Random Boolean networks have been used widely to explore aspects of gene regulatory networks. A modified form of the model through which to systematically explore the effects of increasing the number of gene states has previously been introduced. In this paper, these discrete dynamical networks are coevolved within coupled, rugged fitness landscapes to explore their behaviour. Results suggest the general properties of the Boolean model remain with higher valued logic regardless of the update scheme or fitness sampling method. Introducing topological asymmetry in the coevolving networks is seen to alter behaviour.