Year of Publication


Degree Name

Doctor of Philosophy (PhD)

Document Type

Doctoral Dissertation




Pharmaceutical Sciences

First Advisor

Dr. Markos Leggas


Introduction: Genetic rearrangements in Ewing sarcoma, prostate, and leukemia cells result in activation of oncogenic ETS transcription factor fusions. Mithramycin (MTM) has been identified as an inhibitor of EWS-FLI1 transcription factor, a gene fusion product responsible for oncogenesis in Ewing sarcoma. Despite preclinical success, a phase I/II clinical trial testing MTM therapy in refractory Ewing sarcoma was terminated. Liver and blood toxicities resulted in dose de-escalation and sub-therapeutic exposures. However, the promise of selectively targeting oncogenic ETS transcription factors like EWS-FLI1 prompted us to undertake the discovery of more selective, less toxic analogues of MTM. MTM is a potent inhibitor of ubiquitous SP1 transcription factor, likely inducing non-specific toxicity. In collaboration with two medicinal chemistry groups, two semi-synthetic efforts were implemented to develop novel analogues of MTM. The first effort utilized the biosynthetic product mithramycin SA (MTMSA) to modify C3-side chain. The second effort utilized an oxime linker directly formed on MTM’s C3-side chain (MTM-oxime; MTMox). Here I present the pharmacological assessment of over 75 novel MTM analogues towards selectively targeting oncogenic ETS transcription factors, like EWS-FLI1, over ubiquitous transcription factors, like SP1.

Methods: Novel MTM analogues were evaluated for selective cytotoxicity against ETS fusion-dependent cell lines. Selectively cytotoxic analogues were evaluated for inhibitory effects on several gene promoters in TC-32 reporter cell lines, a Ewing sarcoma cell line dependent on EWS-FLI1, transfected with luciferase reporter vector. Cloned reporter vectors incorporated NR0B1 (EWS-FLI1 binding), β-actin (SP1 binding) and CMV (non-specific) gene promoters. Furthermore, gene (mRNA) and protein expression changes of EWS-FLI1 and SP1, as well as regulated target genes, namely NR0B1, VEGFA and BCL-2 were evaluated with MTM analogue treatments. The MTM analogues with most selective activity in vitro were administered to mice by intravenous bolus dose for pharmacokinetic analysis. The MTM analogues with highest systemic exposure from each semi-synthetic effort, namely MTMSA-Trp-A10 and MTMox-24, were further evaluated. Metabolic stabilities in whole blood, plasma, and tumor cell matrices, and across multiple species were compared with MTM. Moreover, intrinsic hepatic clearances were estimated using mouse liver microsomes. Tumor and liver distributions were estimated in tumor bearing mice. Additionally, the effect of organic anionic transporter polypeptides (OATP) on distribution of MTM was investigated. Maximum tolerated doses were evaluated for lead MTM analogues, having both selective activities in vitro and high systemic exposure, compared to MTM. Complete blood cell counts and plasma alanine aminotransferase activity were measured to evaluate dose-dependent blood and liver toxicities, respectively. ETS fusion-dependent and non-dependent cell lines were implanted subcutaneously into immunocompromised mice for efficacy studies. Average tumor volumes and survival were tracked for mice receiving treatment, compared to MTM and vehicle treatment.

Results: Evaluation of MTM analogues from both semi-synthetic efforts revealed that conjugation of MTM C3-side chain with tryptophan (Trp) and/or phenylalanine (Phe) improved selective cytotoxicity against ETS fusion-dependent cell lines. This was highlighted by MTMSA-Trp-A2 (also refer to as MTMSA-Phe-Trp) and MTMSA-Trp-A10 (also refer to as MTMSA-Trp-Trp), with selective indices of 19.1 and 15.6, respectively, compared to MTM (1.5). Similarly, MTMox-23 (also refer to as MTMox-Phe-Trp) and MTMox-20 (also refer to as MTMox-Trp) had selectivity indices of 4.6 and 4.5, respectively. These selectively cytotoxic MTM analogues inhibited EWS-FLI1-mediated transcription 10-fold more effectively than both non-specific CMV-mediated and SP1-mediated (via β-actin promoter) transcription in TC-32 reporter cell lines. Moreover, gene (mRNA) and protein expression of EWS-FLI1 and regulated gene, NR0B1, were inhibited with MTM analogue treatment (GI50, 6-hour) in TC-32 cells. Similarly, SP1 and target genes, VEGFA and BCL-2, gene (mRNA) and protein expressions were also inhibited with MTM analogue treatment (GI50, 6-hour) in TC-32 cells.

Conjugation of Trp and/or Phe to C3-side chain of MTM increased systemic exposure in vivo. Most impressively, the addition of two Trp residues, namely MTMSA-Trp-A10 and MTMox-24 (also refer to as MTMox-Trp-Trp), resulted in systemic exposure increases of 218- and 42-fold, respectively, after intravenous (IV) bolus dose. Metabolically, tryptophan/phenylalanine conjugated MTM analogues are liable to esterase activity on carboxy-methyl functional group. Very rapid de-methylation in biological matrix was observed with MTMox-24, compared to MTMSA-Trp-A10, suggesting a regiospecific effect. However, esterase activity was limited to rodent matrices and demethylation occurred at significantly diminished rates in non-human primate and human plasma. MTM analogues were not susceptible to p450-mediated metabolism, with negligible loss in mouse liver microsome assay compared to verapamil control. MTM (1mg/kg) and MTMox-24 (6mg/kg) were detected in subcutaneously implanted (flank) LL2 tumors and liver homogenates after IV bolus dose. Interestingly, MTMSA-Trp-A10 (2mg/kg) was not. Despite a 3-fold increase in systemic exposure with rifampin oral pretreatment, an OATP inhibitor, exposure of MTM was unaffected in Oatp knockout mouse model. Exposure of MTM in liver tissue was 8.4-fold higher compared to tumor tissue with low tissue clearance. This agrees with the lack of metabolism observed in liver microsomes and may provide a mechanism for clinically observed liver toxicity.

MTMSATrp-A10 had a single maximum tolerated dose (MTD) of 0.75mg/kg, compared to 1mg/kg for MTM, administered by IV bolus. In contrast, MTM-oxime analogues (MTMox-20, -23, -24 and -25) had single maximum tolerated doses of 20 – 25mg/kg. These increased tolerances are the result of additive differences in whole blood stability, cytotoxicity and systemic exposure. At a dose of 0.75mg/kg, administered every 3 days, MTMSA-Trp-A10 did not result in an efficacious result in tumor xenograft studies. These studies remain under further investigation, but the result may indicate high plasma protein binding of MTMSA-Trp-A10 and lack of free fraction available within tumor. The most selective MTM-oxime analogue in vitro, MTMox-23, significantly inhibited TC-32 (EWS-FLI1+) tumor xenograft growth (p=0.0025, day 16, one-way ANOVA multiple comparisons test) compared to MTM (p=0.1174, day 16) and extending survival for 17 days out of 48 days on study (p=0.0003, Log Rank (Mantel-Cox) single comparison test) with treatment at MTD every 3 days, compared to vehicle. Additionally, the MTM-oxime analogue with highest systemic exposure, MTMox-24, also significantly inhibited TC-32 (EWS-FLI1+) tumor xenograft growth (p=0.0003, day 21, one-way ANOVA multiple comparisons test) compared to MTM (p=0.032, day 21) and extending survival for 12 days out of 37 days on study (p=0.0004, Log Rank (Mantel-Cox) single comparison test) with treatment, compared to vehicle.

Conclusion: These studies in whole highlight the importance of exposure (pharmacokinetics; PK), toxicity and efficacy (pharmacodynamics; PD) relationships. The cytotoxicity and high systemic exposure of MTMSA-Trp-A10 directly contributes to its lower tolerated dose. However, despite a similar tolerated dose to MTM, systemic exposure remains 163-fold higher at the MTD. High systemic exposure may be attributed to high plasma protein binding, but also reduces the exposure of free MTMSA-Trp-A10 within the tumor tissue, which drives the efficacious response. In contrast, the less cytotoxic and rapidly de-methylated MTM-oxime analogues allow for 25-fold higher tolerances in mice. This unique metabolism and clearance may prevent exposures required to induced systemic blood and liver toxicities induced by MTM. Moreover, at these highly tolerated doses, the initial systemic exposure at MTD is highest among analogues tested, which resulted in an efficacious response with MTMox-23 and MTMox-24 treatment in tumor xenograft models. It remains to be determined if these PK/PD relationships can be reproduced in additional animal models, including human, without inducing toxicity. Nonetheless, these initial studies in mice demonstrate that a more selective, more tolerated analogue of MTM has potential for clinical success in treating ETS fusion-dependent tumors.

Digital Object Identifier (DOI)

Funding Information

Funding for this research was provided by the Department of Defense grant # W81XWH-16-1-0477, The DanceBlue Foundation, and the Center for Pharmaceutical Research and Innovation.

Available for download on Monday, October 18, 2021