This work is concerned with the optimization aspects of networked distributed parameter systems. It is assumed that the information exchange between the networked systems, each of which is governed by an evolution equation in an abstract space, is a priori given and the design objective is to choose the leader and the synchronization controllers so that all the followers track the leader in an appropriate norm, and that all networked system agree with each other. The optimization problem is formulated as a minimization problem of a quadratic index penalizing the distance from synchronization and the tracking errors. The optimal value of the quadratic index, as it is parameterized by the synchronization gains, is expressed in terms of the solution to an associated operator Lyapunov equation. Adding another level of optimization, one can optimally select the leader and the associated optimal synchronization gains. Such a leader selection reduces to zeroing a row of an associated gain matrix for the aggregate system. Numerical studies of four networked parabolic PDEs are presented to provide additional insight on the effects of optimal leader selection and optimal gain selection on the synchronization controller performance and the collective behavior of the networked PDE systems.