Amelioration of Cisplatin Nephrotoxicity by Fe–Pt Coordination Exchange‐Triggered Ferroptosis Suppression

ABSTRACT

Ferroptosis is an iron‐dependent form of regulated cell death that has emerged as one of the mechanisms underlying cisplatin‐induced nephrotoxicity. Here, we report an iron‐chelating polymeric micelle for controlled cisplatin activation and ferroptosis suppression to address the above issue. The block copolymer, methoxyl polyethylene glycol‐poly(glutamic acid) (mPEG‐PGlu) was coupled with an iron chelator, deferiprone (Dfp) to generate the tailored amphiphilic conjugate, mPEG‐P(Glu‐Dfp) that could coordinate with activated cisplatin and self‐assemble into micelles (Pt@M_Dfp). The cisplatin‐loaded mPEG‐PGlu micelles (Pt@M_Glu) and free cisplatin were employed as controls. Upon endocytosis, the intracellular labile iron could replace cisplatin due to the higher affinity between Dfp and Fe²⁺. This coordination exchange resulted in controlled cargo release and suppressed lipid peroxidation and inflammation in renal tubular cells (HK‐2). The proof‐of‐concept was also validated in the subcutaneous 4T1 mouse breast tumor model using plasma and urine biomarkers of ferroptosis, inflammation, and renal injury as the indices, which concurred well with the histological staining of kidney tissue and fibrosis assessment. Moreover, due to the controlled cargo activation, Pt@M_Dfp significantly enhanced the in vivo antitumor efficacy compared to Pt@M_Glu and free cisplatin. Collectively, the coordination exchange approach is promising in controlled delivery of cisplatin for efficacy enhancement.

Transporter‐mediated renal deposition of cisplatin and iron‐dependent ferroptotic cell death are important contributors to cisplatin‐induced nephrotoxicity. This work reports an iron‐chelating polymeric micelle that concurrently achieves sustained cisplatin release/activation and ferroptosis suppression, thereby potently inhibiting cisplatin‐induced renal injury. The coordination exchange approach is promising for controlled delivery of cisplatin to enhance efficacy.

1. Introduction

As a highly effective chemotherapeutic agent, cisplatin has been widely used as the first‐line drug for treating diverse cancers, such as bladder, head and neck, ovarian, and lung cancers [1]. Its mechanism of action involves the interaction with DNA to form intra‐strand crosslinked adducts that interfere with DNA replication and further induce apoptotic cell death [2]. However, despite the well‐established efficacy of cisplatin, the dose‐dependent nephrotoxicity has largely limited its clinical utility [3]. This renal toxicity primarily manifests as acute renal parenchymal injury, which is typically characterized by proximal tubular necrosis, loss of brush border microvilli, and infiltration of immune cells into the renal interstitium [4, 5]. In clinical practice, hydration therapy is often employed to mitigate cisplatin nephrotoxicity by reducing renal tubular accumulation of cisplatin, but it is assoica unsatisfactory outcomes with persistent nephrotoxicity in 40% of patients [6]. Therefore, the discovery of novel therapeutic interventions to mitigate cisplatin‐induced nephrotoxicity while preserving its antitumor efficacy represents a critical unmet clinical need.

The uptake of cisplatin by proximal tubular cells is mediated by copper transporter 1 and organic cation transporter 2 [7, 8]. The latter is predominantly expressed in the S3 segment of the proximal tubule [9]. These transporters facilitate cisplatin accumulation and retention in renal tissues. Meanwhile, mechanistic investigations have revealed that cisplatin‐associated nephrotoxicity arise from the interplay of multiple mechanisms, among which ferroptosis plays a significant role [10, 11, 12]. Ferroptosis is an iron‐dependent regulated cell death modality, characterized by dysregulated lipid peroxidation and intracellular redox dyshomeostasis [13]. The ferroptotic cascade is featured with three hallmarks: (i) accumulation of ferrous ions, (ii) peroxidation of phospholipids containing polyunsaturated fatty acyl moieties, and (iii) compromised glutathione peroxidase 4 (GPX4)‐mediated elimination of lipid peroxides [14]. It has been shown that less than 10% of administered cisplatin forms covalent adducts with nuclear DNA, while the majority undergoes preferential coordination with intracellular sulfur nucleophiles, including low‐molecular‐weight thiols (glutathione) and cysteine‐rich proteins (metallothioneins) [15, 16]. Glutathione depletion could inactivate GPX4, the master regulator of ferroptosis. Moreover, cisplatin simultaneously triggers ferritinophagy‐mediated lysosomal degradation of iron storage proteins, thereby elevating Fe²⁺ concentration in the labile iron pool (LIP) through lysosomal ferritin degradation [17]. The surging Fe²⁺ facilitates the Fenton reaction, producing hydroxyl radicals that propagate a non‐enzymatic lipid peroxidation cascade through targeting the bis‐allylic protons in membrane phospholipids [18]. The combined depletion of cellular reducing equivalents and Fe²⁺‐mediated free radical generation results in a self‐amplifying cycle of redox imbalance, simultaneously driving the oxidative stress cascade and ferroptotic cell death. Therefore, the mitigation of cisplatin‐induced nephrotoxicity can be achieved through reducing intracellular cisplatin concentration in renal tubular cells, coupled with concurrent ferroptosis suppression [12]

Nanoscale drug delivery systems offer a viable approach to modulate drug biodistribution and pharmacokinetic profiles, thereby enhancing therapeutic efficacy while minimizing the off‐target effects [19]. Current clinical investigations highlight that polymer‐based micellar formulations (NC‐6004) and liposomal cisplatin analogs (Lipoplatin) can enhance tumor growth inhibition through optimized biodistribution and concomitant attenuation of nephrotoxicity [20, 21]. Additionally, the intracellular Fe²⁺ level in LIP is directly correlation with the extent of lipid peroxidation. Therefore, targeting LIP homeostasis and suppressing ferroptosis offer a promising therapeutic approach to mitigate cisplatin‐associated nephrotoxicity. Iron‐chelating agents and radical scavengers have emerged as primary ferroptosis inhibitors, and their renoprotective efficacy has been approved in various disease models [22, 23]

In the current work, it was postulated that the employment of iron‐chelating vehicles could achieve efficient cisplatin loading via vehicle‐Pt coordination and Fe²⁺‐triggered coordination exchange‐induced cargo release; meanwhile, the iron chelation could further attenuate cisplatin‐induced ferroptosis and nephrotoxicity (Scheme 1). Hence, this work aimed to engineer an amphiphilic iron‐chelating polymeric micelle for concurrent cisplatin activation and nephrotoxicity attenuation. The tailored polymer comprises a block copolymer, methoxyl poly(ethylene glycol)‐poly(glutamic acid), and an iron‐chelating moiety, deferiprone (Dfp), that is, mPEG‐P(Glu‐Dfp). Dfp is a clinically approved iron chelator for the treatment of various overload diseases [24]. The “kill two birds with one stone” strategy in this work offers a promising means for the efficient delivery of cisplatin with reduced adverse effects.

SCHEME 1.

Schematic illustration of polymeric self‐assembling iron chelators load activated cisplatin via reversible coordination, mitigating cisplatin‐induced nephrotoxicity through intracellular iron chelation, and facilitating drug release for tumor suppression. GSH, glutathione; GSSG, Glutathione disulfide; GR, glutathione reductase; GPX4, Glutathione peroxidase 4; PLOOH, phospholipid hydroperoxides; PLOO·, phospholipid peroxyl radicals.

2. Results and Discussion

2.1. Preparation and Characterization of Pt@M_Dfp Micelles

The biodegradable block copolymer (mPEG‐P(Glu)) exhibits a precisely tunable structure, including molar mass, hydrophilic‐to‐hydrophobic segment ratio, and the ease of functional group modification in the side chains. In the current work, mPEG‐P(Glu) and its iron‐chelating counterpart, mPEG‐P(Glu‐Dfp), were synthesized (Scheme S1). The conjugation between mPEG‐P(Glu) and Dfp was achieved via the amination reaction. The successful synthesis of the tailored polymers and their precursors was confirmed by proton nuclear magnetic resonance (¹H NMR) spectroscopy (Figures S1‐S4). The self‐assembly of mPEG‐P(Glu) and mPEG‐P(Glu‐Dfp) produced two types of micelles, M_Glu and M_Dfp, respectively. Loading the cisplatin derivative, cis‐[Pt(NH₃)₂(H₂O)₂](NO₃)₂ [25, 26] into mPEG‐P(Glu) and mPEG‐P(Glu‐Dfp) generated two types of coordination micelles, Pt@M_Glu and Pt@M_Dfp, respectively (Figure 1A). The molecular weights of polymers were analyzed by gel permeation chromatography (GPC) and ¹H NMR. The average degree of polymerization regarding the PGlu block was ca. 20, and the Dfp conjugation efficiency was 100% (Figure 1B, C).

FIGURE 1.

The physicochemical properties of cisplatin‐loaded iron‐chelating micelles. (A) Illustration of polymeric micelles‐aided cytosolic cisplatin delivery. (B) Gel permeation chromatography (GPC) chromatograms and (C) molecular weight of iron‐chelating and control polymers. The Job's plots between activated cisplatin and different ligands: (D) free Dfp (100 µM), (E) mPEG‐P(Glu‐Dfp) (Dfp: 10 µM), and (F) M_Dfp micelles (Dfp: 100 µM). (G) Critical micelle concentration (CMC) of M_Glu and M_Dfp micelles. (H) Hydrodynamic size, (I) transmission electron microscopy (TEM) images, and (J) zeta potential of M_Glu, Pt@M_Glu, M_Dfp, and Pt@M_Dfp micelles. Scale bar: 200 nm. (K) X‐ray photoelectron spectroscopy (XPS) of Pt@M_Dfp micelles. (L) Pt loading in Pt@M_Glu and Pt@M_Dfp micelles. (M) Kinetic stability of Pt@M_Dfp micelles. (N‐P) Cumulative release of Pt²⁺ from Pt@M_Glu and Pt@M_Dfp micelles at different pH conditions (7.4, 6.0, and 5.0) with or without Fe²⁺ (20 µM). Data are presented as mean ± SD (n = 3). Statistical significance was assessed using an unpaired two‐tailed Student's t‐test for panels G, L, and N‐P; one‐way ANOVA followed by Tukey's post hoc test for panels H and J. n.s. indicates no significant difference, *p < 0.05, **p < 0.01, ***p < 0.001.

The Job's plot analysis demonstrated a 2:1 coordination ratio between free Dfp and activated cisplatin, which was the same for free mPEG‐P(Glu‐Dfp) without self‐assembly (Figure 1D, E). However, M_Dfp micelles exhibited a reduced coordination ratio with Pt²⁺ upon micelle formation (< 2:1), possibly because of the restriction of Dfp movement in the core of micelles (Figure 1F) [25]. The critical micelle concentration (CMC) of mPEG‐P(Glu) and mPEG‐P(Glu‐Dfp) was 0.7 ± 0.1 and 0.5 ± 0.1 µM, respectively (Figure 1G) (p = 0.06805). The trend may still suggest that Dfp modification slightly increased the hydrophobicity of the polymer segment, thereby promoting micelle formation to some extent. Irrespective of cargo loading, the hydrodynamic size of all micelles was less than 200 nm, whereas the particle size dramatically decreased post cargo loading (Figure 1H), which could be explained by the coordination or electrostatic interaction‐induced micelle crosslinking. The transmission electron microscopy (TEM) analysis showed a spherical morphology for all micelles, and the surface charge of micelles shifted from negative to neutral upon cargo loading (p < 0.001) (Figure 1I, J).

2.2. Fe²⁺‐triggered cargo release from micelles

Surface elemental composition analysis via X‐ray photoelectron spectroscopy (XPS) confirmed the presence of carbon, oxygen, and platinum in Pt@M_Dfp and Pt@M_Glu, and the valency of Pt was 2 for both micelles (Figures 1K and S10A). Quantitative elemental analysis using inductively coupled plasma mass spectrometry (ICP‐MS) demonstrated identical platinum loading (ca. 7.0%) for both micellar systems (Figure 1L). With the hydrodynamic diameter and polydispersity index (PDI) as the indices, both Pt@M_Dfp and Pt@M_Glu maintained kinetic stability over 2 days (Figures 1M and S10B). The release of activated cisplatin (Pt²⁺) from Pt@M_Glu and Pt@M_Dfp micelles was pH‐dependent, and the percentage of drug release was higher at lower pH values (Figure 1N‐P).

There was no significant difference between Pt@M_Glu and Pt@M_Dfp regarding the degree of cargo release at 48 h under neutral conditions (pH 7.4) (p > 0.05). However, Pt@M_Dfp released more cargo in an acidic environment (pH 6.0 and 5.0) compared to Pt@M_Glu (p < 0.05). The cumulatively released cargo from Pt@M_Dfp increased from 21.7% ± 1.3% (pH 7.4) to 35.8% ± 2.0% (pH 5.0) at 48 h. Likewise, 24.4% ± 1.0% and 28.8% ± 1.7% of Pt²⁺ was released from Pt@M_Glu at pH 7.4 and pH 5.0, respectively. The drug release from Pt@M_Dfp was a consequence of reduced coordination between Pt²⁺ and Dfp upon pH reduction, which was analogous to that between Fe²⁺ and Dfp. In contrast, the mechanism of drug release from Pt@M_Glu resulted from ionization suppression of the carboxyl group (pKa: 4.0) under low‐pH conditions. The Fe²⁺ concentration in the labile iron pool, approximately 20 µM, was employed for the in vitro release experiment. Irrespective of pH, the presence of Fe²⁺ (20 µM) resulted in a dramatically faster cargo release from Pt@M_Dfp compared to that from Pt@M_Glu (p < 0.001) because of the higher Dfp/Fe²⁺ binding affinity than that of Dfp/Pt²⁺. Such behavior was more evident under acidic conditions. Upon Fe²⁺ triggering, the cumulatively released cargo from Pt@M_Dfp was 43.4% ± 3.0% (pH 7.4) and 65.0% ± 3.4% (pH 5.0), respectively.

The coordination exchange from Dfp/Pt²⁺ to Dfp/Fe²⁺ was a thermodynamically favored process. The extent of drug release from Pt@M_Dfp was consistent with the shift in micelle particle size. At pH 7.4, the hydrodynamic diameter of Pt@M_Dfp remained constant over 48 h in the absence of Fe²⁺; a noticeable increase in particle size coupled with a broader size distribution was observed at pH 6.0 and 5.0 (Figure S11A‐C). In contrast, the addition of Fe²⁺ largely induced the size increase of Pt@M_Dfp under neutral conditions, and the same phenomenon was observed at pH 6.0 and 5.0 (Figure S11D‐F). Ultraviolet‐visible (UV–vis) spectrophotometry revealed an enhancement of absorption of Dfp/Fe²⁺ moiety at 400–550 nm upon metal replacement (Figure S11G‐I), suggesting that Fe²⁺ could induce the release of activated cisplatin from Pt@M_Dfp through a coordination exchange mechanism.

To establish the fluorescent probes for intracellular mechanistic studies, Pt‐Dap, Pt‐Bodipy, and M_Dfp‐Rhb were synthesized and characterized by ¹H NMR, with Pt‐Dap and Pt‐Bodipy further confirmed by ESI‐MS (Scheme S2, Figures S5‐S9). To investigate intracellular coordination exchange, Pt‐Bodipy@M_Dfp‐Rhb was employed as a FRET system, in which a decrease in the FRET ratio reflects the dissociation of the donor (Pt‐Bodipy) and the acceptor (M_Dfp‐Rhb). The FRET ratio kinetically decreased, which was more evident upon Fe²⁺ pre‐treatment, indicating enhanced dissociation of the Pt cargo from the Dfp‐containing polymer. Baf‐A1, an inhibitor of endolysosomal acidification, was used to assess the influence of an acidic intracellular environment on cisplatin release in HK‐2 cells. Co‐incubation with Baf‐A1 increased the FRET ratio, suggesting that endolysosomal acidification also facilitates cargo release (Figure S12). Together, these results support that labile Fe²⁺ and acidic endolysosomal conditions promote intracellular Pt release, while the covalently conjugated Dfp remains associated with the polymer and retains its iron‐chelating function.

Lysosomal colocalization of Pt‐Bodipy@M_Dfp and Pt‐Bodipy@M_Glu was also monitored in HK‐2 cells (Figures S13, S14). Pt‐Bodipy@M_Dfp showed substantial colocalization with LysoTracker Deep Red at 2, 4, and 6 h post incubation. As expected, Fe²⁺ pre‐treatment significantly reduced the Pearson's correlation coefficients, indicating the decreased level of Pt‐Bodipy in the lysosome (Figure S15A). In contrast, such a phenomenon was not observed for Pt‐Bodipy@M_Glu (Figure S15B). These formulation‐dependent differences support a Dfp‐mediated Fe²⁺/Pt²⁺ coordination‐exchange process that facilitates intracellular Pt release.

2.3. Iron chelation suppresses ferroptosis and inflammation in HK‐2 cells

HK‐2 renal tubular epithelial cells were used as an in vitro model to validate the hypothesis that Pt@M_Dfp mitigates renal toxicity through iron chelation‐mediated controlled cargo release and ferroptosis inhibition. The cytotoxicity of M_Dfp, M_Glu, and free Dfp was first assessed using the standardized MTT (3‐(4′,5′‐dimethylthiazol‐2′‐yl)‐2,5‐diphenyl tetrazolium bromide) assay. The cells maintained viability above 90% upon incubation with formulations at concentrations up to 60 µM (Figure S16A‐C). Both free Dfp and mPEG‐P(Glu‐Dfp) at a dose of 50–200 µM equivalent Dfp could efficiently rescue the HK‐2 cells from cisplatin (15 µM)‐induced cytotoxicity (Figure S16D, E). Because of the controlled cargo release, both micelles were less toxic than the free cisplatin in HK‐2 cells; Pt@M_Dfp treatment resulted in a higher cell viability than Pt@M_Glu at the same platinum concentration (Figure 2A), which was in line with the live/dead cell staining (Pt: 15 µM) (Figure S16F). Such a difference was presumed as a consequence of Dfp‐induced iron chelation in Pt@M_Dfp because ferrous ions play an essential role in ferroptosis initiation and propagation.

FIGURE 2.

Tailored micelles mitigate cisplatin‐induced lipid peroxidation and inflammation in HK‐2 cells. The HK‐2 cells were treated with culture medium (Control), cisplatin (Pt), Pt@M_Glu, and Pt@M_Dfp micelles for 6 h or 24 h, followed by further analysis. (A) The viability of HK‐2 cells in response to formulation treatment. The intracellular (B) malondialdehyde (MDA), (C) 4‐hydroxynonenal (4‐HNE), (D) reduced nicotinamide adenine dinucleotide phosphate (NADPH), (E) reduced glutathione (GSH), (F) ratio between GSH and its oxidized form (GSSG), (G) reduced thioredoxin (trx), and (H) total thiol level post sample treatment. (I) Fluorescent imaging of lipid peroxides by liperfluo probe (E_x/E_m: 514 nm/535−650 nm) and (J) quantification of fluorescence intensity of liperfluo. Confocal imaging of (K) reactive oxygen species (ROS) and (L) intracellular Fe²⁺ by DCFH‐DA probe (E_x/E_m: 488 nm/505‐550 nm) and FerroOrange (E_x/E_m: 543 nm/570‐620 nm), respectively, scale bar: 25 µm. The intracellular quantification of inflammation‐related cytokines post formulation treatment, including (M) prostaglandin‐endoperoxide synthase 2 (PTGS2), (N) interleukin‐6 (IL‐6), (O) tumor necrosis factor‐alpha (TNF‐α), and (P) interleukin‐10 (IL‐10). All imaging data were collected post treatment for 6 h, and all other data were collected post formulation incubation for 24 h. Data are presented as mean ± SD (n = 4 for panel A and n = 3 for panels B‐P). Statistical significance was assessed using one‐way ANOVA followed by Tukey's post hoc test. n.s. indicates no significant difference, *p < 0.05, **p < 0.01, ***p < 0.001.

Pt@M_Glu and Pt@M_Dfp significantly reduced the intracellular level of malondialdehyde (MDA) and 4‐hydroxynonenal (4‐HNE), the end‐products of lipid peroxidation (Figure 2B, C). The potency of Pt@M_Dfp was superior to Pt@M_Glu regarding ferroptosis inhibition because Pt@M_Dfp dually showed the ability of controlled release and iron chelation, whereas Pt@M_Glu was not capable of chelating Fe²⁺. Similarly, Pt@M_Dfp and Pt@M_Glu markedly increased the level of canonical ferroptosis biomarkers in HK‐2 cells, including the reduced nicotinamide adenine dinucleotide phosphate (NADPH), reduced glutathione (GSH), the ratio of GSH against its oxidized form (GSSG), the reduced thioredoxin (trx), and the total thiol (Figure 2D‐H). Cisplatin incubation could significantly boost intracellular lipid peroxides and reactive oxygen species (ROS) in HK‐2 cells; Pt@M_Dfp and Pt@M_Glu robustly suppressed the generation of lipid peroxides and ROS, and the potency of Pt@M_Dfp was higher than that of Pt@M_Glu (Figures 2I‐K and S17A). Cisplatin‐induced ferroptosis and oxidative stress were associated with elevated cytosolic Fe²⁺ (Figures 2L and S17B), consistent with a previous report [27]. The corresponding mechanism may be upregulation of heme oxygenase‐1, which further degrades heme to increase cytosolic iron concentration [28]. Pt@M_Dfp, other than Pt@M_Glu, notably restrained the cisplatin‐induced Fe²⁺ escalation, which concurred well with its potency in suppressing lipid peroxidation and oxidative stress.

The micellar vehicle could also effectively mitigate cisplatin‐induced inflammation in HK‐2 cells. There was a significant reduction of intracellular inflammation biomarkers post Pt@M_Dfp and Pt@M_Glu treatment compared to the free drug group, including prostaglandin‐endoperoxide synthase 2 (PTGS2), tumor necrosis factor alpha (TNF‐α), and interleukin‐6 (IL‐6) (Figure 2M‐O). Consistently, the expression of the anti‐inflammatory cytokine, interleukin‐10 (IL‐10), was upregulated in cisplatin‐incubated HK‐2 cells by Pt@M_Dfp and Pt@M_Glu (Figure 2P). There exists a positive loop between oxidative stress and inflammation [29]. The anti‐inflammatory role of Pt@M_Dfp may partly arise from its ability to repress lipid peroxidation by controlling cargo release and chelating redox‐active Fe²⁺. By contrast, Pt@M_Glu exhibited compromised potency in blunting ferroptosis and inflammation compared to Pt@M_Dfp due to its lack of iron‐chelating capability.

Fer‐1 increased the viability of cisplatin‐treated HK‐2 cells in a dose‐dependent manner, with 30 µM showing the strongest rescue effect (Figure S18). Fer‐1 and Dfp also partially restored intracellular GSH levels and reduced MDA accumulation compared with cisplatin alone (Figure S19). Consistently, cisplatin decreased the expression of GPX4, SLC7A11, and FTH1, whereas Pt@M_Dfp markedly restored these ferroptosis‐related proteins compared with Pt@M_Glu (Figure S20). Together with the observed changes in labile Fe²⁺, ROS, and lipid peroxides, these results further support the involvement of ferroptosis in cisplatin‐induced HK‐2 cell injury and the ferroptosis‐suppression effect of Pt@M_Dfp.

2.4. Iron‐Chelating Micelles Enhance The Antitumor Efficacy of Cisplatin

The murine breast cancer cell line (4T1) was selected as the model. M_Dfp and M_Glu almost showed no effect on the viability of 4T1 cells below 100 µM (Figure S21A,B). The canonical ferroptosis inhibitors (ferrostatin‐1, vitamin E, and Dfp) at the recommended dose could not enhance the cytotoxicity of cisplatin in 4T1 cells (Figure S21C,D) [30], indicating ferroptosis was not primarily involved in cisplatin‐induced cell death in this model. All three cisplatin formulations demonstrated dose‐dependent cytotoxicity against 4T1 cells after 24 h of incubation, and the micellar formulation were less toxic than the free drug because drug release from micelles is required (Figure S21E). After 48 h of incubation, the cytotoxicity of these three formulations was similar in 4T1 cells (Figure 3A). Because apoptosis is the primary mechanism of cisplatin‐induced tumor cell death, we assessed apoptotic biomarkers in 4T1 cells post formulation treatment. There was no significant difference regarding the activity of caspase 9 and caspase 3, as well as the concentration of cytochrome c in formulation‐treated 4T1 cells (p > 0.05) (Figure 3B‐D). Mitochondrial cytochrome c release initiates apoptosis by binding apoptotic protease‐activating factor‐1 (Apaf‐1) and caspase 9, ultimately activating the executioner caspase 3 [31]. Flow cytometry and live‐dead staining analyses showed that micellar formulations exhibited the same potency as free cisplatin, and the iron‐chelating moiety did not affect cisplatin's anticancer potency at the cellular level (Figures 3E,F, and S22).

FIGURE 3.

Effects of iron chelation on the antitumor efficacy of cisplatin in vitro and in vivo. (A) The cytotoxicity post treatment with different formulations, including Pt, Pt@M_Glu, and Pt@M_Dfp micelles. Determination of caspase 9 activity (B), caspase 3 activity (C), cytosolic cytochrome c (Cyt c) concentration (D) in 4T1 cell treated by different formulations. (E, F) The flow cytometry analysis of 4T1 cell apoptosis post formulation treatment. The dose of Pt was fixed at 20 µM for the in vitro assay. (G) Kinetic tumor growth inhibition curves, (H) excised tumor weights at the end of the efficacy study, (I) Kaplan‐Meier survival curves, and (J) Kinetic variation of mice body weight during the course of treatment. Histological (K) and apoptotic (L) staining of the tumor tissues at the end of the efficacy study. The mice were treated by different formulations (PBS, Pt, Pt@M_Glu, and Pt@M_Dfp groups; the Pt dose was fixed at 5 mg kg⁻¹). Scale bar: 50 µm (H&E: hematoxylin and eosin); 25 µm (TUNEL: terminal deoxynucleotidyl transferase dUTP nick end labeling). Data are presented as mean ± SD (n = 5 for panels A and H, n = 3 for panels B‐F, and n = 6 for panels G, I‐L). Panels B‐D, F, and H were analyzed by one‐way ANOVA followed by Tukey's post hoc test; panels G and J were analyzed by two‐way repeated‐measures ANOVA followed by Bonferroni‐adjusted pairwise comparisons among treatment groups at day 10; and panel I was analyzed by pairwise log‐rank tests with Holm‐Bonferroni correction. n.s. indicates no significant difference, *p < 0.05, **p < 0.01, ***p < 0.001.

We further assessed the in vivo antitumor efficacy of cisplatin‐loaded coordination micelles using a subcutaneous 4T1 tumor mouse model, in which the efficacy of renal protection of tailored micelles could be simultaneously assessed (Figure S23). All formulations were administered intravenously, and the cargo dose was kept identical. Regarding tumor growth inhibition, Pt@M_Glu and Pt@M_Dfp were more potent than free cisplatin, consistent with the previously reported nanoscale cisplatin delivery [32]. Intriguingly, Pt@M_Dfp was superrior to Pt@M_Glu in terms of the in vivo antitumor efficacy (Figure 3G, H), consistent with the faster Fe²⁺‐triggered cargo release (Figure 1N‐P). The tumor mass at the end of the efficacy experiment and mouse survival rate analysis further corroborated the robust antitumor efficacy of iron‐chelating micelles in vivo; the survival rates of the Pt@M_Dfp‐treated mice remained above 30% over 45 days, whereas all mice in the other groups died by this time point (Figure 3I). The interplay between antitumor potency and the adverse effects of formulations affects the health status of mice. Free cisplatin exhibited significant side effects and poor antitumor efficacy, resulting in lower body weight in mice compared to Pt@M_Dfp (Figure 3J). Unsurprisingly, mice treated with the negative control (phosphate‐buffered saline/PBS) exhibited the highest body weight, which could be explained by the uncontrolled tumor growth. We also employed the hematoxylin and eosin (H&E) staining together with the terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining to analyze the tumor tissue at the end of the efficacy experiment [33]. Both the histological and immunofluorescent images demonstrated the highest potency of Pt@M_Dfp among all formulations in terms of apoptosis induction and tumor eradication (Figure 3K, L). To further evaluate Pt release in vivo, plasma total Pt and free Pt were quantified after intravenous administration of cisplatin (Pt), Pt@M_Glu, or Pt@M_Dfp. Both micellar formulations maintained higher total plasma Pt concentrations than free cisplatin, indicating prolonged systemic circulation. In contrast, the free Pt remained markedly lower at the early time points, followed by gradual release over time (Figure S24). These results support the controlled in vivo release of cisplatin from the micellar formulations.

2.5. Iron Chelation Suppresses Cisplatin‐Induced Ferroptosis In Vivo

At the end of the efficacy experiment, we quantified ferroptosis biomarkers in the kidney tissue of mice. Free cisplatin treatment significantly increased the indices of lipid peroxidation (MDA and 4‐HNE) (p < 0.001); both types of micelles could significantly reduce the end‐products of lipid peroxides, with Pt@M_Dfp being more potent than Pt@M_Glu (p < 0.05) (Figure 4A, B). Consistently, the intracellular antioxidant markers in the kidney (e.g., GSH, trx, total thiols, and NADPH) dramatically dropped post cisplatin dosing (p < 0.001), and such trend was markedly reversed by Pt@M_Dfp and Pt@M_Glu (Figure 4C‐F). Consistently, Pt@M_Dfp was superior to Pt@M_Glu in terms of ferroptosis suppression in the kidney of mice (p < 0.05). The in vivo data concurred well with the in vitro analysis (Figure 2). The ability of iron‐chelating excipients and systems to mitigate lipid peroxidation has been previously demonstrated [23, 34].

FIGURE 4.

Iron chelation mitigates cisplatin‐induced renal lipid peroxidation in vivo. Quantitative analysis of malondialdehyde (MDA) (A), 4‐hydroxynonenal (4‐HNE) (B), reduced glutathione (GSH) (C), reduced thioredoxin (trx) (D), total thiols (E), reduced nicotinamide adenine dinucleotide phosphate (NADPH) (F), interleukin‐6 (IL‐6) (G), prostaglandin‐endoperoxide synthase 2 (PTGS2) (H), and interleukin‐10 (IL‐10) (I) in the kidney tissue of mice treated by different formulations, including phosphate‐buffered saline (PBS), free cisplatin (Pt), and two types of micelles (Pt@M_Glu and Pt@M_Dfp). The plasma level of MDA (J), 4‐HNE (K), reduced trx (L), and total thiols (M) was also analyzed. (N) Prussian blue staining of the kidney tissue (scale: 50 µm). The sampling time was the 5^th day post‐dosing for all samples. Data are presented as mean ± SD (n = 6). Statistical significance was assessed using one‐way ANOVA followed by Tukey's post hoc test. n.s. indicates no significant difference, *p < 0.05, **p < 0.01, ***p < 0.001.

Ferroptosis is often linked with inflammation, forming a positive loop [29, 35]. Free cisplatin substantially boosted the pro‐inflammatory cytokines (e.g., IL‐6 and PTGS2) and downregulated IL‐10, a critical anti‐inflammatory cytokine (p < 0.01) (Figure 4G‐I). Likewise, both micellar formulations ameliorated the inflammatory phenotype, presumably as a consequence of controlled cargo release. Pt@M_Dfp showed greater anti‐inflammatory activity than Pt@M_Glu, which corresponded to its inherent iron‐chelating and ferroptosis‐suppressing function capability. The quantitative analysis of representative biomarkers in plasma (e.g., MDA, 4‐HNE, trx, total thiols, and PTGS2) corroborated the superior potency of Pt@M_Dfp in suppressing ferroptosis and inflammation compared to Pt@M_Glu (Figures 4J‐M and S25), which was consistent with the indices observed in the kidney.

We further employed Prussian blue staining to analyze ferric iron deposition in kidney tissue post formulation treatment (Figure 4N). This technique makes labile iron visible through the reaction between potassium ferrocyanide and ferric ions, resulting in the formation of an insoluble bright blue pigment [36]. Of note, free cisplatin increased renal iron deposition, which was reversed by both micellar formulations. As expected, Pt@M_Dfp exhibited a higher potency than Pt@M_Glu due to the presence of an iron‐chelating moiety. Previous work reported the renal accumulation of hemopexin in a cisplatin‐induced kidney injury model, leading to an increased hemoglobin deposition on proximal tubular cells and elaboration of iron toxicity [37]. Apparently, the employment of a hydrophilic iron chelator, deferoxamine, attenuated the cisplatin‐induced nephrotoxicity. The data in the current work agreed well with the published work.

2.6. Iron Chelation Mitigates Cisplatin‐Induced Nephrotoxicity In Vivo

The primary manifestations of cisplatin‐associated nephrotoxicity were tubular injury and decreased glomerular filtration rate. Urine and plasma samples were collected on the fifth day post dosing [3]. We first assessed key renal tubular injury markers in plasma, that is kidney injury molecule 1 (Kim⁻¹) and neutrophil gelatinase‐associated lipocalin (NGAL) (Figure 5A, B) [38]. The essential biochemical indicators of kidney injury, such as blood urea nitrogen (BUN), plasma creatinine, and cystatin C, were also quantified (Figure 5C‐E). Urea nitrogen is a waste product associated with protein degradation and can be removed by the kidneys. Creatinine is also a waste compound resulting from the breakdown of creatine stored in muscles, and the level of creatinine is a reflection of the estimated glomerular filtration rate (eGFR). Similarly, the cystatin C test is another method for estimating glomerular filtration rate (eGFR). As expected, free cisplatin significantly boosted the above five biomarkers, indicating severe tissue injury (p < 0.001). Despite the absence of data regarding Pt distribution in key organs, Pt@M_Glu significantly attenuated the level of these markers, which was presumed as a consequence of low level of free Pt, as evidenced by the pharmacokinetic study (Figure S24). Similar renoprotective effects have been reported for other cisplatin nanocarriers [39, 40, 41, 42].

FIGURE 5.

Iron chelation mitigates cisplatin‐induced nephrotoxicity in vivo. Quantitative analysis of kidney injury molecule‐1 (Kim‐1) (A), neutrophil gelatinase‐associated lipocalin (NGAL) (B), blood urea nitrogen (BUN) (C), creatinine (D), cystatin C (E), interleukin‐6 (IL‐6) (F), tumor necrosis factor‐α (TNF‐α) (G), and interleukin‐10 (IL‐10) (H) in mice plasma. Quantitative analysis of Kim‐1 (I), NGAL (J), N‐acetyl‐β‐D‐glucosaminidase (NAG) (K), urine albumin‐to‐creatinine ratio (UACR) (L), and albumin (Alb) (M) in the mice's urine. Hematoxylin and eosin (H&E) (N) and Masson (O) staining of the mouse kidney (scale: 50 µm). All samples were collected on the 5^th day post dosing. Data are presented as mean ± SD (n = 6). Statistical significance was assessed using one‐way ANOVA followed by Tukey's post hoc test. n.s. indicates no significant difference, *p < 0.05, **p < 0.01, ***p < 0.001.

Importantly, Pt@M_Dfp exhibited a better potency of renal protection against cisplatin‐induced injury than the counterpart (Pt@M_Glu) because the former not only acted as a vehicle for cargo delivery, but also a therapeutic material for iron chelation and ferroptosis suppression. Cisplatin nephrotoxicity is usually characterized by elevated inflammation, with upregulation of pro‐inflammatory IL‐6 and TNF‐α, and downregulation of anti‐inflammatory IL‐10 in the plasma (p < 0.001) (Figure 5F‐H). The positive loop between ferroptosis and inflammation has been well reported previously [35, 38]. Pt@M_Dfp and Pt@M_Glu significantly attenuated the cisplatin‐induced inflammation, presumably as a result of minimized intracellular level of cisplatin via sustained release (p < 0.05). Consistently, the superior anti‐inflammatory potency of Pt@M_Dfp was due to its dual action mode that is delivery vehicle and ferroptosis inhibition.

We further analyzed the biomarkers of kidney injury in urine, including Kim⁻¹, NGAL, N‐acetyl‐𝛽‐glucosaminidase (NAG), urinary albumin/creatinine ratio (UACR), and albumin (Alb) (Figure 5I‐M). As a sensitive indicator of kidney injury, NAG is a proximal tubule lysosomal enzyme that breaks down glycosylated bonds and cleaves N‐acetylglucosamine from glycoproteins [43]. The urinary albumin and UACR are valuable tools for detecting kidney injury, particularly at the early stage. The effects of free cisplatin and two micelles on these urinary biomarkers were consistent with the plasma indices, further demonstrating the efficacy of Pt@M_Dfp in attenuating cisplatin‐induced nephrotoxicity.

Histological analysis of renal tissue via hematoxylin and eosin (H&E) staining revealed distinct morphological alterations among experimental groups (Figure 5N). The cisplatin‐treated cohort exhibited significant structural damage characterized by compromised glomerular architecture and renal tubular degeneration. Renal specimens from the Pt@M_Glu intervention group displayed mild apical cytoplasmic vacuolization in tubular epithelial cells. Notably, both the PBS group and the Pt@M_Dfp‐treated group preserved the glomerular morphology and intact renal tubular epithelium. Furthermore, Masson staining was used to evaluate fibrosis in the renal tissue. Significant fibrosis was observed in the renal tissue of mice treated with free cisplatin compared to the PBS control (Figure 5O). Pt@M_Glu partially alleviated fibrosis, whereas fibrosis was not visible in the Pt@M_Dfp group.

Collectively, these data proved the superior protection ability of iron‐chelating micelles against cisplatin‐induced renal damage. Ferroptosis is increasingly recognized as an important mechanism underlying cisplatin‐induced nephrotoxicity [12]. Previous work has employed a diverse range of active agents to suppress ferroptosis and alleviate cisplatin‐induced kidney injury [44, 45, 46, 47]. In contrast to these ferroptosis inhibitor co‐delivery approaches, the current work developed a polymeric iron‐chelating vehicle for concurrent cisplatin delivery and ferroptosis regulation. The key to this design is the stability difference between the Dfp‐iron and Dfp‐cisplatin complexes, enabling coordination‐exchange‐triggered cargo release and iron sequestration‐induced ferroptosis inhibition.

3. Conclusion

Cisplatin has been associated with severe adverse effects, particularly kidney injury, limiting its wide application in oncology. Ferroptosis has emerged as an important contributor to cisplatin‐induced nephrotoxicity, providing a potential target for reducing the off‐target effects of cisplatin. Here, we report an iron‐chelating polymeric micelle for this purpose. Based on the differentiated coordination affinity between Dfp/Pt²⁺ and Dfp/Fe²⁺, and the elevated iron level in renal tubular cells post cisplatin dosing, the micellar system could sustain cargo release to reduce free drug concentration in the cytosol. Meanwhile, this process could attenuate ferroptosis by sequestering labile iron and mitigating non‐enzymatic lipid peroxidation. The proof of concept was successfully demonstrated in cisplatin‐treated HK‐2 cells and mouse models via quantitative analysis of indices related to ferroptosis, inflammation, fibrosis, and renal damage. Intriguingly, the controlled cisplatin delivery and attenuation of renal toxicity significantly increased the antitumor efficacy in 4T1 tumor‐bearing mice. A limitation of this study is the lack of organ‐level Pt biodistribution data, which merits further investigation. The current work sets up a paradigm for addressing the adverse effects of cisplatin through the coordination exchange approach, which may be employed for the enhanced delivery of cisplatin.

Author Contributions

Tiantian Chen: Conceptualization, methodology, investigation, formal analysis, data curation, and visualization. Panyu Du, Yuning Liu, and Xuefei Chen: Investigation, sample analysis, and data curation for the animal studies. Zheng Wang: Methodology and resources. Xin Li: Investigation, visualization and funding acquisition. Xiujie Zhao: Supervision and writing – review and editing. Yanjun Zhao: Conceptualization, project administration, supervision, funding acquisition, and writing – review and editing. All authors reviewed and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (22175130) and the Tianjin Natural Science Foundation (24JCQNJC01740).

Ethics Statement

All animal studies were conducted in accordance with the applicable institutional guidelines for the care and use of laboratory animals. The in vivo antitumor efficacy, survival, and nephrotoxicity studies were reviewed and approved by the Ethics Committee of Tianjin Nankai Hospital, China (Approval No. NKYY‐DWLL‐2023‐195). The additional in vivo pharmacokinetic study was reviewed and approved by the Animal Ethical and Welfare Committee (AEWC) of Tianjin Jinke Bona Biotechnology Co., Ltd., China (Approval No. GENNK‐20260156). All animal procedures were performed under the respective approved protocols.

Statistical Analysis

All data were presented as the mean ± standard deviation (SD) without pre‐processing. All data were checked in terms of normality (Kolmogorov–Smirnov test) and homogeneity of variances (Levene's test) prior to analysis. For comparisons between two independent groups, an unpaired two‐tailed Student's t‐test was employed. For comparisons among three or more independent groups, one‐way analysis of variance (ANOVA) followed by Tukey's post hoc multiple‐comparisons test was employed. For longitudinal tumor‐volume and body‐weight data, two‐way repeated‐measures ANOVA (treatment × time) was used, with Greenhouse‐Geisser correction applied when the assumption of sphericity was violated. Bonferroni‐adjusted pairwise comparisons between treatment groups were performed at day 10. For survival analysis, Kaplan‐Meier curves were compared using pairwise log‐rank tests with Holm‐Bonferroni correction. The critical p‐value was set at 0.05. Data organization and preliminary calculations were performed using Microsoft Excel 2024, while statistical analyses and graph preparation were conducted using Origin 2022.

Conflicts of Interest

The authors declare no conflicts of interests.

Supporting information

Supporting File: adhm71575‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.