Chemosensitizing activity of peptide from Lentinus squarrosulus (Mont.) on cisplatin-induced apoptosis in human lung cancer cells

The limitations of cisplatin, a standard chemotherapy for lung cancer, have been documented with serious adverse effects and drug resistance. To address the need for novel therapy, this study firstly reveals the potential of peptide from Lentinus squarrosulus (Mont.) as a chemotherapeutic adjuvant for cisplatin treatment. The purified peptide from L. squarrosulus aqueous extracts was obtained after eluting with 0.4 M NaCl through FPLC equipped with anion exchange column. Preincubation for 24 h with 5 µg/mL of the peptide at prior to treatment with 5 µM cisplatin significantly diminished %cell viability in various human lung cancer cells but not in human dermal papilla and proximal renal cells. Flow cytometry indicated the augmentation of cisplatin-induced apoptosis in lung cancer cells pretreated with peptide from L. squarrosulus. Preculture with the peptide dramatically inhibited colony formation in lung cancer cells derived after cisplatin treatment. Strong suppression on integrin-mediated survival was evidenced with the diminution of integrins (β1, β3, β5, α5, αV) and down-stream signals (p-FAK/FAK, p-Src/Src, p-Akt/Akt) consequence with alteration of p53, Bax, Blc-2 and Mcl-1 in cisplatin-treated lung cancer cells preincubated with peptide from L. squarrosulus. These results support the development of L. squarrosulus peptide as a novel combined chemotherapy with cisplatin for lung cancer treatment.


Results
Purified peptide isolated from L. squarrosulus extracts. The peptide pellets were precipitated out from the homogenized aqueous solution of L. squarrosulus fruiting bodies by adding 40-80% (NH 4 ) 2 SO 4 . These crude peptide extracts were then resolubilized in PBS (pH 7.4) and further purified through FPLC coupled with anion exchange column. Figure 1a shows a representative FPLC chromatogram of different peptide fractions which were eluted through a stepwise increase in concentration of NaCl solution (0.1, 0.2, 0.3, 0.4 and 0.5 M). The eluted peptide profile was composed of six peaks including unbound peptides. The composition of each peptide fraction was evaluated through SDS-PAGE analysis. Among the six peptide fractions, the most purified peptide appears in the fraction eluted with 0.4 M NaCl, which presented an intensively stained single band of protein at ~ 37 kDa (Fig. 1b). It is worth noting that protein bands appearing with less intensity and various molecular weight were presented in the other peptide fractions. Therefore, the peptide of 0.4 M NaCl fraction was selected for further investigations on chemosensitizing effect to cisplatin.
Proteomics analysis of L. squarrosulus purified fraction. Despite hundreds of peptide fragments separated through LC-MS/MS, the search results reported only one matched query. Fragments with observed mass at 657.9400 were uniquely matched (p < 0.05) with the peptide sequence (R.YGFTEVAGNFQQHNFGR.G) from hypothetical protein GSI_05267 of Ganoderma sinense ZZ0214-1 and the same set of peptides was also matched to hypothetical protein L227DRAFT_578502 of Lentinus tigrinus ALCF2SS1-6 as shown in Fig. 1c. The peptide sequence of matched fragments was then 100% identity matched to "A0A2G8SFL4_9APHY-Extracellular metalloproteinase Ganoderma sinense ZZ0214-1" and "A0A5C2S0B7_9APHY-Extracellular metalloproteinase Lentinus tigrinus ALCF2SS1-6" from UniProt database.

Cytotoxicity of L. squarrosulus peptide.
To clarify the optimum concentrations for further use in the study, the cytotoxic effect of 0.4 M NaCl L. squarrosulus peptide fraction was investigated in both human lung cancer and normal cells. After incubation with various concentrations (0-50 µg/mL) of 0.4 M NaCl peptide for 24 h, the reduction of %cell viability was significantly noted in human lung cancer H460 cells treated with 10 and 50 µg/mL of the peptide (Fig. 2a). Corresponding with the viability assay results, apoptosis as characterized by the bright blue fluorescence of Hoechst33342 nuclear stain was evidently increased in the cells cultured with 10-50 µg/mL of 0.4 M NaCl peptide (Fig. 2b). Notably, there were no red fluorescence cells signifying necrosis in all peptide-treated lung cancer cells.
Because hair loss and renal failure are major side effects of cisplatin [30][31][32] , the cytotoxicity of 0.4 M NaCl peptide was also investigated in human dermal papilla DPCs cells and human proximal renal HK-2 cells. Figure 2c,e www.nature.com/scientificreports/ depict the viability in DPCs and HK-2 cells as determined through MTT assay. The %cell viability remained approximately 100% in both cell types after treatment with 0-10 µg/mL of L. squarrosulus peptide for 24 h. Although significant reduction of cell viability was detected, apoptosis and necrosis were barely observed in DPCs and HK-2 cells cultured with peptide of 0.4 M NaCl fraction at 50 µg/ml for 24 h (Fig. 2d,f). Remarkably, the peptide extract possesses selective cytotoxicity towards lung cancer cells which is evident in the lower %cell viability and increased apoptosis observed in human lung cancer H460 cells after treatment with 0.4 M NaCl peptide at 10-50 µg/ml compared with both DPCs and HK-2 cells.
To clearly clarify chemosensitizing effect, pretreatment of cancer cells with chemotherapeutic adjuvant at nontoxic concentration is suggested 33 . Therefore, 5 µg/mL of 0.4 M NaCl peptide which did not exhibit cytotoxicity in cancer and non-cancer cells was selected as the optimum non-toxic concentration for further experiments related to sensitization to cisplatin-induced cell death in human lung cancer cells.
L. squarrosulus peptide sensitizes cisplatin-induced apoptosis in human lung cancer cells. The chemosensitizing effect of L. squarrosulus peptide was determined in time-dependent investigation in human lung cancer cells. Lung cancer H460 cells were pretreated with 0.4 M NaCl peptide prior to exposure of 5 µM  www.nature.com/scientificreports/ cisplatin. The cell viability of H460 cells further decreased in a time-dependent manner in response to pretreatment (3-24 h) of 0.4 M NaCl peptide (5 µg/mL) as indicated in Fig. 3a. Figure 3b presents the time-dependent increment of %apoptosis in H460 cells preincubated with 0.4 M NaCl peptide. Correspondingly, nuclear costaining with Hoechst33342 and propidium iodide (PI) also showed that apoptosis cell death was remarkably noted in L. squarrosulus peptide-pretreated H460 cells in a time (6-24 h) -dependent manner (Fig. 3c). Flow cytometry with annexin V-FITC/PI was performed to further clarify mode of cell death and evaluate the sensitizing effect of L. squarrosulus peptide on cisplatin-induced apoptosis in human lung cancer H460 cells. Labelled annexin V-FITC allows the accurate tracking of the early apoptosis cells due to specific binding to the exposed phosphatidylserine on the surface of cell membrane 34 . Annexin V-FITC is used in conjugation with PI to further track the cells undergoing late apoptosis and necrosis, cell integrity is compromised thereby allowing PI internalization 35 . The obtained histograms indicated that majority of the cisplatin-treated and peptide-treated H460 were viable cells, comparable to the untreated cell population, as indicated by the large accumulation of non-stained cells (Annexin V-/PI-) in the three groups (Fig. 3d). On the other hand, there was remarkable increase in the percentage of cells in both early (Annexin V + /PI-) and late (Annexin V + /PI +) apoptosis in human lung cancer cells pretreated with 5 µg/mL of 0.4 M NaCl peptide for 6-24 h following with further 24 h treatment of 5 µM cisplatin (Fig. 3e). These results indicate that the L. squarrosulus peptide augments cisplatininduced apoptosis when non-toxic concentrations were used for both cisplatin and the peptide, thus providing further evidence of a strong chemosensitization effect of L. squarrosulus peptide.
Diminution of colony formation in cisplatin-treated H460 cells precultured with L. squarrosulus peptide. The ability of lung cancer cells to form new colonies after cisplatin treatment alone or with 0.4 M NaCl fraction from L. squarrosulus extracts was investigated through clonogenic assay, which is an in vitro assay often used to gauge the effect of a treatment on cell survival and proliferation 36 . Viable H460 cells were collected after preincubation with 0-5 µg/mL of 0.4 M NaCl peptide for 24 h following with the further incubation of 5 µM cisplatin for another 24 h. A single cell suspension was prepared and seeded onto 6-well plate at a density of 250 cells/well, and the culture was maintained for 7 days. Figure 3f depicts that in comparison with L. squarrosulus peptide-treated H460 cells and the untreated control group which presented evidently higher colony number, colony formation was diminished in cisplatin-treated cells. Remarkably, the reduction of colony formation was greater in cisplatin-treated H460 cells which were preincubated with L. squarrosulus peptide (Fig. 3g).

L. squarrosulus peptide modulates apoptosis and integrin-mediating survival pathway.
The alteration on apoptosis-regulating proteins was examined in human lung cancer cells cultured with L. squarrosulus peptide for varying time periods (0-24 h). Augmentation of tumor suppressor p53 and pro-apoptosis Bax proteins as well as downregulation of anti-apoptosis Mcl-1 protein were time-dependently observed in H460 cells incubated with 0.4 M NaCl peptide at 5 µg/mL for 3-24 h (Fig. 4a). The alteration of these apoptosisregulating proteins correlated well with the chemosensitizing activity of L. squarrosulus peptide. Moreover, the expression of Bcl-2 anti-apoptosis protein was significantly decreased in lung cancer cells after treatment with the peptide for 12-24 h compared with non-treated control cells (Fig. 4b).
The evaluation on modulatory role of L. squarrosulus peptide on survival signaling pathways in human lung cancer cells was further performed. Western blot analysis revealed incrementally decreased expression of survival-related proteins including FAK, p-FAK, Src, p-Src, Akt, p-Akt in H460 cells cultured with 0.4 M NaCl peptide at 5 µg/mL for 6-24 h (Fig. 4c,d). Moreover, the diminished levels of β1, β3, β5, α5 and αV integrins, which are up-stream regulatory molecules, were time-dependently observed following the treatment of L. squarrosulus peptide (Fig. 4e,f). These results indicated that L. squarrosulus peptide at non-toxic concentration effectively alters apoptosis and integrin-mediating survival signals in human lung cancer cells.
Anticancer mechanisms of cisplatin promoted by L. squarrosulus peptide. The results in Fig. 4 show that incubation with L. squarrosulus peptide alone modulated the expression levels of signaling proteins involved in apoptosis and survival pathways in human lung cancer cells. These proteins were likewise investigated in L. squarrosulus peptide-preincubated H460 cells which received cisplatin treatment. Culture with 5 µM of cisplatin alone for 24 h significantly altered the level of Bcl-2 family proteins including Bax, Mcl-1 and Bcl-2 showing that intrinsic apoptosis pathway is activated in response to cisplatin treatment (Fig. 5a). The upregulation of p53 and Bax as well as the reduction of Mcl-1 and Bcl-2 were significantly greater in the 0.4 M NaCl peptide-pretreated H460 cells subjected to cisplatin treatment in comparison to the cisplatin-treated cells (Fig. 5b). Similarly, the suppression on integrin-mediating survival pathway was also indicated in cisplatin-treated H460 cells. The decreased levels of integrin β1, β3, β5 and α5 and the corresponding diminution of survival signaling proteins including FAK, p-FAK, Src, p-Src, Akt and p-Akt were more pronounced in the cisplatin treated-H460 cells that were precultured with 0.4 M NaCl peptide compared with the cells treated with only cisplatin (Fig. 5c,d,e,f). Intriguingly, the downregulation of integrin αV in human lung cancer cells was only observed in response to pretreatment with L. squarrosulus peptide followed by the exposure to cisplatin.

Selective chemosensitizing effect of L. squarrosulus peptide in human lung cancer cells.
To further assess the chemosensitizing effect previously observed in H460 (p53 and KRas wild-type) (Fig. 3), other human lung cancer cells including H292 (p53 wild-type), H23 (p53 and KRas mutant) and A549 (KRas mutant) were either pretreated with 5 µg/mL of 0.4 M NaCl peptide for 24 h or left untreated before further incubation with non-toxic concentration of cisplatin at 5 µM for 24 h. Figure 6a,c,e, respectively revealed the greater reduction of %cell viability in cisplatin-treated H23, H292 and A549 lung cancer cells which were preincubated with L. squarrosulus peptide compared with the cells treated with cisplatin alone. While there were no observable apop- www.nature.com/scientificreports/ tosis and necrosis cells detected after nuclear staining with Hoechst33342 and PI in lung cancer cells cultured with cisplatin or 0.4 M NaCl peptide alone for 24 h, apoptosis was strongly augmented in cisplatin-treated cells that were preincubated with L. squarrosulus peptide (Fig. 6b,d,f). Due to notable cisplatin-modulated serious adverse effects such as hair loss and nephrotoxicity [30][31][32] , the effect of L. squarrosulus peptide on cisplatin-induced toxicity was also evaluated in human dermal papilla DPCs and proximal renal HK-2 cells. Figure 6g,h revealed that 24 h pretreatment with 0.4 M NaCl peptide at 5 µg/mL did not cause significant alteration in viability and death in DPCs incubated with cisplatin (5 µM) for 24 h when compared with untreated control and cells with cisplatin treatment alone. Furthermore, there was no cisplatininduced toxicity as reflected by the MTT viability assay and Hoechst33342/PI costaining performed in HK-2 cells precultured with L. squarrosulus peptide prior exposure to cisplatin (Fig. 6i,j). These results suggested that L. squarrosulus peptide selectively sensitized cisplatin toxicity in human lung cancer cells.
Additionally, chemosensitizing effect of the peptide from L. squarrosulus was validated in three-dimensional cancer spheroids. The in vitro multicellular cancer spheroids better reflect in vivo tumor behavior and thus the spheroid culture has been recognized as a good model for evaluation of anticancer activity 37,38 . Similar to the effects seen in two-dimension cell culture, preincubation with 0.4 M NaCl peptide at 5 µg/mL for 24 h prior to treatment with cisplatin remarkably diminished %cell viability of cancer spheroids derived from lung cancer H460 (Fig. 7a), H23 (Fig. 7c), H292 (Fig. 7e) and A549 (Fig. 7g) cells. Meanwhile, culture alone either with 0.4 M NaCl peptide (5 µg/mL) or cisplatin (5 µM) for 24 h did not alter viability in all lung cancer spheroids. Nuclear staining obviously depicted apoptosis cell death evidenced by the bright blue fluorescence of Hoechst33342 in the spheroids treated with L. squarrosulus peptide following with cisplatin treatment (Fig. 7b,d,f,h). These results strongly support the chemosensitizing effect of L. squarrosulus peptide on cisplatin-induced apoptosis in human lung cancer cells.

Discussion
Cisplatin is one of the most clinically effective anticancer agents and it remains to be a component of first-line treatment for advanced non-small cell lung cancer 39 . However, drug resistance and numerous undesirable side effects limit the use of cisplatin. To combat these problems and obtain greater therapeutic outcome, the combination of cisplatin with other drugs has been an available option 40 . Chemosensitization by natural compounds to increase cisplatin potency and decrease substantial toxicity is gaining interest as a novel anticancer strategy 27 . Thus, this study evaluated the chemosensitizing effect of the peptide extracted from L. squarrosulus, an edible mushroom, on cisplatin-induced cell death in human lung cancer cells. Time-dependent chemosensitizing activity was evidenced by the augmentation of cisplatin-induced apoptosis detected with Hoechst33342/PI costaining (Fig. 3b,c) and annexin V-FITC/PI assay through flow cytometry (Fig. 3d,e) of human lung cancer cells pretreated with 5 µg/ml of purified 0.4 M NaCl peptide from L. squarrosulus extracts for 6-24 h.
Cisplatin cytotoxicity is mediated by the propagation signals produced in response to DNA damage which in turn activate downstream proteins such as p53 and pro-apoptosis Bax, ultimately resulting in apoptosis 40,41 . The dysregulation of apoptosis, especially through the intrinsic pathway controlled by the Bcl-2 family proteins, could contribute to the resistance to cisplatin-induced toxicity observed in cancer cells 41 . Indeed, the elevated level of anti-apoptosis proteins in the Bcl-2 family has been well recognized as an underlying mechanism of chemotherapeutic resistance 42 . The downregulation of p53 and Bax in lung cancer cells is also associated with increased cisplatin resistance and reduced apoptosis 10 . Moreover, tumor progression is tightly correlated with decreased level of tumor suppressor p53 protein 43 . The treatment with L. squarrosulus peptide alone in human lung cancer cells caused not only the reduction of anti-apoptosis Mcl-1 and Bcl-2 proteins but also the augmentation of p53 and Bax in a time-dependent manner (Fig. 4a). Besides that, pretreatment with the peptide dramatically enhanced the effects of cisplatin treatment at non-toxic concentration, such as reduction of Mcl-1 and Bcl-2 levels as well as the elevation of p53 and Bax in cisplatin-treated lung cancer cells (Fig. 5a).
Integrins are heterodimeric transmembrane adhesion proteins which serve as survival initiating molecules. Cellular signaling generated from the interaction between ECM and integrins activates down-stream survival proteins including FAK and Src which subsequently propagates the PI3K/Akt pathway and its strong survivalrelated effects 18 . Downregulation of integrins (β1, β3, β5, α5 and αV) corresponds to the suppression of activated  www.nature.com/scientificreports/ forms of survival signaling molecules, p-FAK, p-Src and p-Akt in lung cancer cells cultured with L. squarrosulus peptide (Fig. 4c-f). Interestingly, it has been continuously reported that various natural peptides could effectively repress the integrin-regulated survival pathways 44,45 . Integrins, especially the β1, β3, β5, α5 and αV subunit, and the related survival proteins have been shown to contribute to uncontrollable lung cancer progression and www.nature.com/scientificreports/ chemotherapeutic failure, as their dysregulation could promote rapid cloning of cancer cells and increase cell survival against various stresses such as cisplatin toxicity 19,20,46 . Coinciding with decreased survival signaling, the capability of individual cells to generate new cancer colonies was clearly more suppressed in cisplatin-treated lung cancer cells which were then preincubated with L. squarrosulus peptide as compared to both untreated control and cisplatin-treated groups (Fig. 3e,g). This inhibitory activity of L. squarrosulus peptide on colony formation correlated well with observed lower level of integrin-regulated survival proteins in lung cancer cells preincubated with L. squarrosulus following with cisplatin treatment in comparison with cells cultured only with cisplatin ( Fig. 5c-f).
Integrin signaling allows the survival of lung cancer cells against drug-induced apoptosis. Thus, targeting of integrins presents good therapeutic potential for lung cancer cells treatment 24,46 . Downregulation of integrins and associated survival signals induces apoptosis and suppresses survival in diverse lung cancer cell, as observed in both p53 wild-type and mutants lung cancer cells 46,47 . Moreover, activation of integrins (β1, β3, αV) is also associated with aggressive features of oncogenic KRas dependent lung cancer cells and antagonizing these integrins has been shown to sensitize the resistant lung cancer cells to chemotherapy [48][49][50] . The presented results indicate that L. squarrosulus peptide may provide chemosensitizing activity against wild-type and mutant lung cancer cells through inhibition of integrins as demonstrated by the peptide's effects on H460 (p53 and KRas wild-type), H292 (p53 wild-type), H23 (p53 and KRas mutant) and A549 (KRas mutant) (Figs. 3, 6, 7).
In line with the results presented in this study, there is much evidence which suggest that integrin is a likely target molecule affected by anticancer peptides 44,45,51 . The disruption on integrin-ECM interaction can trigger the internalization and degradation of transmembrane integrins 52,53 . Although downregulated level of integrins and consequent repression of downstream survival signals support that integrin is a potential therapeutic target of L. squarrosulus peptide, other related mechanisms of anticancer peptides such as perturbation of cell membrane and penetration to intracellular organelles are worthy of further elucidation 54 .
Not only selective anticancer activity against human lung cancer cells (Fig. 2) but also specific chemosensitizing effect of the obtained L. squarrosulus peptide was clearly evident as the results indicate that the enhancement of cisplatin-induced apoptosis occurred only in human lung cancer cells but not in non-cancer DPCs and HK-2 cells (Fig. 6). These results suggest that the L. squarrosulus peptide may provide protective effect against cisplatin side effects such as hair loss and nephrotoxicity which correlate with increased cell death in dermal papilla and proximal renal cells, respectively 31,32 . The selective chemo-enhancing activity against human lung cancer cells of L. squarrosulus peptide would thus be beneficial in reducing the dose-limiting cisplatin toxicity observed in normal cells.
Although initial assessment suggests the eluted 0.4 M NaCl fraction is composed of a ~ 37 kDa peptide (Fig. 1), the proteomics analysis revealed many amino acid sequences that did not match with available peptides in the library. Further improvement in purification process might facilitate the identification of the novel chemosensitizing peptide. However, the highest recorded similarity score in the proteomics analysis corresponded to extracellular metalloproteinase Ganoderma sinense and Lentinus tigrinus (Fig. 1c), confirming the mushroom origins of this chemotherapeutic adjuvant peptide and encouraging greater attention to these natural sources.
In summary, the current study revealed selective chemosensitizing effect of L. squarrosulus peptide extract to cisplatin-induced apoptosis in human lung cancer cells via suppression of integrin/FAK/Src/Akt survival signaling, downregulation of anti-apoptosis proteins (Mcl-1, Bcl-2), and upregulation of p53 as well as pro-apoptosis Bax proteins (Fig. 8). Subsequently, the data presented herein would support the development of L. squarrosulus peptides as a novel chemosensitizer.

Preparation of extracted peptide.
• Isolation of crude peptides The edible mushrooms, L. squarrosulus, were homogenized with deionized water (3 mL/g). Proteins were salted out slowly on ice by adding (NH 4 ) 2 SO 4 to the aqueous filtrate until 40-80% final saturation. Next, centrifugation at 6,500 rpm for 1 h (4 °C) was performed to acquire crude protein pellets which were resolubilized with phosphate buffer solution (PBS; pH 7.4) containing Na 2 HPO 4  www.nature.com/scientificreports/ the solution was dialyzed overnight with the phosphate buffer at 4 °C to eliminate (NH 4 ) 2 SO 4 as previously described 29 . • Peptide purification Fast protein liquid chromatography (FPLC) with HiTrap DEAE FF 1 mL anion exchange column pre-equilibrated with PBS (pH 7.4) was used for further peptide purification. After elution of the unbound peptides, the different adsorbed peptides were released from the positively charged adsorbent via stepwise increase of NaCl concentration (0.1, 0.2, 0.3, 0.4 and 0.5 M) in PBS at a flow rate of 1 mL/min. The eluted peptides were further concentrated via lyophilization. The eluates were then analyzed for quantity and homogeneity of peptides by BCA assay kit and SDS-PAGE (sodium dodecyl sulfate polyacrylamide gel electrophoresis), respectively.
• Determination of protein content BCA assay kit was used to measure the total protein content of each concentrated peptide extract, according to manufacturer's instructions. The developed product was evaluated through microplate reader (Anthros, Durham, NC, USA) at 570 nm. The protein concentration was calculated based on the calibration curve of the BSA standard.
• Evaluation on homogeneity of purified peptide fraction The homogeneity of purified peptide fraction was determined by SDS-PAGE using 15% (w/v) gel as previously described 29 . Briefly, the mixtures containing 15 µg protein content of each peptide fraction and loading dye were added onto SDS-PAGE after heating at 95 °C for 5 min. After complete separation, the gel was stained with coomassie brilliant blue R-250 solution overnight and then destained with isopropanol: acetic acid: water (10: 10: 80% v/v) solution. The most purified fraction from L. squarrosulus extracts presenting only one major protein band via SDS-PAGE analysis was selected for further peptide identification. The SDS-PAGE gel area of interested protein was cut and washed with sterile deionized water. After destaining with 25 mM NH 4 HCO 3 in 50% methanol, the gel was washed three times with sterile deionized water. Before incubation with 10 mM dithiothreitol (in 10 mM NH 4 HCO 3 ) at room temperature for 1 h, the gel was dehydrated by soaking in acetonitrile. After adding 5 mM dithiothreitol (in 50 mM iodoacetamide), the gel was subsequently incubated for 1 h in the dark. Then, the gel was washed with acetonitrile (2 time × 5 min) and digested with 10 ng/μL of proteomics-grade trypsin (Sigma-Aldrich Chemical, St. Louis, MO, USA) in 50% acetonitrile/10 mM NH 4 HCO 3 at 37 °C for 12 h. The supernatant was collected and concentrated through vacuum-drying. The peptide sample was kept at -80 °C until performing proteomics analysis. Before the sample was subjected to mass spectrum analysis, it was solubilized in 0.1% formic acid. LC-MS/MS with nanocolumn was set at a flow rate of 300 nL/min using the HCT Ultra PTM Discovery System (Bruker Daltonics Ltd., Hamburg, Germany) coupled to an UltiMate 3000 LC System (Dionex Ltd., Sunnyvale, CA, USA). A multistep gradient of 10-70% acetonitrile (in 0.1% formic acid) was used as mobile phase to elute peptides 55  Cytotoxicity assay. Cells were seeded onto 96-well plate at a density of 1 × 10 4 cells/well for overnight.
After indicated treatment, the culture medium was replaced with 0.4 mg/mL of MTT and the cells were further incubated for 3 h at 37 °C in dark place. Then, the supernatant was removed and DMSO was added to dissolve the formazan product. A microplate reader (Anthros, Durham, NC, USA) was used to measure the intensity of formazan color at 570 nm. The absorbance ratio of treated to non-treated control cells was calculated and presented as percent cell viability.
Nuclear staining assay. Mode of cell death was evaluated via nuclear staining assay. The treated cells were costained with Hoechst33342 (0.02 µg/mL) and PI (0.01 µg/mL) at 37 °C for 30 min. A fluorescence microscope (Olympus IX51 with DP70, Olympus, Tokyo, Japan) was used to visualize the cells and characterize the mode of death. The Hoechst33342 dye was allowed for the detection of nuclear condensation which characterizes apoptosis while the necrosis cells were distinguished by their uptake of PI.
Flow cytometric analysis. The percentages of cells undergoing apoptosis and necrosis were also quantified through flow cytometry after double staining using annexin V-FITC/PI assay kit, following manufacturer's instructions. Briefly, lung cancer H460 cells were cultured at the density of 1.5 × 10 5 cells/well in 6-well plate. The cells were preincubated with the peptide extract for varying times (0, 6, 12, 24 h) before further culture with 5 µM of cisplatin for 24 h. Then, the cells were detached and centrifuged at 5,000 rpm (4 °C) for 5 min. The annexin V-FITC and PI working solution were added into respective single cell suspensions which were prepared in binding buffer solution. The proportions of living, apoptosis and necrosis cells in each prepared suspension were determined with Guava easyCyte 5 benchtop flow cytometer with the GuavaSoft version 2.7 software (Merck, Darmstadt, Germany).
Clonogenic assay. Human lung cancer cells (1.5 × 10 5 cells/well) in 6-well plate were pretreated with peptide (5 µg/mL) for 24 h before exposure with 5 µM of cisplatin. After 24 h, single cell suspensions of 250 viable cells derived from each treatment were seeded again onto 6-well plate. The colony formation in each plate was examined after 7 days of incubation under 5% CO 2 humidified atmosphere at 37 °C. The resulting cancer colonies were counted after staining with 0.05% w/v crystal violet in 4% formaldehyde 36 . Establishment of three-dimensional multicellular cancer spheroids. Human lung cancer cells were seeded at a density of 1 × 10 4 cells/well and maintained with appropriate media containing 10% FBS, 2 mM l-glutamine and 100 units/mL penicillin/streptomycin in 96-well round bottom ultra-low attachment plate (Corning, Tewksbury, MA, USA). After 4-day incubation, the lung cancer spheroids were subjected to www.nature.com/scientificreports/ the evaluation of chemosensitizing activity. To examine viability of lung cancer cells, the three-dimensional spheroids were incubated with PrestoBlue reagent (10 µL) for 3 h. Then, the fluorescence intensity of resorufin, the reduced form of resazurin was detected with fluorescence microplate reader (CLARIOstar Plus Microplate Reader, BMG Labtech, Baden-Wurttemberg, Germany) at excitation wavelength of 570 nm and emission wavelength of 610 nm. The percent cell viability was derived from the relative fluorescence intensity between treated to non-treated spheroids. Moreover, mode of cell death was observed under fluorescence microscope (Olympus IX51 with DP70, Olympus, Tokyo, Japan) after costaining tumor spheroids with Hoechst33342 (0.02 µg/mL) and PI (0.01 µg/mL) at 37 °C for 30 min 56 .
Western blot analysis. Human lung cancer H460 cells were seeded onto 6-well plate at the density of 1.5 × 10 5 cells/well. After indicated treatments, the cells were incubated with lysis buffer containing 1X radioimmunoprecipitation assay (RIPA) buffer (Thermo scientific, Rockford, IL, USA) and cocktail protease inhibitor (Roche Applied Science, Indianapolis, IN, USA) for 45 min on ice. To collect the clear supernatant, the cell lysates were centrifuged at 10,000 rpm at 4 °C for 15 min. BCA protein assay kit was used to determine total protein content. Equal amount of protein from each sample was adjusted to appropriate volume using lysis buffer and then mixed with loading buffer. The samples were heated at 95 °C for 5 min to speed up protein denaturation. Then, 35 µg protein from each sample was loaded onto a 10% SDS-PAGE. After separation, the cellular proteins were transferred onto 0.45 µm nitrocellulose membranes (Bio-Rad Laboratories, Hercules, CA, USA). The membranes were blocked with 5% non-fat dry milk in Tris buffered saline with Tween 20 (TBST) at room temperature for 45 min. The primary antibodies were added onto the membranes and incubated at 4 °C overnight. Then, the membranes were washed with TBST for three times (7 min). The membranes were further incubated with specific horseradish peroxidase (HRP)-linked secondary antibodies for 2 h at 25 °C. Finally, the signals from specific proteins were detected using chemiluminescent substrates. The analyst/PC densitometric software (Bio-Rad Laboratory, Hercules, CA, USA, version 6.0.1, 2017) was used to quantify the density of protein signal.

Statistical analysis.
All the data were averaged from three independent experiments. Statistical data analysis was carried out by one-way ANOVA and Tukey HSD post hoc test using SPSS Statistic 22 version (Armonk, NY, USA). Statistical significance was considered at p < 0.05.