The Human hnRNP-M Proteins: Structure and Relation with Early Heat Shock-Induced Splicing Arrest and Chromosome Mapping (original) (raw)

Journal Article

,

Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS/INSERM/ULP

,

BP 163, 67404 Illkirch Cedex, C.U. De Strasbourg, France

Search for other works by this author on:

,

Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS/INSERM/ULP

,

BP 163, 67404 Illkirch Cedex, C.U. De Strasbourg, France

Search for other works by this author on:

,

Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS/INSERM/ULP

,

BP 163, 67404 Illkirch Cedex, C.U. De Strasbourg, France

Search for other works by this author on:

,

Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS/INSERM/ULP

,

BP 163, 67404 Illkirch Cedex, C.U. De Strasbourg, France

Search for other works by this author on:

,

1

U 242 de l'INSERM, Hôpital des Enfants, Groupe Hospitalier de la Timone

,

13385 Marseille Cedex 5, France

Search for other works by this author on:

,

Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS/INSERM/ULP

,

BP 163, 67404 Illkirch Cedex, C.U. De Strasbourg, France

Search for other works by this author on:

Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS/INSERM/ULP

,

BP 163, 67404 Illkirch Cedex, C.U. De Strasbourg, France

* To whom correspondence should be addressed

Search for other works by this author on:

• PDF
• Split View
• • Article contents
  • Figures & tables
  • Video
  • Audio
  • Supplementary Data
• Cite

Cite

Renata Gattoni, Dominique Mahé, Philippe Mähl, Nadine Fischer, Marie-Geneviève Mattei, James Stévenin, Jean-Paul Fuchs, The Human hnRNP-M Proteins: Structure and Relation with Early Heat Shock-Induced Splicing Arrest and Chromosome Mapping, Nucleic Acids Research, Volume 24, Issue 13, 1 July 1996, Pages 2535–2542, https://doi.org/10.1093/nar/24.13.2535
Close

Navbar Search Filter Mobile Enter search term Search

Abstract

With anti-hnRNP monoclonal antibody 6D12 we previously showed in HeLa cells that as early as 10 min after the onset of a heat shock at 45°C, a 72.5–74 kDa antigen doublet leaves the hnRNPs and strongly associates with the nuclear matrix, the effect being reversed after a 6 h recovery at 37°C. cDNA cloning and sequencing enabled us to identify these antigens as hnRNP-M proteins and further to show that the correct sequence differs by an 11 amino acid stretch from the originally published sequence. We also show that monoclonal antibodies raised against synthetic hnRNP-M peptides can directly inhibit in vitro splicing. Furthermore, stressing cells at 45°C for 10 min is sufficient to abolish the splicing capacity of subsequently prepared nuclear extracts which, interestingly, do not contain the hnRNP-M proteins any more. Taken together, our data suggest that these proteins are involved in splicing as well as in early stress-induced splicing arrest. Further in situ hybridization assays located the hnRNP-M encoding gene on human chromosome 19.

Introduction

The stress response is a strategy the cell has developed to withstand an environment deviating from normal physiological conditions. This cellular response, which may be triggered by heat shock, ultimately results in shutting off the synthesis of most proteins and launching the synthesis of heat shock proteins, so as to avoid injury or cell death ( 1 , 2 ). At the post-translational level, heat shock is known to transiently inhibit pre-mRNA splicing ( 3 , 4 ), thus most likely preventing errors otherwise resulting in the synthesis of abnormal proteins which would be harmful to the cell ( 5 ).

Pre-mRNA splicing occurs within heterogeneous nuclear RNP complexes (hnRNP) ( 6–9 ) which are associated with the nuclear matrix ( 10–13 ). Following heat shock at the upper temperature range of the stress response, one would expect hnRNPs to be transiently altered so as to inhibit splicing. Analysis of the general characteristics of hnRNP complexes did not show any alteration ( 14–16 ). Screening our anti-hnRNP monoclonal antibody library by immunocytofluorescence of normal and stressed HeLa cells essentially led to the same conclusion, except for two antibodies which revealed differences between both types of cells ( 15 , 16 ). Based on this observation we further showed that heat shock rapidly induces subtle and reversible rearrangements within hnRNP complexes. Indeed, as early as 10 min after the onset of a heat shock at 45°C two antigen doublets of 35–37 and 72.5–74 kDa, respectively recognized by monoclonal antibodies 2H9 ( 15 ) and 6D12 ( 16 ), rapidly leave the hnRNP complexes and strongly bind to the nuclear matrix, this effect being reversed after cells recovered for 6 h at 37°C. At this point, our data correlated with a possible role of these antigens in turning splicing on and off.

In the present investigations centered on the 72.5–74 kDa antigens, we first addressed the question whether these antigens really intervene in splicing and then further determined to what extent the splicing machinery is altered by such a brief heat shock. The latter aspect is all the more interesting since most published data describe heat-shock effects on splicing following stress periods of 1–2 h ( 3 , 4 , 17 ). Since the original 6D12 monoclonal antibody (IgM) was of limited use in such functional studies, we cloned and sequenced the cDNA encoding this antigen doublet and further raised monoclonal and polyclonal anti-peptide antibodies. We also show that these antigens are in fact hnRNP-M proteins and that their previously published sequences ( 18 ) contain some limited errors. Anti-peptide antibodies were then tested for their effect on in vitro splicing. Having in mind our previous data showing that a 10 min heat shock at 45°C causes the 72.5–74 kDa antigens to transiently leave the hnRNP complexes ( 16 ), we also tested the splicing capacity of nuclear extracts from normal and stressed cells, which were further analyzed so as to detect possible differences. Additional in situ hybridization allowed chromosomal localization of the hnRNP-M encoding gene.

Materials and Methods

Monoclonal antibody 6D12

Monoclonal antibody 6D12 (IgM) which is specific for the 72.5–74 kDa hnRNP antigens, was obtained as described ( 16 ), after immunizing Balb/c mice with hnRNP complexes purified from HeLa S3 cells. Both hybridoma culture supernatant and ascites fluids were used as antibody sources. Antibody was purified from ascites fluid on an ABx column, according to the manufacturer (J.T.Baker).

cDNA cloning, sequencing and expression

cDNA cloning was performed by using standard molecular biological techniques ( 19 ). A λgt11 MCF-7 cell random-primed cDNA library was immunoscreened with monoclonal antibody 6D12. Inserts from positive plaques were used to further screen a λZAP II HeLa cell oligo-dT primed cDNA library, so as to isolate full-size cDNAs. After self-excision in pBluescript SK − and synthesis of single-stranded DNA from inserts in both orientations, sequences were determined from both strands ( 20 ).

The 6D12 cDNA in pBluescript SK − was expressed in E.coli as a β-galactosidase fusion protein. Following IPTG induction, bacterial proteins were analyzed by Western blotting.

Monoclonal anti-peptide antibodies

Potentially immunogenic amino acid stretches of antigen 6D12 were determined from the hydropathy ( 21 ) and mobility ( 22 ) plots drawn from the predicted primary structure. Synthetic peptides with an additional Cys residue were prepared and coupled to ovalbumin with m -maleimidobenzoyl -N -succinimide ester.

Hybridomas were established after fusing X-63 myeloma cells with splenic cells from Balb/c mice ( 16 ) immunized with peptide-ovalbumin. Monoclonal antibodies were purified from ascites liquids on ABx columns (J.T.Baker).

Cell culture, heat shock and nuclear extracts

HeLa S3 cells were grown in Eagle suspension medium supplemented with 10% newborn calf serum and antibiotics (penicillin, 100 IU/ml; streptomycin, 100 µg/ml). For heat treatment, cells were concentrated to 5 × 10 6 per ml and transferred to culture flasks which were immersed in a water bath for 10 min at 45°C. Nuclear extracts were prepared according to Dignam et al. ( 23 ) and dialyzed against buffer D for splicing assays.

For complementation assays, when investigating the role of U-snRNPs, nuclear extracts from normal cells were incubated at 30°C for 15 min with 2000 U/ml micrococcal nuclease, in the presence of 1 mM CaCl 2 . The nuclease was inactivated with 2 mM EGTA. Controls showed that this treatment leads to complete hydrolysis of snRNAs.

In vitro splicing and the antibody effect

Splicing substrates were prepared from three constructions: plasmid Sp4, generated by inserting almost the entire adenovirus-2 natural E1A sequence (positions 533–1342) between the Sma I and Xba I sites of vector pSP65 ( 24 ); plasmid Sp1, derived from the previous one and containing only part of the E1A sequence (positions 1006–1336); a plasmid containing the rabbit β-globin gene (positions 64–732) ( 25 ). Capped precursor RNA was synthesized using SP6 RNA polymerase in the presence of m 7 G(5′)ppp(5′)G and [α- 32 P]CTP and purified as described ( 24 ).

Splicing reactions were carried out for 2 h at 30°C in 25 µl assays including 10 5 c.p.m. of 32 P-labelled pre-mRNA (∼4 ng) and 10 µl nuclear extract (6–8 mg/ml protein) in the presence of 2.6 mM MgCl 2 and 60 mM KCl, as described ( 24 ). For studying the direct effect of antibody on splicing, nuclear extracts were preincubated at 0°C for 30 min with 10 µg purified monoclonal antibody. For controls, the nuclear extract was incubated with an equal amount of non-relevant 5G4 antibody (IgG; from our library) or of specific anti-peptide antibody, the latter being previously either heat denatured at 65°C for 15 min or preincubated for 2 h at room temperature in the presence of a 50× or 100× molar excess of specific or non-specific peptide. After phenol extraction and purification, RNA was analyzed on 5.2% polyacrylamide gels in 8 M urea.

Recombinant protein

The 6D12 cDNA coding sequence was modified so as to have the initiation codon within an Nde I site and the stop codon followed by a Bam HI site and either cloned in a pET 3b expression vector (Novagen) or in a pET 3d vector modified so as to encode an N-terminal His-tagged protein.

For recombinant protein purification, bacteria were freeze-thawed, resuspended in 20 mM Tris-HCl, pH 7.9, 0.5 M NaCl and the lysate sonicated and centrifuged. The inclusions were washed with buffer D and dissolved in 6 M urea or guanidine-HCl in buffer D; the protein was renaturated by step dialysis against the same buffer containing 2 and 0.5 M urea or guanidine-HCl. Recombinant protein from the supernatant was prepared as follows: His-tagged protein was purified on Ni 2+ chelation resin columns, the protein being eluted with 0.4 M imidazole in 20 mM Tris-HCl, pH 7.9, 0.5 M NaCl and dialyzed against buffer D; non-tagged protein was immunopurified on Sepharose-antibody columns, the protein being eluted with 3.5 M NaSCN or 4.5 M MgCl 2 and dialyzed against buffer D.

Western blot analysis

Proteins were resolved by SDS-PAGE and electrophoretically transferred to nitrocellulose sheets, the blots being further probed with the relevant first antibody and 125 I-labelled anti-mouse Ig F(ab') 2 ( 16 ).

Double immunocytofluorescence and confocal microscopy

Schaffner HeLa cells ( 26 ) (10 5 cells/ml) in Dulbecco's modified Eagles medium supplemented with 10% fetal calf serum and antibiotics (100 IU/ml penicillin, 100 µg/ml streptomycin) were grown on glass slides for 3 days, washed with PBS and fixed for 4 min with 2% paraformaldehyde in PBS. For heat treatment, cell cultures were immersed in a water bath for 10 min at 45°C prior to fixation. The cells were permeabilized for 20 min with 0.1% Triton X-100, 0.05% NaN 3 in PBS. After rinsing with PBS/ NaN 3 , the cells were incubated overnight with a mixture of 1F7 anti-D2 peptide monoclonal antibody (1 µg/ml) and rabbit anti-p80 coilin antibody (1:1000) in PBS. The cells were washed with PBS/Triton/NaN3 and incubated for 45 min with a mixture of Cy3-conjugated goat anti-mouse IgG (1:200) and FITC-conjugated donkey anti-rabbit IgG (1:50; Jackson) in PBS/Triton/NaN 3 . After washing the cells, DNA was counterstained with Hoechst 33258. The preparations were mounted in glycerol/PBS (4:1) containing 5% propylgallate as an anti-oxidant. Immunofluorescence images from 0.6 µm optical sections were obtained using a Leica confocal laser scanning microscope equipped with a PL APO 100/1.4 oil immersion lens and a krypton/argon laser. Excitation wavelengths of 488 and 568 nm were selected for FITC and Cy3, respectively. Pseudo-coloured images of the signals were generated and further superimposed using an image processing program.

Figure 1

Nucleotide and deduced amino acid sequence. The open reading frame of cDNA clone 6D12 encodes a 691 amino acid protein with the same sequence as hnRNP-M1/M2 proteins, except for a short amino acid stretch (reverse type box) which must replace the previously reported GPACERQRGS sequence ( 18 ). This error is due to three additional C residues (*), the first one creating a unique Bss HII restriction site (grey box).

Gene mapping by in situ hybridization

In situ hybridization was essentially performed as described ( 27 ). Phytohaemagglutinin-stimulated human lymphocytes were cultured for 72 h, in the presence of 5-bromodeoxyuridine (60 µg/ml) for the last 7 h. The 6D12-coding cDNA in Bluescript II SK − plasmid was tritium labeled by nick translation (1.5 × 10 8 d.p.m./µg) and hybridized to metaphase spreads (100 ng/ml). The slides were coated with Kodak-NTB 2 nuclear track emulsion, exposed for 18 days at 4°C and developed. R-banding was performed using the fluorochrome/photolysis/Giemsa method.

Results

cDNA isolation, nucleotide sequence and deduced amino acid sequence

By immunoscreening the λgt11 expression library with monoclonal antibody 6D12 we isolated a 1.1 kb cDNA fragment which was further used to screen the γZAP II library. We finally obtained four cDNAs of 2.4 kb, each revealing a single 2.4 kb band by Northern blotting. The 2.4 kb cDNA (GenBank accession no. L32611) contains 2394 bp comprising a 5′-untranslated region of 9 bp, an open reading frame of 2073 bp encoding a protein of 691 amino acids and a 3′-untranslated region including a 53 bp poly(A) tail. Database searches (Blitz EMBO) showed that the encoded 6D12 protein is close to the hnRNP-M1/M2 proteins described by Datar et al. ( 18 ). However, the 6D12 protein has 691 amino acids versus 690 for the M1/M2 proteins, the difference being due to three additional C (positions 80, 91 and 114) ( Fig. 1 ), the one at position 80 generating a unique cleavable Bss HII restriction site (G/CG C GC) which is absent in the previously published sequence. Finally, by sequencing the relevant region in the original pHCM4 cDNA clone ( 18 ), we recently confirmed the presence of these additional nucleotides, which modify the sequence of a short amino acid stretch (positions 24–34; Fig. 1 , reverse type box) located outside the three consensus RNA-binding domains (CS-RBD) and the ERMG repeat region.

The hnRNP-M proteins are involved in splicing in vitro

Our previous studies suggested that the 72.5–74 kDa antigens might be involved in turning splicing on and off ( 16 ). To investigate this possibility, we first had to check whether these antigens are really involved in splicing. Therefore, we tested the ability of specific antibodies to directly inhibit in vitro splicing of an 854 nt pre-mRNA synthesized from plasmid Sp4 which contains adenoviral E1A sequence. In the salt conditions we used (see Materials and Methods), this pre-mRNA yields a 13S mRNA and a 114 nt intron.

Preincubating a nuclear extract with monoclonal antibody 6D12 had a partial but reproducible inhibitory effect on in vitro splicing when compared with non-related monoclonal antibody 5G4 from our library ( Fig. 2 A, lanes 3 and 4) or with IgG from non-immune mice (not shown). Since the only partial effect might be explained by the low affinity of this antibody, we then tested two monoclonal antibodies we raised against peptide D2 (amino acids 34–46; Fig. 1 ). These antibodies, named D2-1F7 and D2-1B9 recognize the entire hnRNP-M family ( 18 ) and behave as does monoclonal antibody 6D12 with respect to Western blot and immunocytofluorescence analyses ( 16 ). Preincubating nuclear extracts with either of the two monoclonal anti-peptide antibodies led to nearly total inhibition of 13S mRNA accumulation. Figure 2 A (lane 5 or 7) shows the inhibition observed when using 10 µg monoclonal antibody D2-1F7 (IgG 1 ). Antibody D2-1B9 (IgG 1 ) and also an anti-D2 rabbit polyclonal antibody led to the same result (not shown). Since neither mature mRNA nor free exon 1 are accumulated, inhibition occurs during spliceosome assembly. To exclude artefacts, we first preincubated nuclear extracts with heat-denatured antibody ( 6 ), which had no effect on splicing ( Fig. 2 A; lanes 8 and 7). Further assays in which monoclonal antibody D2-1F7 was preincubated with peptide D2 (50× or 100× molar excess) showed a significant relief of the inhibition ( Fig. 2 B; lanes 4 and 2). In contrast, non-specific peptide D9 had no effect (lane 3). Additional controls in which nuclear extracts were preincubated with either peptide D2 or D9 alone did not significantly alter splicing efficiency (lanes 5 and 6). Taken together, our data show that the direct splicing inhibition we observe is due to specific antibody-antigen recognition.

Experiments based on specific depletion and aimed at further complementation were also attempted, although we and others have shown that hnRNP-M proteins are tightly associated with hnRNP complexes ( 16 , 18 ) and cannot easily be solubilized by salt treatment. However, we attempted to solubilize them in the presence of 0.4, 0.6 or 1 M NaCl. Depletion of salt-treated nuclear extracts was by chromatography on purified antibody coupled to CNBr-Sepharose or on biotinylated antibody bound to Sepharose-streptavidin beads. Even at 1 M NaCl and whatever the antibody we used (original 6D12 or anti-peptide), depletion never exceeded 50%, which did not alter the splicing capacity of these nuclear extracts (not shown). Submitting such extracts to a second round of depletion did not improve that rate, thus showing that the antigen still present in the extract remains unavailable to the antibodies.

Figure 2

Inhibitory effect of anti-hnRNP-M antibodies on splicing in vitro. Competition with synthetic peptides. (A) Nuclear extracts from HeLa cells were preincubated for 30 min on ice in the presence or absence of purified antibody (10 µg). After adding the labeled Sp4 pre-mRNA, splicing reactions were performed for 2 h and the products further analyzed on 5.2% polyacrylamide gels. Lane 1, initial pre-mRNA. Lanes 2 and 6, control splicing reaction without antibody. Lanes 3–5, 7 and 8, splicing reaction after preincubating nuclear extracts with non-relevant 5G4 monoclonal antibody (lane 3), original 6D12 monoclonal antibody (lane 4), native (lanes 5 and 7) or heat denatured (65°C for 10 min) (lane 8) D2-1F7 anti-peptide antibody. (B) Purified antibody (10 µg) was incubated for 2 h at room temperature in the presence or absence of synthetic peptide (10 µg). Preincubation with nuclear extract and splicing reaction were as in (A). Lane 1, control splicing reaction without antibody. Lanes 2–4, splicing reaction after preincubating the nuclear extract with D2-1F7 anti-peptide antibody (lane 2) or with the same antibody previously incubated with either an unrelated peptide D9 (lane 3) or specific peptide D2 (lane 4). Lanes 5 and 6, splicing reaction after preincubating nuclear extracts with peptide D9 (lane 5) or D2 (lane 6). Tr, transcript; E1, exon 1; E1:E2, final product. The asterisk shows the final product of a minor reaction occurring on the Sp4 pre-mRNA ( 24 ).

Early effect of heat shock on in vitro splicing

It has been shown by others that nuclear extracts from HeLa cells heat shocked for 1–2 h at the upper temperature range of the stress response are inactive in splicing in vitro ; in such conditions the structure of U-snRNPs, which are essential splicing components ( 28 , 29 ), was found to be altered ( 3 , 17 ). Whether a 10 min heat

shock at 45°C is already leading to inactive nuclear extracts was checked by using three different pre-mRNAs. Figure 3 shows that splicing in vitro is indeed strongly reduced (β-globin pre-mRNA, lane 2; Sp1 pre-mRNA, lane 4) or almost totally inhibited (Sp4 pre-mRNA, lane 8). Analysis of splicing complex formation revealed non-specific H-complexes, but no pre-spliceosomes or spliceosomes (not shown). Further assays showed that the splicing ability of such extracts is significantly restored by complementation with normal cell extracts pretreated with micrococcal nuclease to hydrolyse U-snRNPs ( Fig. 3 , lanes 7–9). Thus, the lack of splicing activity of extracts from briefly stressed cells does not seem to be due to U-snRNP alteration.

Figure 3

Rapid effect of heat shock on splicing in vitro. Complementation assays. Nuclear extracts (NE) were prepared from normal HeLa cells (−) or cells heat shocked (+) at 45°C for 10 min and further assayed for splicing capacity using labeled pre mRNAs: β-globin (693 nt, lanes 1 and 2); Sp1 (378 nt, lanes 3 and 4); Sp4 (858 nt, lanes 5–8); lanes 6 and 7 being controls where nuclear extracts from normal cells were pretreated at 45°C for 10 min (A, lane 6) or with micrococcal nuclease (MN, lane 7). Lanes 9 and 10, complementation assays of nuclear extracts from heat shocked cells with either a nuclear extract from normal cells pretreated with micrococcal nuclease (NE-MN, lane 9) or with recombinant protein 6D12 (P-6D12, lane 10); splicing capacity was tested using Sp4 pre-mRNA. Splicing products were analyzed on 5.2% polyacrylamide gels. The different pre-mRNAs were aligned on the figure.

To find out what might be the origin of this loss of splicing ability, we performed a comparative protein analysis of nuclear extracts from briefly stressed and normal cells. Global protein analysis by SDS-PAGE showed no significant difference between nuclear extracts from normal and stressed cells ( Fig. 4 , panel 1). We then analyzed the extracts by Western blotting with monoclonal antibodies from our anti-hnRNP library (including anti SR splicing factor 9G8 antibody) as well as with other antibodies like anti-hnRNP C (4F4) and anti-hnRNP U (3G6). Among the analyzed antigens no significant difference was detected, except for two of them which were nearly absent in extracts from stressed cells, namely the hnRNP-M proteins ( Fig. 4 , panel 2) and a 35–37 kDa antigen doublet whose sequence and functional analysis will be submitted elsewhere. Following stress, these proteins in fact become unextractible under our conditions of nuclear extract preparation and therefore are found in the nuclear pellet, which is consistent with the shift from hnRNP complexes to the nuclear matrix we observed previously in the cell ( 16 ).

Figure 4

Western blot analysis and quantitative comparison of hnRNP-M antigens present in nuclear extracts from normal and stressed cells. Equal amounts of nuclear extracts from normal cells (−) and cells heat shocked at 45°C for 10 min (+) were resolved on a SDS-10% polyacrylamide gel. 1, Coomassie blue staining. 2, Western blot analysis using D2-1F7 anti-peptide antibody.

We then attempted to restore the splicing activity of extracts from briefly stressed cells by complementation with purified hnRNP-M protein. Since hnRNP-M proteins could not be efficiently immunopurified from nuclear extract because they strongly associate with other proteins and pre-mRNA, we expressed cDNA 6D12 in E.coli ( Fig. 5 ). Control Western blotting showed that recombinant protein is recognized by the original 6D12 as well as by the anti-peptide monoclonal antibodies. In complementation assays, recombinant protein, with or without His-tag, purified from the soluble fraction or from inclusions, was unable to restore the splicing activity of nuclear extracts from stressed cells ( Fig. 3 , lane 10). Controls in which the protein was added to a nuclear extract from normal cells did not alter splicing efficiency. Further complementation assays of nuclear extracts from stressed cells with 9G8 ( 30 ) or SC-35 ( 31 ) SR factors, known to efficiently complement S100 extracts, were not successful either, thus showing that the loss of splicing activity of nuclear extracts from briefly heat shocked cells is not due to the inactivation of SR splicing factors, or not only to this (data not shown).

HnRNP-M proteins are also detected in nuclear bodies

We have previously shown by immunocytofluorescence that the 72.5–74 kDa (hnRNP-M) proteins are not accessible to monoclonal antibody 6D12 in normal HeLa cells, whereas in heat shocked cells (45°C for 10 min) the nuclei displayed a bright signal distributed as small granules all over the interchromatin space ( 16 ). This behavior and this distribution were observed in HeLa cells cultured in Dulbecco's medium supplemented with 7.5% newborn calf serum and 2.5% fetal calf serum. However, later on during our investigations, cells were routinely grown in culture medium supplemented with 10% fetal calf serum. Interestingly, under these conditions hnRNP-M proteins also become accessible in normal cells. In fact, monoclonal antibodies 6D12, D2-1F7 and D2-1B9 revealed the previously described small granule pattern in both normal and heat shocked cells; in addition, up to three brightly stained spherical nuclear bodies are also detected in part of the interphasic cells ( Fig. 6 A and D). These structures have nothing in common with the numerous irregularly shaped nuclear speckles which are known to contain many splicing factors. Indeed, our anti SR splicing factor 9G8 antibody ( 30 ), which reveals the characteristic nuclear speckled pattern, does not stain these nuclear bodies. In fact these bodies rather resemble coiled bodies, which are believed to contain splicing factor U2AF ( 32 , 33 ). This prompted us to carry out a double immunocytofluorescence assay, using a rabbit polyclonal anti-coilin antibody to label coiled bodies ( 34 , 35 ) ( Fig. 6 B and E) and an anti-hnRNP-M monoclonal antibody ( Fig. 6 A and D). Image overlays show that the nuclear bodies revealed with the anti-hnRNP-M antigens are not coiled bodies ( Fig. 6 C and F).

Figure 5

Gel analysis of recombinant 6D12 protein expressed in E.coli. Lane 1, total extract from non-induced bacteria. Lane 2, total extract from IPTG-induced bacteria. Lane 3, purified recombinant 6D12 antigen obtained from inclusions. Staining was with Coomassie blue.

Since the same nuclear signal is observed before and after heat shock in cells grown in 10% FCS, the question arose whether or not the other aspects of the behavior of hnRNP-M proteins are also serum dependent. We then assayed nuclear extracts from cells grown in 10% fetal calf serum for in vitro splicing capacity and for the presence of hnRNP-M proteins. In fact, a possible serum dependence can be excluded, as the results we obtained were the same as those described above: nuclear extracts from normal cells are entirely active in splicing and contain the hnRNP-M proteins, whereas extracts from heat shocked cells are inactive and do not contain these proteins. Moreover, these experiments show that the structural modifications triggered by heat shock are clearly distinct from the modifications in antigen accessibility induced by fetal calf serum.

Southern blot analysis and chromosomal mapping

Southern blot analysis was performed on human cellular DNA digested with either Eco RI (one site in 6D12 cDNA) or Bgl II (no sites) and further blotted on Highbond N + . Hybridization under normal stringency conditions using as a probe a full 6D12 cDNA labeled with 32 P by random priming ( 19 ) revealed two Eco RI bands of 14 and 15 kb and two Bgl II bands. Cellular DNAs originating from ∼20 different patients all revealed this pattern, no polymorphism being observed (not shown).

Figure 6

Confocal microscopy. Distribution of hnRNP-M proteins and p80 coilin in HeLa cells. HeLa cells were grown in Dulbecco's medium supplemented with 10% fetal calf serum instead of 7.5% newborn calf serum and 2.5% fetal calf serum. Normal cells (A—C) and cells heat shocked at 45°C for 10 min (D—F) were processed for double immunocytofluorescence as described in Materials and Methods. D2-1F7 anti-peptide monoclonal antibody was used to detect hnRNP-M proteins (A and D, red staining) and anti-coilin to label the coiled bodies (B and E, green dots). (C) and (F) are confocal overlays of (A) and (B) and (D) and (E), respectively. The image inserted in (D) shows a nucleus examined with a conventional photomicroscope fitted for epifluorescence and further photographed on Technical Pan 2415 films (Eastman Kodak). The arrows point to nuclear bodies containing hnRNP-M proteins and the arrowhead to a nucleolus. Bar 10 µm.

Chromosomal location of the gene encoding the hnRNP-M proteins was determined by in situ hybridization in human lymphocytes. Within 100 metaphasic cells examined, 153 silver grains were associated with chromosomes and among them 34 (22%) were located on chromosome 19. This distribution was non-random as 28 grains out of 34 (82%) were located on the p13.3 band of the short arm of chromosome 19, which allows us to map the 6D12 gene to the 19 p13.3 band of the human genome ( Fig. 7 ). Moreover, these results, along with the Southern blot experiments, point to a unique hnRNP-M locus in the human genome.

Discussion

Following a 10 min heat shock at 45°C, we previously showed that a 72.5–74 kDa antigen doublet transiently leaves the hnRNP population and strongly binds the nuclear matrix, whereas the general properties of hnRNP complexes remain unaltered ( 16 ). Here we study these antigens in more detail, trying to find out whether they may intervene in stress-induced splicing arrest.

cDNA cloning led to the isolation of four 2.4 kb cDNAs, all encoding an authentic 6D12 protein. cDNA sequencing allowed us to show that these proteins are hnRNP-M1/M2 proteins ( 18 ) and also that the sequence published by Datar et al. (pHCM4 cDNA clone; 18 ) is lacking 3 nt, thus modifying the sequence of an 11 amino acid stretch. Curiously, like pHCM4-cDNA, our cDNAs also have short 5′-untranslated regions (9 nt), which raises the question as to whether the N-terminus is really complete; indeed, co-migration of recombinant protein with native protein from a nuclear extract might not be sufficient evidence, since a difference of just a few additional amino acids is undetectable.

Figure 7

Location of the 6D12 gene to human chromosome 19 by in situ hybridization. All procedures were as described in Materials and Methods. The diagram of chromosome 19 indicates the distribution of labeled sites.

In contrast to our original 6D12 monoclonal antibody (IgM), three anti-D2 peptide antibodies (two monoclonals and one polyclonal) we raised allowed direct inhibition of in vitro splicing of an Sp4 pre-mRNA. This demonstration relies on the utilization of three different antibodies and further on two controls showing first that heat-denatured antibodies do no more inhibit splicing and second, and most importantly, that the addition of specific peptide can relieve this antibody-mediated inhibition. Taken together, these data already provide further clues in favour of either direct or indirect involvement of the hnRNP-M proteins in in vitro splicing, although standard depletion/complementation could not be used. Here it is worthwhile mentioning that the difficulties we encountered reflect those other investigators have had when trying to demonstrate a role for hnRNP proteins in splicing. Indeed, although the structure of numerous hnRNP protein families has been established, only two of them have been assigned a function in splicing: hnRNP-A/B proteins, which regulate alternative splicing ( 36 , 37 ), and hnRNP-C proteins ( 6 ). Also to be mentioned is galectin-3, which has similarities to hnRNP proteins and was shown to also be involved in splicing ( 38 ). The hnRNP-C proteins, which exhibit a single RNA binding domain ( 39 ), were further shown to interact with the polypyrimidine stretch of the 3′-end of introns ( 40 ) so as to display the RNA sequence in a configuration suitable for splicing reactions ( 41 ), which clearly points to a chaperone activity. Since several other hnRNP proteins also revealed strong RNA annealing activity ( 36 , 42 ), and therefore potential ability to modulate pre-mRNA configuration, one might think of chaperone activity as a general concept for hnRNP protein function. Therefore, hnRNP-M proteins which preferentially bind poly(G) and poly(U) homopolymers ( 18 ) might also be candidates for such a function.

As it turned out that normal nuclear extracts could not be efficiently depleted for further complementation assays, we tried to take advantage of the fact that in vivo , following a 10 min stress at 45°C, the hnRNP-M proteins transiently leave the hnRNP complexes and bind to the nuclear matrix ( 16 ). Our assumption that this process might also be viewed as an in vivo depletion of hnRNP or splicing complexes was correct in that nuclear extracts from briefly stressed cells no longer contain the hnRNP-M proteins. Since these extracts are splicing deficient, we show that the switch of hnRNP-M proteins from hnRNP complexes to nuclear matrix ( 16 ) correlates with splicing inhibition, which provides further evidence strengthening the idea that hnRNP-M proteins are required for splicing in vitro . Furthermore, and although our experiments essentially rely on cellular subfractionation techniques, it remains that the heat shock was to living cells, therefore, our results most likely reflect the in vivo situation, which might not be the case when nuclear extracts are submitted to elevated temperatures ( 43 ). As the hnRNP-M proteins are associated with hnRNP complexes when splicing is on (normal cells) and dissociated from these complexes when splicing is off (heat shocked cells), our data also provide additional evidence in favour of these proteins being involved in turning splicing on and off in vivo . From our data one can also appreciate the prominent role of the nuclear matrix, which increasingly appears as a dynamic nuclear scaffold organizing and regulating gene expression ( 44 ), including transcription ( 45 ) and also splicing, in which nuclear matrix proteins were shown to be involved ( 46 , 47 ).

Whether the absence of the hnRNP-M proteins can alone cause splicing inactivation is a question not yet solved, since nuclear extracts from stressed cells could not be complemented with recombinant protein. Several reasons linked to the recombinant protein might explain this failure: the protein might not be in the right conformation, it might not have the required post-translational modifications and finally it might not be representative of all hnRNP-M species. Further, we identified an additional hnRNP protein family which is also missing in the nuclear extract from stressed cells. These are 35–37 kDa proteins, which behave as do the hnRNP-M (or 6D12) proteins ( 16 ), since they are associated with hnRNP complexes in normal cells and switch to the nuclear matrix following a brief heat shock ( 15 ). Their structure and function will be presented elsewhere.

In contrast to what has been shown following 1–2 h heat shocks ( 3 , 4 , 17 ), early induced splicing arrest cannot be ascribed to U-snRNP disruption, since normal extracts, in which U-snRNPs are hydrolyzed, still restore the splicing capacity of nuclear extracts from cells stressed for 10 min at 45°C. This is in agreement with immunocytofluorescence data showing that anti-snRNP monoclonal antibodies did not reveal any difference in the nuclear speckled pattern between normal cells and cells stressed at 45°C for 10 min (unpublished observation), whereas after 15 min the signal is uniformly distributed throughout the nucleoplasm ( 48 ). So, whether various stress conditions trigger different mechanisms to shut off splicing is not as yet known. Assays we performed by submitting HeLa cells to 45°C heat shocks for up to 1 h all showed the behaviour previously observed by immunocytofluorescence, i.e. the appearance of a strong nuclear signal, lasting for 2 h after cells were returned to 37°C and then decreasing over another 6 h until disappearance ( 16 ). However, we observed dramatic differences in survival capacity, since cells stressed for longer than 15–20 min, although they apparently did well at 37°C for the next 24 h, were all dead after two days (unpublished data). It is therefore likely that data obtained after submitting cells to long-lasting stresses at high temperature actually describe processes which are not fully reversible and therefore might well be involved in cell death.

Finally, we incidentally observed by immunocytofluorescence that hnRNP-M protein behavior or accessibility is also sensitive to other environmental parameters than temperature, namely serum quality or concentration, since growing cells in 10% fetal calf serum results in a nuclear signal in both normal and stressed cells, which also display nuclear bodies not seen before ( 16 ). Although the nuclear bodies we detect are not coiled bodies, it is nevertheless interesting to notice that, like U-snRNPs, splicing factor U2AF ( 32 , 33 , 49 ) and hnRNP-K and -L proteins ( 50 , 51 ), hnRNP-M proteins are also found in nuclear bodies. More generally, recent data show that splicing components are also present in nuclear compartments different from the pre-mRNA processing sites ( 52–54 ). As to the hnRNP-M proteins, variations in their nuclear distribution or accessibility may reflect a fine tuning of their function, thus allowing the cell to adjust to environmental changes of different kinds and of various intensity and duration.

Acknowledgements

The authors express their appreciation to G.Duval, N.Jung, VSchultz and Y.Lutz for producing anti-peptide polyclonal and monoclonal antibodies and to J.L.Vonesch for carrying out confocal microscopy analysis. They thank A.Staub, P.Eberling and F.Ruffenach for synthesizing oligonucleotides and peptides, J.M.Garnier and T.Lerouge for technical advice and B.Boulay and J.M.Lafontaine for the photographic work. They are grateful to Dr G.Dreyfuss (Howard Hughes Medical Institute) for the gift of antibodies 4F4 and 3G6, to Dr M.S.Swanson (University of Florida) for the gift of the pHCM4 cDNA clone, to Dr A.I.Lamond (EMBL) for the gift of anti-coilin antibody and to Drs A.Hanauer and J.L.Mandel for providing human DNA blots. This work was supported by funds from the Centre National de la Recherche Scientifique, the Institut National de la Santé et de la Recherche Médicale, the Centre Hospitalier Universitaire Régional and by special grants from the Association pour la Recherche sur le Cancer. Dominique Mahé is sponsored by the Ministère de la Recherche et de l'Espace and by the Ligue contre le Cancer.

References

1

,

Annu. Rev. Genet.

,

1988

, vol.

22

(pg.

631

-

677

)

2

,

Stress Proteins in Biology and Medicine

,

1990

Cold Spring Harbor, NY

Cold Spring Harbor Laboratory Press

3

,

EMBO J.

,

1988

, vol.

7

(pg.

3504

-

3518

)

4

,

Cell

,

1986

, vol.

45

(pg.

185

-

193

)

5

,

Science

,

1988

, vol.

242

(pg.

1544

-

1548

)

6

,

Science

,

1986

, vol.

231

(pg.

1534

-

1539

)

7

,

Annu. Rev. Biochem.

,

1993

, vol.

62

(pg.

289

-

321

)

8

,

Nucleic Acids Res.

,

1986

, vol.

14

(pg.

4171

-

4185

)

9

,

Trends Cell Biol.

,

1993

, vol.

3

(pg.

151

-

155

)

10

,

Chromosomal Nonhistone Proteins

,

1984

, vol.

IV

Boca Raton, FL

CRC Press

(pg.

119

-

190

)

11

,

J. Cell Biol.

,

1990

, vol.

110

(pg.

569

-

580

)

12

,

J. Mol. Biol.

,

1982

, vol.

154

(pg.

103

-

119

)

13

,

Mol. Cell Biol.

,

1987

, vol.

7

(pg.

111

-

120

)

14

,

Biochim. Biophys. Acta

,

1986

, vol.

867

(pg.

9

-

15

)

15

,

Exp. Cell. Res.

,

1988

, vol.

175

(pg.

109

-

124

)

16

,

J. Cell Biol.

,

1989

, vol.

109

(pg.

1921

-

1935

)

17

,

J. Biol. Chem.

,

1990

, vol.

265

(pg.

20377

-

20383

)

18

,

Nucleic Acids Res.

,

1993

, vol.

21

(pg.

439

-

446

)

19

,

Molecular Cloning: A Laboratory Manual

,

1989

2nd Edn.

Cold Spring Harbor, NY

Cold Spring Harbor Laboratory Press

20

,

Proc. Natl. Acad. Sci. USA

,

1977

, vol.

74

(pg.

5463

-

5467

)

21

,

J. Mol. Biol.

,

1982

, vol.

157

(pg.

105

-

132

)

22

,

Nature

,

1984

, vol.

311

(pg.

123

-

126

)

23

,

Nucleic Acids Res.

,

1983

, vol.

11

(pg.

1475

-

1489

)

24

,

Cell

,

1987

, vol.

50

(pg.

31

-

39

)

25

,

Genes Dev.

,

1991

, vol.

5

(pg.

1847

-

1858

)

26

,

Cell

,

1981

, vol.

27

(pg.

299

-

308

)

27

,

Hum. Genet.

,

1985

, vol.

69

(pg.

268

-

271

)

28

,

Annu. Rev. Cell Biol.

,

1991

, vol.

7

(pg.

559

-

599

)

29

,

Biochim. Biophys. Acta

,

1990

, vol.

1087

(pg.

265

-

292

)

30

,

EMBO J.

,

1994

, vol.

13

(pg.

2639

-

2649

)

31

,

Nature

,

1990

, vol.

343

(pg.

437

-

441

)

32

,

EMBO J.

,

1991

, vol.

10

(pg.

195

-

206

)

33

,

EMBO J.

,

1991

, vol.

10

(pg.

207

-

214

)

34

,

J. Exp. Med.

,

1991

, vol.

173

(pg.

1407

-

1419

)

35

,

Trends Cell Biol.

,

1993

, vol.

3

(pg.

198

-

204

)

36

,

Cell

,

1992

, vol.

68

(pg.

365

-

375

)

37

,

EMBO J.

,

1994

, vol.

13

(pg.

5483

-

5495

)

38

,

Proc. Natl. Acad. Sci. USA

,

1995

, vol.

92

(pg.

1213

-

1217

)

39

,

Mol. Cell. Biol.

,

1987

, vol.

7

(pg.

1731

-

1739

)

40

,

EMBO J.

,

1988

, vol.

7

(pg.

3519

-

3529

)

41

,

EMBO J.

,

1992

, vol.

11

(pg.

3289

-

3295

)

42

,

EMBO J.

,

1994

, vol.

13

(pg.

213

-

221

)

43

,

Biochem. Biophys. Res. Commun.

,

1993

, vol.

190

(pg.

223

-

228

)

44

,

Science

,

1995

, vol.

268

(pg.

1564

-

1565

)

45

,

EMBO J.

,

1993

, vol.

12

(pg.

1059

-

1065

)

46

,

J. Cell Biol.

,

1994

, vol.

127

(pg.

593

-

607

)

47

,

Proc. Natl. Acad. Sci. USA

,

1994

, vol.

91

(pg.

1505

-

1509

)

48

,

EMBO J.

,

1991

, vol.

10

(pg.

3469

-

3481

)

49

,

J. Cell Biol.

,

1993

, vol.

4

(pg.

841

-

852

)

50

,

J. Cell Biol.

,

1989

, vol.

109

(pg.

2575

-

2587

)

51

,

Nucleic Acids Res.

,

1993

, vol.

21

(pg.

1193

-

1198

)

52

,

J. Cell Biol.

,

1994

, vol.

126

(pg.

877

-

899

)

53

,

Nature

,

1994

, vol.

372

(pg.

727

-

728

)

54

,

Nature

,

1994

, vol.

372

(pg.

809

-

812

)

© 1996 Oxford University Press

I agree to the terms and conditions. You must accept the terms and conditions.

Submit a comment

Name

Affiliations

Comment title

Comment

You have entered an invalid code

Thank you for submitting a comment on this article. Your comment will be reviewed and published at the journal's discretion. Please check for further notifications by email.

Citations

Views

Altmetric

Metrics

Total Views 906

627 Pageviews

279 PDF Downloads

Since 2/1/2017

Citations

62 Web of Science

×

Email alerts

Citing articles via

• Latest

• Most Read

• Most Cited

More from Oxford Academic