Abstract
The work performed in this thesis is part of a larger project (“Computational design of stable enzymes”) involving several research teams, which aimed to improve PROPKA (http://propka.ki.ku.dk) and to provide the scientific community with a computational protocol and associated PROPKA program, which could be used for predicting mutations with expectation of increased thermostability at a certain pH value or a shifted pH activity optimum.
The ability of a Bacillus circulans xylanase (BCX) mutant (N35D/A115E) to induce a decrease in pH activity optimum was evaluated by a pH dependent xylanase activity study and compared to wild type (WT). BCX N35D/A115E was created by site-directed mutagenesis followed by expression and purification. The kinetic characterization for BCX WT and N35D/A115E protein were determined as a function of pH using o-nitrophenyl-β-D-xylobioside as a substrate. The N35D/A115E substitutions caused a 1.4 pH unit decrease in the pH activity optimum but the overall activity of the mutant was reduced by approximately 75 % compared to WT protein at their respective pH optima.
Additionally BCX was used as a model for evaluating PROPKA. Six designed mutants were generated in an attempt to create ionic interactions predicted to increase the thermostability of BCX. The melting temperature (Tm) for each variant was investigated by circular dichroism and differential scanning calorimetry and compared to that of WT. Effects of the mutations ranged from no change in stability to a 6 °C decrease in stability. The structures of the six mutants were solved by high-resolution X-ray crystallography using molecular replacement.
The six BCX mutants crystallized in different crystallization conditions and from conditions that are different from previously published crystallization conditions for BCX WT. Five different crystal forms were obtained for the six BCX variants. P47Δ/F48R/N151D, F48R/N151D, V57D/A59R and N25R/N181D all crystallized in a P21 space group. Although belonging to the same space group, the size of the unit cell for F48R/N151D differed from the three other, and despite unit cell similarity the V57D/A59R variant is not isomorphous with P47Δ/F48R/N151D and N25R/N181D. The structure of Y94K/Q133R and Y94K/T97E/Q133R was solved in a P3121 and P1 space group, respectively. The five crystal forms obtained in this study have previously not been observed for BCX. The structures of the six mutants were found to be very similar to the structure for the WT. However, backbone deviations were observed at some locations and in the backbone near a few of the mutation sites. Crystal packing analysis and B-factor analysis indicated that the observed backbone deviations most likely were due to flexible regions or the effects of crystal packing contacts, but in the case of BCX P47Δ/F48R/N151D the difference seen near P47Δ/F48R was most likely due to the deletion of proline.
In the BCX F48R/N151D structure the predicted salt bridge was not formed. A double conformation of Arg48 was observed in the two BCX P47Δ/F48R/N151D molecules found in the asymmetric unit. The predicted salt bridge was formed with one of the conformations in one of the XVII molecules. In BCX V57D/A59R and N25R/N181D the predicted salt bridge was formed in only one of the two molecules found in the asymmetric unit. In BCX Y94K/Q133R and BCX Y94K/T97E/Q133R introduced side chains were supposed to be involved in an ionic bonding network. Salt bridge interactions were observed but they were found to be somewhat different from the xylanase structure the mutants were inspired from possibly due to deviations of side chain conformations and differences in the sequence.
All the BCX mutants except for BCX Y94K/Q133R were shown to have one or more of introduced residues that were in contact with residues in symmetry equivalent molecules, thus possibly affecting their side chain conformations in the crystal structure. Therefore is not known to what extent the observed side chain orientations are present in solution. Nevertheless, it does not change the facts that none of mutants were stabilized.
The structures suggested that the BCX variants most likely were destabilized due to loss of interactions on removal of WT side chains. Moreover, introduced side chains and their intended partners were observed to be quite mobile in each of the mutants, thus indicating that they do not tend to come together to form predicted structurally localized salt bridges and therefore they might not contribute significantly to a greater protein stability, or to the contrary they might possibly have contributed negatively to the stability.
The ability of a Bacillus circulans xylanase (BCX) mutant (N35D/A115E) to induce a decrease in pH activity optimum was evaluated by a pH dependent xylanase activity study and compared to wild type (WT). BCX N35D/A115E was created by site-directed mutagenesis followed by expression and purification. The kinetic characterization for BCX WT and N35D/A115E protein were determined as a function of pH using o-nitrophenyl-β-D-xylobioside as a substrate. The N35D/A115E substitutions caused a 1.4 pH unit decrease in the pH activity optimum but the overall activity of the mutant was reduced by approximately 75 % compared to WT protein at their respective pH optima.
Additionally BCX was used as a model for evaluating PROPKA. Six designed mutants were generated in an attempt to create ionic interactions predicted to increase the thermostability of BCX. The melting temperature (Tm) for each variant was investigated by circular dichroism and differential scanning calorimetry and compared to that of WT. Effects of the mutations ranged from no change in stability to a 6 °C decrease in stability. The structures of the six mutants were solved by high-resolution X-ray crystallography using molecular replacement.
The six BCX mutants crystallized in different crystallization conditions and from conditions that are different from previously published crystallization conditions for BCX WT. Five different crystal forms were obtained for the six BCX variants. P47Δ/F48R/N151D, F48R/N151D, V57D/A59R and N25R/N181D all crystallized in a P21 space group. Although belonging to the same space group, the size of the unit cell for F48R/N151D differed from the three other, and despite unit cell similarity the V57D/A59R variant is not isomorphous with P47Δ/F48R/N151D and N25R/N181D. The structure of Y94K/Q133R and Y94K/T97E/Q133R was solved in a P3121 and P1 space group, respectively. The five crystal forms obtained in this study have previously not been observed for BCX. The structures of the six mutants were found to be very similar to the structure for the WT. However, backbone deviations were observed at some locations and in the backbone near a few of the mutation sites. Crystal packing analysis and B-factor analysis indicated that the observed backbone deviations most likely were due to flexible regions or the effects of crystal packing contacts, but in the case of BCX P47Δ/F48R/N151D the difference seen near P47Δ/F48R was most likely due to the deletion of proline.
In the BCX F48R/N151D structure the predicted salt bridge was not formed. A double conformation of Arg48 was observed in the two BCX P47Δ/F48R/N151D molecules found in the asymmetric unit. The predicted salt bridge was formed with one of the conformations in one of the XVII molecules. In BCX V57D/A59R and N25R/N181D the predicted salt bridge was formed in only one of the two molecules found in the asymmetric unit. In BCX Y94K/Q133R and BCX Y94K/T97E/Q133R introduced side chains were supposed to be involved in an ionic bonding network. Salt bridge interactions were observed but they were found to be somewhat different from the xylanase structure the mutants were inspired from possibly due to deviations of side chain conformations and differences in the sequence.
All the BCX mutants except for BCX Y94K/Q133R were shown to have one or more of introduced residues that were in contact with residues in symmetry equivalent molecules, thus possibly affecting their side chain conformations in the crystal structure. Therefore is not known to what extent the observed side chain orientations are present in solution. Nevertheless, it does not change the facts that none of mutants were stabilized.
The structures suggested that the BCX variants most likely were destabilized due to loss of interactions on removal of WT side chains. Moreover, introduced side chains and their intended partners were observed to be quite mobile in each of the mutants, thus indicating that they do not tend to come together to form predicted structurally localized salt bridges and therefore they might not contribute significantly to a greater protein stability, or to the contrary they might possibly have contributed negatively to the stability.
| Original language | English |
|---|
| Publisher | Department of Chemistry, Faculty of Science, University of Copenhagen |
|---|---|
| Number of pages | 182 |
| Publication status | Published - 2013 |