Bistability is the existence of two stable steady states consistent with the same operating conditions. The mechanism for bistability is generally presumed to lie in feedback. However, as suggested by mathematical analysis, bistability can be found in certain enzyme reactions without any apparent feedback. Simulations indicated that the human dihydrofolate reductase (DHFR) catalysis with an ordinary mechanism without apparent feedback has the capacity for bistability in certain open-reactor configurations.
A primary objective of this thesis is to explore the existence of bistability in the human DHFR reaction in simulations. It showed that the human DHFR reaction exhibited bistability in the context of a partially open continuous flow stirred tank reactor (CFSTR) in which DHFR was trapped within a membrane compartment. Simulations on the dynamics of the human DHFR reaction when coupled with the other two enzyme reactions in the thymidylate synthesis pathway in which DHFR works indicated that the bistability of DHFR reaction was propagated along the pathway.
However, it was observed in exploratory experiments that the membranes used in a laboratory realization of the same partially open CFSTR hindered the mass transfer of the substrate and product molecules. Additional simulations were performed with the human DHFR reaction in a partially open CFSTR considering the hindrance effect. In this case, the human DHFR reaction had both the capacity for bistability and periodic oscillations.
As an alternative to the partially open CFSTR, a classical fully open CFSTR was considered. The simulation indicated that the human DHFR reaction in a fully open CFSTR also had the capacity for bistability. In the same fully open CFSTR, and in the presence of an extra feed stream of methotrexate, which is an inhibitor of DHFR, the human DHFR reaction exhibited bistability with respect to the changes in methotrexate concentration.
Another objective of this thesis is to lay the foundation for experimental verification of the existence of bistability in the human DHFR reaction in a partially open CFSTR, a classical fully open CFSTR, and a partially open CFSTR with DHFR tethered to chitin beads.
Preliminary experiments with a partially open CFSTR encountered many challenges mainly due to the usage of small molecular weight cutoff membranes. It turned out to be difficult for the DHFR-in-solution reaction to reach a steady state within the partially open CFSTR.
Genetically modified human DHFR with a chitin binding domain, was successfully immobilized onto chitin beads. However, the rate constants of immobilized human DHFR might deviate from those of free human DHFR. Therefore, kinetics studies of the immobilized human DHFR should be considered in preparation for any flow experiments with DHFR tethered to chitin beads in the future.
Experiments showed that the human DHFR reaction in a fully open CFSTR could reach a steady state quickly. However, there remain concerns about whether the two distinct stable steady states found in simulations would be sufficiently distinct as to be convincingly detected with currently available instruments.