Fake news causes an adverse effect on the regular public order and has become easier to propagate with the popularity of online social networks. The threat of fake news propagation makes it important to explore the vital nodes, which are defined as nodes with large branch sizes and hence generating a wider influence than others in this work. Previous studies about identifying vital nodes are mainly from single propagation of fake news networks, which do not consider that users may participate in different propagation networks. Here we identify vital nodes with the feature named the Ck-value that combines structural feature out-degree in a single network and multi-network user activeness. The Ck-value could reflect the branch size with a strong correlation, even at the early stage of propagation, and percolation based on Ck-value is more efficient than other indicators such as node activeness and out-degree. Thus, this research may provide a better understanding of vital nodes in the fake news propagation from topology properties, and further inspires innovative ways to identify vital nodes of fake news propagation.