Ferrichrome inhibited the growth of six colorectal cancer cells in vitro and in vivo
To clarify the tumor-suppressive effect in colorectal cancer cells in vitro, ferrichrome was administered to Caco2/bbe, HT29, SK-CO-1, HCT116, SW480 and SW620 cells. The characteristics of each cell line are shown in Table 1. The growth of each cell line was suppressed by ferrichrome treatment (Figure 1A).
To assess the tumor-suppressive effect of ferrichrome in vivo, SW620 cells and HCT116 cells were transplanted to nude mice. The tumor volume and weight were significantly decreased by the intraperitoneal injection of 5 mg/kg ferrichrome to the SW620 tumor (Figure 1B). However, no marked change in the body weight was observed on ferrichrome treatment (Figure 1C). Likewise, a tumor-suppressive effect of the intraperitoneal injection of 5 mg/kg ferrichrome was observed for the HCT116 tumor (Figure 1D), suggesting that ferrichrome exerts anti-tumor effects against colorectal cancer in vivo.
Assessing the antitumor effect of ferrichrome in patient-derived sporadic colorectal neoplasms
To clarify the anti-tumor effect of ferrichrome on cancer cells associated with the adenoma-carcinoma sequence pathway, 9 adenoma and 9 carcinoma organoids derived from patients were constructed from endoscopic biopsied specimens. The details of the patients who underwent endoscopic biopsies are described in Table 2. The MTT/resazurin assay revealed that the growth suppression rate of each ferrichrome-treated organoid was 68.4% (A1), 52.6% (A2), 67.4% (A3), 81.6% (A4), 56.1% (A5), 59.2% (A6), 55.9% (A7), 41.6% (A8), 45.7% (A9), 68.2% (C1), 52.3% (C2), 35.0% (C3), 85.3% (C4), -7.7% (C5), 73.8% (C6), 73.0% (C7), 80.7% (C8) and 40.5% (C9) (Table 3) (The dose dependency was confirmed in A1 organoid (Supplemental figure 1)). Ferrichrome significantly inhibited the growth in seven adenoma organoids (A1, A3, A4, A5, A6, A7, A8) and seven carcinoma organoids (C1, C2, C3, C4, C6, C7, C8) but not in the four other organoids (A2, A9, C5, C9).
We previously showed that ferrichrome induced colorectal cancer apoptosis through the activation of the DDIT3 signaling [17]. The change in the DDIT3 mRNA expression in the 18 ferrichrome-treated organoids was therefore assessed. Real-time PCR revealed the ratio of DDIT3 mRNA (ferrichrome treated organoids/ferrichrome un-treated organoids) to be 1.54 (A1), 1.62 (A2), 2.80 (A3), 1.89 (A4), 1.88 (A5), 1.95 (A6), 2.03 (A7), 1.39 (A8), 1.86 (A9), 2.28 (C1), 1.80 (C2), 4.21 (C3), 1.54 (C4), 1.04 (C5), 1.50 (C6), 3.07 (C7), 6.21 (C8), and 1.63 (C9) (Table 3). The expression of DDIT3 was significantly induced in 14 organoids treated with 1 mg/mL of ferrichrome (A1, A2, A3, A4, A5, A7, A8, C1, C2, C3, C4, C7, C8, C9) (p<0.05) and tended to be induced in the remaining 1 organoid (A6) (p=0.0626), in which ferrichrome significantly inhibited the tumor growth in the MTT assay. In contrast, DDIT3 was not induced in the three organoids (A9, C5, C6) in which ferrichrome exhibited no inhibitory effect in the MTT assay. This suggests that the ferrichrome-induced tumor inhibition was mediated by the upregulation of DDIT3. To confirm whether or not ferrichrome induced apoptosis in patient-derived sporadic colorectal cancer, a TUNEL assay was performed in ferrichrome-treated organoids. The proportion of TUNEL-positive organoids was significantly higher in the ferrichrome group than in the control group (Figure 1E). These results suggest that ferrichrome induces apoptosis and thus suppresses the growth of sporadic colorectal neoplastic cells of organoids.
Ferrichrome exerted an anti-tumor effect in a colitis-associated cancer model in vivo
To investigate the anti-tumor effect of ferrichrome on colorectal cancer associated with the colitis-associated cancer pathway, an AOM-DSS carcinogenesis mouse model was constructed. Ferrichrome (5 mg/kg) was intraperitoneally administered every other day. The tumor size was significantly reduced by the administration of ferrichrome in the AOM-DSS carcinogenesis mouse model (Figure 2A). No significant change in the body weight was noted in the AOM-DSS+ferrichrome or AOM-DSS+PBS group (Figure 2B). A Western blotting analysis showed the accumulation of cleaved caspase-3 and poly(ADP-ribose)polymerase (PARP) in the AOM-DSS+ferrichrome group compared to the AOM-DSS+PBS group, suggesting that tumor cell apoptosis had been induced by ferrichrome treatment (Figure 2C).
To clarify the ferrichrome effect on the precancerous condition of colitis-associated cancer, a DSS-colitis mouse model was constructed. Ferrichrome treatment did not change the colon length or expression of proinflammatory cytokines, including TNF alpha, IFN-gamma and IL-1beta (Figure 2D, E). These data indicated that ferrichrome exerts an anti-tumor effect on colitis-associated cancer but not on the precancerous, non-neoplastic phase of the pathway.
The anti-tumor effect of ferrichrome was stronger than that of 5-FU and cisplatin
To compare the anti-tumor effect of ferrichrome and currently available anti-tumor agents, SW620 cells were treated with these agents. An SRB assay revealed that the tumor-suppressive effect of ferrichrome was superior to that of 5-FU and cisplatin (Figure 3A). To assess the antitumor effects of ferrichrome and 5-FU in vivo, a mouse xenograft model of SW620 cells was constructed, and experimental agents were directly injected into the tumor. The tumor volume and weight were significantly suppressed by the administration of either 0.5 mg / kg of ferrichrome or 5-FU in comparison to the control group, and the tumor suppressive effect of ferrichrome was superior to that of 5-FU (Figure 3B). These data indicated that ferrichrome exerted a strong tumor-suppressive effect, and its effect was superior to that of currently available anti-tumor agents, including 5-FU and cisplatin, both in vitro and in vivo.
The effect of ferrichrome in combination with 5-FU is stronger that of 5-FU alone in vivo
The prognosis of colorectal cancer patients with an insufficient response to 5-FU or cisplatin treatment is poor, so enhancing the effect of 5-FU is clinically important for improving the outcome of such patients. To assess the combination effect of ferrichrome and currently available anti-tumor agents, the mixture of ferrichrome and cisplatin or 5-FU was administered to SW620 cells. The SRB assay revealed that the anti-tumor effect of low-dose (0.2 µg/mL) ferrichrome was equal to or stronger than that of low-dose (0.2 µg/mL) 5-FU or cisplatin. The combination of ferrichrome and 5-FU highly suppressed the tumor growth compared to 5-FU alone, but the combination of ferrichrome and cisplatin showed no additional anti-tumor effect (Figure 3C, D). Ferrichrome (0.5 mg/kg) and 5-FU (5 mg/kg) were then intraperitoneally injected into SW620 transplanted nude mice. The concentration of 5-FU was determined based on the FOLFOX regimen generally used in colorectal cancer patients. The tumor volume and weight were reduced by treatment with ferrichrome or 5-FU. The tumor-suppressive effect was significantly augmented by the combination of ferrichrome and 5-FU compared to single-agent administration of either one (Figure 4A). Likewise, ferrichrome (5 mg/kg) and 5-FU (5 mg/kg) were intraperitoneally administered to the AOM-DSS carcinogenesis model mice. The average tumor size of the combination treatment group was smaller in comparison to the single treatment group (Control; 399.8±190.2, Ferrichrome; 144.5±168.9, 5-FU; 162.0±72.7, ferrichrome and 5-FU; 92.2±84.4), suggesting an additional effect of ferrichrome on colorectal cancer cells when the effect of 5-FU is insufficient (Figure 4B).
Ferrichrome is an anti-tumor molecule with few adverse effects
To assess the acute toxicity of excessive amounts of ferrichrome administration, 300 mg/kg ferrichrome, which is 60-fold the anti-tumor amount used in the AOM-DSS and tumor-transplanted models, was administered to the mice via tail vein injection. There were no dead mice on day 1 of observation (data not shown). Serum was then collected from the inferior vena cava of 300 mg/kg ferrichrome-treated mice, and biochemical tests were performed. There were no significant changes in the test values of AST, ALT or creatinine. Furthermore, we assessed the serum iron contents because ferrichrome is an Fe3+ chelator derived from bacteria [19]. However, the iron content was not markedly changed by ferrichrome treatment (Figure 5). Therefore, the acute toxic level of ferrichrome is thought to be over 300 mg/kg via intravenous administration in vivo.