Marineosins A and B are two spiroaminal-ring containing tripyrrole compounds isolated from the marine actinomycete, Streptomyces CNQ-617, and were found to possess potent and selective cytotoxic activity against leukemia and melanoma. Marineosins belong to the prodiginines class of natural products, examples of which are undecylprodiginine and streptorubin B. Unlike marineosins, prodiginines structures are characterized by the presence of fully conjugated tripyrrole nucleus linked to an alkyl chain (that lacks any oxygen). Cyclic prodiginines arise from an oxidative cyclization of the alkyl chain onto the tripyrrole, a step catalyzed by Rieske-oxygenase like enzymes such as RedG. The biosynthesis of prodiginines is directed via the red gene cluster. The unique structural differences between marineosin and other prodiginines spurred the proposal of a number of hypotheses for its biosynthesis, none of which have been experimentally tested. A red gene cluster homolog which has only one extra dehydratase-encoding gene; marA has been identified from the genomic library of Streptomyces CNQ-617, and the identified cluster was proposed to direct the biosynthesis of marineosin. In this study, the identified putative gene cluster was expressed in the heterologous host, S. venezuelae, and marineosin production in the new strain; JND2 was confirmed via LC/MS and 1H-NMR. The new engineered strain also produces a myriad of marineosin related shunt metabolites and pathway intermediates. This study hence presents the first identified gene cluster proved to direct the biosynthesis of marineosin; the mar gene cluster and proves that the cloned cluster encodes most, if not all the enzymes required to direct the biosynthesis of marineosin. Deletion of the Rieske-oxygenase encoding gene; marG (a RedG homolog) from the mar gene cluster led to the accumulation of 2-hydroxyundecylprodiginine; G410 with an m/z 410.28 and molecular formula C25H35O2N3. This data proves that MarG is not responsible for the introduction of the spiromaminal ring oxygen on the alkyl chain, but is required for catalyzing macrocyclic ring formation between C-8 and C-9 of G410. Undecylprodiginine production in marG deletion mutant was not observed which indicates that undecylprodiginine is likely not an intermediate along the pathway for marineosin biosynthesis, and indicates that the spiroaminal ring oxygen is introduced early in the pathway, possibly due to the incorporation of a 3-hydroxy-butyric acid starter unit. Deletion of the dehydratase-encoding gene; marA, from the mar gene cluster led to the accumulation of compounds JN408 and JN422 with m/z 408.26 and 422.24 and molecular formulae C25H33O2N3, and C25H31O3N3, respectively. Purification and structure elucidation of JN408 proves it to be an oxidized marineosin analog which has fully aromatic tripyrrole rings while; purification and structure elucidation of JN422 proves it to be a 9-keto-JN408 derivative. Both JN408 and JN422 compounds have a spiroaminal ring which indicates that MarA does not catalyze spiroaminal ring formation but catalyzes the reduction of pyrrole ring B of JN408 to yield marineosin. Therefore, we are proposing that MarA acts as a dehydrogenase, rather than a dehydratase. We are proposing that the intramolecular spiroaminal ring formation is catalyzed by either MarG or occurs non-enzymatically. JN422 is a shunt metabolite produced due to promiscuous activity of either MarG or an unidentified oxidase in the mar cluster, possibly MarT. From the data generated in this study, we present the first experimentally supported pathway for the biosynthesis of marineosin and the opportunity to generate novel compounds with potentially useful biological activities.