Rosamines targeting the cancer oxidative phosphorylation pathway

Abstract

Reprogramming of energy metabolism is pivotal to cancer, so mitochondria are potential targets for anticancer therapy. A prior study has demonstrated the anti-proliferative activity of a new class of mitochondria-targeting rosamines. This present study describes in vitro cytotoxicity of second-generation rosamine analogs, their mode of action, and their in vivo efficacies in a tumor allografted mouse model. Here, we showed that these compounds exhibited potent cytotoxicity (average IC50<0.5 µM), inhibited Complex II and ATP synthase activities of the mitochondrial oxidative phosphorylation pathway and induced loss of mitochondrial transmembrane potential. A NCI-60 cell lines screen further indicated that rosamine analogs 4 and 5 exhibited potent antiproliferative effects with Log10GI50 = -7 (GI50 = 0.1 µM) and were more effective against a colorectal cancer sub-panel than other cell lines. Preliminary in vivo studies on 4T1 murine breast cancer-bearing female BALB/c mice indicated that treatment with analog 5 in a single dosing of 5 mg/kg or a schedule dosing of 3 mg/kg once every 2 days for 6 times (q2d×6) exhibited only minimal induction of tumor growth delay. Our results suggest that rosamine analogs may be further developed as mitochondrial targeting agents. Without a doubt proper strategies need to be devised to enhance tumor uptake of rosamines, i.e. by integration to carrier molecules for better therapeutic outcome.

Figure 1. Structural variations of rosamine.

(A) of the rosamine dyes variously functionalized at the meso position as previously reported by Lim et al. (Anticancer Drugs 2009, 20: 461–468), the ones shown here with meso- thiofuran or 4-iodophenyl had superior anticancer activities in cellular assays. (B) Second-generation targets featured in this work.

Figure 2. Schematic illustration of the synthesis of rosamines.

(A) The starting material of xanthone ditriflate was prepared in solution by triflation of the phenols, followed by animation of the triflate with piperidine to give symmetrical cyclic amines substitution or; (B) by stepwise addition of piperidine and morpholine to give unsymmetrical cyclic amines substitution.

Figure 3. NCI-60 cell lines screen.

GI₅₀ (50% growth inhibition) mean graphs showing the activity patterns of 4, 5 and methyl violet (NSC271967) in the NCI-60 cell line screens. Both the rosamines exhibited potent antiproliferative effects with Log₁₀GI₅₀ = −7 (GI₅₀ = 0.1 µM) and were particularly more effective against colorectal cancer panel. COMPARE analyses indicated 4 and 5 have similar pattern of activity as methyl violet with Pearson correlation coefficient values of 0.767 and 0.72, respectively.

Figure 4. Loss of mitochondrial transmembrane potential.

Representative event of mitochondrial transmembrane potential loss in HSC-2 cells treated with 2 and 5 at 0.1 µM. Following 8 h of treatment with 2 and 5, the percentage of cell population with mitochondrial transmembrane potential loss increased to 19% and 34%, respectively. The percentage depolarized cell of untreated control at 8 h remains at 8%, while for positive control, cells treated with 5 µM of carbonyl cyanide 3-chlorophenylhydrazone (CCCP) for 5 min results in 70% depolarized cell population. *Difference with P-value<0.05 compared to control at 0 h.

Figure 5. Inhibition of mitochondrial oxidative phosphorylation complexes.

The dose-response inhibition of mitochondrial oxidative phosphorylation Complex 1 (A), Complex II (B), Complex IV (C) and ATP synthase (D) activities by rosamine 2 (solid line) and 5 (dotted line). The activity of Complex II was partially inhibited by 5 with IC₅₀ value of 9.6±0.1 µM whereas for 2, inhibition was observed but with undetermined IC₅₀ value. Both 2 and 5 also inhibited the ATP synthase activities with IC₅₀ values of 3.9±0.3 and 3.0±0.8 µM respectively. The activity of Complex I and Complex IV were not affected by the rosamines at the treated concentrations (highest at 10 µM). IC₅₀ values depict concentration that inhibits the complexes activity by 50%. ND - indicate non-determined IC₅₀ values based on the concentration used.

Figure 6. In vivo antitumor effects of rosamine.

(A) The relative tumor volume (RTV) – time profile of 4T1 murine breast carcinoma in Balb/C mice following intravenous dosing of rosamine 5 or saline as vehicle control. Each point represents median ±95% confidence interval of RTV to staging day (n = 8). The terminal % T/C value for mice receiving 5 mg/kg and 3 mg/kg (q2d×6) of 5 were 72% and 66% respectively. The two doubling tumor growth delay (T-C)/C (dotted line, RTV = 4) for mice receiving a single bolus of 5 mg/kg and multiple doses of 3 mg/kg (q2d×6) of 5 were 22% and 38%, respectively. T and C refer to RTV for treatment and control groups, respectively. *Difference with P-value<0.05 compared to control animal. (B) Percentage of mean body weight of mice received 5 mg/kg or 3 mg/kg (q2d×6) of 5 compared to untreated mice. Body weight loss was observed in treatment groups but none of these mice experienced weight loss of more than 15%.