<p>Plenty of researches have studied microbial diversity and interactions in various ecosystems. However, the geographic pattern of microbial interaction at a large space scale and the underlying factors triggering its changes remain unclear. In this study, we integrated the prokaryotic and micro-eukaryotic communities from Chinese coastal wetlands crossing 2200 km to disclose microbial interactions. Cross-kingdom co-occurrence networks were constructed and their topological features were analyzed. Multiple bacteria involved in sulfur metabolism were identified as important module hubs in meta-community network. At the sub-network level of each sample, sulfur-related taxa were found to be high-degree keystones, and genus <i>Thioprofundum</i> stands out consistently across all samples. Furthermore, topological features and β-diversities of sub-networks showed a distinct geographic pattern that highly associated with topological characteristics, geographic distances, and keystone taxa edaphic features. These findings were consistent with node degree latitudinal distribution. These results disclosed the generalities and specificities of network properties across coastal wetlands on a large space scale. By integrating parameters and network structures, a putative model about the possible mechanisms driving co-occurrence network dynamics was developed. This work provides new insights into the keystone soil microbes and the biogeographic patterns of soil microbial interactions in coastal wetland ecosystem.</p>