Folding of apomyoglobin: Analysis of transient intermediate structure during refolding using quick hydrogen deuterium exchange and NMR

Abstract

The structures of apomyoglobin folding intermediates have been widely analyzed using physical chemistry methods including fluorescence, circular dichroism, small angle X-ray scattering, NMR, mass spectrometry, and rapid mixing. So far, at least two intermediates (on sub-millisecond- and millisecond-scales) have been demonstrated for apomyoglobin folding. The combination of pH-pulse labeling and NMR is a useful tool for analyzing the kinetic intermediates at the atomic level. Its use has revealed that the latter-phase kinetic intermediate of apomyoglobin (6 ms) was composed of helices A, B, G and H, whereas the equilibrium intermediate, called the pH 4 molten-globule intermediate, was composed mainly of helices A, G and H. The improved strategy for the analysis of the kinetic intermediate was developed to include (1) the dimethyl sulfoxide method, (2) data processing with the various labeling times, and (3) a new in-house mixer. Particularly, the rapid mixing revealed that helices A and G were significantly more protected at the earlier stage (400 µs) of the intermediate (former-phase intermediate) than the other helices. Mutation studies, where each hydrophobic residue was replaced with an alanine in helices A, B, E, F, G and H, indicated that both non-native and native-like structures exist in the latter-phase folding intermediate. The N-terminal part of helix B is a weak point in the intermediate, and the docking of helix E residues to the core of the A, B, G and H helices was interrupted by a premature helix B, resulting in the accumulation of the intermediate composed of helices A, B, G and H. The prediction-based protein engineering produced important mutants: Helix F in a P88K/A90L/S92K/A94L mutant folded in the latter-phase intermediate, although helix F in the wild type does not fold even at the native state. Furthermore, in the L11G/W14G/A70L/G73W mutant, helix A did not fold but helix E did, which is similar to what was observed in the kinetic intermediate of apoleghemoglobin. Thus, this protein engineering resulted in a changed structure for the apomyoglobin folding intermediate.

Figure 1.

Primary and ternary structures of holoMb. (A) The structure of Mb (PDB: 1mbc) was expressed by MOLMOL.⁸⁸⁾ The locations of helices are indicated as upper-case letters. (B) Amino-acid sequence of sperm whale Mb was indicated with the underlines for the location of helices.⁶⁴⁾

Figure 2.

Mass spectrometry of the pH-pulse labeled samples. The pH-pulse labeling (pH 10.1, 20 ms) was employed after the initiation of the folding and each folding time (6.4 ms–6 s). The peaks for N, I, and U correspond with the native, intermediate, and unfolded states. The region of m/z for 7 + charge state is indicated. This figure was reproduced with modifications based on the original literature.⁴⁶⁾

Figure 3.

NMR analyses of the pH-pulse labeled samples. The pH-pulse labeling (pH 10.1, 20 ms) was employed after the initiation of folding and at each folding time. The transition of the proton occupancies (values for the ratio of folding which were calculated as a standard of the peak height for the longest refolding time) as a function of folding time is indicated for helix B residues (D27–K34). The region at residues of I28–F33 is folded in the burst phase intermediate, whereas both ends of helix B (D27 and K34) are folded in the slower phase from the I2 intermediate to the native state. This figure was produced with modifications based on the original literature.⁴⁶⁾

Figure 4.

Difference of the number of probe residues between (A) the new DMSO method (94 residues) and (B) conventional holoMb method (52 residues). The protection results derived from the experiments of pH-pulse labeling were mapped onto the crystal structure of Mb (PDB: 1mbc). The proton occupancies in the I2 intermediate of apoMb for each residue as elucidated using both methods were calculated. High, medium, and low protections are expressed by red, green, and blue spheres, respectively. The total numbers of spheres for the probe residues were 94 (A) and 52 (B), respectively. The residue numbers for the probe residues are given. The locations of helices are indicated as upper-case letters in all following figures. Figure 4A was reproduced based on the original literature.³⁹⁾ Figure 4B was produced based on the reported data.⁵⁰⁾

Figure 5.

Difference of the structures between (A) kinetic and (B) equilibrium intermediates. (A) The proton occupancy of the kinetic intermediate (I2) and (B) the protection factor (protected ratio normalized by the intrinsic exchange rate for HD exchange) for the equilibrium intermediate of apoMb. The locations of helices are shown at the top of the figures, represented as boxes. Protections of the (C) I2 (kinetic) and (D) equilibrium intermediates were mapped onto the crystal structure of Mb. High, medium, and low protections for residues are expressed by red (black residue number), green (green residue number), and blue spheres, respectively. The parts indicating the different protections between two intermediates are circled by color rings. These plots and mappings on the structure were reproduced with some modifications from the original paper.³⁹⁾

Figure 6.

(A) I2 and (B) equilibrium intermediates for two helix B stabilized mutants. (A) Differences in the proton occupancy for the I2 intermediate between the mutant and WT. Differences of the I2 intermediates between the G23A/G25A mutant (red squares) or H24L/H119F mutant (blue diamonds) and WT are shown. The numbers of residues, which display the difference of more than 0.1 for proton occupancy, are indicated by colors of red for G23A/G25A, blue for H24L/H119F or green for both. Green arrows are used for the more protected regions for both mutants compared to WT. Gray arrows indicated the locations of mutated regions. (B) Differences of protection factors for the equilibrium intermediates between the mutant and WT were shown in the way similar to (A). The numbers of residues in which the change of more than two times or less than 0.5 was observed for the protection factor are indicated. Blue arrows indicated the more protected regions for the H24L/H119F mutant than for WT. These plots were revised with some modifications based on the original literature.⁵⁵⁾

Figure 7.

Differences of the I2 intermediates between two helix F stabilized mutants: (A) P88K/S92K and (B) P88K/A90L/S92K/A94L. Experimental data of the proton occupancy for P88K/S92K (red circles) and P88K/A90L/S92K/A94L (blue circles) are shown. The proton occupancy of the WT is overlaid (open black circles). Some residue numbers are indicated. The regions for helix F residues (black ring) and GH-turn (blue ring) are highlighted for the P88K/A90L/S92K/A94L mutant. (C) AABUF plots of P88K/S92K (red line only for helix F residues) and P88K/A90L/S92K/A94L (blue for entire region) are indicated. (D) Differences of the proton occupancies between the P88K/A90L/S92K/A94L mutant and the P88K/S92K mutant. The more protected region, including helices G and H for the P88K/A90L/S92K/A94L mutant, is indicated by a blue open square. The less protected regions including helix B and a part of helix E for the P88K/A90L/S92K/A94L mutant are indicated by red open squares. The numbers of residues that indicated more (+0.1) or less (−0.1) protection in proton occupancy by mutations are shown. All plots were reproduced with some modifications based on the original paper.⁸⁰⁾

Figure 8.

Difference of the I2 intermediate structure between (A) apoLb and (B) apoMb. Mapping of the proton occupancy derived from the pH-pulse labeling experiment on the crystal structure of Mb and Lb (PDB: 1BIN). (A) The kinetic intermediate for apoLb and (B) the kinetic intermediate I2 for apoMb are indicated with high (red sphere), medium (green sphere), and low protections (blue sphere). The locations of differing protections between apoLb and apoMb intermediates are indicated by blue or red rings for helices A and B or for a part of helix E. The residue numbers (black, green, and blue) of the probe residues are indicated for high, medium, and low protections, respectively. Figure 8A⁸⁴⁾ and Figure 8B⁵⁰⁾ were produced based on the reported data.

Figure 9.

Protein engineering for the production of the apoLb-like apoMb intermediate. (A) AABUF plots for the L11G (red), W14G (blue), A71L (green), G73W (pink), and L11G/W14G/A71L/G73W (orange) mutations and the WT (black dash) based on the sequence. The area with the higher score is predicted as the protected region in the intermediate. (B) Experimental data for proton occupancy with dots and (C) prediction from the amino-acid sequence with lines for the I2 intermediate for the L11G/W14G/A71L/G73W mutant (blue) and WT (black). All plots were revised with some modifications based on the reported literature.⁸⁶⁾

Figure 10.

Mapping of the stepwise helix formation of apoMb onto the crystal structure of Mb. The sequential order of the helix formation from unfolded state to native state is supposed to be helices G (magenta),⁷⁰⁾ A (red),^59,60) H (blue),^23,60) B (orange),^39,50,70) E (green),^37,50,59) C (yellow),⁵⁹⁾ and D (aquamarine). Helix F and the GH-turn are also colored (light gray and light blue, respectively). The structure of Mb was expressed by MOLMOL.⁸⁸⁾ References are Vahidi et al. (2012),⁷⁰⁾ Vahidi et al. (2013),⁵⁹⁾ Uzawa et al. (2008),⁶⁰⁾ Jennings et al. (1993),²³⁾ Nishimura et al. (2005),³⁹⁾ Nishimura et al., (2002),⁵⁰⁾ and Xu et al. (2012).³⁷⁾