903
Journal of Cell Science 107, 903-912 (1994)
Printed in Great Britain © The Company of Biologists Limited 1994
Nuclear export of signal recognition particle RNA is a facilitated process that
involves the Alu sequence domain
Xiao-Ping He, Nelly Bataillé* and Howard M. Fried†
Department of Biochemistry and Biophysics, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA
*Present address: INSERM U298, CHRU Angers, 49 033 Angers Cedex 01, France
†Author for correspondence
SUMMARY
The signal recognition particle is a cytoplasmic RNAprotein complex that mediates translocation of secretory
polypeptides into the endoplasmic reticulum. We have used
a Xenopus oocyte microinjection assay to determine how
signal recognition particle (SRP) RNA is exported from the
nucleus. Following nuclear injection, SRP RNA accumulated in the cytoplasm while cytoplasmically injected SRP
RNA did not enter the nucleus. Cytoplasmic accumulation
of SRP RNA was an apparently facilitated process
dependent on limiting trans-acting factors, since nuclear
export exhibited saturation kinetics and was completely
blocked either at low temperature or by wheat germ agglutinin, a known inhibitor of nuclear pore-mediated
transport. At least one target for trans-acting factors that
promote nuclear export of SRP RNA appears to be the Alu
INTRODUCTION
Transport of endogenous macromolecules into and out of the
nucleus is generally regarded as a vectorial, energy-requiring
process dependent upon interactions between specific
molecular signals in the transported substances and receptors
for these signals (Feldherr, 1992; Forbes, 1992; Gerace, 1993).
Proteins that enter the nucleus possess discrete nuclear localization sequences that, while lacking significant sequence
identity (Dingwall and Laskey, 1991; Richter and Standiford,
1992), all appear to interact with a common set of cytoplasmic
and nuclear pore complex (NPC) receptors to accomplish their
nuclear import (Yamasaki and Lanford, 1992; see also
Imamoto et al., 1992, and references therein). Discovery of
signals and receptors for RNA transport, however, has revealed
a more heterogeneous picture (Fried, 1992; Izaurralde and
Mattaj, 1992). Since all RNA molecules likely exist in association with one or more proteins, information specifying RNA
transport could reside in the RNA, in its associated proteins,
or within both moieties. For example, efficient mRNA
transport requires a 5′-m7G cap while other evidence suggests
that particular hnRNP proteins, the poly(A) tail, and even premRNA splicing also have roles in mRNA export (Dargemont
and Kühn, 1992; Izaurralde and Mattaj, 1992; Piñol-Roma and
element of the molecule, since a transcript consisting of
only the Alu sequence was exported from the nucleus in a
temperature-dependent manner and the Alu transcript
competed in the nucleus for transport with intact SRP
RNA. Although the identities of trans-acting factors
responsible for SRP RNA transport are at present
unknown, we suggest that proteins contained within the
cytoplasmic form of SRP are candidates. Consistent with
this idea were the effects of a mutation in SRP RNA that
prevented binding of two known SRP proteins to the Alu
sequence.
Key words: RNA transport, nuclear pore complex,
ribonucleoprotein, Xenopus laevis, oocyte microinjection
Dreyfuss, 1993; Riedel and Fasold, 1992). Nuclear export of
U snRNAs transcribed by RNA polymerase II is similarly
enhanced by a m7G cap; re-entry of U snRNAs into the nucleus
requires hyper-methylation of the cap (to m3G) as well as
binding of the Sm proteins (U1 and U2 snRNA), is moderately
enhanced by the m3G cap and Sm proteins (U4 and U5), or is
totally independent of cap/Sm elements, relying instead on a
specific RNA sequence (U3) (Baserga et al., 1992; Izaurralde
and Mattaj, 1992; Terns et al., 1993; Wersig et al., 1992). A
particular cap binding protein has been implicated in nuclear
export of m7G-capped RNAs (Izaurralde et al., 1992) and
evidence for a nuclear localization signal in the Sm core has
been presented (Fischer et al., 1993); yet to be identified are
trans-acting transport factors that recognize the m3G cap.
Still other RNAs have been examined for their transport
requirements. Nuclear export of tRNA occurs through specific
recognition (Zasloff, 1983) but just what is recognized is
unclear, since nucleotide substitutions in virtually every
portion of a tRNA impaired transport (Tobian et al., 1985).
Nuclear export of 5 S rRNA (which appears to occur only in
oocytes) requires binding to either ribosomal protein L5 or the
5 S gene transcription factor TFIIIA (Guddat et al., 1990);
exactly how these two quite dissimilar proteins direct 5 S RNP
transport is unknown. Likewise, 5 S rRNA is re-imported into
904
X.-P. He, N. Bataillé and H. M. Fried
the nucleus by a facilitated process but what trans-acting
factors mediate re-import is also unknown (Allison et al.,
1993). Finally, ribosomal subunits may not possess any
discrete transport signal but appear to be recognized for export
from the nucleus by virtue of a general biochemical property
(Bataillé et al., 1990).
Small RNAs such as U snRNA, tRNA and 5 S rRNA are
convenient subjects for transport studies, since their primary
and secondary structures can be easily manipulated and since
they likely interact with limited numbers of proteins whose
identities and possible roles in transport should be discernible.
Still, U snRNA transport is complicated by the involvement of
the m7G cap, the m3G cap and Sm proteins, while tRNA
transport may be intimately associated with processing. Thus,
in the study reported here, we have begun to investigate mechanisms of RNA transport using another small RNA, namely
SRP RNA (previously denoted 7SL RNA). SRP RNA (300
nucleotides) together with six proteins (SRP9, 14, 19, 54, 68
and 72) comprise the signal recognition particle (SRP), whose
function is to mediate translocation of secretory polypeptides
into the endoplasmic reticulum (Rapoport, 1991; Walter and
Lingappa, 1986). SRP RNA consists of an Alu sequence
(approx. residues 1-105 and 256-300) related to the human and
rodent families of Alu interspersed repetitive sequences (Ullu
et al., 1982; Ullu and Tschudi, 1984) interrupted by a unique
S sequence (residues 106 through 255). The Alu domain and
associated proteins SRP9 and SRP14 make up SRP’s translational arrest activity; the S domain is associated with SRP19
and SRP54, the latter protein mediating signal-sequence recognition, as well as SRP68 and SRP72, which promote polypeptide translocation across the ER membrane (Siegel and Walter,
1988; see Fig. 1 for a summary of SRP structure-function).
Since SRP functions in the cytoplasm, one step in SRP biosynthesis must involve nuclear export of SRP RNA. Furthermore,
since SRP RNA is not capped, spliced, polyadenylated or
known to be modified in any other way (except perhaps for
removal of several 3′-terminal uridylate residues from the
primary transcript), nuclear export of SRP RNA most likely
depends upon trans-acting factors that recognize specific structural features of the molecule. Thus, the simplicity of SRP
RNA makes it a convenient subject for identifying nuclear
transport mechanisms.
To determine which structural features of SRP RNA are
required for its nuclear efflux and what nuclear proteins
interact with SRP RNA to promote its export, we have used a
Xenopus oocyte microinjection assay. We found that nuclearinjected SRP RNA accumulated in the cytoplasm by an apparently receptor-mediated process, since transport was temperature dependent and inhibited by wheat germ agglutinin
(WGA), a known inhibitor of NPC-mediated transport. Competition experiments demonstrated that transport was
dependent upon limiting cellular components. Information
directing transport of SRP RNA appears to reside, at least
partly, within the Alu portion of the molecule, since a transcript
corresponding to this sequence exhibited properties of
mediated transport and competed for the same limiting components. While the identity of any trans-acting factor
mediating SRP RNA transport is at present unknown, proteins
found in the cytoplasmic form of SRP are possible candidates.
This idea was supported by effects of a mutation in SRP RNA
that blocked binding of SRP9/14.
MATERIALS AND METHODS
Cloning and construction of SRP RNA transcription
templates
The coding sequence for Xenopus SRP RNA was obtained from a
Xenopus SRP RNA gene, isolated by screening a λ phage Xenopus
genomic clone library with a hybridization probe consisting of a PvuIStyI restriction fragment derived from the S domain of a human SRP
RNA gene (Ullu and Weiner, 1984). (An initial attempt to obtain this
same restriction fragment from a Xenopus SRP RNA cDNA was
unsuccessful, owing to several errors in the published sequence; the
revised sequence contains G, G and C at positions 71, 136 and 180,
respectively.) We verified that the cloned gene was an expressed gene
by injecting it into Xenopus oocyte nuclei together with [α-32P]GTP;
an RNA transcript of the correct size was produced. The sequence
surrounding the 3′ end of SRP RNA was determined to
be...GTTCTTTTCTTTTTA..., where the underlined bases are the end
of the mature RNA found in SRP. Since SRP RNA is a product of
RNA polymerase III, which usually terminates transcription following
a run of T residues, and since the La protein is found associated with
the oligo(U) stretch at the end of all newly synthesized Pol III transcripts, we presume that the primary in vivo SRP RNA transcript terminates with three additional U residues, which, like many Pol III
transcripts, are subject to 3′ end trimming (Hendrick et al., 1981;
Rinke and Steitz, 1982). To synthesize this primary transcript in vitro
for injection into oocytes, the coding sequence was amplified by the
polymerase chain reaction (PCR) and cloned into the phagemid
pUC119 (Vieira and Messing, 1987). The 5′ PCR primer, TAATACGACTCACTATAG*CCGGGCGCTGCTGTGG, contained at its 5′
end the T7 bacteriophage promoter (underlined), whose transcription
start site (G*) was coincident with the first base in SRP RNA. The 3′
primer, TTT*AAAAGAACTGTGTCTCG, contained a DraI restriction site (underlined) so that the template cleaved with this enzyme
(at *) would produce a run-off transcript with a 3′ end identical to the
presumed in vivo primary transcript. The cloned PCR product was
sequenced in its entirety.
A template for synthesis of an RNA molecule corresponding to the
S domain of SRP RNA was constructed by amplifying the sequence
between residues 99 and 250. The 5′ PCR primer, TAATACGACTCACTATAG*ATCGGGTGTCCGCA, contained the T7 promoter
(underlined) with a transcription start site (G*) coincident with residue
99 of SRP RNA; the 3′ primer, GAATGCT*CACAGGCGCGATCC,
contained a BsmI restriction site (underlined; this site is not present
in the SRP RNA gene) so that the template cleaved with this enzyme
(at *) would produce a run-off transcript whose 3′ end would correspond to residue 250 of SRP RNA. The PCR product was cloned into
pUC119 and sequenced. The template for in vitro transcription of the
Alu domain of SRP RNA was obtained in two steps. Two mutagenic
oligonucleotide primers were annealed simultaneously to the singlestranded form of pUC119 containing the intact T7-SRP RNA gene
(see above). Second-strand synthesis followed by transformation of
Escherichia coli was carried out as described (Kunkel et al., 1991) to
derive a mutant SRP coding sequence in which the GT at positions
96 and 97 were changed to CG, creating an NruI restriction site, and
the GT at positions 248 and 249 were changed to TC, creating a BstBI
restriction site. The plasmid was cleaved with BstBI, the ends filled
in with E. coli DNA polymerase Klenow fragment, followed by
cleavage with NruI and religation. The resultant SRP RNA gene thus
was deleted from residues 96 through 249 and contained a three-base
insertion (CGC) following residue 95. This same plasmid, from which
the Alu domain RNA could now be synthesized, was further mutagenized to convert the GUA at positions 24-26 to AGG (see Results);
the three-base substitution created an AvrII restriction site, providing
the means to screen isolates rapidly for the desired mutation. A
control template containing the murine U6 snRNA sequence was
produced by PCR amplification of the U6 gene (provided by Jack
SRP RNA nuclear transport
Keene, Duke University). The 5′ primer, TAATACGACTCACTATAG*GG†TGCTCGCTTCGGC, contained the T7 promoter
(underlined) while the transcription start site (*) was separated from
the authentic U6 start by a G, as the inclusion of two additional G
residues enhanced the efficiency of T7 transcription from this particular template. The 3′ primer, TTT*AAAATATGGAACGCTTCAC,
contained a DraI cleavage site, so as to produce a run-off transcript
whose 3′ end was identical to authentic U6 snRNA. For some experiments purified total yeast tRNA (Sigma Chemical Co.) was used as
a control.
Computer-predicted RNA secondary structure
Predicted RNA secondary structures were calculated using the FOLD
program of Zuker and Stiegler (1981) on a VAX computer. The output
from FOLD was converted to a graphical image using LoopViewer,
a Macintosh computer program written by Don Gilbert (Biology
Department, Indiana University).
RNA synthesis
Transcription templates were cleaved with the appropriate restriction
enzyme, phenol extracted, and recovered by ethanol precipitation.
32P-labeled RNA was synthesized from 1 µg of linearized template in
a reaction containing 40 mM Tris-HCl (pH 7.5), 6 mM MgCl2, 2 mM
spermidine, 10 mM NaCl, 10 mM DTT, 500 µM each of ATP, GTP
and CTP, 12.5 µM UTP, 50 µCi [α-32P]UTP (10 mCi/mmole), 20
units RNasin (Promega), and 20 units T7 RNA polymerase. Reactions
were incubated for 1 hour at 37˚C, after which they were subjected
to phenol-chloroform extraction and ethanol precipitation. A sample
of the radiolabeled RNA was used to determine yield and subjected
to electrophoresis under denaturing conditions in 6% polyacrylamide/8 M urea gels to verify that a product of correct size was
produced. The gels were dried and subjected to autoradiography.
Unlabeled RNA was synthesized according to Weitzmann et al.
(1990) in a reaction containing 40 mM Tris-HCl (pH 8.0), 8 mM
Mg(OAc)2, 25 mM NaCl, 2 mM spermidine, 10 mM DTT, 2.5 mM
each ribonucleotide triphosphate, 4 pmoles/ml of template, 1000
units/ml RNasin, 5 units/ml yeast inorganic pyrophosphatase (Sigma),
and 500-1000 units/ml T7 RNA polymerase. The reactions were
incubated for from 5 hours to overnight at 37˚C, after which 25 units
of RNase free DNase were added for an additional 30 minutes at 37˚C.
The reaction mixtures were extracted with phenol-chloroform, and
RNA was recovered by ethanol precipitation. Because of the large
amount of unincorporated ribonucleotide that precipitated in ethanol,
the RNA sample was dissolved in 100 mM NaCl, 10 mM Tris-HCl,
1 mM EDTA (pH 7.5) and centrifuged in a Centricon 30 concentrator (Millipore). The concentrated RNA was diluted with fresh buffer
and re-concentrated. The final yield of RNA was determined spectrophotometrically and a sample was subjected to electrophoresis in
a formaldehyde-agarose gel, stained after electrophoresis with
ethidium bromide, to assess RNA integrity. Typically, these largescale reactions produced about 5 mg of SRP RNA, 3 mg of Alu
domain RNA, 1 mg S domain RNA, and 1.5 mg U6 snRNA per ml
of reaction.
Oocyte injections and RNA analysis
Late stage Xenopus oocytes were injected as described previously
(Bataillé et al., 1990). Briefly, oocytes, oriented with their animal pole
upward, were centrifuged at 650 g for 10 minutes, causing the nuclei
to migrate to the surface and creating a ring within the animal pole
pigment that served as a target for microinjection. All samples were
mixed with polyvinylpyrrolidone-coated colloidal gold to monitor the
actual site of injection; following injection and dissection of the
oocytes to separate nuclei and cytoplasm, only nuclei that were pink
(from the colloidal gold) were analyzed. Samples for nuclear injections were 20 nl or less while up to 50 nl of sample was injected into
the cytoplasm. For experiments in which samples were injected into
both the cytoplasm and nuclei of the same cells, the oocytes were cen-
905
trifuged and injected first with the cytoplasmic sample; these injections required about 20 minutes to complete, after which the nuclei
were injected (the ring-like target was generally visible for only about
20 minutes). Following injection and various periods of incubation,
cells were fixed for 30 minutes in 1% trichloroacetic acid at 4˚C.
Oocytes were then manually enucleated and the samples frozen.
Samples were extracted by homogenization in 50 mM Tris-HCl (pH
7.5), 5 mM EDTA, 1.5% SDS, 300 mM NaCl, 1.5 mg/ml proteinase
K, and 20 µg/ml yeast tRNA (Fischer et al., 1991), using a small
plastic conical rod (Kimble) designed to fit a microcentrifuge tube.
Following homogenization, the samples were extracted with phenolchloroform and ethanol precipitated twice. The RNA was dissolved
in water at a concentration of one oocyte nucleus or cytoplasm/µl.
Samples (1 µl) were then subjected to electrophoresis under denaturing conditions in 6% polyacrylamide/8 M urea gels. The gels were
dried and subjected to autoradiography. For experiments in which
quantitation was desired, samples were loaded in alternate wells so as
to be well separated and the resulting autoradiograph was used as a
template to excise the radiolabeled RNA bands from the dried gel,
which were then assayed by scintillation counting.
RESULTS
Nuclear-injected SRP RNA accumulates in the
cytoplasm of Xenopus oocytes
To investigate requirements for its cytoplasmic accumulation,
we joined the sequence for Xenopus SRP RNA to a T7 bacteriophage promoter and transcribed [32P]SRP RNA in vitro;
the derived transcript was designed to be identical to the
predicted in vivo primary transcription product (see Materials
and Methods). The RNA was isolated and injected into
Xenopus oocytes. After various time periods, cells were
manually enucleated and the distribution of SRP RNA in the
nuclear and cytoplasmic fractions was determined by denaturing gel electrophoresis. Fig. 2A shows that SRP RNA injected
into the nucleus (~0.5 fmol injected/nucleus) steadily accumulated in the cytoplasm over a 24 hour period. Consistent with
its normal subcellular location, SRP RNA injected into the
cytoplasm did not enter the nucleus. The graph in Fig. 2B
shows a quantitative representation of another experiment
similar to the one depicted in the gel. The peak of cytoplasmic
accumulation usually occurred about 12 hours following
injection; by that time about 80% of the RNA had reached the
cytoplasm (results of 14 measurements, s.d.=10%). Infrequently, considerably less RNA accumulated in the cytoplasm
(e.g. see Fig. 3A, 19˚C), presumably due to more marked differences in transport potential between some batches of
oocytes. We would also point out that the percentage of RNA
in the cytoplasm is slightly overestimated, especially at longer
times of incubation, since the RNA was somewhat unstable in
the nucleus and degraded slowly. This nuclear degradation was
evident from the finding that when RNA was injected into the
nucleus the total amount of RNA recovered at later time points
was always less than at earlier time points, but when RNA was
injected into the cytoplasm essentially 100% of the RNA was
always recovered, even at late time points; at 4˚C the RNA was
prevented from nuclear degradation (e.g. Fig. 3A).
Transport of SRP RNA is a facilitated NPC-mediated
process
To determine whether cytoplasmic accumulation of injected
906
X.-P. He, N. Bataillé and H. M. Fried
A
Cyto. inject.
Nuclear injection
Uninjected
SRP
RNA
100
B
B
% in cytoplasm
80
60
40
20
0
0
6
12
18
24
30
Time (hours)
Fig. 1. Structure-function relationships in the signal recognition
particle. Top: phylogenetically-based predicted secondary structure
of Xenopus SRP RNA. The vertical arrows indicate the approximate
boundary between the Alu and S sequences. The scissors span the
bases between which cleavage was carried out to delete the S
sequence and join the segments of the Alu sequence together for
synthesis of Alu RNA (see Materials and Methods). Bottom: diagram
of SRP. The shaded rectangular blocks represent major helices and
domains of SRP RNA, the numbers corresponding to the large
numbers in the top drawing. The ovals represent SRP proteins. The
lower figure is adapted from Zwieb (1991) and reproduced with
permission,  Oxford University Press. The assignment of SRP
functional domains is from Siegel and Walter (1988).
SRP RNA occurred by a facilitated process or by free
diffusion, we examined the distribution of RNA in cells maintained at 4˚C. Fig. 3A shows that transport of SRP RNA to the
cytoplasm was blocked completely at low temperature. Inhibition of transport was reversible, since cells maintained at 4˚C
for 6 hours showed substantial cytoplasmic accumulation of
SRP RNA after being returned to 19˚C for an additional 6
hours. Reversible inhibition suggests that low temperature
blocked RNA transport by inhibiting the activity of a specific
transporter, e.g. the NPC, rather than by preventing transport
by way of non-specific cellular damage.
Supplementary evidence for a mediated process was
obtained by examining the effect of wheat germ agglutinin
(WGA) on transport. WGA, which binds a group of O-linked
N-acetylglucosamine (GlcNAc)-modified proteins in the NPC,
has been shown to inhibit facilitated nuclear import of proteins
and snRNPs as well as export of ribosomal subunits, mRNA,
U snRNAs and tRNA (reviewed by Starr and Hanover, 1992;
see also Bataillé et al., 1990; Dargemont and Kühn, 1992;
Fischer et al., 1991; Michaud and Goldfarb, 1992; Neuman de
Vegvar and Dahlberg, 1990). As Fig. 3B shows, co-injection
of WGA (~15 µM final intranuclear concentration) completely
blocked cytoplasmic accumulation of SRP RNA. When
GlcNAc was also included in the sample before injection,
transport of SRP RNA was essentially unaffected. The sup-
Fig. 2. Subcellular distribution of SRP RNA following
microinjection into Xenopus oocytes. (A) [32P]SRP RNA was
injected into either the nucleus or the cytoplasm, cells were
incubated at 19˚C for various periods of time, and processed as
described in Materials and Methods to determine the location of the
injected RNA. N, nucleus; C, cytoplasm; numbers above pairs of N
and C are incubation times in hours. (B) Quantitation of an
experiment similar to that shown in A.
pression by GlcNAc of WGA’s ability to inhibit export demonstrated that inhibition was specific for glycosyl residues, most
likely those present in NPC proteins, and not the result of a
non-specific effect of WGA.
In summary, the reversible inhibition of transport at low
temperature and GlcNAc-specific inhibition by WGA
indicated that nuclear efflux of injected SRP RNA occurred by
a facilitated, presumably NPC-mediated process.
Transport of SRP RNA is dependent upon a limiting
component
The preceding experiments provided evidence that injected
SRP RNA is transported by a mediated process, implying that
the molecule is recognized by one or more trans-acting
transport factors. Further evidence for involvement of transacting components was derived from competition experiments
in which a fixed amount (0.5 fmol/10 nM final concn) of
[32P]SRP RNA was injected into oocyte nuclei along with
various amounts of unlabeled SRP RNA. As can be seen in
Fig. 4A, an increasing amount of unlabeled SRP RNA progressively diminished appearance of labeled RNA in the
cytoplasm. Although it would appear that this result was more
pronounced at 6 hours than at 12, we emphasize again that the
later time points are under-represented in their nuclear content
of RNA due to slow degradation in that compartment (see
above), thus artificially inflating the percentage of RNA in the
cytoplasm. Competition for transport was specific, as either an
equivalent molar excess of unlabeled U6 snRNA (320×), or
A
19°C
4°C
4-19°C
SRP RNA nuclear transport
907
A
100
No Comp.
32 X
160 X
320 X
% in cytoplasm
80
60
40
20
B
0
0
6
12
Time (hours)
B
100
five times that amount (1800×), was without effect (Fig. 4B).
U6 snRNA was chosen as a non-specific competitor because
normally it is not transported to the cytoplasm (Vankan et al.,
1990) and its continued presence in the nucleus during the
course of the experiment should have revealed any ill effects
of injecting large amounts of RNA. (Using a radiolabeled transcript, we verified that the in vitro synthesized U6 snRNA was
stable and that it remained in the nucleus (data not shown).)
Given that the total amount of SRP RNA for each injection
was known, the data in Fig. 4A were used to calculate an
estimated apparent Km and Vmax for SRP RNA transport in
oocytes (since the 12 hour time points are more affected by the
slow turnover of the RNA, transport rates were calculated from
the slopes of the lines between time zero and 6 hours only).
Fig. 4C shows the rate of transport as a function of the amount
of RNA injected. The transport rate approached a plateau as
the concentration of SRP RNA increased; such saturation
kinetics are indicative of a facilitated transport mechanism in
which the amount of the mediating component is limiting. A
double reciprocal plot of the data in Fig. 4C (not shown)
yielded an apparent Km=1 µM and Vmax=3×107 molecules
transported/minute per nucleus.
The Alu domain of SRP RNA is at least one target
for factors that mediate SRP RNA transport
Saturation kinetics and inhibition of transport by either low
60
No Comp.
320 X
1800 X
40
20
0
0
6
12
Time (hours)
C
500
_4
400
fmole/min X 10
Fig. 3. Transport of SRP RNA is a facilitated process. (A) Effect of
low temperature on transport of SRP RNA. Nuclei were injected
with [32P]SRP RNA and maintained at 19˚C or 4˚C for the times (in
hours) indicated by the numbers, after which cells were enucleated
and the distribution (N, nuclear; C cytoplasmic) of RNA determined.
The last two lanes on the far right of the gel show results for cells
incubated at 4˚C for 6 hours and then transferred to 19˚C for an
additional 6 hours. (B) Effect of wheat germ agglutinin on transport
of SRP RNA. [32P]SRP RNA was mixed with WGA, or with WGA
+ N-acetylglucosamine (GlcNAc) and injected into nuclei. Assuming
a nuclear volume of 40 nl (Gurdon and Wickens, 1983), WGA was
added to the RNA at an amount calculated to yield a final
intranuclear concentration of ~15 µM; GlcNAc was added to a final
intranuclear concentration of 20 mM.
% in cytoplasm
80
300
200
100
0
0
0.5 fmole
50
100
150
200
fmole injected
Fig. 4. SRP RNA nuclear export is saturable. (A) Extent of transport
of nuclear-injected [32P]SRP RNA (0.5 fmol) co-injected with
various amounts of unlabeled SRP RNA. (䊏) no unlabeled RNA;
(䊐) 16 fmol (32×) unlabeled RNA; (䊉) 80 fmol (160×); (䉭) 160
fmol (320×). (B) Extent of transport of nuclear injected [32P]SRP
RNA (0.5 fmol) co-injected with various amounts of unlabeled U6
snRNA. (■) No U6 RNA; (䊐) 160 fmol (320×) unlabeled U6;
(䊉) 900 fmol (1800×). (C) Rate of transport of SRP RNA as a
function of total SRP RNA injected (data from A). Note: the point
near the origin of the graph is not zero but 0.5 fmol.
temperature or WGA provided evidence that SRP RNA is
transported by a mediated process, implying that some specific
RNA sequence or structure interacts with a transport factor or
factors. Yet, locating the requisite interaction sites by random
sequence alterations in SRP RNA would be somewhat
laborious using a microinjection assay. Also, since SRP RNA
X.-P. He, N. Bataillé and H. M. Fried
is neither capped nor known to be modified in any other way,
there are no obvious unique molecular signatures that would
be targets for alteration. Recall, however, that SRP RNA
consists of two sequence elements, Alu and S, organized into
three RNA-protein functional domains (Fig. 1). Further, the
three activities of SRP are both biochemically and physically
separable (Gundelfinger et al., 1983; Siegel and Walter, 1986,
1988) making SRP a modular RNP whose individual domains
act independently. Being modular in the arrangement of its
cytoplasmic functions, we wondered whether one or the other
sequence domains of SRP RNA might not also harbor nuclear
transport information. Thus, we created templates for in vitro
synthesis of either the Alu or S portion of SRP RNA (see
Materials and Methods) for injection into oocytes. The Alu
sequence was created by deleting SRP RNA residues 98
through 251 from the original transcription plasmid; the joint
created by the deletion was predicted to create a five-nucleotide
loop at the point where the Alu sequence folds back on to itself
(see Fig. 1). An S transcript was produced by PCR-cloning the
desired sequence (residues 99 through 250) adjacent to the T7
promoter (see Materials and Methods). The secondary
structure predictions for the Alu and S transcripts derived from
these plasmids were exactly superimposable upon those
domains in SRP RNA.
Upon injection of the SRP S domain, most of the RNA was
rapidly degraded in the nucleus, although a small amount did
appear in the cytoplasm (Fig. 5). Thus, from nuclear injection
experiments we were unable to determine whether the S
domain has any role in SRP RNA transport. However, the Alu
transcript was stable when injected into the nucleus and in Fig.
6A it can be seen that the molecule accumulated in the
cytoplasm. The results in Fig. 6A also reveal that the extent of
cytoplasmic accumulation of the Alu transcript after 12 hours
was somewhat less than for intact SRP RNA in the same experiment. Indeed, from three such experiments in which Alu accumulation in the cytoplasm was quantitated we found that 32%
(s.d.=15%) was exported by 12 hours (compared to 80% for
SRP RNA; see above). Significantly, however, cytoplasmic
accumulation of the Alu domain was blocked at 4˚C (Fig. 6B)
and inhibition of Alu transport was reversed by shifting cells
incubated at 4˚C back to 19˚C. These results indicate that the
Alu transcript required facilitated transport to enter the
cytoplasm.
The Alu domain of SRP RNA competes with the
intact molecule for transport
Since the Alu domain of SRP RNA exhibited a property of
facilitated nuclear efflux (inhibition at low temperature), we
expected that, like full-sized SRP RNA, an excess of Alu RNA
injected into the nucleus should diminish export of SRP RNA.
The graph in Fig. 7 shows that increasing amounts of unlabeled
Alu domain RNA co-injected into the nucleus did compete with
labeled SRP RNA for export. However, as is clearly evident,
the Alu domain competed less efficiently than intact SRP RNA.
At an 1800-fold molar excess of Alu RNA (18 µM final concn),
A
SRP
RNA
Alu
RNA
B
19°C
4°C
4-19°C
908
SRP
RNA
Alu
RNA
Fig. 5. Fate of S domain RNA injected into the nucleus. Equimolar
amounts of S domain and intact SRP RNA were co-injected. Cells
were incubated at 19˚C and processed as described in Materials and
Methods. Incubation times, indicated by numbers, are in hours.
Fig. 6. Subcellular distribution of Alu RNA following nuclear
injection. (A) SRP and Alu RNAs were injected separately. (B) Left
half of panel: SRP and Alu RNAs were co-injected and cells were
incubated at 19˚C; right half of panel: RNAs were co-injected and
cells were incubated at 4˚C. The last two lanes on the far right are
samples from cells incubated at 4˚C for 6 hours and then transferred
to 19˚C for an additional 6 hours.
SRP RNA nuclear transport
909
100
% in cytoplasm
80
SRP
RNA
60
40
20
0
0
6
12
Time (hours)
Fig. 7. Alu domain RNA competes with SRP RNA for nuclear
export. [32P]SRP RNA (0.5 fmol) was injected into nuclei along with
various amounts of Alu domain RNA. (䊏) No Alu RNA; (䊐) 90 fmol
(180×) unlabeled Alu RNA; (䊉) 360 fmol (720×); (䉭) 900 fmol
(1800×).
there was a 50% reduction in SRP RNA transported by 6 hours;
only a 32-fold excess of SRP RNA was required to decrease
transport of the labeled RNA by about 33% (Fig. 4A). Lessefficient competition by the Alu domain RNA was not unexpected, since, as indicated above, this molecule was transported less efficiently to begin with. Furthermore, since we
cannot rule out the possibility that other parts of SRP RNA
also possess information that promotes transport, it is possible
that continued transport of SRP RNA in the presence of an
1800-fold excess of Alu RNA was the result of residual
transport mediated by the remainder of SRP RNA and associated trans-acting factors. Nonetheless, since the same amount
of U6 snRNA was without effect, we conclude that at least the
Alu domain plays a role in SRP RNA export.
Evidence for nuclear trans-acting factors specific
for the Alu domain
The foregoing experiments suggested that the Alu sequence
element of SRP RNA contributes to its facilitated transport.
Thus, we attempted to obtain evidence for interaction of the
Alu sequence with nuclear trans-acting factors, possibly
proteins that play a role in SRP RNA transport. Note that
transport could depend upon unidentified factors that interact
with the Alu domain in the nucleus or, by analogy to 5 S RNP
and ribosomal subunit transport (Guddat et al., 1990; Warner,
1990), perhaps proteins found in the cytoplasmic form of SRP
might migrate into the nucleus to assemble with the RNA as a
pre-requisite to transport. However, since the only proteins
known to bind the SRP RNA Alu sequence are SRP9 and
SRP14 (in the form of a heterodimer), we chose to focus on a
possible nuclear function of SRP9/14. As described below, we
created nucleotide alterations in SRP RNA that were expected
to prevent binding of these proteins and determined what effect
these mutations had on transport.
Hydroxyl radical cleavage has shown that four regions
(totaling about 35 bases) spanning residues 1-68 of SRP RNA
are in close contact with SRP9/14 (Strub et al., 1991). Also,
mutations at position 4 in Schizosaccharomyces pombe SRP
RNA, which corresponds to position 24 in vertebrate SRP
RNA, were reported to cause a conditional growth defect (Liao
et al., 1992). Since the G residue at position 24 is protected
Alu
RNA
Fig. 8. (A) A three-nucleotide substitution in the SRP Alu domain
reduces its accumulation in the cytoplasm. Wild-type SRP RNA was
co-injected with either wild-type (WT) Alu RNA (left half of panel)
or the same Alu RNA containing nucleotide substitutions at residues
24-26 (M1; right half of panel). (B) Mutant M1 RNA was injected
directly into the cytoplasm to assess its stability in that compartment.
from hydroxyl radical cleavage and is also 100% conserved in
eukaryotic SRP RNAs, it seemed likely that this base is critical
for protein binding. Thus, we created a G→A substitution at
position 24 and at the same time changed bases 25 and 26 from
UA to GG (located in the single-stranded loop between stems
3 and 4 in Fig. 1), thereby producing an AvrII restriction site
to facilitate screening for the desired mutant. Since other
portions of SRP RNA may contribute to its transport, we tested
the G24U25A26→ΑGG mutation within the Alu transcript
alone, so that the presence of other potential contributory
sequences would not obscure any defect. Fig. 8A shows that
following nuclear injection the cytoplasmic abundance of the
mutant Alu sequence, M1, was reduced compared to the wildtype sequence (in two experiments the percentage of the total
amount of recovered mutant RNA that reached the cytoplasm
was 1/2 to 2/3 of that of the total amount of recovered wildtype Alu in the cytoplasm); full-sized SRP RNA co-injected
into the same nuclei was efficiently exported. The reduced
cytoplasmic abundance of the mutant Alu RNA did not result
from degradation following nuclear export, since M1 was
stable when injected directly into the cytoplasm (Fig. 8B). Furthermore, we have determined that mutant M1 RNA is indeed
unable to bind SRP9/14 in vitro (C. Zwieb, H. Xiao-Ping and
H. Fried, unpublished results). On the surface then, these
results, i.e. lack of in vitro SRP9/14 binding and reduced cytoplasmic accumulation of the RNA, would suggest that association of SRP9/14 with the Alu domain in the nucleus confers
upon SRP RNA, in part, its ability to be transported. However,
it is not possible to state definitively that SRP9/14 is involved
in transport. This is because the M1 RNA was also slightly
more unstable in the nucleus than its wild-type counterpart
(compare in particular lanes labeled 12, N in Fig. 8A). The rate
of transport is concentration dependent but from the simple
910
X.-P. He, N. Bataillé and H. M. Fried
analysis presented here it cannot be determined to what degree
the slightly lower nuclear concentration of M1 affected its
transport. Nevertheless, whether slow degradation alone or
degradation plus an inherently impaired transport ability is the
basis for reduced cytoplasmic accumulation of the mutant
RNA, either condition is linked to an inability of SRP RNA to
bind SRP9/14. Thus while these observations do not prove a
role for SRP proteins in transport, they are at least consistent
with the suggestion that SRP proteins interact with SRP RNA
in the nucleus.
DISCUSSION
In this report we present evidence that SRP RNA injected into
the Xenopus oocyte nucleus is transported to the cytoplasm by
a facilitated, nuclear pore-mediated process and that at least
part of the information enabling facilitated transport is
contained within the Alu domain of the RNA molecule. The
conclusion of facilitated transport is based on the findings that:
(a) SRP RNA transport exhibited apparent saturation kinetics;
(b) transport was reversibly inhibited at low temperature; and
(c) transport was blocked specifically and completely by the
NPC inhibitor wheat germ agglutinin. The conclusion that the
Alu element of SRP RNA contributes to its transport is based
on the findings that: (a) the Alu domain RNA entered the
cytoplasm and its entry was reversibly inhibited at low temperature; and (b) the Alu domain competed with SRP RNA for
transport when both were present in the nucleus.
From competition experiments, we estimated an apparent
Km of 1 µM for transport of SRP RNA injected into oocytes.
This value is higher than that reported for tRNA transport (0.1
µM; Zasloff, 1983) using the same oocyte assay. SRP RNA
synthesized in vitro is known to exist in at least four different
conformations (Zwieb and Ullu, 1986). Thus, it is possible that
only a fraction of the injected RNA molecules was competent
to bind a requisite transport protein and, during the course of
the experiment, it was necessary for the remaining RNA to
rearrange into a form competent for protein binding and
transport. Since the Km calculation assumes a single conformational species, a smaller value would be obtained if indeed
only a fraction of the RNA was fit for export. We also
estimated a Vmax for transport of 3×107 molecules/minute per
nucleus; by comparison, the maximum rate for tRNA was
reported to be 19×108 molecules/minute (Zasloff, 1983), for
mRNA, 1.6×108 molecules/minute (Dargemont and Kühn,
1992), and for ribosomal subunits 0.17×108 to 3×108
molecules per minute (Bataillé et al., 1990). A smaller Vmax
indicates that transport of injected SRP RNA was less efficient
than that of these other molecules. Nonetheless, we believe that
the rather approximate kinetic parameters reported here are
physiologically relevant. Cells accumulate fewer SRP
complexes than either ribosomes or tRNA molecules and, since
transport of SRP RNA may in fact be equivalent to assembly
of SRP (see below), less efficient transport of SRP RNA may
be a manifestation of a small amount of unassembled SRP
proteins.
Competition experiments also revealed that the Alu domain
was most likely transported from the nucleus at least in part by
the same pathway or mechanism as SRP RNA, since the two
molecules competed with one another in that compartment. On
a molar basis the Alu transcript was less efficient that the fullsized molecule in competition with the radiolabeled tracer SRP
RNA. One possible explanation for this effect is that other
portions of SRP RNA may also contribute to transport, so that
titration of Alu-specific factors would be insufficient to block
transport of the full-sized molecule totally. We also noted that
the Alu domain by itself was transported inherently less efficiently compared to SRP RNA. Possibly, in intact SRP RNA
the Alu domain and associated factors act synergistically with
factors bound elsewhere in the RNA to promote transport.
Electron microscopic and other studies have shown that the
cytoplasmic form of SRP is a rod-shaped particle divided into
three domains, with the bulk of the RNA located at the two
termini (Andrews et al., 1987). If we assume for the moment
that transport occurs by way of the known SRP proteins interacting with SRP RNA in the nucleus (see below) it seems reasonable to assume that the nuclear conformation of SRP is
similar to its cytoplasmic form. Perhaps, then, the tri- or
bimodal nature of the particle is better, since it contains
multiple ‘RNP’ units, which can be recognized for transport,
while the artificial RNP that we presume assembles in the
nucleus upon Alu RNA injection is a less favorable, single RNP
monomer. Resolution of this question must await a better
understanding of the fundamental mechanism whereby the
transport apparatus, e.g. NPC, interacts with RNPs to translocate them to the cytoplasm.
While we do not know the identity of any trans-acting factor
involved in SRP RNA transport, we speculate that perhaps
some or all of the known SRP proteins migrate into the nucleus
to assemble with SRP RNA as a pre-requisite for transport.
Such a scheme would be analogous to biogenesis of ribosomal
subunits (Warner, 1990) and oocyte-specific 5 S rRNPs
(Guddat et al., 1990), in which proteins associated with the
cytoplasmic forms of the RNPs first enter the nucleus to
assemble with their cognate RNAs and in which such assembly
is required for transport. Although inconclusive, this idea is
also suggested by the results of the site-directed mutagenesis
experiment described above. Our findings that a threenucleotide substitution in the Alu domain prevented SRP9/14
binding in vitro and lead to destabilization, and possibly diminished transport, of SRP RNA in the nucleus suggests that
SRP9/14 might interact with SRP RNA in that compartment.
Obviously, more direct experiments are necessary to determine
whether SRP proteins have a role in SRP RNA nuclear
transport. Interestingly, the findings that the Alu portion of SRP
alone was inefficiently exported and that titration of Aluspecific transport factors was enough to slow but not prevent
SRP RNA suggest a biological rationale for the involvement
of both RNA domains in transport. Efficient release of SRP
RNA from the nucleus would depend on there being an
adequate supply of all SRP proteins, in much the same way as
the normal cytoplasmic accumulation of a ribosomal subunit
requires a full complement of each protein in the subunit
(Woolford and Warner, 1991).
Finally, besides SRP proteins, the only other protein known
to interact with SRP RNA is La (Chambers et al., 1983), a 50
kDa phosphoprotein involved in transcription termination by
RNA polymerase III (Gottlieb and Steitz, 1989), which is associated only transiently with newly synthesized transcripts. La
is found predominantly in the nucleus but there are indications
of its occurrence in the cytoplasm (Hendrick et al., 1981; J.
SRP RNA nuclear transport
Keene, personal communication), so it could be asked whether
La has a role in Pol III RNA export. For 5 S rRNA, La dissociation precedes binding of ribosomal protein L5 and export of
the RNA complex (Guddat et al., 1990). Using anti-La sera,
we did not detect association of SRP RNA with La in the
cytoplasm (unpublished results). Thus, were La to have a role
in SRP RNA export, it would have to dissociate immediately
upon the entry of RNA into the cytoplasm.
We thank Brian Kay (University of North Carolina) and Igor Dawid
(NIH) for providing the λ phage Xenopus genomic clone library, Elisabetta Ullu (Yale University) for providing the Xenopus SRP RNA
cDNA and the human SRP RNA gene clones, Jack Keene (Duke University) for a U6 snRNA gene, and Stephanie Harrison and Jörg
Grosshans for assistance in the initial stages of cloning and sequencing the Xenopus SRP RNA gene. We thank Christian Zwieb for
providing the phylogenetically based secondary structure model of
SRP RNA and for carrying out SRP RNA-protein binding assays,
Michael Terns (University of Wisconsin) for advice on construction
of the T7-U6 snRNA template, Don Gilbert (Indiana University) for
distributing LoopViewer and Peter Walter, Carl Feldherr, Michael
Caplow and Avrom Caplan for valuable discussions. This work was
supported by National Science Foundation research grant DMB8903639 to H.M.F.
REFERENCES
Allison, L. A., North, M. T., Murdoch, K. J., Romaniuk, P. J., Deschamps,
S. and le Maire, M. (1993). Structural requirements of 5 S rRNA for nuclear
transport, 7 S ribonucleoprotein particle assembly, and 60 S ribosomal
subunit assembly in Xenopus oocytes. Mol. Cell. Biol. 13, 6819-6831.
Andrews, D. W., Walter, P. and Ottensmeyer, F. P. (1987). Evidence for an
extended 7SL RNA structure in the signal recognition particle. EMBO J. 6,
3471-3477.
Baserga, S., Gilmore-Herbert, M. and Yang, X. W. (1992). Distinct
molecular signals for nuclear import of the nucleolar snRNA, U3. Genes
Dev. 6, 1120-1130.
Bataillé, N., Helser, T. and Fried, H. M. (1990). Cytoplasmic transport of
ribosomal subunits microinjected into the Xenopus laevis oocyte nucleus: a
generalized, facilitated process. J. Cell Biol. 111, 1571-1582.
Chambers, J. C., Kurilla, M. G. and Keene, J. D. (1983). Association
between the 7 S RNA and the lupus La protein varies among cell types. J.
Biol. Chem. 258, 11438-11441.
Dargemont, C. and Kühn, L. C. (1992). Export of mRNA from microinjected
nuclei of Xenopus laevis oocytes. J. Cell Biol. 118, 1-9.
Dingwall, C. and Laskey, R. A. (1991). Nuclear targeting sequences - a
consensus? Trends Biochem. Sci. 16, 478-481.
Feldherr, C. M. (1992). Overview of signal-mediated transport. In Nuclear
Trafficking (ed. C. M. Feldherr). Academic Press, San Diego, CA.
Fischer, U., Darzynkiewicz, E., Tahara, S. M., Dathan. N. A., Lührmann,
R. and Mattaj, I. W. (1991). Diversity in the signals required for nuclear
accumulation of U snRNPs and variety in the pathways of nuclear transport.
J. Cell Biol. 113, 705-714.
Fischer, U., Sumpter, V., Sekine, M., Satoh, T. and Lührmann, R. (1993).
Nucleo-cytoplasmic transport of U snRNPs: definition of a nuclear location
signal in the Sm core domain that binds a transport receptor independently of
the m3G cap. EMBO J. 12, 573-583.
Forbes, D. J. (1992). Structure and function of the nuclear pore complex. Annu.
Rev. Cell Biol. 8, 495-527.
Fried, H. M. (1992). Transport of ribosomal proteins and rRNA, tRNA, and
snRNA. In Nuclear Trafficking (ed. C. M. Feldherr), pp. 291-329. Academic
Press, San Diego, CA.
Gerace, L. (1993). Molecular trafficking across the nuclear pore complex.
Curr. Opin. Cell Biol. 4, 637-645.
Gottlieb, E. and Steitz, J. A. (1989). Function of the mammalian La protein:
evidence for its action in transcription termination by RNA polymerase III.
EMBO J. 8, 851-861.
Guddat, U., Bakken, A. H. and Pieler, T. (1990). Protein-mediated nuclear
911
export of RNA: 5 S rRNA containing small RNPs in Xenopus oocytes. Cell
60, 619-628.
Gundelfinger, E. D., Krause, E., Melli, M. and Dobberstein, B. (1983). The
organization of the 7SL RNA in the signal recognition particle. Nucl. Acids
Res. 11, 7363-7374.
Gurdon, J. B. and Wickens, M. P. (1983). The use of Xenopus oocytes for the
expression of cloned genes. Meth. Enzymol. 101, 370-386.
Hendrick, J. P., Wolin, S. L., Rinke, J., Lerner, M. R. and Steitz, J. A.
(1981). Ro scRNPs are a subclass of La RNPs. Further characterization of the
Ro and La small ribonucleoproteins from uninfected mammalian cells. Mol.
Cell. Biol. 1, 1138-1149.
Imamoto, N., Matsuoka, Y., Kurihara, T., Kohno, K., Miyagi, M.,
Sakiyama, F., Okada, Y., Tsunasawa, S. and Yoneda, Y. (1992).
Antibodies against 70-kD heat shock cognate protein inhibit mediated
nuclear import of karyophilic proteins. J. Cell Biol. 119, 1047-1061.
Izaurralde, E. and Mattaj, I. W. (1992). Transport of RNA between nucleus
and cytoplasm. Semin. Cell Biol. 3, 279-288.
Izaurralde, E., Stepinski, J., Darzynkiewicz, E. and Mattaj, I. W. (1992). A
cap binding protein that may mediate nuclear export of RNA polymerase IItranscribed RNAs. J. Cell Biol. 118, 1287-1295.
Kunkel, T. A., Bebenek, K. and McClary, J. (1991). Efficient site-directed
mutagenesis using uracil-containing DNA. Meth. Enzymol. 204, 125-139.
Liao, X., Selinger, D., Althoff, S., Chiang, A., Hamilton, D., Ma, M. and
Wise, J. A. (1992). Random mutagenesis of Schizosaccharomyces pombe
SRP RNA: lethal and conditional lesions cluster in presumptive protein
binding sites. Nucl. Acids Res. 20, 1607-1615.
Michaud, N. and Goldfarb, D. (1992). Microinjected U snRNAs are imported
to oocyte nuclei via the nuclear pore complex by three distinguishable
targeting pathways. J. Cell Biol. 116, 851-861.
Neuman de Vegvar, H. E. and Dahlberg, J. E. (1990). Nucleocytoplasmic
transport and processing of small nuclear RNA precursors. Mol. Cell. Biol.
10, 3365-3375.
Piñol-Roma, S. and Dreyfuss, G. (1993). hnRNP proteins: localization and
transport between the nucleus and the cytoplasm. Trends Cell Biol. 3, 151155.
Rapoport, T. A. (1991). Protein transport across the endoplasmic reticulum
membrane: facts, models, mysteries. FASEB J. 5, 2792-2798.
Richter, J. D. and Standiford, D. (1992). Structure and regulation of nuclear
localization signals. In Nuclear Trafficking (ed. C. M. Feldherr), pp. 90-120.
Academic Press, San Diego, CA.
Riedel, N. and Fasold, H. (1992). Role of mRNA transport in
posttranscriptional control of gene expression. In Nuclear Trafficking (ed. C.
M. Feldherr), pp. 231-290. Academic Press, San Diego, CA.
Rinke, J. and Steitz, J. A. (1982). Precursor molecules of both human 5 S
ribosomal RNA and transfer RNAs are bound a cellular protein reactive with
anti-La lupus antibodies. Cell 29, 149-159.
Siegel, V. and Walter, P. (1986). Removal of the Alu structural domain from
signal recognition particle leaves its protein translocation activity intact.
Nature 320, 81-84.
Siegel, V. and Walter, P. (1988). Functional dissection of the signal
recognition particle. Trends Biochem. Sci. 13, 314-316.
Starr, C. M. and Hanover, J. A. (1992). Structure and function of nuclear pore
glycoproteins. In Nuclear Trafficking (ed. C. M. Feldherr), pp. 175-201.
Academic Press, San Diego, CA.
Strub, K., Moss, J. and Walter, P. (1991). Binding sites of the 9- and 14kilodalton heterodimeric protein subunit of the signal recognition particle
(SRP) are contained exclusively in the Alu domain of SRP RNA and contain
a sequence motif that is conserved in evolution. Mol. Cell. Biol. 11, 39493959.
Terns, M. P., Lund, E. and Dahlberg, J. E. (1992). 3′-end-dependent
formation of U6 small nuclear ribonucleoprotein particles in Xenopus laevis
oocyte nuclei. Mol. Cell. Biol. 12, 3032-3040.
Tobian, J. A., Drinkard, L. and Zasloff, M. (1985). tRNA nuclear transport:
defining the critical regions of human tRNAiMet by point mutagenesis. Cell
43, 415-422.
Ullu, E., Murphy, S. and Melli, M. (1982). Human 7SL RNA consists of a 140
nucleotide middle-repetitive sequence inserted in an Alu sequence. Cell 29,
195-202.
Ullu, E. and Tschudi, C. (1984). Alu sequences are processed 7SL RNA
genes. Nature 312, 171-172.
Ullu, E. and Weiner, A. M. (1984). Human genes and pseudogenes for the 7SL
RNA component of signal recognition particle. EMBO J. 3, 3303-3310.
Vankan, P., McGuigan, C. and Mattaj, I. W. (1990). Domains of U4 and U6
912
X.-P. He, N. Bataillé and H. M. Fried
snRNAs required for snRNP assembly and splicing complementation in
Xenopus oocytes. EMBO J. 9, 3397-3404.
Vieira, J. and Messing, J. (1987). Production of single-stranded plasmid
DNA. Meth. Enzymol. 153, 3-11.
Walter, P. and Lingappa, V. R. (1986). Mechanism of protein translocation
across the endoplasmic reticulum membrane. Annu. Rev. Cell. Biol. 2, 499516.
Warner, J. R. (1990). The nucleolus and ribosome formation. Curr. Opin.
Cell. Biol. 2, 521-527.
Weitzmann, C. J., Cunningham, P. R. and Ofengand, J. (1990). Cloning, in
vitro transcription, and biological activity of Escherichia coli 23S ribosomal
RNA. Nucl. Acids. Res. 18, 3515-3520.
Wersig, C., Guddat, U., Pieler, T. and Bindereif, A. (1992). Assembly and
nuclear transport of the U4 and U4/U6 snRNPs. Exp. Cell Res. 199, 373-377.
Woolford, J. L. and Warner, J. R. (1991). The ribosome and its synthesis. In
The Molecular and Cellular Biology of the Yeast Saccharomyces: Genome
Dynamics, Protein Synthesis and Energetics, vol. 1 (ed. J. R. Broach, J. R.
Pringle and E. W. Jones), pp. 587-626. Cold Spring Harbor Laboratory Press,
New York.
Yamasaki, L. and Lanford, R. E. (1992). Nuclear transport receptors:
specificity amid diversity. In Nuclear Trafficking (ed. C. M. Feldherr), pp.
121-174. Academic Press, San Diego, CA.
Zasloff, M. (1983). tRNA transport from the nucleus in a eukaryotic cell:
carrier-mediated translocation process. Proc. Nat. Acad. Sci. USA 80, 64366440.
Zuker, M. and Stiegler, P. (1981). Optimal computer folding of large RNA
sequences using thermodynamics and auxiliary information. Nucl. Acids
Res. 9, 133-148.
Zwieb, C. and Ullu, E. (1986). Identification of dynamic sequences in the
central domain of 7SL RNA. Nucl. Acids Res. 14, 4639-4657.
Zwieb, C. (1991). Interaction of protein SRP19 with signal recognition particle
RNA lacking individual RNA-helices. Nucl. Acids Res. 19, 2955-2960.
SRP RNA nuclear transport
(Received 15 October 1993 - Accepted 12 January 1994)
913