Domaino1s: Guiding LLM Reasoning for Explainable Answers in High-Stakes Domains