In the views about the aspect of existence of biological information resources, they have a client-server architecture based on simple HTTP protocol developed until now to support convenience of the use for researchers. But when taking consideration of the specificities of the research, it needs to be accessed further various types of information and to be automated. And the way of the legacy web based connection is also needed. Because of above listed, we have implemented a multi-agent system for supporting distribution, autonomy, coordination, reusability to PDB database which maintains protein structure information. Also applying web service technique to it, which is a recently come and already used in industry, we have made the processing of SOAP typed request and response message possible from web clients through the agent service-to-web service gateway. By combining two distinct featured techniques, it is possible to development of application systems about the higher leveled web resources and specially, to increase of the usefulness and the help to integrate biological information resources efficiently.