Bound water at protein-protein interfaces: partners, roles and hydrophobic bubbles as a conserved motif

Abstract

Background: There is a great interest in understanding and exploiting protein-protein associations as new routes for treating human disease. However, these associations are difficult to structurally characterize or model although the number of X-ray structures for protein-protein complexes is expanding. One feature of these complexes that has received little attention is the role of water molecules in the interfacial region.

Methodology: A data set of 4741 water molecules abstracted from 179 high-resolution (≤ 2.30 Å) X-ray crystal structures of protein-protein complexes was analyzed with a suite of modeling tools based on the HINT forcefield and hydrogen-bonding geometry. A metric termed Relevance was used to classify the general roles of the water molecules.

Results: The water molecules were found to be involved in: a) (bridging) interactions with both proteins (21%), b) favorable interactions with only one protein (53%), and c) no interactions with either protein (26%). This trend is shown to be independent of the crystallographic resolution. Interactions with residue backbones are consistent for all classes and account for 21.5% of all interactions. Interactions with polar residues are significantly more common for the first group and interactions with non-polar residues dominate the last group. Waters interacting with both proteins stabilize on average the proteins' interaction (-0.46 kcal mol(-1)), but the overall average contribution of a single water to the protein-protein interaction energy is unfavorable (+0.03 kcal mol(-1)). Analysis of the waters without favorable interactions with either protein suggests that this is a conserved phenomenon: 42% of these waters have SASA ≤ 10 Å(2) and are thus largely buried, and 69% of these are within predominantly hydrophobic environments or "hydrophobic bubbles". Such water molecules may have an important biological purpose in mediating protein-protein interactions.

Figure 1. Molecular model of human placental RNase inhibitor (hRI)- human angiogenin (hAng) complex (1a4y).

(A) Connolly-type surface representation with blue for hRI and green for hAng; (B) Interface region; water molecules colored blue are Relevant (≥ 0.25) with respect to hRI, green with respect to hAng, magenta with respect to both hRI and hAng, and white with respect to neither (see Table 2); (C) Of particular interest is the “hydrophobic bubble” enclosing the non-Relevant waters HOH59, HOH71 and HOH72. Note that these three waters are encompassed within a region of the cavity (rendered with white dots by VICE [84]) that is of hydrophobic character (green contours) as indicated by focused HINT complement maps . HOH52 and HOH74 are also in the cavity but in a polar region (magenta contours). The pocket map is set on the surface of hRI; the structure and surface for hAng has been deleted for clarity.

Figure 2. Molecular model of human TGFβ Type II receptor extracellular domain (hβIIR)-TGF β3 (hβ3) complex (1ktz).

(A) Connolly-type surface representation with blue for hβIIR and green for hβ3. (B) Interface region; water molecules colored blue are Relevant (≥ 0.25) with respect to hβIIR, green with respect to hβ3, magenta with respect to both hβIIR and hβ3, and white with respect to neither (see Table 2).

Figure 3. Relative fractions of waters with Relevance to neither (white), one (gray) and both (black) proteins for.

(A) full data set of 4741 waters from 179 protein X-ray structures of resolutions ≤ 2.3 Å; (B) reduced data set of 2605 waters from 87 structures of resolutions ≤ 1.90 Å; (C) reduced data set of 2136 waters from 92 protein structures of resolutions between 1.91 Å and 2.3 Å; and (D) 109 waters from 16 structures with resolutions between 2.4 Å and 3.5 Å.

Figure 4. Selection of Relevance threshold.

Fraction of waters Relevant to neither (white), one (gray) and both (black) proteins. Relevance was previously trained so that waters having total values 0.50 or greater with respect to all other molecules are conserved; 0.25 (blue line) is the corresponding Relevance with respect to one molecule (protein). Values less than 0.25 are thus not statistically meaningful, while the rapid decrease in the number of bridging waters for thresholds between 0.15 and 0.40 argues for a low threshold. Thus, the selected 0.25 threshold meets both criteria.

Figure 5. Histograms illustrating distribution of HINT scores for water molecules.

(A) All waters in data set; (B) water molecules with Relevance to neither protein; (C) waters with Relevance to one protein; (D) waters with Relevance to both proteins. Note that 500 HINT score units is approximately −1.0 kcal mol⁻¹.

Figure 6. Average HINT interaction scores for waters at protein-protein interfaces.

(A) Scores averaged over all water molecules for interactions with protein backbone atoms; (B) scores averaged over all water molecules for interactions with protein sidechain atoms; (C) scores normalized by weighted count of residue types (Table 4) with protein backbone atoms; and (D) scores normalized by weighted count of residue types with protein sidechain atoms.

Figure 7. Color heat maps depicting Res₁-H₂O-Res₂ interactions for water molecules found at protein-protein interfaces.

All maps are linearly scaled over the maximum range of values for that data set. (A) Total HINT score between waters and Res₁/Res₂: upper left – all waters in data set (minimum score -71,358, maximum score 114,632); upper right – waters in set with Relevance to neither protein (minimum -41,868, maximum 3,685); lower left – waters in set with Relevance to one protein (minimum -26,470, maximum 50,220); lower right – waters in set with Relevance to both proteins (minimum -3,534, maximum 60,727). (B) Weighted count of Res₁/Res₂ with water interactions: upper left – all waters in data set (minimum count 0.1, maximum count 242.7); upper right – waters in set with Relevance to neither protein (minimum 0.0, maximum 74.0); lower left – waters in set with Relevance to one protein (minimum 0.1, maximum 113.3); lower right – waters in set with Relevance to both proteins (minimum 0.0, maximum 114.5). (C) Average HINT score (normalized by weighted count) between waters and Res₁/Res₂: upper left – all waters in data set (minimum average score -601.6, maximum average score 540.5); upper right – waters in set with Relevance to neither protein (minimum -624.3, maximum 483.0); lower left – waters in set with Relevance to one protein (minimum -633.7, maximum 499.7); lower right – waters in set with Relevance to both proteins (minimum -875.1, maximum 680.9). Cells colored black represent cases where the weighted count was zero, and the HINT score normalization yields an undefined value.

Figure 8. Dendograms indicating clustering of residues with respect to average HINT score (normalized by weighted count) in Res₁-H₂O-Res₂ interactions.

(A) for all waters; (B) for waters with Relevance to neither protein; (C) for waters with Relevance to one protein; and (D) for waters with Relevance to both proteins.

Figure 9. Water as a nano-scale buffer.

(A) increasing the pH of the system is compensated by a reorientation of the bridging water molecule; (B) direct unmediated interactions are less able to compensate for changes in pH.

Figure 10. Motifs for water molecules in ring region (overlap of shaded zones).

(A) water without interactions with either protein may be stabilized in situ by other water molecules; (B) under favorable conditions water may bridge between proteins and be Relevant with respect to both.

Figure 11. Water in chain of three water molecules.

HOH2331 (red) from protein complex 1kxq is Relevant with respect to waters HOH828 and HOH2288 (blue), which are each, in turn, Relevant to the proteins in the complex.

Figure 12. Average interaction type scores for waters with Relevance to zero, one and two proteins.