That's a difficult question to answer! Generally, wild-type activity can always be improved since it is near the top of its fitness peak. It is still very difficult to rationally design enzyme improvements, since we know so little about how the mutations will affect enzyme architecture which will influence its activity. You can limit your search to the active site pocket, as some have done, and mutate randomly to generate your library for assessment of activity. However, there are numerous studies showing that mutating distal amino acids can have a profound effect on enzyme catalysis. Your strategy will depend on how much you known about the enzyme, your available resources, etc. There are lots of papers and resources (which can be found via Google) that can help you with the design of your strategy. A good strategy takes a lot of thought and time, so don't be tempted to rush into it.

Angela, thank you for your suggestion and explanation.

I agree with Angela.

Notice that evolution has lead to very efficient molecules, enzymes in the case of cutinase. Then in 99% of cases the mutant you prepare will lead to decrease in activity and not to an increase.

Thank you for your time and response, Prof. @Rafael.

Hi Hasan,

As Rafael and Angela have pointed out, there's little point in trying to improve on what has already evolved. However, very often we would like enzymes to function well in environments or with substrates that are unlike those originally encountered. I'd recommend starting your project by defining the conditions that you need your ideal cutinase to function in (or substrate to act on). Then determine how these differ to the natural environment of this enzyme. Your project will become more feasible and potentially more useful if you aim to improve the catalytic rate under new conditions, such as different: temperature, solvent, pH, or substrate, or the presence of denaturants or inhibitors. You can then target your literature searching to find out how others have improved enzymes for similar traits.

You're probably aware that there are many approaches available depending on what you have to work with: sequence, structure, or previous mutagenesis. This will determine whether you can dive into design, targeted random mutagenesis, or directed evolution. There are also a host of chemical modifications you can make if you only want to improve stability.

Good luck.

The relationship between activity and the residues at the active site of a protein is, as mentioned by Angela and Bradley, a very complex aspect within the realm of Quantum Chemistry (QC), and in many senses the issue is still unsolved. Although the Schrödinger's equation (SE, the equation behind QC) is the fundamental equation of physics providing the most precise theoretical predictions in physics. Its exact solution for chemical systems is only known for a very limited number of cases involving only two particles (i.e. the Hydrogen atom).

The SE can be approximated by numerical methods, but these are extremely demanding, and as the number of electrons involved grows, more simplifications are required to make calculations possible.

With proteins it is currently possible to calculate the electronic structure of a large molecule, but using a, so called "low level of theory" (semiempirical QM methods and localized molecular orbitals).

Predictions made using force-fields, which ignore QC, and relly on biased approximations of classical mechanics. These tend to fail in the details 80% of the times or more, but may provide an overall of protein Molecular Dynamics (MD). Because these do not simulate the bond formation or breaking, their use in the study of the QC of a system is very limited.

However, a carefully planned set of MD simulations may offer you insight into the protein dynamics and may uncover contacts important for the enzyme stability and/or catalysis. In turn, this information may help you to predict which residues are more likely to influence the active site behaviour. However, it is possible to do short simulations (short in terms of system's simulated time, but very long in computational wall time) , combining MD of the protein and QC of the atoms around the active site. Nevertheless, you need to start with some very specific question (restricted system), because an open question would require a computational cost too high.

For the above reasons, the experimental approaches are preferred and the best you can do is scan the literature for experiments correlating mutations in other cutinases having effect of their kinetic parameters. Because several cutinases share the catalytic triad of serine proteases, you may extend your search to tripsin, chymotripsin, elastase and many other serine-protease enzymes, there is a very extensive literature on this kind of hydrolases. Yet, that depends on the cutinase of your interest, it may not be of such kind.

You may combine the data in the literature with MD simulations to figure if a particular mutation reported for some other enzyme would have in your enzyme the same effect that was reported for the protein in the reference. That is to say you would need to compare simulations of your protein and the protein reported with and without the mutation, to see if the perturbation in the protein's MD behaviour is alike.

As mentioned by Angela, mutations affecting activity are not restricted to active site residues and as example I add a very recent reference you might find enlightening:

Saavedra HG , Wrabl JO , Anderson JA , Li J , & Hilser VJ (2018) Dynamic allostery can drive cold adaptation in enzymes. Nature 558(7709):324--328. , DOI: https://doi.org/10.1038/s41586-018-0183-2

Best wishes,

Rogelio