to the ribosome
in vitro
and prevents translation
in vivo
when placed within the ribosomal mRNA binding track; (ii) an extended toeprint signal, indicative of large structured mRNAs being accomodated by the ribosomal mRNA track, is missing in all but one mRNA construct with tRNA
fMet
as the initiator tRNA; (iii) although bases of the UGA codon are accessible to chemical probes, decoding by tRNA
Sec
is sterically hindered due to the presence of the mRNA stem-loop structure.
Taken together these data indicate that the
fdhF
hairpin has to be unfolded during elongation prior to entering the ribosomal
mRNA track
in vivo
, as the mRNA hairpin interferes with binding of the mRNA to the ribosome, as
well as with decoding of selenocysteinyl-tRNA
Sec
at the UGA codon. However, since defined initiation complexes were used to mimic distinct steps of elongation, it still
has to be demonstrated that the mode of interaction of an mRNA with the
ribosome is similar during the initiation and elongation phases. In addition,
we omitted any initiation or elongation factors in our toeprinting assays, which might affect binding of mRNAs to the ribosome (
30
). In that respect, our study can only be a first approach in trying to
elucidate the interaction of the structured
fdhF
mRNA with the ribosome.
The consequences of our model would imply that the
fdhF
hairpin is partially unfolded during decoding of the UGA selenocysteine codon by selenocysteinyl-tRNA
Sec
, as the UGA codon is adjacent to the stem of the
fdhF
hairpin (Fig.
1
). Would this interfere with binding of special elongation factor SELB to the
fdhF
hairpin during decoding? Special elongation factor SELB was shown to be
structurally divided into an N-terminal domain, which shares extensive sequence homology with EF-Tu and binds to selenocysteinyl-tRNA
Sec
and GTP, and a separate C-terminal domain, required for binding exclusively to only the upper half of the
mRNA hairpin, including the loop region (
23
,
31
). Therefore, we propose a model where the upper part of the
fdhF
mRNA hairpin which binds to the C-terminus of SELB can be placed `outside' the ribosomal mRNA track during
decoding of the UGA codon (Fig.
7
A). Upon decoding of the UGA codon by selenocysteinyl-tRNA
Sec
, SELB might dissociate from the mRNA hairpin, whereupon the mRNA secondary
structure could be unfolded by the elongating ribosome. It will be interesting
to investigate whether other mRNA secondary structures, like for example
pseudoknots, are also unfolded by the ribosome prior to entering the ribosomal
mRNA track.
We thank B.Ehresmann and T.Maier for critical reading of the manuscript and
M.Famulok for providing DNA oligonucleotide AH93. This work was supported by
the Deutsche Forschungsgemeinschaft (grant Bo 305/20-1) and the Fonds der Deutschen Chemie.


Strain
Plasmid
Miller units
FM434
pATG-TGA
6
FM464 ([Delta]selC)
pATG-TGA
8
FM434
pATG-(NNN)
7
-TGA
1450


REFERENCES
Return
