|Home | About | Journals | Submit | Contact Us | Français|
Conceived and designed the experiments: DL. Performed the experiments: YW JL MX YP WY JH. Analyzed the data: YW JL. Contributed reagents/materials/analysis tools: DL. Wrote the paper: YW DL.
The conformational conversion of the host-derived cellular prion protein (PrPC) into the disease-associated scrapie isoform (PrPSc) is responsible for the pathogenesis of transmissible spongiform encephalopathies (TSEs). Various single-point mutations in PrPCs could cause structural changes and thereby distinctly influence the conformational conversion. Elucidation of the differences between the wild-type rabbit PrPC (RaPrPC) and various mutants would be of great help to understand the ability of RaPrPC to be resistant to TSE agents.
We determined the solution structure of the I214V mutant of RaPrPC(91–228) and detected the backbone dynamics of its structured C-terminal domain (121–228). The I214V mutant displays a visible shift of surface charge distribution that may have a potential effect on the binding specificity and affinity with other chaperones. The number of hydrogen bonds declines dramatically. Urea-induced transition experiments reveal an obvious decrease in the conformational stability. Furthermore, the NMR dynamics analysis discloses a significant increase in the backbone flexibility on the pico- to nanosecond time scale, indicative of lower energy barrier for structural rearrangement.
Our results suggest that both the surface charge distribution and the intrinsic backbone flexibility greatly contribute to species barriers for the transmission of TSEs, and thereby provide valuable hints for understanding the inability of the conformational conversion for RaPrPC.
The conformational conversion of the prion protein, from the normal cellular form (PrPC) to the abnormal scrapie isoform (PrPSc), is responsible for the pathogenesis of transmissible spongiform encephalopathies (TSEs) . The “protein-only” hypothesis is supported by much compelling evidence presented recently –. To elucidate the detailed mechanism of the conformational change, understanding of the structural basis of the prion protein is required. The three-dimensional structures of PrPC among species have been extensively determined using NMR techniques so far –. Great effort has also been made to explore the architecture of PrPSc or PrPSc-like filaments –.
Noticeably, all known forms of inherited human TSEs, including Creutzfeldt-Jakob disease (CJD), Gerstmann-Straussler-Scheinker syndrome (GSS) and fatal familial insomnia (FFI), are closely associated with dominant mutations in human PrPC (hPrPC) , . In addition to apparently disease-causing mutations in humans, polymorphisms in sheep appear to exert substantial effects on variations of PrPC in disease susceptibility –. Particularly, rabbits are one of the few mammalian animals reported to be resistant to TSE agents . Multiple amino acid residues throughout the rabbit PrPC (RaPrPC) sequence contribute to the inability of RaPrPC to form PrPSc. Several mouse PrPC (mPrPC) mutants, with one single residue substituted by the corresponding residue in RaPrPC, are completely inhibited to undergo the conversion to the disease-related isoform . All these facts indicate a significant influence on the property of the PrPC structure as the result of one important residue mutation. Therefore, detailed comparisons of three-dimensional structures of wild-type PrPC and its mutants could provide valuable insights into the underlying molecular mechanism of prion conversions.
In our previous work we have demonstrated that the S173N substitution leads to distinct structural changes for RaPrPC . However, whether similar changes could be observed in other single-point mutants remains to be addressed. Since mPrPC with the V214I substitution is prohibited to convert to the abnormal form , we thereby expect that the I214V substitution would cause structural changes more or less for RaPrPC. In this present work, we determined the solution structure of the I214V mutant of RaPrPC(91–228) using multi-dimensional heteronuclear NMR techniques. In addition, we performed 15N relaxation measurements to detect the backbone dynamics of its structured C-terminal domain (121–228). Furthermore, we investigated its structural stability using CD spectroscopy. Our results reveal significant structural changes caused by the single-residue mutation, which may be of benefit to understand the detailed molecular mechanism of the conformational conversion for prion proteins.
A family of 200 structures is calculated and the structural statistics are presented in Table 1. The diagram representing 15 lowest-energy structures for the I214V mutant of RaPrPC(91–228) in solution is shown in Fig. 1A, together with a ribbon cartoon of average secondary structure elements displayed in Fig. 1B. The I214 mutant contains two short antiparallel β-sheets (S1: 128–130, S2: 160–162) and three α-helices (H1: 144–155, H2: 175–186, H3:199–227), with a disulfide bond (C178–C213) stabilizing helices 2 and 3. The N-terminal loop 91–120 is highly disordered. Loop 165–172 is well defined owing to the long-range NOEs from residues at the end of helix 3. The overall structure of the I214V mutant appears to be identical to that of the wild-type (Fig. 1C).
We evaluated the effect of the I214V substitution on the surface charge distribution of RaPrPC(91–228). Both the mutant and the wild-type carry neutral charge at the right substituted site 214 (Fig. 2A, B). The I214V substitution does not change the electrostatic potential at site 214 due to the similar non-charged feature of the two amino acids. Unexpectedly, significant changes are observed at many other unsubstituted sites. For example, the wild-type displays a neutral charge distribution at site 124, while the I214V mutant carries positive charge in the same position (Fig. 2C, D). Site 164 also alters electrostatic potential from positive to neutral after the mutation (Fig. 2C, D). The I214V substitution results in a prominent shift of the surface charge distribution.
The hydrogen bond network usually makes a significant contribution to maintain secondary and tertiary structures of a protein. Stable hydrogen bonds are defined herein if they exist in at least 9 structures among the 15 lowest-energy structures. Totally, 41 hydrogen bonds are detected in the I214V mutant of RaPrPC(91–228), much less than the number 55 in the wild-type. In particular, only 4 hydrogen bonds participate in sustaining the tertiary structure of the mutant, compared with 8 in the wild-type (Table 2). As less hydrogen bonds are formed in the mutant, one could expect that the RaPrPC molecule is somehow readily to experience the conformational conversion after the mutation.
To access the conformational stability of the I214V mutant of RaPrPC(121–228), we performed urea-induced unfolding transition experiments using far-UV CD spectroscopy. The I214V mutant is rich in α-helix in the absence of urea (Fig. 3A), indicating a well-folded state. This mutant loses its secondary structure entirely under the condition of 9 M urea (Fig. 3A), implying a completely unfolded state. The mean residue ellipticity at 222 nm versus the urea concentration is shown in Fig. 3B, with a solid line showing the theoretical curve on the basis of a two-state mechanism. The denaturation was not reversible in our experimental condition (Fig. 3B), thus we determined the apparent thermodynamic parameters for the equilibrium unfolding of the I214V mutant (Table 3). is the concentration of urea required to denature 50% of a protein, and presents the apparent estimated free energy of unfolding extrapolated to zero concentration of denaturant. The measured values of and for the mutant are less than those for the wild-type (Table 3), indicating reduced conformational stability of RaPrPC after the I214V substitution. The coefficient m is also different from that for the wild-type (Table 3).
To analyze the backbone dynamics of the I214V mutant of RaPrPC(121–228), we performed 15N relaxation measurements using NMR spectroscopy. As a whole, 103 assigned residues are used except 5 residues with unobservable resonance signals. We utilized the peak height for curve fitting so as to avoid the effect of partially overlapped resonance peaks. The relaxation rates R1, R2 and heteronuclear NOEs versus residue number are shown in Fig. 4. The R1 values do not change much with the sequence, mostly between 1.1 and 1.5 s−1. Different from the R1 distribution, the R2 values are relatively variable with residue number, ranging from 3.0 to 18.1 s−1 approximately. The residue D166 shows the largest R2 value over 18.1 s−1, and G130 also displays distinctly large R2 value as high as 16.0 s−1. Almost all residues exhibit positive NOE values except the first three residues. Typically, NOE values for the two antiparallel β-sheets and three α-helices are obviously higher than those for the loop fragments encompassing residues 121–127, 137–141 and 188–197. The secondary structure elements display NOE values larger than 0.7, indicative of more restricted dynamics in these regions.
We adopted the spectral density function approach – to interpret the 15N relaxation data for the I214V mutant of RaPrPC(121–228). The calculated values of reduced spectral densities J(ωN), J(0) and J(0.87ωH) versus residue number are shown in Fig. 5A. The middle frequency spectral densities J(ωN) exhibit relatively invariable values ranging from 0.25 to 0.38 ns·rad−1, reflecting insensitivity to variations in backbone motion. Residues in helix 3 show a slight reduction in J(ωN) spectral densities compared to helices 1 and 2. It has been suggested that such a reduction reflects slightly anisotropic tumbling of PrPC .
The low frequency spectral density J(0) is sensitive both to fast internal motions on the pico- to nanosecond (ps-ns) time scale and to slow motions (Rex) on the micro- to millisecond (µs-ms) time scale. Rapid internal motions tend to reduce the J(0) value, while slow internal motions usually lead to anomalously large J(0) values . The plot of J(0) versus residue number shows the trend of R2 values change with the sequence since J(0) values are normally dominated by R2 values. In contrast to J(ωN), the spectral densities J(0) cover a wider range of values. Residues with lower J(0) values are mainly observed in loop regions 121–127, 138–141 and 189–197, implying sub-nanosecond flexibility of the N-H bond vector. Dissimilarly, loop 165–172 is not as flexible as the other three loop fragments. Residues in secondary structure elements have relatively higher J(0) values. However, the residue H186 in helix 2 displays a fairly small J(0) value, implying significant internal motion on the ps-ns time scale. D166 exhibits the largest J(0) value around 7.0 ns·rad−1, and G130 in β-sheet 1 also shows a remarkable J(0) value as large as 5.9 ns·rad−1, both indicative of slow µs-ms Rex motions. Furthermore, a few residues in helices 2 and 3 are likely to undergo slow Rex motions as well.
The high frequency spectral density J(0.87ωH) is sensitive only to fast internal motions on the sub-nanosecond time scale. Fast motions are reflected in relatively large values of J(0.87ωH). Aside from residues in loop regions 121–127, 138–141 and 189–197, the C-terminal end of helix 3 and the particular residue H186 show large J(0.87ωH) values, implying fast ps-ns motions. In addition, small J(0.87ωH) values and large J(0) values are observed in most residues in secondary structure elements, indicative of restricted motions on the fast ps-ns time scale.
The overall distributions of the reduced spectral densities J(ωN), J(0) and J(0.87ωH) for the I214V mutant are similar to those for the wild-type reported previously . We compared the values of the spectral densities between the wild-type and the I214V mutant, and the differences versus residue number are shown in Fig. 5B. The most significant changes are observed in J(0). Almost all residues exhibit distinctly negative values in ΔJ(0). Generally, the smaller the value of J(0), the greater the sub-nanosecond flexibility of N-H bond vectors . Thus the I214V mutation leads to increased internal dynamics.
The modelfree formalism is usually used to analyze internal motions of a protein –. The D/D ratio of the rotational diffusion tensor for the I214V mutant of RaPrPC(121–228) was calculated to be 1.24±0.01 using the r2r1_diffusion program (kindly provided by Prof. A.G. Palmer 3rd), suggesting that the axially symmetric model is suitable for the data fitting. The calculated order parameter S2 versus residue number is plotted in Fig. 6A. This parameter describes the amplitude of the sub-nanosecond timescale motions (0≤S2≤1). Residues in loop 135–141 and 189–197 display the lowest S2 values, which imply that these regions are highly disordered. In contrast to these two loops, loop 165–172 show higher S2 values, indicating an ordered fragment. The S2 values for residues in secondary structure elements are relatively larger, which is consistent to the J(0) distribution. The mean values of S2 for helices 1, 2 and 3 are 0.88, 0.88 and 0.82, respectively. H186 in helix 2 shows a fairly small S2 value as low as 0.56, indicative of significant internal flexibility incompatible with the secondary structure property. We further compared the difference in S2 values between the wild-type and the mutant (Fig. 6B). Almost all residues show negative ΔS2 values, providing solid evidence that the I214V substitution significantly enhances the backbone flexibility of RaPrPC. The ΔS2 values are mapped onto the mean structure of the I214V mutant (Fig. 6C).
Single-point mutation may result in functional alteration owing to the global or local structural change in the protein. Several mPrPC mutants exhibit local structural changes and show distinctly different behavior from the wild-type , . Three single-residue mutants, including hPrPC(M166V), hPrPC(S170N) and hPrPC(R220K), show variations in the length and quality of definition of helix 3 . In addition, loop 166–172, which is lack of backbone amide resonances in hPrPC and mPrPC, is well-defined in both hPrPC (S170N) and mPrPC (V166A) mutants , .
In this present work, we have determined the solution structure of the I214V mutant of RaPrPC(91–228). The three-dimensional structure of the mutant is almost identical to that of the wild-type (Fig. 1), however, both the altered hydrogen bond network and the changed surface charge distribution demonstrate that the I214V substitution could lead to marked structural changes (Fig. 2 and Table 2). Significant change in surface-restricted charges has also been observed in hPrPC (E200K)  and RaPrPC(S173N) mutants . It is well known that electrostatic interaction could distinctly influence on both the binding specificity and affinity of a protein with substrates. Thus, the mutation-induced change of surface charge distribution is likely to affect the binding of PrPC with many potential chaperones, including nucleic acids , , protein X , , sugars, lipids, etc., which would modulate the conformational conversion. Interestingly, both hPrPC(E200K)  and RaPrPC(S173N) mutants  exhibit distinct charge alterations at the right substituted sites, whereas the I214V mutant described herein remains the neutral charge at site 214 as wild-type RaPrPC carries (Fig. 2). These observations suggest that the charge alteration at the substituted site is not a precondition for the global electrostatic potential shift. The change of surface charge distribution resulting from single-point mutation could be more significant than expected.
It has been demonstrated that both the tertiary structure and internal motions of a protein determine the protein function . Therefore, evaluation of the differences in dynamics between wild-type PrPC and its mutants is important to address the molecular mechanism of prion conversions. Both molecular dynamics simulations and NMR relaxation measurement experiments have suggested that backbone flexibility of PrPC is associated with the conformational conversion , , , although some studies show that mutations do not lead to significant changes in backbone dynamics of proteins , . Our results demonstrate that the I214V substitution leads to decreased J(0) and S2 values (Fig. 5 and and6),6), implicating an increase in backbone flexibility on the ps-ns time scale, similar to the S173N substitution . Usually, fragments with high flexibility are indicative of low energy barrier to structural rearrangement. In addition, H186 in helix 2 displays a fairly small S2 value and a much lower J(0) value after the I214V mutation (Fig. 5 and and6),6), implying prominent internal motions which potentially make a significant contribution to destabilization of helix 2. Thus, it could be expected that the conformational change is somewhat easier to occur for the I214V mutant than for the wild-type. On the other hand, dynamics could modulate binding energy to maintain the moderate affinity required for biological function , . The increased dynamics might enhance or inhibit the binding ability of PrPC with other chaperones, which would exert profound influence on prion conversions.
We have assessed the change of thermodynamic stability of RaPrPC after the I214V substitution. The coefficient m is implicated to be approximately proportional to the difference in solvent-accessible surface areas between the folded and unfolded states. Hydrophobic interactions stabilizing the protein native state can be interpreted, to a large extent, in terms of the change in accessible surface area upon protein folding –. The distinctly changed m value is indicative of differences in intramolecular hydrophobic interactions between the wild-type and the mutant (Table 3). Furthermore, similar to the S173 substitution , the I214V substitution leads to marked decreases in both and values during urea-induced transitions (Fig. 3 and Table 3), implying that the I214V mutant adopts a less stable conformation compared with the wild-type. Our experimental observations are supported by both molecular dynamics simulations reported recently  and the altered hydrogen bond network described above (Table 2). It has also been shown that internal dynamics especially the ps-ns motions has an intimate correlation with thermodynamics , . The increased backbone flexibility, indicated by the decreased J(0) and S2 values (Fig. 5 and and6),6), implying the conformational instability of the I214V mutant. Therefore, one could expect that PrPC with low structural stability would be inclined to undergo conformational conversion.
In conclusion, the I214V substitution results in distinct structural changes for RaPrPC. Both the unique surface charge distribution and the intrinsic backbone are likely to make significant contributions to species barriers for the transmission of TSEs. Our results provide valuable hints for understanding the inability of RaPrPC to form PrPSc, and shed light on the detailed molecular mechanism of the conformation conversion for prion proteins.
The I214V mutants of RaPrPC(91–228) and RaPrPC(121–228) were constructed using site-directed mutagenesis PCR. Plasmid construction, protein expression and purification, NMR spectroscopy, structure calculation, 15N relaxation measurements, dynamics analysis and CD experiments were conducted as described previously , . Nearly complete backbone and side-chain resonance assignments for the I214V mutant of RaPrPC(91–228) have been obtained. The chemical shift data are available at the Biological Magnetic Resonance Data Bank with accession number 16616. The atomic coordinates have been deposited into the Protein Data Bank with PDB ID 2JOM.
We would like to thank Prof. Daiwen Yang (National University of Singapore) for helpful discussions on dynamics analysis and Prof. Gengfu Xiao (Wuhan University) for providing RaPrPC gene.
Competing Interests: The authors have declared that no competing interests exist.
Funding: This work was supported by grants from the National Natural Science Foundation of China (Nos. 30730026, 30570352) and National Science & Technology Major Project “Key New Drug Creation and Manufacturing Program,” China (No. 2009ZX09301-001). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.