In Mobile Social Networks (MSNs), the single-path routing might not have enough of a chance to transmit content to the destination (i.e., low network throughput), due to limited contact opportunities. Meanwhile, the multiplepath routing improves the network throughput at the cost of higher system resource consumption (e.g., energy and storage). Therefore, there exists a trade-off between the network throughput and the system resource consumption. Moreover, we should consider user features in MSNs, i.e., some of the nodes would like to help the other nodes with the same social features (e.g., neighbors, classmates) during content transmission, regardless of their resource consumption. These nodes are called interested nodes. The remaining nodes, called uninterested nodes, will be reluctant to transmit contents to save their resources. To achieve high network throughput and control the system resource consumption of uninterested nodes, we propose a novel multiple-path twostage routing algorithm, InterestSpread, to transmit contents in the MSNs as follows. (1) In the first stage, we limit the content transmission into a relay candidate set. The contact information, bandwidth information, and social features are leveraged together to select such a set. (2) In the second stage, a classical max-flow method is used to get maximum throughput in the relay candidate set. The simulation based on real human and synthetic traces indicate that our algorithm achieves a good trade-off between throughput and the system recourse consumption.