In this paper, we jointly design antenna beamforming and power allocation to maximize the sum rate for multiple users in a hierarchical cognitive radio (CR) system, where an underlying microcellular system adopts dynamic spectrum access techniques to reuse the same spectrum of a macrocellular system. One challenge in this kind of hierarchical CR systems is the interference management between the macrocell and the microcell. We suggest utilizing the antenna beamforming technique to overcome the interference between the macrocell and the microcell, while maximizing the achievable sum rate of the underlying CR microcellular system. We formulate an optimization problem to achieve the objective of maximizing sum rate subject to the constraints of the maximum allowable interference power to the primary system, the minimum required signal-to-interference plus noise power ratio (SINR) for the secondary users, and the maximum transmission power for the underlying CR base station (BS). An iterative rate maximization (IRM) algorithm is developed to find the optimal solution. This developed methodology provides an important insight into the design of an optimal hierarchical CR system for various numbers of users as well as cell coverage.