Stability Prediction for Mutations in the Cytosolic Domains of Cystic Fibrosis Transmembrane Conductance Regulator

Abstract

Cystic Fibrosis (CF) is caused by mutations to the Cystic Fibrosis Transmembrane Conductance Regulator (CFTR) chloride channel. CFTR is composed of two membrane spanning domains, two cytosolic nucleotide-binding domains (NBD1 and NBD2) and a largely unstructured R-domain. Multiple CF-causing mutations reside in the NBDs and some are known to compromise the stability of these domains. The ability to predict the effect of mutations on the stability of the cytosolic domains of CFTR and to shed light on the mechanisms by which they exert their effect is therefore important in CF research. With this in mind, we have predicted the effect on domain stability of 59 mutations in NBD1 and NBD2 using 15 different algorithms and evaluated their performances via comparison to experimental data using several metrics including the correct classification rate (CCR), and the squared Pearson correlation (R²) and Spearman's correlation (ρ) calculated between the experimental ΔT_m values and the computationally predicted ΔΔG values. Overall, the best results were obtained with FoldX and Rosetta. For NBD1 (35 mutations), FoldX provided R² and ρ values of 0.64 and -0.71, respectively, with an 86% correct classification rate (CCR). For NBD2 (24 mutations), FoldX R², ρ, and CCR were 0.51, -0.73, and 75%, respectively. Application of the Rosetta high-resolution protocol (Rosetta_hrp) to NBD1 yielded R², ρ, and CCR of 0.64, -0.75, and 69%, respectively, and for NBD2 yielded R², ρ, and CCR of 0.29, -0.27, and 50%, respectively. The corresponding numbers for the Rosetta's low-resolution protocol (Rosetta_lrp) were R² = 0.47, ρ = -0.69, and CCR = 69% for NBD1 and R² = 0.27, ρ = -0.24, and CCR = 63% for NBD2. For NBD1, both algorithms suggest that destabilizing mutations suffer from destabilizing vdW clashes, whereas stabilizing mutations benefit from favorable H-bond interactions. Two triple consensus approaches based on FoldX, Rosetta_lpr, and Rosetta_hpr were attempted using either "majority-voting" or "all-voting". The all-voting consensus outperformed the individual predictors, albeit on a smaller data set. In summary, our results suggest that the effect of mutations on the stability of CFTR's NBDs could be largely predicted. Since NBDs are common to all ABC transporters, these results may find use in predicting the effect and mechanism of the action of multiple disease-causing mutations in other proteins.

Figure 1.

Crystal structures of CFTR NBDs used for $\text{ΔΔG}$ calculations: (A) NBD1 (PDB code: 2PZE); (B) NBD2 (9sol construct, 6UK1). The stabilizing and destabilizing mutations are represented as balls and sticks in blue and pink colors, respectively.

Figure 2.

FoldX - $\text{ΔTm}$ correlations. (A) The correlation between the experimental $\text{ΔTm}$ values and the FoldX predicted $\text{ΔΔG}$ values for all 35 NBD1 mutations. (B) The correlation between the experimental $\text{ΔTm}$ values and FoldX vdW clashes term for all 35 NBD1 mutations. Red points correspond to neutral mutations, which have small $\text{ΔTm}$ values between −0.5°C to 0.5°C. Extreme points are circled in black.

Figure 3.

Correlation graphs between the experimental and predicted $\text{ΔTm}$ values for 2-term MLR model composed of FoldX energy terms: Training set (A) and test set (B). Upon removal of the extreme points circled in black, correlation decreased to: (A) 0.59, p = 3.3e⁻⁵; and (B) 0.70, p = 0.001.

Figure 4.

FoldX - $\text{ΔTm}$ correlations. The correlation between the experimental $\text{ΔTm}$ values and the FoldX predicted $\text{ΔΔG}$ values for all 24 NBD2 mutations. Upon removal of the extreme points circled in black correlation slightly improved to 0.53 (p = 9.15e⁻⁵) for all mutations and to 0.53 (p = 0.0006) excluding neutral mutations. The points coloured in red denote neutral mutations having small $\text{ΔTm}$ values between −0.5°C to 0.5°C.

Figure 5.

Rosetta_hrp - $\text{ΔTm}$ correlations. (A) The correlation between the experimental $\text{ΔTm}$ values and the Rosetta_hrp predicted $\text{ΔΔG}$ values for all 35 NBD1 mutations. (B) The correlation between the experimental $\text{ΔTm}$ values and Rosetta_hrp ‘fa_rep’ energy term for all 35 NBD1 mutations. Upon removal of the extreme points circled in black correlation reduced to: (A) ${\text{R}}^2$ = 0.47, p = 7.09e⁻⁶ for all mutations and ${\text{R}}^2$ = 0.47, p = 3.95e⁻⁵ excluding neutral mutations; (B) ${\text{R}}^2$ = 0.37, p = 0.0001. Upon additional removal of points circled in red (R560T in (A) and R560T and R560K in (B)), correlation increased back to: (A) ${\text{R}}^2$ = 0.56, p = 4.75e⁻⁷ for all and ${\text{R}}^2$ = 0.56, p = 4.13e⁻⁶ excluding neutral mutations; (B) ${\text{R}}^2$ = 0.66, p = 1.76e⁻⁸. Finally, upon removal of only the outliers circled in red, the correlation increased to: (A) ${\text{R}}^2$ = 0.72, p = 2.09e⁻¹⁰ for all and ${\text{R}}^2$ = 0.72, p = 5.86e⁻⁹ excluding neutral mutations; (B) ${\text{R}}^2$ = 0.75, p = 1e⁻¹⁰. Red points are neutral mutations.

Figure 6.

Correlation between the experimental and predicted $\text{ΔTm}$ values for 1-term (equation 4) and 2-term (equation 5) MLR models derived from the Rosetta_hrp energy terms. (A): Training set 1-term model. Upon removal of the extreme point circled in black correlation reduced to 0.59, p = 4.69e⁻⁵; (B): Test set 1-term model. Upon removal of the extreme points circled in black the correlation reduced to 0.52, p = 0.028; (C): Training set 2-term model. Upon removal of the extreme point circled in black correlation reduced to 0.63, p = 1.91e⁻⁵; (D): Test set 2-term model.

Figure 7.

The correlation of $\text{ΔTm}$ with severity of folding defect in CFTR. Mutations are classified according to their types: premature stop codon (I), trafficking/folding defect (II), gating defect (III), reduced Chloride ion transport (IV), splicing defect (V) and not CF causing (none). CFTR folding is represented as band C divided by (band B + band C). Solid symbols represent data obtained from Yang et al., open symbols represent data taken from Sosnay et al. $\text{ΔTm}$ data were obtained for a WT NBD1 construct comprising residues 387–646, excluding residues 405–436. A correlation between the extent of folding and thermal stabilization/destabilization is clearly visible.