Suppose we are testing $H_0: \theta = \theta_0$ against $H_a: \theta > \theta_0$. If we use the Score test statistic $S(\theta_0) = \frac{U(\theta_0)^2}{I(\theta_0)}$, where $U$ is the score function and $I$ is Fisher information, what is its asymptotic distribution under $H_0$?