Fundus imaging plays a crucial role in diagnosing ocular diseases, where well-exposed images with rich details enhance the reliability of clinical assessments. However, achieving optimal exposure is challenging due to clinical environment limitations and variability in patient cooperation. To address this, we propose a transformer-based model, Illumination Guidance and Retinal Structural Modeling Network (IGRSM-Net), for effective fundus image enhancement. The framework simultaneously predicts illumination information and models retinal structures to generate enhanced images. Specifically, we design an Illumination Guidance Enhancement Block (IGEB) that leverages illumination representations to guide the modeling of global interactions in regions with varying exposure levels. Additionally, to enhance retinal clarity, our framework integrates structural modeling to compensate for retinal features, which are adaptively fused with the enhanced image using a multi-scale fusion strategy. Extensive experiments demonstrate that our method achieves competitive performance compared to the state-of-the-art (SOTA) approaches, particularly in improving exposure levels and preserving vascular texture details.