Limits...
Hydrogen bond networks determine emergent mechanical and thermodynamic properties across a protein family.

Livesay DR, Huynh DH, Dallakyan S, Jacobs DJ - Chem Cent J (2008)

Bottom Line: Nevertheless, significant differences are found in molecular cooperativity correlations that can be explained by the detailed nature of the hydrogen bond network.This inference is consistent with well-known results that show allosteric response within a family generally varies significantly.Identifying the hydrogen bond network as a critical determining factor for these large variances may lead to new methods that can predict such effects.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Computer Science and Bioinformatics Research Center, University of North Carolina at Charlotte, Charlotte, NC, USA. drlivesa@uncc.edu

ABSTRACT

Background: Gram-negative bacteria use periplasmic-binding proteins (bPBP) to transport nutrients through the periplasm. Despite immense diversity within the recognized substrates, all members of the family share a common fold that includes two domains that are separated by a conserved hinge. The hinge allows the protein to cycle between open (apo) and closed (ligated) conformations. Conformational changes within the proteins depend on a complex interplay of mechanical and thermodynamic response, which is manifested as an increase in thermal stability and decrease of flexibility upon ligand binding.

Results: We use a distance constraint model (DCM) to quantify the give and take between thermodynamic stability and mechanical flexibility across the bPBP family. Quantitative stability/flexibility relationships (QSFR) are readily evaluated because the DCM links mechanical and thermodynamic properties. We have previously demonstrated that QSFR is moderately conserved across a mesophilic/thermophilic RNase H pair, whereas the observed variance indicated that different enthalpy-entropy mechanisms allow similar mechanical response at their respective melting temperatures. Our predictions of heat capacity and free energy show marked diversity across the bPBP family. While backbone flexibility metrics are mostly conserved, cooperativity correlation (long-range couplings) also demonstrate considerable amount of variation. Upon ligand removal, heat capacity, melting point, and mechanical rigidity are, as expected, lowered. Nevertheless, significant differences are found in molecular cooperativity correlations that can be explained by the detailed nature of the hydrogen bond network.

Conclusion: Non-trivial mechanical and thermodynamic variation across the family is explained by differences within the underlying H-bond networks. The mechanism is simple; variation within the H-bond networks result in altered mechanical linkage properties that directly affect intrinsic flexibility. Moreover, varying numbers of H-bonds and their strengths control the likelihood for energetic fluctuations as H-bonds break and reform, thus directly affecting thermodynamic properties. Consequently, these results demonstrate how unexpected large differences, especially within cooperativity correlation, emerge from subtle differences within the underlying H-bond network. This inference is consistent with well-known results that show allosteric response within a family generally varies significantly. Identifying the hydrogen bond network as a critical determining factor for these large variances may lead to new methods that can predict such effects.

No MeSH data available.


Related in: MedlinePlus

H-bond contact maps of ligated LAOBP (green squares) and apo LAOBP (green x's). Most differences occur between non-secondary structure H-bonds, which primarily involve sidechain interactions and within coil regions.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2533333&req=5

Figure 6: H-bond contact maps of ligated LAOBP (green squares) and apo LAOBP (green x's). Most differences occur between non-secondary structure H-bonds, which primarily involve sidechain interactions and within coil regions.

Mentions: In addition to the bPBP-ligand complex structures, LAOBP [34] and GBP [47] have been crystallized in the open (apo) forms. Differences within the apo vs. ligated LAOBP hydrogen bond networks are plotted in Fig. 6. Juxtaposed to the variability in Cp discussed above, comparison of the ligated and apo structures reveals several clear trends. First, in both cases the Tm of the apo structure is reduced compared to its ligated counterpart (Fig. 5c). This predicted down shift for Tm when LAOBP and GBP is ligated is fully consistent with experimental observations [27]. The mDCM also predicts Cpmax to be substantially lowered upon ligand binding, which arises due to the reduced likelihood of enthalpic fluctuations. This effect is shown in Fig. 7a, where Cpmax is plotted versus total H-bond energy for the four bPBP-ligand complexes. The values of the apo structures are superimposed onto the plot of the four complexes, which confirms that the observed changes within Cpmax upon ligand removal are strongly associated with the loss of H-bonds. Of course, addition/removal of the ligand is also associated with hydrophobic interactions, free ligand entropy and its chemical potential, all of which are absent within the mDCM. Nevertheless, the importance of the H-bond network as a dominate factor has been pointed out by Cooper [48], showing how the observed Cp changes between folded and unfolded protein conformations arise through fluctuations within the H-bond network. Consistent with our results, the same Cpmax reduction within the apo form of HBP is observed within the experimental DSC curves [27]. The reduction of Cpmax is accompanied by a lower energy barrier (see Fig. 5d). Interestingly, the mDCM predicts a linear relationship between Cpmax and free energy barrier height over the six cases studied (Fig. 7b). In fact, the mDCM predicts an unfolding transition with virtually no barrier for apo GBP, suggesting that unfolding/folding is a continuous transition (second order transition). It will be quite interesting to know if future experiments are consistent with this prediction.


Hydrogen bond networks determine emergent mechanical and thermodynamic properties across a protein family.

Livesay DR, Huynh DH, Dallakyan S, Jacobs DJ - Chem Cent J (2008)

H-bond contact maps of ligated LAOBP (green squares) and apo LAOBP (green x's). Most differences occur between non-secondary structure H-bonds, which primarily involve sidechain interactions and within coil regions.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2533333&req=5

Figure 6: H-bond contact maps of ligated LAOBP (green squares) and apo LAOBP (green x's). Most differences occur between non-secondary structure H-bonds, which primarily involve sidechain interactions and within coil regions.
Mentions: In addition to the bPBP-ligand complex structures, LAOBP [34] and GBP [47] have been crystallized in the open (apo) forms. Differences within the apo vs. ligated LAOBP hydrogen bond networks are plotted in Fig. 6. Juxtaposed to the variability in Cp discussed above, comparison of the ligated and apo structures reveals several clear trends. First, in both cases the Tm of the apo structure is reduced compared to its ligated counterpart (Fig. 5c). This predicted down shift for Tm when LAOBP and GBP is ligated is fully consistent with experimental observations [27]. The mDCM also predicts Cpmax to be substantially lowered upon ligand binding, which arises due to the reduced likelihood of enthalpic fluctuations. This effect is shown in Fig. 7a, where Cpmax is plotted versus total H-bond energy for the four bPBP-ligand complexes. The values of the apo structures are superimposed onto the plot of the four complexes, which confirms that the observed changes within Cpmax upon ligand removal are strongly associated with the loss of H-bonds. Of course, addition/removal of the ligand is also associated with hydrophobic interactions, free ligand entropy and its chemical potential, all of which are absent within the mDCM. Nevertheless, the importance of the H-bond network as a dominate factor has been pointed out by Cooper [48], showing how the observed Cp changes between folded and unfolded protein conformations arise through fluctuations within the H-bond network. Consistent with our results, the same Cpmax reduction within the apo form of HBP is observed within the experimental DSC curves [27]. The reduction of Cpmax is accompanied by a lower energy barrier (see Fig. 5d). Interestingly, the mDCM predicts a linear relationship between Cpmax and free energy barrier height over the six cases studied (Fig. 7b). In fact, the mDCM predicts an unfolding transition with virtually no barrier for apo GBP, suggesting that unfolding/folding is a continuous transition (second order transition). It will be quite interesting to know if future experiments are consistent with this prediction.

Bottom Line: Nevertheless, significant differences are found in molecular cooperativity correlations that can be explained by the detailed nature of the hydrogen bond network.This inference is consistent with well-known results that show allosteric response within a family generally varies significantly.Identifying the hydrogen bond network as a critical determining factor for these large variances may lead to new methods that can predict such effects.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Computer Science and Bioinformatics Research Center, University of North Carolina at Charlotte, Charlotte, NC, USA. drlivesa@uncc.edu

ABSTRACT

Background: Gram-negative bacteria use periplasmic-binding proteins (bPBP) to transport nutrients through the periplasm. Despite immense diversity within the recognized substrates, all members of the family share a common fold that includes two domains that are separated by a conserved hinge. The hinge allows the protein to cycle between open (apo) and closed (ligated) conformations. Conformational changes within the proteins depend on a complex interplay of mechanical and thermodynamic response, which is manifested as an increase in thermal stability and decrease of flexibility upon ligand binding.

Results: We use a distance constraint model (DCM) to quantify the give and take between thermodynamic stability and mechanical flexibility across the bPBP family. Quantitative stability/flexibility relationships (QSFR) are readily evaluated because the DCM links mechanical and thermodynamic properties. We have previously demonstrated that QSFR is moderately conserved across a mesophilic/thermophilic RNase H pair, whereas the observed variance indicated that different enthalpy-entropy mechanisms allow similar mechanical response at their respective melting temperatures. Our predictions of heat capacity and free energy show marked diversity across the bPBP family. While backbone flexibility metrics are mostly conserved, cooperativity correlation (long-range couplings) also demonstrate considerable amount of variation. Upon ligand removal, heat capacity, melting point, and mechanical rigidity are, as expected, lowered. Nevertheless, significant differences are found in molecular cooperativity correlations that can be explained by the detailed nature of the hydrogen bond network.

Conclusion: Non-trivial mechanical and thermodynamic variation across the family is explained by differences within the underlying H-bond networks. The mechanism is simple; variation within the H-bond networks result in altered mechanical linkage properties that directly affect intrinsic flexibility. Moreover, varying numbers of H-bonds and their strengths control the likelihood for energetic fluctuations as H-bonds break and reform, thus directly affecting thermodynamic properties. Consequently, these results demonstrate how unexpected large differences, especially within cooperativity correlation, emerge from subtle differences within the underlying H-bond network. This inference is consistent with well-known results that show allosteric response within a family generally varies significantly. Identifying the hydrogen bond network as a critical determining factor for these large variances may lead to new methods that can predict such effects.

No MeSH data available.


Related in: MedlinePlus