|
|
|
|||||||
|
|||||||||
J Biol Chem, Vol. 273, Issue 51, 34603-34610, December 18, 1998
From the Requirements for intron recognition during
pre-mRNA splicing in plants differ from those in vertebrates
and yeast. Plant introns contain neither conserved branch points nor
distinct 3' splice site-proximal polypyrimidine tracts characteristic
of the yeast and vertebrate introns, respectively. However, they are
strongly enriched in U residues throughout the intron, property
essential for splicing. To understand the roles of different sequence
elements in splicing, we are characterizing proteins involved in intron recognition in plants. In this work we show that Nicotiana
plumbaginifolia, a dicotyledonous plant, contains two genes
encoding different homologs of the large 50-65-kDa subunit of the
polypyrimidine tract binding factor U2AF, characterized previously in
animals and Schizosaccharomyces pombe. Both plant
U2AF65 isoforms, referred to as NpU2AF65a and
NpU2AF65b, support splicing of an adenovirus pre-mRNA
in HeLa cell nuclear extracts depleted of the endogenous U2AF factor.
Both proteins interact with RNA fragments containing plant introns and
show affinity for poly(U) and, to a lesser extend, poly(C) and poly(G). The branch point or the 3' splice site regions do not contribute significantly to intron recognition by NpU2AF65. The
existence of multiple isoforms of U2AF may be quite general in plants
because two genes expressing U2AF65 have been identified in
Arabidopsis, and different isoforms of the U2AF small
subunit are expressed in rice.
Accurate and orderly splicing of nuclear pre-mRNAs requires
that exon and intron sequences are effectively distinguished from each
other and that matching 5' and 3' splice sites (5'ss and 3'ss)1 are selected with
precision and juxtaposed before the catalytic steps. Several short
sequence elements in pre-mRNA contribute to intron recognition and
splice site selection. The 5'ss and 3'ss represent the most universally
conserved and functionally important elements. The 5'ss, containing the
nearly invariant GU dinucleotide, is recognized through base pairing by
the U1 small nuclear ribonucleoprotein (snRNP) early in spliceosome
assembly and by other snRNPs and factors at later stages of the
reaction. The precise function of the conserved AG at the 3' intron
border is less well understood (reviewed in Refs. 1-4). Although the role of splice sites appears to be similar in different eukaryotes, the
relative contribution of other signals differs significantly between
different organisms. In vertebrates and insects, the characteristic and
functionally important feature of introns is a polypyrimidine tract,
which is located between the branch point and the 3'ss. The pyrimidine
tract is recognized by the heterodimeric protein U2AF (U2 snRNP
auxiliary factor) early in spliceosome assembly, and this interaction
is essential for positioning the U2 snRNP at the branch site, the
sequence of which is highly degenerate in metazoa (5-7). Another
factor, mBBP/SF1, interacting directly with the branch point region,
may also assist U2 snRNP to establish a base pairing interaction with
the branch site (8-10). The distinguishing feature of pre-mRNA
introns in the yeast Saccharomyces cerevisiae is a highly
conserved branch site sequence, UACUAAC (reviewed in Refs. 2 and 11).
Early in spliceosome assembly, UACUAAC is specifically recognized by
the branch point-bridging protein BBP, the yeast ortholog of mBBP/SF1
(8-10). Most introns in S. cerevisiae lack 3'ss-proximal
polypyrimidine tracts, and specific sequences positioned 3' of the
branch site do not contribute to the first step of splicing despite the
fact that a factor, Mud2p, sharing some structural and functional
similarity with the large subunit of the metazoan U2AF, is present in
yeast (8, 12). Some yeast introns contain uridine stretches upstream of
the 3'ss, but these sequences function during the second step of
splicing (13, 14).
Requirements for intron recognition in higher plants differ from those
in yeast and vertebrates. Plant introns contain neither conserved
branch point sequences nor distinct 3'ss-proximal polypyrimidine tracts
similar to those present in yeast and vertebrate introns, respectively
(reviewed in Refs. 4, 15, and 16). With respect to both consensus and a
position relative to the 3'ss, the branch sites in plant introns
resemble those found in metazoa (17, 18). A characteristic feature of
plant introns is their UA or U richness, a property essential for
splicing (19-25). The requirement for UA richness is particularly
strong in dicotyledonous plants, which have been studied in the most
detail. It is not known how, at the molecular level, the UA-rich
elements, usually distributed throughout the whole length of introns,
contribute to splicing. Experiments performed with UA-deficient
synthetic introns have shown that stimulation of splicing by U-rich
sequences does not depend on their position in the intron (19, 26).
Short U-rich elements such as UUUUUAU or its multimers activated
splicing whether inserted near the 5'ss or in the middle of the intron
(hence upstream of the branch site) or in the vicinity of the 3'ss.
Other studies have shown that UA-rich elements play an important role
in the definition of intron borders; the 3' and 5' splice sites
preferentially selected for splicing were those present at the
transition regions from the UA-rich to GC-rich sequences (Refs. 21, 22,
and 27; reviewed in Ref. 15). Mutational analysis of the 3'ss-proximal UA-rich elements indicated that, as in the case of splicing activation discussed above (26), U rather than A residues play a key role in
intron border definition (21); preference for U residues, although less
pronounced, was also seen upon mutagenesis of the 5'ss-proximal and
internal UA elements (28, 29). U-richness was also found to be an
essential feature for splicing in maize (Ref. 25 and references
therein). Stimulation of splicing by UA-rich elements from upstream
locations in the intron and their importance for selection of both 5'
and 3' splice sites make it unlikely that these elements simply act as
plant counterparts to the metazoan polypyrimidine tracts. The latter
are always located between the branch point and the 3'ss and usually
function only in the definition of the 3' intron border (reviewed in
Refs. 1 and 3).
To obtain more insight into the function of UA-rich elements in plant
splicing, we are characterizing proteins likely to be involved in
intron recognition in plants. A group of ~50-kDa proteins that
interact with plant introns and show specificity for oligouridylates has been previously identified in nuclear extracts of Nicotiana plumbaginifolia (26); their role in splicing in vivo
remains to be established. In this work we characterize two isoforms of the large subunit of the splicing factor U2AF in N. plumbaginifolia and find that the small subunit of this factor is
also expressed in different isoforms in plants. Structural and
functional similarity between the plant and human U2AF subunits
suggests that plant U2AF, like its metazoan counterpart, interacts with
and defines the 3'ss-proximal region of the intron. Identification of
the U2AF-like factor in plants raises the interesting question as to
how the plant splicing machinery distinguishes between the 3'-proximal
and upstream U-rich elements during intron definition.
Isolation and Characterization of cDNA Clones--
The
PCR-derived cDNA clones encoding different RBD-type RNA binding
domains found in proteins of N. plumbaginifolia are
described by Mieszczak et al. (30). The cDNA insert
encoding the domain number 8 was labeled with
[ Plasmid Constructions--
Constructs for overexpression of
glutathione S-transferase (GST) fusion proteins were
prepared as follows. The coding region of NpU2AF65a was
amplified by PCR using the oligonucleotides
AGGAGGGAACCATGGGGGACTATGA and GTGGTTAAAGGGATCCCTCATAGT and cloned
into the HincII-digested pBluescript KS vector. The
NpU2AF65a cDNA fragment was excised from the
plasmid by digestion with XhoI and EcoRV, the
ends were filled in using the Klenow DNA polymerase, and the purified
DNA fragment was ligated into the SmaI-digested pGEX-2T vector (Amersham Pharmacia Biotech). To restore the coding frame, the plasmid was cut with ClaI, treated with the
Klenow enzyme, and religated. This construct, named
pGST-U2AF65a, contains an additional 33-bp sequence that
encodes the SSLSRYRQEGT peptide positioned between the GST and the
NpU2AF65a coding regions. The coding region of
NpU2AF65b was amplified by PCR using the oligonucleotides
TGGACTAAGGGATCCAGTCAAGAG and
GGGTGGGTAGAATTCACCATCATA, which introduce a
BamHI site in front of the ATG codon and an EcoRI
site in place of the stop codon, respectively (sequences of introduced
restriction sites are printed in bold). The PCR product was cut with
BamHI and EcoRI and ligated into the
corresponding sites of pGEX-2T, resulting in pGST-U2AF65b.
Plasmid pGST-hU2AF (in pGEX-3X; Amersham Pharmacia Biotech), containing
human GST-U2AF65 cDNA (5), was a kind gift of Dr. M. Green (University of Massachusetts, Worcester, MA).
To prepare plasmids for generation of the RNase mapping probes, the
fragment downstream of the SphI site located at position 1,370 of the U2AF65a cDNA in pNpU2AF65a and
the fragment downstream of the SacI site located at position 1,589 of the U2AF65b cDNA in pNpU2AF65b/3'
were removed by cutting with appropriate restriction enzymes and
religating. The two resulting plasmids are referred to as pU2AF65a/probe and pU2AF65b/probe. Plasmids
used for generation of the UV cross-linking substrates are pGS7 (19),
encoding syn7 RNA, and pGS/Rca. pGS/Rca was
derived by replacing the KpnI-XhoI DNA fragment
of pGS7 by a DNA fragment encoding intron 3, flanked by 10 nucleotides
of exon 2 and 11 nucleotides of exon 3, of the Arabidopsis
Rubisco activase (Rca) gene. The Rca gene
fragment was amplified by PCR using pRcam (17) as a template and
oligonucleotides TTGACGGTACCTAACATCAA and
CATATACTCGAGATAAGTGGAAC, which introduce the KpnI
site upstream of the 5'ss and the XhoI site downstream of
the 3'ss, respectively.
To express the hemagglutinin (HA) epitope-tagged protein, the coding
region of NpU2AF65a was amplified by PCR with
oligonucleotides AGGAGGGAACCATGCGGGACTATGA and
GTGGTTAAAGGGATCCCTCATAGTC, which introduce the
NcoI site upstream of the ATG codon and the BamHI
site at the position of the stop codon, respectively. The PCR product
was cut with NcoI and BamHI and cloned into
NcoI/BamHI-cleaved pGGS.5 (31), yielding plasmid
pNpU2AF65a/HA from which NpU2AF65a is expressed
as a C-terminal fusion with the influenza hemagglutinin nonapeptide
tag, YPYDVPDYA. The identity of all constructs was verified by sequencing.
RNase A/T1 Mapping--
RNA from different organs of N. plumbaginifolia plants and from cells grown in suspension (26) was
isolated following the guanidinium chloride method described by Arrand
(32). Plasmids pU2AF65a/probe and
pU2AF65b/probe were linearized with PvuII and
transcribed by T3 RNA polymerase using [ Overexpression and Purification of GST Fusion Proteins--
The
bacterial host for the expression of the N. plumbaginifolia
and human GST fusion proteins was BL21 (DE3). One liter of LB medium
supplemented with 2% glucose and 100 µg/ml ampicillin was inoculated
with an overnight culture grown at 37 °C to have a starting OD at
600 nm of 0.05-0.1. Cells were grown at 37 °C to an OD at 600 nm of
0.7-0.9, and the expression of transgenes was induced with 0.1 mM isopropyl- UV Cross-linking--
Plasmids pGS7 and pGS/Rca were
linearized with PstI and transcribed with T7 RNA polymerase
(Promega) using [ In Vitro Splicing Assays--
HeLa cell nuclear extracts
prepared as described in Dignam et al. (34) were depleted of
U2AF by chromatography on oligo(dT)-cellulose exactly as described in
Valcárcel et al. (35). The 9-µl in vitro
splicing reactions contained 15 µg of the nuclear extract protein,
the amounts of recombinant proteins indicated in the legend of Fig.
4B, and 20,000 cpm (10 fmol) of the AdML pre-mRNA. Incubations were for 40 min at 30 °C in 13.3 mM
HEPES-KOH, pH 8.0, containing 3 mM MgCl2, 66.6 mM KCl, 1.1 mM ATP, 22.2 mM
creatine phosphate, 0.9 units/µl of RNasin, 0.13 mM
EDTA, 13.3% glycerol, 3.3% polyvinyl alcohol, 0.03% Nonidet P-40,
and 0.66 mM dithiothreitol. The reactions were stopped by
digestion with proteinase K, and the RNAs, purified by
phenol/chloroform extraction and ethanol precipitation, were analyzed
by electrophoresis on a 13% polyacrylamide denaturing gel.
Immunolocalization in N. plumbaginifolia
Protoplasts--
Protoplasts prepared from N. plumbaginifolia leaves were transfected by the polyethylene glycol
method (33) using 20 µg of plasmid DNA/6 × 105
protoplasts. Protoplasts were processed for indirect immunofluorescence 18-20 h after transfection as described by Genschik et al.
(31), using the rabbit polyclonal anti-HA antibody HA11 (Berkley
Antibody Co.) at 1:80 dilution and fluorescein
isothiocyanate-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch
Laboratories) at 1:100 dilution. The slides were examined with a Zeiss
Axiophot microscope and a Leica TCS 4D confocal scanning laser microscope.
Characterization of the cDNAs Encoding NpU2AF65a
and b--
We have previously identified, using a PCR-based approach,
15 cDNA sequences encoding different RBD-type RNA binding domains in N. plumbaginifolia (30). One of the characterized
RBD domains (domain 8; Ref. 30) showed strong sequence similarity to
one of the RBD domains of the large 65-kDa subunit of the human
splicing factor U2AF. The 2,204-bp-long cDNA encoding the
full-length protein, named NpU2AF65b, containing this RBD
domain was isolated by screening the N. plumbaginifolia
cDNA library followed by the 5'-rapid amplification of cDNA
ends (see "Experimental Procedures"). The NpU2AF65b
cDNA was used for an additional screen of the cDNA library. Four cDNA clones encoding a protein closely related to
NpU2AF65b, referred to as NpU2AF65a, were
identified and sequenced. The longest cDNA of class a is 2,108 bp long.
Conceptual translation of the NpU2AF65a and b cDNAs
yields proteins of 555 and 573 amino acids, respectively. In both
clones, the ATG codon is preceded by a stop codon in the same coding
frame. The 5'-untranslated regions of the NpU2AF65a and b
cDNAs are 62 and 138 bp, respectively. However, the 5'-terminal regions of the cDNAs may be incomplete. Northern analysis has indicated that NpU2AF65b mRNA is approximately 2,600 nucleotides long (data not shown). The 3'-untranslated regions of the
NpU2AF65a and b cDNAs are 381 and 347 bp long, respectively.
The NpU2AF65a and b proteins are 85% identical and 92%
similar and have the same structural organization. In both proteins, the approximately 150-170-amino acid N-terminal domain, rich in RS and
RD repeats, is separated by the 50-amino acid-long linker region from
the three RBD-type RNA binding domains present at the C-terminal part.
Most differences between the two proteins are confined to the
N-terminal RS/RD domain. When compared with NpU2AF65a,
NpU2AF65b contains a 22-amino acid-long insertion close to
the N terminus and is missing four amino acids, DRDG, that follow
position 74 in NpU2AF65a (Fig.
1A). Data base searches
indicated strong homology of both plant proteins with the large
subunits of the splicing factor U2AF from other organisms. Both
N. plumbaginifolia proteins are most related to large U2AF
subunits from mammals and Drosophila (approximately 41 and
49-50% identity and similarity, respectively). The lowest
conservation, 32% identity and 45% similarity, is with the homolog
from Schizosaccharomyces pombe. This high sequence conservation and similar domain organization indicate that the two
N. plumbaginifolia proteins are U2AF65 homologs
(see also below).
When this work was completed, sequences encoding likely counterparts of
the human U2AF65 appeared as part of the
Arabidopsis genome-sequencing project. The complete sequence
of one Arabidopsis U2AF65 isoform, deduced from
the genomic sequence (GenBank accession number Z99708) and referred to
as AtU2AF65a, is included in the alignment shown in Fig.
1A. Homology of AtU2AF65a with U2AF large
subunits from other species is similar to that of the N. plumbaginifolia proteins. In addition, the partial genomic sequence (GenBank accession number AC002292), which encodes a second
AtU2AF65 isoform, has been identified. The deduced protein
fragment shows 75% identity and 81% similarity with amino acids
459-568 of AtU2AF65a. The conservation of plant and
metazoan proteins is higher in RNA binding domains than in the RS/RD
domain, where the only common feature is the presence of RS and RD
repeats without any conserved arrangement. A notable characteristic of
the plant U2AF65 proteins is the bias in favor of RD over
RS repeats in the N-terminal domain (Fig. 1B). The two
regions of U2AF65 shown to be responsible for the
interaction with the small subunit of U2AF (36, 37) and with the
splicing factor UAP56 (7) in animals are also conserved in plant
proteins (Fig. 1, C and D).
Expression of the NpU2AF65 Genes--
Southern
analysis performed with N. plumbaginifolia genomic DNA
digested with different restriction enzymes and with
NpU2AF65a and b cDNAs as probes has indicated that,
consistent with the results of the cDNA library screening, the
N. plumbaginifolia genome contains two genes encoding
different U2AF65 isoforms (data not shown). Expression of
the NpU2AF65a and b genes in different tissues of N. plumbaginifolia was studied by RNase A/T1 mapping using antisense
RNA probes specific for NpU2AF65a and b mRNAs
(see "Experimental Procedures"). Both mRNAs are present at
similar levels in leaves, stems, and apical buds and also in N. plumbaginifolia cells grown in suspension (Fig.
2). Hence, it appears that the
U2AF65 genes are constitutively expressed in different
plant tissues, in agreement with the expected general function of U2AF
in pre-mRNA processing.
Cellular Localization of NpU2AF65 by Indirect
Immunofluorescence--
An epitope-tagging approach combined with
indirect immunofluorescence was used to determine the cellular
localization of the N. plumbaginifolia U2AF65a
protein. Plasmids expressing NpU2AF65a with the influenza
HA epitope fused to the C terminus were transfected into mesophyll
protoplasts of N. plumbaginifolia. The protoplasts were
processed for immunofluorescence microscopy, using a rabbit anti-HA
antibody and a fluorescein isothiocyanate-conjugated goat anti-rabbit
antibody. Analysis of the slides by fluorescence (Fig. 3A) or confocal (Fig.
3B) microscopy indicated that the fusion protein localizes
to the nucleus. The staining coincided with nuclei visualized with the
dye 4',6'-diamidino-2-phenylindole and was not visible when the anti-HA
antibody was omitted (data not shown). Protoplast staining was
completely different when the same HA epitope was fused to the
Arabidopsis cyclic nucleotide phosphodiesterase, a protein
with cytoplasmic localization (Fig. 3, A and B;
Ref. 31).
Activity of NpU2AF65 in the HeLa Cell Splicing
Extract--
NpU2AF65a and b proteins were expressed as
N-terminal GST fusions in E. coli. Recombinant proteins were
purified by affinity chromatography on glutathione-Sepharose 4B and by
gel filtration on a Superose 12 column. The fraction of full-length
fusion protein was 70-80% total protein as judged by
SDS-polyacrylamide gel electrophoresis (Fig.
4A).
Because an in vitro splicing system originating from plant
cells is not available, we tested whether plant U2AF65
isoforms can substitute for the human factor in the HeLa cell-free splicing extract depleted of the endogenous U2AF65 (5, 35).
It was found that both NpU2AF65a and b GST fusion proteins
markedly stimulate splicing of the AdML pre-mRNA in the depleted
extract, although their activity is lower than that of the homologous
factor (Fig. 4B). The addition of higher amounts of plant
proteins was found to be slightly inhibitory (Fig. 4B,
lanes 8, 9, and 12), possibly because of some
squelching effect (35).
Binding Specificity of the NpU2AF65 Isoforms--
UV
cross-linking and competition experiments were performed to
characterize the RNA binding properties of the plant
U2AF65. Both recombinant NpU2AF65 isoforms
could be UV cross-linked to RNA fragments containing either intron 3 of
the Arabidopsis Rubisco activase gene (Rca; Ref.
38) (Fig. 5; for the sequences of the
Rca intron, see Fig. 6) or a
synthetic intron syn7 (19) (data not shown). To determine nucleotide
specificity of the NpU2AF65 isoforms, four different
homoribopolymers were tested as competitors, using Rca RNA
as a 32P-labeled substrate. With both isoforms, poly(U) was
found to be the most potent competitor, having a strong effect even at a 5-fold molar excess, whereas with poly(G) and poly(C), similar inhibition required approximately a 25-fold excess of the polymer. Poly(A) did not compete for the Rca RNA binding even when
added at 400-fold molar excess (Fig. 5).
During splicing in metazoa, U2AF65 interacts with
polypyrimidine tracts located specifically between the branch point and
the 3'ss (5, 39). Because U- or UA-rich sequences are usually distributed throughout the intron (Refs. 19, 21, 22, and 24; reviewed
in Refs. 4 and 15), it is not obvious how U2AF would distinguish
between the 3'ss-proximal and upstream U-rich elements in plant
introns. We used the Rca gene intron 3 to find out whether
3'ss or branch point nucleotides might contribute to plant
U2AF65 binding at the downstream region of the intron. The
Rca intron 3 is the only natural plant intron for which the
branch point utilized during splicing in planta has been
directly mapped (17). This intron contains 74% UA but has 2 particularly U-rich stretches, 1 of 9 U residues interrupted by 1 C,
positioned downstream of the 5'ss, and another consisting of 8 consecutive Us, located upstream of the 3'ss (Fig. 6A). Five
different 27-mer oligoribonucleotides (Fig. 6A) were used in
competition experiments to analyze NpU2AF65 binding to the
Rca intron. As shown in Fig. 6B, oligonucleotides containing the 3'ss-proximal U stretch, in combination with either nearly optimal or debilitated 3'ss or with the wild-type or the consensus branch point, competed with equal efficiencies for the binding of NpU2AF65b to the intron. In contrast, the U-rich
oligoribonucleotide representing the 5'ss region of the intron was a
significantly weaker competitor (Fig. 6B). The same results
were obtained with the NpU2AF65a isoform. Hence, it appears
that the region positioned between the branch point and the 3'ss of the
intron may have some special features other than the branch point or
the 3'ss that contribute to the interaction with
NpU2AF65.
Sequences of Plant U2AF35-Like Proteins--
In
metazoa and S. pombe, U2AF is a heterodimer consisting of
one large and one small subunit. By searching data bases, we have
identified five different rice Expressed Sequence Tags encoding protein
fragments having similarity with the small subunit of the human factor,
U2AF35. Complete sequencing of the cDNAs revealed, that
they encode two different isoforms of the rice U2AF35
counterpart, termed OsU2AF35a and OsU2AF35b
(Fig. 7). One of the identified cDNAs
encodes a variant of OsU2AF35a, likely resulting from
alternative splicing of the OsU2AF35a mRNA, which lacks
16 amino acids (positions 256-281; indicated by a broken
line in Fig. 7) in the RS domain. OsU2AF35a and b are
predicted to be 290 and 301 amino acids long, respectively. The two
rice proteins are 77% identical and 82% similar. Their identity with
the human and S. pombe homologs is 55-59 and 55%, respectively. A gene encoding a U2AF35-like protein has
also been recently identified as part of the Arabidopsis
sequencing project (GenBank accession numbers AB008264 and AB007647).
The encoded protein is included in the alignment shown in Fig. 7.
Plant U2AF small subunits are 40-60 amino acids longer than
counterparts in other organisms, but the overall structure of all the
proteins is similar. Like metazoan and S. pombe (36, 40,
41), plant factors contain two zinc finger-like domains, the pseudo RBD
domain (42),2 and the RS/RD
domain, with excess of RS over RD in the C-terminal part of the
protein. Although OsU2AF35b, like the metazoan proteins,
contains very glycine-rich regions near the C terminus, such regions
are missing from OsU2AF35a and AtU2AF35
(Fig. 7).
In this work, we demonstrate that two different isoforms of
U2AF65 are expressed in N. plumbaginifolia. Two
different isoforms of U2AF65 appear to be encoded also in
the Arabidopsis genome, and at least two classes of
U2AF35 genes are active in rice, one of them potentially
encoding two alternatively spliced products. Hence, it appears that
plants quite generally express multiple isoforms of both U2AF subunits. The biological significance, if any, of this finding needs to be
established. In all organisms studied to date, only one form of large
and small U2AF subunits has been identified, although a protein-sharing
limited similarity with U2AF35, named Urp, which interacts
with U2AF65 and SR proteins, has been identified in mammals
(43). In Caenorhabditis elegans, the alternatively spliced
RNA is produced in addition to the usual U2AF65 mRNA.
However, this variant RNA is apparently not translated and may provide
a regulatory function (44).
The plant U2AF65- and U2AF35-like proteins are
highly conserved with metazoan counterparts, at both domain
organization and primary sequence levels. The major difference between
plant and metazoan U2AF65 proteins is in the N-terminal RS
domain, which has a very strong bias in favor of RD over RS dipeptides
in plant proteins. The RD repeats are also found, although in smaller
numbers, in RS domains of U2AF65 in other organisms (Fig.
1B) and in the U1 snRNP 70 kDa protein of metazoa (45) and
plants (46). The function of RD repeats is not known, but it has been
proposed that several additional RD-rich proteins may participate in
splicing in mammals (47). As regards the small U2AF subunits,
potentially interesting is the observation that only one of the
characterized rice U2AF35 isoforms, OsU2AF35b,
contains a glycine-rich domain. Glycine-rich sequences, although not
essential for the function of the Drosophila factor in
vivo, are present at a similar location in all known metazoan
U2AF35 subunits (36, 40).
Studies carried out mainly in metazoa have demonstrated an essential
role for U2AF in 3'ss selection. U2AF specifically binds to the
pyrimidine tract between the branch point and the 3'ss (5, 39). It
recruits U2 snRNP to the branch point (5, 6) and is implicated in
establishing bridging contacts between the 5' and 3' intron regions (8,
48, 49). In these functions, the U2AF is assisted by other proteins.
U2AF65 interacts cooperatively with the branch
point-bridging protein, mBBP/SF1, which specifically recognizes the
branch site (8-10). Furthermore, U2AF65 recruits to
pre-mRNA a DEAD-box protein, UAP56, required for the U2
snRNP-branch point interaction (7). U2AF, particularly its small
subunit, also appears to be involved in a number of interactions with
constitutive and alternative splicing factors of the SR family
(48-51). Both N. plumbaginifolia U2AF65
isoforms were found to stimulate splicing of the AdML pre-mRNA in
the HeLa cell extract depleted of the endogenous U2AF, indicating functional conservation between plant and human factors. This finding,
together with the strong structural similarity between plant and
metazoan U2AF subunits, suggests that functions of U2AF during splicing
in plants are similar to those in metazoa. This conclusion is further
supported by our identification of the Arabidopsis and/or
rice Expressed Sequence Tags, which encode proteins having strong
similarity to the mammalian factors interacting with U2AF65
such as UAP56 or
mBBP/SF1.3
UA- or U-rich sequences distributed usually along the entire length of
plant introns are essential for both efficient splicing and definition
of the 5' and 3' intron borders (see the introduction). If U2AF in
plants indeed interacts with the region upstream of the 3'ss and helps
U2 snRNP to recognize the branch site, how does the factor distinguish
the 3'ss-proximal U-rich element from other similar sequences present
elsewhere in the intron? Competition experiments with different
oligoribonucleotides spanning the 3' region of the Rca
intron indicated that branch point or 3'ss sequences do not contribute
to the binding of the isolated NpU2AF65 to the
3'ss-proximal U-rich element. However, recent experiments have shown
that in yeast and mammals, the branch point region is recognized at the
early step of spliceosome assembly by the branch point-bridging factor,
BBP. The mammalian factor, mBBP/SF1, and U2AF65 interact
cooperatively during binding to the adjacent branch site and
polypyrimidine sequences, and this interaction may be crucial for both
contacting the U1 snRNP and recruiting the U2 snRNP to the branch point
(8-10). It is possible that a similar cooperative interaction also
occurs in plants.
An additional factor contributing to the specificity of U2AF binding
could be the properties of the 3'ss-proximal region itself. U-rich
regions positioned between the branch point and the 3'ss may differ
structurally from apparently similar sequences present at other intron
locations. Indeed, we have found that oligoribonucleotides representing
the branch point-to-3'ss region compete for NpU2AF65
binding to the Rca intron more efficiently than the 5'ss
oligonucleotide, despite all these oligonucleotides being equally U
rich. Compilations of plant intron sequences indicated that in the
3'ss-proximal region (positions -20/ Cross-linking experiments with the nuclear extracts from N. plumbaginifolia have previously identified a group of ~50-kDa proteins that specifically interact with plant intron sequences in vitro (26). Cloning of the most efficiently
cross-linkable 50-kDa protein revealed that it is a nuclear protein
containing three RBD domains. Binding of this protein to plant introns
is competed by poly(U) but not by other homopolymers
(26).4 Hence, nucleotide
specificity of this protein differs from that of NpU2AF65.
Further characterization of intron-binding proteins and splicing factors and preparation of splicing extracts originating from plant
cells will be required to better understand the mechanism of intron
recognition in plants.
We thank the Japanese Rice
Research Program of the National Institute of Agrobiological Resources
and The Institute of the Society of Techno-Innovation in Agriculture,
Forestry, and Fisheries, Tsukuba 251, Japan for making the rice
Expressed Sequence Tag clones available to us and A. Steinmetz for
temporarily hosting C. D. in his laboratory at the Institute of
Plant Molecular Biology of CNRS in Strasbourg. We also thank M. Hemmings-Mieszczak and U. Klahre for their initial input to cDNA
cloning, F. Dragon for help with some experiments, J. Petruska for
technical assistance, and F. Dragon, M. Echeverria, and M. Lambermon
for critical reading of the manuscript.
*
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement" in
accordance with 18 U.S.C. Section
1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) Y18348, Y18349, Y18350, and Y18351.
§
Present Address: INSERM U381, 3, avenue Molière, 67200 Strasbourg, France.
The abbreviations used are:
3'ss and 5'ss, 3' and 5' splice site, respectively; BBP, branch point-bridging
protein; GST, glutathione S-transferase; PCR, polymerase
chain reaction; snRNP, small nuclear ribonucleoprotein; U2AF, U2 snRNP
auxiliary factor; Rubisco, ribulose-bisphosphate carboxylase/oxygenase; bp, base pair(s); HA, hemagglutinin
2
A. R. Krainer, personal communication.
3
Z. J. Lorkovic and W. Filipowicz,
unpublished observations.
4
M. Hemmings-Mieszczak, U. Klahre, M. Lambermon,
and W. Filipowicz, unpublished results.
Multiple Forms of the U2 Small Nuclear Ribonucleoprotein
Auxiliary Factor U2AF Subunits Expressed in Higher Plants*
§,
,
Friedrich Miescher-Institut, CH-4002 Basel,
Switzerland and ¶ Gene Expression Programme, European
Molecular Biology Laboratory, Meyerhofstrasse 1, D-69117 Heidelberg, Germany
ABSTRACT
Top
Abstract
Introduction
Procedures
Results
Discussion
References
INTRODUCTION
Top
Abstract
Introduction
Procedures
Results
Discussion
References
EXPERIMENTAL PROCEDURES
Top
Abstract
Introduction
Procedures
Results
Discussion
References
-32P]dATP (3000 Ci/mmol; Amersham Pharmacia Biotech)
by the random priming method and used as a probe for screening of the
N. plumbaginifolia cDNA 
-ZAP II library (4.5 × 105 plaque-forming units) as described by Mieszczak
et al. (30). The isolated 1,379-bp-long cDNA, present in
pBluescript SK(
) and named pNpU2AF65b/3', was sequenced
on both strands. The missing 5'-end region of the cDNA was isolated
following the 5'-rapid amplification of cDNA ends protocol using
the Marathon kit (CLONTECH) and two nested
oligonucleotide primers, GGCATTACTAGCTTCCTCTACT and
TATGATGCCATCCAATGCCATGGCA, complementary to the regions located
at positions 1,044-1,023 and 1,066-1,041 of the complete
NpU2AF65b cDNA. A 1,044-bp isolated fragment containing
the 5' portion of the cDNA was cloned into the HincII
site of pBluescript KS and sequenced. The full-length
NpU2AF65b cDNA clone was reconstructed by ligating the
two cDNA halves using the SphI site present at position
850 of the complete cDNA. The final construct contains 2,204 bp of
cDNA cloned in the BspD I site of pBluescript SK(). To
isolate NpU2AF65a cDNA, the same cDNA library
(4 × 105 plaque-forming units) was screened using the
3'-terminal 1,379-bp fragment of NpU2AF65b cDNA as a
probe. The phagemid DNA of the four purified clones was excised using
the ExAssist/SOLR system (Stratagene), and their inserts were
sequenced. The longest 2,108-bp cDNA, present in pBluescript
SK(), is called pNpU2AF65a. The five rice Expressed
Sequence Tag cDNA clones (accession numbers D46624, D48800, C19537,
C19669, and C28752) encoding U2AF small subunits were obtained from
MAFF DNA Bank (Tsukuba, Japan) and sequenced. The nucleotide sequences
of N. plumbaginifolia U2AF65a and b cDNAs
are deposited in the EMBL/GenBank/DDBJ Nucleotide Sequence
Libraries under accession numbers Y18351 and Y18350, respectively.
Sequences of rice U2AF35a and b cDNAs are deposited
under accession numbers Y18349 and Y18348, respectively.
-32P]CTP (800 Ci/mmol). Probes were purified by gel electrophoresis, and RNase A/T1
mapping was performed as reported previously (33) using 10 µg of RNA
per assay.
-D-thiogalactopyranoside for 1-2 h at 30 °C. Cells were harvested by a 10-min centrifugation at
3,000 × g at 4 °C and resuspended in 20 ml of ice-cold lysis buffer (20 mM Tris-HCl, pH 7.5, 1 M NaCl, 0.2 mM EDTA, 1 mM dithiothreitol, and 3 mM phenylmethylsulfonyl fluoride) and subsequently broken with a French press (800 p.s.i.). After centrifugation (25,000 × g for 30 min at 4 °C), the supernatant was applied onto
glutathione-Sepharose 4B and further processed according to
manufacturer's instructions (Amersham Pharmacia Biotech). Proteins
were further purified by gel filtration on the Superose 12 column to
eliminate contaminating shorter protein fragments retained on the
affinity column. Pooled fractions were dialyzed against 2 liters of 20 mM HEPES-KOH buffer, pH 7.9, containing 100 mM
KCl, 0.5 mM dithiothreitol, and 0.2 mM EDTA.
The protein concentration was estimated by the Bradford assay.
-32P]UTP (800 Ci/mmol) as a label.
RNAs were purified by gel electrophoresis. RNA (2 fmol; ~40,000 cpm),
with or without a competitor added, was denatured by brief heating at
95 °C and chilling on ice. RNA samples were incubated with 100 ng of
GST-U2AF65 fusion proteins for 10 min at 20 °C in 10 mM HEPES-KOH buffer, pH 7.9, containing 2 mM
MgCl2, 50 mM KCl, 0.025% Nonidet P-40, 1 mM dithiothreitol, 10% glycerol and 5 units of RNasin
(Promega) in a total volume of 10 µl. UV cross-linking was performed
in Eppendorf tubes placed on ice in the Stratalinker (Stratagene) by
15-min irradiation at 2.3 joules/cm2 (~10 cm distance
from the UV source). Ten µg of RNase A was added and after 45 min at
37 °C, the samples were analyzed on 10% SDS-polyacrylamide gels.
RESULTS
Top
Abstract
Introduction
Procedures
Results
Discussion
References
View larger version (108K):
[in a new window]
Fig. 1.
Sequences of plant U2AF65
proteins. A, alignment of two N. plumbaginifolia U2AF65 isoforms with the
Arabidopsis and human homologs. Assignment of the
exon/intron borders in the Arabidopsis U2AF65
(GenBank accession number Z99708) is incorrect. The protein sequence
used for alignment is based on a more probable splicing pattern of the
gene. The RNP1 (8 amino acids) and RNP2 (6 amino acids) motifs of the
three RBD domains are underlined. B, numbers of RS and RD
repeats present in RS/RD domains of plant, animal, and fission yeast
U2AF65 proteins. C, sequence alignment of
U2AF65 domains implicated in the interaction with the U2AF
small subunit. Alignment is based on that presented by Rudner et
al. (37). D, alignment of U2AF65 domains
implicated in the interaction with the splicing factor UAP56 (7).
Multiple sequence alignments in A, C, and
D were performed with the PileUP program using the complete
multiple alignment protocol with default parameters. Alignment in
A was improved manually. Identical amino acids and amino
acids conserved in at least 50% of sequences are indicated by
black and gray boxes, respectively.
View larger version (40K):
[in a new window]
Fig. 2.
Expression of NpU2AF65 genes
analyzed by RNase A/T1 mapping. A, schematic representation
of U2AF65a and b probes, which are complementary to
nucleotides 1,050-1,370 and 1,170-1,590 in NpU2AF65a and
NpU2AF65b mRNA, respectively. B, analysis of
protected fragments on a denaturing gel. Mapping was performed with RNA
isolated from apical buds (1), leaves (2), stems
(3), and cells grown in suspension (4).
Lanes a and b represent mappings performed with
U2AF65a- and U2AF65b-specific probes,
respectively. Two lanes on the left ( ) correspond to
control mappings performed with tRNA added instead of plant RNA.
Lane M, size markers. nt, nucleotides.
View larger version (43K):
[in a new window]
Fig. 3.
Immunolocalization of the
NpU2AF65a in protoplasts of N. plumbaginifolia. Protoplasts were transfected with
constructs expressing fusions of the HA epitope-tagged
NpU2AF65a (U2AF65) or cyclic nucleotide
phosphodiesterase (CPDase) from Arabidopsis.
Protoplasts were processed for indirect immunofluorescence as described
under "Experimental Procedures" and examined either by fluorescence
(A) or by laser scanning confocal (B) microscopy.
Bars represent 40 µm (A) and 15 µm
(B). The arrows in A indicate the
nuclei of productively transfected cells visualized with either
fluorescein isothiocyanate (FITC)-conjugated secondary
antibody (left panel) or with 4',6'-diamidino-2-phenylindole
staining (right panel). In B, the panels
a and b represent immunostaining of protoplasts
transfected with U2AF65 and CPDase, respectively.
View larger version (59K):
[in a new window]
Fig. 4.
Preparation of recombinant
Np2AF65 a and b isoforms and their activity in the HeLa
cell in vitro splicing system. A, Coomassie
Blue-stained gel of the purified GST-U2AF65a and b
isoforms. Two µg of protein was loaded in each lane.
Positions of molecular weight markers (low molecular weight calibration
kit, Amersham Pharmacia Biotech) are indicated (M).
B, in vitro splicing. Reactions were carried out
with the complete HeLa cell nuclear extract (NE, lanes
1 and 2) or with the extract depleted of endogenous
U2AF (lanes 3-12). The depleted extract was
complemented with recombinant GST fusion proteins: human
U2AF65 (lanes 4-6),
NpU2AF65a (lanes 7-9), or
NpU2AF65b (lanes 10-12). Increasing
amounts of added proteins correspond to 25, 75, and 225 ng/assay. The
lower panel represents a longer exposure of the bottom
portion of the gel to show the exon 1 intermediate.
View larger version (42K):
[in a new window]
Fig. 5.
Competition of different homoribopolymers for
cross-linking of the 32P-labeled Rca RNA to
recombinant NpU2AF65 isoforms. UV cross-linking of
either the U2AF65a (upper panel) or
U2AF65b (lower panel) isoform was performed in
the absence ( ) or presence of increasing amounts of the indicated
homoribopolymers. Homoribopolymers were added at 5- (lanes 2, 6, 10, and 14), 25- (3, 7, 11, and
15), 100- (lanes 4, 8, 12, and 16),
and 400-fold (lanes 5, 9, 13, and 17) excess
(calculated in mol of nucleotides) over Rca RNA. After UV
cross-linking and digestion with RNase A, products were analyzed on
10% polyacrylamide/SDS gels.
View larger version (38K):
[in a new window]
Fig. 6.
Competition of different oligoribonucleotides
with the UV cross-linking of NpU2AF65b to Rca
RNA. A, sequences of the Rca intron 3 and
oligoribonucleotides used as competitors. UA-rich stretches of four or
more nucleotides in the Rca intron are
underlined. The branch site and 3'ss sequences are in
bold. Changes introduced into oligoribonucleotide sequences
are underlined. The mapped branch point A residue is
indicated with an asterisk. Flanking exon sequences are in
lowercase. opt, optimized; mut, mutant;
wt, wild type; cons, consensus; B,
unlabeled competitors specified at the top of the gel were added at
100- (lanes 2, 5, 8, 12, and 15),
1,000- (lanes 3, 6, 9, 13, and 16), and
10,000-fold (lanes 4, 7, 11, 14, and 17) molar
excess over Rca RNA. Lane 1, cross-linking
without competitor added. For additional details, see "Experimental
Procedures" and Fig. 5.
View larger version (80K):
[in a new window]
Fig. 7.
Alignment of two rice U2AF35
isoforms OsU2AF35a and b, and Arabidopsis
U2AF35 with those from metazoa and S. pombe. Sequence alignment was performed as described in Fig. 1.
The AtU2AF35 presented is a translational product
originating from two adjacent genomic clones (GenBank accession numbers
AB008264 and AB007647). The identity of the amino acid, marked by
X, at position 73 is not certain. The putative zinc fingers
are indicated below the alignment (CCCH), and the RNP1 and RNP2 of the
pseudo RBD domain are underlined. The glycine-rich regions
in human, Drosophila, and OsU2AF35b are printed
in bold and underlined. The sequence absent in
the OsU2AF35a variant protein is indicated by the
broken line. Os, Oryza sativa (rice).
For other abbreviations, see Fig. 1.
DISCUSSION
Top
Abstract
Introduction
Procedures
Results
Discussion
References
6 relative to the 3'ss AG),
enrichment in U residues is a few percent higher than in the regions
positioned further upstream (Refs. 52, 53; reviewed in Ref. 4).
However, it is possible that nucleotides other than U also contribute
to interactions with plant U2AF. In vitro competition
experiments indicated that poly(G) and poly(C) compete, although less
strongly than poly(U), with the binding of U2AF65 to the
Rca intron.
ACKNOWLEDGEMENTS
FOOTNOTES
To whom correspondence should be addressed: Friedrich
Miescher-Institut, P.O. Box 2543, CH-4002 Basel, Switzerland. Tel.: 57-6976993 or 6978234; Fax: 57-6973976; E-mail: Filipowi{at}FMI.CH; FedEx
address: Maulbeerstrasse 66, CH-4058 Basel.
REFERENCES
Top
Abstract
Introduction
Procedures
Results
Discussion
References
Copyright © 1998 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. GULLEROVA, A. BARTA, and Z. J. LORKOVIC AtCyp59 is a multidomain cyclophilin from Arabidopsis thaliana that interacts with SR proteins and the C-terminal domain of the RNA polymerase II RNA, April 1, 2006; 12(4): 631 - 643. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B.-B. Wang and V. Brendel Molecular Characterization and Phylogeny of U2AF35 Homologs in Plants Plant Physiology, February 1, 2006; 140(2): 624 - 636. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. J. Webb and J. A. Wise The Splicing Factor U2AF Small Subunit Is Functionally Conserved between Fission Yeast and Humans Mol. Cell. Biol., May 15, 2004; 24(10): 4229 - 4240. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Lopato, C. Forstner, M. Kalyna, J. Hilscher, U. Langhammer, K. Indrapichate, Z. J. Lorkovic', and A. Barta Network of Interactions of a Novel Plant-specific Arg/Ser-rich Protein, atRSZ33, with atSC35-like Splicing Factors J. Biol. Chem., October 11, 2002; 277(42): 39989 - 39998. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. H. L. Lambermon, Y. Fu, D. A. W. Kirk, M. Dupasquier, W. Filipowicz, and Z. J. Lorkovic UBA1 and UBA2, Two Proteins That Interact with UBP1, a Multifunctional Effector of Pre-mRNA Maturation in Plants Mol. Cell. Biol., June 15, 2002; 22(12): 4346 - 4357. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. J. Lorkovic and A. Barta Genome analysis: RNA recognition motif (RRM) and K homology (KH) domain RNA-binding proteins from the flowering plant Arabidopsis thaliana Nucleic Acids Res., February 1, 2002; 30(3): 623 - 635. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Golovkin and A. S. N. Reddy An SC35-like Protein and a Novel Serine/Arginine-rich Protein Interact with Arabidopsis U1-70K Protein J. Biol. Chem., December 17, 1999; 274(51): 36428 - 36438. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Hartmann, O. Nayler, F. W. Schwaiger, A. Obermeier, and S. Stamm The Interaction and Colocalization of Sam68 with the Splicing-associated Factor YT521-B in Nuclear Dots Is Regulated by the Src Family Kinase p59fyn Mol. Biol. Cell, November 1, 1999; 10(11): 3909 - 3926. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |