| Nucleic Acids Research | Pages |
Oligonucleotide binding specificities of the hnRNP C protein tetramer
Introduction
Materials And Methods
C-protein purification and quantification
Fluorescence titrations
Data analysis
Competitor RNAs
Binding isotherms using competitor RNAs
Results
C protein displays no significant affinity for U-rich polypyrimidine tracts
The affinity of C protein for a SELEX-identified winner sequence is not solely due to contiguous uridines
Discussion
Acknowledgement
References
Oligonucleotide binding specificities of the hnRNP C protein tetramer
ABSTRACT
INTRODUCTION
In the presence of nuclease inhibitors the majority of pre-mRNA molecules released from nuclei by sonic disruption or by low salt extraction exist as oligomeric arrays of 20 nm ribonucleoprotein particles, termed heterogeneous nuclear ribonucleoprotein (hnRNP) complexes (for reviews see 1-3). In the presence of mild nuclease activity these complexes are converted to 20 nm monoparticles that sediment near 40S in density gradients and are termed 30-40S hnRNP core particles. The major proteins recovered from core particles (the A, B and C proteins) apparently exist as three different heterotetramers (4,5). The (A2)3B1 and (C1)3C2 tetramers have been isolated and partially characterized (4,6). Although protein A1 has been shown by crosslinking experiments to exist as homotrimers in monoparticles (7-9), in the absence of RNA stable (A1)3B2 tetramers have not been isolated. The core particle proteins are among the most abundant nuclear proteins (10,11) and they possess the intrinsic activity in vitro to reconstitute hnRNP particles every 700 nt along the length of long RNA substrates (6,12,13).
The nucleation of monoparticle assembly and the contiguous placement of hnRNP particles every 700 nt on long substrates appears to result from the unique properties of the C protein tetramer. For example, the C protein tetramer binds 230 nt of RNA in a highly cooperative salt-resistant manner and each contiguous grouping of three tetramers fold 700 nt lengths of RNA into a unique 19S triangular complex in vitro (6,14). The triangular C protein-RNA complex is a physiologically relevant structure as it exists as a remnant structure following dissociation of hnRNP particles in 300 mM salt. Additionally, purified triangular C protein-RNA complexes nucleate stoichiometric monoparticle assembly in vitro, while, in the absence of C protein, the basic A and B proteins bind RNA to form aggregate complexes that sediment in a heterodisperse manner. These findings indicate that C protein possesses three interrelated functions in pre-mRNA maturation. More specifically, it orders and constrains 230 nt lengths of the elongating transcript, it functions as a protein ruler (through the combinatorial folding of three tetramers) to define the length of RNA to be packaged in hnRNP particles and the 19S triangular complex functions to direct the stoichiometric assembly of monoparticles. The ATP-dependent displacement of C protein and other hnRNPs from splicing-competent pre-RNAs during spliceosome assembly (15) is generally consistant with a sequence-independent RNA chaperonin-like role for C protein in pre-mRNA maturation (3).
Other experimental findings, however, have been interpreted as evidence that C protein functions directly in pre-mRNA splicing through sequence-specific RNA binding events. For example, in oligonucleotide binding experiments it was observed that C protein does not dissociate from Sepharose-linked oligo(U) at 2 M NaCl (16). In experiments where defined RNAs were added to splicing-competent nuclear extracts and exposed to ribonuclease T1, U-rich polypyrimidine tracts were recovered from immunoprecipitates of C protein (17). Additionally, through UV irradiation experiments it has been reported that C protein specifically crosslinks to a U-rich sequence downstream of the cleavage site for polyadenylation (18) and to repeats of AUUUA in the 3[prime]-untranslated region of several mRNAs (19). Through more recent UV crosslinking experiments, it has also been reported that C protein specifically crosslinks to a region of four uridylates in the 5[prime] stem-loop structure of U2 snRNA (20). Finally, using a randomized sequence pool of 20 nt oligomers as in vitro substrates for C protein binding, it was observed that following eight rounds of binding and amplification (the SELEX procedure) the selected oligonucleotides possessed regions of five or more contiguous uridines (21). These non-equilibrium binding studies have led to the suggestion that C protein specifically binds in vivo to the polypyrimidine tract of introns (and perhaps to snRNAs) and functions to recruit snRNPs and other splicing factors to the 3[prime] consensus sequence for spliceosome assembly (17,21-23).
In contrast to the binding studies mentioned above, it has been observed more recently that when cDNAs for human C1 and C2 are expressed in bacterial cells, stable homo C14 and C24 tetramers assemble spontaneously. Under equilibrium conditions these proteins bind single-stranded RNA through a highly cooperative binding mode such that the overall binding affinity for poly(A) and poly(C), of sufficient length to saturate the tetramer (>230 nt), is significantly greater than for poly(U) (14,24). It thus became important to determine if, under equilibrium binding conditions, native C protein displays the same binding properties as recombinant protein and if short U-rich sequences are bound by C protein in preference to other substrates. In the studies described here a competition binding assay conducted under equilibrium conditions was used to investigate the relative affinity of native C protein for several short RNA sequences that have been postulated to function as target sequences for specific binding (17-19,21,25). Included in these studies were RNA oligos corresponding to 3[prime] splice sites possessing authentic polypyrimidine tracts, a uridylate-rich 3[prime]-untranslated region common to several mRNAs, several other uridylate-rich sequences, randomly generated sequences, an intronic sequence with no known splicing function, a SELEX-identified `winner' sequence previously shown to bind with high affinity to bacterially expressed C protein and to variants of the winner sequence (Table 1). As reported here, in comparison with random oligonucleotide sequences of the same length and to intronic sequences, native C protein did not show significantly higher affinity for any of the previously reported C protein target sequences or for uridine-rich oligonucleotides. Also, while the experiments described here confirm the enhanced affinity of C protein for a SELEX-identified winner sequence, they demonstrate that the interaction is not solely due to the presence of contiguous uridines.
MATERIALS AND METHODS
C-protein purification and quantification
C protein was purified from HeLa S3 cells under native conditions from density gradient-purified 40S hnRNP core particles as described previously (26,27). In these experiments an additional elution from the strong anion exchange column, Mono-Q (Pharmacia HR 5/5), was utilized to achieve electrophoretic homogeneity. C protein at concentrations of 1-2 mg/ml was equilibrated in binding buffer A (10 mM Tris, pH 8.0, 1 mM EDTA) and various concentrations of NaCl depending on the titration. Protein concentrations were determined using the BCA assay (Pierce). The BCA assay was accurate to within ±5% as determined through quantitative amino acid analysis. Because C protein contains no tryptophan, quantitative amino acid analyses were also utilized to demonstrate the absence of tryptophan in the purified native C protein preparations.
Fluorescence titrations
Equilibrium binding isotherms were obtained by measuring the change in the fluorescence signal (enhancement) that occurs when C protein binds the fluorescent probe RNA poly[r([epsis]A)]. Fluorescence measurements were performed with an SLM Aminco Bowman Series 2 Luminescence Spectrometer. Excitation and emission slits were fixed at 4 and 8 nm respectively. Titrations were conducted by exciting a fixed amount (0.5-2 µM) of the probe RNA at 310 nm and measuring the change in its emission at 410 nm as a function of increasing protein concentration. The probe RNA was prepared by treating poly(A) with chloroacetaldehyde (28). Fluorescence measurements were corrected for dilution due to the addition of protein sample and for background fluorescence. In all cases, the dilution of probe RNA did not exceed 5% of the total sample volume. Inner filter corrections were not necessary, since the quantities of RNA used in these titrations have negligible absorbance at 310 nm.
Data analysis
Binding parameters for the interaction of native C protein with the probe RNA under equilibrium conditions were derived by simulating theoretical curves to fit the experimental data using the non-cooperative model of McGhee and von Hippel for the interaction of a ligand (protein) with a nucleic acid (29).
| [nu](i)/L = K[1 - n[nu]][1 - n[nu]][1 - (n - 1)[nu]]-1 |
In the above equation, [nu] is the binding density (mol protein bound/mol nucleotide), L is the free ligand (protein) concentration and K is the association constant. The lack of sigmoidicity in the isotherms for the interaction of C protein with probe RNA suggests that the cooperativity parameter [omega], if it exists, is <10n, where n is the ligand binding site size (30). Since cooperativity can only be determined when [omega] is substantially greater than n, we used the simplest model (non-cooperative) to fit the experimental data.
To reduce the number of fitted parameters we independently determined n from stoichiometric titrations of the probe RNA. In addition, the site size has previously been determined through ultrastructural and hydrodynamic studies using various synthetic pre-mRNAs and binding substrates (6,31). Likewise, maximum fluorescent enhancement (Emax) was obtained from experimental binding isotherms. To generate a theoretical set of binding parameters we varied [nu] from 0 to 1/n (the maximum binding density at 100% saturation is 1/n). Theoretical enhancement values (Ecalc) were calculated for any given Emax, K, n and [omega] using the relationship
| Ecalc = n[nu]Emax |
Total protein concentrations corresponding to calculated E values were determined from the equation
| Lt calc = Lb + Lf = (n[nu]Rt)/n + Lf |
where Lt and Lb are the total and bound ligand respectively and Rt is the total RNA concentration at any one point during the titration. L was calculated by solving the McGhee-von Hippel equation for free ligand concentration. Lb was calculated in terms of (n[nu]Rt/n). Sets of theoretical data were iteratively generated until the sum of the squared deviation between the experimental and calculated data were minimized. In addition, we found that visual inspection of the binding isotherms are consistant with this analysis.
To confirm binding parameters obtained through fitting routines, non-linear regression analyses were also performed on the experimental data using a modification of methods described elsewhere (28). Root mean square errors in the optimization were evaluated using an iterative algorithm which numerically searches for the value of the binding density, [nu](i), which corresponds to the ith experimental concentration of C protein, Lt exp(i). An initial value for the C protein bound to RNA, Lb(i), is obtained from
| Lb(i) = [E(i)/Emax × n[nu](i)] × Rt |
where Rt is the concentration of RNA, Emax is the maximum enhancement and E(i) is the experimentally determined enhancement. Thus, the parameters K, Emax and n were optimized in a non-linear least squares algorithm (32). The binding site sizes n determined from these fits were within 10% of the value determined from stoichiometric titrations.
| [nu](i) = [Lb(i)/Rt] |
The binding density [nu](i) is calculated from the above expression. The McGhee-von Hippel non-cooperative equation is solved for free C protein concentration. A new value for Lb(i) is calculated knowing the binding density at each stage of the titration. The corresponding value for Lt is calculated from
| Lt - 1 = Lb + L |
Lt - 1 calc(i) is compared with the experimental value Lt exp(i) and [nu](i) is iteratively incremented until the difference between the experimental and calculated Lt values is reduced (<0.01% error). The value of the enhancement observed Eobs(i), which corresponds to the final Lt - 1 calc(i) [i.e. Lt exp(i)] is calculated from the following expression
| Eobs(i) = [Lb(i)] × Emax × n/Rt |
Therefore Eobs(i) is calculated for a value of Lt exp(i).
Competitor RNAs
Except for r(X)20, r(U)20, r(G)20 and Ad2E1a, the oligonucleotides used in these studies were prepared commercially by Midland Certified Reagent and by Cruachem. The excepted oligonucleotides were prepared at the W.M.Keck Foundation at Yale University. Each oligonucleotide was purified via HPLC prior to use. The sequence of each oligonucleotide is shown in Table 1. The oligonucleotide denoted Ad2 IVS1 is the polypyrimidine tract from IVS1 of the adenovirus-2 major late transcript. The polypyrimidine tract termed Ad2 E1A is from the pre-mRNA encoding the E1A oncoprotein of adenovirus-2 (33). The oligonucleotide denoted U2AF65 consensus was deduced from a series of SELEX-identified winner sequences for the splicing factor U2AF65, known to specifically bind polypyrimidine tracts as an early event is spliceosome assembly (34). The polypyrimidine tract termed B3P3 is present in the second intron of human [alpha]-tropomyosin. It has been found to be a high affinity substrate for PTB (polypyrimidine tract binding protein) (35). The oligonucleotide termed AUUUA repeats is a 20 nt portion of the [Delta]2R1 sequence from the 3[prime]-untranslated region of the mRNA for granulocyte-macrophage colony stimulating factor (GM-CSF) (19). The 20 nt [beta]-globin intronic sequence termed [beta]-intron is present in IVS1 of human [beta]-globin pre-mRNA. This oligonucleotide is not a polypyrimidine tract and it has no known function in splicing (23). The oligonucleotide termed r(U)20 is a 20 nt homoribopolymer of uridine and r(G)20 is a 20 nt homoribopolymer of guanosine. The oligonucleotide r(X)20 is a 20 nt sequence produced under conditions where the nucleotide at each position was randomly chosen (36,37). The oligonucleotide termed winner was identified through the SELEX procedure using bacterially produced C protein. Additionally, this sequence was shown by a filter binding assay to be a high affinity substrate for C protein (21). The oligonucleotides termed winner-G and WinRan are identical to the winner sequence except that the five contiguous uridines were replaced with five guanosines or with UAACC.
Table 1.
| Description | Sequence | Apparent affinity (M-1) |
| Ad2 IVS1 | rGUCCCUUUUUUUUCCACAG/C | 6.4 ×105 (± 5.3 × 104) |
| Ad2 E1a 3[prime] | rUGAUUUUUUUAAAAG/GUCCU | 4.3 × 105 (± 3.0 × 104) |
| U2AF65 consensus | rUUUUUUCCCCUUUUUUUUCC | 7.0 × 105 (± 4.7 × 104) |
| B3P3 | rCUUUCUCUUUCUCUCUCCCU | 1.4 × 105 (± 1.6 × 104) |
| AUUUA repeats | rCAUUUAUUUAUUUAUUUAAG | 6.3 × 105 (± 2.3 × 104) |
| [beta]-intron | rGAUCACUUGUGUCAACACAG | 2.4 × 106 (± 1.1 × 106) |
| r(X)20 | 2.4 × 105 (± 1.9 × 104) | |
| r(U)20 | 6.5 × 105 (± 1.6 × 104) | |
| r(G)20 | 4.8 × 107 (± 2.4 × 107) | |
| Winner | rAGUAUUUUUGUGGA | 3.1 × 107 (± 1.3 × 107) |
| Winner-G | rAGUAGGGGGGUGGA | 1.1 × 106 (± 4.5 × 104) |
| WinRan | rAGUAUAACCGUGGA | 2.1 × 105 (± 3.7 × 104) |
Binding isotherms using competitor RNAs
The concentration of each oligo was calculated using the nearest neighbor approximation method described in Borer et al. (38). For competition assays a control titration of the probe RNA with purified native C protein was carried out in the absence of competitor. To determine if the added oligonucleotides affect probe fluorescence, titrations were also performed with increasing amounts of each oligonucleotide. Titrations were then performed with C protein in the presence of the different oligonucleotide competitors (described in Table 1). The competitor RNAs were used in 11- to 50-fold excess (total nucleotides) depending on the strength of the competitor as a substrate for C protein binding (determined through test titrations).
An expression for the apparent equilibrium constant (Kapp) for the interaction of C protein with each oligonucleotide was derived considering the following equilibria (36,39)
| P + [epsis]A [harr] P[epsis]A |
and
| P + comp [harr] [Pcomp] |
where P, [epsis]A and comp are the steady-state concentrations of protein, poly[r([epsis]A)] and competitor RNA respectively. P[epsis]A and Pcomp are the equilibrium concentrations of protein-([epsis]A) or protein-competitor RNA complex. The mass action expression defining the Kapp for the interaction of protein with poly[r([epsis]A)] (KeA) or competitor RNA (Kcomp) are
| KeA = [P[epsis]A]/[P]free[[epsis]A]free | 1 |
and
| Kcomp = [Pcomp]/[P]free[comp]free | 2 |
In a competition titration, the free protein concentration for both mass action expressions is the same (equations 1 and 2). As a result, both expressions can be equated through [P]free. The solution for Kcomp is shown in equation 3 below.
| Kcomp = [Pcomp] × KeA × [[epsis]A]free/[P[epsis]A][comp]free | 3 |
In the above equation KeA was derived from a fit of data from a control titration (no competitor present) using the McGhee-von Hippel non-cooperative binding equation (29).
RESULTS
In the experiments described here the relative affinity of purified native C protein for various RNA substrates was determined through a competition binding assay conducted under equilibrium binding conditions. The fluorescent RNA poly[r([epsis]A)] was used as a `reporter' or probe substrate. In typical experiments the probe RNA (0.7-2.8 µM) was irradiated at 310 nm and the emission spectrum was monitored at 410 nm. In competition binding experiments, the magnitude of fluorescence attenuation in the presence of competitor RNA correlates with C protein binding to, and its relative affinity for, the competitor RNA (see Materials and Methods).
The stoichiometry of the native C protein-poly[r([epsis]A)] interaction at 100 mM salt was determined by titration with highly purified native C protein obtained from exponentially growing HeLa cells. Non-linear regression analyses of the binding isotherm yielded a binding stoichiometry of 260 nt/tetramer (Fig.
Figure 1. Equilibrium binding parameters for the interaction of native C protein with the fluorescent probe RNA poly[r([epsis]A)] at a concentration of 0.45 µM. This control titration was conducted in buffer A (10 mM Tris, 1 mM EDTA, pH 8.0) at 25°C containing 100 mM NaCl. The equilibrium binding parameters were obtained by simulating theoretical curves (smooth line) to fit the experimental data (open circles) using the McGhee-von Hippel non-cooperative model assuming a site size (n) of 230 nt. Through this method, the intrinsic association for the interaction of native C protein (C1)3C2 with poly[r([epsis]A)] is 5.8 × 106/M. The binding parameters were further defined through a non-linear regression analysis of the experimental data. In close agreement with the theoretical curve, this analysis yielded an intrinsic affinity of 5.6 × 106/M and a binding site size of 260 nt. To determine the relative binding affinity of native C protein for short target sequences under equilibrium conditions, a competition binding assay was utilized. The RNAs used as competitor substrates in these experiments are shown in Table 1 and described in Materials and Methods. To more definitively compare the relative strengths of these oligos as substrates for C protein binding, an 11- to 50-fold excess of competitor to probe (in nucleotides) was used in the binding assays. Through this approach, binding isotherms generated in the presence of competitor RNAs are less likely to superimpose on the probe curve and subtle differences in weak binding affinities are amplified. The control titration (with no competitor) is apparent as the upper solid tracing in Figures Figure 2. Determination of the relative binding affinity of C protein for the [beta]-globin intronic sequence (open circles), for two adenoviral polypyrimidine tracts (Ad2 IVS1, open triangles; Ad2 E1a, open diamonds) and for the randomly generated oligonucleotide r(X)20 (open squares). The solid tracing shows the binding isotherm for the control titration (probe RNA only). The tracings of the competitor titrations here and in Figures 4-7 are interpolations through the data points. In these competition binding experiments the probe RNA was 0.48 µM. Concentrations of the competitor RNAs were as follows: [beta]-globin, 25.0 µM; Ad2 IVS1, 23.1 µM; Ad2 E1A, 23.5 µM; r(X)20, 26.7 µM. To determine if C protein displays an enhanced affinity for bona fide polypyrimidine tracts, competition titrations were conducted with polypyrimidine tracts from IVS1 of the adenovirus-2 major late transcript, from the adenovirus-2 oncoprotein E1A 3[prime] splice site (33) and from IVS2 of human [alpha]-tropomyosin (the B3P3 element) (35). Also tested was the consensus polypyrimidine tract deduced from the sequence of several SELEX-identified winner sequences for U2AF65 binding (21). In some experiments a 20 nt oligo(U) sequence [r(U)20] was also used as a competitor. In all experiments the 20 nt [beta]-globin intronic sequence (described above) and randomly generated r(X)20 were used as competitor controls. Figure Figure 3. Histogram of the relative affinity of native C protein for the various competitor RNAs used in these studies. Note that the histograms denoting the relative affinities for r(G)20 and for the C protein winner sequence have been truncated (see scale at left), thus the apparent magnitude of their difference from other competitor RNAs is greatly under-represented. Figure 4. Determination of the relative binding affinity of C protein for the [beta]-globin intronic sequence (open circles), for the U2AF65 `winner' sequence (open triangles), for the randomly generated oligonucleotide r(X)20 (open squares) and for the B3P3 sequence (open diamonds). The solid tracing represents the binding isotherm for the control titration (probe RNA only). In these competition binding experiments the probe RNA was 0.48 µM. Concentrations of the competitor RNAs were as follows: [beta]-Globin, 25.0 µM; U2AF65 winner, 23.4 µM; r(X)20, 26.7 µM; B3P3, 23.2 µM. It has previously been suggested from UV crosslinking and immunoprecipitation experiments that C protein binds a 3[prime]-untranslated region in the mRNA for GM-CSF (19) and for the [beta]A4 amyloid protein (25). Presumably this specificity is due to interactions between C protein and repeats of an AUUUA sequence element. To determine if C protein displays enhanced affinity for this sequence under equilibrium conditions, a 20 nt region of the [Delta]2R1 oligonucleotide (possessing the AUUUA repeats) described in Hamilton et al. (19; Table 1) was used as a competitor in the equilibrium binding assay. The binding isotherm for this oligonucleotide reveals an affinity of 6.3 × 105/M, which is less than the affinity of C protein for the [beta]-globin intronic sequence (2.4 × 106/M) and is not significantly greater than (~2-fold) the affinity of C protein for the randomly generated sequence r(X)20 (2.4 × 105/M) (Figs Figure 5. Determination of the relative affinity of C protein for the [beta]-globin intronic sequence (open circles), the randomly generated sequence r(X)20 (open squares) and the oligonucleotide possessing AUUUA repeats (open diamonds). The solid tracing represents the binding isotherm for the control titration (probe RNA alone). In these competition binding experiments the probe RNA was 0.48 µM. Concentrations of the competitor RNAs were as follows: [beta]-globin, 25.0 µM; r(X)20, 26.7 µM; AUUUA repeats, 23.3 µM. Using the selection and amplification procedure of Tuerk and Gold (41) recombinant C protein was found to select a group of oligonucleotides from an infinite pool of randomly generated sequences that all possess at least one group of five or more contiguous uridines (21). From these findings it was assumed that this feature is the determinant of the affinity of C protein for the selected oligonucleotides. To determine if native C protein binds a SELEX-identified `winner' sequence under equilibrium conditions with especially high affinity, the winner shown in Table 1 was utilized in the competition binding assay. In the filter binding studies of Gorlach et al. (21) it was found that C protein binds this oligonucleotide with high affinity. To gain further insight into the role of contiguous uridines in C protein binding, variants of the winner sequence lacking contiguous uridines were also tested, as well as 20 nt homoribopolymers of uridine and guanosine. In Figures Figure 6. Determination of the relative affinity of C protein for the [beta]-globin intronic sequence (open circles), for r(G)20 (open diamonds), for r(U)20 (open triangles) and for r(X)20 (open squares). The solid tracing represents the binding isotherm for the control titration (probe RNA alone). In these competition binding experiments the probe RNA was 0.48 µM. Concentrations of the competitor RNAs were as follows: [beta]-globin, 25.0 µM; r(G)20, 25.5 µM; r(U)20, 24.0 µM; r(X)20, 26.7 µM.
Figure 7. Determination of the relative affinity of C protein for a SELEX-identified winner sequence (open circles), for the winner sequence containing five contiguous G residues (open diamonds) and for the winner sequence possessing the sequence UAACC in place of the five contiguous uridines (open squares). The solid tracing represents the binding isotherm for the control titration (probe RNA only). In these competition binding experiments the probe RNA was 0.48 µM. Concentrations of the competitor RNAs were as follows: winner sequence, 5.2 µM; winner-G, (24.4 µM); winner-UAACC, 23.2 µM. The findings described above demonstrate that under equilibrium conditions native C protein does not bind with significantly enhanced affinity to short oligonucleotide sequences previously suggested to function as target sequences for C protein binding in vivo. Namely, to the uridine-rich polypyrimidine tracts of introns (17,21,23,42) or to AUUUA repeats present in the 3[prime]-untranslated regions of some mRNAs (19,25). As summarized in Figure In the SELEX studies on C protein it was reported that each of 28 SELEX-identified oligonucleotides possessed at least one group of five or more contiguous uridines and it was suggested that this element was the determinant of C protein binding (21). However, several observations described here indicate that contiguous uridines do not alone define a high affinity substrate for C protein nor do they alone define the high affinity interaction of C protein with the winner sequence. First, C protein has no enhanced affinity under equilibrium conditions for bona fide polypyrimidine tracts possessing runs of five or more uridines nor does it display enhanced affinity for poly(U)20. Secondly, C protein binds poly(G)20 with essentially the same affinity as it binds the winner sequence. Finally, when the five U residues are replaced with G residues there is a several fold drop in affinity yet C protein continues to bind the variant with affinities similar to or slightly higher than its affinities for the uridine-rich substrates tested. Taken together, these findings suggest that the winner oligos exist as unique structures with steric nucleotide positionings that optimize C protein interactions with short RNAs in solution. The physiological relevance of this interaction remains unclear because bona fide polypyrimidine sequences (possessing five or more contiguous uridylates) and other uridine-rich oligos are not preferentially bound by C protein under equilibrium conditions. It is also possible that unique RNA secondary structures in pools of short oligonucleotides may be absent in longer pre-mRNAs containing the same sequence elements. In this context, the winner sequence does not correspond to known repeated sequence elements in pre-mRNAs. A finding perhaps relevant to the above discussion is the recent report that C protein binds the 5[prime] stem-loop of U2 snRNA (20). In this study it was found that elements within the stem structure and the four contiguous uridines in the loop region both function in binding. Also, in this context, it has been shown that the high affinity and salt-resistant RNA binding properties of C protein are not due to its N-terminal RRM but to a highly basic region preceding a leucine zipper motif (a bZIP-like motif, termed the bZLM) (24). It is therefore possible that C protein may bind to pre-mRNAs in a sequence-independent manner through its bZIP-like motif and to short RNAs possessing unique structure (i.e. snRNA) through its canonical RRM. The previous report that C protein binds poly(U)14 with a significantly higher affinity than the SELEX winner (21) is in sharp contrast to the results described here. It should be pointed out that the competition binding assay used here evaluates the relative ability of suggested target oligonucleotides to compete as substrates with a pre-mRNA length probe RNA. As shown previously, the SELEX winner is bound with high affinity by the bZLMs of C protein but not by its N-terminal RRM (14). It is therefore possible that poly(U)14 is bound with high affinity by the RRM of C protein (43) but that it competes poorly with long RNA substrates bound mostly by the tetramer bZLMs. Previous conclusions that C protein specifically binds short uridine-rich sequences in pre-mRNA may in part be due to methodological phenomena. For example, it has been observed by several groups that, in comparison with other hnRNPs, C protein readily crosslinks to RNA upon brief UV irradiation at sites possessing one or more uridines (5,19,20,44-47). This, however, would be expected if C protein is distributed along the length of nascent transcripts (through cooperative binding) and if uridine is especially photoreactive. The latter is well documented, as uridine is ~100-fold more labile to UV activation than are the other ribonucleotides (48-50). Estimates of binding preference based on salt-resistant dissociation from immobilized homoribopolymers are problematical because these approaches primarily evaluate differences in intrinsic affinity and in the electrostatic component of particular interactions. Additionally, cooperative binding cannot be evaluated through non-equilibrium approaches and studies conducted with substrates substantially shorter than n (the ligand binding site) are not likely to reflect the physiologically relevant binding equilibrium. In this context, we have utilized the r(U)20 and `winner' oligos described above as control substrates in a separate study to characterize the interaction of native C protein with various snRNAs (51). C protein was observed to bind snRNA substrates with affinities proportional to substrate length and at molar RNA concentrations 500-fold lower than were required in the present studies to detect binding to short oligonucleotides. More specifically, in conventional band shift assays binding of C protein to the oligonucleotides used here could not be detected (51). In addition to the findings described above that question the ability of C protein to identify uridine-rich sequences in pre-mRNAs, other findings point to the likely function of C protein as a sequence-independent RNA chaperonin that orders pre-mRNA and nucleates hnRNP assembly. Previous ultrastructural and hydrodynamic studies on native C protein and previous equilibrium binding studies on recombinant homo C14 and C24 demonstrate that a single tetramer binds ~230 nt of RNA and that three tetramers fold 700 nt increments of RNA into a unique triangular structure that nucleates correct 40S core particle assembly (6,14,31). It has been estimated that a typical HeLa cell contains ~22 × 106 molecules of C protein (11) and that there are ~0.4 × 106 nascent transcripts with average lengths between 6 and 10 kb. Because almost all C protein is nuclear (52), the molar concentration of C protein in the nucleus (12 µm spherical diameter) should be >10 µM. At these histone-like concentrations there is far more C protein in the nucleus than necessary to saturate the pool of pre-mRNAs with an occluded site size of 230 nt. C protein is a highly stable tetramer possessing at least eight RNA binding motifs (24). While this occurrence may be well suited for binding and ordering long lengths of RNA, it complicates arguments for specificity for short sequences. More specifically, an enhanced affinity for a short target sequence must be sufficiently high to override sequence-independent interactions at other sites in the oligomer and to negate cooperative binding. It has been estimated that, for a lattice binding ligand, its sequence-specific binding mode must minimally be an order of magnitude higher than its non-specific interactions, if the latter are to possess physiological relevance (29). This is especially significant in the case of an oligomer binding a lattice that, unlike DNA, lacks spatially repeating sequence elements. Supported by NIH grant GM 48567 to W.M.L.
C protein displays no significant affinity for U-rich polypyrimidine tracts
The affinity of C protein for a SELEX-identified winner sequence is not solely due to contiguous uridines
DISCUSSION
ACKNOWLEDGEMENT
REFERENCES
This article has been cited by other articles:
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 2 Jul 1998
Copyright©Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
J. E. Brunner, J. H. C. Nguyen, H. H. Roehl, T. V. Ho, K. M. Swiderek, and B. L. Semler
Functional Interaction of Heterogeneous Nuclear Ribonucleoprotein C with Poliovirus RNA Synthesis Initiation Complexes
J. Virol.,
March 15, 2005;
79(6):
3254 - 3266.
[Abstract]
[Full Text]
[PDF]
J. H. Kim, K. Y. Paek, K. Choi, T.-D. Kim, B. Hahm, K.-T. Kim, and S. K. Jang
Heterogeneous Nuclear Ribonucleoprotein C Modulates Translation of c-myc mRNA in a Cell Cycle Phase-Dependent Manner
Mol. Cell. Biol.,
January 15, 2003;
23(2):
708 - 720.
[Abstract]
[Full Text]
[PDF]
M. Holcik, B. W. Gordon, and R. G. Korneluk
The Internal Ribosome Entry Site-Mediated Translation of Antiapoptotic Protein XIAP Is Modulated by the Heterogeneous Nuclear Ribonucleoproteins C1 and C2
Mol. Cell. Biol.,
January 1, 2003;
23(1):
280 - 288.
[Abstract]
[Full Text]
V. Markovtsov, J. M. Nikolic, J. A. Goldman, C. W. Turck, M.-Y. Chou, and D. L. Black
Cooperative Assembly of an hnRNP Complex Induced by a Tissue-Specific Homolog of Polypyrimidine Tract Binding Protein
Mol. Cell. Biol.,
October 15, 2000;
20(20):
7463 - 7479.
[Abstract]
[Full Text]
D. J. Williamson, S. Banik-Maiti, J. DeGregori, and H. E. Ruley
hnRNP C Is Required for Postimplantation Mouse Development but Is Dispensable for Cell Viability
Mol. Cell. Biol.,
June 1, 2000;
20(11):
4094 - 4105.
[Abstract]
[Full Text]
O. Sella, G. Gerlitz, S.-Y. Le, and O. Elroy-Stein
Differentiation-Induced Internal Translation of c-sis mRNA: Analysis of the cis Elements and Their Differentiation-Linked Binding to the hnRNP C Protein
Mol. Cell. Biol.,
August 1, 1999;
19(8):
5429 - 5440.
[Abstract]
[Full Text]
[PDF]
This Article 
Abstract
Print PDF (108K)
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (13)
Request Permissions
Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Articles by Soltaninassab, S. R.
Articles by LeStourgeon, W. M.
Search for Related Content
PubMed
PubMed Citation
Articles by Soltaninassab, S. R.
Articles by LeStourgeon, W. M.
Social Bookmarking
What's this?