We argue in this paper that the symbiosis between information extraction (IE) and ontologies is in fact a necessary one. On the one hand, IE needs ontologies as part of the understanding process for extracting the relevant information. One the other hand, ontologies can only be effectively and efficiently populated if one resorts to IE techniques. We thus suggest to arrange IE and ontology learning and population (OLP) in a cyclic process in which IE, which is an ontology-driven process, also helps to extend the ontology. The ontology enriched in this way can then in turn be used in the next IE iteration. This paper is illustrated by examples taken in the biology domain, a domain in which there are critical needs for content-based exploration of the scientific literature. It takes the example of the ExtraPloDocs project, which aims at extracting gene-protein interaction information from the bibliography in genomics.