Multiple-player games involving cooperative and adversarial agents are a type of problems of great practical significance. In this letter, we consider an attack-defense game with a single attacker and multiple defenders. The attacker attempts to enter a protected region, while the defenders attempt to defend the same region and capture the attacker outside the region. We propose a distributed pursuit-defense strategy for the defenders' cooperative defense against the attacker. Inside a bounded, convex, two-dimensional space, the defenders choose among an area-decreasing, a distance-decreasing, or a pursuing strategy. We prove that our strategy guarantees the attacker to be captured before entering the protected region in a finite time. We also demonstrate with simulations that a human-controlled attacker is unable to enter a protected region when multiple defenders are using our pursuit-defense strategy.