Enhancing Modality-Agnostic Representations via Meta-learning for Brain Tumor Segmentation