"

From 2014.igem.org

Contents 1 Introduction 1.1 Michaelis Menten kinetics and Competitive Enzyme Kinetics 2 Results 3 Discussion

Introduction

Figure: Michaelis-Menten kinetics with competitive product inhibition. Reaction mechanism

Michaelis Menten kinetics and Competitive Enzyme Kinetics

Results

To analyze the effect of circularization on the thermostability of the lysozyme variants, the heat shock dependent reaction rate parameters $v_{max}$ for all lysozyme variants had to be identified. For this purpose we analyzed the observed substrate degradation dynamics for the different lysozyme variants by ODE modeling. As detailed in the introduction, the enzymatic reaction mechanism of the lambdaphage lysozyme can be described by Michaelis-Menten kinetics with product inhibition. Furthermore, experiments on pH-dependent lysozyme degradation have shown that lysozyme exists in two distinct states when challenged with pH changes: the normal, functional state and a detured, nonfunctional state (REF). We hypothesized that lysozyme deformation under heat shock conditions could be described by a similar shift from a functional conformation to a distinct, denatured state. Consequently, enzymatic activity after heat shock was assumed to be exerted by only one, homogeneous, population of functional lysozymes, differing in size depending on heat shock intensity. Because the structure of the active enzyme species was assumed to be identical independent of the applied heat shock, the kinetic parameters of the enzymatic reactions could be assumed to be independent of heat shock intensity. Therefore, based on this model of enzyme denuration, enzymatic activity after heat shock could be assumed to be only dependent on the remaining fraction of functional lysozymes.

This model was fitted to all available data, using simultaneous multi-model fitting where appropriate. The model could emulated the substrate degradation dynamics for all lysozyme variants (Fig 1). Profile likelihood-based identifiability analysis was employed to verify practical identifiability of the relevant kinetic parameters. While the kinetic parameters representing enzyme affinity for the substrate and the inhibitors could not be identified in the model, the maximal reaction rate vmax where identifiable in all cases (Fig 2). The complete result of the profile likelihood analysis can be found here.

To compare thermostability of the different lysozyme variants, we analyzed the relationship between heat shock intensity and loss of enzymatic activity. As a measure for enzymatic activity, we used the normalized maximal reaction rate (the ratio of the enzymatic activity after heat shock and the basal enzymatic activity after incubation at 37 degree Celsius). Heat-shock dependent loss of enzymatic activity differed considerably between the different lysozyme variants (Fig 3). For a direct comparison of lysozyme variant thermostability we sought a robust statistic characterizing heat-shock resistance. This statistic should incorporate the threshold heat-shock intensity upon which significant loss of activity occurs as well as the steepness of the heat-shock intensity dependent loss of activity. We decided to focus on the heat-shock intensity window where most of the enzymatic activity was lost (X degree °C to Y °C). Loss of enzymatic activity was characterized by the average enzymatic activity in this window (Fig. Y).

Figure: Introduction of heat shock dependent reaction rates does not significantly improve the model fit. It was tested whether the model fit could be improved by assuming that heat shock induced enzyme deformation occurs gradually and not in distinct stages. In this case, the kinetic parameters of the enzymatic activity are dependent on the heat shock intensity. Exemplary measurements of peptidoglycan degradation by the linear lysozyme at 37 degree Celsius are shown for the simplified model assuming heat shock independent kinetic parameters (a) and the full model with heat shock dependent kinetic parameters (b). These data illustrate the general observation that the model fit was not significantly improved (see text for details).

Finally, we tested whether the mechanistic assumption of a distinct transition between a single active and inactive state upon heat shock had affected the quality of the model fit. The alternative hypothesis concerning the mechanism of enzyme deformation would allow for continuous changes of the lysozyme structure in response to heat shock intensity. Thus, a gradual shift towards more deconformed structures would be expected for higher heat shock intensities. This would result in different kinetic parameters for the same lysozyme species under differing heat shock treatment. To test the effect of implementing this alternative deconformation mode in the model, model fitting was repeated with independent kinetic parameters for different heat shock intensities. Manual inspection of the fitting results did not show a better fit to the data. However, freeing the kinetic parameters resulted in a loss of parameter identifability. Therefore, the increased number of kinetic parameters was considered to negatively affect the usability of the model and the original, parameter-reduced, model structure was retained for analysis.

Figure: Introduction of heat shock dependent reaction rates interferes with parameter identifiability in the model. Likelihood profiles of vmax after 10 min incubation at 37 degrees Celsius and 42 degrees celsius are shown for the linear lysozyme (a) and a circularized lysozyme with the sg1 linker (b). Likelihood profiles for all parameters are documented here.

Discussion

Using dynamic ODE modeling, we could extract the heat-shock dependent maximal reaction rates of different lysozyme variants from simple substrate degradation measurements. The vmax parameters were identifiable in spite of the complex reaction mechanism of the lysozyme. This allowed us to compute a normalized enzymatic activity for all lysozyme variants after a variety of different heat shock challenges. By comparing these enzymatic activities, thermostability of the different lysozymes variants could be directly compared.

Our findings show that circularization of the lysozyme can have a considerable influence on enzyme thermostability. Similar findings have been reported for a variety of other proteins (sources). Here, we extend previous findings by demonstrating that the effect of circularization strongly depends on the chosen linker structure. Suboptimal linker design can decrease thermostability. The most evident example in the findings presented here is the sho2 linker which was chosen for testing as an example for linkers too short to bridge the natural distance between the C- and N-terminus of the lysozyme. In silico guided design of optimized linker sequences on the other hand can indeed result in increased thermostability, as demonstrated by the ord1 and ord3 linkers. These linkers where chosen as examples for linkers with a very low likelihood of crossing the active center of the enzyme. The implications of this analysis for the linker design are discussed in more detail in the documentation of the linker design software (here).