Nucleic Acids Research, Vol. 19, No. 23
6573-6578
Use of single-stranded DNA oligonucleotides in
programming ribosomes for translation
Robert D.Ricker and Akira Kaji*
University of Pennsylvania, School of Medicine, Department of Microbiology, Philadelphia,
PA 19104-6076, USA
Received July 17, 1991; Revised and Accepted November 5, 1991
ABSTRACT
Single-stranded DNA (ssDNA) oligomers were
compared to synthetic RNA oligomers in their ability
to program E. coli ribosomes in vitro. AUG and dATGcontaining oligomers promoted the non-enzymatic
binding of fmet-tRNA to ribosomes, with similar
dependence on time and magnesium concentration;
only at 10 mM Mg + + or at low oligomer concentration
was RNA slightly preferred in complex formation.
These initiation complexes were biologically active in
that fmet-tRNA, bound in response to ssDNA or RNA,
was fully reactive with puromycin. While dAUG could
not function as an initiation codon, p-dAUG functioned
as well as AUG or dATG. However, dUAA and p-dUAA
could not replace UAA in directing release-factor (RF)
activity, and dTAA functioned only to a slight extent.
Release factors had specificity for termination
complexes containing dATGTAA, dATGTAG, or
dATGTGA. At M g + + concentrations of 15 mM or
higher, these hexamers directed peptidyl transferasedependent fmet-tRNA hydrolysis in the absence of RF.
We suggest this RF-independent activation of peptidyl
transferase as a unique system for studying the
mechanism of termination. Overall, these results
indicate that ssDNA can be used in place of RNA for
certain studies of protein synthesis.
INTRODUCTION
In vitro studies of protein synthesis have been carried out
traditionally by adding RNA to a cell extract, or purified
ribosomes and individual components. The RNA used has varied
from full-length phage mRNA (1 - 3 ) to synthetic RNA oligomers
(4-9). However, in several studies, DNA has been used in place
of RNA. As early as 1963, it was shown that ribosomes could
associate with single- stranded <£X174 or T2 DNA to form
'polysomes' (10). Later, it was shown that certain preparations
of single-stranded DNA could serve directly as a template for
protein synthesis (11), and that any denatured DNA was an
effective template in the presence of neomycin (12). The sequence
of the polypeptide products of this translation were consistent
1
To whom correspondence should be addressed
with the genetic code and led to an idea about the degree of
accuracy and codon context effects (13). The characteristics of
this translation are very similar to that for mRNA. In fact, the
translation of poly(dTG) at 5 mM M g + + was shown to require
fmet-tRNA and initiation factors (13). Although neomycin was
necessary for effective translation, the binding of fmet-tRNA and
DNA to the ribosome, and even limited elongation, occured in
the absence of antibiotic (13,14). More recently, initiation factordependent complexes of ribosomes, template, and fmet-tRNA
were made using a synthetic ssDNA oligomer analogous to a
ribosome binding site (15). Using a similar sequence, but without
the dATG codon, it was shown that neither ssDNA nor the
corresponding RNA was dependent upon initiation factors for
its binding to 30S ribosomal subunits (16). Using extension
inhibition analysis (1), 30S subunits in the presence of tRNAfjnet,
bound at the true initiator ATG of a synthetic mini-gene (17).
In addition, the complex blocked reverse transcriptase at the same
position as that found for mRNA— +16 (referring to the A of
ATG as +1) (1)(17). Finally, ssDNA corresponding to tRNA
has been synthesized and aminoacylated (18,19).
The recent development of better chemistries and automated
solid-phase synthesis has significantly increased the availability
of RNA oligonucleotides; but, at best, the methods are expensive
and tedious. In contrast, ssDNA synthesis has developed to a
point where it is routine. This availability of ssDNA has made
it important to better characterize its ability to program ribosomes
for protein synthesis. In this work, we have made comparisons
between RNA and ssDNA in their ability to program ribosomes
for initiation and termination of translation. One of the unexpected
features of ssDNA termination was observed with hexamers
containing an initiation and termination codon. These hexamers
did not require release factor to direct a moderate amount of fmet
release at Mg + + concentrations of 15 mM and above.
MATERIALS AND METHODS
Preparation of Components for Complex Formation
Ribosomes were isolated from E. coli MRE600 and were washed
4 times with 1M NH4C1, essentially as described previously
(20). The f[35S]met-tRNA was synthesized by aminoacylation of
6574 Nucleic Acids Research, Vol. 19, No. 23
tRNA f met (Boehringer-Mannheim Biochemicals) with
[35S]-methionine (Amersham). The reaction mixture was then
fractionated by mixed-mode HPLC as described previously (4).
The final f[35S]met-tRNA was concentrated and desalted by
repeated ultrafiltration and dilution in a Centricon-10 unit from
Amicon.
Nucleotides
AUGUAA was a kind gift from Dr. Ikehara and Dr. Ohtsuka,
Osaka University, Japan. AUG and UAA were obtained from
Boehringer Mannheim Biochemicals. UAG and UGA were
purchased from Miles Pharmaceuticals. The single-stranded DNA
oligomers were synthesized and purified by the DNA Synthesis
Service, University of Pennsylvania. Oligomers containing
deoxy-U or a 5'-phosphate, were obtained from Synthecell Corp.
(Rockville, MD). dATGTAA was 5'-end labeled using
[Y-32P]ATP, essentially as described by Maniatis (21).
Preparation of Initiation Complexes
Reactions for the formation of ribosome complexes with
f[35S]met-tRNA and various oligonucleotides were carried out
in 20 pi of standard complex-formation buffer containing 100
mM Tris-HCl (pH 7.4), 67 mM NH4C1, 5 mM 0mercaptoethanol, and 10 mM MgAc2 (unless otherwise
specified). Ribosomes (80 pmol) and the appropriate oligomer
(250 pmol or as indicated) were pre-incubated at 30 c C for 5 min.
HPLC-purified f[35S]met-tRNA (1 or 10 pmol) was then added
and the incubation continued for 30 min. Ribosome-bound
f[35S]met-tRNA was determined by counting radioactivity
retained on nitrocellulose filters (9) after washing with iTce-cold
complex-formation buffer.
were prepared essentially as described (22). Fraction IE was used
for mixtures of RF-1 and RF-2. Fractions IV and V were used
for RF-1 and RF-2, respectively.
RESULTS
Characteristics of RNA and ssDNA in the Non-enzymatic
Formation of Initiation-complex
ssDNA and RNA oligonucleotides were compared in their ability
to stimulate f[35S]met-tRNA binding to ribosomes over a wide
range of Mg + + concentrations. It can be seen in Figure 1 that
trimers (dATG and AUG, panel A) and AUGUAA (Fig. IB,
A) had similar binding curves under conditions of oligomer
excess, with a plateau being reached near 15 mM Mg + + .
dATGTAA (Fig IB, A) appears not to compare with the other
oligomers, but this is not the case (Fig IB, • ) and will be shown
(see section on unusual termination). Analysis of data obtained
around 10 mM Mg + + revealed a slight preference for RNA
over ssDNA in complex formation. This difference in
effectiveness was observed mainly with hexamers, and this fact
is demonstrated again by the time course shown in Figure 2B.
Figure 1 also shows that the non-specific binding of f[35S]mettRNA (binding in the absence of oligonucleotide) at 10 mM
Mg + + was nearly non-existent. At 15 mM Mg + + , nucleotidedirected binding was nearly maximal, while non-specific binding
Use of various antibiotics in assay of the complexes
Puromycin (Pm) reacts with fmet-tRNA bound to the A-site of
the ribosome to yield fmet-Pm. Pm was added after complex
formation to a final concentration of 1 mM, and the mixture was
incubated at room temperature for 10 min. Aliquots of 10 or 20
/JL\ were diluted to 150 /xl with the complex-formation buffer and
were extracted with 600 /il of ethyl acetate. The sample was
vortexed for 15 s and centrifuged in a microfuge for 30 s to
separate the phases. A 0.5 ml portion of the organic layer was
removed and counted after the addition of 4 ml Ecolite(+) (ICN,
Biomedicals). Chloramphenicol (Cm) and sparsomycin (Sp) were
used to inhibit peptidyl transferase. They were added during
complex formation to final concentrations of 10~6 and 10~4 M,
respectively.
Assay of Release-Factor Activity
Ribosome complexes for the assay of release factor were formed
as the initiation complexes above except that they were
immediately diluted 5-fold with ice-cold RF-assay buffer (50 mM
Tris-HCl pH 7.4, 50 mM NH^Ac, and 35 mM MgAc2) or
complex-formation buffer (as before, but 30 mM MgAc2)
immediately after the binding of fmet-tRNA. Ribosomes were
programmed with various RNA or ssDNA oligomers at the
M g + + concentrations indicated. Release factor-mediated activity
was quantitated by the addition of 1 /tl of assay material to 20
fd of the diluted complex in the presence or absence of 1 \>% UAA.
Incubation was carried out in 22 fil total volume at 24°C for
15 min. A 10 /il aliquot was assayed for released fI35S]met by
acidification with the addition of 150 ix\ of 0.1 N HC1 and ethyl
acetate extraction as described for Pm, above. Release factors
0
10
20
30
40
Mg + + (mM)
Figure 1. Effect of Mg + + concentration on complex formation between fmettRNA and ribosomes programmed with ssDNA or RNA oligomers. Complexes
of ribosomes, f[35S]met-tRNA, and RNA or ssDNA oligomers were formed as
described in Materials and Methods. The standard complex-formation buffer was
adjusted to yield the final Mg + + concentration indicated. The curves indicate
fmet-tRNA bound to ribosomes after complex formation for 30 min at 30°C.
Figure 1A shows trimers AUG, O—O; dATG. • — • : and a control without
oligonucleotide. *—*. Figure IB shows the hexamers AUGUAA. A—A;
dATGTAA, A — • ; and the combined total of fmet-tRNA bound to dATGTAAprogrammed ribosomes, and fmet released from this complex. • — • .
Nucleic Acids Research, Vol. 19, No. 23 6575
remained relatively low. In contrast, at 20 mM Mg + + , nonspecific binding rose to one-half that of the specific binding. Much
of the remaining data was gathered at 10 mM Mg + + in order
to compare ssDNA and RNA under conditions of high template
specificity for fmet-tRNA, and low background of fmet-tRNA
hydrolysis (see section on unusual termination).
The data in Figure 2 indicate the time course for complex
formation between fmet-tRNA, ribosomes, and ssDNA or RNA
oligomers. The experiment was carried out at 10 mM Mg + + ,
conditions at which there was little fmet-tRNA binding in the
absence of nucleotide. AUG, dATG, and AUGUAA were very
similar in both the rate and extent of complex formation (Fig. 2A,
O and • ; Fig. 2B, A) while dATGTAA was less effective at
this Mg + + concentration (Fig. 2B, A). At 12, 15, and 20 mM
Mg + + , respectively, the binding of f[35S]met-tRNA to
ribosomes occurred more quickly and with less dependence on
template (data not shown).
Figure 3 indicates the amount of complex formed versus the
amount of oligonucleotide added, under conditions of limiting
oligomer and 10 mM Mg + + . It is clear from this figure that,
AUG and AUGUAA were more effective than the corresponding
ssDNAs in directing fmet-tRNA binding to ribosomes. In
addition, the trimers, whether RNA or DNA, were more effective
than their respective hexamers.
Early studies on the ability of RNA oligomers to act as
codewords showed that oligomers with a 5'-phosphate were more
effective than those which had no phosphate; while, oligomers
containing a 3'-phosphate were inactive (9). The 5'-phosphate
was especially helpful when the oligomers were very short. The
ssDNA oligomers used here so far do not contain phosphate on
either the 3' or 5' end. In order to determine whether
5'-phosphorylation or replacement of dT by dU might make an
oligomer more effective for initiation, dAUG and p-dAUG were
used. As can be seen in Table 1, substitution of dU for dT nearly
eliminated the oligomer's ability to direct binding of fmet-tRNA
to ribosomes. However, the phosphorylated form of dAUG
directed complex formation to the same extent as either rAUG
or dATG. This is in agreement with the above study indicating
that 5'-phosphorylation increases recognition of an oligomer by
ribosomes (9). Our results indicating no oligo(dT)io-directed
binding of phe-tRNA to ribosomes (data not shown) are consistent
with earlier poly(rT) and poly(dT) experiments (13,23) indicating
their rigid 2 structure and inability to bind to ribosomes.
Unusual Termination Reaction Catalyzed by ssDNA
Figure IB shows that dATGTAA did not appear to stimulate
fmet-tRNA-ribosome complex formation to the same extent as
its RNA counterpart. In an effort to explain the low-efficiency
programming by this oligomer, we examined the possibility that
the termination reaction (assayed by release of fmet from fmettRNA bound to ribosomes) may be directed by dATGTAA in
the absence of RF. Since this hydrolysis is dependent upon fmettRNA binding to ribosomes, the combined value of fmet release
and final fmet-tRNA bound, represents total fmet-tRNA binding.
This value was measured and then plotted in Figure IB. It is clear
that the total amount of fmet-tRNA binding in response to
dATGTAA is very similar to the amount of binding in response
to AUGUAA.
In the experiment shown in Table 2, ribosomes were
programmed by various oligonucleotides in the presence or
Figure 3. fmet-tRNA binding to ribosomes programmed with various amounts
of ssDNA or RNA oligomers. Complexes were formed, at 10 mM M g + + for
30 min at 30°C, as described in Materials and Methods. f[35S]met-tRNA binding
was directed by RNA or ssDNA oligonucleotides at the indicated concentrations
(AUG, O - O ; dATG, • - • ; AUGUAA, A - A ; and dATGTAA, A - A ) .
Table 1. fmet-tRNA binding programmed by ssDNA or RNA oligomers
Figure 2. Time course of complex formation between fmet-tRNA and ribosomes
programmed with ssDNA or RNA oligomers. Complexes of ribosomes,
f35S]met-tRNA, and RNA or ssDNA oligomers were formed at 10 mM Mg + +
and 30°C, as described in Materials and Methods. AJiquots of the reaction mixture
were taken at the times indicated, and the amount of fmet-tRNA bound to ribosomes
is shown. Figure 2A shows fmet-tRNA bound in response to trimers AUG,
O—O; dATG, • — • ; and without oligonucleotide, *—*. Figure 2B indicates
fmet-tRNA bound in response to AUGUAA, A—A; and dATGTAA, A—A.
Oligomer
Used to
Program
Ribosomes
Ribosome
Bound
f[S35]met-tRNA
AUG
dATG
dAUG
p-dAUG
0.0161
0.0144
0.0003
0.0141
Complexes were formed, at 10 mM M g + + but were later diluted 5 fold in
termination assay buffer (to a final of 30 mM Mg + + ) as described in Materials
and Methods. f[ S]met-tRNA (1 pmol/20 pi reaction) binding was directed by
the various single-stranded oligomers listed in incubations at 37°C for 30 min.
For each reaction, the amount of ribosome-bound fmet-tRNA is shown. In each
case, the amount of fmet-tRNA bound in the absence of nucleotide was subtracted.
6576 Nucleic Acids Research, Vol. 19, No. 23
Table 2. Release of fmet from fmet-tRNA bound to ribosomes programmed with
ssDNA or RNA oligomers in the presence or absence of a peptidyl-transferase
inhibitor
Table 3. Stimulation of dATGTAA binding to ribosomes by fmet-tRNA
Addition of
f[35S]met-tRNA
fI35S]met-tRNA
Bound (cpm//d)*
"PdATGTAA
Bound (cpm//il)*
1958
3464
2000
Released
fmet
(pmol//d)
Ribosome
Bound
fmet-tRNA
(pmol/fil)
AUG
0.034
0.029
0.283
0.281
AUG +UAA
0.040
0.025
0.258
0.287
dATG
0.037
0.024
0.268
0.262
Table 4. RF-dependent release of fmet from fmet-tRNA ribosome complexes
programmed with ssDNA or RNA oligomers
+
0.035
0.027
0.242
0.281
Codon
Added
+
0.039
0.026
0.257
0.278
Oligomer
Used to
Program
Ribosomes
dATG
+
0.174
0.032
0.1O9
0.245
UAA
dTAA
dUAA
p-dUAA
UAA
dTAA
none
none
none
Codons
Used for
fmet-tRNA
Binding
dATG +TAA
AUGUAA
dATGTAA
Cm
Complexes were formed at 20 mM Mg + + , as described in Materials and
Methods. f[35S]met-tRNA (10 pmol/20 /tl reaction) binding was directed by
various nucleotides as indicated in the table. Each incubation was carried out
for 30 min at 37°C in the presence or absence of chloramphenicol (Cm).
absence of chloramphenicol (Cm). These complexes were formed
in the presence of 20 mM Mg + + and 10 pmol of f[35S]mettRNA per 20 fil reaction in order to more easily detect fmet
release. It has been shown that transfer of the carboxyl group
from peptidyl-tRNA to H 2 O, rather than to the NH2 group of
aminoacyl-tRNA, takes place in the presence of termination
codons; this reaction is apparently catalyzed by peptidyl
transferase (24), which is sensitive to inhibition by Cm and Sp
(25-28). It can be seen in Table 2 that, while the presence of
Cm did not significantly influence the amount of fmet-tRNA
bound in response to most of the oligonucleotides, binding
directed by dATGTAA was greatly increased. This increase was
to the same extent that fmet release was decreased by Cm. Similar
results were obtained using sparsomycin with dATGTAG and
dATGTGA (data not shown). These results are noteworthy since
they indicate RF-independent hydrolysis of fmet-tRNA by
peptidyl transferase.
•Crossover of isotopes into the opposite channel has been corrected and background
has been subtracted. dATGTAA was phosphorylated with 7[32P]ATP and
incubated with ribosomes in the presence or absence of f[35S]met-tRNA (1
pmol/20 y\ reaction) and 20 mM M g + + for 30 min at 37°C, as described in
Materials and Methods.
AUG
dATGTAA
AUGUAA
dATGTTT
For
Release
RF-stimulated
fI35S]met
Release
(pmol/ml)
74
13
3
oO
56
5
53
18
0
Complexes of fI35S]met-tRNA (10 pmol/20 ii\ reaction) and ribosomes
programmed with the indicated oligonucleotides, were formed at 30 mM Mg + +
for 30 min at 37 "C and used in the release factor assay as described in the Materials
and Methods. Numbers for RF-dependent release were calculated from the
radioactive material detected in the ethyl acetate-extractable portion of the reaction
mixture. Background values of f[35S]met release in the absence of termination
codon or absence of RF, were subtracted for each type of complex.
Presence of fmet-tRNA Stimulates Binding of dATGTAA to
Ribosomes
It has been shown that fmet-tRNA and mRNA work cooperatively
in their binding to the ribosome (3). As indicated in Table 3,
initiation complexes were formed with [32P]dATGTAA and
f(35S]met-tRNA in order to demonstrate direct binding of
ssDNA to ribosomes and to determine the effect of fmet-tRNA
on this binding. It is clear from the table that, as in the case of
RNA, the presence of fmet-tRNA in the initiation complex
stabilizes dATGTAA on the ribosome and results in increased
binding of this oligonucleotide.
a RF mixture, UAA, and fmet-tRNA complexes formed by AUG
or dATG. Very little termination activity was detected when
dTAA replaced UAA. Use of AUGUAA alone, also directed
low levels of activity, consistent with previous data (5,29). In
contrast, the amount of RF-dependent fmet release from
dATGTAA complexes was similar to that for AUG or dATG
complexes. It should be noted that this value is over and above
the RF-independent fmet release by dATGTAA. dATGTTT was
used as a control and did not stimulate fmet release. These results
indicate that RF can recognize ribosome complexes programmed
by either RNA or DNA initiation codons and can accurately
distinguish between sense and nonsense DNA codons. For the
same reasons as those described in the initiation studies, the
oligomers used for termination were synthesized with
5'-phosphates and with dU replacing dT, to examine if dUAA
and p-dUAA were effective in interacting with RF and the
ribosome. The data in Table 4 clearly indicate that neither of
these modifications can support significant RF activity. In
addition, complexes formed by these RNA or ssDNA oligomers
reacted similarly with Pm, releasing all of the f[35]S-met from
f[35S]met-tRNA bound with oligomer-dependence (data not
shown).
Comparison of RNA and ssDNA-Directed Initiation
complexes in their Interaction with RF and Pm
The biological activity of ssDNA-programmed complexes was
further studied through their interactions with release factor. As
shown in Table 4, similar termination activity was observed using
RF-1 and RF-2 recognize dTAG and dTGA, respectively
It has been shown that RF-1 and RF-2 specifically react with
UAG and UGA, respectively, while UAA is effective with both
release factors (30). The experiment shown in Table 5 was
designed to examine whether the DNA counterparts of UAG and
Nucleic Acids Research, Vol. 19, No. 23 6577
Table 5. Selective recognition of single-stranded DNA termination codons by
Release Factors 1 and 2
Oligomer
Used to
Program
Ribosomes
Codon
Added
for
Release
dATG
dATG
dATG
dATG
UAA
UAG
dATGTAG
dATGTGA
[35S]fmet Release by:
RF-1
RF-2
(pmol/ml)
36
60
37
7
19
2
0
12
Complex of ribosomes, dATG, and f[35S]met-tRNA (1 pmol/20/jl reaction) were
formed at 10 mM Mg + + and were later diluted 5 fold in termination assay buffer
(to a final of 30 mM Mg + + ) as described in the Materials and Methods. The
release factor assays were carried out using the dATG-programmed ribosomes
and by adding the termination codons indicated. Partially purified fractions
containing RF-1 (3.1 ^g) or RF-2 (0.15 fig) were then added and the reactions
carried out as described in the Materials and Methods.
UGA would retain this specificity towards RF-1 and RF-2. The
complexes were formed using dATG and 10 mM Mg + + .
dATGTAG and dATGTGA were added only during the
termination assay in order to reduce the RF-independent fmettRNA hydrolysis during formation of the initiation complex.
Upon addition of dATGTAG or dATGTGA for termination, RFindependent release also begins. The values of fmet release shown
in the table indicate enzymatic release over and above this nonenzymatic release. As seen in the table, RF-2 was specific for
UAA and dTGA (linked to dATG). In a similar fashion, RF-1
released fmet specifically from complexes with UAA, UAG, and
dTAG (linked to dATG). We conclude from these experiments
that RF-1 and 2 can recognize ssDNA termination codons in a
specific manner, one similar to that of the RNA counterparts.
In addition, these results show that the hexamers were effective
in stimulating the termination reaction even when they were not
used to form the initial complex.
DISCUSSION
Our data revealed the following similarities between ssDNA and
RNA oligonucleotides used to direct fmet-tRNA binding to
ribosomes: First, under conditions of oligomer excess, they are
very similar in their Mg + + dependence and rate of complex
formation. Also, the fmet-tRNA bound in response to either type
of oligomer was fully reactive to Pm and is, therefore, bound
functionally at the P-site. Finally, RF-1 and 2 can recognize
ssDNA hexamers containing dTAG or dTGA, respectively. In
contrast, the following differences were noted: First, at Mg + +
concentrations of 10 mM or less, RNA is preferred slightly over
ssDNA in the formation of initiation complexes. When oligomer
concentrations are limiting, RNA directs more fmet-tRNA
binding to ribosomes than does an equal amount of ssDNA. In
contrast to AUG, dAUG (containing the artificial nucleotide dU)
had almost no initiation activity, suggesting that it is not
recognized by ribosomes. Unlike UAA, dTAA codes for only
a small amount of RF-dependent fmet-tRNA hydrolysis but has
full activity when covalently bound to dATG. ssDNA hexamers
containing these covalently-linked initiation and termination
sequences stimulate peptidyl transferase-dependent fmet-tRNA
hydrolysis in the absence of RF, at Mg + + concentrations of
15 mM or more.
Mg + + , polyamines, and various other cations, have a major
effect upon the components of a translation system (31 -33). This
is the result of increased hydrophobic interaction due to the
neutralization of negative charges (particularly those in the
phosphate backbone of mRNA). Unlike polyamines, Mg + +
performs this role without spanning between the charges and does
not cause high-molecular-weight complexes. High Mg + +
concentration has, therefore, been used to allow the nonenzymatic binding of both mRNA and aa-tRNA to ribosomes,
and to stabilize ribosome/mRNA/aa-tRNA complexes. The
variation of M g + + concentration is a useful tool in controlling
translation systems (34-36), allowing a comparison of ssDNA
and RNA interaction with ribosomes and aa-tRNA. Like the RNA
system (9,35), complex formation with ssDNA oligomers
occurred very well at 20 mM Mg + + . Reduction of the Mg + +
concentration towards 10 mM, however, resulted in better
discrimination between RNA and ssDNA, and in reduction of
non-specific fmet-tRNA binding. This occurred even in the
absence of factors.
In the presence of 15 mM Mg + + or higher, ssDNA hexamers
which contained sequences analogous to termination codons (in
addition to dATG) appeared impaired in their ability to program
fmet-tRNA binding to ribosomes. However, in the presence of
peptidyl-transferase inhibitors, the amount of fmet-tRNA binding
directed by these hexamers was similar to that observed with all
of the other codons. This indicates that RF-independent hydrolysis
of ribosome-bound fmet-tRNA had taken place. This hydrolysis
was unexpected since the ribosomes were well washed and were,
therefore, free of release factors (shown by absence of activity
in the presence of UAA). This complements recent findings that
termination-codon recognition is an event which is independent
of RF binding (37) and involves a specific interaction of the
termination codon with sequences on the 16S rRNA (37—40).
It has been shown that oligomers containing dA5 are only stable
in the B conformation, which cannot bind to rU (41). As a result,
dTAA alone might associate only weakly with the proposed
termination-activating region of the 16S rRNA. While the B
conformation may result in low affinity, the large amount of
termination activity directed by dATGTAA indicates that this
conformation may be very similar to that required for natural
activation of peptidyl transferase. Covalent attachment of the
termination codon triplets to dATG directs the triplet to the active
region of the 16S rRNA. It is, therefore, possible that RF
stimulates peptidyl-transferase activity by stabilizing RNA
termination codons in a B-like conformation. Use of dTAA, dTAG, or dTGA (each covalently bound to dATG) to carry out
termination in the absence of RF, may present an ideal system
for studying the molecular mechanism of peptidyl-transferase
activation. Our data adds to evidence that ribosomes have an
intrinsic ability to carry out many functions of protein synthesis
in the absence of factors (42).
Further studies into the use of ssDNA for programming
ribosomes should include a careftil analysis of the elongation steps
of translation. While ssDNA may promote one or two steps of
elongation, extensive elongation has only been shown to occur
in the presence of neomycin (43). Ribosome-releasing factor acts
to separate ribosomes, mRNA, and deacylated tRNA at the
termination of translation (44,45). It would be of interest to
determine its activity upon complexes formed with ssDNA. It
should be kept in mind that these two steps, elongation and
ribosome release, are very sensitive to the affinity of the message
for the ribosome, and the altered structure of ssDNA might
prevent it from being functional in these steps.
6578 Nucleic Acids Research, Vol. 19, No. 23
REFERENCES
1. Gold,L., Hartz,D., McPheeters,D.S. and Traut.R. (1988) Methods Enzymol.,
164, 419-425.
2. Ryoji.M., Berland.R. and Kaji.A. (1981) Proc. Natl. Acad. Sci. USA, 78,
5973-5977.
3. Jay,G. and Kaempfer,R. (1974) Proc. Natl. Acad. Sci. USA, 71, 3199-3203.
4. Ricker,R.D. and Kaji,A. (1988) Anal. Bioch., 175, 327-333.
5. Ganoza.M.C, Buckingham,K., Hader.P. and Neilson,T. (1984) J Biol.
Chem., 259, 14101-14104.
6. Ganoza,C.M., Sullivan,P., Cunningham,C, Hader,P., Kofoid.E.C. and
Neilson.T. (1982) J Biol. Chem., 257, 8228-8232.
7. Thach,S.S. and Thach,R.E. (1971) Proc. Natl. Acad. Sci. USA, 68,
1791-1795.
8. Caskey.C.T., Scolnik,E., Caryk,T. and Nirenberg,M. (1968) Science, 162,
135-138.
9. Nirenberg,M. and Leder,P. (1964) Science, 145, 1399-1407.
10. Takanami,M. and Okamato.T. (1963) Biochem. Biophys. Res. Commun.,
13, 297-302.
11. Holland,J.J. and McCarthys J. (1964) Proc. Natl. Acad. Sci. USA, 52,
1554-1561.
12. McCarthy,B.J. and Holland,J.J. (1965) Proc. Natl. Acad. Sci. USA, 54,
880-886.
13. Morgan,A.R., Wells,R.D. and Khorana.H.G. (1967) J Mol. Biol., 26,
477-497.
14. Bretscher.M.S. (1969) J Mol. Biol., 42, 595-598.
15. Jay,E., Seth.A.K. and Jay.G. (1980) J Biol. Chem., 255, 3809-3812.
16. Calogero,R.A., Pon,C.L., Canonaco.M.A. and Gualerzi,C.O. (1988) Proc.
Natl. Acad. Sci. USA, 85, 6427-6431.
17. Shabarova.Z.A., Merenkova.I.N., Oretskaya,T.S., Sokolova.N.I.,
Skripkin,E.A., Alexeyeva.E.V., Balakin.A.G. and Bogdanov.A.A. (1991)
Nucleic Acids Res., 19, 4247-4251.
18. Khan.A.S. and Roe,B.A. (1988) Science, 241, 7 4 - 7 9 .
19. Khan.A.S. and Roe.B.A. (1990) FASEB. J., 4, A2002.
20. Kaji.A. (1968) Methods Enzymol. 12B, 692-699.
21. Maniatis.T., Fritsch,E.F. and Sambrook,J. (1982) Molecular Cloning: A
Laboratory Manual. Cold Spring Harbor University Press, Cold Spring
Harbor, New York.
22. Caskey,C.T., Scolnik.E., Tompkins,R., MUman.G. and Goldstein.J. (1971)
Methods Enzymol. 20, 367-379.
23. Szer,W. and Ochoa.S. (1964) J Mol. Biol., 8, 823-834.
24. Caskey.C.T., Beaudet.A.L., Scolnik,E. and Rosman,M. (1971) Proc. Natl.
Acad. Sci. USA, 68, 3163-3167.
25. Goldberg.I.H. (1979) In Hahn,F.E. (ed.), Antibiotics. Springer-Verlag, NewYork, Vol. I, pp. 264-271.
26. Tompkins.R.K., Scolnik,E.M. and Caskey,C.T. (1970) Proc. Nail. Acad.
Sci. USA, 65, 702-708.
27. Herner,A.E., Goldberg.I.H. and Cohen,L.B. (1969) Bioch., 8, 1335-1344.
28. Monro,R.E. and Vasquez.D. (1967) J Mol. Biol., 28, 161-165.
29. Buckingham.K., Chung.D.G., Neilson.T. and Ganoza,M.C. (1987) Biochim.
Biophys. Ada, 909, 9 2 - 9 8 .
30. Scolnik.E., Tompkins.R., Caskey.T. and Nirenberg,M. (1968) Proc. Natl.
Acad. Sci. USA, 61, 768-774.
31. Aikawa,J.K. (1981) Magnesium: Its Biologic Significance. CRC Press, Inc.,
Boca Raton, Fl.
32. Loftfield,R.B., Eigner.E.A. and Pastuszyn.A. (1981) In Morris,D.R. and
Marton,L.J. (eds.), Polyamines in Biology and Medicine. Marcel Dekker,
Inc., New York, pp. 208-221.
33. Bachrach.U. (1973) Function of Naturally Occurring Polyamines. Academic
Press, New York.
34. Snavely,M.D. (1990) In Sigel.H. and Sigel,A. (eds.), Metal Ions in Biological
Systems. Marcel Dekker, Inc., New York, Vol. 26, pp. 155-175.
35. Bartetzko.A. and Nierhaus,K.H. (1988) Methods Enzymol. 164,650-658.
36. Noll.M. and Noll.H. (1976) J Mol. Biol., 105, 111-127.
37. Hanfler.A., Kleuvers,B. and Goringer.H.U. (1990) Nucleic Acids Res., 18,
5625-5632.
38. Murgola,E.J., Hijazi.K.A., Goringer.H.U. and Dahlberg.A.E. (1988) Proc.
Natl. Acad. Sci. USA, 85, 4162-4165.
39. Murgola,E.J., Goringer.H.U., Dahlberg.A.E. and Hijazi.K.A. (1989) In
Cech,T.R. (ed.), Molecular Biology of RNA. Alan R. Liss, Inc., New York,
pp. 221-229.
40. Lang.A., Friemert,C. and Gassen,H.G. (1989) Ear. J. Biochem., 180,
547-554.
41. Martin,F.H. and Tinoco,I.Jr. (1980) Nucleic Acid Res., 8, 2295-2299.
42. Pestka.S. (1968) J Biol. Chem., 243. 2810-2820.
43. Bretscher,M.S. (1970) Cold Spring Harb. Symp. Quant. Biol., 34, 651 -653.
44. Ichikawa.S. and Kaji.A. (1989) J Biol. Chem., 264, 20054-20059.
45. Ogawa,K. and Kaji.A. (1975) Bioch. Biophys. Acta, 402, 288-296.