Deferasirox was purchased from Aladdin Biotechnology (Shanghai, Chain). Cell culture medium, trypsin, penicillin-streptomycin and fetal bovine serum were purchased from Gibco (Guangzhou, China). Cell counting kit-8 (CCK8) was purchased from Yeasen (shanghai, China). Annexin V-FITC/PI apoptosis Kit, cell cycle kit, mitochondria membrane potential detection (JC-1) kit and calcein-AM were purchased from Beyotime (Shanghai, China).

Firstly, the hyperbranched polylysine was synthesized by following method (Figure 1): Lysine·HCl (27.40 g, 150 mmol) and KOH (8.42 g, 150 mmol) was completely stirred by mortar until well mixed. The mixture was transferred into open 1 L round bottom flask and stirred under 240°C with 3 mol% H₃BO₃ as catalyst. The flask was opened to allow water formed in the reaction to escape. The reaction was stopped and cooled to room temperature after 5 h. The crud product was collected by dissolving in methanol. The KCl formed during the reaction was filtered off. The methanol was evaporated and the product was dissolved in water. The aqueous solution was then lyophilized to afford polylysine solid in 85% yield.

The DFX loading reaction was carried out as follows (Figure 1): the hyperbranched polylysine (274.0 mg, 1.50 mmol) synthesized in the former step was dissolved in DMSO (30 ml), DIPEA (100 μl) was then added to the solution. DFX (280.0 mg, 0.75 mmol) in DMSO (20 ml) and EDC (152.2 mg, 0.080 mmol) was added. The two solution was then mixed in one flask and stirred at room temperature for 5 h. The solution was dialyzed (Spectra/pro MWCO = 1000) against acetonitrile (500 ml). The precipitate was collected and washed with acetonitrile for three times in a glass filter. The crude product was dissolved in minimum of DMSO and dialyzed against water (500 ml, three times). The product was lyophilized to appear as white to yellow solid (yield = 56%).

¹H NMR was recorded at room temperature on a Bruker Avance III 600 MHz nuclear magnetic resonance spectrometer; FTIR: IR spectra was recorded by using FT-IR Spectrometer Platform INVENIO; UV-Vis: UV-Vis spectra was measured by Agilent Cary UV-Vis spectormeter; Lyophilized: Lypholization was carried on a Freeze Dryer Lyophilizer VriTis Benchtop 4KBTZL.

In order to evaluate drug loading efficiency of PL-DFX, calibration curve of DFX has been illustrated by the UV absorbance of 0, 2, 4, 6, 8, 10 µg/ml DFX dissolved in PBS. The absorbances of diluted solutions were measured at 245 nm using UV/Visible spectrophotometer. The plot of UV absorbance versus DFX concentration was subjected to linear regression analysis. The 1.00 mg of PL-DFX was weighed precisely, the amide bond linked polylysine and DFX was hydrolyzed by aqueous solution of HCl (2 M), the DFX was separated by chromatographic column and dissolved in 50 ml PBS to determine the drug concentration. The drug loading efficiency was calculated based on equation LE (%) = We/Wm × 100%.

The human gastric carcinoma cell (HGC-27) was cultured in RPMI 1640 (Gibco). The human colorectal carcinoma cells (DLD-1 and HCT-116) were cultured in RPMI 1640 (Gibco) and McCoy’s 5A respectively. The human renal tubular epithelial cell (HK2) was cultured in DMEM (Gibco). All media were supplemented with 10% fetal bovine serum (FBS, Gibco) and 1% penicillin-streptomycin (Gibco). All cells were cultured and incubated in a humidified atmosphere at 37°C with 5% CO₂.

Cell viability was analyzed by cell-counting kit-8 (CCK8) assay. Briefly, cells were seeded into 96-well plates at a density of 5000–10000 cells per well and incubated overnight. Then drugs were added at different concentrations. 48 h after treatment, cell viability was measured according to the manufacturer’s instructions.

Evaluation of apoptosis assay was performed by using Apoptosis Kit. According to the manufacturer’s instructions, cells were seeded into 6-well plates and incubated with PL-DFX (25 μM) at 37°C for 0 h, 24 h and 48 h. Then the cells were collected and stained with Annexin V-FITC and PI for 20 min. After staining, the cells were evaluated by flow cytometry.

HGC-27 cells were seeded into 6-well plates at the density of 5×10⁵ per well and incubated for 24 h at 37°C. Then using a sterile pipette tip to scratch the cells. After washing 3 times with PBS, the medium containing DFX (12.5 μM) and PL-DFX (12.5 μM) was added to the wells. Finally, photos of wound healing were taken at 0 h and 24 h, respectively.

Transwell chambers were placed into a 24-well plate and 60 μl of diluted Matrigel was added to each chamber. Once the Matrigel was solidified at 37°C, 200 μl of cell suspension (5×10⁴ cells per chamber) which contained DFX (40 μM) and PL-DFX (40 μM) was added into the chamber, and 700 μl of medium containing 10% FBS in the lower chamber. Then cells were incubated at 37°C for 48 h. After washed two times with PBS, cells were fixed with 4% paraformaldehyde for 30 min and dyed with crystal violet for 30 min. Removed the excess dye, the chamber were dried at room temperature and photos were taken.

The cellular LIP was measured as described previously (Prus and Fibach, 2008; Komoto et al., 2021). Briefly, cells were seeded into 6-well plates and incubated with DFX (25 μM) and PL-DFX (12.5 μM) for 48 h. After washed with PBS, these cells were incubated with calcein-AM (0.5 μM) at 37°C for 15 min protected from light. The mean fluorescence intensity was measured by using flow cytometry.

Cells were seeded into 6-well plates and cultured at 37°C overnight. Then cells were incubated with DFX (25 μM) and PL-DFX (12.5 μM) for 24 h. At the end of incubation, cells were collected and fixed in 75% ethanol for 24 h at 4°C. Then the cells were washed with PBS and stained with propidium iodide (PI) solution containing RNase A at 37°C for 30 min protected from the light. Finally, the cells were analyzed by flow cytometry.

Cells were seeded into 6-well plates and incubated overnight. Then cells were treated with DFX (50 μM) and PL-DFX (25 μM) for 24 h. After treatment, cells were stained with JC-1 dyeing working solution for 20 min and washed three times with JC-1 buffer according to the manufacturer’s protocol. Finally, cells were collected and analyzed by flow cytometry.

This study was approved by the ethical committee of the Seventh Affiliated Hospital of Sun Yat-Sen University and performed in compliance with the Declaration of Helsinki. Written informed consents were obtained from all patients.

The clinical specimens were rinsed in PBS containing penicillin-streptomycin 5 times and then sheared into 1–3 mm³ small piece. The tissue fragments were digested for 2 h and supernatant was taken. After centrifugation, supernatant was removed and the remaining was resuspended with DMEM containing FBS. Centrifuged again, the cells were resuspended with DMEM and Matrigel at the ratio of 1:1 by volume. The mixed liquid was added into prewarmed 96-well plates with 10 μl per well then incubated at 37°C for 30 min. Once the Matrigel was solidified, 100 μl of medium was added to each well and cells were cultured at 37°C with 5% CO₂.

Organoids were digested with TripLE (Gibco) at 37°C for 30 min and then centrifuged at 7000 rpm for 1min. The supernatant was discarded and cells were washed one time with DMEM. Then cells were resuspended in DMEM and mixed with Matrigel (Corning) at the ratio of 1:1 by volume. Subsequent steps were described above.

Organoids were digested, centrifuged and resuspended as previously described. The cell suspension was added into prewarmed 96-well plates with 5 μl per well and then incubated at 37°C for 30 min. Once the Matrigel was solidified, 75 μl of organoid-conditioned medium was added and cells were cultured at 37°C and 5% CO₂. Two days later, different concentrations of PL-DFX (75 μl) were added to the wells and organoids were cultured for an additional 120 h. Then, 10 μl of CCK8 reagent was added into each well and incubated for another 4–6 h, and the absorbance of each well was measured at 450 nm by microplate reader (BioTek, SynergyH1, United States).

Data was analyzed by GraphPad 8, and results were presented as mean ± SD. Comparisons between two independent groups were performed by using Student’s t test. *, ** and *** indicate that p-value < 0.05, 0.01 and 0.001, respectively.

The iron nanochelator PL-DFX composed of deferasirox and hyperbranched polylysine. In the choice of DFX and polylysine linkage site, we have carefully selected the carboxyl group on the opposite site to the“iron-catching”domain of DFX in order to avoid its iron decrease potency (Figure 2A). Firstly, the hyperbranched polylysine was synthesized, followed by DFX loading via amide bond formation based on the protocol in the method part. The purified product was characterized by ¹H NMR (Bruker 600 MHz. MeOD, 298K) and FTIR (Figures 2B,C). According to the FTIR spectrum of PL-DFX, peaks at 3296.88 cm⁻¹ represented the O-H streching, and the peaks at 1643.38 and 1625.29 cm⁻¹ represented the benzene streching in the DFX molecule. Additionally, the multi peaks at 7.291–7.356 ppm of ¹H NMR spectrum belong to the aromatic ring proton of DFX, which indicated the drug loading was successful.

Moreover, according to the UV absorbance of DFX (Figure 2D), the drug loading efficiency calculated was 34%. Size and zeta potential are key parameters of nanoparticle efficacy. Nanoparticle with around 100 nm diameter and positive charge can be more easily uptaken by tumor cells (Albanese et al., 2010). The transmission electron microscopy (TEM) demonstrated that PL-DFX was spherical in morphology and monodisperse nanoparticles (Figure 2E). The diameter and zeta potential of PL-DFX, as illustrated by Figure 2F, were 120 nm and 23.8 mV, respectively.

Accumulating evidence has confirmed the antitumor effects of DFX in vitro and in vivo (Ford et al., 2013; Amano et al., 2020; Zhou et al., 2022). Polylysine-based delivery platform can enhance the anti-tumor effects of drugs (Thambi et al., 2016; Toshiyama et al., 2019). Herein, CCK8 assay was used to evaluate the cytotoxicity of DFX and PL-DFX in HGC-27, DLD-1, and HCT-116 cells. IC50 value was also calculated from the dose-response curves shown. As illustrated in Figures 3A–C, cytotoxicity of DFX and PL-DFX against these tumor cells was in a concentration-dependent manner, and cell viability decreased with increasing concentrations of drugs. Compared to DFX, cells incubated with the same concentration of PL-DFX (at equivalent concentrations of DFX) showed lower viability. PL-DFX inhibited HGC-27, DLD-1 and HCT-116 with IC50 values of 18.73 μM, 34.80 μM and 12.58 μM, which were significantly lower than IC50 values of DFX (87.23 μM, 448.7 μM and 168.5 μM, respectively). The bright field images also showed that PL-DFX exhibited greater ability to inhibit cell proliferation than DFX (Figure 3E). The higher the PL-DFX concentration, the lesser number of cells, and the cells became small and round.

Polylysine, as a biodegradable drug delivery platform, not only enhances the cytotoxicity of free drugs in tumor cells, while showing no higher toxicity in non-tumor cells (Du et al., 2020; Zhu et al., 2021). Due to nephrotoxicity of DFX, the cytotoxicity of DFX and PL-DFX was also investigated in human renal tubular epithelial cells (HK2). PL-DFX did not display enhanced cytotoxicity in HK2 cells, which was similar with DFX (Figure 3D). These results confirmed that the antitumor effect and biosafety of PL-DFX were superior to DFX.

Besides, we further evaluated the pro-apoptotic effects of PL-DFX against HGC-27 and DLD-1 by using Annexin V-FITC/PI kit. As illustrated in Figures 4A–D, the percentage of overall apoptotic cells incubated with PL-DFX for 48 h was 34.46% in HGC-27 and 25.7% in DLD-1, which was significantly higher than that at 24 h (24.53% in HGC-27 and 16.85% in DLD-1) and 0 h (7.58% in HGC-27 and 5.42% in DLD-1). These results indicated the time-dependent cytotoxicity of PL-DFX. In summary, polylysine carrier could enhance the anti-tumor effects of deferasirox in the tumor cells.

Next, we also explored the impact of DFX and PL-DFX on cell migration and invasion in vitro by wound healing and transwell invasion assays. As shown in Figures 5A,B, after 24 h incubation, the wound healing rate of PL-DFX group was 2.53%, which was significantly lower than that of control and DFX groups (43.65% and 25.66%, respectively). The transwell assay also displayed that the cell count of invasive HGC-27 cells incubated with PL-DFX for 24 h was significantly lower than that of control and DFX groups (Figures 5C,D). The above observation indicated that the ability of DFX loaded on polylysine to inhibit migration and invasion was greater than free deferasirox.

The cellular labile iron pool (LIP) was measured by using the calcein-AM (Prus and Fibach, 2008; Komoto et al., 2021). Although, calcein-AM has been frequently used to evaluate the cell viability and calcium, its fluorescence is quenched when binding with cellular iron. Iron chelators can inhibit the formation of complexation and increase the fluorescence intensity of calcein. Therefore, the changes in fluorescence intensity of calcein indicates the change in intracellular iron levels. As shown in Figures 6A,C, compared with the control, HGC-27 and DLD-1 cells incubated with DFX and PL-DFX had increased calcein fluorescence, indicating reduction of intracellular iron levels and similar iron-chelating ability of both. Moreover, after incubation with PL-DFX, supplement of Fe²⁺ can partially reverse the cytotoxicity of PL-DFX in HGC-27 and DLD-1 cells (Figures 6B,D). Herein, in the structure of PL-DFX nanoparticles, the conjugation between polylysine and deferasirox do not affect iron chelating ability of deferasirox.

We have demonstrated that PL-DFX has similar iron chelating ability with DFX. Next, the impact of PL-DFX on biological functions of HGC-27 and DLD-1 cells was performed. Iron is essential for the activity of ribonucleotide reductase, and iron depletion can inhibit the DNA synthesis and cause G1/S arrest (Chantrel-Groussard et al., 2006; Yu et al., 2007). Herein, as shown in Figure 7, similar to the previous study, compared to the control group, cells incubated with DFX displayed obviously phase G1/S arrest. As expected, PL-DFX also resulted in an increased proportion of cells in G1/S phase.

Iron plays a key role in mitochondria biological function and biosynthesis (Paul et al., 2017), and iron deficiency can compromise mitochondria function (Hoes et al., 2018). Decreased mitochondria membrane potential (MMP) is a hallmark of mitochondria dysfunction. Thus, the impact of DFX and PL-DFX on mitochondria membrane potential was investigated by using JC-1. As illustrated in Figure 8, iron deficiency induced by DFX resulted in decreased MMP. Similar to DFX, PL-DFX also depolarized the MMP in HGC-27 and DLD-1 cells.

Cell line is an important platform for drug development and screening. However, variations between cell lines and original tumors are responsible for the failure of the drug-based clinical trials (Kamb, 2005; Caponigro and Sellers, 2011; Barretina et al., 2012). Organoid, which established from patient-derived tumor tissue, resembles the original tumors in terms of biological characteristics and heterogeneity, and organoid-based drug screening methods have yield satisfactory results (Drost and Clevers, 2018). Therefore, we utilized the organoids which established from patient-derived gastric and colorectal cancer tissues to evaluate the clinical efficacy of PL-DFX. As shown in Figures 9A,B, the viability of organoids decreased with increasing the concentrations of PL-DFX. Besides, the PL-DFX treated organoids displayed smaller in size and fewer in number in compared to control organoids in both gastric cancer organoids (GC) and colorectal cancer organoids (CC) (Figure 9C).