Stopped-flow kinetics support a two-state folding mechanism of the RNP domain

Unfolding of the RNP domain variants under equilibrium appeared to proceed without stable intermediates. Additional unfolding transitions for WT-RNP and dS-RNP by near-UV CD spectroscopy were monitored to follow unfolding of the secondary structures. Unexpectedly, the transition midpoints detected by CD spectroscopy were shifted to higher urea concentrations than those detected by fluorescence (Supporting Information Fig. S2). This non-coincidence of CD and fluorescence signals is presumably caused by the surface-exposed position of the tyrosine residues (Supporting Information Fig. S3). Consequently, unfolding transitions were repeated in the presence of 0.1M acrylamide as a fluorescence quencher. Under these conditions, transition midpoints were shifted to even lower urea concentrations (Supporting Information Fig. S2). The result confirmed the assumption that low urea conditions are sufficient to increase the solvent exposure of the tyrosine residues.

To elucidate the folding pathway of the RNP domain in more detail, folding kinetics of WT-RNP and dS-RNP were analyzed by stopped-flow measurements (Fig. 5). A coincidence of fluorescence and CD signals during folding and unfolding was observed, supporting the assumption of a two-state folding mechanism. Neither in the folding nor in the unfolding branch, any significant deviations from linearity were detectable. Since stopped-flow kinetics present a robust assay for assessing if structure formation occurs via a two-state pathway,²¹ these results were considered as an additional proof for the absence of stable intermediates. Moreover, the ΔG_D values of the equilibrium studies agreed with the kinetic data for both proteins (Table​(Table11).

Open in a separate window

Figure 5

Chevron plots for the unfolding and refolding of RNP domain variants. (A) WT-RNP and (B) dS-RNP. Measurements were performed using fluorescence (circles) or CD measurements (squares). Data obtained from unfolding experiments are shown as black symbols and from refolding experiments as open symbols. The fits were calculated from the fluorescence data. Thermodynamic values are listed in Table I.

Notably, the folding rate, of dS-RNP was increased by factor two, while the unfolding rate, showed a 10-fold acceleration when compared to the wild-type domain (Table​(Table1).1). The faster folding rate is likely to be caused by the artificial disulfide bond, which may add to early contacts of native structures. Residues that form interactions in the native or transition state upon folding maybe constrained in close proximity, leading to faster folding. The accelerated unfolding rate could be caused by the replacement of both cysteines in WT-RNP by serines which probably lead to the faster disintegration of the native structure. Considering the low stability of ΔCys-RNP and SH-RNP and the need to add stabilizing salts during equilibrium measurements at room temperature, folding kinetics of these variants were not carried out. In summary, in the case of ΔCys-RNP and SH-RNP, a two-state folding process could not be deduced.

DNA technology for expression of mutant constructs

Mutations were introduced into a previously described expression construct⁸ for WT-PABPN1 in pET11a (Novagen) using the QuickChange II Site-Directed Mutagenesis Kit (Agilent Technologies). As a basis for disulfide-linked PABPN1 variants, the two existing cysteines of WT-PABPN1 were replaced by serines resulting in ΔCys-PABPN1. Forward primers for the replacement of cysteines 195 and 205 were 5′ G CTA GAA GCA CAC TTT CAT GGC AGT GGT TCA GTC AAC C 3′ and 5′ GC GTA ACT ATA CTC AGT GAC AAA TTT AGT GG 3′, respectively, and the corresponding complementary reverse primers. Based on ΔCys-PABPN1, codons for variant A185C/V201C-PABPN1 were introduced. Primers for A185C were 5′ GGC AAT GTG GAC TAT GGT GCA ACA TGT GA A GAG CTA GAA GC 3′ and for V201C, 5′ GGT TCA GTC AAC CGC TGT ACT ATA CTC TGT GAC AAA TTT AGT GGC C3′ including the complementary oligonucleotides. The RNP domain and mutant variants encompassing residues 167–254¹³ were amplified by PCR (Eppendorf Mastercycler Gradient) with Pwo polymerase from peqGOLD. Forward primer was 5′ TAT GTG CAT ATG GAG GCT GAT GCC C 3′ and reverse primer 5′ AT TAT GGA TCC TTA TTA GAT GCC TGG TCT GTT GG 3′. The PCR products were inserted into pET11a (Novagen) via NdeI and BamHI restriction sites.

Recombinant protein production and purification

PABPN1 and RNP domain variants were produced in E. coli BL21 (DE3) Rosetta 2 pLysS cells in the presence of chloramphenicol and carbenicillin (Applichem). Full-length PABPN1 was produced in complex medium in a 10 L bioreactor (Biostat C-DCU; Sartorius Stedim) according to an established protocol³⁶ with minor modifications: The feeding solution contained glycerol instead of glucose. Before induction with 1 mM isopropyl 1-thio-β-d-galactopyranoside (IPTG), the cultivation temperature was reduced to 30°C to minimize protein aggregation. Recombinant gene expression was continued for 3 h. Purification of PABPN1 was performed as described.^8,37 Buffers for ΔCys-PABPN1 lacked DTT. For the production of A185C/V201C-PABPN1, the wet cell pellet before cell disruption was resuspended in buffer containing 1 mM DTT. The elution buffer from the Blue Sepharose column contained 5 mM oxidized glutathione to trigger the formation of the intramolecular disulfide bond. From this oxidized form, referred to as dS-PABPN1, the reduced species SH-PABPN1 was obtained by overnight reduction on ice with 100 mM DTT.

For recombinant production of the RNP domain, cells were grown at 37°C in terrific broth (TB) in shake flasks. Induction with IPTG was done at OD₆₀₀ = 0.4–0.5 and cultivation continued for 3 h. All buffers for preparation of WT-RNP except the final storage buffer contained 1 mM DTT. For production of ΔCys-RNP and oxidized species of A185C/V201C-RNP, DTT was omitted. The reduced form of A185C/V201C-RNP was maintained by the presence of 5 mM Tris-(2-carboxyethyl)-phosphine hydrochloride (TECP).

Per gram wet cell pellet, 4 mL 50 mM Tris/Cl pH 8.0, 5 mM EDTA supplemented with complete Mini EDTA-free protease inhibitor (Roche) and 0.1 mg mL⁻¹ lysozyme (Applichem) was added. Cell disruption was performed by high-pressure homogenization. Subsequently, 10 U Benzonase (Novagen) per gram cells and MgCl₂ to a final concentration of 8 mM were added. Cell extracts containing WT-RNP and ΔCys-RNP were subjected to a HiPrep Q XL 16/10 (GE Healthcare). The flow-through was collected, and (NH₄)₂SO₄ was adjusted to a final concentration of 1.2M. After 1 h incubation at 4°C, precipitates were removed by centrifugation. The supernatant was loaded on a HiPrep Phenyl HP 16/10 column (GE Healthcare). Elution was achieved by a linear gradient from 1.2M to 0M (NH₄)₂SO₄ in 50 mM Tris/HCl pH 8.0, 5 mM EDTA. RNP domain containing fractions were dialyzed against 20 mM MOPS pH 6.5, loaded on a HiPrep SP FF 16/10 column (GE Healthcare) and eluted by a gradient from 0M to 1.5M NaCl. Fractions containing pure protein were applied to a HiLoadSuperdex 75 prep grade 16/60.

For production of A185C/V201C-RNP which accumulated in the insoluble fraction after cell lysis, per gram wet pellet, 5 mL solubilization buffer containing 100 mM Tris/Cl pH 8.0, 6M GdmCl, 5 mM EDTA, 100 mM DTT was added. Thereafter, dialysis against 10 mM Na-acetate, pH 4.0, 5M GdmCl was performed. Renaturation at a final protein concentration of 0.2 mg mL⁻¹ was performed by rapid dilution in 100 mM Tris/Cl pH 8.5, 1M l-arginine, 5 mM EDTA. To obtain oxidized species A185C/V201C-RNP, 5 mM oxidized and 2 mM reduced glutathione were added to the renaturation buffer. Reduced species was produced by renaturation in the presence of 10 mM DTT. Final purification procedures, including HiPrep SP FF 16/10 and size exclusion chromatography were done as described above. Protein purity was evaluated via SDS-PAGE and UV absorption. Finally, the RNP domains were dialysed against storage buffer (20 mM Tris/HCl pH 7.5) and kept at −70°C.

Circular dichroism and fluorescence spectroscopy

Circular dichroism (CD) spectra were recorded using a Jasco J810 spectropolarimeter and 0.1 mm cuvettes at 20°C. Protein concentration was between 0.5 and 1.0 mg mL⁻¹. Spectra were buffer-corrected. Fluorescence spectra were measured on a Jobin Yvon Fluoromax-3 in 1 cm cuvettes at 20°C. The protein concentration was 10 µM. Excitation wavelength was 274 nm and the bandwidth was 5 nm. Emission bandwidth was 8 nm and spectra were accumulated 10 times, averaged and buffer corrected. Native protein was buffered in 20 mM Tris/HCl pH 7.5. For denaturation, 6M guanidinium chloride was added to the buffer.

Denaturant- and heat-induced unfolding transitions

For urea-induced unfolding, RNP domain variants at a concentration of 2.5 µM were incubated in the presence of varying urea concentrations in 20 mM Tris/HCl pH 7.5 overnight. Urea-induced unfolding was monitored by fluorescence measurements on a Jobin Yvon Fluoromax-3 at 20°C in 1 cm cuvettes. Since no tryptophans are present in the RNP domain, emission at 307 nm was averaged over 30 s upon excitation of tyrosines at 274 nm. Excitation and emission band widths were 5 nm and 8 nm, respectively. Unfolding transitions were evaluated assuming a two-state model by nonlinear least-squares fitting.³⁸ Heat-induced unfolding/refolding signal changes were monitored by CD at 208 nm using a Jasco J810 spectropolarimeter and 1 mm cuvettes. Protein concentration was 24.5 µM in 20 mM KH₂PO₄, pH 7.5. The heating rate was 1 K min⁻¹. For calculation of T_m, CD signals were fitted according to a published method.³⁹

Stopped-flow measurements

Folding and unfolding rates were followed by fluorescence and CD spectroscopy at 20°C. In both methods 20 mM Tris/HCl, pH 7.5 was used as a buffer. Stock solutions with native (0M urea) or denatured (8M urea) protein were mixed before measurement at a ratio of 1 + 10 in buffer containing different urea concentrations to observe folding/unfolding kinetics. Fluorescence data were collected with an Applied Photophysics SX20 stopped-flow fluorimeter at a protein concentration of 5 µM. Excitation was set at 280 nm, emission was recorded above 305 nm using a cut-off filter. CD experiments were performed using an Applied Photophysics Pi-Star-180 spectrometer. Signals were detected at 235 nm with a slit width of 1 nm. The final protein concentration was 25 µM. Data were fitted as published previously.²¹

Valuable advice on the RNP domain purification was provided by Till Scheuermann. The assistance of Franziska Schikora and Victoria Junghans is gratefully acknowledged. We also thank Cordelia Schiene-Fischer for providing CD stopped-flow equipment.