Structural organization of mRNA complexes with major core mRNP protein
http://www.100md.com
《核酸研究医学期刊》
Institute of Protein Research, Russian Academy of Sciences, Pushchino, Moscow Region 142290, Russia and 1 Faculty of Physics and Faculty of Chemistry of Moscow State University, Moscow 119992, Russia
* To whom correspondence should be addressed. Tel/Fax: +7 095 924 04 93; Email: ovchinn@vega.protres.ru
ABSTRACT
YB-1 is a universal major protein of cytoplasmic mRNPs, a member of the family of multifunctional cold shock domain proteins (CSD proteins). Depending on its amount on mRNA, YB-1 stimulates or inhibits mRNA translation. In this study, we have analyzed complexes formed in vitro at various YB-1 to mRNA ratios, including those typical for polysomal (translatable) and free (untranslatable) mRNPs. We have shown that at mRNA saturation with YB-1, this protein alone is sufficient to form mRNPs with the protein/RNA ratio and the sedimentation coefficient typical for natural mRNPs. These complexes are dynamic structures in which the protein can easily migrate from one mRNA molecule to another. Biochemical studies combined with atomic force microscopy and electron microscopy showed that mRNA–YB-1 complexes with a low YB-1/mRNA ratio typical for polysomal mRNPs are incompact; there, YB-1 binds to mRNA as a monomer with its both RNA-binding domains. At a high YB-1/mRNA ratio typical for untranslatable mRNPs, mRNA-bound YB-1 forms multimeric protein complexes where YB-1 binds to mRNA predominantly with its N-terminal part. A multimeric YB-1 comprises about twenty monomeric subunits; its molecular mass is about 700 kDa, and it packs a 600–700 nt mRNA segment on its surface.
INTRODUCTION
In eukaryotic cells, all mRNAs and their precursors are permanently associated with various proteins to form heterogeneous nuclear RNPs (hnRNPs) in the nucleus and translatable (polysomal) and untranslatable free messenger RNPs (mRNPs) in the cytoplasm (1,2). The hnRNP proteins pack pre-mRNAs (mRNA precursors), regulate the pre-mRNA processing and participate in mRNA transport from the nucleus to the cytoplasm (3). Cytoplasmic mRNP proteins participate in the mRNA packing and regulate its translation, life span and intracellular localization (1,4–6). Some common features are reported for hnRNPs and mRNPs (2,7): (i) a broad sedimentation distribution that reflects heterogeneity of mRNAs and pre-mRNAs in size; RNP sedimentation coefficients are usually 2.5–3 times higher than those of RNA; (ii) a constant high protein/RNA weight ratio that is 3 for hnRNPs and free cytoplasmic mRNPs and 2 for translatable polysomal mRNPs; this ratio determines mRNP unique buoyant density values in CsCl that are 1.39–1.41 g/cm3 and about 1.45 g/cm3, respectively; (iii) a high sensitivity of pre-mRNAs and mRNAs to endoribonucleases, which is indicative of surface location of the RNAs.
The unique structure of hnRNPs is known to be formed mostly by hnRNP major proteins of the A-, B- and C-groups (8,9). Similarly, in mRNPs, the role of packing proteins may be played by their own universal major proteins present in many copies and capable of binding to any mRNA (1,10). In mRNPs from various tissues and cell lines, there are two clearly observed major proteins—p50 (Mr 50 000–55 000) and p70 (Mr 70 000–80 000) (11–15). The protein p70 displays a high affinity for poly(A) sequences and is localized on the mRNA poly(A) tail (16,17). This protein has been termed the poly(A) binding protein (PABP). PABP was detected mostly in polysomal mRNPs with poly(A)+mRNA. Therefore, PABP can hardly be a universal packing protein that determines unique physico-chemical properties of all mRNPs. The other major mRNP protein, p50 (a 36 kDa protein with anomalous electrophoretic mobility in SDS–PAGE typical of a 50 kDa protein), seems much more suitable for the discussed function. It is detected both in free and polysomal mRNPs and is associated with all kinds of mRNA nucleotide sequences. Its affinity for mRNA is very high (Kd 4 nM) (18). A comparative analysis of the primary structure of p50 preparations from mRNPs of Xenopus laevis oocytes (19), murine oocytes and spermatocytes (20,21), rabbit reticulocytes (22), and goldfish oocytes (23) showed strong homology of all these proteins with the human transcription factor YB-1 and classified them as members of the family of cold shock domain proteins (CSD proteins).
The mRNA transition from its untranslatable to translatable state is accompanied by a 2-fold decrease of the mRNA-associated YB-1 (24). In the cell-free translation system with the increasing YB1/mRNA weight ratio up to 2, YB-1 progressively stimulates protein synthesis at the initiation stage (24–26). A further growth of this ratio causes protein synthesis inhibition both in the cell-free system (24) and in the cell (27); at YB1/mRNA weight ratios from 5 to 6 in the cell-free system, the translation was inhibited completely. The mRNA saturation with YB-1 prevents mRNA interaction with translation initiation factors, thereby causing mRNA transition from polysomes to free mRNPs (28). The stored untranslatable mRNPs of amphibious and mammalian gametes contain significant amounts of YB-1 homologous proteins (29–31). The involvement of stored mRNAs from Xenopus oocytes in translaton just after fertilization is accompanied by dephosphorylation of FRGY2, the YB-1 homolog, resulting in a decrease of its affinity to mRNA (32,33). Then a gradual degradation of the protein occurs and it is substituted by its somatic homolog FRGY1 (34).
YB-1 considerably increases the mRNA life span in cells and cell lysates, probably by protecting mRNA from exonucleases (35). Such an activity of YB-1 could contribute to the storage of mRNA in a form of free mRNPs.
YB-1 and its closest homologs consist of three domains: a short N-terminal Ala- and Pro-rich domain (A/P domain) followed by a central cold shock domain (CSD) and a C-terminal domain (CTD). The CSD domain has RNP-1 and RNP-2 consensus motifs and displays DNA- and RNA-binding activity. The CTD domain is formed by alternating clusters of positively and negatively charged amino acid residues, non-specifically binds RNA and can undergo oligomerization (36). It was shown that free (mRNA-unbound) YB-1 and FRGY2 form homomultimeric complexes whose molecular mass, in the case of YB-1, reaches 800 kDa (22,37,38).
The properties of YB-1 and its separate domains allow us to suggest that in translatable mRNPs with a comparatively low YB-1/mRNA ratio, YB-1 molecules associate with mRNA as monomers through their both RNA-binding domains, thus giving rise to a relatively unfolded mRNA conformation. In untranslatable mRNPs with a high YB-1/mRNA ratio, the interaction between YB-1 molecules and mRNA is realized only through CSDs, which displace CTDs from the complex with mRNA. The displaced CTDs are likely to interact with one another. As a result, YB-1 molecules form a multimeric compact protein complex with a tightly packed mRNA on its surface fully accessible to endoribonucleases. However, within particles with such a conformation the mRNA termini are most probably buried, which makes the mRNA resistant to exoribonucleases and the mRNA termini inaccessible for interaction with translation initiation factors, and first of all, with eIF4F and PABP (10,35,39).
The current study was focused on mRNA–YB-1 complexes and its aim was to find out whether (i) YB-1 alone, without other mRNP proteins, is capable of forming complexes with typical properties of natural mRNPs; (ii) YB-1 is monomeric when binding to mRNA at a low YB-1/mRNA ratio; (iii) both RNA-binding domains of YB-1 participate in this binding; (iv) mRNA-associated YB-1 undergoes multimerization at higher YB-1/mRNA ratios; (v) in these conditions, YB-1 uses only one (which of the two?) or both of its RNA-binding domains to associate with mRNA.
We have shown that the YB-1-saturated mRNA forms particles whose sedimentation coefficient and buoyant density are typical of natural free mRNPs. At a low YB-1/mRNA ratio, YB-1 binds to mRNA as a monomer using its both RNA-binding domains; as this ratio is increased, YB-1 becomes multimeric and predominantly uses CSD for its interaction with mRNA. The relatively short globin mRNA is packed on the surface of only one multimeric protein globule, whereas larger YB-1-associated RNAs form polysome-like ‘beads-on-a-string’-type structures where the number of globules involved is proportional to the length of RNA used.
MATERIALS AND METHODS
Preparation of recombinant YB-1
The synthesis of recombinant rabbit YB-1 was induced in Escherichia coli BL21(DE3) transformed with plasmid pET-3-1-YB-1 (p50) by isopropyl-?-D-thiogalactopyranoside (IPTG), and the protein was purified by chromatography on heparin–Sepharose 4B and Superose 12 HR 10/30 columns (Amersham Biosciences) as described in (25,40). The YB-1 preparations were concentrated and dialyzed against a buffer containing 10 mM HEPES–KOH, pH 7.6 and 150 ml KCl, and stored at –70°C in aliquots. The protein concentration was determined by staining with a Micro-BCA kit (Pierce).
Synthesis of RNAs
The sequence of -globin cDNA (a gift of Dr J. Ilan) containing a coding part, the sequence of 3'-UTR and 50 nt poly(dA) was inserted into NcoI and BamHI sites of the pET-28a vector (Novagen), resulting in the pET-28a–-globin plasmid. The plasmid was linearized with BamHI and 660 nt -globin mRNA was transcribed by T7 polymerase as described previously (41). To obtain a hybrid 2Luc mRNA, a DNA fragment (a gift of Dr I. Shatsky), containing two full-length cDNAs of luciferases from Renilla reinformis and Photinus pyralis, separated by a polylinker, was cloned into the EcoRV and XbaI sites of the pSP72 vector (Promega). The prepared plasmid pSP72-2Luc was linearized by XhoI and 2Luc mRNA, including both the luciferase sequences, was synthesized by T7 polymerase. 2Luc mRNA was about 3000 nt long. The full-size TMV RNA of about 6000 nt long, isolated from viral particles by phenol extraction, was a gift of Dr E. Karger.
To prepare uniformly labeled mRNAs, the concentration of UTP was decreased and UTP was added to the transcription mixture (>2000 Ci/mmol; Radioisotop, Russia). Homogeneity of all RNAs used was verified by gel electrophoresis in denaturing conditions (data not shown).
Sucrose gradient analysis
10 μg (280 pmol) YB-1 was mixed with different amounts of -globin mRNA, which is a mixture of a constant amount of 32P-labeled mRNA (100 000 c.p.m.) and a variable amount of unlabeled mRNA, in 25 μl of buffer A (10 mM HEPES–KOH, pH 7.6, 100 mM KCl) containing 0.5 units/μl RNasin. The samples were incubated for 15 min at 30°C and layered onto 5–20% (w/v) linear sucrose gradients in buffer A. Centrifugation was carried out at 45 000 r.p.m. for 4 h at 4°C in an SW 60 rotor (Beckman, USA). Gradients were monitored for absorbance at 254 nm during collection of the fractions from the bottom. Radioactivity in the fractions was estimated by Cherenkov counting in a LS-100 radioactivity counter (Beckman, USA). Then in the case of -globin mRNA, the proteins in the fractions were precipitated with 10% TCA and subjected to SDS–PAGE. 28S and 18S rRNAs, as well as 9S globin mRNA, were used as sedimentation coefficient markers, the latter obtained from rabbit reticulocyte polysomes as described earlier (42).
CsCl density gradient centrifugation
mRNA complexes were formed in the following way: 0.2 μg (1 pmol) 32P-labeled -globin mRNA (100 000 c.p.m.) were mixed with increasing amounts of YB-1 in 15 μl of buffer A. The samples were incubated for 15 min at 30°C. Then they were fixed with 4% freshly neutralized formaldehyde for 1 h at 37°C. For the gradient to be formed, a ‘light’ ( = 1.232 g/cm3) and a ‘heavy’ ( = 1.718 g/cm3) solutions of CsCl in the buffer containing 10 mM HEPES–KOH, pH 7.6, 1 mM MgCl2, 4% formaldehyde were prepared. Then 2.2 ml of the ‘heavy’ CsCl solution was poured in a 5 ml centrifuge tube and on top of it 2.8 ml of the ‘light’ CsCl solution preliminarily mixed with the mRNP sample was placed. The gradients were centrifuged at 36 000 r.p.m. for 24 h at 4°C in an SW55-Ti rotor of a Beckman L5 centrifuge and fractionated from the bottom. The CsCl densities in selected fractions were calculated from their refraction indexes and the radioactivity was estimated by Cherenkov counting. To determine the protein distribution, the fractions were filtered through a nitrocellulose membrane (Whatman, England) and the protein was detected using anti-YB-1 antibodies at a 1:500 dilution as described in (25).
Chemical cross-linking
To obtain mRNP complexes with different molar ratios of YB-1/mRNA, 3 μg (85 pmol) YB-1 was mixed with premixes of a constant amount of -globin 32P-labeled mRNA (200 000 c.p.m.) with a varying amount of unlabeled mRNA in 20 μl of buffer A. The samples were incubated for 10 min at 30°C. Then glutaraldehyde (Merck) up to 0.15% was added to the reaction mixture and the incubation was continued for 15 min at 37°C. The reactions were stopped with 70 mM glycine, pH 8.0. mRNA was destroyed by incubation with 20 U RNase T1 (USB) and 5 μg RNase A (USB) and by subsequent boiling with 20 mM MgCl2 for 20 min. The resulting 32P-labeled proteins were resolved by 4–16% SDS–PAGE. The gel was stained with Coomassie brilliant blue R-250 (Pierce), dried and exposed to Kodak X-omat AR film at –70°C overnight.
UV cross-linking and cyanogen bromide cleavage
32P-labeled -globin mRNA (15 pmol, 1 500 000 c.p.m.) was incubated with increasing amounts of YB-1 in buffer A in the final volume of 25 μl for 15 min at 30°C. The samples were layered onto Parafilm M (American Can Company, USA), placed on ice and exposed to UV light (254 nm) in a transilluminator-cross-linker (Vilber-Lourmant, France). The distance from the source to the cross-linked sample was about 3 cm and the radiation dose was 0.75 J/cm2. With such a dose, individual YB-1 monomers were not yet noticeably cross-linked to themselves, but efficiently cross-linked to mRNA (data not shown). Then 20 U RNase T1 and 10 μg RNase A were added to the samples and the mixture was incubated for 30 min at 37°C. After treatment with ribonucleases, the samples were incubated with 1.5 mg BrCN in 80 μl of 70% formic acid at room temperature in the dark overnight. The samples were 4-fold diluted with H2O and lyophilized. The resulting 32P-labeled protein fragments were resolved in 15% SDS–PAGE. The gel was stained with Coomassie brilliant blue, dried and exposed to the X-ray film as mentioned above. Gel slices containing the labeled YB-1 fragments were excised after autoradiography and the radioactivity of the slices was determined by Cherenkov counting.
Atomic force microscopy and electron microscopy
To prepare YB-1–RNA complexes, 5 μg -globin mRNA, 2Luc mRNA or TMV RNA were mixed with 2.5 μg YB-1 (for non-saturated complexes) or with 40 μg YB-1 (for saturated complexes). The complexes were incubated in buffer A for 20 min at 30°C, and fixed with 0.15% glutaraldehyde as described above. To separate YB-1–RNA complexes from the protein unbound to mRNA, the preparations were centrifuged in 5–20% sucrose gradient for 3 h (-globin mRNA), 1.5 h (2Luc mRNA) and 1 h (TMV mRNA). The obtained RNPs were collected by centrifugation in a TLA 100.1 rotor at 90 000 r.p.m. for 2.5 h, dissolved in buffer A and used for atomic force microscopy (AFM) and electron microscopy (EM) analysis.
Free multimeric YB-1 (2 μg) was fixed with 0.15% glutaraldehyde in buffer A for 15 min at 37°C as described above.
AFM imaging experiments were performed in tapping mode using a Nanoscope IIIa multimode scanning probe microscope (Digital Instruments, USA). Standard silicon 125 μm cantilevers (NanoProbe) with 300–350 kHz resonant frequencies and 100 μm non-contact cantilevers with 200–300 kHz resonant frequencies (State Research Institute for Problems in Physics, Russia) were used for imaging in air. Experiments in liquid employed a tapping mode liquid cell (Digital Instruments, USA) and silicon nitride cantilevers (Nanoprobe) with a spring constant of 0.6 N/m. The cantilever oscillation frequency was 8–10 kHz. For image processing and presentation, software Femtoscan 001 (43) was applied. Samples were prepared by a direct adsorption technique, using aminopropylsilatran-treated (44) mica (APS-mica) plates as substrates. The preparation was carried out as previously described in (45). An aliquot of 10 μl of the mRNPs (1–10 ng/μl taking the concentration of RNA) or fixed YB-1 (10 ng/μl) were applied onto the substrate and left for 5 min for adsorption. Samples designated for imaging in the liquid cell were mounted directly onto the scanner, whereas those prepared for experiments in air were rinsed with tridistilled water and dried in airflow.
For EM studies, the samples were prepared by shadow casting and by negative staining. The mRNPs (50 ng/μl of RNA) or YB-1 (50 ng/μl) were adsorbed onto a thin carbon film and shadowed with platinum at an angle of tan(1/10) with an electron gun (46). Negative staining with 2% aqueous uranyl acetate was performed using the single-layer carbon technique (47). The samples were prepared also by the method of a preshadowed carbon replica (46). In this case, the samples were prepared as those for AFM. The negatively stained or metal-shadowed samples were examined in a JEM-100C electron microscope at 80 kV.
RESULTS
YB-1 associates with -globin mRNA to form complexes with sedimentation coefficient and buoyant density in CsCl typical for natural mRNPs
To estimate sedimentation coefficients of YB-1–mRNA complexes, YB-1 was mixed with 32P-labeled -globin mRNA at YB-1/mRNA molar ratios from 3:1 to 60:1 (weight ratios from 0.5 to 10) and analyzed using centrifugation in sucrose concentration gradient. Figure 1A shows the results of sedimentation distribution of complexes formed at YB-1/mRNA molar ratios from 12:1 to 60:1. An addition of increasing amounts of YB-1 to mRNA caused a gradual increase of the sedimentation coefficient of the formed complexes. The maximal saturation of -globin mRNA with YB-1 was observed at the 36:1 molar ratio. The formed saturated complex displayed sedimentation homogeneity, and its sedimentation coefficient was about 28S, i.e. it was three times as high as the sedimentation coefficient of the used RNA, which is typical of natural free mRNPs (informosomes) (7). Up to a YB-1/mRNA molar ratio of 12:1, the entire protein was mRNA-bound (data not shown), but at higher ratios, some free YB-1 appeared, though the sedimentation coefficient of the complex remained below its maximum. The appearance of unbound protein before mRNA saturation could be explained by an insufficiently high constant of its binding to mRNA. The minor component sedimenting faster than the main mRNP complex might be a complex of YB-1 with a stable globin mRNA conformer. Indeed, the used globin mRNA contained a minor component with a higher sedimentation coefficient and a lower electrophoretic mobility at native gel electrophoresis; this component disappeared upon gel electrophoresis under denaturing conditions (7 M urea) (data not shown).
Figure 1. Sedimentation and density analysis of YB-1 complexes with -globin mRNA. (A) Sedimentation distribution of YB-1 complexes with -globin mRNA in sucrose gradient. YB-1 was mixed with 32P-labeled -globin mRNA at indicated molar ratios. The formed complexes were layered on 5–20% sucrose gradient and centrifuged in a SW-60 rotor at 45 000 r.p.m. for 4 h at 4°C. UV absorbance at 254 nm (solid line) and radioactivity (open circles) are shown. YB-1 distribution over gradient fractions revealed by SDS–gel electrophoresis is shown below each figure. (B) Distribution of YB-1 complexes with -globin mRNA in CsCl density gradient. YB-1 complexes with 32P-labeled -globin mRNA formed at indicated molar ratios were fixed with 4% formaldehyde. Then the samples were centrifuged in CsCl gradient in a SW-55Ti rotor at 36 000 r.p.m. for 24 h at 4°C. Radioactivity (open circles) and CsCl density (filled triangles) are shown. YB-1 distribution over gradient fractions revealed by immunoblotting with antibodies to YB-1 is shown below each figure.
To determine the buoyant density of YB-1–mRNA complexes, YB-1 was in vitro incubated with 32P-labeled -globin mRNA, fixed with formaldehyde and centrifuged in the CsCl density gradient to equilibrium (Figure 1B). Within the YB-1/mRNA range from 12:1 to 48:1, there was one homogeneous peak whose buoyant density decreased with the increasing YB-1/mRNA ratio and reached 1.39 g/cm3 upon mRNA saturation at the molar ratio of 36:1. Excess of mRNA-unbound YB-1 was detected in fractions with a characteristic protein buoyant density of 1.35 g/cm3 on the right shoulder of the peak of the mRNP complex. The buoyant density of YB-1-saturated complexes was just the same as that of natural free mRNPs (7).
mRNA-bound YB-1 is monomeric at a low YB-1/mRNA ratio and multimeric when this ratio is high
Free YB-1 is known to be homomultimeric (22,38). A question arises as to whether it binds to mRNA as a multimer or as a monomer that subsequently restores its multimeric form upon mRNA saturation, as suggested earlier (39). The experiments described above on centrifugation both in sucrose gradient and in CsCl density gradient showed a unimodal sedimentation and density distribution of the complexes formed at various YB-1/mRNA ratios and a gradual shift of the sedimentation coefficient and buoyant density with changes in this ratio. Such a behavior of the complexes is in agreement with the suggestion that YB-1 binds to mRNA as a monomer, although the relatively low resolving capacity of these techniques does not allow ruling out the possibility that YB-1 associates with mRNA in small cooperative groups.
To have a more precise knowledge about quaternary structure of mRNA-associated YB-1 at various YB-1/mRNA ratios, free YB-1 or YB-1 bound to 32P-labeled -globin mRNA was fixed with 0.15% glutaraldehyde. The fixed preparations were subjected to exhaustive RNase treatment and analyzed by SDS–PAGE (Figure 2A). As seen, the initial free YB-1 is a multimer only slightly displaced from the separating gel start after electrophoresis under denaturing conditions. Its relative molecular mass far exceeds the highest marker (205 kDa), which is consistent with the sedimentation and gel filtration data on its large size (up to 800 kDa) (22,40). The protein obtained from the complexes with mRNA with a low YB-1/mRNA molar ratio (1.5:1) moved as a monomer during SDS–gel electrophoresis. An increase of the YB-1/mRNA ratio to 3:1 and more gave rise to cross-linked oligomers of gradually increasing size. Finally, at YB-1/mRNA exceeding 24:1, large multimers appeared close in size to a free fixed protein. This supports our previous suggestion that YB-1 changes its quaternary structure on mRNA depending on the YB-1/mRNA ratio.
Figure 2. (A) Alteration in the multimeric state of YB-1 within its complexes with -globin mRNA formed at different ratios of YB-1/mRNA. Complexes of YB-1 with 32P-labeled -globin mRNA were formed at indicated YB-1/mRNA ratios and fixed with 0.15% glutaraldehyde. RNA was destroyed by RNases A and T1 and by boiling with Mg2+ ions. The protein was analyzed using SDS–gel electrophoresis. Left, YB-1 without mRNA: lane 1, without fixation; lane 2, after fixation with glutaraldehyde, Coomassie stained. Middle and right, YB-1 from YB-1/mRNA complexes formed at indicated ratios of the components: lane 1, without addition of RNA; lanes 1–3, without fixation; lanes 4–9, after fixation with glutaraldehyde. Middle, Coomassie stained; right, radioautograph. (B) Analysis of domain YB-1 contacts with -globin mRNA within mRNP complexes. The scheme presents the domain organization of YB-1 and distribution of methionine residues. Complexes of YB-1 and 32P-labeled -globin mRNA were formed at indicated molar ratios, exposed to UV and treated with RNases A and T1. YB-1 was cleaved with cyanogen bromide at methionine residues. Left, gel autoradiograph after SDS–gel electrophoresis of YB-1 fragments resulting from cleavage is shown. Right, the graph demonstrates changes in the relative amount of radioactive RNA cross-linked with fragment I and fragment II of YB-1.
As mentioned above, YB-1 has two RNA-binding domains: the N-terminal CSD with consensus motifs RNP-1 and RNP-2 and the CTD with alternating clusters of positively and negatively charged amino acid residues (Figure 2B, upper scheme). We studied contacts of these domains with mRNA in YB-1–mRNA complexes with various YB-1/mRNA ratios. For this purpose, YB-1 was UV-cross-linked with 32P-labeled -globin mRNA. After exhaustive RNase digestion of mRNA, YB-1 labeled with radioactive RNA fragments was cleaved by BrCN at methionine residues concentrated in the first half of the CTD. The cleavage yielded two fragments: the larger N-fragment (designated as fragment I) contained the N-terminal A/P domain, the CSD and the beginning of the CTD, while the smaller one (fragment II) comprised the second half of CTD. The fragments were separated by SDS–gel electrophoresis, and radioactivity of each was determined (Figure 2B). As seen, the ratio between the RNA amount cross-linked with fragment I and the RNA amount cross-linked with fragment II considerably increased with increasing the YB-1/mRNA ratio, in particular in the range from 12:1 to 48:1. It is in this range that one can observe transition from active mRNA translation to complete protein synthesis inhibition both in a cell-free translation system (24,25,28) and in vivo in cultivated cells (27). This allows conclusion that at a comparatively low YB-1/mRNA ratio, typical of polysomal mRNPs, YB-1 binds to mRNA through its both RNA-binding domains, whereas high YB-1/mRNA ratios (24:1 and more) make it predominantly use its N-terminus with CSD. The CTDs are probably displaced from RNA by CSDs.
YB-1–mRNA complexes are dynamic, and YB-1 easily migrates from one mRNA molecule to another
In cell lysates, natural mRNPs are known to be stable enough to prevent YB-1 redistribution after addition of exogenous RNA (48). The informosome-like particles formed by exogenous RNA in cell lysates at physiological temperature proved to be as stable as natural mRNPs (49). However, in the cell, mRNPs are quite dynamic, and the set of mRNA-associated proteins changes upon mRNA transition from the nucleus to cytoplasm and from free mRNPs to polysomes (15,50,51). Moreover, it was shown that in translatable mRNPs in vivo proteins can leave mRNA and come back to it again after passing the ribosome (52).
In this study, we examined whether YB-1 can migrate from its complexes with mRNA to other mRNA molecules, i.e. whether its in vitro complex with mRNA is stable or dynamic. For this purpose, YB-1 was incubated with 32P-labeled -globin mRNA (molar ratio 36:1), then unlabeled -globin mRNA was added to the ratio of 9:1 and incubation was prolonged for additional time. Samples were analyzed in sucrose gradient (Figure 3A and B). If YB-1 could form stable complexes with mRNA and could not migrate to other RNA molecules, upon addition of unlabeled mRNA, the entire radioactivity would remain in the initial peak together with the major part of YB-1. But, as revealed by comparison of Figure 3A and B, the entire radioactive mRNA has shifted to a lighter complex and sedimented together with unlabeled added mRNA and almost all YB-1. So, YB-1 has been uniformly distributed over the initial and newly added mRNA. So, YB-1 can easily redistribute between mRNA molecules and consequently its complexes with RNA are dynamic. May be, natural mRNPs, as well as informosome-like particles formed in cell lysates with exogenous RNA at physiological temperature, are stabilized due to additional (minor) proteins and/or possible covalent modifications of YB-1.
Figure 3. Analysis of stability of YB-1 complexes with -globin mRNA. Complexes of YB-1 with 32P-labeled -globin mRNA were formed at YB-1/mRNA molar ratios of 36:1. The sample was incubated for 15 min at 30°C, then unlabeled -globin mRNA was added to a half of the sample to the final ratio of 9:1, and the preparations were incubated for additional 15 min at the same temperature. The preparations were layered onto 5–20% sucrose gradient and centrifuged in a SW-60 rotor at 45 000 r.p.m. for 4 h at 4°C. (A) Distribution of the complex formed by YB-1 with -globin mRNA at a molar ratio of 36:1. (B) Distribution of the complex formed by YB-1 with -globin mRNA at a molar ratio of 36:1 to which unlabeled -globin mRNA was added to the YB-1/mRNA ratio of 9:1. UV absorbance (solid line) at 254 nm and radioactivity (open circles) profiles are shown. YB-1 distribution over gradient fractions revealed by SDS–gel electrophoresis is shown below the plots.
AFM and EM studies of YB-1 and YB-1–RNA complexes
As it was shown in biochemical experiments, free YB-1 forms large multimeric complexes up to 800 kDa (22), see also Figure 2. Here, we visualized free YB-1 by EM and AFM (Figure 4). The multimeric protein was pre-fixed with glutaraldehyde. Independent of the preparation techniques (negative staining and shadowing) and the type of substrate (carbon film and mica), EM reveals the same shape and dimensions of the YB-1 multimer. This protein looks like quite homogeneous granules 9–10 nm high and 35–40 nm in diameter (Figure 4A and B). AFM images revealed particles with 8–10 nm height and 30–40 nm diameter, measured at half of the particle height (Figure 4C), which is in good agreement with EM data. The observed di- and trimers are probably the result of granule sticking to one another. It can be concluded that multimeric YB-1 is a flattened particle whose diameter is 3–4 times its height. It is not excluded that in solution YB-1 is a spherical particle that is flattened upon interaction with the surface. To obtain a YB-1 image in the form of a monomer, the protein was dissociated with 1 M LiCl. Presumably, high salt concentrations do not allow the monomerized YB-1 in such a way to give distinct AFM and EM images upon drying on the substrate (data not shown). In order to overcome this problem, monomerized YB-1 was imaged by AFM in solution in a high salt buffer (Figure 4D). Under these conditions, the protein molecules were detected as particles with an average height of about 4 nm.
Figure 4. EM and AFM images of free multimeric and monomeric YB-1. (A) Electron micrograph of multimeric YB-1. The protein, fixed with 0.15% glutaraldehyde, was adsorbed onto a carbon film and shadowed with platinum-carbon. The same images of shadowed fixed YB-1 were obtained at its adsorbtion onto mica (not shown). (B) Electron micrograph of negatively stained fixed YB-1 deposited on mica. (C) AFM image of fixed YB-1 on APS-mica imaged in air. An aliquot of 5 μl of fixed YB-1 (0.01 μg/μl) was deposited onto the substrate, rinsed with water, dried and imaged. (D) AFM image of YB-1 in high salt solution on APS-mica substrate. YB-1 (0.05 μg/μl) was incubated in a high salt buffer (10 mM HEPES–KOH pH 7.6, 1 M LiCl) for 20 min at 30°C. Cross-sections made along the marked lines in (C) and (D) are shown under the respective AFM images. These cross-sections represent the height of the complexes and its variations. Triangles in the fields indicate the ends of the section lines.
As shown above in biochemical experiments, in YB-1–mRNA complexes, YB-1 associates with RNA as a monomer when the YB-1/mRNA ratio is relatively low. As the YB-1 content in the complex is increased, YB-1 forms a multimer with a molecular mass comparable to that of a free multimeric protein. To visualize YB-1 in complex with RNA, we studied such complexes formed with RNAs of various lengths. Apart from -globin mRNA (660 nt), we used 2Luc mRNA (3000 nt). The complexes were pre-fixed with glutaraldehyde. First, unsaturated complexes were studied by means of AFM (Figure 5). In these complexes, the YB-1/RNA mass ratio was 0.5:1, which corresponds to a molar ratio of 3:1 for globin mRNA and to 14:1 for 2Luc mRNA. Since mRNA has a highly developed secondary structure and monomeric YB-1 is small in size, it was impossible to distinguish separate protein monomers within the complex. Nevertheless, it is clearly seen (especially for the more extended 2Luc mRNA) that the complexes are quite unfolded and that their heights are below 2 nm, which is indicative, most probably, that YB-1 is monomeric in these complexes. The average complex height is 1.7 nm for globin mRNA and 1.9 nm for 2Luc mRNA. The fact that these values are lower than the height of free monomerized YB-1 measured on APS-mica in solution (4 nm) can be explained by both the YB-1 interaction with mRNA and by the effect of YB-1 drying on APS-mica. The latter effect, previously reported for several other proteins (53,54), is more probable, since in our AFM experiments monomerized YB-1 formed 1.5–2.0 nm thick monolayers upon drying on APS-mica surface (data not shown). Thus, AFM images of unsaturated mRNPs show no large multimeric YB-1 on mRNAs, thereby proving its dissociation upon binding to RNA, which complies with the biochemical data on monomerization of YB-1 in complexes with a low YB1/mRNA ratio (see Figure 2A).
Figure 5. AFM images of unsaturated YB-1–mRNP complexes in air. Unsaturated complexes of YB-1 with -globin mRNA (A) and 2Luc mRNA (B) were fixed with glutaraldehyde, deposited on APS-mica and imaged as described in Materials and Methods. Left, typical field; right, magnified images of individual complexes and cross-sections made along the marked lines on the fields. Triangles in the fields indicate the ends of the section lines.
In the next series of experiments, we studied saturated complexes of YB-1 with the same mRNAs. These complexes were formed at the YB-1/RNA mass ratio 8:1, which corresponds to a molar ratio of 48:1 for globin mRNA and to 224:1 for 2Luc mRNA. After fixation, the complexes were separated from the RNA-unbound protein as described in Materials and Methods. As compared to unsaturated complexes, the saturated ones appeared to be much more compact (Figure 6). The YB-1–globin mRNA complexes were seen as 6.2 ± 0.9 nm high globules, i.e. they were three times as high as the largest unsaturated complexes (Figure 6A). Saturated complexes look either like single globules or represent dimers, resulting, probably, from the association of the two globules. The distribution of the number of globules per complex shows that the major part of the complexes contains only one globule (Figure 6A, right). The dimensions of these globules imply that each contains several YB-1 molecules and support the conclusion on YB-1 multimerization upon mRNA saturation.
Figure 6. AFM images of saturated mRNP complexes. Fixed saturated complexes of YB-1 with -globin mRNA (A), 2Luc mRNA (B) and TMV RNA (C) were deposited on APS-mica and imaged in air. Left, typical field; right, cross-sections made along the marked lines on the fields. Triangles in the fields indicate the ends of the section lines. Histograms illustrating the distribution of number of multimeric globules per complex are shown on the right.
AFM experiments with 2Luc mRNA that is 4.5 times longer than globin mRNA revealed formation of mRNPs composed of several successive globular particles 6.8 ± 0.9 nm high (Figure 6B). These particles' dimensions are very close to those of ones formed on globin mRNA, which allows considering them to be YB-1 multimers, as well. Unlike globin mRNA, most of 2Luc mRNA complexes consisted of four successive globules, the center-to-center distance of adjacent globules being 20–25 nm and the average mRNP length being 130 ± 20 nm (Figure 6B, right). Figure 7A presenting their magnified images shows the structure of these complexes in more detail. EM images (Figure 7B) allow measuring the real diameter of separate globules of the complex that is about 20 nm. Thus, like free multimeric YB-1, separate globules of RNA-associated YB-1 are flattened and have a diameter several times exceeding their height. However, the size of globules of mRNA-bound protein differs from that of free multimeric protein (specifically, in height). This clearly shows that the formation of saturated complexes is a two-step event: first, the initially multimeric YB-1 dissociates to monomers upon its binding to mRNA; and second, the YB-1 multimers are formed again as RNA is saturated with the protein.
Figure 7. Images of saturated complexes formed with 2Luc mRNA at high magnification. (A) AFM images of individual complexes in air. Right, cross-sections made along the marked lines on the corresponding images. (B) Electron micrograph of the negatively stained complexes.
Since globin mRNA (660 nt) allows formation mainly of one multimeric globule and 2Luc mRNA (3000 nt) forms four of them, it may be suggested that one globule is formed at a 600–700 nt fragment of RNA. In order to verify this suggestion, we studied saturated complexes with still more extended TMV RNA (6000 nt). Like with other RNAs used, YB-1 forms multimers with the same globule height of 6.5 ± 0.8 nm on TMV RNA (Figure 6C). Most of these complexes contain eight multimeric globules (see the diagram in Figure 6C), thus confirming our suggestion. The fact that a considerable number of YB-1–TMV RNA complexes contain less than eight globules can be explained by a higher probability of their random degradation in the course of sample storage and preparation. The measured average length of the saturated YB-1–TMV RNA complexes was 270 ± 50 nm. According to EM data, the average specific length of single-stranded RNA is 260 nm/kb (55) and, therefore, the TMV RNA length is 1560 nm. According to AFM, the length of a fully unfolded denatured TMV RNA is 1500 nm (56). Hence, saturation of TMV RNA with YB-1 is accompanied by its 4- to 5-fold compactization.
Thus, within the formed saturated complexes mRNA is noticeably compact, and the general structure of these complexes is of a ‘beads-on-a-string’ type, where one ‘bead’ is formed on a 600–700 nt long segment of the RNA chain.
DISCUSSION
In this study, we analyzed complexes formed by YB-1, the major protein of cytoplasmic mRNPs, with mRNA at various molar ratios, including those typical for translatable (polysomal) and untranslatable (free) mRNPs. It has been found that saturation of globin mRNA with YB-1 gives rise to the formation of particles with basic physico-chemical parameters of natural free mRNPs, namely, a sedimentation coefficient three times as high as that of the used -globin mRNA and a buoyant density in CsCl equal to 1.39 g/cm3. This value exactly coincides with that of natural free mRNPs and corresponds to the 3:1 YB-1/mRNA weight ratio. Besides, earlier we showed that YB-1-bound mRNA is extremely sensitive to endoribonucleases, which is also characteristic of natural mRNPs and shows surface location of mRNA in the particles (22).
YB-1 and its homologs are universal major mRNP proteins of not only mammalian cells but of cells from other organisms too. It was shown recently that two isoforms of the homologous proteins Ct-p40 and Ct-p50 are associated with mRNA from the insect Chironomus tentans (57). Two proteins of this family were detected in amphibia Xenopus laevis (34). One of these proteins, FRGY1, packs mRNA of mature somatic cells and shows high homology to mammalian YB-1. The other, FRGY2, having two isoforms p54/p56, packs masked mRNA in oocytes. FRGY2 is far less homologous to YB-1 beyond CSD, although its isoelectric point, amino acid composition and charged amino acid residue distribution in CTD coincide with those of YB-1. As shown earlier, FRGY2 also forms with mRNA, at its saturation, RNA–protein complexes with a buoyant density of 1.4 g/cm3 (58,59). This suggests that in various multicellular organisms from insects to man, YB-1 and its homologs are the major structure-forming mRNP proteins. Hence, the complexes between rabbit YB-1 and mRNA described here can serve as a simplest model system for studying the general principles of structural organization of natural mRNPs in all kinds of eukaryotic cells.
Free YB-1 and its homologs are multimeric (22,37,38). In this connection, a question arises as to whether YB-1 remains multimeric when binding to mRNA or, as we suggested earlier, it dissociates in monomers and afterwards restores its initial multimeric form when mRNA has been protein-saturated. Our experiments on rabbit YB-1 showed that in forming a YB-1–mRNA complex with a low YB-1/mRNA ratio, YB-1 binds to mRNA predominantly as a monomer, which points to dissociation of the YB-1 multimer upon interaction with mRNA. According to AFM, such complexes are relatively unfolded.
As the YB-1 amount on mRNA is increased, homo-oligomeric YB-1 complexes are formed and grow to their maximal size; so, mRNA with a higher molecular weight has several protein globules close in size, and the more extended mRNA, the greater the number of globules. The binding of YB-1 to mRNA was found to be a non-cooperative process. The known protein/RNA weight ratio (3:1) and the number of protein globules on mRNA molecules of various lengths allow calculating the molecular mass of a protein globule, that was about 700 kDa and hence, was formed by about twenty YB-1 molecules. In saturated complexes, one YB-1 molecule occupies on average 35 nt of mRNA, and the surface of one protein globule can house a 700 nt mRNA segment, the RNA compaction coefficient being about 5.
Recent EM studies of saturated complexes of FRGY2 with TFIIIA mRNA (1.5 kb) and histone H1 mRNA (0.75 kb) (59) showed them as tight thickened structures some of which contained large particles with a diameter of 10–20 nm. Although the presented images hardly allow estimating the number of particles per individual RNA, the general view of the saturated complexes is in agreement with our AFM and EM images. The diameter of some RNA-housed particles (10–20 nm), reported by the authors in (59), is slightly less than that of YB-1 particles (20–25 nm) in our studies. This discrepancy can be explained by an essential difference between FRGY2 and YB-1 that are only 64% homologous. Nevertheless, it can be assumed that the formation of large successive mRNA-associated particles, whose number is proportional to the RNA length, is the general mode of mRNA packing by YB-1 homologs in the eukaryotic cytoplasm.
This model of structural organization of YB-1/mRNA complexes resembles the model for organization of hnRNPs proposed previously (8,60). Both the models imply the formation of large (1000 kDa) multimeric protein complexes that pack RNA on their surface in the ‘beads-on-a-string’ mode, the only difference being that in the case of cytoplasmic mRNPs, close in size globular protein structures are formed by YB-1 alone, rather than by six various proteins in a certain proportion in the case of hnRNPs.
It follows from our experiments that the mode of interaction between YB-1 and mRNA depends on the YB-1/mRNA ratio: at a low value of this ratio, monomeric YB-1 binds to mRNA through its two RNA-binding domains, CSD and CTD. As the YB-1 amount on mRNA is increased, the CTD binding decreases, and the association with mRNA is realized mostly through the YB-1 N-terminus with CSD. This can be explained by a successful CSD competition for mRNA that leads to CTD displacement. The displaced CTDs can interact with one another, thereby restoring the multimeric form of YB-1 that it had in the free state. It is reasonable to suggest that the CTDs are packed inside the formed protein globules, whereas the CSDs are located on their surface (Figure 8). In saturated complexes, the surface-attached mRNA can contact only the CSDs.
Figure 8. Scheme showing involvement of YB-1 in the formation of polysomal and free cytoplasmic mRNPs. For details, see text.
Surface location of mRNA is essential for its recognition by different regulatory proteins, which is necessary for its selective activation, selective degradation or specific localization in the cell. As shown previously, such location makes mRNA very sensitive to endonucleases (22). But its life span, both in the cell and in cell lysates, is strongly increased by YB-1 (35). This paradox can be easily explained by the fact that degradation of mRNA in the cell usually starts from its ends and is realized by exonucleases, and in saturated complexes, the mRNA ends are probably buried and inaccessible to exonucleases. This should be verified experimentally. The inaccessibility of mRNA ends to proteins hinders their interaction with translation initiation factors and leads to protein synthesis inhibition. It was demonstrated in lysates and in systems reconstituted from pure components of the protein-synthesizing apparatus that at a high YB-1/mRNA ratio, YB-1 does prevent interaction of mRNA with translation initiation factors (28,35).
As mentioned above, the mRNA transition to a translatable state is accompanied both by the binding of PABP to poly(A) tail sequence of poly(A)+ mRNA and by an approximately 2-fold decrease in the YB-1 amount on mRNA. The mechanism of this transition has remained unclear yet. It can be postulated that a decrease in the amount of YB-1 on mRNA within polysomes would change the interaction of YB-1 with mRNA in the same way as in our model system. In such a case, YB-1 binds to mRNA not only through its CSD, but through CTD as well, which is accompanied by dissociation of multimetic YB-1 complexes into monomers and promotes mRNP unfolding and exposure of mRNA termini. This contributes to mRNA interaction with translation initiation factors and mRNA involvement in the process of translation (Figure 8).
It is notable that complexes of recombinant YB-1 with mRNA are reversible, and the protein migrates easily to the newly added mRNA. The situation within the intact lysate may be different. It is known that after incubation in physiological conditions, both natural mRNPs and those formed on the exogenous mRNA added to lysates are irreversible and do not lose their proteins during incubation even with a high excess of exogenous RNA (48,49). This suggests the existence of factors, which stabilize the mRNP structure. Such factors could be covalent modifications of YB-1 in an eukaryotic cell that are lacking in the protein expressed in E.coli. Minor masking mRNP proteins that may play the role of peculiar clips, as proposed earlier (1), can be in the capacity of such factors.
ACKNOWLEDGEMENTS
We thank Dr J. Ilan for providing -globin cDNA, Dr I. Shatsky for luciferase cDNA and Dr E. Karger for TMV RNA. We also thank T. Kuvshinkina and E. Serebrova for their help in preparing this manuscript. This work was supported in part by the Russian Academy of Sciences (Programmes on ‘Molecular and Cellular Biology’ and ‘Basic Sciences to Medicine’) and by a grant from the President of the Russian Federation (#1959.2003.4).
REFERENCES
Spirin,A.S. ( (1996) ) Masked and translatable messenger ribonucleoproteins in higher eukaryores. In Hershey,J.W.B., Mathews,M.B. and Sonenberg,N. (eds), Translational Control. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 319–334.
Samarina,O.P. ( (1996) ) hnRNP particles. Bioessays, , 18, , 595–601.
Dreyfuss,G., Kim,V.N. and Kataoka,N. ( (2002) ) Messenger-RNA-binding proteins and the messages they carry. Nature Rev. Mol. Cell. Biol., , 3, , 195–205.
Keene,J.D. ( (2001) ) Ribonucleoprotein infrastructure regulating the flow of genetic information between the genome and the proteome. Proc. Natl Acad. Sci. USA, , 98, , 7018–7024.
Kloc,M., Zearfoss,N.R. and Etkin,L.D. ( (2002) ) Mechanisms of subcellular mRNA localization. Cell, , 108, , 533–544.
Tourriere,H., Chebli,K. and Tazi,J. ( (2002) ) mRNA degradation machines in eukaryotic cells. Biochimie, , 84, , 821–837.
Preobrazhensky,A.A. and Spirin,A.S. ( (1978) ) Informosomes and their protein components: the present state of knowledge. Prog. Nucleic Acid Res. Mol. Biol., , 21, , 1–38.
Huang,M., Rech,J.E., Northington,S.J., Flicker,P.F., Mayeda,A., Krainer,A.R. and LeStourgeon,W.M. ( (1994) ) The C-protein tetramer binds 230 to 240 nucleotides of pre-mRNA and nucleates the assembly of 40S heterogeneous nuclear ribonucleoprotein particles. Mol. Cell. Biol., , 14, , 518–533.
McAfee,J.G., Huang,M., Soltaninassab,S.R., Rech,J.E., Iyengar,S. and LeStourgeon,W.M. ( (1997) ) The packaging of pre-mRNA. In Krainer,A.R. (ed.), Eukaryotic mRNA Processing. Oxford University Press, pp. 68–102.
Ovchinnikov,L.P., Skabkin,M.A., Ruzanov,P.V. and Evdokimova,V.M. ( (2001) ) Major mRNP proteins in the structural organization and function of mRNA in eukaryotic cells. Mol. Biol. (Mosk.), , 35, , 548–558.
Morel,C., Kayibanda,B. and Scherrer,K. ( (1971) ) Proteins associated with globin messenger RNA in avian erythroblasts: isolation and comparison with the proteins bound to nuclear messenger-like RNA. FEBS Lett., , 18, , 84–88.
Blobel,G. ( (1972) ) Protein tightly bound to globin mRNA. Biochem. Biophys. Res. Commun., , 47, , 88–95.
Kumar,A. and Pederson,T. ( (1975) ) Comparison of proteins bound to heterogeneous nuclear RNA and messenger RNA in HeLa cells. J. Mol. Biol., , 96, , 353–365.
Jain,S.K., Pluskal,M.G. and Sarkar,S. ( (1979) ) Thermal chromatography of eukaryotic messenger ribonucleoprotein particles on oligo (dT)-cellulose. Evidence for common mRNA-associated proteins in various cell types. FEBS Lett., , 97, , 84–90.
Dreyfuss,G. ( (1986) ) Structure and function of nuclear and cytoplasmic ribonucleoprotein particles. Annu. Rev. Cell Biol., , 2, , 459–498.
Blobel,G. ( (1973) ) A protein of molecular weight 78,000 bound to the polyadenylate region of eukaryotic messenger RNAs. Proc. Natl Acad. Sci. USA, , 70, , 924–928.
van Venrooij,W.J., van Eekelen,C.A., Jansen,R.T. and Princen,J.M. ( (1977) ) Specific poly-A-binding protein of 76,000 molecular weight in polyribosomes is not present on poly A of free cytoplasmic mRNP. Nature, , 270, , 189–191.
Minich,W.B., Maidebura,I.P. and Ovchinnikov,L.P. ( (1993) ) Purification and characterization of the major 50-kDa repressor protein from cytoplasmic mRNP of rabbit reticulocytes. Eur. J. Biochem., , 212, , 633–638.
Murray,M.T., Schiller,D.L. and Franke,W.W. ( (1992) ) Sequence analysis of cytoplasmic mRNA-binding proteins of Xenopus oocytes identifies a family of RNA-binding proteins. Proc. Natl Acad. Sci. USA, , 89, , 11–15.
Tafuri,S.R. and Wolffe,A.P. ( (1993) ) Dual roles for transcription and translation factors in the RNA storage particles of Xenopus oocytes. Trends Cell Biol., , 3, , 94–98.
Gu,W., Tekur,S., Reinbold,R., Eppig,J.J., Choi,Y.C., Zheng,J.Z., Murray,M.T. and Hecht,N.B. ( (1998) ) Mammalian male and female germ cells express a germ cell-specific Y-Box protein, MSY2. Biol. Reprod., , 59, , 1266–1274.
Evdokimova,V.M., Wei,C.L., Sitikov,A.S., Simonenko,P.N., Lazarev,O.A., Vasilenko,K.S., Ustinov,V.A., Hershey,J.W. and Ovchinnikov,L.P. ( (1995) ) The major protein of messenger ribonucleoprotein particles in somatic cells is a member of the Y-box binding transcription factor family. J. Biol. Chem., , 270, , 3186–3192.
Katsu,Y., Yamashita,M. and Nagahama,Y. ( (1997) ) Isolation and characterization of goldfish Y box protein, a germ-cell-specific RNA-binding protein. Eur. J. Biochem., , 249, , 854–861.
Minich,W.B. and Ovchinnikov,L.P. ( (1992) ) Role of cytoplasmic mRNP proteins in translation. Biochimie, , 74, , 477–483.
Evdokimova,V.M., Kovrigina,E.A., Nashchekin,D.V., Davydova,E.K., Hershey,J.W. and Ovchinnikov,L.P. ( (1998) ) The major core protein of messenger ribonucleoprotein particles (p50) promotes initiation of protein biosynthesis in vitro. J. Biol. Chem., , 273, , 3574–3581.
Pisarev,A.V., Skabkin,M.A., Thomas,A.A., Merrick,W.C., Ovchinnikov,L.P. and Shatsky,I.N. ( (2002) ) Positive and negative effects of the major mammalian messenger ribonucleoprotein p50 on binding of 40S ribosomal subunits to the initiation codon of beta-globin mRNA. J. Biol. Chem., , 277, , 15445–15451.
Davydova,E.K., Evdokimova,V.M., Ovchinnikov,L.P. and Hershey,J.W. ( (1997) ) Overexpression in COS cells of p50, the major core protein associated with mRNA, results in translation inhibition. Nucleic Acids Res., , 25, , 2911–2916.
Nekrasov,M.P., Ivshina,M.P., Chernov,K.G., Kovrigina,E.A., Evdokimova,V.M., Thomas,A.A., Hershey,J.W. and Ovchinnikov,L.P. ( (2003) ) The mRNA-binding protein YB-1 (p50) prevents association of the eukaryotic initiation factor eIF4G with mRNA and inhibits protein synthesis at the initiation stage. J. Biol. Chem., , 278, , 13936–13943.
Darnbrough,C.H. and Ford,P.J. ( (1981) ) Identification in Xenopus laevis of a class of oocyte-specific proteins bound to messenger RNA. Eur. J. Biochem., , 113, , 415–424.
Richter,J.D. and Smith,L.D. ( (1984) ) Reversible inhibition of translation by Xenopus oocyte-specific proteins. Nature, , 309, , 378–380.
Herbert,T.P. and Hecht,N.B. ( (1999) ) The mouse Y-box protein, MSY2, is associated with a kinase on non-polysomal mouse testicular mRNAs. Nucleic Acids Res., , 27, , 1747–1753.
Sommerville,J. ( (1990) ) RNA-binding phosphoproteins and the regulation of maternal mRNA in Xenopus. J. Reprod. Fertil. Suppl., , 42, , 225–233.
Braddock,M., Muckenthaler,M., White,M.R., Thorburn,A.M., Sommerville,J., Kingsman,A.J. and Kingsman,S.M. ( (1994) ) Intron-less RNA injected into the nucleus of Xenopus oocytes accesses a regulated translation control pathway. Nucleic Acids Res., , 22, , 5255–5264.
Wolffe,A.P., Tafuri,S., Ranjan,M. and Familari,M. ( (1992) ) The Y-box factors: a family of nucleic acid binding proteins conserved from Escherichia coli to man. New Biol., , 4, , 290–298.
Evdokimova,V., Ruzanov,P., Imataka,H., Raught,B., Svitkin,Y., Ovchinnikov,L.P. and Sonenberg,N. ( (2001) ) The major mRNA-associated protein YB-1 is a potent 5' cap-dependent mRNA stabilizer. EMBO J., , 20, , 5491–5502.
Matsumoto,K. and Wolffe,A.P. ( (1998) ) Gene regulation by Y-box proteins: coupling control of transcription and translation. Trends Cell Biol., , 8, , 318–323.
Tafuri,S.R. and Wolffe,A.P. ( (1992) ) DNA binding, multimerization, and transcription stimulation by the Xenopus Y box proteins in vitro. New Biol., , 4, , 349–359.
Gaudreault,I., Guay,D. and Lebel,M. ( (2004) ) YB-1 promotes strand separation in vitro of duplex DNA containing either mispaired bases or cisplatin modifications, exhibits endonucleolytic activities and binds several DNA repair proteins. Nucleic Acids Res., , 32, , 316–327.
Evdokimova,V.M. and Ovchinnikov,L.P. ( (1999) ) Translational regulation by Y-box transcription factor: involvement of the major mRNA-associated protein, p50. Int. J. Biochem. Cell Biol., , 31, , 139–149.
Ustinov,V.A., Skabkin,M.A., Nashchekin,D.V., Evdokimova,V.M. and Ovchinnikov,L.P. ( (1996) ) Major p50 protein of the somatic cell cytoplasmic mRNP: expression in Escherichia coli, isolation, and some properties of the recombinant protein. Biochemistry (Mosc.), , 61, , 559–566.
Pokrovskaya,I.D. and Gurevich,V.V. ( (1994) ) In vitro transcription: preparative RNA yields in analytical scale reactions. Anal. Biochem., , 220, , 420–423.
Aviv,H. and Leder,P. ( (1972) ) Purification of biologically active globin messenger RNA by chromatography on oligothymidylic acid-cellulose. Proc. Natl Acad. Sci. USA, , 69, , 1408–1412.
Filonov,A.S. and Yaminsky,I.V. ( (1997) ) SPM Control and Image Processing Software. Chichester: Advance Technologies Center, Moscow.
Shlyakhtenko,L.S., Potaman,V.N., Sinden,R.R. and Lyubchenko,Y.L. ( (1998) ) Structure and dynamics of supercoil-stabilized DNA cruciforms. J. Mol. Biol., , 280, , 61–72.
Kiselyova,O.I., Galyamov,M.O., Nasikan,N.S., Yaminsky,I.V., Karpova,O.V. and Novikov,V.K. ( (2002) ) Scanning probe microscopy of biomacromolecules: nucleic acids, proteins and their complexes. In Buzanaeva,E.V. and Scharff,P. (eds), Frontiers of Multifunctional Nanosystems. Kluwer Academic Publishers, Dordrecht, pp. 321–330.
Vasiliev,V.D. and Koteliansky,V.E. ( (1979) ) Freeze-drying and high-resolution shadowing in electron microscopy of Escherichia coli ribosomes. Meth. Enzymol., , 59, , 612–629.
Valentine,R.C., Shapiro,B.M. and Stadtman,E.R. ( (1968) ) Regulation of glutamine synthetase. XII. Electron microscopy of the enzyme from Escherichia coli. Biochemistry, , 7, , 2143–2152.
Ovchinnikov,L.P., Avanesov,A.C. and Spirin,A.S. ( (1969) ) Informosomes from loach embryos. Mol. Biol. (Mosk.), , 3, , 465–472.
Ovchinnikov,L.P., Voronina,A.S., Stepanov,A.S., Belitsina,N.V. and Spirin,A.S. ( (1968) ) Informosome-like complexes were formed by RNA adding to animal cell homogenates. Mol. Biol. (Mosk.), , 2, , 752–763.
Bag,J. ( (1984) ) Cytoplasmic mRNA–protein complexes of chicken muscle cells and their role in protein synthesis. Eur. J. Biochem., , 141, , 247–254.
Daneholt,B. ( (2001) ) Assembly and transport of a premessenger RNP particle. Proc. Natl Acad. Sci. USA, , 98, , 7012–7017.
Greenberg,J.R. ( (1981) ) The polyribosomal mRNA–protein complex is a dynamic structure. Proc. Natl Acad. Sci. USA, , 78, , 2923–2926.
Lyubchenko,Y.L., Oden,P.I., Lampner,D., Lindsay,S.M. and Dunker,K.A. ( (1993) ) Atomic force microscopy of DNA and bacteriophage in air, water and propanol: the role of adhesion forces. Nucleic Acids Res., , 21, , 1117–1123.
Lyubchenko,Y.L. and Shlyakhtenko,L.S. ( (1997) ) Visualization of supercoiled DNA with atomic force microscopy in situ. Proc. Natl Acad. Sci. USA, , 94, , 496–501.
Glass,J. and Wertz,G.W. ( (1980) ) Different base per unit length ratios exist in single-stranded RNA and single-stranded DNA. Nucleic Acids Res., , 8, , 5739–5751.
Drygin,Y.F., Bordunova,O.A., Gallyamov,M.O. and Yaminsky,I.V. ( (1998) ) Atomic force microscopy examination of tobacco mosaic virus and virion RNA. FEBS Lett., , 425, , 217–221.
Soop,T., Nashchekin,D., Zhao,J., Sun,X., Alzhanova-Ericsson,A.T., Bjorkroth,B., Ovchinnikov,L. and Daneholt,B. ( (2003) ) A p50-like Y-box protein with a putative translational role becomes associated with pre-mRNA concomitant with transcription. J. Cell Sci., , 116, , 1493–1503.
Kick,D., Barrett,P., Cummings,A. and Sommerville,J. ( (1987) ) Phosphorylation of a 60 kDa polypeptide from Xenopus oocytes blocks messenger RNA translation. Nucleic Acids Res., , 15, , 4099–4109.
Matsumoto,K., Tanaka,K.J., Aoki,K., Sameshima,M. and Tsujimoto,M. ( (2003) ) Visualization of the reconstituted FRGY2-mRNA complexes by electron microscopy. Biochem. Biophys. Res. Commun., , 306, , 53–58.
Samarina,O.P., Lukanidin,E.M., Molnar,J. and Georgiev,G.P. ( (1968) ) Structural organization of nuclear complexes containing DNA-like RNA. J. Mol. Biol., , 33, , 251–263.(Maxim A. Skabkin, Olga I. Kiselyova1, Ko)
* To whom correspondence should be addressed. Tel/Fax: +7 095 924 04 93; Email: ovchinn@vega.protres.ru
ABSTRACT
YB-1 is a universal major protein of cytoplasmic mRNPs, a member of the family of multifunctional cold shock domain proteins (CSD proteins). Depending on its amount on mRNA, YB-1 stimulates or inhibits mRNA translation. In this study, we have analyzed complexes formed in vitro at various YB-1 to mRNA ratios, including those typical for polysomal (translatable) and free (untranslatable) mRNPs. We have shown that at mRNA saturation with YB-1, this protein alone is sufficient to form mRNPs with the protein/RNA ratio and the sedimentation coefficient typical for natural mRNPs. These complexes are dynamic structures in which the protein can easily migrate from one mRNA molecule to another. Biochemical studies combined with atomic force microscopy and electron microscopy showed that mRNA–YB-1 complexes with a low YB-1/mRNA ratio typical for polysomal mRNPs are incompact; there, YB-1 binds to mRNA as a monomer with its both RNA-binding domains. At a high YB-1/mRNA ratio typical for untranslatable mRNPs, mRNA-bound YB-1 forms multimeric protein complexes where YB-1 binds to mRNA predominantly with its N-terminal part. A multimeric YB-1 comprises about twenty monomeric subunits; its molecular mass is about 700 kDa, and it packs a 600–700 nt mRNA segment on its surface.
INTRODUCTION
In eukaryotic cells, all mRNAs and their precursors are permanently associated with various proteins to form heterogeneous nuclear RNPs (hnRNPs) in the nucleus and translatable (polysomal) and untranslatable free messenger RNPs (mRNPs) in the cytoplasm (1,2). The hnRNP proteins pack pre-mRNAs (mRNA precursors), regulate the pre-mRNA processing and participate in mRNA transport from the nucleus to the cytoplasm (3). Cytoplasmic mRNP proteins participate in the mRNA packing and regulate its translation, life span and intracellular localization (1,4–6). Some common features are reported for hnRNPs and mRNPs (2,7): (i) a broad sedimentation distribution that reflects heterogeneity of mRNAs and pre-mRNAs in size; RNP sedimentation coefficients are usually 2.5–3 times higher than those of RNA; (ii) a constant high protein/RNA weight ratio that is 3 for hnRNPs and free cytoplasmic mRNPs and 2 for translatable polysomal mRNPs; this ratio determines mRNP unique buoyant density values in CsCl that are 1.39–1.41 g/cm3 and about 1.45 g/cm3, respectively; (iii) a high sensitivity of pre-mRNAs and mRNAs to endoribonucleases, which is indicative of surface location of the RNAs.
The unique structure of hnRNPs is known to be formed mostly by hnRNP major proteins of the A-, B- and C-groups (8,9). Similarly, in mRNPs, the role of packing proteins may be played by their own universal major proteins present in many copies and capable of binding to any mRNA (1,10). In mRNPs from various tissues and cell lines, there are two clearly observed major proteins—p50 (Mr 50 000–55 000) and p70 (Mr 70 000–80 000) (11–15). The protein p70 displays a high affinity for poly(A) sequences and is localized on the mRNA poly(A) tail (16,17). This protein has been termed the poly(A) binding protein (PABP). PABP was detected mostly in polysomal mRNPs with poly(A)+mRNA. Therefore, PABP can hardly be a universal packing protein that determines unique physico-chemical properties of all mRNPs. The other major mRNP protein, p50 (a 36 kDa protein with anomalous electrophoretic mobility in SDS–PAGE typical of a 50 kDa protein), seems much more suitable for the discussed function. It is detected both in free and polysomal mRNPs and is associated with all kinds of mRNA nucleotide sequences. Its affinity for mRNA is very high (Kd 4 nM) (18). A comparative analysis of the primary structure of p50 preparations from mRNPs of Xenopus laevis oocytes (19), murine oocytes and spermatocytes (20,21), rabbit reticulocytes (22), and goldfish oocytes (23) showed strong homology of all these proteins with the human transcription factor YB-1 and classified them as members of the family of cold shock domain proteins (CSD proteins).
The mRNA transition from its untranslatable to translatable state is accompanied by a 2-fold decrease of the mRNA-associated YB-1 (24). In the cell-free translation system with the increasing YB1/mRNA weight ratio up to 2, YB-1 progressively stimulates protein synthesis at the initiation stage (24–26). A further growth of this ratio causes protein synthesis inhibition both in the cell-free system (24) and in the cell (27); at YB1/mRNA weight ratios from 5 to 6 in the cell-free system, the translation was inhibited completely. The mRNA saturation with YB-1 prevents mRNA interaction with translation initiation factors, thereby causing mRNA transition from polysomes to free mRNPs (28). The stored untranslatable mRNPs of amphibious and mammalian gametes contain significant amounts of YB-1 homologous proteins (29–31). The involvement of stored mRNAs from Xenopus oocytes in translaton just after fertilization is accompanied by dephosphorylation of FRGY2, the YB-1 homolog, resulting in a decrease of its affinity to mRNA (32,33). Then a gradual degradation of the protein occurs and it is substituted by its somatic homolog FRGY1 (34).
YB-1 considerably increases the mRNA life span in cells and cell lysates, probably by protecting mRNA from exonucleases (35). Such an activity of YB-1 could contribute to the storage of mRNA in a form of free mRNPs.
YB-1 and its closest homologs consist of three domains: a short N-terminal Ala- and Pro-rich domain (A/P domain) followed by a central cold shock domain (CSD) and a C-terminal domain (CTD). The CSD domain has RNP-1 and RNP-2 consensus motifs and displays DNA- and RNA-binding activity. The CTD domain is formed by alternating clusters of positively and negatively charged amino acid residues, non-specifically binds RNA and can undergo oligomerization (36). It was shown that free (mRNA-unbound) YB-1 and FRGY2 form homomultimeric complexes whose molecular mass, in the case of YB-1, reaches 800 kDa (22,37,38).
The properties of YB-1 and its separate domains allow us to suggest that in translatable mRNPs with a comparatively low YB-1/mRNA ratio, YB-1 molecules associate with mRNA as monomers through their both RNA-binding domains, thus giving rise to a relatively unfolded mRNA conformation. In untranslatable mRNPs with a high YB-1/mRNA ratio, the interaction between YB-1 molecules and mRNA is realized only through CSDs, which displace CTDs from the complex with mRNA. The displaced CTDs are likely to interact with one another. As a result, YB-1 molecules form a multimeric compact protein complex with a tightly packed mRNA on its surface fully accessible to endoribonucleases. However, within particles with such a conformation the mRNA termini are most probably buried, which makes the mRNA resistant to exoribonucleases and the mRNA termini inaccessible for interaction with translation initiation factors, and first of all, with eIF4F and PABP (10,35,39).
The current study was focused on mRNA–YB-1 complexes and its aim was to find out whether (i) YB-1 alone, without other mRNP proteins, is capable of forming complexes with typical properties of natural mRNPs; (ii) YB-1 is monomeric when binding to mRNA at a low YB-1/mRNA ratio; (iii) both RNA-binding domains of YB-1 participate in this binding; (iv) mRNA-associated YB-1 undergoes multimerization at higher YB-1/mRNA ratios; (v) in these conditions, YB-1 uses only one (which of the two?) or both of its RNA-binding domains to associate with mRNA.
We have shown that the YB-1-saturated mRNA forms particles whose sedimentation coefficient and buoyant density are typical of natural free mRNPs. At a low YB-1/mRNA ratio, YB-1 binds to mRNA as a monomer using its both RNA-binding domains; as this ratio is increased, YB-1 becomes multimeric and predominantly uses CSD for its interaction with mRNA. The relatively short globin mRNA is packed on the surface of only one multimeric protein globule, whereas larger YB-1-associated RNAs form polysome-like ‘beads-on-a-string’-type structures where the number of globules involved is proportional to the length of RNA used.
MATERIALS AND METHODS
Preparation of recombinant YB-1
The synthesis of recombinant rabbit YB-1 was induced in Escherichia coli BL21(DE3) transformed with plasmid pET-3-1-YB-1 (p50) by isopropyl-?-D-thiogalactopyranoside (IPTG), and the protein was purified by chromatography on heparin–Sepharose 4B and Superose 12 HR 10/30 columns (Amersham Biosciences) as described in (25,40). The YB-1 preparations were concentrated and dialyzed against a buffer containing 10 mM HEPES–KOH, pH 7.6 and 150 ml KCl, and stored at –70°C in aliquots. The protein concentration was determined by staining with a Micro-BCA kit (Pierce).
Synthesis of RNAs
The sequence of -globin cDNA (a gift of Dr J. Ilan) containing a coding part, the sequence of 3'-UTR and 50 nt poly(dA) was inserted into NcoI and BamHI sites of the pET-28a vector (Novagen), resulting in the pET-28a–-globin plasmid. The plasmid was linearized with BamHI and 660 nt -globin mRNA was transcribed by T7 polymerase as described previously (41). To obtain a hybrid 2Luc mRNA, a DNA fragment (a gift of Dr I. Shatsky), containing two full-length cDNAs of luciferases from Renilla reinformis and Photinus pyralis, separated by a polylinker, was cloned into the EcoRV and XbaI sites of the pSP72 vector (Promega). The prepared plasmid pSP72-2Luc was linearized by XhoI and 2Luc mRNA, including both the luciferase sequences, was synthesized by T7 polymerase. 2Luc mRNA was about 3000 nt long. The full-size TMV RNA of about 6000 nt long, isolated from viral particles by phenol extraction, was a gift of Dr E. Karger.
To prepare uniformly labeled mRNAs, the concentration of UTP was decreased and UTP was added to the transcription mixture (>2000 Ci/mmol; Radioisotop, Russia). Homogeneity of all RNAs used was verified by gel electrophoresis in denaturing conditions (data not shown).
Sucrose gradient analysis
10 μg (280 pmol) YB-1 was mixed with different amounts of -globin mRNA, which is a mixture of a constant amount of 32P-labeled mRNA (100 000 c.p.m.) and a variable amount of unlabeled mRNA, in 25 μl of buffer A (10 mM HEPES–KOH, pH 7.6, 100 mM KCl) containing 0.5 units/μl RNasin. The samples were incubated for 15 min at 30°C and layered onto 5–20% (w/v) linear sucrose gradients in buffer A. Centrifugation was carried out at 45 000 r.p.m. for 4 h at 4°C in an SW 60 rotor (Beckman, USA). Gradients were monitored for absorbance at 254 nm during collection of the fractions from the bottom. Radioactivity in the fractions was estimated by Cherenkov counting in a LS-100 radioactivity counter (Beckman, USA). Then in the case of -globin mRNA, the proteins in the fractions were precipitated with 10% TCA and subjected to SDS–PAGE. 28S and 18S rRNAs, as well as 9S globin mRNA, were used as sedimentation coefficient markers, the latter obtained from rabbit reticulocyte polysomes as described earlier (42).
CsCl density gradient centrifugation
mRNA complexes were formed in the following way: 0.2 μg (1 pmol) 32P-labeled -globin mRNA (100 000 c.p.m.) were mixed with increasing amounts of YB-1 in 15 μl of buffer A. The samples were incubated for 15 min at 30°C. Then they were fixed with 4% freshly neutralized formaldehyde for 1 h at 37°C. For the gradient to be formed, a ‘light’ ( = 1.232 g/cm3) and a ‘heavy’ ( = 1.718 g/cm3) solutions of CsCl in the buffer containing 10 mM HEPES–KOH, pH 7.6, 1 mM MgCl2, 4% formaldehyde were prepared. Then 2.2 ml of the ‘heavy’ CsCl solution was poured in a 5 ml centrifuge tube and on top of it 2.8 ml of the ‘light’ CsCl solution preliminarily mixed with the mRNP sample was placed. The gradients were centrifuged at 36 000 r.p.m. for 24 h at 4°C in an SW55-Ti rotor of a Beckman L5 centrifuge and fractionated from the bottom. The CsCl densities in selected fractions were calculated from their refraction indexes and the radioactivity was estimated by Cherenkov counting. To determine the protein distribution, the fractions were filtered through a nitrocellulose membrane (Whatman, England) and the protein was detected using anti-YB-1 antibodies at a 1:500 dilution as described in (25).
Chemical cross-linking
To obtain mRNP complexes with different molar ratios of YB-1/mRNA, 3 μg (85 pmol) YB-1 was mixed with premixes of a constant amount of -globin 32P-labeled mRNA (200 000 c.p.m.) with a varying amount of unlabeled mRNA in 20 μl of buffer A. The samples were incubated for 10 min at 30°C. Then glutaraldehyde (Merck) up to 0.15% was added to the reaction mixture and the incubation was continued for 15 min at 37°C. The reactions were stopped with 70 mM glycine, pH 8.0. mRNA was destroyed by incubation with 20 U RNase T1 (USB) and 5 μg RNase A (USB) and by subsequent boiling with 20 mM MgCl2 for 20 min. The resulting 32P-labeled proteins were resolved by 4–16% SDS–PAGE. The gel was stained with Coomassie brilliant blue R-250 (Pierce), dried and exposed to Kodak X-omat AR film at –70°C overnight.
UV cross-linking and cyanogen bromide cleavage
32P-labeled -globin mRNA (15 pmol, 1 500 000 c.p.m.) was incubated with increasing amounts of YB-1 in buffer A in the final volume of 25 μl for 15 min at 30°C. The samples were layered onto Parafilm M (American Can Company, USA), placed on ice and exposed to UV light (254 nm) in a transilluminator-cross-linker (Vilber-Lourmant, France). The distance from the source to the cross-linked sample was about 3 cm and the radiation dose was 0.75 J/cm2. With such a dose, individual YB-1 monomers were not yet noticeably cross-linked to themselves, but efficiently cross-linked to mRNA (data not shown). Then 20 U RNase T1 and 10 μg RNase A were added to the samples and the mixture was incubated for 30 min at 37°C. After treatment with ribonucleases, the samples were incubated with 1.5 mg BrCN in 80 μl of 70% formic acid at room temperature in the dark overnight. The samples were 4-fold diluted with H2O and lyophilized. The resulting 32P-labeled protein fragments were resolved in 15% SDS–PAGE. The gel was stained with Coomassie brilliant blue, dried and exposed to the X-ray film as mentioned above. Gel slices containing the labeled YB-1 fragments were excised after autoradiography and the radioactivity of the slices was determined by Cherenkov counting.
Atomic force microscopy and electron microscopy
To prepare YB-1–RNA complexes, 5 μg -globin mRNA, 2Luc mRNA or TMV RNA were mixed with 2.5 μg YB-1 (for non-saturated complexes) or with 40 μg YB-1 (for saturated complexes). The complexes were incubated in buffer A for 20 min at 30°C, and fixed with 0.15% glutaraldehyde as described above. To separate YB-1–RNA complexes from the protein unbound to mRNA, the preparations were centrifuged in 5–20% sucrose gradient for 3 h (-globin mRNA), 1.5 h (2Luc mRNA) and 1 h (TMV mRNA). The obtained RNPs were collected by centrifugation in a TLA 100.1 rotor at 90 000 r.p.m. for 2.5 h, dissolved in buffer A and used for atomic force microscopy (AFM) and electron microscopy (EM) analysis.
Free multimeric YB-1 (2 μg) was fixed with 0.15% glutaraldehyde in buffer A for 15 min at 37°C as described above.
AFM imaging experiments were performed in tapping mode using a Nanoscope IIIa multimode scanning probe microscope (Digital Instruments, USA). Standard silicon 125 μm cantilevers (NanoProbe) with 300–350 kHz resonant frequencies and 100 μm non-contact cantilevers with 200–300 kHz resonant frequencies (State Research Institute for Problems in Physics, Russia) were used for imaging in air. Experiments in liquid employed a tapping mode liquid cell (Digital Instruments, USA) and silicon nitride cantilevers (Nanoprobe) with a spring constant of 0.6 N/m. The cantilever oscillation frequency was 8–10 kHz. For image processing and presentation, software Femtoscan 001 (43) was applied. Samples were prepared by a direct adsorption technique, using aminopropylsilatran-treated (44) mica (APS-mica) plates as substrates. The preparation was carried out as previously described in (45). An aliquot of 10 μl of the mRNPs (1–10 ng/μl taking the concentration of RNA) or fixed YB-1 (10 ng/μl) were applied onto the substrate and left for 5 min for adsorption. Samples designated for imaging in the liquid cell were mounted directly onto the scanner, whereas those prepared for experiments in air were rinsed with tridistilled water and dried in airflow.
For EM studies, the samples were prepared by shadow casting and by negative staining. The mRNPs (50 ng/μl of RNA) or YB-1 (50 ng/μl) were adsorbed onto a thin carbon film and shadowed with platinum at an angle of tan(1/10) with an electron gun (46). Negative staining with 2% aqueous uranyl acetate was performed using the single-layer carbon technique (47). The samples were prepared also by the method of a preshadowed carbon replica (46). In this case, the samples were prepared as those for AFM. The negatively stained or metal-shadowed samples were examined in a JEM-100C electron microscope at 80 kV.
RESULTS
YB-1 associates with -globin mRNA to form complexes with sedimentation coefficient and buoyant density in CsCl typical for natural mRNPs
To estimate sedimentation coefficients of YB-1–mRNA complexes, YB-1 was mixed with 32P-labeled -globin mRNA at YB-1/mRNA molar ratios from 3:1 to 60:1 (weight ratios from 0.5 to 10) and analyzed using centrifugation in sucrose concentration gradient. Figure 1A shows the results of sedimentation distribution of complexes formed at YB-1/mRNA molar ratios from 12:1 to 60:1. An addition of increasing amounts of YB-1 to mRNA caused a gradual increase of the sedimentation coefficient of the formed complexes. The maximal saturation of -globin mRNA with YB-1 was observed at the 36:1 molar ratio. The formed saturated complex displayed sedimentation homogeneity, and its sedimentation coefficient was about 28S, i.e. it was three times as high as the sedimentation coefficient of the used RNA, which is typical of natural free mRNPs (informosomes) (7). Up to a YB-1/mRNA molar ratio of 12:1, the entire protein was mRNA-bound (data not shown), but at higher ratios, some free YB-1 appeared, though the sedimentation coefficient of the complex remained below its maximum. The appearance of unbound protein before mRNA saturation could be explained by an insufficiently high constant of its binding to mRNA. The minor component sedimenting faster than the main mRNP complex might be a complex of YB-1 with a stable globin mRNA conformer. Indeed, the used globin mRNA contained a minor component with a higher sedimentation coefficient and a lower electrophoretic mobility at native gel electrophoresis; this component disappeared upon gel electrophoresis under denaturing conditions (7 M urea) (data not shown).
Figure 1. Sedimentation and density analysis of YB-1 complexes with -globin mRNA. (A) Sedimentation distribution of YB-1 complexes with -globin mRNA in sucrose gradient. YB-1 was mixed with 32P-labeled -globin mRNA at indicated molar ratios. The formed complexes were layered on 5–20% sucrose gradient and centrifuged in a SW-60 rotor at 45 000 r.p.m. for 4 h at 4°C. UV absorbance at 254 nm (solid line) and radioactivity (open circles) are shown. YB-1 distribution over gradient fractions revealed by SDS–gel electrophoresis is shown below each figure. (B) Distribution of YB-1 complexes with -globin mRNA in CsCl density gradient. YB-1 complexes with 32P-labeled -globin mRNA formed at indicated molar ratios were fixed with 4% formaldehyde. Then the samples were centrifuged in CsCl gradient in a SW-55Ti rotor at 36 000 r.p.m. for 24 h at 4°C. Radioactivity (open circles) and CsCl density (filled triangles) are shown. YB-1 distribution over gradient fractions revealed by immunoblotting with antibodies to YB-1 is shown below each figure.
To determine the buoyant density of YB-1–mRNA complexes, YB-1 was in vitro incubated with 32P-labeled -globin mRNA, fixed with formaldehyde and centrifuged in the CsCl density gradient to equilibrium (Figure 1B). Within the YB-1/mRNA range from 12:1 to 48:1, there was one homogeneous peak whose buoyant density decreased with the increasing YB-1/mRNA ratio and reached 1.39 g/cm3 upon mRNA saturation at the molar ratio of 36:1. Excess of mRNA-unbound YB-1 was detected in fractions with a characteristic protein buoyant density of 1.35 g/cm3 on the right shoulder of the peak of the mRNP complex. The buoyant density of YB-1-saturated complexes was just the same as that of natural free mRNPs (7).
mRNA-bound YB-1 is monomeric at a low YB-1/mRNA ratio and multimeric when this ratio is high
Free YB-1 is known to be homomultimeric (22,38). A question arises as to whether it binds to mRNA as a multimer or as a monomer that subsequently restores its multimeric form upon mRNA saturation, as suggested earlier (39). The experiments described above on centrifugation both in sucrose gradient and in CsCl density gradient showed a unimodal sedimentation and density distribution of the complexes formed at various YB-1/mRNA ratios and a gradual shift of the sedimentation coefficient and buoyant density with changes in this ratio. Such a behavior of the complexes is in agreement with the suggestion that YB-1 binds to mRNA as a monomer, although the relatively low resolving capacity of these techniques does not allow ruling out the possibility that YB-1 associates with mRNA in small cooperative groups.
To have a more precise knowledge about quaternary structure of mRNA-associated YB-1 at various YB-1/mRNA ratios, free YB-1 or YB-1 bound to 32P-labeled -globin mRNA was fixed with 0.15% glutaraldehyde. The fixed preparations were subjected to exhaustive RNase treatment and analyzed by SDS–PAGE (Figure 2A). As seen, the initial free YB-1 is a multimer only slightly displaced from the separating gel start after electrophoresis under denaturing conditions. Its relative molecular mass far exceeds the highest marker (205 kDa), which is consistent with the sedimentation and gel filtration data on its large size (up to 800 kDa) (22,40). The protein obtained from the complexes with mRNA with a low YB-1/mRNA molar ratio (1.5:1) moved as a monomer during SDS–gel electrophoresis. An increase of the YB-1/mRNA ratio to 3:1 and more gave rise to cross-linked oligomers of gradually increasing size. Finally, at YB-1/mRNA exceeding 24:1, large multimers appeared close in size to a free fixed protein. This supports our previous suggestion that YB-1 changes its quaternary structure on mRNA depending on the YB-1/mRNA ratio.
Figure 2. (A) Alteration in the multimeric state of YB-1 within its complexes with -globin mRNA formed at different ratios of YB-1/mRNA. Complexes of YB-1 with 32P-labeled -globin mRNA were formed at indicated YB-1/mRNA ratios and fixed with 0.15% glutaraldehyde. RNA was destroyed by RNases A and T1 and by boiling with Mg2+ ions. The protein was analyzed using SDS–gel electrophoresis. Left, YB-1 without mRNA: lane 1, without fixation; lane 2, after fixation with glutaraldehyde, Coomassie stained. Middle and right, YB-1 from YB-1/mRNA complexes formed at indicated ratios of the components: lane 1, without addition of RNA; lanes 1–3, without fixation; lanes 4–9, after fixation with glutaraldehyde. Middle, Coomassie stained; right, radioautograph. (B) Analysis of domain YB-1 contacts with -globin mRNA within mRNP complexes. The scheme presents the domain organization of YB-1 and distribution of methionine residues. Complexes of YB-1 and 32P-labeled -globin mRNA were formed at indicated molar ratios, exposed to UV and treated with RNases A and T1. YB-1 was cleaved with cyanogen bromide at methionine residues. Left, gel autoradiograph after SDS–gel electrophoresis of YB-1 fragments resulting from cleavage is shown. Right, the graph demonstrates changes in the relative amount of radioactive RNA cross-linked with fragment I and fragment II of YB-1.
As mentioned above, YB-1 has two RNA-binding domains: the N-terminal CSD with consensus motifs RNP-1 and RNP-2 and the CTD with alternating clusters of positively and negatively charged amino acid residues (Figure 2B, upper scheme). We studied contacts of these domains with mRNA in YB-1–mRNA complexes with various YB-1/mRNA ratios. For this purpose, YB-1 was UV-cross-linked with 32P-labeled -globin mRNA. After exhaustive RNase digestion of mRNA, YB-1 labeled with radioactive RNA fragments was cleaved by BrCN at methionine residues concentrated in the first half of the CTD. The cleavage yielded two fragments: the larger N-fragment (designated as fragment I) contained the N-terminal A/P domain, the CSD and the beginning of the CTD, while the smaller one (fragment II) comprised the second half of CTD. The fragments were separated by SDS–gel electrophoresis, and radioactivity of each was determined (Figure 2B). As seen, the ratio between the RNA amount cross-linked with fragment I and the RNA amount cross-linked with fragment II considerably increased with increasing the YB-1/mRNA ratio, in particular in the range from 12:1 to 48:1. It is in this range that one can observe transition from active mRNA translation to complete protein synthesis inhibition both in a cell-free translation system (24,25,28) and in vivo in cultivated cells (27). This allows conclusion that at a comparatively low YB-1/mRNA ratio, typical of polysomal mRNPs, YB-1 binds to mRNA through its both RNA-binding domains, whereas high YB-1/mRNA ratios (24:1 and more) make it predominantly use its N-terminus with CSD. The CTDs are probably displaced from RNA by CSDs.
YB-1–mRNA complexes are dynamic, and YB-1 easily migrates from one mRNA molecule to another
In cell lysates, natural mRNPs are known to be stable enough to prevent YB-1 redistribution after addition of exogenous RNA (48). The informosome-like particles formed by exogenous RNA in cell lysates at physiological temperature proved to be as stable as natural mRNPs (49). However, in the cell, mRNPs are quite dynamic, and the set of mRNA-associated proteins changes upon mRNA transition from the nucleus to cytoplasm and from free mRNPs to polysomes (15,50,51). Moreover, it was shown that in translatable mRNPs in vivo proteins can leave mRNA and come back to it again after passing the ribosome (52).
In this study, we examined whether YB-1 can migrate from its complexes with mRNA to other mRNA molecules, i.e. whether its in vitro complex with mRNA is stable or dynamic. For this purpose, YB-1 was incubated with 32P-labeled -globin mRNA (molar ratio 36:1), then unlabeled -globin mRNA was added to the ratio of 9:1 and incubation was prolonged for additional time. Samples were analyzed in sucrose gradient (Figure 3A and B). If YB-1 could form stable complexes with mRNA and could not migrate to other RNA molecules, upon addition of unlabeled mRNA, the entire radioactivity would remain in the initial peak together with the major part of YB-1. But, as revealed by comparison of Figure 3A and B, the entire radioactive mRNA has shifted to a lighter complex and sedimented together with unlabeled added mRNA and almost all YB-1. So, YB-1 has been uniformly distributed over the initial and newly added mRNA. So, YB-1 can easily redistribute between mRNA molecules and consequently its complexes with RNA are dynamic. May be, natural mRNPs, as well as informosome-like particles formed in cell lysates with exogenous RNA at physiological temperature, are stabilized due to additional (minor) proteins and/or possible covalent modifications of YB-1.
Figure 3. Analysis of stability of YB-1 complexes with -globin mRNA. Complexes of YB-1 with 32P-labeled -globin mRNA were formed at YB-1/mRNA molar ratios of 36:1. The sample was incubated for 15 min at 30°C, then unlabeled -globin mRNA was added to a half of the sample to the final ratio of 9:1, and the preparations were incubated for additional 15 min at the same temperature. The preparations were layered onto 5–20% sucrose gradient and centrifuged in a SW-60 rotor at 45 000 r.p.m. for 4 h at 4°C. (A) Distribution of the complex formed by YB-1 with -globin mRNA at a molar ratio of 36:1. (B) Distribution of the complex formed by YB-1 with -globin mRNA at a molar ratio of 36:1 to which unlabeled -globin mRNA was added to the YB-1/mRNA ratio of 9:1. UV absorbance (solid line) at 254 nm and radioactivity (open circles) profiles are shown. YB-1 distribution over gradient fractions revealed by SDS–gel electrophoresis is shown below the plots.
AFM and EM studies of YB-1 and YB-1–RNA complexes
As it was shown in biochemical experiments, free YB-1 forms large multimeric complexes up to 800 kDa (22), see also Figure 2. Here, we visualized free YB-1 by EM and AFM (Figure 4). The multimeric protein was pre-fixed with glutaraldehyde. Independent of the preparation techniques (negative staining and shadowing) and the type of substrate (carbon film and mica), EM reveals the same shape and dimensions of the YB-1 multimer. This protein looks like quite homogeneous granules 9–10 nm high and 35–40 nm in diameter (Figure 4A and B). AFM images revealed particles with 8–10 nm height and 30–40 nm diameter, measured at half of the particle height (Figure 4C), which is in good agreement with EM data. The observed di- and trimers are probably the result of granule sticking to one another. It can be concluded that multimeric YB-1 is a flattened particle whose diameter is 3–4 times its height. It is not excluded that in solution YB-1 is a spherical particle that is flattened upon interaction with the surface. To obtain a YB-1 image in the form of a monomer, the protein was dissociated with 1 M LiCl. Presumably, high salt concentrations do not allow the monomerized YB-1 in such a way to give distinct AFM and EM images upon drying on the substrate (data not shown). In order to overcome this problem, monomerized YB-1 was imaged by AFM in solution in a high salt buffer (Figure 4D). Under these conditions, the protein molecules were detected as particles with an average height of about 4 nm.
Figure 4. EM and AFM images of free multimeric and monomeric YB-1. (A) Electron micrograph of multimeric YB-1. The protein, fixed with 0.15% glutaraldehyde, was adsorbed onto a carbon film and shadowed with platinum-carbon. The same images of shadowed fixed YB-1 were obtained at its adsorbtion onto mica (not shown). (B) Electron micrograph of negatively stained fixed YB-1 deposited on mica. (C) AFM image of fixed YB-1 on APS-mica imaged in air. An aliquot of 5 μl of fixed YB-1 (0.01 μg/μl) was deposited onto the substrate, rinsed with water, dried and imaged. (D) AFM image of YB-1 in high salt solution on APS-mica substrate. YB-1 (0.05 μg/μl) was incubated in a high salt buffer (10 mM HEPES–KOH pH 7.6, 1 M LiCl) for 20 min at 30°C. Cross-sections made along the marked lines in (C) and (D) are shown under the respective AFM images. These cross-sections represent the height of the complexes and its variations. Triangles in the fields indicate the ends of the section lines.
As shown above in biochemical experiments, in YB-1–mRNA complexes, YB-1 associates with RNA as a monomer when the YB-1/mRNA ratio is relatively low. As the YB-1 content in the complex is increased, YB-1 forms a multimer with a molecular mass comparable to that of a free multimeric protein. To visualize YB-1 in complex with RNA, we studied such complexes formed with RNAs of various lengths. Apart from -globin mRNA (660 nt), we used 2Luc mRNA (3000 nt). The complexes were pre-fixed with glutaraldehyde. First, unsaturated complexes were studied by means of AFM (Figure 5). In these complexes, the YB-1/RNA mass ratio was 0.5:1, which corresponds to a molar ratio of 3:1 for globin mRNA and to 14:1 for 2Luc mRNA. Since mRNA has a highly developed secondary structure and monomeric YB-1 is small in size, it was impossible to distinguish separate protein monomers within the complex. Nevertheless, it is clearly seen (especially for the more extended 2Luc mRNA) that the complexes are quite unfolded and that their heights are below 2 nm, which is indicative, most probably, that YB-1 is monomeric in these complexes. The average complex height is 1.7 nm for globin mRNA and 1.9 nm for 2Luc mRNA. The fact that these values are lower than the height of free monomerized YB-1 measured on APS-mica in solution (4 nm) can be explained by both the YB-1 interaction with mRNA and by the effect of YB-1 drying on APS-mica. The latter effect, previously reported for several other proteins (53,54), is more probable, since in our AFM experiments monomerized YB-1 formed 1.5–2.0 nm thick monolayers upon drying on APS-mica surface (data not shown). Thus, AFM images of unsaturated mRNPs show no large multimeric YB-1 on mRNAs, thereby proving its dissociation upon binding to RNA, which complies with the biochemical data on monomerization of YB-1 in complexes with a low YB1/mRNA ratio (see Figure 2A).
Figure 5. AFM images of unsaturated YB-1–mRNP complexes in air. Unsaturated complexes of YB-1 with -globin mRNA (A) and 2Luc mRNA (B) were fixed with glutaraldehyde, deposited on APS-mica and imaged as described in Materials and Methods. Left, typical field; right, magnified images of individual complexes and cross-sections made along the marked lines on the fields. Triangles in the fields indicate the ends of the section lines.
In the next series of experiments, we studied saturated complexes of YB-1 with the same mRNAs. These complexes were formed at the YB-1/RNA mass ratio 8:1, which corresponds to a molar ratio of 48:1 for globin mRNA and to 224:1 for 2Luc mRNA. After fixation, the complexes were separated from the RNA-unbound protein as described in Materials and Methods. As compared to unsaturated complexes, the saturated ones appeared to be much more compact (Figure 6). The YB-1–globin mRNA complexes were seen as 6.2 ± 0.9 nm high globules, i.e. they were three times as high as the largest unsaturated complexes (Figure 6A). Saturated complexes look either like single globules or represent dimers, resulting, probably, from the association of the two globules. The distribution of the number of globules per complex shows that the major part of the complexes contains only one globule (Figure 6A, right). The dimensions of these globules imply that each contains several YB-1 molecules and support the conclusion on YB-1 multimerization upon mRNA saturation.
Figure 6. AFM images of saturated mRNP complexes. Fixed saturated complexes of YB-1 with -globin mRNA (A), 2Luc mRNA (B) and TMV RNA (C) were deposited on APS-mica and imaged in air. Left, typical field; right, cross-sections made along the marked lines on the fields. Triangles in the fields indicate the ends of the section lines. Histograms illustrating the distribution of number of multimeric globules per complex are shown on the right.
AFM experiments with 2Luc mRNA that is 4.5 times longer than globin mRNA revealed formation of mRNPs composed of several successive globular particles 6.8 ± 0.9 nm high (Figure 6B). These particles' dimensions are very close to those of ones formed on globin mRNA, which allows considering them to be YB-1 multimers, as well. Unlike globin mRNA, most of 2Luc mRNA complexes consisted of four successive globules, the center-to-center distance of adjacent globules being 20–25 nm and the average mRNP length being 130 ± 20 nm (Figure 6B, right). Figure 7A presenting their magnified images shows the structure of these complexes in more detail. EM images (Figure 7B) allow measuring the real diameter of separate globules of the complex that is about 20 nm. Thus, like free multimeric YB-1, separate globules of RNA-associated YB-1 are flattened and have a diameter several times exceeding their height. However, the size of globules of mRNA-bound protein differs from that of free multimeric protein (specifically, in height). This clearly shows that the formation of saturated complexes is a two-step event: first, the initially multimeric YB-1 dissociates to monomers upon its binding to mRNA; and second, the YB-1 multimers are formed again as RNA is saturated with the protein.
Figure 7. Images of saturated complexes formed with 2Luc mRNA at high magnification. (A) AFM images of individual complexes in air. Right, cross-sections made along the marked lines on the corresponding images. (B) Electron micrograph of the negatively stained complexes.
Since globin mRNA (660 nt) allows formation mainly of one multimeric globule and 2Luc mRNA (3000 nt) forms four of them, it may be suggested that one globule is formed at a 600–700 nt fragment of RNA. In order to verify this suggestion, we studied saturated complexes with still more extended TMV RNA (6000 nt). Like with other RNAs used, YB-1 forms multimers with the same globule height of 6.5 ± 0.8 nm on TMV RNA (Figure 6C). Most of these complexes contain eight multimeric globules (see the diagram in Figure 6C), thus confirming our suggestion. The fact that a considerable number of YB-1–TMV RNA complexes contain less than eight globules can be explained by a higher probability of their random degradation in the course of sample storage and preparation. The measured average length of the saturated YB-1–TMV RNA complexes was 270 ± 50 nm. According to EM data, the average specific length of single-stranded RNA is 260 nm/kb (55) and, therefore, the TMV RNA length is 1560 nm. According to AFM, the length of a fully unfolded denatured TMV RNA is 1500 nm (56). Hence, saturation of TMV RNA with YB-1 is accompanied by its 4- to 5-fold compactization.
Thus, within the formed saturated complexes mRNA is noticeably compact, and the general structure of these complexes is of a ‘beads-on-a-string’ type, where one ‘bead’ is formed on a 600–700 nt long segment of the RNA chain.
DISCUSSION
In this study, we analyzed complexes formed by YB-1, the major protein of cytoplasmic mRNPs, with mRNA at various molar ratios, including those typical for translatable (polysomal) and untranslatable (free) mRNPs. It has been found that saturation of globin mRNA with YB-1 gives rise to the formation of particles with basic physico-chemical parameters of natural free mRNPs, namely, a sedimentation coefficient three times as high as that of the used -globin mRNA and a buoyant density in CsCl equal to 1.39 g/cm3. This value exactly coincides with that of natural free mRNPs and corresponds to the 3:1 YB-1/mRNA weight ratio. Besides, earlier we showed that YB-1-bound mRNA is extremely sensitive to endoribonucleases, which is also characteristic of natural mRNPs and shows surface location of mRNA in the particles (22).
YB-1 and its homologs are universal major mRNP proteins of not only mammalian cells but of cells from other organisms too. It was shown recently that two isoforms of the homologous proteins Ct-p40 and Ct-p50 are associated with mRNA from the insect Chironomus tentans (57). Two proteins of this family were detected in amphibia Xenopus laevis (34). One of these proteins, FRGY1, packs mRNA of mature somatic cells and shows high homology to mammalian YB-1. The other, FRGY2, having two isoforms p54/p56, packs masked mRNA in oocytes. FRGY2 is far less homologous to YB-1 beyond CSD, although its isoelectric point, amino acid composition and charged amino acid residue distribution in CTD coincide with those of YB-1. As shown earlier, FRGY2 also forms with mRNA, at its saturation, RNA–protein complexes with a buoyant density of 1.4 g/cm3 (58,59). This suggests that in various multicellular organisms from insects to man, YB-1 and its homologs are the major structure-forming mRNP proteins. Hence, the complexes between rabbit YB-1 and mRNA described here can serve as a simplest model system for studying the general principles of structural organization of natural mRNPs in all kinds of eukaryotic cells.
Free YB-1 and its homologs are multimeric (22,37,38). In this connection, a question arises as to whether YB-1 remains multimeric when binding to mRNA or, as we suggested earlier, it dissociates in monomers and afterwards restores its initial multimeric form when mRNA has been protein-saturated. Our experiments on rabbit YB-1 showed that in forming a YB-1–mRNA complex with a low YB-1/mRNA ratio, YB-1 binds to mRNA predominantly as a monomer, which points to dissociation of the YB-1 multimer upon interaction with mRNA. According to AFM, such complexes are relatively unfolded.
As the YB-1 amount on mRNA is increased, homo-oligomeric YB-1 complexes are formed and grow to their maximal size; so, mRNA with a higher molecular weight has several protein globules close in size, and the more extended mRNA, the greater the number of globules. The binding of YB-1 to mRNA was found to be a non-cooperative process. The known protein/RNA weight ratio (3:1) and the number of protein globules on mRNA molecules of various lengths allow calculating the molecular mass of a protein globule, that was about 700 kDa and hence, was formed by about twenty YB-1 molecules. In saturated complexes, one YB-1 molecule occupies on average 35 nt of mRNA, and the surface of one protein globule can house a 700 nt mRNA segment, the RNA compaction coefficient being about 5.
Recent EM studies of saturated complexes of FRGY2 with TFIIIA mRNA (1.5 kb) and histone H1 mRNA (0.75 kb) (59) showed them as tight thickened structures some of which contained large particles with a diameter of 10–20 nm. Although the presented images hardly allow estimating the number of particles per individual RNA, the general view of the saturated complexes is in agreement with our AFM and EM images. The diameter of some RNA-housed particles (10–20 nm), reported by the authors in (59), is slightly less than that of YB-1 particles (20–25 nm) in our studies. This discrepancy can be explained by an essential difference between FRGY2 and YB-1 that are only 64% homologous. Nevertheless, it can be assumed that the formation of large successive mRNA-associated particles, whose number is proportional to the RNA length, is the general mode of mRNA packing by YB-1 homologs in the eukaryotic cytoplasm.
This model of structural organization of YB-1/mRNA complexes resembles the model for organization of hnRNPs proposed previously (8,60). Both the models imply the formation of large (1000 kDa) multimeric protein complexes that pack RNA on their surface in the ‘beads-on-a-string’ mode, the only difference being that in the case of cytoplasmic mRNPs, close in size globular protein structures are formed by YB-1 alone, rather than by six various proteins in a certain proportion in the case of hnRNPs.
It follows from our experiments that the mode of interaction between YB-1 and mRNA depends on the YB-1/mRNA ratio: at a low value of this ratio, monomeric YB-1 binds to mRNA through its two RNA-binding domains, CSD and CTD. As the YB-1 amount on mRNA is increased, the CTD binding decreases, and the association with mRNA is realized mostly through the YB-1 N-terminus with CSD. This can be explained by a successful CSD competition for mRNA that leads to CTD displacement. The displaced CTDs can interact with one another, thereby restoring the multimeric form of YB-1 that it had in the free state. It is reasonable to suggest that the CTDs are packed inside the formed protein globules, whereas the CSDs are located on their surface (Figure 8). In saturated complexes, the surface-attached mRNA can contact only the CSDs.
Figure 8. Scheme showing involvement of YB-1 in the formation of polysomal and free cytoplasmic mRNPs. For details, see text.
Surface location of mRNA is essential for its recognition by different regulatory proteins, which is necessary for its selective activation, selective degradation or specific localization in the cell. As shown previously, such location makes mRNA very sensitive to endonucleases (22). But its life span, both in the cell and in cell lysates, is strongly increased by YB-1 (35). This paradox can be easily explained by the fact that degradation of mRNA in the cell usually starts from its ends and is realized by exonucleases, and in saturated complexes, the mRNA ends are probably buried and inaccessible to exonucleases. This should be verified experimentally. The inaccessibility of mRNA ends to proteins hinders their interaction with translation initiation factors and leads to protein synthesis inhibition. It was demonstrated in lysates and in systems reconstituted from pure components of the protein-synthesizing apparatus that at a high YB-1/mRNA ratio, YB-1 does prevent interaction of mRNA with translation initiation factors (28,35).
As mentioned above, the mRNA transition to a translatable state is accompanied both by the binding of PABP to poly(A) tail sequence of poly(A)+ mRNA and by an approximately 2-fold decrease in the YB-1 amount on mRNA. The mechanism of this transition has remained unclear yet. It can be postulated that a decrease in the amount of YB-1 on mRNA within polysomes would change the interaction of YB-1 with mRNA in the same way as in our model system. In such a case, YB-1 binds to mRNA not only through its CSD, but through CTD as well, which is accompanied by dissociation of multimetic YB-1 complexes into monomers and promotes mRNP unfolding and exposure of mRNA termini. This contributes to mRNA interaction with translation initiation factors and mRNA involvement in the process of translation (Figure 8).
It is notable that complexes of recombinant YB-1 with mRNA are reversible, and the protein migrates easily to the newly added mRNA. The situation within the intact lysate may be different. It is known that after incubation in physiological conditions, both natural mRNPs and those formed on the exogenous mRNA added to lysates are irreversible and do not lose their proteins during incubation even with a high excess of exogenous RNA (48,49). This suggests the existence of factors, which stabilize the mRNP structure. Such factors could be covalent modifications of YB-1 in an eukaryotic cell that are lacking in the protein expressed in E.coli. Minor masking mRNP proteins that may play the role of peculiar clips, as proposed earlier (1), can be in the capacity of such factors.
ACKNOWLEDGEMENTS
We thank Dr J. Ilan for providing -globin cDNA, Dr I. Shatsky for luciferase cDNA and Dr E. Karger for TMV RNA. We also thank T. Kuvshinkina and E. Serebrova for their help in preparing this manuscript. This work was supported in part by the Russian Academy of Sciences (Programmes on ‘Molecular and Cellular Biology’ and ‘Basic Sciences to Medicine’) and by a grant from the President of the Russian Federation (#1959.2003.4).
REFERENCES
Spirin,A.S. ( (1996) ) Masked and translatable messenger ribonucleoproteins in higher eukaryores. In Hershey,J.W.B., Mathews,M.B. and Sonenberg,N. (eds), Translational Control. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 319–334.
Samarina,O.P. ( (1996) ) hnRNP particles. Bioessays, , 18, , 595–601.
Dreyfuss,G., Kim,V.N. and Kataoka,N. ( (2002) ) Messenger-RNA-binding proteins and the messages they carry. Nature Rev. Mol. Cell. Biol., , 3, , 195–205.
Keene,J.D. ( (2001) ) Ribonucleoprotein infrastructure regulating the flow of genetic information between the genome and the proteome. Proc. Natl Acad. Sci. USA, , 98, , 7018–7024.
Kloc,M., Zearfoss,N.R. and Etkin,L.D. ( (2002) ) Mechanisms of subcellular mRNA localization. Cell, , 108, , 533–544.
Tourriere,H., Chebli,K. and Tazi,J. ( (2002) ) mRNA degradation machines in eukaryotic cells. Biochimie, , 84, , 821–837.
Preobrazhensky,A.A. and Spirin,A.S. ( (1978) ) Informosomes and their protein components: the present state of knowledge. Prog. Nucleic Acid Res. Mol. Biol., , 21, , 1–38.
Huang,M., Rech,J.E., Northington,S.J., Flicker,P.F., Mayeda,A., Krainer,A.R. and LeStourgeon,W.M. ( (1994) ) The C-protein tetramer binds 230 to 240 nucleotides of pre-mRNA and nucleates the assembly of 40S heterogeneous nuclear ribonucleoprotein particles. Mol. Cell. Biol., , 14, , 518–533.
McAfee,J.G., Huang,M., Soltaninassab,S.R., Rech,J.E., Iyengar,S. and LeStourgeon,W.M. ( (1997) ) The packaging of pre-mRNA. In Krainer,A.R. (ed.), Eukaryotic mRNA Processing. Oxford University Press, pp. 68–102.
Ovchinnikov,L.P., Skabkin,M.A., Ruzanov,P.V. and Evdokimova,V.M. ( (2001) ) Major mRNP proteins in the structural organization and function of mRNA in eukaryotic cells. Mol. Biol. (Mosk.), , 35, , 548–558.
Morel,C., Kayibanda,B. and Scherrer,K. ( (1971) ) Proteins associated with globin messenger RNA in avian erythroblasts: isolation and comparison with the proteins bound to nuclear messenger-like RNA. FEBS Lett., , 18, , 84–88.
Blobel,G. ( (1972) ) Protein tightly bound to globin mRNA. Biochem. Biophys. Res. Commun., , 47, , 88–95.
Kumar,A. and Pederson,T. ( (1975) ) Comparison of proteins bound to heterogeneous nuclear RNA and messenger RNA in HeLa cells. J. Mol. Biol., , 96, , 353–365.
Jain,S.K., Pluskal,M.G. and Sarkar,S. ( (1979) ) Thermal chromatography of eukaryotic messenger ribonucleoprotein particles on oligo (dT)-cellulose. Evidence for common mRNA-associated proteins in various cell types. FEBS Lett., , 97, , 84–90.
Dreyfuss,G. ( (1986) ) Structure and function of nuclear and cytoplasmic ribonucleoprotein particles. Annu. Rev. Cell Biol., , 2, , 459–498.
Blobel,G. ( (1973) ) A protein of molecular weight 78,000 bound to the polyadenylate region of eukaryotic messenger RNAs. Proc. Natl Acad. Sci. USA, , 70, , 924–928.
van Venrooij,W.J., van Eekelen,C.A., Jansen,R.T. and Princen,J.M. ( (1977) ) Specific poly-A-binding protein of 76,000 molecular weight in polyribosomes is not present on poly A of free cytoplasmic mRNP. Nature, , 270, , 189–191.
Minich,W.B., Maidebura,I.P. and Ovchinnikov,L.P. ( (1993) ) Purification and characterization of the major 50-kDa repressor protein from cytoplasmic mRNP of rabbit reticulocytes. Eur. J. Biochem., , 212, , 633–638.
Murray,M.T., Schiller,D.L. and Franke,W.W. ( (1992) ) Sequence analysis of cytoplasmic mRNA-binding proteins of Xenopus oocytes identifies a family of RNA-binding proteins. Proc. Natl Acad. Sci. USA, , 89, , 11–15.
Tafuri,S.R. and Wolffe,A.P. ( (1993) ) Dual roles for transcription and translation factors in the RNA storage particles of Xenopus oocytes. Trends Cell Biol., , 3, , 94–98.
Gu,W., Tekur,S., Reinbold,R., Eppig,J.J., Choi,Y.C., Zheng,J.Z., Murray,M.T. and Hecht,N.B. ( (1998) ) Mammalian male and female germ cells express a germ cell-specific Y-Box protein, MSY2. Biol. Reprod., , 59, , 1266–1274.
Evdokimova,V.M., Wei,C.L., Sitikov,A.S., Simonenko,P.N., Lazarev,O.A., Vasilenko,K.S., Ustinov,V.A., Hershey,J.W. and Ovchinnikov,L.P. ( (1995) ) The major protein of messenger ribonucleoprotein particles in somatic cells is a member of the Y-box binding transcription factor family. J. Biol. Chem., , 270, , 3186–3192.
Katsu,Y., Yamashita,M. and Nagahama,Y. ( (1997) ) Isolation and characterization of goldfish Y box protein, a germ-cell-specific RNA-binding protein. Eur. J. Biochem., , 249, , 854–861.
Minich,W.B. and Ovchinnikov,L.P. ( (1992) ) Role of cytoplasmic mRNP proteins in translation. Biochimie, , 74, , 477–483.
Evdokimova,V.M., Kovrigina,E.A., Nashchekin,D.V., Davydova,E.K., Hershey,J.W. and Ovchinnikov,L.P. ( (1998) ) The major core protein of messenger ribonucleoprotein particles (p50) promotes initiation of protein biosynthesis in vitro. J. Biol. Chem., , 273, , 3574–3581.
Pisarev,A.V., Skabkin,M.A., Thomas,A.A., Merrick,W.C., Ovchinnikov,L.P. and Shatsky,I.N. ( (2002) ) Positive and negative effects of the major mammalian messenger ribonucleoprotein p50 on binding of 40S ribosomal subunits to the initiation codon of beta-globin mRNA. J. Biol. Chem., , 277, , 15445–15451.
Davydova,E.K., Evdokimova,V.M., Ovchinnikov,L.P. and Hershey,J.W. ( (1997) ) Overexpression in COS cells of p50, the major core protein associated with mRNA, results in translation inhibition. Nucleic Acids Res., , 25, , 2911–2916.
Nekrasov,M.P., Ivshina,M.P., Chernov,K.G., Kovrigina,E.A., Evdokimova,V.M., Thomas,A.A., Hershey,J.W. and Ovchinnikov,L.P. ( (2003) ) The mRNA-binding protein YB-1 (p50) prevents association of the eukaryotic initiation factor eIF4G with mRNA and inhibits protein synthesis at the initiation stage. J. Biol. Chem., , 278, , 13936–13943.
Darnbrough,C.H. and Ford,P.J. ( (1981) ) Identification in Xenopus laevis of a class of oocyte-specific proteins bound to messenger RNA. Eur. J. Biochem., , 113, , 415–424.
Richter,J.D. and Smith,L.D. ( (1984) ) Reversible inhibition of translation by Xenopus oocyte-specific proteins. Nature, , 309, , 378–380.
Herbert,T.P. and Hecht,N.B. ( (1999) ) The mouse Y-box protein, MSY2, is associated with a kinase on non-polysomal mouse testicular mRNAs. Nucleic Acids Res., , 27, , 1747–1753.
Sommerville,J. ( (1990) ) RNA-binding phosphoproteins and the regulation of maternal mRNA in Xenopus. J. Reprod. Fertil. Suppl., , 42, , 225–233.
Braddock,M., Muckenthaler,M., White,M.R., Thorburn,A.M., Sommerville,J., Kingsman,A.J. and Kingsman,S.M. ( (1994) ) Intron-less RNA injected into the nucleus of Xenopus oocytes accesses a regulated translation control pathway. Nucleic Acids Res., , 22, , 5255–5264.
Wolffe,A.P., Tafuri,S., Ranjan,M. and Familari,M. ( (1992) ) The Y-box factors: a family of nucleic acid binding proteins conserved from Escherichia coli to man. New Biol., , 4, , 290–298.
Evdokimova,V., Ruzanov,P., Imataka,H., Raught,B., Svitkin,Y., Ovchinnikov,L.P. and Sonenberg,N. ( (2001) ) The major mRNA-associated protein YB-1 is a potent 5' cap-dependent mRNA stabilizer. EMBO J., , 20, , 5491–5502.
Matsumoto,K. and Wolffe,A.P. ( (1998) ) Gene regulation by Y-box proteins: coupling control of transcription and translation. Trends Cell Biol., , 8, , 318–323.
Tafuri,S.R. and Wolffe,A.P. ( (1992) ) DNA binding, multimerization, and transcription stimulation by the Xenopus Y box proteins in vitro. New Biol., , 4, , 349–359.
Gaudreault,I., Guay,D. and Lebel,M. ( (2004) ) YB-1 promotes strand separation in vitro of duplex DNA containing either mispaired bases or cisplatin modifications, exhibits endonucleolytic activities and binds several DNA repair proteins. Nucleic Acids Res., , 32, , 316–327.
Evdokimova,V.M. and Ovchinnikov,L.P. ( (1999) ) Translational regulation by Y-box transcription factor: involvement of the major mRNA-associated protein, p50. Int. J. Biochem. Cell Biol., , 31, , 139–149.
Ustinov,V.A., Skabkin,M.A., Nashchekin,D.V., Evdokimova,V.M. and Ovchinnikov,L.P. ( (1996) ) Major p50 protein of the somatic cell cytoplasmic mRNP: expression in Escherichia coli, isolation, and some properties of the recombinant protein. Biochemistry (Mosc.), , 61, , 559–566.
Pokrovskaya,I.D. and Gurevich,V.V. ( (1994) ) In vitro transcription: preparative RNA yields in analytical scale reactions. Anal. Biochem., , 220, , 420–423.
Aviv,H. and Leder,P. ( (1972) ) Purification of biologically active globin messenger RNA by chromatography on oligothymidylic acid-cellulose. Proc. Natl Acad. Sci. USA, , 69, , 1408–1412.
Filonov,A.S. and Yaminsky,I.V. ( (1997) ) SPM Control and Image Processing Software. Chichester: Advance Technologies Center, Moscow.
Shlyakhtenko,L.S., Potaman,V.N., Sinden,R.R. and Lyubchenko,Y.L. ( (1998) ) Structure and dynamics of supercoil-stabilized DNA cruciforms. J. Mol. Biol., , 280, , 61–72.
Kiselyova,O.I., Galyamov,M.O., Nasikan,N.S., Yaminsky,I.V., Karpova,O.V. and Novikov,V.K. ( (2002) ) Scanning probe microscopy of biomacromolecules: nucleic acids, proteins and their complexes. In Buzanaeva,E.V. and Scharff,P. (eds), Frontiers of Multifunctional Nanosystems. Kluwer Academic Publishers, Dordrecht, pp. 321–330.
Vasiliev,V.D. and Koteliansky,V.E. ( (1979) ) Freeze-drying and high-resolution shadowing in electron microscopy of Escherichia coli ribosomes. Meth. Enzymol., , 59, , 612–629.
Valentine,R.C., Shapiro,B.M. and Stadtman,E.R. ( (1968) ) Regulation of glutamine synthetase. XII. Electron microscopy of the enzyme from Escherichia coli. Biochemistry, , 7, , 2143–2152.
Ovchinnikov,L.P., Avanesov,A.C. and Spirin,A.S. ( (1969) ) Informosomes from loach embryos. Mol. Biol. (Mosk.), , 3, , 465–472.
Ovchinnikov,L.P., Voronina,A.S., Stepanov,A.S., Belitsina,N.V. and Spirin,A.S. ( (1968) ) Informosome-like complexes were formed by RNA adding to animal cell homogenates. Mol. Biol. (Mosk.), , 2, , 752–763.
Bag,J. ( (1984) ) Cytoplasmic mRNA–protein complexes of chicken muscle cells and their role in protein synthesis. Eur. J. Biochem., , 141, , 247–254.
Daneholt,B. ( (2001) ) Assembly and transport of a premessenger RNP particle. Proc. Natl Acad. Sci. USA, , 98, , 7012–7017.
Greenberg,J.R. ( (1981) ) The polyribosomal mRNA–protein complex is a dynamic structure. Proc. Natl Acad. Sci. USA, , 78, , 2923–2926.
Lyubchenko,Y.L., Oden,P.I., Lampner,D., Lindsay,S.M. and Dunker,K.A. ( (1993) ) Atomic force microscopy of DNA and bacteriophage in air, water and propanol: the role of adhesion forces. Nucleic Acids Res., , 21, , 1117–1123.
Lyubchenko,Y.L. and Shlyakhtenko,L.S. ( (1997) ) Visualization of supercoiled DNA with atomic force microscopy in situ. Proc. Natl Acad. Sci. USA, , 94, , 496–501.
Glass,J. and Wertz,G.W. ( (1980) ) Different base per unit length ratios exist in single-stranded RNA and single-stranded DNA. Nucleic Acids Res., , 8, , 5739–5751.
Drygin,Y.F., Bordunova,O.A., Gallyamov,M.O. and Yaminsky,I.V. ( (1998) ) Atomic force microscopy examination of tobacco mosaic virus and virion RNA. FEBS Lett., , 425, , 217–221.
Soop,T., Nashchekin,D., Zhao,J., Sun,X., Alzhanova-Ericsson,A.T., Bjorkroth,B., Ovchinnikov,L. and Daneholt,B. ( (2003) ) A p50-like Y-box protein with a putative translational role becomes associated with pre-mRNA concomitant with transcription. J. Cell Sci., , 116, , 1493–1503.
Kick,D., Barrett,P., Cummings,A. and Sommerville,J. ( (1987) ) Phosphorylation of a 60 kDa polypeptide from Xenopus oocytes blocks messenger RNA translation. Nucleic Acids Res., , 15, , 4099–4109.
Matsumoto,K., Tanaka,K.J., Aoki,K., Sameshima,M. and Tsujimoto,M. ( (2003) ) Visualization of the reconstituted FRGY2-mRNA complexes by electron microscopy. Biochem. Biophys. Res. Commun., , 306, , 53–58.
Samarina,O.P., Lukanidin,E.M., Molnar,J. and Georgiev,G.P. ( (1968) ) Structural organization of nuclear complexes containing DNA-like RNA. J. Mol. Biol., , 33, , 251–263.(Maxim A. Skabkin, Olga I. Kiselyova1, Ko)