Limits...
Free Energy Perturbation Calculation of Relative Binding Free Energy between Broadly Neutralizing Antibodies and the gp120 Glycoprotein of HIV-1

View Article: PubMed Central - PubMed

ABSTRACT

Direct calculation of relative binding affinities between antibodies and antigens is a long-sought goal. However, despite substantial efforts, no generally applicable computational method has been described. Here, we describe a systematic free energy perturbation (FEP) protocol and calculate the binding affinities between the gp120 envelope glycoprotein of HIV-1 and three broadly neutralizing antibodies (bNAbs) of the VRC01 class. The protocol has been adapted from successful studies of small molecules to address the challenges associated with modeling protein–protein interactions. Specifically, we built homology models of the three antibody–gp120 complexes, extended the sampling times for large bulky residues, incorporated the modeling of glycans on the surface of gp120, and utilized continuum solvent-based loop prediction protocols to improve sampling. We present three experimental surface plasmon resonance data sets, in which antibody residues in the antibody/gp120 interface were systematically mutated to alanine. The RMS error in the large set (55 total cases) of FEP tests as compared to these experiments, 0.68 kcal/mol, is near experimental accuracy, and it compares favorably with the results obtained from a simpler, empirical methodology. The correlation coefficient for the combined data set including residues with glycan contacts, R2 = 0.49, should be sufficient to guide the choice of residues for antibody optimization projects, assuming that this level of accuracy can be realized in prospective prediction. More generally, these results are encouraging with regard to the possibility of using an FEP approach to calculate the magnitude of protein–protein binding affinities.

No MeSH data available.


Experimental versus FEP/REST relative binding affinity values for alanine scan cases showing the combined data set from VRC01 (circles), VRC03 (squares), and VRC-PG04 (triangles). Correlation values with the largest experimental ΔΔG value excluded are given in parentheses.
© Copyright Policy - CC BY
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC5383735&req=5

f0030: Experimental versus FEP/REST relative binding affinity values for alanine scan cases showing the combined data set from VRC01 (circles), VRC03 (squares), and VRC-PG04 (triangles). Correlation values with the largest experimental ΔΔG value excluded are given in parentheses.

Mentions: Figure 6 presents the combined data for all three antibodies compared to the experiment. Table 4 summarizes separately the RMSE and correlation results for all three bNAbs and for the full data set with and without glycan contacts. VRC03 contained the largest computed outliers as compared to the experiment in the set, which may have been the result of the particularly strong chain of contacts between a glycan and the antibody light and heavy chain. This may have had some effect on other residues on the heavy chain not in a direct sequence of contact. The RMSE of the experimental results between re-measurements of the same system is estimated to be around 0.45 kcal/mol for a typical case (see the Supplementary Data Table S2 for details). The overall uncertainty range in our predictions with this protocol is comparable to the experimental levels, as is confirmed by the analysis of multiple independent FEP simulations, which gives an error estimate (full width) of about 0.51 kcal/mol (see Supplementary Data Table S7 for details). With the random error in the computed values and measurements both roughly 0.5 kcal/mol, the overall RMSE due to both theory and experiment would be expected to be around 0.7 kcal/mol, which is roughly what is found in the data presented here. While individual cases likely still exhibit systematic errors associated with the sampling or protein force field, the magnitude of the overall RMSE for the three data sets, 0.70 kcal/mol, suggests that it will be increasingly difficult to distinguish experimental from computational error unless the experimental error can be reduced.


Free Energy Perturbation Calculation of Relative Binding Free Energy between Broadly Neutralizing Antibodies and the gp120 Glycoprotein of HIV-1
Experimental versus FEP/REST relative binding affinity values for alanine scan cases showing the combined data set from VRC01 (circles), VRC03 (squares), and VRC-PG04 (triangles). Correlation values with the largest experimental ΔΔG value excluded are given in parentheses.
© Copyright Policy - CC BY
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC5383735&req=5

f0030: Experimental versus FEP/REST relative binding affinity values for alanine scan cases showing the combined data set from VRC01 (circles), VRC03 (squares), and VRC-PG04 (triangles). Correlation values with the largest experimental ΔΔG value excluded are given in parentheses.
Mentions: Figure 6 presents the combined data for all three antibodies compared to the experiment. Table 4 summarizes separately the RMSE and correlation results for all three bNAbs and for the full data set with and without glycan contacts. VRC03 contained the largest computed outliers as compared to the experiment in the set, which may have been the result of the particularly strong chain of contacts between a glycan and the antibody light and heavy chain. This may have had some effect on other residues on the heavy chain not in a direct sequence of contact. The RMSE of the experimental results between re-measurements of the same system is estimated to be around 0.45 kcal/mol for a typical case (see the Supplementary Data Table S2 for details). The overall uncertainty range in our predictions with this protocol is comparable to the experimental levels, as is confirmed by the analysis of multiple independent FEP simulations, which gives an error estimate (full width) of about 0.51 kcal/mol (see Supplementary Data Table S7 for details). With the random error in the computed values and measurements both roughly 0.5 kcal/mol, the overall RMSE due to both theory and experiment would be expected to be around 0.7 kcal/mol, which is roughly what is found in the data presented here. While individual cases likely still exhibit systematic errors associated with the sampling or protein force field, the magnitude of the overall RMSE for the three data sets, 0.70 kcal/mol, suggests that it will be increasingly difficult to distinguish experimental from computational error unless the experimental error can be reduced.

View Article: PubMed Central - PubMed

ABSTRACT

Direct calculation of relative binding affinities between antibodies and antigens is a long-sought goal. However, despite substantial efforts, no generally applicable computational method has been described. Here, we describe a systematic free energy perturbation (FEP) protocol and calculate the binding affinities between the gp120 envelope glycoprotein of HIV-1 and three broadly neutralizing antibodies (bNAbs) of the VRC01 class. The protocol has been adapted from successful studies of small molecules to address the challenges associated with modeling protein–protein interactions. Specifically, we built homology models of the three antibody–gp120 complexes, extended the sampling times for large bulky residues, incorporated the modeling of glycans on the surface of gp120, and utilized continuum solvent-based loop prediction protocols to improve sampling. We present three experimental surface plasmon resonance data sets, in which antibody residues in the antibody/gp120 interface were systematically mutated to alanine. The RMS error in the large set (55 total cases) of FEP tests as compared to these experiments, 0.68 kcal/mol, is near experimental accuracy, and it compares favorably with the results obtained from a simpler, empirical methodology. The correlation coefficient for the combined data set including residues with glycan contacts, R2 = 0.49, should be sufficient to guide the choice of residues for antibody optimization projects, assuming that this level of accuracy can be realized in prospective prediction. More generally, these results are encouraging with regard to the possibility of using an FEP approach to calculate the magnitude of protein–protein binding affinities.

No MeSH data available.