In Vivo Antitumor Activity of Allicin in a Pediatric Neuroblastoma Patient-derived Xenograft (PDX) Mouse Model

Abstract

Background/aim: Allicin is a small-molecule natural product found in garlic (Allium sativum). We previously showed that allicin inhibits ornithine decarboxylase (ODC) in vitro and induces apoptotic cell death in pediatric neuroblastoma (NB) cancer cell cultures. However, its potency as an anticancer agent in vivo has not been sufficiently explored.

Materials and methods: In this study, we used cell proliferation assays, immunoblotting techniques, and light microscopy to study NB tumor cell cultures and human primary neonatal skin fibroblast control cells as well as a MYCN-amplified NB patient-derived xenograft (PDX) mouse tumor model to study the efficacy of allicin in vivo.

Results: Allicin strongly inhibits NB tumor cell proliferation in a dose-dependent manner while non-cancerous human primary neonatal skin fibroblast control cells were largely unaffected. Importantly, two intra-tumoral injections of allicin over a two-week trial period significantly reduced the NB tumor burden in mice compared to controls (N=4-9 mice/group). Excised tumor tissues revealed that allicin treatment increased the cyclin-dependent kinase inhibitor p27^Kip1 protein levels, suggesting that in vivo, allicin increases p27^Kip1-mediated G₁/S cell cycle arrest.

Conclusion: Our findings warrant further preclinical development of allicin as a potential anticancer agent, especially for those types of cancers that are treatable by intra-tumoral injections, including neuroblastoma, glioblastoma, and medulloblastoma.

Keywords: Allicin; childhood cancer; in vivo antitumor activity; intra-tumoral injection; natural products; neuroblastoma; patient-derived xenograft (PDX).

Figure 1

Dose-dependent effect of allicin in neuroblastoma cell cultures. The chemical allicin (B) is a natural product extracted from garlic (A). Increasing concentrations of allicin inhibited NB cell proliferation in a dose dependent manner (C-F) with the greatest effect in SK-N-FI (C) and KELLY (F) cells. Data represents the mean±the standard deviation (S.D.) of a single experiment done in quadruplicate (N=4). *Statistically significant decrease in cell proliferation as compared to control (p=0.0001 or less).

Figure 2

Effect of allicin on human primary control cells. Human primary neonatal skin fibroblasts treated with allicin (0, 6, 12, 25 μM) for 24 h. Allicin did not affect human primary neonatal skin fibroblast control cell proliferation after 24 h of treatment (A). Light micrographs show that allicin does not induce the detachment of adherent cells and the formation of rounded cell phenotypes, a hallmark of apoptosis (B). Allicin did not induce apoptosis in human primary neonatal skin fibroblast control cells, as indicated by the lack of cleaved PARP using the immunoblotting detection method (C). Likewise, there were no changes in the cell cycle marker p27^Kip1 (C). Data represents the three independent experiments (N=16)±the standard deviation (B). Western blot images are representative of 2 (p27^Kip1) or 3 (PARP) independent experiments.

Figure 3

Antitumor efficacy of allicin in vivo. Mice harboring NB patient-derived xenograft (PDX) tumors were treated by intra-tumoral injections with 0.05, 0.2 and 0.5/0.2 mg allicin per injection when tumors were palpable (100 mm³, day 0) and seven days later (day 7). Tumors treated with allicin exhibited a dose-dependent decrease in tumor volume. Tumors treated with 0.2 mg (N=9, green line) and 0.5/0.2 mg (N=4, yellow line) allicin were statistically significantly smaller than control tumors (N=7, blue line) starting at day 7 and continued throughout the study until day 14. Treatment with 0.05 mg allicin (N=8, red line) had no effect on tumor volume. The data represents the mean tumor volume±standard error (S.E.) at each time point. *Denotes a statistically significant decrease as compared to control (p<0.05). Black arrows denote the days of intra-tumoral allicin injections.

Figure 4

Allicin treatment increases cyclin-dependent kinase inhibitor p27^Kip1 protein in NB tumors. Excised tumors were analyzed for p27^Kip1 protein levels by immunoblotting (A, C). Each number represents an individual mouse. GAPDH was used as a loading control. p27^Kip1/GAPDH signal ratios were determined using Image Studio Lite (Ver 5.2) and normalized to control (B, D). Bar graphs represent the mean relative signal±S.D. Tumors treated with 0.2 mg (A, B) and 0.5/0.2 mg (C, D) allicin had increased amounts of p27^Kip1 as compared to control. *Denotes a statistically significant increase as compared to control (p<0.05).