Cell, Vol. 119, 491–502, November 12, 2004, Copyright 2004 by Cell Press
She2p Is a Novel RNA Binding Protein
with a Basic Helical Hairpin Motif
viewed in Tekotte and Davis , Vale ). Selec-
the budding of Saccharomyces cerevisiae represents
one of the best-characterized examples of mRNA trans-
location (reviewed in Chartrand et al. , Darzacq et
Upon nutrient deprivation, diploid S. cerevisiae cells
undergo meiosis, which results in four haploid spores,
two pairs belonging to each opposite mating type. Only
cells of opposing mating type are able to mate and form
a diploid cell. To ensure that cells within a population
are equally distributed between the two mating types,
and, thus, capable of reproduction, the mother cell but
not the daughter cell switches its mating type after each
et al. ). Yeast regulates mating-type switching by
inhibiting expression of HO endonuclease in daughter
cells via specific transcriptional repression of HO endo-
nuclease (Ash1p). Exclusive daughter cell expression of
Ash1p is achieved by active localization of ASH1 mRNA
to the bud tip of the daughter cell (Long et al., 1997;
Takizawa et al., 1997). At the bud tip, the mRNA is trans-
lated into Ash1p, which represses transcription of the
inducer of mating-type switch (i.e., the HO endonucle-
ase; Bobola et al., 1996; Sil and Herskowitz, 1996).
Genetic screening has identified three core compo-
nents of the ASH1mRNA translocation complex, includ-
ing She1p/Myo4p, She2p, and She3p (Jansen et al.,
1996). She1p/Myo4p is a type V unconventional myosin
that conveys ASH1 mRNA-containing RNPs along the
actin network (Bertrand et al., 1998; Bobola et al., 1996;
Jansen et al., 1996; Long et al., 1997; Mu ¨nchow et al.,
1999; Reck-Peterson et al., 2001; Takizawa et al., 1997).
She2p facilitates ASH1 mRNA binding by recognizing
four independent zip code elements (Bo ¨hl et al., 2000;
Chartrand et al., 1999, 2002; Gonzalez et al., 1999; Long
et al., 2000). Myo4p and She2p are connected by the
adaptor protein She3p (Bo ¨hl et al., 2000; Long et al.,
2000; Mu ¨nchow et al., 1999;Takizawa and Vale, 2000). A
recently published study showed that the Pumilio family
protein Puf6p, which is not deemed to be part of the
core translocation machinery, associates with the ASH1
mRNP complex to repress translation of ASH1 mRNA
during translocation (Gu et al., 2004).
As in the paradigm common to metazoans (reviewed
in Dreyfuss et al. , Kloc et al. ), She2p may
escort ASH1 mRNA from the nucleus, during transloca-
tion as part of a translationally silent mRNP, until ASH1
Bo ¨hl et al., 2000; Kruse et al., 2002; Takizawa and Vale,
2000). After translocation into the cytoplasm, the adap-
tor protein She3p binds the She2:ASH1 mRNA complex.
Together with She1p/Myo4p, the complex assembles
into a larger mRNP (Long et al., 2000; Takizawa and
Vale, 2000). Thereafter, the ASH1 mRNP is transported
into the bud and translation is activated.
terizing a large number of RNA binding motifs (reviewed
in Dreyfuss et al. ), unsuccessful attempts to de-
tect similarity between the amino acid sequence of
Dierk Niessing,1,4Stefan Hu ¨ttelmaier,3
Daniel Zenklusen,3Robert H. Singer,3
and Stephen K. Burley1,2,4,*
1Laboratories of Molecular Biophysics and
2Howard Hughes Medical Institute
The Rockefeller University
1230 York Avenue
New York, New York 10021
3Department of Anatomy and Structural Biology
Albert Einstein College of Medicine
1300 Morris Park Avenue
Bronx, New York 10461
Selective transport of mRNAs in ribonucleoprotein
particles (mRNP) ensures asymmetric distribution of
information within and among eukaryotic cells. Actin-
dependent transport of ASH1 mRNA in yeast repre-
sents one of the best-characterized examples of mRNP
translocation. Formation of the ASH1 mRNP requires
recognition of zip code elements by the RNA binding
protein She2p. We determined the X-ray structure of
She2p at 1.95 A˚resolution. She2p is a member of a
previously unknown class of nucleic acid binding pro-
teins, composed of a single globular domain with a
five? helixbundlethatforms asymmetrichomodimer.
After demonstrating potent, dimer-dependent RNA
binding in vitro, we mapped the RNA binding surface
of She2p to a basic helical hairpin in vitro and in vivo
and present a mechanism for mRNA-dependent initia-
tion of ASH1 mRNP complex assembly.
In eukaryotic cells, mRNA translocation is required for
asymmetric propagation of information (reviewed in Fa-
rina and Singer , Kloc et al. , Lopez de Here-
dia and Jansen , St Johnston ) and repre-
sents a central mechanism for such diverse functions
et al. ), budding of yeast (reviewed in Chartrand
et al. , Darzacq et al. ), development of body
Wickens et al. ), synaptic plasticity (reviewed in
Steward and Worley ), and basal-apical cell polar-
ity (reviewed in Lopez de Heredia and Jansen ,
translocation are usually bound by specialized proteins
and packaged into larger messenger ribonucleoprotein
particles (mRNP), also known as translocons, or loca-
somes (Bertrand et al., 1998). Within such translocons
port along cellular actin or microtubule networks (re-
4Present address: Structural GenomiX, Inc., 10505 Roselle Street,
San Diego, California 92121.
She2p and the amino acid sequences of known RNA
binding proteins (Bo ¨hl et al., 2000; Gonsalvez et al.,
any previously identified class of RNA binding proteins.
(Pieper et al., 2004) to predict its three-dimensional
structure were similarly unsuccessful (data not shown).
dimensional structure of She2p at 1.95 A˚ resolution,
which revealed that the She2p polypeptide chain folds
into a single globular domain consisting of a bundle of
five antiparallel ? helices with a small additional helix
protruding at right angles from the middle of the bundle.
She2p forms a stable dimer that is required for function
in vitro and in vivo. Systematic comparison with pre-
viously determined structures available in the Protein
similarity between She2p and known nucleic acid bind-
ing proteins. We determined that She2p binds its target
surface of She2p in vitro and in vivo, and present a
molecular mechanism for RNA-dependent initiation of
the assembly of ASH1 mRNP complex.
dimer are symmetric, with their pentacle bundles super-
imposing with a root-mean-square deviation (rmsd) of
0.43 A˚ (for 141 common ? carbon atom pairs). The
atomic model of the first monomer lacks residues 184–
191 and 238–239, whereas, in the second monomer,
(residues 1–5 and 240–246 were excluded from the ex-
pression construct, see Experimental Procedures).
with buried solvent-accessible surface area of 663 A˚2/
dimer, which is significantly smaller than interfaces re-
ported for most stable oligomers (Conte et al., 1999;
Dasgupta et al., 1997). Size exclusion chromatography
lytical equilibrium ultracentrifugation revealed the pre-
dominant She2p-oligomerization state at concentra-
tions of 8–32 ?M to be dimeric (data not shown; see
Experimental Procedures). Point mutations predicted to
not result in a change in size exclusion chromatography
behavior and did not affect RNA binding (data not
shown). We conclude, therefore, that the observed
She2p dimer-dimer contact in the crystal is almost cer-
tainly not of physiologic importance.
She2p Represents a Novel Protein Fold
Systematic comparison of She2p with structures pre-
viously deposited in the PDB using the DALI server
(Holm and Sander, 1998) revealed no detectable similar-
ity to known nucleic acid binding proteins. Instead, the
? helix core of She2p bears some resemblance to the
central domain of fumarase/aspartase homologous su-
marase C (PDB ID, 1fur-A; Z score, 8.6; rmsd ? 4.2 A˚
for 169 ? carbon pairs), adenylosuccinate lyase (PDB
ID, 1c3c-A; Z score, 7.7; rmsd ? 3.9 A˚for 170 ? carbon
pairs), the eye lens protein ?-2 crystallin (PDB ID,
1auw-A; Z score, 7.7; rmsd ? 4.0 A˚for 172 ? carbon
pairs), and ?-carboxy-cis, cis-muconate cycloisomer-
ase (PDB ID, 1q5n; Z score, 7.6; rmsd ? 4.3 A˚for 178
? carbon pairs). Although class II fumarases like fumar-
ase C oligomerize via their central domain (Weaver et
al., 1995), their subunit interactions are distinct from
those observed for the She2p dimer.
Amino acid sequence identities within the fumarase/
aspartase homologous superfamily tend to be high
(Woods et al., 1986) (e.g., E. coli and human fumarase C
are 59% identical), but none of the proteins enumerated
above are more than 12% identical to She2p. Sequence
identities between She2p and fumarase/aspartase su-
perfamily homologs from S. cerevisiae do not exceed
11%. A signature sequence motif found within the cen-
tral domain of fumarase/aspartase superfamily mem-
bers, which consists of a highly conserved, functionally
important region (Estevez et al., 2002), is also lacking
in She2p. Unlike She2p, proteins of the fumarase/aspar-
tase superfamily consist of three domains, and their
ity has not been reported for any member of the fumar-
In summary, similarities between She2p and the cen-
tral domains of fumarase/aspartase superfamily mem-
nant full-length She2p expressed in E. coli. Results of
limited proteolysis combined with matrix-assisted laser
desorption/ionization mass spectrometry (MALDI-MS)
suggested that N- and C-terminal segments of the
She2ppolypeptide chaindonotadopt stablesecondary
structures. A total of 20 different N- and/or C-terminal
truncations of She2p with various additional amino acid
alterations were expressed in E. coli, purified to homo-
geneity, and used in crystallization trials (see Experi-
mental Procedures). Diffraction-quality crystals were
obtained from a truncated form of She2p (residues 6–
239) with all four Cys residues (Cys14, Cys68, Cys106,
and Cys180) changed to serine to overcome oxidation
during purification. The X-ray crystal structure was de-
termined via multiple isomorphous replacement (Table
1; Ke, 1997). The final refinement model was obtained
at 1.95 A˚resolution, with excellent crystallographic sta-
tistics (Rwork? 19.1%, Rfree? 24.1%; Bru ¨nger, 1992) and
stereochemistry (see Table 1 and Experimental Proce-
tion and refinement).
She2p Is a Symmetric Homodimer
The She2p polypeptide chain folds into a single domain
consisting of a bundle of five antiparallel ? helices
(monomer dimensions: 70 A˚? 30 A˚? 45 A˚; Figure 1),
which are arrayed like a pentacle around a central axis
and contribute stacking aromatic side chains to the hy-
drophobic core. A small additional ? helix protrudes at
right angles from the middle of the bundle. Within the
crystal, She2p forms a noncrystallographic dimer (di-
mensions: 70 A˚? 54 A˚? 62 A˚), stabilized by significant
buried solvent-accessible surface area (?2015 A˚2/
monomer) and several hydrophobic amino acids con-
tributing to the dimer interface. The two halves of the
Structural and Functional Studies of She2p
Figure 1. Structure of the She2p Homodimer
(A) Stereoview of the She2p homodimer with each monomer in blue or green (PyMOL, DeLano Scientific, CA). The second, less complete
monomer was substituted with a superimposed first monomer to provide a fuller depiction of the She2p homodimer. Vertical line labeled with
“2-fold” indicates the axis of 2-fold noncrystallographic symmetry relating the halves of the homodimer. Arrows on the green subunit denote
(B) GRASP (Nicholls et al., 1993) surface representation of the chemical properties of the solvent-accessible surface of She2p, using a water
probe radius of 1.4 A˚. The surface electrostatic potential is color coded red and blue, representing electrostatic potentials between ? ?14
to ? ?14 kBT, where kBis the Boltzmann constant and T is the temperature. Orientation is identical to (A).
(C) Stereoview of (A) rotated 90? around the vertical axis. Dotted lines represent an eight amino acid gap in the final refinement model.
(D) GRASP surface representation of (B) rotated 90? around the vertical axis.
(E) Stereoview of (A) rotated 90? around the horizontal axis, with the top of She2p turned toward the reader. Rainbow coloration of one
monomer follows the polypeptide chain, with the N terminus in blue and the C terminus in red. The second monomer is colored brown.
(F) GRASP surface representation of (B) rotated 90? around the horizontal axis, with the hydrophobic upper surface of She2p facing the reader.
Figure 2. Structural Prediction and Analysis of She2p RNA Binding
(A) Secondary structure prediction (Mfold; Zucker, 1998) for the ASH1 mRNA E3 zip code element as previously described by Chartrand et
(B) Filter binding results with She2p (wild-type) and the E3 zip code element of ASH1 mRNA (Kd? 210 nM ? 40). Since C-terminal residues
of She2p are required for efficient RNA binding (data not shown), no binding studies were performed with the N- and C-terminally truncated
version of She2p used for crystallization.
bers are strictly limited to the overall structural arrange-
ment of their respective five ? helix core, with no
ogy. We conclude, therefore, that She2P represents a
novel protein fold of distinct evolutionary origin that is
not currently represented in the PDB.
She2p Binds RNA via a Positively
Charged Surface Feature
Analysis of the solvent-accessible molecular surface
(Nicholls et al., 1991) of She2p revealed an overall nega-
tive electrostatic potential, with the exception of a dis-
tinct surface area with positive electrostatic potential
tified RNA binding residues (Gonsalvez et al., 2003) to
the surface feature with positive electrostatic potential
(Figure 3A, marked with asterisk; Figure 3B). In addition,
we produced and analyzed various mutant forms of
She2p binding to ASH1 RNA. Only those amino acids
located within the surface feature with positive electro-
static potential were required for RNA binding (Figures
3A and 3B).
Together, these results define a basic helical hairpin
RNA binding motif consisting of two antiparallel ? heli-
ces separated by a loop (Figure 3C). Within this local
tertiary structural feature, we find the following se-
quence motif: N-x(6)-[KR]-[KR]-loop-x(7)-R-x(4)-K-x(2)-
cates either a lysine or an arginine, and “loop” marks
the position where two ? helices are connected by a
loop. In Figure 1A, this region of the polypeptide chain
is located on the right (green) monomer in the upper
half of the two foremost ? helices (marked with arrows)
and for the left (blue) monomer in the corresponding
position on the rear surface of the homodimer (not
Within the same dimer subunit (A) and adjacent to
the positively charged RNA binding motif, we sought
additional residues necessary for nucleic acid binding.
None of the following A subunit mutations (His33→Ala,
Arg49→Ala, Asp67→Ala, Thr235→Tyr) affected RNA
binding. Mutation of the following residues from the B
subunit of the homodimer (Gln119→Ala, Lys123→Ala,
mer RNA binding motif, does not affect RNA binding
thebasichelical hairpinservesasthe primaryanchorfor
ASH1 mRNA to each half of the homodimer.
She2p Binds ASH1 RNA with nM Affinity
The core function of She2p is to bind ASH1 mRNA and
initiate assembly of the ASH1 mRNP (Bo ¨hl et al., 2000;
Chartrand et al., 1999; Gonzalez et al., 1999; Long et
al., 2000). To date, no quantitative information on She2p
RNA binding is available, raising the question as to
whether or not RNA binding by She2p suffices to recruit
ASH1 mRNA into mRNP complexes in vivo. We deter-
mined the affinity of wild-type She2p for the single zip
code element E3 of ASH1 mRNA using filter binding
assays (see Figure 2A for the results of RNA secondary
structure prediction). She2p binds its target RNA with
A calculation of She2p abundance in vivo (4070 mole-
cules per cell, Ghaemmaghami et al. ; cell volume,
30 ?m3, Tyson et al. ) suggests that She2p is pres-
is localized in yeast cells, the effective concentration
is almost certainly higher. Although this calculation is
unlikely to provide an accurate estimate of She2p con-
centration in cells, it suffices to support our conclusion
that She2p is present at concentrations at or above its
RNA binding Kd. She2p should, therefore, be able to
recruit target mRNA into mRNP complexes in vivo. It is
remarkable that She3p association with the She2p:
ASH1 mRNA complex further increases the affinity of
the growing complex for ASH1 mRNA (Bo ¨hl et al., 2000).
This effect would serve to ensure recruitment of ASH1
mRNA into a productive mRNP complex at She2p con-
centrations less than the Kdand to enhance the effi-
ciency of ASH1 mRNA translocation in mating-type
Structural and Functional Studies of She2p
Figure 3. She2p Dimer Binds RNA via a Conserved, Basic Helical Hairpin
(A) Table summarizing RNA binding properties of mutant forms of She2p from filter binding experiments (this study) and yeast three-hybrid
studies (asterisk; Gonsalvez et al., 2003): “wt,” “?/?,” and “–” denote wild-type, reduced, and not-detectable RNA binding affinity, respectively.
The mutant form She2p Arg63→Ala was previously described as being defective in RNA binding (Gonsalvez et al., 2003). In our hands, She2p
Arg63→Ala was insoluble. Instead, we generated She2p Arg63→Asp and determined that this soluble mutant form of She2p does not bind RNA.
(B) Enlarged view of GRASP surface representation shown in Figure 1B, rotated 10? around the horizontal axis and 15? counterclockwise
around the vertical axis. Labels shown in green with italics overlay residues required for RNA binding. Black labels overlay the sites of
mutations that do not affect RNA binding (see Figure 3A).
(C) Schematic of the basic helical hairpin of She2p viewed from above and slightly rotated from the orientation shown in Figure 1E. The
polypeptide backbone for residues 25–68 is colored brown, with the exception of green color-coded basic residues for which the side chains
are also depicted.
(D) Sequence alignment of She2p. Secondary structural elements were obtained from the X-ray structure. Dashes denote residues that were
either disordered or excluded from the recombinant form of She2p used for crystallization. Full, half-filled, and empty circles denote sites of
mutations resulting in wild-type, reduced, and negligible RNA binding, respectively (see also Figures 3A and 3B). “*” denotes amino acids
that are required for ASH1 mRNA binding (Figures 5C and 5E), interaction with She3p (Leu130; Gonsalvez et al., 2003), and translocation of
ASH1 RNP in vivo (Leu130; Gonsalvez et al., 2003), which are located outside the RNA binding surface feature with positive electrostatic
potential. Sequence similarity is encoded by a yellow-to-green color gradient, representing 40%–100% identity (BLOSUM62).
Sequence Conservation in She2p Reveals
Regions of Functional Importance
We used the RNA binding sequence motif found within
the basic helical hairpin to search for similar motifs in
proteins of 18 other yeasts (http://www.yeastgenome.
org; partially and completely available genomes). Pat-
tern matching detected nine independent amino acid
sequences, all of which proved to be related to S. cere-
visiae She2p (e value cutoff for full-length She2p ? 1 ?
10?30). Among these matches were She2p orthologs
found in two distantly related yeasts, Kluyveromyces
waltii (Kellis et al., 2004) and Ashbya gossypii (Dietrich
et al., 2004), which also encode homologs of all known
core proteins comprising the ASH1 mRNP complex.
We used ClustalX (Thompson et al., 1997) to perform
an amino acid sequence alignment of She2p from
S. cerevisiae and its orthologs from S. castellii, K. waltii,
and A. gossypii (Figure 3D). The resulting sequence
evisiae She2p of 51.4% (S. castellii), 42.6% (K. waltii),
and 40.2% (A. gossypii). The remarkable level of se-
quence identity and the pattern of amino acid differ-
ences among these yeast strains allow us to conclude
that all known She2p proteins share the same three-
dimensional structure (Sander and Schneider, 1991).
Comparative protein structure modeling (Pieper et al.,
2004) with our structure of She2p from S. cerevisiae
was used by Sali and coworkers to produce rea-
sonably accurate structural models of all known She2p
homologs. The resulting structural models are avail-
able to academic users via http://salilab.org/modeller/
Of the seven She2p RNA binding residues identified
in this work and Gonsalvez et al. (2003) (Figures 3A and
3B), six are identical in all species, and one residue
demonstrates a conservative exchange in S. castellii
and A. gossypii (Arg44→Lys). This conservative ex-
change has been shown in S. cerevisiae to functionally
substitute for arginine (Figure 3A) (Gonsalvez et al.,
basic helical hairpin motif supports our contention that
residues of functional importance are highly conserved
among She2p family members. As more genomic se-
quences of yeasts become available, additional con-
us to discern other functionally important surface re-
We attempted to extend our analysis of candidate
proteins bearing a S. cerevisiae She2p-like basic helical
hairpin motif (Figures 3A–3C) by searching a protein
sequence database that includes proteins from non-
yeast species (Swissprot; http://motif.genome.jp; vari-
able sequence length for the loop). The search yielded
seven candidate proteins (50S ribosomal protein L13P,
Sulfolobus solfataricus; transcription factor ICE1, Arabi-
viral core protein VP6, bluetongue virus; CbbX protein
homolog, Guillardia theta; mannitol-1-phosphate-5-
dehydrogenase, Mycoplasma mycoides; and transient
receptor potential channel, Caenorhabditis elegans), of
which four are thought to bind nucleic acids. Two of
these knownnucleic acid bindingproteins, the50S ribo-
somal protein L13P and the transcription factor ICE1,
are members of known structural classes that do not
resemble She2p and do not in fact contain basic helical
hairpins. No structural information is available for any
of the five remaining candidate proteins. Secondary
structure predictions were used to look for evidence of
matching the basic helical hairpin sequence motif from
She2p. At best, the results of these analyses were am-
biguous. Given our current, somewhat limited, knowl-
edge of the universe of three-dimensional protein struc-
tures, it appears that the basic helical hairpin motif
occurs only within She2p proteins and their homologs
She2p, we selected two single amino acid substitutions
(Cys68→Tyr and Ser120→Tyr, respectively) intended to
produce unfavorable steric clashes at the dimer inter-
face and thereby prevent She2p subunit joining. Analy-
tography indicated a significantly lower molecular
weight for both mutant proteins as compared to wild-
type She2p (data not shown). Measurement of RNA
binding affinities revealed that both mutant proteins
were unable to bind ASH1 RNA efficiently (Figure 4A).
(Circular dichroism [CD] spectroscopy documented
identical secondary structural composition for mutant
and wild-type She2p [data not shown], making it ex-
tremely unlikely that loss of RNA binding activity is an
artifact of mutant She2p unfolding or misfolding.) Our
results, therefore, document that She2p dimerization is
required for ASH1 mRNA binding.
She2p Function In Vivo
In yeast cells, She2p binding to ASH1 mRNA is required
for assembly of ASH1 mRNP and translocation to the
daughter cell tip. Mutant forms of She2p that do not
bind ASH1 RNA in vitro should not, therefore, translo-
cate ASH1 mRNA to the bud tip.
In order to validate our in vitro findings, a SHE2 dele-
tion strain was transformed with plasmids expressing
either wild-type or various she2 mutant alleles under
tion properties of different she2 alleles were tested with
fluorescent in situ hybridization using Cy3-labeled
probes againstASH1 mRNA.In astrain expressingwild-
type She2p, ASH1 mRNA is localized to the bud tip of
the daughter cell, whereas, in a she2 deletion yeast
strain, ASH1 mRNA is found distributed throughout
mother and daughter cell (Figures 4B and 4G and Fig-
ures 4C and 4H, respectively). In vitro, the mutant form
She2p Lys60→Ala is not able to bind RNA (Figure 3A).
Confirming the requirement of She2p RNA binding for
in vivo function, She2p Lys60→Ala does not localize
ASH1 mRNA in vivo (Figures 4D and 4I). The mutation
Asp67→Ala, which lies outside of the She2p RNA bind-
ing domain and supports RNA binding in vitro, yields a
mutantformof She2pthatiscompetent forASH1mRNA
localization to the bud tip of the daughter cell (Fig-
ures 4E and 4J). Finally, one of the two mutations
(Ser120→Tyr) that interfere with dimerization and RNA
binding (Figure 4A) yields a form of She2p that also
cannot localize ASH1 mRNA in vivo (Figures 4F and 4K).
In summary, the yeast translocation studies support
our previous findings by showing a direct correlation
between in vitro She2p-RNA binding efficiency and
ASH1 mRNA translocation. Our in vivo yeast studies
mation for translocation, which itself is necessary for
RNA binding activity in vitro.
A Conserved, Uncharged Surface Feature Is
Required for She2p Function
tionalimportance, weanalyzeda surfacerepresentation
of the She2p-sequence alignment (Figures 3D and 5A–
5D). In addition to the previously defined RNA binding
surface feature (Figures 5A–5C; enclosed with dotted
Dimerization of She2p Is Required for RNA binding
Since She2p forms a stable dimer in solution, we sought
to establish whether or not She2p dimerization is re-
quired for RNA binding. Guided by our structure of
Structural and Functional Studies of She2p
Figure 4. Defective She2p Function Interferes with RNA Binding and ASH1 mRNA Localization at the Bud of the Daughter Cell
(A) Filter binding experiments with dimerization-defective forms of She2p and the E3 zip code element of ASH1 mRNA. Both dimer mutant
forms of She2p fail to bind RNA efficiently.
(B–F) In situ hybridization with Cy3-labeled oligonucleotides against ASH1 mRNA after transformation of SHE2 deletion strain with constructs
expressing either the wild-type SHE2 allele (B), plasmid without SHE2 (C), or various mutant she2 alleles (D–F). ASH1 mRNA is depicted in
red and DAPI nuclear staining in blue. Images depict representative cell staining.
(G–K) Images show identical views to those depicted in (B)–(F) acquired with Normarski optics.
lines), the small protuberant helix and its connection to
the five ? helix bundle is conserved. Surface features
at the side (Figure 5B) and the bottom of She2p (Figure
5D) show little and no conservation, respectively. The
upper surface region of She2p, which is largely un-
charged (Figure 1F), contains hydrophobic amino acids,
and consists of surface features from both dimer sub-
units, shows a high degree of sequence conservation
In order to characterize the functional relevance of
the conserved upper surface region of the She2p dimer,
we generated two independent amino acid exchanges
within this area (Thr47→Tyr and Leu130→Tyr, respec-
binant She2p Thr47→Tyr and She2p Leu130→Tyr
showed reduced RNA binding activities for both mutant
proteins (Figure 5E), demonstrating a functional require-
ment for the upper surface feature of She2p. Compara-
tiveanalysis ofwild-type She2pand She2pLeu130→Tyr
by CD spectroscopy documented no significant differ-
ences in secondary structural composition (data not
It is remarkable that Golsalvez et al. (2003) isolated a
yeast strain bearing a Leu130→Ser mutation in She2p
that does not translocate ASH1 mRNA in yeast (see
Supplemental Material, Table S1, in Gonsalvez et al.
). In addition, they showed that recombinant
Leu130→Ser mutant She2p is unable to bind She3p. In
contrast, other RNA binding-defective mutants de-
scribed by Gonsalvez et al. (2003) (Figure 3A, marked
with asterisk) retain She3p binding. Our findings with
Leu130→Ser mutants (Gonsalvez et al., 2003), indicate
that the upper, uncharged surface region of She2p is
required for ASH1 mRNA binding and translocation
ity involves the binding of one RNA zip code element
resulting in a ratio of dimer to zip code element of 1:2
to zip code element ratio of 1:1 (Figure 5G).
we performed filter binding experiments with near stoi-
chiometric ratios of She2p and ASH1 mRNA E3 zip code
one She2p dimer binds to one RNA zip code element.
Our studies demonstrate that She2p represents an RNA
binding protein of novel three-dimensional structure.
tion is required for ASH1 mRNA binding in vitro and
mRNA translocation in yeast. We showed that She2p
binds ASH1 mRNA in vitro with sufficient affinity to sup-
port effective recruitment into functional mRNP com-
plexes in vivo and identified a helical hairpin RNA bind-
ing motif, which is displayed on a conserved surface
feature with positive electrostatic potential. Moreover,
our results, together with structure-based interpretation
of previously published findings, demonstrate that a
The E3 zip code element, which was used in filter
binding experiments, is thought to contain an extended
stem-loop secondary structure (Figure 2A). Binding
studies with mutated E3 zip code elements showed that
the predicted RNA loop region at the tip of the putative
double-stranded stem is dispensable (Chartrand et al.,
1999). Mutations altering the secondary structure of the
putative double-stranded stem result in a loss of She2p
binding, whereas She2p binding is not impaired when
bases on both complementary strands are exchanged,
without destabilizing the predicted secondary structure
One She2p Dimer Binds to One ASH1 mRNA
Zip Code Element
She2p dimer formation is required for ASH1 mRNA zip
code element binding, suggesting two possible modes
Figure 5. Sequence Conservation Identifies Functional Surface Features
(A) GRASP surface representation of sequence conservation as shown in Figure 3D with identical orientation as in Figure 1B. The dotted line
encloses surface residues responsible for RNA binding.
(B) GRASP surface representation of (A) with orientation as in Figure 1D.
(C) GRASP surface representation of (A) with orientation as in Figure 1F. “T” and “L” denote Thr47 and Leu130 on both dimer subunits (see
(D) GRASP surface representation of (A), with orientation of (C) rotated 180? around the horizontal axis (bottom view instead of view on upper
surface of She2p).
(E) Filter binding experiments with mutant forms of She2p that contain amino acid exchanges in its upper, uncharged surface region (see
also [C]), and the E3 zip code element of ASH1 mRNA. She2p Thr47→Tyr and She2p Leu130→Tyr show strong and modest reduction in
affinity for ASH1 mRNA, respectively.
(F and G) Schematics depicting potential modes of She2p dimer binding to the four independent zip code elements of ASH1 mRNA, as
discussed in the main body of the paper.
(H) Filter binding experiments with 100 nM wild-type She2p and increasing, near-stoichiometric amounts of E3 zip code element of ASH1
mRNA (up to 2.5-fold excess). The projected maximum bound RNA concentration of 64 nM indicates a ratio of She2p dimer to RNA zip code
element of 1:1.28, suggesting that one She2p dimer binds one zip code element (see Figure 5G).
(Bo ¨hl et al., 2000; Chartrand et al., 1999, 2002; Gonzalez
et al., 1999). These results together with the findings
of our structural and mechanistic studies suggest that
She2p binds ASH1 mRNA by recognizing RNA second-
ary structure via interactions with its residues displayed
by the basic helical hairpin (Figure 3C) and on the con-
served hydrophobic feature on the upper surface of the
homodimer depicted in Figure 1F.
Since She2p homodimer has two RNA binding motifs,
it is intriguing to speculate that one She2p homodimer
binds two independent zip code elements. Two She2p
homodimers would suffice for efficient binding to all
four known zip code elements of one ASH1 transcript
In contrast to this model, our filter binding experi-
ments under near-stoichiometric concentrations indi-
cate that one zip code element binds around one She2p
homodimer, requiring four She2p homodimers to bind
all four zip code elements of ASH1 mRNA (Figures 5G
and 5H). In addition to residues from the RNA binding
surface feature with positive electrostatic potential, mu-
tations on the upper, uncharged surface feature of
She2p impair RNA binding. These findings suggest that
one ASH1 mRNA zip code element arches over the up-
neously to the basic helical hairpins of both monomers.
To date, the smallest identified ASH1 zip code element
has a length of 77 bases (ASH1 U element) and is be-
lieved to form a double-stranded stem-loop structure
(Bo ¨hl et al., 2000; Gonzalez et al., 1999). Manual align-
ment of modeled dsRNA to the RNA binding surface
features of She2p dimer indicates that a 77 base dsRNA
would be large enough to arch over the upper surface
region of She2p and interact with both positively
Structural and Functional Studies of She2p
linker region, unless stated otherwise.
Initial, poor-quality crystals were obtained by hanging drop vapor
diffusion against 0.1 M Na-HEPES (pH 7.5), 1.5 M LiSO4, and 2 mM
TCEP at20?C. Analyses of She2p,both from solution andfrom these
disulfide bond formation. Alkylation of She2-P with N-ethyl-malei-
mide or iodoacetamide under anaerobic conditions combined with
electrospray ionization mass spectrometry (ESI-MS) (Vinkemeier et
al., 1996) suggested that all four cysteines in She2p are solvent
exposed and capable of forming intermolecular disulfide bonds.
Cys14, Cys68, Cys106, and Cys180 were simultaneously mutated
to serines. Based on results from limited proteolysis combined with
MALDI-MS and ESI-MS (data not shown; Cohen and Chait, 2001),
we generated multiple N- and C-terminal truncations of She2p and
tested a total of 20 truncated forms of She2p in crystallization trials.
Diffraction-quality crystals of She2p (residues 6–239, with Cys14,
Cys68, Cys106, and Cys180 mutated to serines) were obtained at
a protein concentrationof 2–2.5 mg/ml via hangingdrop vapor diffu-
sion against 50–100 mM Tris(pH 8.5), 160–200 mM LiSO4, 20%–25%
PEG 4000 (w/v), and 2.4 mM ?-octyl-glucoside at 20?C. Diamond-
shaped crystals appeared after 2–4 days, which grew to a typical
size of 0.05 ? 0.1 ? 0.1 mm3. Crystal cryoprotection was achieved
by adding ethylene glycol to a final concentration of 30% (v/v). No
crystals were obtained from Se-Met-substituted She2p. Experimen-
tal phases were measured from crystals soaked in crystallization
solution with various heavy metals present at concentrations rang-
ing from 1 to 100 mM for 1–4 days.
charged RNA binding surface features simultaneously
(data not shown).
This mode of RNA binding, in which a She2p dimer
uses both subunits to bind one zip code element (Figure
for RNA binding. Moreover, it is consistent with the re-
II and V, which showed that ca. four motor proteins are
required to move cargo continuously (Reck-Peterson et
al., 2001). Four myosin V motors are also thought to
be required for processivity of ASH1 mRNA transport
(Chartrand et al., 2002).
ASH1 mRNP complex assembly requires She3p join-
ing the She2p:RNA complex. We speculate that She3p
interacts with a heterologous surface region, consisting
of She2p and dsRNA. The requirement for a heterolo-
gous surface region for She3p binding could explain
why She3p interacts efficiently with She2p:RNA and not
with She2p alone (Long et al., 2000) and why She3p
stabilizes She2p-RNA interaction (Bo ¨hl et al., 2000).
Such an mRNA-dependent assembly mechanism of the
ASH1 mRNP may ensure that the translocation complex
does notengage in afutile transportation ofcore factors
into the bud cell in the absence of cargo mRNA.
A surface representation of sequence conservation
and the upper, uncharged surface region of She2p are
conservation and functional importance (Figure 3D), we
speculate that one or both of these conserved regions
participates in formation of a (heterologous) surface for
She3p binding. The uncharged region of She2p with its
hydrophobic amino acids represents an unusual feature
on the surface of an otherwise charged protein (Figures
1B, 1D, and 1F). Interactions between proteins and nu-
cleic acids are generally mediated by charged amino
acids (Nadassy et al., 1999; Nagai, 1996), whereas pro-
tein-protein interactions can involve both uncharged,
hydrophobic surface areas and charged amino acids
possible that, in addition to its requirement for efficient
ASH1 mRNA binding, the uncharged surface region of
She2p mediates interactions with She3p. A previously
published report that She2p bearing a mutation in the
uncharged surface region (Leu130→Ser) fails to interact
with She3p (see Supplemental Material, Table S1, in
Gonsalvez et al. ) is consistent with this specula-
tion.Future structuralandfunctional studieswill benec-
essary to understand precisely how She2p binds ASH1
mRNA and how interactions with She3p influence the
formation of a functional localization complex.
Data Collection, Structure Determination, and Refinement
All diffraction data were collected at the SGX-CAT insertion device
tory) under standard cryogenic conditions and processed and
scaled using DENZO and SCALEPACK (Otwinowski and Minor,
1997). A total of 24 datasets was collected (Table 1; for all crystals
examined in the different experiments, space group C2; unit cell
a ? 97.2 A˚, b ? 103.7 A˚, c ? 57.7 A˚, ? ? 110.5?, ? ?1%; two
She2 molecules/asymmetric unit). Ta6Br12cluster binding sites were
located using direct methods (SnB, Weeks and Miller ), and
four independent IrCl3sites were obtained by Fourier difference
syntheses from initial Ta6Br12-derived phases. Definitive experimen-
tal phases were obtained via multiple isomorphous replacement
using CCP4/MLPHARE (Collaborative Computational Project, 1994)
(three Ta6Br12sites [one data set]; four independent IrCl3sites [three
datasets]) for two independent datasets at 1.95 A˚(Table 1, Nativea)
and 2.35 A˚(Table 1, Nativeb) resolution, respectively. After density
sitymapsthat showedinterpretabledensityfeatures incomplemen-
tary regions of each map. After merging the electron density maps,
ten large ?-helical features were immediately recognizable within
the asymmetric unit. The merged electron density map revealed the
noncrystallographic symmetry (NCS) operator relating dimer sub-
units. After NCS averaging and further density modification, the
secondary structural features for most of the pentacle core of the
She2p homodimer were readily apparent. Automated electron den-
sity map interpretation with MAID (Levitt, 2001) yielded atomic mod-
els encompassing an average of 103 residues per molecule (206
residues per asymmetric unit). Iterative rounds of refinement and
model building gave the current refinement model, with the working
Rfactor ? 19.1% and an Rfreevalue of 24.1% (residues 6–183 and
192–237 for the first molecule in the asymmetric unit and residues
13–76, 92–177, and 204–237 for the second molecule in the asym-
metric unit, plus 302 isolated density features modeled as water
?) combinations in the most favorable region of the Ramachandran
diagram and an overall G factor of ?0.1, which is significantly better
than average for structures obtained at ?2 A˚resolution.
Protein Preparation, Quality Assurance, and Crystallization
cDNAs encoding various forms of She2-P (full-length, 246 residues)
S-transferase fusion proteins in the vector pGEX6p-1. Each ex-
rose chromatography, followed by subtractive glutathione-Sepha-
rose and anion exchange chromatographies and a final size
exclusion chromatography step. Protein purity and quality were as-
sessed by gel electrophoresis, size exclusion chromatography,
MALDI-MS (Cohen and Chait, 2001), and dynamic light scattering
(Ferre ´-D’Amare ´ and Burley, 1997). Purified proteins contained the
Sedimentation equilibrium measurements were made to determine
the oligomerization state of She2p in solution. Data were collected
at the Keck Biophysics Facility at Northwestern University (Evans-
ton, IL) with an XL-A analytical ultracentrifuge (Beckmann, Fullerton,
Table 1. Data Statistics
Data Sets Nativea
IrCl3#1 IrCl3#2IrCl3# 3
Completeness (%), overall/outer shell
?I/?(I)?, overall/outer shell
Rsym(I) (%), overall/outer shell
Overall figure of merit
Phasing power (iso) (Nativea)
Phasing Power (iso) (Nativeb)
aCrystal soaked in 1mM K2PtI6.
bCrystal soaked in 1 mM (NH4)2WS4. Rsym(I) ? ?|I ? ?I?|/?I, where I ? observed intensity and ?I? ? average intensity obtained from multiple
observations of symmetry related reflections. Phasing power ? rmsd (|FH|/E), with |FH| ? heavy atom structure factor amplitude, and E ?
residual lack of closure. Data set “Nativea” was used for refinement and iterative model building.
UTHSC San Antonio, TX). She2p full-length (Cys14→Ser, Cys68→
Ser, Cys106→Ser, and Cys180→Ser) in 150 mM KCl and 10 mM
Tris/HCl (pH 7.5) was examined at concentrations of 7.9 ?M; 15.9
?M; 23.8 ?M; and 31.7 ?M (corresponding to optical densities at
280 nm, 0.2, 0.4, 0.6, and 0.8) at 17,000; 25,500; 34,000; 42,500; and
50,000 rpm. The major oligomerization state of She2p was dimeric,
constants (data not shown). For stoichiometric RNA binding experi-
ments, 100 nM wild-type She2p was titrated against near-stoichio-
metric concentrations of E3 zip code element up to a total of 2.5-
fold excess. The binding buffer used in this study contained 1 mM
MgOAc. The transcript used for RNA binding studies consists of
the E3 zip code element (capital letters) and vector sequence (small
Circular Dichroism Spectroscopy
Data were collected and processed with a Circular Dichroism Spec-
trometer Model 215 (Aviv Instruments). She2 full-length (wild-type,
HCl (pH 7.5), and 0.1 mM TCEP were examined at a concentration of
7.9 ?M (optical density at 280 nm, 0.2) between 192 and 250 nM
(step size, 1 nM).
Use of the Advanced Photon Source was supported by the U.S.
Department of Energy, Office of Science, Office of Basic Energy
Sciences, under Contract No. W-31-109-Eng-38. Use of the SGX
Collaborative Access Team (SGX-CAT) beamline facilities at Sector
31 of the Advanced Photon Source was provided by Structural Ge-
nomiX, Inc., which constructed and operates the facility. We are
grateful to Dr. J.B. Bonanno for technical support and advice during
data analyses and structure determination; to Xiuhua Meng for ex-
cellent support with filter binding experiments; to Dr. A.K. Padyana
for help during structure determination; to Dr. S. Wasserman for
assistance with X-ray diffraction data collection; to Dr. Brian Noland
for assistance with Circular Dichroism measurements; and to K.
Bain, M. Buchanan, D. Phanstiel, Dr. X. Gao, M. Maletich, M. Riedy,
and Dr. X. Zhao for technical support. We also thank Drs. S. Antony-
samy, R.C. Deo, C. Groft, C. Kissinger, H.A. Lewis, G. Louie, J.
Marcotrigiano, A.K. Padyana, F. Park, and X. Zhao for many useful
discussions. We are thankful to J. Kosh and Dr. C. Stamper at
Keck Biophysics Facility of Northwestern University (IL) and Dr. B.
Demeler at The University of Texas Health Science Center at San
Antonio (TX) for help with analytical ultracentrifugation. D.N. was
Planck Society. S.K.B. was an investigator in the Howard Hughes
Medical Institute. This work was supported by NIGMS grant
GM61262 (S.K.B.) and NIGMS grant GM57071 (R.H.S.).
Yeast Expression of SHE2 Mutants
A SHE2 expression cassette for full-length She2p (wild-type and
mutants described in Figures 4B–4K) was constructed by ligating
two PCR products into XhoI endonuclease sites. The PCR products
were amplified from genomic DNA spanning 500 bp upstream and
downstream of the SHE2 open reading frame and adding an EcoRI
restriction site after the start codon and an HA epitope to the 3?
end of the coding sequence, respectively. This expression cassette
was ligated at KpnI and XmaI endonuclease sites into pRS414,
creating thepRS414 SHE2HA cassette.Wild-type andmutant SHE2
were amplified by PCR and ligated at EcoRI and XhoI endonuclease
into a SHE2 deletion strain, and expression levels were estimated
via Western blotting using an anti-HA antibody.
In Situ Hybridization
Yeast cells were grown to mid-log phase and processed for in situ
hybridization as described in Long et al. (2000) using a pool of Cy3-
labeled oligonucleotides against ASH1 mRNA.
RNA Binding Experiments
Radiolabeled E3 zip code element of ASH1 mRNA was added to
serial dilutions of full-length She2p (wild-type or mutant forms) in
200mM KoAc,50 mMTris (pH7.4), 5mM MgOAc,1 mMDTT, and30
ug/ml yeast tRNA, as described in Farina et al. (2003). Radioactivity
retained on a nitrocellulose filter was measured. Best fit analyses
and calculation of equilibrium dissociation constants from a plot
of the fraction of bound RNA versus protein concentration were
performed using the Langmuir Isotherm, assuming a single binding
site. For all She2p variants, three independent experiments for each
of at least six different dilutions were performed.
Comparative RNA binding experiments for wild-type She2p were
also performed with 0.25, 0.5, 1, and 5 mM MgOAc, showing no
significant differences in their respective equilibrium dissociation
Received: July 19, 2004
Revised: September 20, 2004
Accepted: September 30, 2004
Published: November 11, 2004
and Long,R.M. (1998).Localization ofASH1 mRNAparticles inliving
yeast. Mol. Cell 2, 437–445.
Bobola, N., Jansen, R.P., Shin, T.H., and Nasmyth, K. (1996). Asym-
metric accumulation of Ash1p in postanaphase nuclei depends on
Structural and Functional Studies of She2p
a myosin and restricts yeast mating-type switching to mother cells.
Cell 84, 699–709.
Holm, L., and Sander, C. (1998). Touring protein fold space with
Dali/FSSP. Nucleic Acids Res. 26, 316–319.
Jansen, R.P., Dowzer, C., Michaelis, C., Galova, M., and Nasmyth,
K. (1996). Mother cell-specific HO expression in budding yeast de-
pends on the unconventional myosin myo4p and other cytoplasmic
proteins. Cell 84, 687–697.
Ke, H. (1997). Overview of Isomorphous Replacement Phasing. In
Methods in Enzymology, C.W. Carter and R.M. Sweet, eds. (San
Diego, CA: Academic Press), pp. 448–461.
Kellis, M.,Birren, B.W., andLander, E.S. (2004). Proofand evolution-
ary analysis of ancient genome duplication in the yeast Saccharo-
myces cerevisiae. Nature 428, 617–624.
Kloc, M., Zearfoss, N.R., and Etkin, L.D. (2002). Mechanisms of
subcellular mRNA localization. Cell 108, 533–544.
Kruse, C., Jaedicke, A., Beaudouin, J., Bo ¨hl, F., Ferring, D., Gu ¨ttler,
T., Ellenberg, J., andJansen, R.P. (2002). Ribonucleoprotein-depen-
dent localization of the yeast class V myosin Myo4p. J. Cell Biol.
Laskowski, R.J., MacArthur, M.W., Moss, D.S., and Thornton, J.M.
(1993). PROCHECK: a program to check stereochemical quality of
protein structures. J. Appl. Crystallogr. 26, 283–290.
Levitt, D.G. (2001). A new software routine that automates the fitting
of protein X-ray crystallographic electron-density maps. Acta Crys-
tallogr. D. Biol. Crystallogr. 57, 1013–1019.
Long, R.M., Singer, R.H., Meng, X., Gonzalez, I., Nasmyth, K., and
Jansen, R.P. (1997). Mating type switching in yeast controlled by
asymmetric localization of ASH1 mRNA. Science 277, 383–387.
Long, R.M., Gu, W., Lorimer, E., Singer, R.H., and Chartrand, P.
(2000). She2p is a novel RNA-binding protein that recruits the
Myo4p-She3p complex to ASH1 mRNA. EMBO J. 19, 6592–6601.
Lopez de Heredia, M., and Jansen, R.P. (2004). mRNA localization
and the cytoskeleton. Curr. Opin. Cell Biol. 16, 80–85.
Mohr, E., and Richter, D. (2001). Messenger RNA on the move:
implications for cell polarity. Int. J. Biochem. Cell Biol. 33, 669–679.
Mu ¨nchow, S., Sauter, C., and Jansen, R.P. (1999). Association of
the class V myosin Myo4p with a localised messenger RNA in bud-
ding yeast depends on She proteins. J. Cell Sci. 112, 1511–1518.
Nadassy, K., Wodak, S.J., and Janin, J. (1999). Structural features of
protein-nucleic acid recognition sites. Biochemistry 38, 1999–2017.
Nagai, K. (1996). RNA-protein complexes. Curr. Opin. Struct. Biol.
Nicholls, A., Sharp, K., and Honig, B. (1991). Protein folding and
association: insights from the interfacial and thermodynamic prop-
erties of hydrocarbons. Proteins 11, 281–296.
Nicholls, A., Bharadwaj, R., and Honig, B. (1993). GRASP: Graphical
Representation and Analysis of Surface Properties. Biophys. J.
Bo ¨hl, F., Kruse, C., Frank, A., Ferring, D., and Jansen, R.P. (2000).
She2p, a novel RNA-binding protein tethers ASH1 mRNA to the
Myo4p myosin motor via She3p. EMBO J. 19, 5514–5524.
Bru ¨nger, A.T. (1992). Free R value: a novel statistical quantity for
assessing the accuracy of crystal structures. Nature 355, 472–475.
tural elements required for the localization of ASH1 mRNA and of
a green fluorescent protein reporter particle in vivo. Curr. Biol. 9,
Chartrand, P., Singer, R.H., and Long, R.M. (2001). RNP localization
and transport in yeast. Annu. Rev. Cell Dev. Biol. 17, 297–310.
Chartrand, P., Meng, X.H., Hu ¨ttelmaier, S., Donato, D., and Singer,
tion of translation by localization elements in the mRNA. Mol. Cell
Cohen, S.L., and Chait, B.T. (2001). Mass spectrometry as a tool for
protein crystallography. Annu. Rev. Biophys. Biomol. Struct. 30,
Collaborative Computational Project. (1994). The CCP4 suite: pro-
grams for protein crystallography. Acta Crystallogr. D. Biol. Crys-
tallogr. 50, 760–763.
Conte, L.L., Chothia, C., and Janin, J. (1999). The atomic structure
of protein-protein recognition sites. J. Mol. Biol. 285, 2177–2198.
Darzacq, X., Powrie, E., Gu, W., Singer, R.H., and Zenklusen, D.
ation in yeast. Curr. Opin. Microbiol. 6, 614–620.
Dasgupta, S., Iyer, G.H., Bryant, S.H., Lawrence, C.E., and Bell, J.A.
(1997). Extent and nature of contacts between protein molecules in
crystal lattices and between subunits of protein oligomers. Proteins
Dietrich, F.S., Voegeli, S., Brachat, S., Lerch, A., Gates, K., Steiner,
S., Mohr, C., Pohlmann, R., Luedi, P., Choi, S., et al. (2004). The
Ashbya gossypii genome as a tool for mapping the ancient Sacchar-
omyces cerevisiae genome. Science 304, 304–307.
Ding, B., Itaya, A., and Qi, Y. (2003). Symplasmic protein and RNA
traffic: regulatory points and regulatory factors. Curr. Opin. Plant
Biol. 6, 596–602.
Dreyfuss, G., Kim, V.N., and Kataoka, N. (2002). Messenger-RNA-
binding proteins and the messages they carry. Nat. Rev. Mol. Cell
Biol. 3, 195–205.
Estevez, M., Skarda, J., Spencer, J., Banaszak, L., and Weaver, T.M.
(2002). X-ray crystallographic and kinetic correlation of a clinically
observed human fumarase mutation. Protein Sci. 11, 1552–1557.
Farina, K.L., and Singer, R.H. (2002). The nuclear connection in RNA
transport and localization. Trends Cell Biol. 12, 466–472.
Farina, K.L., Hu ¨ttelmaier, S., Musunuru, K., Darnell, R., and Singer,
ization, granule formation, and cytoskeletal attachment. J. Cell Biol.
Otwinowski,Z., andMinor, W.(1997). Processingof X-raydiffraction
data collected in oscillation mode. In Methods of Enzymology, C.W.
Carter and R.M. Sweet, eds. (San Diego, CA: Academic Press),
Ferre ´-D’Amare ´, A.R., and Burley, S.K. (1997). Dynamic light scatter-
ing in evaluating crystallizability of macromolecules. Methods Enzy-
mol. 276, 157–166.
Pieper, U., Eswar, N., Braberg, H., Madhusudhan, M.S., Davis, F.P.,
Stuart, A.C., Mirkovic, N., Rossi, A., Marti-Renom, M.A., Fiser, A., et
al. (2004). MODBASE, a database of annotated comparative protein
structure models, and associated resources. Nucleic Acids Res.
Ghaemmaghami, S., Huh, W.K., Bower, K., Howson, R.W., Belle,
A., Dephoure, N., O’Shea, E.K., and Weissman, J.S. (2003). Global
analysis of protein expression in yeast. Nature 425, 737–741.
Reck-Peterson, S.L., Tyska, M.J., Novick, P.J., and Mooseker, M.S.
(2001). The yeast class V myosins, Myo2p and Myo4p, are nonpro-
cessive actin-based motors. J. Cell Biol. 153, 1121–1126.
J.R., and Long, R.M. (2003). RNA-protein interactions promote
asymmetric sorting of the ASH1 mRNA ribonucleoprotein complex.
RNA 9, 1383–1399.
protein structures and the structural meaning of sequence align-
ment. Proteins 9, 56–58.
Gonzalez, I., Buonomo, S.B., Nasmyth, K., and von Ahsen, U. (1999).
ASH1 mRNA localization in yeast involves multiple secondary struc-
tural elements and Ash1 protein translation. Curr. Biol. 9, 337–340.
Sil, A., and Herskowitz, I. (1996). Identification of asymmetrically
localized determinant, Ash1p, required for lineage-specific tran-
scription of the yeast HO gene. Cell 84, 711–722.
Gu, W., Deng, Y., Zenklusen, D., and Singer, R.H. (2004). A new
yeast PUF family protein, Puf6p, represses ASH1 mRNA translation
and is required for its localization. Genes Dev. 18, 1452–1465.
Steward, O., and Worley, P. (2001). Localization of mRNAs at synap-
tic sites on dendrites. Results Probl. Cell Differ. 34, 1–26.
St Johnston, D. (1995). The intracellular localization of messenger
RNAs. Cell 81, 161–170.
Takizawa, P.A., and Vale, R.D. (2000). The myosin motor, Myo4p,
binds Ash1 mRNA via the adapter protein, She3p. Proc. Natl. Acad.
Sci. USA 97, 5273–5278.
Takizawa, P.A., Sil, A., Swedlow, J.R., Herskowitz, I., and Vale, R.D.
(1997). Actin-dependent localization of an RNA encoding a cell-fate
determinant in yeast. Nature 389, 90–93.
Tekotte, H., and Davis, I. (2002). Intracellular mRNA localization:
motors move messages. Trends Genet. 18, 636–642.
Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., and
Higgins, D.G. (1997). The CLUSTAL_X windows interface: flexible
strategies for multiple sequence alignment aided by quality analysis
tools. Nucleic Acids Res. 25, 4876–4882.
Tyson, C.B., Lord, P.G., and Wheals, A.E. (1979). Dependency of
size of Saccharomyces cerevisiae cells on growth rate. J. Bacteriol.
port. Cell 112, 467–480.
Vinkemeier, U., Cohen, S.L., Moarefi, I., Chait, B.T., Kuriyan, J., and
Darnell, J.E.,Jr. (1996).DNA bindingof invitro activated Stat1alpha,
Stat1 beta and truncated Stat1: interaction between NH2-terminal
domains stabilizes binding of two dimers to tandem DNA sites.
EMBO J. 15, 5616–5626.
Weaver, T.M., Levitt, D.G., Donnelly, M.I., Stevens, P.P., and Banas-
zak, L.J. (1995). The multisubunit active site of fumarase C from
Escherichia coli. Nat. Struct. Biol. 2, 654–662.
Weeks, C.M., and Miller, R. (1999). The design and implementation
of SnB V2.0. J. Appl. Crystallogr. 32, 120–124.
Wickens, M., Goodwin, E.B., Kimble, J., Strickland, S., and Hentze,
M.W. (2000). Translational Control of Developmental Decisions. In
Translational Control of Gene Expression, N. Sonenberg, J.W.B.
Hershey, and M.B. Mathews, eds. (Cold Spring Harbor, NY: Cold
Spring Harbor Laboratory Press), pp. 295–370.
tural and functional relationships between fumarase and aspartase.
Nucleotide sequences of the fumarase (fumC) and aspartase (aspA)
genes of Escherichia coli K12. Biochem. J. 237, 547–557.
Zucker, M. (1998). On finding all suboptimal foldings of an RNA
molecule. Science 244, 48–52.
The refined atomic coordinates and X-ray structure factors for
She2p have been deposited in the Protein Data Bank under ID