doi:10.1006/jmbi.2001.4632 available online at http://www.idealibrary.com on
J. Mol. Biol. (2001) 309, 727±735
Structural Studies of the tRNA Domain of tmRNA
Scott M. Stagg1, Ashley A. Frazer-Abel2, Paul J. Hagerman3
and Stephen C. Harvey1*
1
Department of Biochemistry
and Molecular Genetics
University of Alabama at
Birmingham, Birmingham
AL 35294, USA
2
Cancer Causation and
Prevention, AMC Cancer
Research Center, 1600 Pierce
St., Denver, CO 80214, USA
3
Department of Biological
Chemistry, University of
California, Davis, School of
Medicine, One Shields Ave.
Davis, CA 95616, USA
tmRNA is a small, stable prokaryotic RNA. It rescues ribosomes that
have become stalled during the translation of mRNA fragments lacking
stop codons, or during periods of tRNA scarcity. It derives its name from
the presence of two separate domains, one that functions as a tRNA, and
another that serves as an mRNA. We have carried out modeling and
transient electric birefringence studies to determine the angle between the
acceptor stem and anticodon stem of the tRNA domain of Eschericia coli
tmRNA. The results of the modeling studies yielded an interstem angle
of 110 , in agreement with the lower end of the range of angles (111 137 ) determined experimentally for various solution conditions. The
range of experimental angles is greater than the angles observed for any
of the tRNA crystal structures, in line with the presence of a shortened D
stem. The secondary structure of the tRNA domain is conserved for all
known tmRNA sequences, so we propose that the angle is also conserved. These results also suggest that the region of tmRNA between P2a
and P2b may interact with the decoding site of the ribosome.
# 2001 Academic Press
*Corresponding author
Keywords: tmRNA; ssrA RNA; transient electric birefringence; RNA
structure; molecular modeling
Introduction
When a ribosome translates an mRNA that lacks
a stop codon, or during periods of tRNA scarcity,
it becomes stalled, unable to release either the
mRNA or the nascent peptide. Prokaryotes have
evolved a molecule, tmRNA, that functions to
recycle these ribosomes (Karzai et al., 2000).
tmRNA has a tRNA-like domain that is aminoacylated with alanine before the molecule recognizes
the stalled ribosome and enters the empty A-site.
Peptidyl transfer attaches the growing peptide
chain to tmRNA, which then moves to the P-site
by translocation. The molecule also has a short
mRNA domain that then enters the A-site and is
translated as a normal message. The mRNA
domain ends in a stop codon, thus allowing the
ribosome to release the nascent peptide and
mRNA, freeing the ribosome for recycling. Interestingly, the mRNA domain encodes an amino acid
sequence recognized by proteases, so the resulting
protein contains a C-terminal sequence that targets
Abbreviations used: TEB, transient electric
birefringence.
E-mail address of the corresponding author:
[email protected]
0022-2836/01/030727±9 $35.00/0
it for degradation. This prevents accumulation of
incomplete translation products.
To perform these complicated functions, tmRNA
must make speci®c and unique interactions with
the ribosome. During normal translation, the selection of the correct tRNA depends on Watson-Crick
base-pairing between the tRNA and the mRNA at
the ribosomal decoding site. With a damaged
message, this mechanism cannot be used for selection of tmRNA during its initial binding to the
ribosome. After binding, tmRNA must promote
transpeptidation, and then translocation. Since it is
much larger than tRNA (363 nucleotides for E. coli
tmRNA, versus 76 for a typical tRNA), it must use
a novel set of interactions to facilitate these transitions. Finally, tmRNA must move its mRNA
domain into the proper position in the decoding
site, and its tRNA domain must be ejected from
the ribosomal P-site, so that the normal translation
mechanisms can take over to ®nish synthesis of the
protein's carboxy terminus. All of these steps
require a unique set of interactions between
tmRNA, the ribosome, and translational cofactors.
Structural studies on tmRNA, in isolation and later
in situ, are needed to elucidate structure-function
relationships in this important molecule.
No data have yet been reported for the threedimensional structure of tmRNA, but the second# 2001 Academic Press
728
ary structure of E. coli tmRNA has been well established through phylogenetic analyses (Williams,
1999; Williams & Bartel, 1996; Zwieb et al., 1999)
and chemical probing (Felden et al., 1997;
Hickerson et al., 1998). These reveal a highly folded
structure of 363 nucleotides with a tRNA domain,
an mRNA domain, and four conserved pseudoknots (Figure 1). The secondary structure of the
tRNA domain is similar to that of canonical
tRNAs, with a normal acceptor stem, hairpin
elements that resemble the D and T stems, and an
extended anticodon stem. There is nothing that
resembles an anticodon loop, although there are
two interesting internal loops in the extended
anticodon stem (Figure 1). tmRNA also has modi®ed bases in positions analogous to those of canonical tRNAs (Felden et al., 1998). Functionally, the
tRNA domain resembles tRNA, since it has a 30 CCA terminus that is aminoacylated (Komine et al.,
1994), and since it is known to interact with
elongation factor Tu in vitro (Rudinger-Thirion
et al., 1999). The only major difference between the
tRNA domain and normal tRNAs is the shortened
D stem, with two base-pairs rather than the canonical four base-pairs.
tRNAs with shortened D-stems are common in
mitochondria, and these have unusual tertiary
structures. These truncated tRNAs have a greater
angle between their acceptor and anticodon arms
than do tRNAs with normal D stems. It was ®rst
argued on theoretical grounds that the interarm
angle would have to be increased in order to maintain the distance from the end of the acceptor stem
to the anticodon loop, since the ``primary axis
Structural Studies of the tRNA Domain of tmRNA
length'' is necessary for the tRNAs to interact simultaneously with the ribosomal transpeptidation
and decoding sites (Steinberg & Cedergren, 1994;
Steinberg et al., 1994). This proposal was later con®rmed experimentally (Frazer-Abel & Hagerman,
1999), and it provided the basis for a broader set of
rules about structural compensation in tRNAs with
unusual secondary structures (Steinberg et al.,
1997). If the angle between the acceptor stem and
anticodon stem were known for the tRNA domain
of tmRNA, one might use the conserved primary
axis length to estimate the position on the extended
tmRNA anticodon stem that corresponds to the
location of the anticodon, thereby identifying the
region of tmRNA that is likely to interact with the
ribosomal decoding site.
Angles between double helices in RNAs can be
quanti®ed using the method of transient electric
birefringence (TEB) (Friederich et al., 1995; Vacano
& Hagerman, 1997; Zacharias & Hagerman, 1995).
In the TEB experiment, a transient electric ®eld is
used to align the molecules, yielding solution birefringence. When the ®eld is turned off, the birefringence decay re¯ects the free rotational diffusion of
the molecules. One compares the rotational diffusion of a test molecule containing a putative bend
with that of a fully base-paired control. The angle
between the adjacent helices of the bent molecule
can be determined from the ratio of the terminal
decay times of these molecules. This technique has
been used to measure the interarm angle in
tRNAPhe, and the result, 89(4) , compares favorably with the crystallographic value of 82 (Friederich et al., 1995). The angles have been
Figure 1. Secondary structure of
E. coli tmRNA featuring the tRNA
domain. Helices are in red, residues
8 and 9 are in yellow, the D loop is
in blue, residues 333-335 are in
orange, and the T loop is in green.
The mRNA domain is shown in
purple.
Structural Studies of the tRNA Domain of tmRNA
measured for various mitochondrial tRNAs
(Frazer-Abel & Hagerman, 1999; Leehey et al.,
1995).
Here, we report both TEB measurements and
molecular modeling studies on tmRNA, with the
goal of determining the angle between the tmRNA
acceptor arm and the arm corresponding to the
stacked D stem/anticodon stem in tRNAs.
Results
Construction of RNA molecules with
extended helices
In order to increase the sensitivity of TEB analysis to the interarm angles of RNA molecules, the
two arms are normally extended. For the tRNA
domain of tmRNA, this was accomplished by the
method of Friedrich et al. (1995). DNA encoding
the sequence for the 50 -half of the molecule (nucleotides 1-27) was cloned into plasmid pGJ122A,
while DNA encoding the sequence for the 30 -half
(nucleotides 326-359) was cloned into plasmid
pGJ122B. The sequences ¯anking the cloning site
of pGJ122A are complementary to the sequences
¯anking the cloning site of pGJ122B. When RNAs
are transcribed from these plasmids, they can be
annealed together to form a bimolecular complex
encompassing the tRNA domain with both the
729
acceptor and anticodon stems extended by 68 basepairs (Figure 2).
A full duplex control must be constructed with
the same contour length as the RNA molecule to
be studied. The accuracy of the TEB measurement
is sensitive to the length of the linear control,
because the interstem angle is determined from the
ratio of the rotational diffusion times of the bent
and linear RNAs, which vary roughly as the cube
of the length (VanHolde, 1971). For the tRNA
domain of tmRNA, the contour length is the sum
of the distances from the end of the acceptor and
anticodon stems to the projected vertex in the
tRNA elbow. From modeling studies, the appropriate length was found to be 37 bp (see Materials
and Methods). A DNA oligonucleotide of this
length with the appropriate restriction sites at the
ends was cloned into plasmids pGJ122A and
pGJ122B, and transcription produced a control
RNA of 173 bp (37 bp plus two arms each containing 68 bp).
Comparison of gel electrophoretic mobilities in
response to Mg2‡
Some general features of RNA structure can be
determined by relative electrophoretic mobilities in
native gels. The extended linear control and tRNA
domain RNAs were run on 8 % polyacrylamide
Figure 2. Schematic of the method for synthesizing the RNA analogous to the tRNA domain of tmRNA but with
extended helices. DNA oligonucleotides were cloned into the HindIII site of plasmids pGJ122A and pGJ122B. The
regions between the T7 promoter and SmaI site in these plasmids are complementary and antiparallel to each other
so that when RNA is transcribed from these plasmids the ends are complementary and will anneal. The resultant
RNA is extended by 68 bp on either end.
730
gels in varying concentrations of MgCl2 (Figure 3).
The tmRNA construct has a lower relative mobility
than the linear control, indicating that the tRNA
domain contains a bend. Furthermore, the relative
mobility of the tmRNA construct decreases with
Mg2‡ concentrations greater than 1 mM, indicating
that the tRNA-like structure is more bent in the
presence of Mg2‡.
Transient electric birefringence measurements
of RNAs with extended helices
General features of TEB have been described
(Hagerman, 2000). In brief, the RNA to be analyzed is placed in a cell between two crossed polarizers. The solution is subjected to a brief (1 ms)
electric ®eld pulse that partially orients the molecules and creates a birefringence signal. When the
®eld is turned off, rotational Brownian motion
causes randomization of RNA orientations, leading
to birefringence decay. With extended RNA molecules like those used in the current investigation,
the rotational diffusion coef®cient depends
strongly on the angle between the two arms of the
molecule. Decay pro®les are recorded for two molecules, one being the test RNA (possessing a central non-helix element), and the other being a
linear control (a full duplex RNA of the same total
contour length as the test RNA). The ratio of the
Structural Studies of the tRNA Domain of tmRNA
terminal decay times for the two molecules yields
the apparent angle between the helix arms of the
test RNA (Vacano & Hagerman, 1997).
In the current work, TEB experiments were performed in a variety of buffers and temperatures. In
the absence of Mg2‡, measurements were taken in
TEB and 2 TEB both at 4 C and 20 C. For all of
these conditions, the angle measured was approximately 128 (Figure 4(a) and Table 1). For TEB
with 1 mM MgCl2 at 4 C, the angle measured was
111 , while at 20 C the angle was 137 (Figure 4(b)
and Table 1).
TEB analysis is very sensitive to RNA aggregation. In some of the gel-mobility assays with 2 mM
MgCl2 a minor band was observed in the tRNA
domain lane at approximately twice the molecular
mass of the RNA, suggesting association (data not
shown). To avoid aggregation, TEB measurements
were not taken with MgCl2 concentrations above
1 mM.
Modeling of the tRNA domain
To date, there are 11 published crystal structures
of tRNAs, some free and some complexed to proteins. These tRNAs have similar tertiary interactions, including several layers of stacked bases.
The tRNA domain of tmRNA has a secondary
structure that closely resembles that of canonical
tRNAs, except for some differences in the D stem
and loop, so we based the tertiary structure of the
model on known tRNA tertiary structures, primarily tRNAPhe (Hingerty et al., 1978) and tRNAAsp
(Westhof et al., 1988) (Figure 5). Models were
developed using both of these as prototype structures, with the subtle differences between them
providing different options for solving stereochemical problems.
tmRNA residues were initially placed by superimposing them on the analogous residues of the
tRNA crystal structures. Since the core of all
Table 1. Angles observed for different buffer conditions
in the TEB analysis of the tRNA domain of tmRMA
Conditions
TEB, 4 C
TEB, 20 C
2 TEB, 4 C
2 TEB, 20 C
TEB, 1 mM MgCl2, 4 C
TEB, 1 mM MgCl2, 20 C
Figure 3. Gel electrophoresis of RNAs with extended
helices. The 8 % polyacrylamide gels were run in TBE
with (a) no MgCl2 and (b) 4 mM MgCl2. Lanes contain
1, Gel Marker I2 (Research Genetics); 2, extended tRNA
domain RNA; 3, 173 bp linear control. (c) Relative electrophoretic mobility of the tRNA domain RNA to the
linear control RNA at the indicated MgCl2 concentrations.
Angle (deg.)
Std. err.a
127.8
127.5
120.5
130.0
111.4
136.7
2.0
3.9
3.1
4.1
2.0
4.0
a
Standard errors. The variance arising from the uncertainty
in the length of the linear control (s12) was estimated by applying an interpolation function derived from a series of experiments on linear controls ranging in size from 170 to 192 bp to
the data for each experimental condition, assuming that 2 bp
corresponds to one standard deviation in length. The error arising from statistical scatter in measured decay times was calculated directly from the data, expressed as the variance of the
mean (s22). The total standard error is the square-root of the
sum of the variances, SE ˆ (s12 ‡ s22)0.5. Note that s12 is the
dominant term.
Structural Studies of the tRNA Domain of tmRNA
Figure 4. Birefringence decay curves for tRNA
domain and linear control RNAs. Graphs are shown for
TEB in the (a) absence and (b) presence of 1 mM MgCl2.
Measurements were made at 4 C (blue) and 20 C (red).
Curves for the linear control are shown as diamonds
while those for the tRNA domain are shown as circles.
known tRNA structures is comprised of several
layers of stacked and coplanar bases, this assured
the proper stacking and planarity for analogous
residues. To assure that all the sequential
50 -phosphate to 30 -hydroxyl connections could be
made, small adjustments were made to base positions and to backbone and glycosidic torsion
angles. While adjusting base positions, efforts were
made to maximize hydrogen bonding.
The elements in the tRNA domain most similar in secondary structure to those of canonical
tRNAs were modeled ®rst. These included the
acceptor stem, the T stem and loop, and resi-
731
dues analogous to 8-9 and 45-48 of canonical
tRNAs. The sequences of the acceptor/T stem
and loop are almost identical with those of
tRNAPhe, so these elements were modeled by
superimposing the tmRNA sequences onto the
tRNAPhe crystal structure. Residues A8, U9,
G333, A334, and C335 were modeled by superimposing them on tRNA residues 8, 9, 45, 46,
and 48. (This was done twice, once using a
tRNAPhe template, and once using tRNAAsp, to
explore slightly different orientations.) These
bases are important, because, in tRNAs, they
form a core of stacked bases connecting the
coaxially stacked T and acceptor stems to the
coaxially stacked D and anticodon stems. The
positions of some of these bases had to be
slightly adjusted in order to make the sequential
50 to 30 connections, but overall they remain
stacked in positions similar to those of the tRNA
crystal structures.
The next elements to be modeled were the D
loop and stem, followed by the anticodon stem.
G13 and G14 are highly conserved in tmRNAs
(Williams, 1999), suggesting that they interact with
the T loop in a manner similar to that observed for
G18 and G19 in normal tRNAs (Ushida et al.,
1994). It is important to note that G13 and G14 are
on the 50 -side of the D loop in the tmRNA secondary structure map, whereas the analogous G18 and
G19 are on the 30 -side in canonical tRNAs. Therefore, residues G13 and G14 were superimposed on
residues G18 and G19 of tRNAPhe. This placement
is supported by the fact that these residues are protected in chemical probing experiments (Hickerson
et al., 1998). The remaining residues of the D loop
and stem were then placed by positioning them so
that all of the sequential 50 to 30 connections could
be made. The D stem is modeled as two base-pairs
based on probing data (Felden et al., 1997;
Hickerson et al., 1998). Sequence analysis neither
con®rms nor disproves these base-pairs, because
G19 and A20 are 100 % conserved and the bases
at position 10 and 11 vary but can pair with G19
and A20 most of the time. There is only one base,
U12, to bridge the gap between the D stem and
G13. Although this is a large gap, U12 can bridge
it adequately. In building the D loop and stem,
small adjustments had to be made to the positions
of previously placed bases in order to make the
connections, but stacking and coplanarity were
conserved in every case.
Finally, the structure was subjected to gentle
energy minimization to correct any unfavorable
bond lengths and angles, and to eliminate unacceptable steric clashes. Figure 5 compares the ®nal
tRNA domain model with the structure of
tRNAAsp. The similarity in the folding of the two
structures is clear.
The angle between the acceptor and anticodon
stems in the model is 110 . This angle is in good
agreement with the TEB data for the tRNA domain
at 4 C in the presence of Mg2‡, adding support to
the model.
732
Structural Studies of the tRNA Domain of tmRNA
Figure 5. Stereo diagrams of the
(a) tRNA domain model and (b)
tRNAAsp. The tRNAs are shown in
similar orientations to show the
similarities in stacking and other
tertiary interactions. The coloring
scheme is the same as that used in
Figure 1.
Possible interaction of the tRNA domain with
the ribosome
The model offers a direct way of predicting
what region of the tmRNA might interact with
the ribosomal decoding site, under the assumption that this region must lie at the same distance from the end of the acceptor stem as does
the center of the anticodon loop in tRNA. This
assumption is based on the conservation of that
distance (the primary axis) in tRNAs with shortened D stems and other unusual secondary
structure features (Steinberg & Cedergren, 1994;
Steinberg et al., 1997). We de®ne the primary
axis as the distance from the phosphate group
of residue 1 to the phosphate group of residue
35 (the second residue in the anticodon). In the
crystal structures of tRNAPhe (Hingerty et al.,
(Basavappa, 1991), and tRNAAsp
1978), tRNAMet
f
(Westhof et al., 1988), the primary axes measure
Ê , 68 A
Ê , and 66 A
Ê , respectively. These dis65 A
tances closely match those of tRNAs modeled
Ê Thermus thermophilus ribosome
into the 7.8 A
Ê and
crystal structure (Cate et al., 1999) (65 A
Ê for the A and P-site tRNAs, respectively),
68 A
Ê cryo-electron microscopy
and into the 11.5 A
reconstruction of the E. coli ribosome (Gabashvili
Ê for the P-site tRNA). Using
et al., 2000) (65 A
Ê , as a
the average of these distances, 66 A
measuring rule, we ®nd that base-pair U27-A326
is at the appropriate distance from residue 1
(Figure 6). This suggests that the region between
P2a and P2b of tmRNA may interact with the
ribosomal decoding site. The exact nature of that
interaction is not known, and it probably
changes as the tmRNA moves from the A-site to
the P-site, but the presence of two unstructured
loops between P2a and P2b (Figure 1) is almost
certainly related to the dynamic nature of
tmRNA's interactions with the decoding site.
Discussion
In the current study, we determined the angle
between the acceptor stem and anticodon stem of
the tRNA domain of tmRNA through independent
experimentation and modeling. Several different
angles were observed in the TEB analysis. In the
absence of Mg‡2 the observed angles clustered
around 127 . In the presence of Mg‡2, two angles
were observed, 111 at 4 C, and 137 at 20 C.
This temperature-dependence in the presence of
Mg‡2 is probably due to loss of Mg‡2 binding at
the higher temperature. This is supported by the
Structural Studies of the tRNA Domain of tmRNA
733
Ê distance from the end
Figure 6. Structures of tRNAPhe (left) and the tRNA domain model (right) showing the 66 A
of the acceptor stem to the beginning of the decoding site. For the tRNA domain model, this distance falls on the
base-pair between U27 and A326, suggesting that the region around these bases interacts with the decoding site.
relative electrophoretic mobility studies, where the
relative mobility of the tRNA domain RNA
increases slightly at 1 mM Mg‡2 compared to that
in the absence of Mg‡2, then decreases with Mg‡2
concentrations greater than 1 mM. This shows that
the angle becomes more acute in the presence of
greater than 1 mM Mg‡2. Together these data support a conclusion that in the absence of Mg‡2 the
tRNA domain has an angle of 127 , but in the
presence of greater than 1 mM Mg‡2 the tRNA
domain takes on a more folded structure with an
angle of 111 .
Independently of the TEB data, the tRNA
domain of tmRNA was modeled by similarity to
the crystal structures of various tRNAs, particularly tRNAPhe and tRNAAsp. The secondary structure of the tRNA domain is very similar to that of
canonical tRNAs with differences only in the
length of the D stem and the sequential placement
of the G18 and G19 analogs in the D loop. Though
the tRNA domain was modeled manually, the
method used is similar to that of automated homology modeling for proteins. Residues were
assigned three-dimensional coordinates by superimposing them on analogous residues in the tRNA
cystal structures. The resulting structure maintained the conserved tertiary interactions of the
known tRNA structures but had an angle of
110 .
The modeling results are in good agreement
with the TEB data. The 110 angle calculated in
the modeled tRNA domain structure matches very
closely the 111 angle seen in the TEB experiment
with 1 mM Mg‡2 at 4 C. To make a model where
the angle matched the larger observed angles
(127 -137 ), several of the tertiary interactions
(speci®cally those involving residues 8, 9, and 333-
335) in the tRNA-like core would have to be broken. Thus, a model where the tRNA domain has
an angle of 110 is most similar to canonical
tRNAs and is the most likely conformation for the
tRNA domain at physiological conditions.
It is likely that other tmRNAs have a similar
angle for their tRNA domains. Sequence analysis
of numerous tmRNA sequences (Williams, 1999;
Zwieb et al., 1999) has shown that the tRNA
domain consists of a normal acceptor stem and T
stem and loop, but a shortened or absent D stem.
Here, we have found that a shortened D stem in
the tRNA domain of tmRNA coincides with a 110 angle between the acceptor stem and the anticodon
stem. Since the non-canonical D stem is phylogenetically conserved, tmRNAs from other species
probably have similar angles. It is likely that the
non-canonical D stem probably has some importance for tmRNA function, perhaps in the way
tmRNA interacts with the ribosome.
In normal translation, an mRNA codon pairs
with a tRNA anticodon at the decoding site of the
ribosome, and cognate pairing provides the signal
for subsequent steps leading to peptidyl transfer.
Since tmRNA has evolved to interact with the ribosome when the mRNA is damaged, the tRNA
domain of tmRNA does not have an anticodon.
The primary axis length from the end of the acceptor stem to the anticodon can be used to determine
the region of the tmRNA that is closest to the
decoding site of the ribosome. This distance is
Ê in normal tRNAs. In our model for the
about 66 A
Ê
tRNA domain of tmRNA, the region that lies 66 A
from the acceptor terminus falls in the extended
anticodon stem, between P2a and P2b. This region
consists of a series of bulges and non-canonical
base-pairs, according to the tmRNA secondary
734
structure (Figure 1). These could form a structure
that interacts with the decoding site and mimics
the codon/anticodon of a normal mRNA/tRNA
complex, signaling the ribosome to proceed with
peptidyl transfer and begin synthesizing the carboxy-terminal region of the incomplete protein.
Materials and Methods
Preparation of plasmids
Construction of the plasmids used for the in vitro
transcription of the RNAs used in this study was as
described (Friederich et al., 1995). Duplex DNAs containing the sequence of residues 1-25 and 326-359 of
the tRNA domain of tmRNA were synthesized by
Research Genetics. These were cloned into the HindIII
site of plasmids pGJ122A and pGJ122B, respectively.
For the linear control, DNA oligonucleotides were synthesized with sequences close to those of the tRNA
domain, thus reducing the chance of sequence-dependent differences in the rotational decay time. These
were likewise inserted into the HindIII sites of the
pGJ122A and pGJ122B.
In vitro transcription and annealing of
RNA molecules
For the in vitro transcription of the RNAs used in this
study, reactions were set up in 1 ml volumes with the
following reagents: 20 mg of template plasmid, 2.5 mM
each NTP, 5 units of pyrophosphatase, 400 units of
RNasin, and 100 units of phage T7 RNA polymerase.
RNA pairs were annealed by mixing equimolar amounts
of each RNA in 100 mM Tris-HCl (pH 6.5), 100 mM
NaCl, 10 mM EDTA (pH 8.0), 5 mM EGTA. The solutions were then heated to 95 C and allowed to cool to
room temperature. For puri®cation, the annealed RNAs
were run out on 6 % (w/v) polyacrylamide gels, and
bands were cut out and eluted.
Determination of the length of the linear control
For determining the length of the linear control, a
reduced representation model of the tRNA domain was
made with YAMMP (Tan & Harvey, 1993). For this
model, normal A-form helix parameters were used as
constraints for Watson-Crick base-pairs. Also the acceptor stem was stacked on the T stem, and the D stem was
stacked on the anticodon stem. Finally, residues 13 and
Ê of residues 51 and
14 were constrained to be within 18 A
52, respectively, which is analogous to the interactions of
G13 and G14 with U342 and C343. The structure with
these constraints was subjected to 10,000 rounds of steepest descents minimization. From the resulting structure,
the distances from the end of the acceptor stem and
anticodon stem to the intersection where they meet was
Ê . This length
measured, giving a total length of 104 A
Ê , the rise per base of a standard
was divided by 2.8 A
A-form helix, to give 37 bp, which is the length of
the insert used for the linear control.
Gel electrophoresis
The 8 % polyacrylamide gels (19:1 (w/w) monomer to
bisacrylamide) were run at room temperature with recirculating buffer. TBE running buffer was used with the
Structural Studies of the tRNA Domain of tmRNA
speci®ed concentration of MgCl2. Gels were run at 100V
for 2.5 hours. The relative mobilities were determined
from the ratio of the mobility of the tRNA domain RNA
to the linear control.
Transient electric birefringence
The application of TEB to the study of angles between
adjacent helices in RNA has been described (Hagerman,
1996, 2000; Vacano & Hagerman, 1997). The speci®c conditions used in this study are as follows: 1.0 ms pulse
width, 1 Hz pulse frequency, and 10 kV/cm orienting
®eld strength. RNAs to be measured were placed in
either TEB (5 mM NaPi, 0.125 mM Na2EDTA) or
2 TEB (10 mM NaPi, 0.25 mM Na2EDTA) buffer as
speci®ed. Birefringence decay pro®les were collected by
taking ®ve sets of averaged 512 pulses for the 20 C,
1 mM MgCl2 data set. For all other cases, ®ve sets of
averaged 125 pulses were taken. The interarm angle, y,
is given by:
y ˆ 180 ÿ f1:46 cosÿ1 …R† ‡ 0:005‰sinÿ1 …1 ÿ R†Š2:3 g
where R is the ratio of the terminal decay time of the test
RNA to that of the linear control (Frazer-Abel &
Hagerman, 1999).
Modeling
The tRNA domain model was constructed as
described in Results using InsightII (Molecular
Simulations, Inc) and MANIP (Massire & Westhof,
1998). In order to relieve any unfavorable bond
lengths, angles, or steric interactions, the resultant
model was subjected to 200 rounds of steepest
descents minimization using the cvff force-®eld in
InsightII. Coordinates for the model are available at
http://uracil.cmc.uab.edu/Publications/
Acknowledgments
We thank Stephen Hajduk and his group for the use
of laboratory space and reagents, and for numerous
helpful discussions. This work was supported by grants
from the NIH (GM53827 to S.C.H. and GM35305 to
P.J.H.).
References
Ê crystal structure of yeast
Basavappa, R. (1991). The 3 A
initiator tRNA: functional implications in initiator/
elongator discrimination. EMBO J. 10, 3105-3111.
Cate, J. H., Yusupov, M. M., Yusupova, G. Z., Earnest,
T. N. & Noller, H. F. (1999). X-ray crystal structures
of 70S ribosome functional complexes. Science, 285,
2095-2104.
Felden, B., Himeno, H., Muto, A., McCutcheon, J. P.,
Atkins, J. F. & Gesteland, R. F. (1997). Probing the
structure of the Escherichia coli 10Sa RNA (tmRNA).
RNA, 3, 89-103.
Felden, B., Hanawa, K., Atkins, J. F., Himeno, H., Muto,
A., Gesteland, R. F., McCloskey, J. A. & Crain, P. F.
(1998). Presence and location of modi®ed nucleotides in Escherichia coli tmRNA: structural mimicry
with tRNA acceptor branches. EMBO J. 17, 31883196.
Structural Studies of the tRNA Domain of tmRNA
Frazer-Abel, A. A. & Hagerman, P. J. (1999). Determination of the angle between the acceptor and anticodon stems of a truncated mitochondrial tRNA.
J. Mol. Biol. 285, 581-593.
Friederich, M. W., Gast, F. U., Vacano, E. & Hagerman,
P. J. (1995). Determination of the angle between the
anticodon and aminoacyl acceptor stems of yeast
phenylalanyl tRNA in solution. Proc. Natl Acad. Sci.
USA, 92, 4803-4807.
Gabashvili, I. S., Agrawal, R. K., Spahn, C. M. T.,
Grassucci, R. A., Svergun, D. I., Frank, J. &
Penczek, P. (2000). Solution structure of the E. coli
Ê resolution. Cell, 100, 537-549.
70 S ribsome at 11.5 A
Hagerman, P. J. (1996). Sometimes a great motion: the
application of transient electric birefringence to the
study of macromolecular structure. Curr. Opin.
Struct. Biol. 6, 643-649.
Hagerman, P. J. (2000). Transient electric birefringence
for determining global conformations of nonhelix
elements and protein-induced bends in RNA.
Methods Enzymol. 317, 440-453.
Hickerson, R. P., Watkins-Sims, C. D., Burrows, C. J.,
Atkins, J. F., Gesteland, R. F. & Felden, B. (1998). A
nickel complex cleaves uridine in folded RNA structures: application to E. coli tmRNA and related
engineered molecules. J. Mol. Biol. 279, 577-587.
Hingerty, B., Brown, R. S. & Jack, A. (1978). Further
re®nement of the structure of yeast tRNAPhe. J. Mol.
Biol. 124, 523-534.
Karzai, A. W., Roche, E. D. & Sauer, R. T. (2000). The
SsrA-SmpB system for protein tagging, directed
degradation and ribosome rescue. Nature Struct.
Biol. 7, 449-455.
Komine, Y., Kitabatake, M., Yokogawa, T., Nishikawa,
K. & Inokuchi, H. (1994). A tRNA-like structure is
present in 10Sa RNA, a small stable RNA from
Escherichia coli. Proc. Natl Acad. Sci. USA, 91, 92239227.
Leehey, M. A., Squassoni, C. A., Friederich, M. W.,
Mills, J. B. & Hagerman, P. J. (1995). A noncanonical tertiary conformation of a human mitochondrial
transfer RNA. Biochemistry, 34, 16235-16239.
Massire, C. & Westhof, E. (1998). MANIP: an iteractive
tool for modelling RNA. J. Mol. Graph. Modell. 16,
197-205.
735
Rudinger-Thirion, J., GiegeÂ, R. & Felden, B. (1999).
Aminoacylated tmRNA from Escherichia coli interacts with prokaryotic elongation factor Tu. RNA, 5,
989-992.
Steinberg, S. & Cedergren, R. (1994). Structural compensation in atypical mitochondrial tRNAs. Nature
Struct. Biol. 1, 507-510.
Steinberg, S., Gautheret, D. & Cedergren, R. (1994). Fitting the structurally diverse animal mitochondrial
tRNAs(Ser) to common three-dimensional constraints. J. Mol. Biol. 236, 982-989.
Steinberg, S., Leclerc, F. & Cedergren, R. (1997). Structural rules and conformational compensations in
the tRNA L-form. J. Mol. Biol. 266, 269-282.
Tan, R. K.-Z. & Harvey, S. C. (1993). Yammp: development of a molecular mechanics program using the
modular programming method. J. Comput. Chem.
14, 455-470.
Ushida, C., Himeno, H., Watanabe, T. & Muto, A.
(1994). tRNA-like structures in 10Sa RNAs of Mycoplasma capricolum and Bacillus subtilis. Nucl. Acids
Res. 22, 3392-3396.
Vacano, E. & Hagerman, P. J. (1997). Analysis of
birefringence decay pro®les for nucleic acid helices
possessing bends: the tau-ratio approach. Biophys. J.
73, 306-317.
VanHolde, K. E. (1971). Physical Biochemistry PrenitceHall, Englewood Cliffs, NJ.
Westhof, E., Dumas, P. & Moras, D. (1988). Restrained
re®nement of two crystalline forms of yeast aspartic
acid and phenylalanine transfer RNA crystals. Acta
Crystallog. sect. A, 44, 112-123.
Williams, K. P. (1999). The tmRNA website. Nucl. Acids
Res. 27, 165-166.
Williams, K. P. & Bartel, D. P. (1996). Phylogenetic
analysis of tmRNA secondary structure. RNA, 2,
1306-1310.
Zacharias, M. & Hagerman, P. J. (1995). Bulge-induced
bends in RNA: quanti®cation by transient electric
birefringence. J. Mol. Biol. 247, 486-500.
Zwieb, C., Wower, I. & Wower, J. (1999). Comparative
sequence analysis of tmRNA. Nucl. Acids Res. 27,
2063-2071.
Edited by I. Tinoco
(Received 8 January 2001; received in revised form 16 March 2001; accepted 16 March 2001)