Heat shock protein 90 (Hsp90) is emerging as an important target in cancer therapeutics. Benzoquinone ansamycins such as geldanamycin (GM, 1) are effective Hsp90 inhibitors, but their clinical development is limited by toxicity, especially liver toxicity, due to conjugation with glutathione. Thus, new non-quinone 1 analogs with improved pharmacological profiles are needed. Previously, we developed non-quinone 1 analogs (e.g., DHQ3) via mutasynthetic and directed biosynthetic methods, which showed stronger Hsp90 ATPase inhibition than 1, and obtained novel tricyclic analogs from genetically engineered Streptomyces hygroscopicus strain AC15. Presently, we describe the fermentation of AC15 (constructed by combinational mutation of the geldanamycin polyketide synthase gene gelA and post-PKS modification gene gel7), and the isolation, structural determination, and bioactivity of new non-quinone 1 analogs DHQ7 (4) and DHQ8 (5) produced by AC15. The structures of 4 and 5 were elucidated using NMR (COSY, HMQC, HMBC) and HR-ESIMS: 4 is a C-15 hydroxylated non-quinone 1 analog with an additional double bond, and 5 is a carbamoylated derivative of 4. Biological evaluations showed that 4 and 5 inhibited yeast Hsp90 ATPase activity with IC50 values of 1.75 and 5.87 μM, respectively, and competed with ATP for Hsp90 binding. Anti-proliferation assays in human ovarian (A2780) and breast (SK-Br3, BT474) cancer cell lines revealed 4 had IC50 values of 13, 15, and 8 μM, while 5 was less effective. Western blot analysis indicated that DHQ3 (3) and 4 degraded Hsp90 client proteins Her2 and c-Raf in a concentration-dependent manner, suggesting their anti-proliferative effects are mediated by Hsp90 inhibition. Although the cellular potency of these non-quinone analogs was lower than that of 1, further semi-synthetic modifications with diverse residues may improve their cellular efficacy.