In the Industrial Internet of Things (IIoT), the integration of Time-Sensitive Networking (TSN) and 5G addresses the demand for deterministic transmission in industrial applications with flexible deployment. In TSN, bandwidth resources are reserved for flow transmission within short time slots, whereas 5G allocates bandwidth across multiple channels with varying quality of service. This discrepancy in resource management hinders 5G from achieving the fine-grained allocation seen in TSN, leading to transmission bottlenecks and limiting seamless integration. To resolve this issue, we propose a fine-grained resource allocation strategy for integrated TSN-5G networks that enables efficient collaboration between TSN and 5G. First, we introduce a TSN-5G resource allocation framework aimed at precise channel and spectrum resource allocation for flows. Next, to address the transmission bottleneck of fine-grained resource allocation in 5G, we divide the flows and transmit them in parallel across multiple channels. Finally, we develop a Proximal Policy Optimization (PPO)-based fine-grained resource allocation algorithm (PFRA) to make resource allocation decisions. Experimental results show that, compared to other algorithms, PFRA significantly improves resource utilization and increases the flow scheduling success rate.