Chemical reaction networks (CRNs) provide a powerful abstraction to formally represent complex biochemical processes. DNA provides a promising substrate to implement the abstract representation (or programming language) of CRNs due to its programmable nature. Prior works that used DNA to implement CRNs either used DNA-only systems or multienzyme DNA circuits. Architectures with DNA-only components had the rationale of being biologically simple systems. Multienzyme systems, on the other hand, aimed at using natural enzymes to improve circuit performance, although, at the cost of increased complexity. In this work, we explore an alternative architecture that lies along the spectrum in between DNA-only systems and multienzyme DNA systems. Our architecture relies on only a strand displacing polymerase enzyme and DNA hybridization reactions for implementing CRNs. First, we briefly introduce the theory and DNA design of simple CRNs and then explore the fundamental properties of polymerase-based strand displacement systems. Finally, we engineer a catalytic amplifier in vitro as a use-case of our framework since such amplifiers require the intricate design of DNA sequences and reaction conditions.