In multiple-input multiple-output systems, direction-of-arrival (DoA) estimation is a critical technique as the basis of precoding and beamforming. With the ever-increasing requirement of high data rate and ubiquitous devices, effective wideband DoA estimation is desirable. In this paper, we propose an iterative coherent signal-subspace method including three steps in each iteration for two-dimensional (2D) DoA estimation with a uniform circular array. The first step selects partial frequency points for the subsequent focusing process to reduce unnecessary complexity. The second step realizes the focusing process, where we provide angle intervals to generate focusing matrices with robustness, and the signal-subspaces at the selected frequency points are focused into a reference frequency. The third step estimates DoAs with the multiple signal classification algorithm where the range of the spatial spectrum is constrained by the aforementioned angle intervals, thereby further reducing the complexity. The key parameters of the proposed method in the current iteration are adjusted based on the previous estimation results. In addition, we derive the Cram\'er-Rao bound of the investigated wideband 2D DoA estimation to provide guidance for practical applications. The simulation results indicate that our proposed method enjoys better performance and implementation efficiency when compared with benchmark methods.