The design of integrated radar-communication (IRC) systems is seen as a promising solution to alleviate the congestion of spectrum resource in recent years. Additionally, with the deployment of large-scaled arrays in base stations, sparse array design via antenna selection has become prominent to reduce system overheads and alleviate mutual coupling effect. In this paper, we propose a joint design of antenna selection and transmit beamformer for IRC systems to realize the dual operations of radar and communications. Specifically, we propose a joint model that achieves a trade-off between minimizing the total number of antennas for IRC and suppressing the peak sidelobe level (PSL) of overall transmit beampattern with well-controlled mainlobe ripples. Unlike the existing schemes that focus on individual beamformer design with varying sparse array configurations, we impose a group sparsity on the beamforming matrix and jointly optimize multiple beamformers to obtain a common sparse array for the constellation of all communication symbols. To solve the resulting non-convex problem, we employ the alternating direction method of multipliers to split the original problem into parallel subproblems which are then solved iteratively. Simulation results illustrate the superiority of the proposed method over the conventional approaches in obtaining a lower PSL using fewer antennas with dual functions.