In the realm of Low-Light Image Enhancement (LLIE), existing research primarily focuses on enhancing images globally. However, many applications require local LLIE, where users are allowed to illuminate specific regions using an input mask, such as creating a protagonist stage or spotlight effect. However, this task has received limited attention currently. This paper aims to systematically define the requirements of local LLIE and proposes a novel strategy to convert current existing global LLIE methods into local versions. The image space is divided into three regions: Masked Area A be enlightened to achieve the desired lighting effects; Transition Area B is a smooth transition from the enlightened area (Area A) to the unchanged region (Area C). To achieve the task of local LLIE, we introduce Region-Aware Normalization for Local Enhancement, dubbed as RANLEN. RANLEN uses a dynamically designed mask-based normalization operation, which enhances an image in a spatially varying manner, ensuring that the enhancement results are consistent with the requirements specified by the input mask. Additionally, a set of region-aware loss terms is formulated to facilitate the learning of the local LLIE framework. Our strategy can be applied to existing global LLIE networks with varying structures. Extensive experiments demonstrate that our approach can produce the desired lighting effects compared to global LLIE, all the while offering controllable local enhancement with various mask shapes.