MOLECULAR AND CELLULAR BIOLOGY,
Copyright ? 1997, American Society for Microbiology
May 1997, p. 2587–2597 Vol. 17, No. 5
Interaction between the Human Nuclear Cap-Binding Protein
Complex and hnRNP F
CHIARA GAMBERI, ELISA IZAURRALDE, CHRISTINA BEISEL, AND IAIN W. MATTAJ*
European Molecular Biology Laboratory, D-69117 Heidelberg, Germany
Received 14 October 1996/Returned for modification 18 December 1996/Accepted 23 January 1997
hnRNP F was identified in a screen for proteins that interact with human CBP80 and CBP20, the compo-
nents of the nuclear cap-binding complex (CBC). In vitro interaction studies showed that hnRNP F can bind
to both CBP20 and CBP80 individually. hnRNP F and CBC bind independently to RNA, but hnRNP F binds
preferentially to CBC-RNA complexes rather than to naked RNA. The hnRNP H protein, which is 78% identical
to hnRNP F and also interacts with both CBP80 and CBP20 in vitro, does not discriminate between naked RNA
and CBC-RNA complexes, showing that this effect is specific. Depletion of hnRNP F from HeLa cell nuclear
extract decreases the efficiency of pre-mRNA splicing, a defect which can be partially compensated by addition
of recombinant hnRNP F. Thus, hnRNP F is required for efficient pre-mRNA splicing in vitro and may
participate in the effect of CBC on pre-mRNA splicing.
The splicing of mRNA precursors takes place in multicom-
osome contains five essential U snRNAs and a large number of
both snRNA-associated and free proteins (32, 49, 55). Splice-
osome assembly in vitro occurs as an ordered stepwise process.
The earliest step appears to be relatively conserved between
Saccharomyces cerevisiae and humans. In yeast it involves the
ATP-independent recognition of the 5? splice site by the U1
snRNP and of the branch point region of the intron by a factor
whose identity is not yet established (1, 34, 60, 61). This com-
plex commits yeast pre-mRNAs to splicing and was thus
named the commitment complex. In humans, the apparently
homologous complex is the E complex, which forms in the
absence of ATP and involves U1 snRNP binding at the 5?
splice site and U2AF binding to the polypyrimidine tract (2,
47). A further component of both complexes is the nuclear
cap-binding protein complex (CBC) (8, 35, 36).
CBC is composed of two subunits, CBP80 and CBP20 (23,
24, 28, 29). It binds either to an RNA cap dinucleotide,
m7GpppN, or to capped RNAs (26, 50). CBC is bifunctional.
In addition to its stimulatory role in the formation of E or
commitment complex, described above, vertebrate CBC has
been shown to play an important role in the export of U
snRNAs from the cell nucleus in the first stage of U snRNP
assembly (23; see reference 25 for a review). The precise role
of CBC in RNA export is not defined, but CBC is translocated
through the nuclear pore complex together with the RNA (67).
Removal of CBC from the RNA in the cytoplasm appears to
involve binding of importin, the nuclear protein import recep-
tor (17, 18), to the CBC-RNA complex. It is unclear at present
if the transport function of CBC is conserved; however, both
vertebrate and yeast CBCs are found in abundant nuclear
complexes together with importin-?, the nuclear localization
signal-binding subunit of the protein import receptor. Disso-
ciation of both yeast and human CBCs from capped RNA
requires the interaction of this complex with importin-? (17),
providing circumstantial evidence of a conserved transport
hnRNP proteins are a diverse family of highly abundant (45)
nuclear proteins whose common property is their association in
the nucleus with poly(A)-containing RNA, i.e., pre-mRNA and
mRNA (11). As a family, these proteins share with CBC the
characteristic of having been functionally implicated in both
pre-mRNA splicing and RNA export from the nucleus. The
involvement of mammalian hnRNP proteins in splicing was
first suggested by the fact that monoclonal antibodies directed
against the hnRNP C proteins inhibited splicing either when
added to mammalian cell nuclear extracts or when used to
deplete the extracts (7). Similarly, monoclonal antibodies
against hnRNP M inhibit splicing upon their addition to HeLa
cell nuclear extracts in vitro (16). More specific roles for ver-
tebrate hnRNP proteins in splicing regulation have also been
discovered. A role for the hnRNP F protein as one factor
required for regulation of neuron-specific splicing of the c-src
N1 exon has been defined (48). A second mammalian hnRNP
protein, hnRNP A1, has been shown to specifically affect splice
site choice by influencing discrimination between alternative 5?
splice sites (5, 44). The Drosophila melanogaster hnRNP pro-
tein, hrp48 (40, 42), has a role in ensuring the germ cell spec-
ificity of splicing of a P element intron (62). Mutation of other
members of the Drosophila family, like hrp40 (30, 39) and
Rb97D (27), has specific effects on dorsoventral axis formation
and spermatogenesis, respectively. Although the molecular ba-
sis for these defects is not clear, they may also result from
alterations in pre-mRNA splicing.
There is also good evidence for a role for mammalian
hnRNP A1 in the nuclear export of mRNA. hnRNP A1, like
several other hnRNP proteins, shuttles continuously between
the nucleus and cytoplasm, and it is found in both compart-
ments in association with poly(A)-containing RNA (52, 53). A
38-amino-acid region of hnRNP A1, called M9, is capable of
directing active export out of, as well as active import into, the
nucleus (45, 63, 68). Further, microinjection of saturating
amounts of hnRNP A1, but not of a mutant protein lacking the
M9 domain, into Xenopus laevis oocyte nuclei results in specific
competitive inhibition of the mRNA export pathway (22).
Taken together, these data suggest that hnRNP A1 is likely to
have an active role in the transport of mRNA to the cytoplasm.
hnRNP A1 is unlikely to be the only member of the family
involved in this function. hnRNP L, for example, has been
proposed to have a stimulatory effect on the nuclear export of
* Corresponding author. Mailing address: European Molecular Bi-
ology Laboratory, Meyerhofstrasse 1, D-69117, Heidelberg, Germany.
Phone: 6221 387 393. Fax: 6221 387 518. E-mail: MATTAJ@EMBL
mRNAs transcribed from genes lacking introns (37). Several
other mammalian hnRNP proteins show shuttling behavior
similar to that of A1 (12) and may also be involved in mRNA
export. There is also accumulating evidence that Npl3p/Nop3p,
a candidate yeast hnRNP protein, also has a role in mRNA
In an attempt to deepen our understanding of the mode of
action of CBC, we carried out a screen for proteins with which
FIG. 1. A three-hybrid screen with CBC selects a C-terminal fragment of hnRNP F. (A) Cartoon showing the principle of the CBC three-hybrid interaction screen
described in the text. The human CBP80 protein was coexpressed in S. cerevisiae with the GAL4DBD-CBP20 fusion. A human GAL4 activation domain-tagged cDNA
library was screened for proteins interacting with the CBC proteins. The interaction between one or both of the CBC proteins and X activates the transcription of the
two reporter genes (HIS3 and lacZ) whose structure is schematically shown at the bottom. Positive clones were selected on the basis of the expression of both reporter
genes. (B) Clone 3 encodes the C-terminal two-thirds of the hnRNP F protein. Diagramatic structure of the partial cDNA (hnRNP F c-ter) rescued from a positive
clone in the three-hybrid screen. The three RNA-binding domains (RBDs) of hnRNP F (20, 43) are shown as stippled boxes, and the RNP1 and RNP2 consensus
sequences are shown as black bars. The other regions of the protein are shown as thick, patterned lines. The amino acid positions at the termini of hnRNP F c-ter and
full-length hnRNP F are indicated.
2588GAMBERI ET AL.MOL. CELL. BIOL.
it interacts. Among several candidate proteins identified was
hnRNP F. Given the known functions of other hnRNP pro-
teins, we examined the interaction between hnRNP F and CBC
in some detail. We show that hnRNP F can interact separately
with both CBP80 and CBP20. hnRNP F binds preferentially to
CBC-RNA complexes compared to naked RNA, providing ev-
idence that the hnRNP F-CBC interaction has functional conse-
quences. We also show that, in addition to its previously defined
specific role in splicing (48), depletion of hnRNP F has a more
general effect on splicing efficiency in vitro.
FIG. 2. CBC and hnRNP F interact in vitro. (A) In vitro-translated35S-labelled hnRNP F c-ter (C-term) and full-length hnRNP F were incubated with E. coli lysates
from strains expressing His-tagged human CBP20 (lanes 1 to 4); the bound fractions were precipitated via Ni-NTA agarose resin, washed with buffer containing 25 mM
imidazole, and eluted by boiling in protein sample buffer. The same in vitro-translated polypeptides were mixed and incubated with E. coli lysates expressing untagged
human CBP80 (lanes 5 to 9), anti-CBP80 immune serum, and protein A-Sepharose. The bound fractions were recovered after washing in low-salt buffer and eluted
as described above. One-fifth of the eluted fraction was resolved by SDS-PAGE. Lanes 11 to 15 contain the input polypeptides (1/50 of the amount in the binding
reaction mixture). Lanes 3 and 7 show negative controls (DHFR), and lanes 4, 8, and 9 are positive controls (CBP80 and CBP20, respectively, for the CBP20 and CBP80
lysates). Lane 9 shows a binding reaction mixture identical to that in lane 8 except for the use of twice as much anti-CBP80 antiserum, a control to show that the amount
of anti-CBP80 antibody used was not limiting. The migration positions of the protein molecular weight marker (MWM) are shown. On the right, lanes 5? to 7? show
a five-times-longer exposure of lanes 5 to 7, respectively. (B) In vitro-translated CBP20 and CBP80 were incubated with increasing amounts of a lysate of E. coli
expressing His-tagged hnRNP F and treated as described for panel A. One-fifth of each of the bound fractions was resolved by SDS-PAGE. Lanes 2 to 4, CBP20 alone;
lanes 5 to 7, CBP80 alone; lanes 8 to 10, a mixture of CBP20 and CBP80; lane 1, DHFR (negative control). Lane 11 shows a positive control: in vitro-translated hnRNP
F selected from a (His6)-CBP20 lysate. One fiftieth of each unbound fraction was loaded to assay for protein degradation (lanes 12 to 22). Lanes 23 to 26 contain 1/50
of the input polypeptides. The migration positions of the protein molecular weight marker (MWM) are on the right.
VOL. 17, 1997 CBC INTERACTS WITH hnRNP F2589
MATERIALS AND METHODS
All restriction enzymes, T4 DNA polymerase, and T4 DNA ligase were pur-
chased from New England BioLabs. The DNA manipulations and cloning pro-
cedures were done as previously described (57). T7, T3, and SP6 RNA poly-
merases were from Promega, and AmpliTaq DNA polymerase was from Perkin-
Elmer Cetus. The cap analog m7GpppG (9) was a gift from E. Darzynkiewicz.
The Protein A Sepharose Fast Flow was purchased from Pharmacia.
RNA probes were labelled according to standard protocols (57) with32P-
labelled nucleotide triphosphates from Amersham International. DNA se-
quences were determined by the dideoxynucleotide chain termination method
(58) with a kit purchased from Pharmacia.
Two- and three-hybrid screens. The bait plasmids were constructed as follows:
the CBP20 open reading frame was amplified with AmpliTaq DNA polymerase
and the oligonucleotides 5?CTCGAATTCATGTCGGGTGGCCTCCTG3? and
5?CACCTCGAGCTGGTTCTGTGCCAGTTTTCC3?, the fragment was then
cleaved with EcoRI and XhoI, and the end was repaired with T4 DNA polymer-
ase and cloned in the plasmid pAS2 (13) at the end-repaired NdeI site to give
pAS2-CBP20. The nucleotide sequence was checked. The cDNA for CBP80, as
a BamHI-XhoI fragment, was first inserted between the yeast alcohol dehydro-
genase promoter and terminator cassettes in the pVT102U plasmid (65), and the
resulting transcription unit was excised with SphI, end repaired with T4 DNA
polymerase, and cloned in the pAS2-CBP20 plasmid at the end-repaired SacI site
to give pAS2-CBC.
The S. cerevisiae Y190 strain (13) was first transformed with the pAS2-CBC
bait plasmid according to a standard lithium acetate protocol (21) to give the
190-CBC strain. Expression of the bait proteins was checked by Western blotting.
The strain was then transformed with a human cDNA library in the vector
pACTII (13) according to a modified protocol (15a).
Transformants (6 ? 106) were selected on SD plates lacking histidine, leucine,
and tryptophan and containing 25 or 50 mM 3-aminotriazole (Sigma) and 65 mg
of X-Gal (5-bromo-4-chloro-3-indolyl-?-D-galactopyranoside) per ml.
Bait plasmids were cured from the positive cells by active selection on cyclo-
heximide-containing media and were then mated to Y187 cells transformed with
unrelated baits (CDK2, SNF1, and p53; the gift of S. Elledge), with the empty
pAS2 plasmid or with the original bait plasmid. Library plasmids were extracted
from positive yeast cells as previously described (54) and transformed into
Escherichia coli cells by electroporation.
Cloning and production of recombinant hnRNP F and hnRNP H proteins.
The hnRNP F cDNA was amplified by PCR from human cDNA with the
GGAG3? and 5?CTCTAGATCTGGTACCGTCATAGCCACCCATGCTGTT
C3? as primers, digested with NcoI and BglII, and cloned in the vector pQE60
(Qiagen). The resulting C-terminal His-tagged protein was expressed in E. coli
M15(pREP4) (66). The bacterial cells were grown in Luria broth containing 100
?g of ampicillin per ml; the optimal induction time was determined by checking
the amount of protein produced in a time course experiment. The pelleted cells
were then lysed in PGK buffer (50 mM sodium phosphate [pH 7.2], 100 mM KCl,
10% glycerol, 0.5% Triton X-100) containing 6 M guanidine–HCl and lysed with
a French press. The recombinant protein was then purified by binding to nickel
nitriloacetate agarose (Ni-NTA; Qiagen) as previously described (23). The ac-
tivity of each eluted fraction was then tested for RNA binding by gel retardation
assays. Only fractions displaying maximal RNA binding were capable of restoring
splicing activity to extracts depleted of hnRNP F (“depleted extracts”), even if
these fractions contained less hnRNP F protein. The hnRNP F cDNA was then
cloned as a BamHI-EcoRI fragment in the pGEX-2T plasmid, and the protein
was expressed and purified from E. coli cells according to standard procedures
(64). The hnRNP F cDNA was also cloned as a BglII fragment in the pBluescript
II SK(?) vector (Stratagene) to give the plasmid pBS-hnRNPFT3 that was used
to transcribe and translate hnRNP F in vitro. The hnRNP F c-ter protein was
produced by PCR amplification of the insert contained in the clone originally
isolated in the screen with the primers 5?TAATACGACTCACTATAGGGAG
ACCACATGGATGATGTATATAACTATCATTTC3? and 5?CTACCAGAAT
TCGGCATGCCGGTAGAGGTGTGGTCA3?. The hnRNP H cDNA was am-
plified with AmpliTaq polymerase from human cDNA with the oligonucleotides
FIG. 3. Interaction between CBC and hnRNP H. (A) The amino acid se-
quence of hnRNP F. Residues that are identical in hnRNP F and hnRNP H are
boxed. Positions of insertion of amino acids in hnRNP H compared to hnRNP F
are also indicated. (B) CBC and hnRNP H interact in vitro. In vitro-translated,
[35S]methionine-labelled hnRNP H was tested for binding to His-tagged CBP20
or untagged CBP80 as described in the legend to Fig. 2. Lanes 1 to 3 show the
proteins selected by (His)6-CBP20 (hnRNP H, CBP80, and DHFR, respectively).
Lanes 4 to 6 show the proteins selected by CBP80 and the anti-CBP80 antiserum
(hnRNP H, CBP20, and DHFR, respectively). Lanes 7 to 10 contain the refer-
ence input polypeptides, hnRNP H, CBP80, CBP20, and DHFR, respectively.
The migration positions of the protein molecular weight markers (MWM) are on
2590 GAMBERI ET AL.MOL. CELL. BIOL.
5?GCAGGATCCATGATGTTGGGCACGGAAGG3? and 5?CGAGCCATGG
TTACCTATGCAATGTTTGATTG3? and cloned as a BamHI-NcoI fragment in
the pRSETA (Invitrogen) vector to give the plasmid pRSET-RNPH. The
hnRNP H cDNA was also cloned as a BamHI-XhoI fragment in the pBluescript
II SK(?) vector to give the plasmid pBS-hnRNPHT3 that was used to transcribe
and translate hnRNP H in vitro.
Production of antibodies against hnRNP F. Polyclonal antibodies were raised
against an N-terminal histidine-tagged recombinant protein containing amino
acids 283 to 415 of hnRNP F according to standard protocols (19). The anti-
bodies were purified over an affinity column made by coupling a glutathione
S-transferase-hnRNP F fusion, encoding the same amino acids, to Affigel 10
Electrophoretic mobility retardation assay. The 77-nucleotide RNA used as a
probe was obtained by in vitro transcription of U1 SII? (59). Binding reactions
were done in a 10-?l volume with 10% glycerol, 20 mM Tris-HCl (pH 7.4), 0.125
mM EDTA, 60 mM KCl, 5 mM dithiothreitol, 5 ?g of yeast tRNA, 2 U of
RNasin per ?l, and approximately 5 ? 104cpm of labelled RNA. Samples were
incubated at 30?C for 30 min and then loaded on 6% native polyacrylamide
(60:1) gels in 1? Tris-borate-EDTA. The gels were run in 0.5? Tris-borate-
EDTA at 10 V/cm and at room temperature.
Pull down assays and immunoprecipitations. The lysates from E. coli strains
expressing recombinant proteins were prepared in PGK buffer (see above) at a
ratio of 1 g (wet weight) of cells to 25 ml of buffer, and the cell suspension was
sonicated extensively and centrifuged for 15 min at 12,000 ? g at 4?C. The
cleared supernatant was then used for the binding assays.
The amount of lysate used in each interaction assay was variable, depending on
the expression level of the different proteins, and was determined empirically.
When the lysates had to be diluted more than twofold for the reaction, a blank
lysate, prepared from bacteria not expressing recombinant proteins, was used as
The proteins were in vitro translated either from RNAs transcribed in vitro
according to standard procedures (57) or from plasmids by using the TNT
coupled transcription-translation system (Promega) according to the manufac-
The assays were performed in a 500-?l volume by slow rotation of the mixtures
for 2 h at room temperature. The bound fraction was washed with PGK buffer
containing 25 mM imidazole four times for 5 min each time and then eluted by
boiling in 2? protein sample buffer (19).
Immunoprecipitations were carried out in IPP150 buffer (10 mM Tris-HCl [pH
8.0], 150 mM NaCl, 0.1% Nonidet P-40) from a mixture including bacterial lysate
containing rCBP80, in vitro-translated proteins, anti-CBP80 immune serum (24),
and protein A-Sepharose (Pharmacia); samples were incubated with slow rota-
tion for 2 h at room temperature. Beads with bound proteins were washed with
IPP150 buffer as described above and eluted by boiling in 2? protein sample
buffer. One-fifth of the bound fractions was resolved by sodium dodecyl sulfate-
polyacrylamide gel electrophoresis (SDS-PAGE) according to standard proce-
dures (19), and the gels were treated with Entensify (Dupont) according to the
recommendations of the manufacturer.
Immunodepletion and in vitro splicing assays. The immunoaffinity matrix was
prepared as previously described (24), except that a different ratio of serum to
Protein A Sepharose Fast Flow was used (1:6.5). The beads for the mock
depletion were prepared by coupling immunoglobulins G from the serum of a
nonimmune rabbit to the resin. HeLa splicing extracts were prepared as de-
scribed previously (10) and were brought to 500 mM KCl prior to depletion.
Depletions were carried out at 500 mM KCl as described previously (24), except
that a 1:2.5 ratio of beads to extract was used. Extracts were then dialyzed against
buffer D containing 100 mM KCl.
The levels of depletion were estimated by Western blotting with anti-hnRNP
F monoclonal antibody 7C2 (43) or with anti-CBP80 immune serum used as
described previously (6, 24).
Splicing assays were carried out essentially as previously described (24), except
that the optimal MgCl2concentration was determined empirically for the extract.
The spliced products were resolved on 10% polyacrylamide–urea gels (40:1). The
pre-mRNA splicing probes for the adenovirus major late and the chicken ?-crys-
tallin mRNAs were transcribed from linearized plasmids (pBSAd1  and
pSP14-15 , respectively), using T3 and SP6 RNA polymerases.
FIG. 4. hnRNP F binds preferentially to a CBC-RNA complex. (A) Recombinant hnRNP F or recombinant human CBC was incubated with a capped RNA
substrate and analyzed by native gel electrophoresis in order to resolve the different RNA and RNA-protein complexes: free RNA (lane 1), hnRNP F-RNA (lanes 2,
3, 5, and 6), and CBC-RNA complexes (lanes 4, 5, and 6). In the presence of both hnRNP F and CBC, an additional slower-migrating complex was formed (lanes 5
and 6). (B) hnRNP H does not bind CBC-RNA complexes preferentially. The experiment is analogous to that used for panel A, except that recombinant hnRNP H
was used instead of hnRNP F. Free RNA (lane 1), CBC-RNA complexes (lanes 5 and 6), hnRNP H-RNA complexes (lanes 3, 4, 5, and 6), or complexes formed by
incubation of RNA with both CBC and hnRNP H (lanes 5 and 6) are indicated.
VOL. 17, 1997 CBC INTERACTS WITH hnRNP F2591
The three-hybrid screen. The active form of CBC requires
the presence of both subunits (23, 24). In order to identify
factors able to interact with the assembled human CBC, we
coexpressed the large subunit, CBP80, in an untagged form in
the GAL4-CBP20-expressing strain (Fig. 1A). This CBC-ex-
pressing strain was then transformed with a human lymphocyte
cDNA library (13), and a total of 6 ? 106transformants was
plated on selective medium. Five blue colonies were picked,
and the plasmids were rescued in E. coli and retransformed
into the CBC-expressing strain or in Y190 strains containing
other, unrelated baits or the empty vector. Three clones were
identified as being positive in the second screen, and these
plasmids were subjected to further analysis. Sequencing of the
inserted cDNAs revealed that two of them are novel clones;
they will be described in more detail elsewhere. The third clone
encoded the C-terminal two-thirds of the hnRNP F protein
(43) (hnRNP F c-ter) (Fig. 1B).
hnRNP F interacts with CBP80 and CBP20. The abilities of
hnRNP F c-ter and of full-length hnRNP F to bind either CBC
or CBP20 or CBP80 individually in vitro were next tested. An
E. coli lysate from a strain expressing (His6)-CBP20 was incu-
bated with in vitro-translated hnRNP F and hnRNP F c-ter.
Dihydrofolate reductase (DHFR) and CBP80 were used as a
negative control and a positive control, respectively. Proteins
bound to CBP20 were selected on Ni-NTA agarose. None of
the proteins bound to Ni-NTA agarose in the absence of
(His6)-CBP20 (data not shown). The bound fraction was re-
covered and separated by SDS-PAGE (Fig. 2A, lanes 1 to 4).
As expected, CBP80 was found to bind to CBP20, while DHFR
was not detectably associated (Fig. 2A, lanes 3 and 4). Both
hnRNP F c-ter and, although with reduced efficiency, full-
length hnRNP F were coprecipitated with (His6)-CBP20 (Fig.
2A, lanes 1 and 2). To test the interaction of hnRNP F with
CBP80, an E. coli lysate from a strain expressing untagged
human CBP80 was incubated with in vitro-translated hnRNP F
c-ter, hnRNP F, DHFR, or CBP20 and anti-CBP80 immune
serum. The bound fractions were recovered by binding to pro-
tein A-Sepharose and then separated electrophoretically (Fig.
2A, lanes 5 to 9). CBP20 was efficiently coprecipitated, and
antibody was in excess (Fig. 2A, lanes 8 and 9). Small quanti-
ties of both hnRNP F c-ter and hnRNP F were specifically
precipitated (Fig. 2A, lanes 5 and 6 and 5? to 7?).
Thus, both hnRNP F and hnRNP F c-ter can interact with
CBP20 and CBP80. It was difficult to assess the relative affin-
ities of hnRNP F for CBP20 and CBP80, due to the differences
in the assay conditions. The weaker signal obtained for the
binding to CBP80 could, for example, be a reflection of inter-
ference of bound antibodies with the interaction between
CBP80 and hnRNP F.
To confirm the binding interactions detected as described
above, we next expressed a (His6)-hnRNP F protein in E. coli
and used the corresponding lysate to examine binding to in
vitro-translated CBP20 and CBP80 by means of Ni-NTA aga-
rose. CBP20 was bound in the presence of various amounts of
hnRNP F-containing bacterial lysate (Fig. 2B, lanes 2 to 4 and
13 to 15). The bound fractions in the case of either CBP80
alone or the mixture of CBP20 and CBP80 are shown (Fig. 2B,
lanes 5 to 7 and 16 to 18 and 8 to 10 and 19 to 21, respectively.)
The negative control, DHFR, did not bind to hnRNP F (Fig.
2B, lane 1). As a positive control, in vitro-translated hnRNP F
binding to (His6)-CBP20 was included (Fig. 2B, lanes 11 and
22). hnRNP F bound CBP20 and CBP80 at comparable levels,
FIG. 5. Identification of hnRNP F- and H-containing RNA complexes. (A) Analysis of hnRNP F-containing complexes. The RNP complexes indicated on the left
were formed as described in the legend to Fig. 4. Identical reactions were performed either in the absence (lanes 1 to 4) or in the presence (lanes 5 to 8) of
affinity-purified polyclonal anti-hnRNP F antibody. (B) Analysis of hnRNP H-containing complexes. The RNP complexes indicated on the left were formed as described
in the legend to Fig. 4. Lane 1, free RNA; lane 2, hnRNP H and RNA; lane 3, CBC and RNA; lane 4, hnRNP H, CBC, and RNA; lane 5, hnRNP H, CBC, RNA,
and affinity-purified anti-hnRNP F antibody.
2592 GAMBERI ET AL.MOL. CELL. BIOL.
whether they were present alone or in combination in the
The addition of increasing amounts of hnRNP F lysate led to
a decreased recovery of both CBPs (Fig. 2B, lanes 4, 7, and 10).
This result is probably explained by the ability of hnRNP F to
dimerize or multimerize, as indicated by the fact that
labelled hnRNP F binds to immobilized (His6)-hnRNP F (data
not shown). In order to exclude the possibility that the inter-
action between CBP20 or CBP80 and hnRNP F was dependent
on binding to RNA, we performed the same assays in the
presence of excess RNase A. Under these conditions, we did
not detect any reduction in the interactions (data not shown).
hnRNP F is very closely related (with identities of 78 and
75%, respectively) to at least two other hnRNP proteins,
FIG. 6. Depletion of HeLa cell extract with anti-hnRNP F antibodies reduces splicing efficiency. (A) Depletion with anti-hnRNP F antibodies. HeLa cell nuclear
extract was passed twice over beads to which affinity-purified antibodies from either nonimmune or immune serum raised against recombinant hnRNP F had been
coupled. Depleted or control extracts were fractionated by SDS-PAGE and analyzed by Western blotting with either monoclonal antibody 7C2, which interacts with
both hnRNP F and hnRNP H (43), or polyclonal anti-CBP80 antiserum (24). Lane 1, nondepleted extract; lanes 2 and 3, mock-depleted extract after one (lane 2) or
two (lane 3) rounds of depletion; lanes 4 and 5, hnRNP F-depleted extract after one (lane 4) or two (lane 5) rounds of depletion. (B) Splicing in hnRNP F-depleted
HeLa nuclear extracts. The splicing of an adenovirus major late (lanes 1 to 3) or a ?-crystallin (lanes 4 to 7) pre-mRNA were analyzed in mock-depleted or depleted
extracts. Lanes 1 and 5, splicing in untreated extract; lanes 3 and 6, splicing in mock-depleted extract; lanes 2 and 7, splicing in hnRNP F-depleted extract. Lane 4
contains the unspliced precursor ?-crystallin RNA. The positions of pre-mRNA, intermediates, and products of the reactions are indicated. (C) Addition of recombinant
hnRNP F (rhnRNP F) increases splicing in a depleted extract. Reactions were performed as described for panel B, using ?-crystallin pre-mRNA as the splicing
substrate. Lane 1, pre-mRNA; lanes 2 and 3, splicing in mock-depleted (m) or depleted (d) extract as indicated; lanes 4 to 9; splicing in mock-depleted (m) or depleted
(d) extracts to which increasing amounts of rhnRNP F protein were added. The splicing reactions were all performed in the same final volume; to achieve identical
reaction volumes, buffer D (10) was added to compensate for the different amounts of recombinant protein added.
VOL. 17, 1997 CBC INTERACTS WITH hnRNP F 2593
hnRNP H and hnRNP H? (20, 43) (Fig. 3A). hnRNP H inter-
action with the CBC proteins was therefore assayed by the
methods used in the experiment whose results are shown in
Fig. 2A. hnRNP H was found to interact with both His-tagged
CBP20 (Fig. 3B, lane 1) and nontagged CBP80 (Fig. 3B, lane
4). Again, DHFR (Fig. 3B, lanes 3 and 6) and CBP80 or
CBP20 (Fig. 3B, lanes 2 and 5, respectively) served as negative
and positive controls, respectively. No DHFR binding was seen
even after longer exposure of the autoradiograph.
CBC affects hnRNP F RNA binding. As an initial test of the
possible functional consequences of interaction between CBC
and hnRNPs F and H, the ability of the proteins to affect each
other’s RNA binding in vitro was tested. Recombinant hnRNP
F and CBC purified from E. coli both bind to a capped RNA
probe (Fig. 4A, lanes 1 to 4). In the presence of CBC, a
significantly larger amount of hnRNP F appeared to bind to
RNA than in its absence (Fig. 4A, lanes 5 and 6). Quantitation
of the effect of CBC on the formation of hnRNP F-containing
complexes in several experiments showed that in the presence
of CBC, between three- and fourfold more hnRNP F com-
plexes were detected. To prove that the additional complex
migrating more slowly than the CBC-RNA complex in lanes 5
and 6, Fig. 4A, was indeed due to hnRNP F binding, the effect
of adding affinity-purified polyclonal antibodies directed
against hnRNP F to the reaction mixture was examined. The
complexes proposed to contain either hnRNP F alone or
hnRNP F plus CBC were either disrupted or further retarded
by the anti-hnRNP F antibodies (Fig. 5A, compare lanes 2 to
4 with lanes 6 to 8), whereas the CBC-RNA complexes were
not greatly affected. In conclusion, hnRNP F binds preferen-
tially to a CBC-RNA complex rather than to RNA alone.
We next determined whether this effect would also be ob-
served with hnRNP H. In contrast to the result obtained with
hnRNP F, the presence of CBC had little or no effect on the
total amount of recombinant hnRNP H-containing RNA com-
plexes (Fig. 4B, lanes 3 to 6). Although some hnRNP H did
bind to CBC-RNA complexes, most was still found in the
hnRNP H-RNA form. The increase in hnRNP H-containing
complexes detectable in the presence of CBC was measured to
be only 1.2-fold. Since the polyclonal antibody raised against
hnRNP F cross-reacts with hnRNP H (data not shown), we
could use this antibody to show that the putative hnRNP H
complexes indeed contained the protein (Fig. 5B, lanes 4 and
5). Thus, even if both hnRNP proteins can interact with the
CBC proteins in solution, only hnRNP F shows preferential
binding to a CBC-RNA complex rather than to naked RNA.
Depletion of hnRNP F affects in vitro splicing efficiency.
CBC has two known functions, in pre-mRNA splicing and U
FIG. 7. Splicing in hnRNP F-depleted extract. (A) Splicing of a ?-crystallin pre-mRNA (lane 1) in mock-depleted (M) or anti-hnRNP F-depleted (D) HeLa cell
nuclear extract in the absence (lanes 2 and 3) or in the presence of increasing quantities (lanes 4 to 7) of purified recombinant hnRNP F. (B) Splicing of an adenovirus
major late pre-mRNA (lane 1) in mock-depleted (M) or anti-hnRNP F-depleted (D) HeLa cell nuclear extract in the absence (lanes 2 and 3) or in the presence of
increasing quantities (lanes 4 to 7) of purified recombinant hnRNP F.
2594GAMBERI ET AL.MOL. CELL. BIOL.
snRNA nuclear export. While hnRNP proteins in general may
affect many aspects of RNA metabolism, the only defined role
for hnRNP F is in a particular example of neuron-specific
pre-mRNA splicing (48). Since the role of CBC in pre-mRNA
splicing seems to be general, it was of interest to determine
whether hnRNP F might also play a more general role. HeLa
cell nuclear extract was therefore depleted of hnRNP F by
repeated passage over beads to which polyclonal antibodies
raised against hnRNP F were bound. This procedure resulted
in a significant depletion of hnRNP F protein as measured by
Western blotting with a monoclonal antibody, 7C2, that inter-
acts with both hnRNP F and hnRNP H (43) (Fig. 6A, lane 5).
As expected from the cross-reactivity mentioned above,
hnRNP H levels in the depleted extract were also reduced. No
detectable reduction in CBP80 (Fig. 6A) was seen, indicating
that the depletion was specific and that no significant fraction
of the CBC in a HeLa cell nuclear extract is tightly associated
with hnRNP F.
The effect of depletion on splicing was initially tested with
two different pre-mRNA substrates, one derived from adeno-
virus (31) and the other derived from the ?-crystallin gene (51,
56). Although mock depletion had some effect on splicing of
both the pre-mRNAs (Fig. 6B, lanes 1, 3, 5, and 6), this effect
was less than that caused by hnRNP F depletion (Fig. 6B, lanes
2 and 7). In neither case was splicing completely inhibited. It
should, however, be noted that hnRNP F was not entirely
removed by depletion (Fig. 6A). Further reduction in hnRNP
F levels could not be achieved under conditions in which pre-
mRNA splicing activity was retained in extracts which had
undergone mock depletion (“mock-depleted extracts”) (data
not shown). The splicing of additional substrates tested, de-
rived from the Xenopus ribosomal protein LI gene or from the
Drosophila transformer or fushi tarazu genes, was also reduced
by the immunodepletion.
If the effect of depletion is specific, then it should be re-
versed by the readdition of hnRNP F protein to the depleted
extract. The splicing of ?-crystallin pre-mRNA was therefore
monitored in mock-depleted or depleted extract to which in-
creasing amounts of recombinant hnRNP F were added.
hnRNP F addition partially restored splicing activity (Fig. 6C,
lanes 2 to 9). Addition of larger quantities of hnRNP F did not
lead to greater recovery of activity but, rather, to increased
inhibition of splicing (data not shown). In view of the incom-
pleteness of both the depletion of activity and of its restoration
by addition of hnRNP F, it was important to document the
reproducibility of the effects seen. In Fig. 7, the effect on
splicing of a ?-crystallin (Fig. 7A) and an adenovirus major late
(Fig. 7B) pre-mRNA are shown. The extracts and preparations
of recombinant hnRNP F used in these experiments are dif-
ferent from those used for Fig. 6, and these experiments dem-
onstrate the reproducibility of the inhibition of hnRNP F ac-
tivity on immunodepletion (Fig. 7 lanes 1 and 3 [both panels])
and the generally partial restoration of activity after addition of
recombinant hnRNP F (Fig. 7 lanes 4 to 7 [both panels]).
When recombinant hnRNP H, which was codepleted with
hnRNP F (Fig. 6A), was added to depleted extract either alone
or in combination with hnRNP F, no effect on splicing activity
was detected (data not shown). In spite of this negative result,
it is possible that additional factors, including members of the
hnRNP F-H subfamily, might be codepleted with hnRNP F
and be responsible for the lack of complete restoration of
activity we saw in most add back experiments. Examination of
the reduction in accumulation of splicing intermediates and
products (see particularly Fig. 6B and C) indicated that both
steps of splicing were affected by hnRNP F depletion. Accu-
mulation of the products of the first step of splicing were not
reduced in parallel with products of the second step, indicating
a more severe effect on the second step than on the first step.
On the other hand, in all of the experiments carried out in
depleted extracts, the amount of first-step products accumu-
lating was insufficient to account for the reduction in second-
step products. The quantities of first-step products observed
were usually virtually identical in mock-depleted and depleted
extracts (Fig. 6B and C). This indicates that the first step in
splicing is also affected by hnRNP F depletion.
Using a variation of the two-hybrid screening method (14),
we have selected several human cDNAs encoding proteins that
interact either with CBP20 or with CBC. One of the proteins
that interacted with CBC was the previously characterized
hnRNP F protein (43). Because of the dual role of CBC in
pre-mRNA splicing and U snRNA nuclear export (8, 24, 35,
36) and since hnRNP proteins have been implicated in both of
these aspects of cellular RNA metabolism (11, 46), we have
characterized this interaction in detail. Interestingly, hnRNP F
can interact with either CBP80 or CBP20 individually. In ad-
dition, hnRNP F appears to interact with CBC as an RNA-
bound heterodimer, as demonstrated by the preferential bind-
ing of hnRNP F to CBC-RNA complexes in vitro.
hnRNP F is a member of a subfamily of hnRNP proteins
which includes at least the F, H, and H? proteins (20, 43).
These proteins are highly related; hnRNPs H and H? are 96%
identical to each other and 78 and 75% identical to hnRNP F,
respectively. In spite of this similarity and in spite of the fact
that hnRNP H interacts with both CBP20 and CBP80 individ-
ually, we did not detect interaction between CBC and hnRNP
H in a more functional context, as measured by the lack of
effect of CBC on hnRNP H binding to RNA. The least similar
regions of the three proteins are located at their C termini, but
further study will be required to define the regions of hnRNPs
F and H required for interaction with the individual CBC
proteins and with RNA-bound CBC.
hnRNP F has previously been shown to be a component of
a complex that is responsible for activation of a splicing event
that results in the inclusion of a neuron-specific exon in the
c-src gene (3, 4, 48). We show here that hnRNP F also plays a
more general role in splicing, since depletion of the protein
results in a reduction of the in vitro splicing efficiency of several
introns tested. In their study of the role of hnRNP F in c-src
splicing, Min et al. (48) failed to see inhibition of splicing of an
adenovirus intron similar to the one used here on addition of
a monoclonal antibody against hnRNP F to HeLa nuclear
extract. Presumably, the different results obtained reflect the
considerable differences in experimental design between the
two studies. In the one case, a specific monoclonal antibody
was added to the splicing extract, and in the other, hnRNP F
protein was immunodepleted from the extract with polyclonal
antibodies. The specificity of at least part of the reduction in
splicing observed on depletion with anti-hnRNP F antibodies
in our study was proven by the partial restoration of splicing
activity observed upon readdition of recombinant hnRNP F.
Since we are unable to totally remove hnRNP F from extracts
under conditions which are consistent with splicing activity in
control extracts, we cannot say whether or not hnRNP F is
essential for in vitro splicing. Previous work has suggested a
general role for hnRNP proteins in pre-mRNA splicing (7, 16),
but hnRNP F is the first individual hnRNP protein for which
such a role has been demonstrated by depletion and adding
back of the protein. In contrast to the lack of compelling
previous evidence for a general role in splicing of a specific
VOL. 17, 1997CBC INTERACTS WITH hnRNP F 2595
hnRNP protein, it is clear that hnRNP F, hnRNP A1, Dro-
sophila hrp40, and probably other Drosophila hnRNP proteins
are involved in the regulation of specific pre-mRNA splicing
events (see Introduction). The existence of both a general (this
study) and at least one intron-specific (48) function for hnRNP
F is reminiscent of the dual role played by some SR splicing
factors, which are required for both constitutive splicing and
the regulation of particular alternative splicing events (re-
viewed in references 15 and 38). Both SR proteins and hnRNP
proteins are families of highly abundant nuclear RNA-binding
proteins, and our data indicate that they may be more func-
tionally similar than has been generally believed.
CBC is required for efficient removal of cap-proximal in-
trons in vitro and in vivo (24, 36). CBC acts to allow efficient
commitment or E complex assembly and, more directly, to
allow efficient binding of U1 snRNP to the 5? splice site (8, 35,
36). It is, however, unclear how this is achieved. CBC does not
appear to interact directly with U1 snRNP, suggesting the
involvement of one or more proteins that mediate its function
in pre-mRNA splicing (36). hnRNP F is an attractive candidate
for such a protein. The fact that hnRNP F binds preferentially
to CBC-RNA complexes rather than to RNA suggests a model
by which CBC might act to seed the formation of a specific
hnRNP–pre-mRNA complex, beginning by CBC favoring the
binding of some hnRNP proteins, like hnRNP F, over others,
like hnRNP H. The RNA-bound hnRNP F would then influ-
ence the binding of subsequent hnRNP proteins, contributing
to the established variability in hnRNP composition on differ-
ent nascent transcripts and pre-mRNAs (2, 41). Further study
will be required to examine the accuracy of this model and its
possible contribution to CBC function. The fact that hnRNP F
depletion affects both steps of splicing whereas the effect of
CBC is restricted to the first step means that, even if hnRNP F
is involved in mediating CBC function, this is unlikely to be its
only function in splicing.
We thank members of the Mattaj and Se ´raphin labs for discussion
and G. J. Arts, L. Engelmeier, P. Fortes, S. Gunderson, M. Luuk-
konen, M. Ohno, M. Polycarpou-Schwarz, J. Salgado, B. Se ´raphin, and
J. Valca ´rcel for comments on the manuscript and useful suggestions.
We also thank G. Dreyfuss for anti-hnRNP F monoclonal antibody
and S. Elledge for the two-hybrid yeast strains, plasmids, and the
human cDNA library. We thank T. Gibson, L. Toldo, and A. Pastore
for help with computational studies and the EMBL Photolab for help
with the imaging.
Chiara Gamberi and Elisa Izaurralde were supported by the Euro-
pean Union Human Capital and Mobility Programme and the Human
Frontiers Science Programme Organisation, respectively.
1. Abovich, N., X. C. Liao, and M. Rosbash. 1994. The yeast MUD2 protein: an
interaction with PRP11 defines a bridge between commitment complexes
and U2 snRNP addition. Genes Dev. 8:843–854.
2. Bennett, M., R. S. Pinol, D. Staknis, G. Dreyfuss, and R. Reed. 1992.
Differential binding of heterogeneous nuclear ribonucleoproteins to mRNA
precursors prior to spliceosome assembly in vitro. Mol. Cell. Biol. 12:3165–
3. Black, D. L. 1991. Does steric interference between splice sites block the
splicing of a short c-src neuron-specific exon in non-neuronal cells? Genes
4. Black, D. L. 1992. Activation of c-src neuron-specific splicing by an unusual
RNA element in vivo and in vitro. Cell 69:795–807.
5. Caceres, J. F., S. Stamm, D. M. Helfman, and A. R. Krainer. 1994. Regu-
lation of alternative splicing in vivo by overexpression of antagonistic splicing
factors. Science 265:1706–1709.
6. Choi, Y. D., and G. Dreyfuss. 1984. Monoclonal antibody characterization of
the C proteins of heterogeneous nuclear ribonucleoprotein complexes in
vertebrate cells. J. Cell Biol. 99:1997–2004.
7. Choi, Y. D., P. J. Grabowski, P. A. Sharp, and G. Dreyfuss. 1986. Hetero-
geneous nuclear ribonucleoproteins: role in RNA splicing. Science 231:
8. Colot, H., F. Stutz, and M. Rosbash. 1996. The yeast splicing factor Mud13p
is a commitment complex component and corresponds to CBP20, the small
subunit of the nuclear cap-binding complex. Genes Dev. 10:1699–1708.
9. Darzynkiewicz, E., I. Ekiel, S. M. Tahara, L. S. Seliger, and A. J. Shatkin.
1985. Chemical synthesis and characterization of 7-methylguanosine ana-
logues. Biochemistry 24:1701–1707.
10. Dignam, J. D., R. M. Lebovitz, and R. G. Roeder. 1983. Accurate transcrip-
tion initiation by RNA polymerase II in a soluble extract from isolated
mammalian nuclei. Nucleic Acids Res. 11:1475–1489.
11. Dreyfuss, G., M. J. Matunis, S. Pin ˜ol-Roma, and C. G. Burd. 1993. hnRNP
proteins and the biogenesis of mRNA. Annu. Rev. Biochem. 62:289–321.
12. Dreyfuss, G., M. S. Swanson, and S. Pin ˜ol-Roma. 1988. Heterogeneous
nuclear ribonucleoprotein particles and the pathway of mRNA formation.
Trends Biochem. Sci. 13:86–91.
13. Durfee, T., K. Becherer, P. L. Chen, S. H. Yeh, Y. Yang, A. E. Kilburn, W. H.
Lee, and S. J. Elledge. 1993. The retinoblastoma protein associates with the
protein phosphatase type 1 catalytic subunit. Genes Dev. 7:555–569.
14. Fields, S., and O. Song. 1989. A novel genetic system to detect protein-
protein interactions. Nature 340:245–246.
15. Fu, X. D. 1995. The superfamily of arginine/serine-rich splicing factors. RNA
15a.Gamberi, C. Unpublished data.
16. Gattoni, R., D. Mahe ´, P. Mahl, N. Fischer, M.-G. Mattei, J. Stevenin, and
J.-P. Fuchs. 1996. The human hnRNP-M proteins: structure and relation
with early heat shock-induced splicing arrest and chromosome mapping.
Nucleic Acids Res. 24:2535–2542.
17. Go ¨rlich, D., R. Kraft, S. Kostka, F. Vogel, E. Hartmann, R. A. Laskey, I. W.
Mattaj, and E. Izaurralde. 1996. Importin provides a link between nuclear
protein import and U snRNA export. Cell, 87:21–32.
18. Go ¨rlich, D., and I. W. Mattaj. 1996. Nucleocytoplasmic transport. Science
19. Harlow, E., and D. Lane. 1988. Antibodies: a laboratory manual. Cold Spring
Harbor Laboratory, Cold Spring Harbor, N.Y.
20. Honore ´, B., H. H. Rasmussen, H. Vorum, K. Dejgaard, X. Liu, P. Gromov,
P. Madsen, B. Gesser, N. Tommerup, and J. E. Celis. 1995. Heterogeneous
nuclear ribonucleoproteins H, H?, and F are members of a ubiquitously
expressed subfamily of related but distinct proteins encoded by genes map-
ping to different chromosomes. J. Biol. Chem. 270:28780–28789.
21. Ito, H., Y. Fukuda, K. Murata, and A. Kimura. 1983. Transformation of
intact yeast cells treated with alkali cations. J. Bacteriol. 153:163–168.
22. Izaurralde, E., A. Jarmolowski, C. Beisel, I. W. Mattaj, G. Dreyfuss, and U.
Fischer. A role for the hnRNP A1 M9-transport signal in mRNA nuclear
export. J. Cell Biol., in press.
23. Izaurralde, E., J. Lewis, C. Gamberi, A. Jarmolowski, C. McGuigan, and
I. W. Mattaj. 1995. A cap-binding protein complex mediating U snRNA
export. Nature 376:709–712.
24. Izaurralde, E., J. Lewis, C. McGuigan, M. Jankowska, E. Darzynkiewicz, and
I. W. Mattaj. 1994. A nuclear cap binding protein complex involved in
pre-mRNA splicing. Cell 78:657–668.
25. Izaurralde, E., and I. W. Mattaj. 1995. RNA export. Cell 81:153–159.
26. Izaurralde, E., J. Stepinski, E. Darzynkiewicz, and I. W. Mattaj. 1992. A cap
binding protein that may mediate nuclear export of RNA polymerase II-
transcribed RNAs. J. Cell Biol. 118:1287–1295.
27. Karsch, M. I., and S. R. Haynes. 1993. The Rb97D gene encodes a potential
RNA-binding protein required for spermatogenesis in Drosophila. Nucleic
Acids Res. 21:2229–2235.
28. Kataoka, N., M. Ohno, K. Kangawa, Y. Tokoro, and Y. Shimura. 1994.
Cloning of a complementary DNA encoding an 80 kilodalton nuclear cap
binding protein. Nucleic Acids Res. 22:3861–3865.
29. Kataoka, N., M. Ohno, I. Moda, and Y. Shimura. 1995. Identification of the
factors that interact with NCBP, an 80 kDa nuclear cap binding protein.
Nucleic Acids Res. 23:3638–3641.
30. Kelley, R. L. 1993. Initial organization of the Drosophila dorsoventral axis
depends on an RNA-binding protein encoded by the squid gene. Genes Dev.
31. Konarska, M. M., and P. A. Sharp. 1987. Interactions between small nuclear
ribonucleoprotein particles in formation of spliceosomes. Cell 49:763–774.
32. Lamm, G. M., and A. I. Lamond. 1993. Non-snRNP protein splicing factors.
Biochim. Biophys. Acta 1173:247–265.
33. Lee, M. S., M. Henry, and P. A. Silver. 1996. A protein that shuttles between
the nucleus and the cytoplasm is an important mediator of RNA export.
Genes Dev. 10:1233–1246.
34. Legrain, P., B. Se ´raphin, and M. Rosbash. 1988. Early commitment of yeast
pre-mRNA to the spliceosome pathway. Mol. Cell. Biol. 8:3755–3760.
35. Lewis, J. D., D. Go ¨rlich, and I. W. Mattaj. 1996. A yeast cap-binding protein
complex (yCBC) acts at an early step in pre-mRNA splicing. Nucleic Acids
36. Lewis, J. D., E. Izaurralde, A. Jarmolowski, C. McGuigan, and I. W. Mattaj.
1996. A nuclear cap-binding complex facilitates association of U1 snRNP
2596 GAMBERI ET AL.MOL. CELL. BIOL.
with the cap-proximal 5? splice site. Genes Dev. 10:1683–1698. Download full-text
37. Liu, X., and J. E. Mertz. 1995. hnRNP L binds a cis-acting RNA sequence
element that enables intron-dependent gene expression. Genes Dev. 9:1766–
38. Manley, J. L., and R. Tacke. 1996. SR proteins and splicing control. Genes
39. Matunis, E. L., R. Kelley, and G. Dreyfuss. 1994. Essential role for a het-
erogeneous nuclear ribonucleoprotein (hnRNP) in oogenesis: hrp40 is ab-
sent from the germ line in the dorsoventral mutant squid. Proc. Natl. Acad.
Sci. USA 91:2781–2784.
40. Matunis, E. L., M. J. Matunis, and G. Dreyfuss. 1992. Characterization of
the major hnRNP proteins from Drosophila melanogaster. J. Cell Biol.
41. Matunis, E. L., M. J. Matunis, and G. Dreyfuss. 1993. Association of indi-
vidual hnRNP proteins and snRNPs with nascent transcripts. J. Cell Biol.
42. Matunis, M. J., E. L. Matunis, and G. Dreyfuss. 1992. Isolation of hnRNP
complexes from Drosophila melanogaster. J. Cell Biol. 116:245–255.
43. Matunis, M. J., J. Xing, and G. Dreyfuss. 1994. The hnRNP F protein:
unique primary structure, nucleic acid-binding properties, and subcellular
localization. Nucleic Acids Res. 22:1059–1067.
44. Mayeda, A., and A. R. Krainer. 1992. Regulation of alternative pre-mRNA
splicing by hnRNP A1 and splicing factor SF2. Cell 68:365–375.
45. Michael, W. M., M. Choi, and G. Dreyfuss. 1995. A nuclear export signal in
hnRNP A1: a signal-mediated, temperature-dependent nuclear protein ex-
port pathway. Cell 83:415–422.
46. Michael, W. M., H. Siomi, M. Choi, S. Pin ˜ol-Roma, S. Nakielny, Q. Liu, and
G. Dreyfuss. 1995. Signal sequences that target nuclear import and nuclear
export of pre-mRNA binding proteins. Cold Spring Harbor Symp. Quant.
47. Michaud, S., and R. Reed. 1991. An ATP-independent complex commits
pre-mRNA to the mammalian spliceosome assembly pathway. Genes Dev.
48. Min, H., R. C. Chan, and D. L. Black. 1995. The generally expressed hnRNP
F is involved in a neural-specific pre-mRNA splicing event. Genes Dev.
49. Moore, M., C. Query, and P. Sharp. 1993. Splicing of precursors to messen-
ger RNAs by the spliceosome, p. 303–357. In R. Gesteland and J. Atkins
(ed.), The RNA world. Cold Spring Harbor Laboratory Press, Cold Spring
50. Ohno, M., N. Kataoka, and Y. Shimura. 1990. A nuclear cap binding protein
from HeLa cells. Nucleic Acids Res. 18:6989–6995.
51. Ohno, M., H. Sakamoto, K. Yasuda, T. S. Okada, and Y. Shimura. 1985.
Nucleotide sequence of a chicken delta-crystallin gene. Nucleic Acids Res.
52. Pin ˜ol-Roma, S., and G. Dreyfuss. 1991. Transcription-dependent and tran-
scription-independent nuclear transport of hnRNP proteins. Science 253:
53. Pin ˜ol-Roma, S., and G. Dreyfuss. 1992. Shuttling of pre-mRNA binding
proteins between nucleus and cytoplasm. Nature 355:730–732.
54. Robzyk, K., and Y. Kassir. 1992. A simple and highly efficient procedure for
rescuing autonomous plasmids from yeast. Nucleic Acids Res. 20:3790.
55. Rymond, B. C., and M. Rosbash. 1992. Yeast pre-mRNA splicing, p. 143–
192. In J. R. Pringle and J. R. Broach (ed.), The molecular and cellular
biology of the yeast Saccharomyces. Cold Spring Harbor Laboratory Press,
Cold Spring Harbor, N.Y.
56. Sakamoto, H., M. Ohno, K. Yasuda, K. Mizumoto, and Y. Shimura. 1987. In
vitro splicing of a chicken delta-crystallin pre-mRNA in a mammalian nu-
clear extract. J. Biochem. (Tokyo) 102:1289–1301.
57. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a
laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold
Spring Harbor, N.Y.
58. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing with
chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463–5467.
59. Scherly, D., W. Boelens, W. van Venrooij, N. A. Dathan, J. Hamm, and I. W.
Mattaj. 1989. Identification of the RNA binding segment of human U1 A
protein and definition of its binding site on U1 snRNA. EMBO J. 8:4163–
60. Se ´raphin, B., L. Kretzner, and M. Rosbash. 1988. A U1 snRNA:pre-mRNA
base pairing interaction is required early in yeast spliceosome assembly but
does not uniquely define the 5? cleavage site. EMBO J. 7:2533–2538.
61. Se ´raphin, B., and M. Rosbash. 1989. Identification of functional U1 snRNA-
pre-mRNA complexes committed to spliceosome assembly and splicing. Cell
62. Siebel, C. W., R. Kanaar, and D. C. Rio. 1994. Regulation of tissue-specific
P element pre-mRNA splicing requires the RNA-binding protein PSI. Genes
63. Siomi, H., and G. Dreyfuss. 1995. A nuclear localization domain in the
hnRNP A1 protein. J. Cell Biol. 129:551–560.
64. Smith, D. B., and K. S. Johnson. 1988. Single-step purification of polypep-
tides expressed in Escherichia coli as fusions with glutathione S-transferase.
65. Vernet, T., D. Dignard, and D. Y. Thomas. 1987. A family of yeast expression
vectors containing the phage f1 intergenic region. Gene 52:225–233.
66. Villarejo, M. R., and I. Zabin. 1974. Beta-galactosidase from termination
and deletion mutant strains. J. Bacteriol. 120:466–474.
67. Visa, N., E. Izaurralde, J. Ferreira, B. Daneholt, and I. W. Mattaj. 1996. A
nuclear cap-binding complex binds Balbiani ring pre-mRNA cotranscription-
ally and accompanies the ribonucleoprotein particle during nuclear export. J.
Cell Biol. 133:5–14.
68. Weighardt, F., G. Biamonti, and S. Riva. 1995. Nucleocytoplasmic distribu-
tion of human hnRNP proteins: a search for the targeting domains in hnRNP
A1. J. Cell Sci. 108:545–555.
VOL. 17, 1997 CBC INTERACTS WITH hnRNP F2597