Azithromycin, a potent autophagy inhibitor for cancer therapy, perturbs cytoskeletal protein dynamics

Background Autophagy plays an important role in tumour cell growth and survival and also promotes resistance to chemotherapy. Hence, autophagy has been targeted for cancer therapy. We previously reported that macrolide antibiotics including azithromycin (AZM) inhibit autophagy in various types of cancer cells in vitro. However, the underlying molecular mechanism for autophagy inhibition remains unclear. Here, we aimed to identify the molecular target of AZM for inhibiting autophagy. Methods We identified the AZM-binding proteins using AZM-conjugated magnetic nanobeads for high-throughput affinity purification. Autophagy inhibitory mechanism of AZM was analysed by confocal microscopic and transmission electron microscopic observation. The anti-tumour effect with autophagy inhibition by oral AZM administration was assessed in the xenografted mice model. Results We elucidated that keratin-18 (KRT18) and α/β-tubulin specifically bind to AZM. Treatment of the cells with AZM disrupts intracellular KRT18 dynamics, and KRT18 knockdown resulted in autophagy inhibition. Additionally, AZM treatment suppresses intracellular lysosomal trafficking along the microtubules for blocking autophagic flux. Oral AZM administration suppressed tumour growth while inhibiting autophagy in tumour tissue. Conclusions As drug-repurposing, our results indicate that AZM is a potent autophagy inhibitor for cancer treatment, which acts by directly interacting with cytoskeletal proteins and perturbing their dynamics.

Transfected cells were selected with corresponding antibiotics and single colonies were isolated.
Autophagic flux assay. To monitor autophagic flux and cell density, the pMRX-IP-GFP-LC3-RFP-LC3ΔG expression vector, a kind gift from Prof. Noboru Mizushima, was transfected stably into A549 cells. 6 The cells were treated with various concentrations of AZM, HCQ, or BafA1, and the fluorescence intensity or cell density were monitored with IncuCyte ZOOM live-cell imaging system (Essen Bioscience). GFP/RFP fluorescence intensity was used to monitor autophagic flux.
Transmission electron microscopy (TEM). A549 cells were treated with 50 μM AZM for 24 h before being prepared for TEM observation as previously described. 7 Tumor xenograft model. Animal protocols were in accordance with the Regulations and Guidelines on Scientific and Ethical Care and Use of Laboratory Animals by Science Council of Japan and were approved by the Tokyo Medical University Animal Care and Use Committee. A549 cells were injected into six-to eight-week-old male nude mice (BALB/cAJcl-nu/nu). A total of 1 x 10 6 cells were suspended in PBS, mixed with an equal volume of Matrigel (CORNING, #354234), and injected subcutaneously into the flank.
Tumor size was measured with calipers twice per week, and body weight of mice was measured every day. After the average tumor volume reached 100 mm 3 , mice were divided into two groups so that the average size of the bearing tumors was the same (not randomized), and treated with either azithromycin (100 μg/g/day) or vehicle only (0.25 % (w/v) methyl cellulose/0.1% Tween 80) via oral gavage. After three weeks of treatment, tumors were excised and either stored at -80°C for protein extraction or fixed in 10% formalin neutral buffer solution. No power calculation for sample sizes was performed.
No blinding treatment was done.
Immunohistochemical analysis. Immunohistochemical analysis of xenograft tumors was performed with anti-p62 (PROGEN, #GP62-C) antibody. Paraffin sections were deparaffinized and rehydrated. Antigen retrieval was performed by autoclaving at 121°C for 10 min in a retrieval solution (10 mM sodium citrate, pH 6.0). Subsequently, sections were blocked with 10% normal goat serum in TBST and incubated with primary antibody velocity of tracks were analyzed with TrackMate through Fiji from 10 different fields. 8,9 Statistical analysis. Statistical analysis for the FACS analysis and image analysis results in Fig. 3B performed a one-way ANOVA followed by Bonferroni's multiple comparison test. Two-sided Student's t-test was performed to compare tumor weight, tracking analysis results, DQ Red-BSA analysis, and image analysis results in Fig. 3D. A p-value of less than 0.05 was considered statistically significant. Variation across experimental groups was analyzed using F-testing. All analyses were performed with GraphPad Prism 5 software (GraphPad Software).