Multiagent systems (MASs) are seemingly a promising way to model and simulate biological systems and networks, going toward a system-level understanding of such complex systems, as promoted by Systems Biology. In this paper we propose an agent-based framework based on the notion of agents and artifacts, implemented on top of the TuCSoN agent coordination model and infrastructure. The approach promotes the engineering of distributed simulations, where biological networks are modelled as multiagent systems. There, agents — representing active biological components such as proteins and enzymes — interact by means of tuple centres — representing the environment that enables and mediates the interaction of such components.