Temozolomide-Perillyl alcohol conjugate impairs Mitophagy flux by inducing lysosomal dysfunction in non-small cell lung Cancer cells and sensitizes them to irradiation

Human non-small cell lung cancer (NSCLC)-derived cell lines A549, SPC-A1, NCI-H460 and NCI-H520 were purchased from American Type Culture Collection (Manassas, VA, USA) and China Center for Type Culture Collection (Wuhan, China). All these cells were grown in Dulbecco’s modified Eagle’s medium (DMEM, Gibco, Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (Gibco, Invitrogen) and antibiotics (penicillin/streptomycin, 100 U/ml) at 37 °C in 5% CO₂.

TMZ-POH and Perillyl alcohol (POH) were provided by Neonc Technologies, Inc. (Los Angeles, USA) and diluted with DMSO to make stock solutions of 100 mM. Temozolomide (TMZ), 3-Methyladenine (3-MA), baflomycin A1 (Baf.A1), carbonyl cyanide m-chlorophenylhydrazone (CCCP), Nicotinamide (NAM), mevalonolactone (MVL), geranylgeraniol (GGOH), catalase (CAT) and N-acetyl-L-cysteine (NAC), Earle’s Balanced Salt Solution (EBSS) (Sigma-Aldrich, Shanghai, China) were dissolved in DMSO or deionized water dependently; In all cases of cell treatment, the final DMSO concentration in the culture medium never exceeded 0.5%. Stock solutions of all drugs were stored at − 20 °C.

Recombinant adenoviral vector carrying the human mRFP-GFP-LC3 gene was purchased from HanBio (Wuhan, China). Cells were plated in 12-well plates at a density of 1 × 10⁴ cells per well. Cells were infected at an MOI of 2 with GFP-mRFP-LC3 gene for 24 h. After washing with PBS twice, cells were treated with TMZ-POH for another 48 respectively.

For non-selective autophagy induction, cells were washed 3 times with pre-warmed PBS and then incubated with EBSS medium (Sigma-Aldrich) at 37 °C in 5% CO₂ for 2 h. For mitophagy induction, cells were incubated with CCCP (10 μM) or NAM (5 mM) for 48 h respectively. For autophagy inhibition, cells were treated with 3-MA (1 mM), or Baf.A1 (10 nM) for 48 h respectively.

Cells following the above treatment were loaded different fluorescence probes including Mito-Tracker Green (MTG) and Lyso-Tracker Red (LTR) (Beyotime, Beijing, China) for the indicated time as described above. After washing 3 times with PBS, the florescence intensities were measured by a FACS Calibur instrument (Becton Dickinson, USA) and the data were analyzed using FlowJo Software 7.6 (Treestar, Inc., CA).

For immunostaining, cells were fixed with 4% paraformaldehyde and permeabilized with 0.1% Triton X-100 for 15 min. After incubation for 1 h with the following primary antibodies: antibodies against human anti-LC3B, anti-SQSTM1, anti-COX-IV, anti-TOM20 (CTS, Cell Signaling Technology, Danvers, MA, USA), anti-EEA1, anti-Parkin (Abcam, Shanghai, China), anti-LAMP1 (Santa Cruz, California, USA) and washing with PBS, cells were incubated for 1 h with Alexa 488-conjugated (1:1000) or Alexa 555-conjugated (1:500) (Abcam) secondary antibodies, washed with PBS. Nuclei were stained by 4′, 6-diamidino-2-phenylindole (DAPI) (Beyotime) for 3 min. Microscopy was done on a confocal laser microscopy (LSM780, Carle Zeiss, Germany) or DeltaVision microscopy (GE Healthecare Life Science, USA).

For quantification of the number of autophagosomes (diameters 0.3–1.0 μM) and SQSTM1 positive dots (diameters 0.3–1.0 μM) and EEA1 positive dots (diameters 0.1–1.0 μM), at least five cells were randomly chosen, all eligible puncta were recorded and analyzed using Fiji ImageJ software [19]. Quantification of GFP and mRFP fluorescence intensity, and colocalization between two different signals were recorded and analyzed using Fiji ImageJ software.

Apoptosis was evaluated by using the Annexin V-FITC Apoptosis Detection Kit (BD Biosciences Pharmingen, San Diego, USA) according to the description provided by the manufacturer. 1.5 × 10⁵ cancer cells grown in six well plates overnight were exposed to indicated drug treatment or 10 Gy irradiation (X-RAD 225, Radsource, Buford, USA) for indicated time, and then the cells were trypsinized, collected and stained with FITC-Annexin V & Propidum Iodide (PI) for 15 min in the dark. The stained cell population were determined using by a FACS Calibur instrument (Becton Dickinson) and the data were analyzed using FlowJo Software 7.6 (Treestar).

Cells were fixed in TEM stationary solution (2.5% glutaraldehyde in 0.2 M HEPES, G1102, Wuhan servicebio technology) at 4 °C for 4 h, rinsed in PBS, and then embedded in 4% agarose. After fixation in 1% osmium tetroxide for 2 h, the specimens were dehydrated using alcohol and embedded in polybed 812 resin (90529–77-4, SPI). After polymerization at 60 °C for 48 h, ultrathin sections were prepared with the Leica Ultracutuct slicer (Leica EM UC6, Germany), stained with uranyl acetate and lead citrate, and analyzed using TEM (HT7700, HITACHI). Count, measure and analysis on TEM picture were carried out using Fiji ImageJ software.

Mitochondrial and cytoplasmic proteins were collected using a cell Mitochondria Isolation kit (Beyotime) in accordance with the manufacturer’s instructions. Briefly, cells were harvested and washed twice with ice-cold PBS, incubated in Lysis Buffer, and then transferred to glass homogenizers and homogenized in the ice for 30–40 times, centrifuged at 1200×g for 5 min to remove any nuclei, membrane fragments and unbroken cells, and the supernatant was further centrifuged at 15,000×g for 10 min at 4 °C. The resulting supernatant contained the cytoplasmic fraction and the pellet contained the mitochondrial fraction. The mitochondrial pellet was further resuspended in a mitochondrial lysis buffer at 4 °C.

Fifty μg quantity of protein was separated on SDS-PAGE and transferred onto PVDF membranes (Millipore, Billerica, MA, USA). Membranes were then blocked with 5% evaporated skimmed milk (Bio-rad, USA) in Tris-buffered saline (50 mM Tris-HCl, pH 7.5, 150 mM NaCl) containing 0.1% Tween-20 for 1 h, and probed overnight at 4 °C with the following primary antibodies: antibodies against human LC3B, SQSTM1/P62, HSP60 (1:1000; Cell Signaling Technology, CST), antibody against RAB7A, RILP, EEA1, LAMP2, ACTB (1:1000, Proteintech, China), antibody against LAMP1 (1:500; Santa Cruz, USA), followed by incubation with horseradish peroxidase coupled secondary anti-mouse or anti-rabbit antibodies (Proteintech) for 1 h at room temperature. The protein bands were visualized using ECL blotting detection reagents (Bio-rad, USA), and developed and fixed onto x-ray films. ACTB was served as a loading control.

RAB7A activity was determined using a RAB7A activity assay kit (NewEast Biosciences, King of Prussia, PA) according to manufacturer recommended protocol. Briefly, lysates containing equal amounts of total proteins were incubated with a mouse monoclonal antibody recognizing GTP bound RAB7A specifically. The bound active RAB7A was pulled down by protein A/G agarose and detected by a rabbit polyclonal anti-RAB7A antibody.

Statistical significance was evaluated with data from at least three independent experiments or at least five duplicates. GraphPad Prism 6.02 (GraphPad Software, San Diego, CA, USA) was used for data analysis. Statistical analysis was carried out using Student t-test and ANOVA. Data are presented as the mean ± SD. Significance was established at P < 0.05.

To illuminate the effect of TMZ-POH on cell autophagy, four NSCLC cell lines including A549, SPC-A1, NCI-H460 (H460) and NCI-H520 (H520) were employed and subjected to 100 μM TMZ, POH, TMZ plus POH (TMZ + POH) and TMZ-POH, respectively. As shown in Fig. 1a and Additional file 1: Figure S1A, autophagy was activated significantly when treated by TMZ-POH rather than other drugs, as evidence from the increases in the amount of LC3B-II, the important markers of autophagy [20] in all detected cell lines, indicating autophagy activation by TMZ-POH is universal independent of cell type. Next, we checked the formation of autophagosomes by staining endogenous LC3B. We found that TMZ-POH treatment increased intracellular autophagosomes compared to its individual constituents and their combination, as demonstrated by accumulation of LC3B-positive spot-like structures in above drug treated four NSCLC cells (Fig. 1b). In addition, TMZ-POH-induced autophagosome accumulation appeared to be concentration-dependent, as the number of autophagic puncta increased with the concentration of TMZ-POH (Additional file 1: Figure S1B). Furthermore, this phenomenon was further confirmed by transmission electron microscope (TEM). Clearly, TMZ-POH treatment significantly increased intracellular autophagic vacuoles shown as double membrane vesicles with visible cytoplasm contents (Fig. 1c).

To rule out the possibility that TMZ-POH promoted excessive autophagic degradation which led to the failure in autophagosome accumulation, we treated cells combined with Baf.A1, a lysosomal inhibitor leading to accumulation of autophagic vacuoles [18]. As shown in Fig. 1d and Additional file 1: Figure. S1C, we found that in absence of Baf.A1, the number of intracellular autophagic puncta (Fig. 1d) and the amount of LC3B-II (Additional file 1: Figure S1C) were significantly increased when treated with TMZ-POH, whereas upon Baf.A1 treatment to block autophagic flux, these differences caused by TMZ-POH were eliminated, indicating a promotion of excessive autophagic degradation was not involved in the process that TMZ-POH induced autophagosome accumulation.

Induction of autophagy can occur through PI3K-AKT pathway which then phosphorylates mTOR [21]. mTOR inhibits autophagy by targeting autophagy related protein (ATG) 13 [22], and in turn transmits signals to downstream effectors such as autophagy-related gene beclin 1 (BECN1). mTOR functions by directly phosphorylating the key translation regulators p70 ribosomal S6 kinase (P70S6K), leading to an increase in translation of a subset of mRNAs [21]. Therefore, we detected whether TMZ-POH accumulated autophagosome dependent on mTOR signaling. Unexpectedly, TMZ-POH seemed to have no obvious effects on phosphorylation of mTOR itself and its specific substrate P70S6K, and the expression of its downstream effector BECN1 in SPC-A1 and NCI-H460 cells, indicating TMZ-POH-induced autophagosome formation is mTOR independent (Additional file 1: Figure S1D and E).

Next, we checked the effect of TMZ-POH on mitochondrial fusion and fission. Immunostaining for COX-IV, a protein localized on the inner mitochondrial membrane was applied followed by treatment with TMZ-POH and its individual constituents. As shown in Fig. 2a, TMZ-POH induced accumulation of fragmented mitochondria with shorter lengths and fewer numbers of branches due to a lack of mitochondrial fusion whereas other drugs induced that of tubular mitochondria in A549 and SPC-A1 cells. Consistently, immunostaining used to elucidate the mitochondrial outer membranes (anti-TOM20) also demonstrated fragmented mitochondria accumulation in TMZ-POH treated group; Nevertheless, nicotinamide (NAM), an amide form of vitamin B3 can induce autophagy for clearing damaged mitochondria [23], induced tubular mitochondria accumulation (Fig. 2b), implying TMZ-POH unlike NAM might play an alternative role in mitochondrial clearance.

In addition, TEM observation showed that the morphology of mitochondria became more round and shorter, and the overall shapes of the mitochondria were grossly distorted, and the inner mitochondrial matrices were either severely damaged or almost absent in TMZ-POH group (Fig. 2c, and d, left), although their sizes seemed unchanged compared to the control group (Fig. 2d, right).

It is well established that mitochondrial fission can promote mitophagy, a selective autophagy to remove damaged and dysfunctional mitochondria to protect the cells from excessive oxidative stress and cell death [24]. As shown in Fig. 3a and Additional file 2: Figure S2A, total protein, cytoplasmic protein and mitochondrial protein were collected and subjected to western blot analysis. TMZ-POH enhanced LC3B-II expression significantly in total and especially in mitochondrial protein and slightly in the cytoplasmic compared to control, indicating autophagosome formation induced by TMZ-POH almost occurred at mitochondria.

To further confirm this conclusion, colocalization of autophagosome (anti-LC3B) with mitochondria (anti-TOM20) was observed in A549 cells. Starvation as a non-selective autophagy inducer failed to improve the colocalization, while TMZ-POH significantly promoted this colocalization (Fig. 3b), which was also visualized and confirmed by TEM (Additional file 2: Figure S2B). Given that Parkin-ubiquitylated mitochondria is necessary for combination between mitochondrion and autophagosome [24], whether Parkin was localized at site of mitochondrion was detected. As shown in Fig. 3c, TMZ-POH promoted Parkin protein accumulated at mitochondria compared to control. Taken together, our data illuminate TMZ-POH induces mitophagosome accumulation.

To validate the effect of TMZ-POH on mitophagy flux, the intracellular SQSTM1, a selective autophagic adaptor and incorporated with LC3B into autophagosomes and degraded by lysosomal hydrolyses [20], was firstly detected. We found TMZ-POH failed to degrade the SQSTM1 protein although it activated autophagosome formation significantly in A549 and H460 cells (Fig. 4a and b). Moreover, the immunostaining confirmed TMZ-POH upregulated SQSTM1 expression and promoted its assembling into aggregates (Fig. 4c).

To explore the underlying mechanisms how TMZ-POH to block the mitophagy flux, we transfected A549 and SPC-A1 cells with the devised fusion protein mRFP-GFP-LC3 via adenovirus vector, which labels autophagosomes yellow because of superposition of GFP and mRFP signals, and autolysosomes red as the low lysosomal pH quenches the GFP signal [20]. As shown in Fig. 4d and Fig. 5a, most of the puncta lost the GFP signal and retained the mRFP signal in starved cells. However, in cells treated by TMZ-POH like Baf.A1 unlike those starved, quenching of the GFP was significantly diminished as indicated by the retention of both the mRFP and GFP signals.

In search of further evidence that TMZ-POH blocked mitophagy flux, mitochondrial population levels were determined by flow cytometry using Mito-Tracker Green (MTG) dye as described previously [25]. As shown in Fig. 5b, NAM induced a decrease of MTG fluorescence level, whereas TMZ-POH like Baf.A1 induced an increase of MTG fluorescence level relative to untreated cells. Besides, colocalization of lysosome (Lyso-Tracker Red, LTR) with mitochondria (MTG) was visualized in A549 cells. As shown in Fig. 5c, carbonyl cyanide m-chlorophenylhydrazone (CCCP), the most commonly used inducer of mitophagy [26], succeed to induce an overlap between MTG and LTR, whereas TMZ-POH failed, indicating mitophagy flux is blocked by TMZ-POH.

Since acidification is required for the maturation and activation of most lysosomal enzymes, the maintenance of acidity is a hallmark of functionally mature lysosomes [27]. We first stained A549 and SPC-A1 cells with LTR dye, which accumulates in acidic cell compartments, and acts as a lysosomal acidic indicator. This displayed a marked decrease in the LTR signal in response to TMZ-POH treatment, which was also observed in cells treated by Baf.A1 (Fig. 6a). As the lysosomal acidification depends on the process of the endocytic pathway [28], early endosome antigen 1 (EEA1) expression, a marker for early endosomes [29] was analyzed. Western blots and immunostaining showed TMZ-POH decreased down-regulated EEA1 expression as well as the early endosome formation significantly compared to other drugs (Fig. 6b, c, Additional file 3: Figure S3A, B). We further investigated the maturation of the lysosomal hydrolase cathepsin D (CTSD), which is synthesized in the endoplasmic reticulum, and finally matures in the lysosome to form 31-kDa and 14-kDa polypeptide [30]. We found that TMZ-POH clearly reduced the 31-kDa mature CTSD (Fig. 6d and Additional file 3: Figure S3C). In addition, LAMP 1 and 2, the important markers for mature lysosome [9], were also significantly downregulated by TMZ-POH, indicating altered endocytic trafficking or perturbed lysosomal maturation (Fig. 6e and Additional file 3: Figure. S3D).

We further investigated autophagosome-lysosome fusion via observing the colocalization of LC3B with LAMP1. As shown in Fig. 6g, starvation led to an overlap of LC3B with LAMP1, whereas TMZ-POH caused a significant decrease in the coefficient of LC3B-LAMP1 colocalization in A549. The impaired autophagosome-lysosome fusion was observed in TMZ-POH treated group but not in the untreated under TEM (Fig. 6f). Taken together, TMZ-POH impairs lysosome function and its fusion with autophagosome.

Given that RAB7A plays crucial roles in both endo-lysosomal maturation and autophagosome-lysosome fusion as described above, RAB7A activity was analyzed. TMZ-POH inhibited RAB7A activity significantly compared to control (Fig. 7a and Additional file 4: Figure S4A), and this inhibition largely resulted from the downregulation of TMZ-POH on RAB7A protein and its effector RILP expression significantly in all detected cells (Fig. 7b-c and Additional file 4: Figure S4B-C). Moreover, RAB7A protein was not affected at the low concentration (12.5 μM) of TMZ-POH, but suppressed significantly by high concentration (25, 50 and 100 μM) TMZ-POH in A549 cells, indicating TMZ-POH downregulated RAB7A dependent on concentration (Fig. 7d and Additional file 4: Figure S4D). RAB7A prenylation by geranylgeranyl transferase I and II (GGTT I/II) depended on mevalonate pathway, which contributes to its initial targeting to membranes for functioning [15]. Hence, we employed mevalonolactone (MVL) which was capable to hydrolyze to mevalonate [31] and geranylgeraniol (GGOH) [32], an intermediate of the mevalonate pathway upstream of RAB GGTase to co-incubate cells along with TMZ-POH. As shown in Fig. 7e and Additional file 4: Figure S4E, both MVL and GGOH were capable to restore TMZ-POH inhibited RAB7A protein. Taken together, our results indicate the downregulation of TMZ-POH on RAB7A might depend on mevalonate pathway.

Previous study has reported the antitumor effect of TMZ was suppressed when autophagy was prevented by VPS34 inhibitor 3-methyladenine (3-MA) to reduce the formation of autophagosomes [33]. Our results showed that TMZ-POH induced obvious apoptosis both in A549 and SPC-A1 cells coincident with our previous studies [4, 5], which was significantly enhanced by 3-MA mediated autophagy inhibition (Fig. 8a-b), indicating inhibition of autophagosome formation failed to protect cells against TMZ-POH-induced cell death, which was not an “autophagic cell death”.

It has been well established autophagy acts as recycling and defensive mechanisms to supply cells and protect them from cell death. Blockage of autophagic flux surely lead to an insufficient energy supply and sensitizing cells to external stimuli. Whether autophagic flux blockage induced by TMZ-POH sensitized cancer cells to death was also verified. A549 and H520 cells were subjected to irradiation first to stimulate cell death, which also was able to induce autophagy and be sensitized when autophagy inhibited [34]. As shown in Fig. 8c-d, an apoptosis assay revealed that 10 Gy irradiation increased cell apoptosis after 72 h of treatment, and treatment of cells with 25 or 50 μM TMZ-POH during the last 48 h further increased cell apoptosis. These results suggest that treatment of cancer cells with TMZ-POH blocks the activation of autophagic flux by irradiation and increases the sensitivity to the cytotoxic action of irradiation, thus proposing TMZ-POH as a potential radiosensitizer.