A transferrin receptor-targeted liposome for iron metabolism regulation-augmented photodynamic therapy in glioblastoma

Abstract

The clinical standard treatments for glioblastoma (GBM) present a therapeutic challenge of high recurrence. 5-aminolevulinic acid (5-ALA)-based photodynamic therapy (PDT) offers a promising alternative owing to the selective biosynthesis of protoporphyrin IX (PpIX) in tumor mitochondria with favorable phototoxicity. However, the highly elevated iron metabolism leads to low biosynthetic efficiency and insufficient accumulation of PpIX, thus limiting the PDT efficacy. Here, we developed a transferrin receptor (TfR)-targeted liposome (AD@LST) that co-delivers 5-ALA and the iron chelator deferoxamine (DFO) for tumor-targeted PpIX biosynthesis and augmented PDT in GBM. AD@LST can efficiently penetrate the blood-brain barrier, and achieve tumor-selective delivery through TfR-mediated uptake and glutathione-responsive drug release. DFO chelates iron ions to downregulate ferrochelatase expression, while concurrently upregulating coproporphyrinogen oxidase and aminolevulinic acid synthase 1 expression, thereby increasing PpIX above physiological concentration (2.54-fold enhancement) in orthotopic GBM. Furthermore, iron deprivation by AD@LST counteracts heme-mediated antioxidant defenses and reduces mitochondrial oxygen consumption, markedly further improving PDT efficacy. In subcutaneous GBM models, AD@LST enables fluorescence imaging-guided repeated PDT with efficient tumor suppression. This liposome demonstrates great potential in clinical translational applicability.

Graphical abstract

We developed a transferrin receptor (TfR)-targeted liposome (AD@LST) to specifically regulate the elevated iron metabolism, achieve tumor-selective protoporphyrin IX (PpIX) biosynthesis, and simultaneously alleviate hypoxia while disrupting redox homeostasis, ultimately enabling fluorescence imaging-guided repeated photodynamic therapy (PDT) of glioblastoma (GBM).

Highlights

• •A transferrin receptor (TfR)-targeted liposome (AD@LST) was developed to co-deliver 5-aminolevulinic acid (5-ALA) and the iron chelator deferoxamine (DFO).
• •AD@LST chelates iron to augment protoporphyrin IX (PpIX) accumulation by bidirectionally regulating heme biosynthesis enzymes (upregulating ALAS1/CPOX and downregulating FECH).
• •Iron depletion enhances photodynamic therapy (PDT) efficacy by alleviating tumor hypoxia (disruption of the electron transport chain) and disrupting redox homeostasis (reduction of heme/CO/HO-1/NADPH levels).
• •AD@LST increases PpIX concentration and prolongs intratumoral retention, enabling single administration and fluorescence imaging-guided repeated PDT of glioblastoma (GBM).

1. Introduction

Glioblastoma (GBM) remains one of the most lethal solid malignancies, characterized by pervasive infiltration, profound resistance to conventional therapy and frequent recurrence [[1], [2], [3]]. Despite intensive standard treatment involving maximal surgical resection followed by radiotherapy and chemotherapy, the median survival rarely exceeds 18 months, and fewer than 10% of patients survive beyond five years [4]. This dismal prognosis stems from several fundamental challenges: (i) the infiltrative nature of glioma cells hinders complete surgical removal; (ii) the blood-brain barrier (BBB) limits effective drug delivery; and (iii) the highly adaptive tumor microenvironment compromises treatment efficacy [[5], [6], [7]]. Consequently, the development of more effective strategies to improve therapeutic outcomes and prevent recurrence remains a critical unmet clinical need in GBM treatment.

Photodynamic therapy (PDT) represents a spatially controlled, light-activated therapeutic approach that induces localized tumor damage while sparing the surrounding healthy brain tissue [[8], [9], [10], [11]]. Although various photosensitizers (PSs), including porphyrin, chlorin, and phthalocyanine derivatives, have entered clinical or preclinical development [[12], [13], [14], [15], [16]], their systemic administration often results in nonspecific biodistribution, limiting tumor accumulation and increasing the risk of off-target toxicity [8,17]. Protoporphyrin IX (PpIX), generated endogenously from 5-aminolevulinic acid (5-ALA), presents a promising alternative due to its favorable phototoxic properties and preferential accumulation within the mitochondria of tumor cells [[18], [19], [20]]. Notably, 5-ALA is approved by the U.S. Food and Drug Administration (FDA) for fluorescence-guided surgery in GBM, and 5-ALA-PDT is actively being investigated as a postoperative adjuvant [21]. However, the clinical efficacy of 5-ALA-PDT is severely constrained by inadequate tumor-specific delivery of 5-ALA and the rapid metabolic conversion of PpIX into heme [[21], [22], [23]]. Therefore, the development of strategies to achieve high biosynthetic efficiency and sufficiently intratumoral accumulation of PpIX remains a persistent research priority.

The intracellular concentration of PpIX is governed by two key factors: the delivery efficiency of 5-ALA and its enzymatic conversion rate to PpIX [24,25]. The inherent hydrophilicity of 5-ALA restricts its passive diffusion and limits bioavailability, necessitating the development of nanoparticle-based drug delivery systems to improve tumor targeting [[26], [27], [28], [29]]. GBM cells frequently overexpress transferrin receptors (TfR), a well-established target for tumor therapy, as iron ions are necessary for tumor proliferation. Transferrin (Tf), the natural ligand of TfR, is commonly utilized to achieve TfR-targeted drug delivery [6,30,31]. Furthermore, ferrous ions serve as the essential metal substrate for the rapid enzymatic reaction catalyzed by ferrochelatase (FECH), which converts PpIX into heme [32,33]. This reaction not only depletes the intracellular concentration of PpIX but also enhances the antioxidant capacity of tumor cells, since heme degradation products, such as carbon monoxide (CO), biliverdin/bilirubin, and iron ions, possess potent antioxidant properties [34,35]. Thus, iron metabolism programming can be exploited to modulate this balance. Overall, TfR targeting may enable precise recognition of tumor cells with active iron metabolism and facilitate specific and efficient intracellular iron metabolism programming.

In this study, we developed a Tf-decorated, glutathione (GSH)-responsive liposome (AD@LST) co-loaded with 5-ALA and the clinically approved iron chelator deferoxamine (DFO) for targeted PpIX biosynthesis and augmented PDT in GBM (Scheme 1). This system employs multiple synergistic strategies to enhance PDT antitumor efficacy: (i) The liposome facilitates TfR-targeted delivery, enabling efficient transport across the BBB and specific recognition of tumor cells. (ii) Enhanced intracellular delivery of 5-ALA, collaborating with DFO-mediated iron chelation, dramatically increases PpIX accumulation by inhibiting FECH-driven heme formation while concurrently upregulating coproporphyrinogen oxidase (CPOX) and ALA synthase 1 (ALAS1) expression [19]. (iii) Iron chelation by DFO attenuates mitochondrial respiration to alleviate tumor hypoxia and disrupts the redox adaptation conferred by heme degradation products, synergistically amplifying PDT efficiency. These results highlight the great potential of AD@LST liposome for clinical translation in GBM treatment.

Scheme 1.

Schematic illustration of AD@LST for tumor-targeted PpIX biosynthesis and augmented PDT through iron metabolism modulation. (A) Schematic of the targeted delivery strategy of AD@LST to TfR-overexpressing tumor cells. (B) Iron metabolism regulatory mechanism of AD@LST for enhancing mitochondrial PpIX accumulation while reducing heme biosynthesis and its downstream antioxidant metabolites. TfR: Transferrin Receptor; 5-ALA: 5-Aminolevulinic Acid; DFO: Deferoxamine; GSH: Glutathione; PpIX: Protoporphyrin IX; DCM: Dichloromethane; ETC: Electron Transport Chain; HO-1: Heme Oxygenase-1; CO: Carbon Monoxide; COPRO III: Coproporphyrin III; CPOX: Coproporphyrinogen Oxidase; FECH: Ferrochelatase.

2. Results and discussions

2.1. TfR is the ideal target for iron-metabolism reprogramming

Iron metabolism is a critical regulatory axis in oncogenesis [36,37]. GBM is characterized by upregulated TfR expression to enhance iron acquisition, thereby supporting malignant progression [38,39]. Analysis of The Cancer Genome Atlas (TCGA) datasets confirmed consistent TfR upregulation in GBM compared to adjacent normal tissues (Fig. 1A). Notably, GBM was identified as one of the most TfR-addicted malignancies (Figure S1), with TfR expression levels escalating alongside tumor grade (Fig. 1B). Moreover, TCGA-GBMLGG overall survival analysis showed that high TfR expression defines a patient subgroup with markedly inferior survival (hazard ratio (HR) = 5.34; Fig. 1C). These findings were validated by western blotting analysis in GBM cell lines (U87 and U251) and by immunohistochemical analysis in clinical specimens from the Human Protein Atlas (Fig. 1D and E), collectively establishing TfR as a key oncogenic driver and a compelling therapeutic target in GBM.

Fig. 1.

Design and characterization of TfR-targeting and GSH-responsive liposome AD@LST. (A) TfR expression in GBM compared with adjacent normal brain tissue. (B) TfR expression across different GBM grades. (C) TCGA-GBMLGG overall survival analysis of patients stratified by TfR expression. (D) Western blot analysis of TfR in glioma cell lines (U87, U251) and non-neoplastic human embryonic kidney (293T) cells. Relative protein levels were determined by normalizing band intensities to GAPDH and then to the 293T group (293T group = 1.00). (E) Representative immunohistochemical staining for TfR in low-grade and high-grade GBM patient specimens, sourced from the Human Protein Atlas (HPA). (F) Correlation analysis between TfR and FECH expression in the TCGA dataset. (G) TEM image, (H) hydrodynamic size distribution, (I) zeta potential and (J) colloidal stability of AD@LST. GSH-responsive (10 mM) release profiles of (K) 5-ALA and (L) DFO from AD@LST. Data are presented as mean ± SD; n = 3.

Mechanistically, ferrous iron enrichment promotes heme biosynthesis through the iron-dependent activation of FECH, which catalyzes the conversion of PpIX to heme, thereby depleting the concentration of PpIX in 5-ALA-PDT [18,40]. Consistent with this, correlation analyses of GBM transcriptomes revealed significant co-upregulation of TfR with key heme-biosynthesis enzymes: FECH (r = 0.29, p < 0.001), ALAS1 (r = 0.24, p = 0.0017), and CPOX (r = 0.23, p = 0.0037) (Fig. 1F; Figure S2A-B). This pattern suggests a coordinated metabolic program that enhances both iron metabolism and heme biosynthesis. Since heme metabolites serve as important substrates for cellular antioxidant processes [41], targeted disruption of iron-dependent metabolic circuits could confer a triple therapeutic benefit: perturbing malignant-cell metabolism, enhancing PpIX accumulation/retention, and attenuating antioxidant defense.

In this study, DFO, a clinically approved iron chelator, was employed to achieve efficient iron metabolism programming in GBM. To determine the optimal DFO dose for maximal PpIX biosynthesis, we evaluated a series of mass ratios of 5-ALA to DFO (1:0.25, 1:0.5, 1:1, 1:2, 1:4). Analysis of PpIX fluorescence signal intensity using the Operetta CLS™ High Content Analysis System (HCAS) showed that increasing the DFO proportion from 1:0.25 to 1:1 markedly enhanced PpIX production in U87 cells (Figure S3A-B), whereas further increases to 1:2 and 1:4 did not result in additional increase in PpIX levels. The liposome of AD@LST was prepared via a thin-film hydration method [42,43]. The encapsulation parameters of 5-ALA and DFO were further optimized, ultimately identifying a mass ratio of 1:1:2 (5-ALA: DFO: empty LST) for final synthesis (Figure S4A-D). The optimal AD@LST formulation showed encapsulation efficiency (EE) of 49.28 ± 4.41% for 5-ALA and 48.19 ± 2.18% for DFO, with loading efficiency (LE) of 16.57 ± 1.48% for 5-ALA and 16.20 ± 0.73% for DFO, respectively. Transmission electron microscopy (TEM) images revealed a spherical structure of AD@LST (Fig. 1G). Dynamic light scattering (DLS) analysis indicated a hydrodynamic diameter of approximately 140.00 ± 2.45 nm and a surface charge of about −6.11 ± 0.20 mV (Fig. 1H and I). The hydrodynamic diameter remained stable between 140.00 ± 2.45 and 150.67 ± 1.12 nm over seven days in phosphate buffer solution, with the polydispersity index (PDI) showing minimal variation, confirming excellent colloidal stability (Fig. 1J). The 5-ALA/DFO release behavior of AD@LST is crucial for the biosynthesis of PpIX in tumor tissues. Notably, AD@LST presented a GSH-responsive drug release property with 91.25 ± 5.35% of 5-ALA and 86.04 ± 6.04% of DFO released within 1 h (Fig. 1K and L), which was sharply contrasted with the slower release value observed in the absence of GSH condition (36.50 ± 4.51% for 5-ALA and 47.61 ± 2.23% for DFO).

2.2. AD@LST enhances PpIX biosynthesis via TfR targeting and iron metabolism reprogramming

To evaluate the TfR-targeting specificity of AD@LST, we examined its binding affinity to TfR in U87 cells incubated with Dil-labeled AD@LST. Competitive binding assays using excess transferrin (Tf) abolished AD@LST membrane association. This result confirmed TfR-specific recognition, as cells without Tf pre-treatment showed clear red fluorescence signal of Dil on cell membrane, whereas U87 cells pre-treated with excess Tf showed negligible red fluorescence signal of Dil (Fig. 2A). Flow cytometric analysis of fluorescein isothiocyanate (FITC)-labeled AD@LST and AD@LS demonstrated that AD@LST exhibited a time-dependent increase of FITC fluorescence signal in U87 cells over 30 min, which was significantly higher than that of AD@LS, validating that Tf decoration enhances active TfR-mediated uptake (Fig. 2B). These results collectively establish AD@LST as a robust liposome for TfR-targeted delivery.

Fig. 2.

AD@LST enhances tumor cell-selective PpIX accumulation through TfR targeting and programmed iron metabolism. (A) Membrane binding of Dil-labeled AD@LST on U87 cells after 1 h at 4°C with or without excess Tf treatment. (B) Time-dependent cellular uptake of FITC-labeled AD@LST and AD@LS measured by flow cytometry (Orange: AD@LS; Red: AD@LST; Gray: Control). (C) Fluorescence images of cytosolic Fe²⁺ (FerroOrange) and mitochondrial Fe²⁺(Mito-FerroGreen) signals. Corresponding semi-quantitative analysis of (D) FerroOrange and (E) Mito-FerroGreen intensity. (F) Time-dependent PpIX fluorescence images and (G) corresponding semi-quantitative analysis. Quantification by ImageJ software. Data were normalized to the control (3 h A@LST) group (100%). (H) PpIX fluorescence images in a co-culture environment of GFP-labeled 293T and U87 cells treated with AD@LST. (I) Quantitative comparison of PpIX intensity between 293T-GFP and U87 cells. Red: PpIX, Green: 293T-GFP. Quantification by ImageJ software. Data were normalized to the 293T group (100%). (J) Western blot analysis of key heme-biosynthesis enzymes (ALAS1, CPOX, FECH). Relative protein expression was determined by normalizing band intensities to GAPDH and then to the control group (control = 1.00). Data are expressed as mean ± SD (n = 3).

We next assessed the ability of AD@LST to reduce intracellular Fe²⁺ levels by probing labile iron pools using FerroOrange (cytosolic) and Mito-FerroGreen (mitochondrial) indicators. AD@LST-treated U87 cells exhibited significantly lower Fe²⁺ fluorescence signal compared to control and A@LST-treated groups (Fig. 2C). Semi-quantitative analysis showed that Fe²⁺ signals were 8.21 (±0.21)-fold higher in the cytosol and 5.43 (±0.77)-fold higher in the mitochondria of the control group than in AD@LST-treated cells (Fig. 2D and E), confirming the potent iron-sequestering ability of AD@LST.

The effect of iron chelation on the accumulation of PpIX in U87 cells was further characterized. Long-term dynamic monitoring of PpIX fluorescence signal in U87 cells revealed that AD@LST induced higher PpIX accumulation compared to A@LST. Specifically, AD@LST increased PpIX levels by 2.51 (±0.13)-fold relative to A@LST at 15 h (Fig. 2F and G). This enhancement was further validated in A431 cutaneous squamous cell carcinoma cells, where AD@LST treatment led to a 5.3-fold higher PpIX fluorescence intensity compared to A@LST at 12 h (Figure S5A-B). To evaluate the specificity of AD@LST for tumor cells, we co-cultured green fluorescent protein (GFP)-labeled human embryonic kidney (293T) cells with U87 cells. The results showed higher PpIX fluorescence signal in U87 cells than in 293T cells following different treatments with either A@LST (2.23 (±0.21)-fold higher) or AD@LST (3.06 (±0.51)-fold higher) (Fig. 2H and I; Figure S6A-B). These findings indicate that AD@LST selectively recognizes tumor cells with high TfR expression and enhances the targeted PpIX biosynthesis. We further analyzed key enzymes in the heme biosynthesis pathway by Western blot analysis. The data demonstrated that AD@LST upregulated ALAS1 and CPOX while downregulating FECH (Fig. 2J). These results suggest that the amplified PpIX signal may result from disrupted heme synthesis feedback and increased PpIX production due to iron chelation.

2.3. AD@LST suppresses mitochondrial respiration and disrupts redox homeostasis

Indeed, DFO-mediated iron chelation can disrupt iron-sulfur cluster biogenesis and impair mitochondrial respiration [[44], [45], [46]]. Real-time cellular metabolism analysis using the Seahorse platform confirmed that AD@LST suppressed oxygen consumption rate (OCR), reducing basal respiration (58.55 ± 2.17%), maximal respiration (62.06 ± 2.49%), ATP production (56.17 ± 2.22%), and spare respiratory capacity (67.12 ± 3.45%) compared to the control group (Fig. 3A and B). This metabolic reprogramming was accompanied by enhanced glycolysis, as shown by a 1.23 (±0.01)-fold increase in lactate production (Figure S7). Importantly, AD@LST increased the activity of CPOX, an oxygen-sensitive enzyme critical for PpIX biosynthesis [19], by 1.58 (±0.10)-fold relative to the control (Fig. 3C). These results demonstrate that AD@LST reduces tumor oxygen consumption while simultaneously promoting PpIX biosynthesis.

Fig. 3.

AD@LST attenuates mitochondrial respiration, reduces cellular oxygen consumption, and disrupts redox homeostasis. (A) Mitochondrial oxygen consumption rate (OCR) in U87 cells assessed via Seahorse mitochondrial stress assay under varying treatments, with (B) quantitative comparisons of basal respiration (G1), maximal respiration (G2), ATP-linked respiration (G3), and spare respiratory capacity (G4). (C) Quantitative analysis of CPOX enzyme activity. (D) Heme concentration. (E) Western blot analysis of HO-1 expression under different treatments. Relative protein expression was determined by normalizing band intensities to GAPDH and then to the control group (control = 1.00). (F) Representative fluorescence images of COP-1 (carbon monoxide probe-1) in treated U87 cells and (G) corresponding semi-quantitative analysis. Quantification by ImageJ software. Data were normalized to the control group (100%). (H) Phototoxicity assessment of chlorin e6 (Ce6) with or without DFO in U87 cells. (I) Schematic diagram depicting the mechanism of enhanced oxygenation and reduced antioxidant defense following iron chelation. Data are presented as mean ± SD; n = 3.

Beyond PSs accumulation and oxygen availability, antioxidant defenses pose a major barrier to PDT efficacy [[47], [48], [49]]. A reduction in heme levels in PDT is particularly critical because heme metabolites, bilirubin and CO, possess potent free radical scavenging capabilities [34,35]. As shown in Fig. 3D, heme concentration was significantly lower in the AD@LST group (7.96 ± 0.13 μg/mL) than in the A@LST group (8.59 ± 0.11 μg/mL, p = 0.039). Consistent with diminished heme levels, the expression of heme oxygenase-1 (HO-1) and the cellular fluorescence signal of CO detected by COP-1 (CO probe-1) were also significantly reduced, by 34.72% (Fig. 3E) and 67.97 ± 2.97% (Fig. 3F and G), respectively. Furthermore, AD@LST treatment increased the NADP⁺/NADPH ratio by 10.50 (±2.56)-fold compared to the A@LST group (Figure S8). The diminished antioxidant capacity and enhanced oxygen availability following DFO treatment were further validated in combination studies using DFO and chlorin e6 (Ce6) to treat U87 cells (Fig. 3H). Overall, DFO-mediated disruption of iron metabolism and aerobic respiration can lead to increased PpIX accumulation and oxygen availability, while decreasing the generation of antioxidants, thus improving PDT efficiency (Fig. 3I).

2.4. AD@LST amplified mitochondrial and DNA damage for augmented PDT

The low dark toxicity and high phototoxicity of AD@LST are essential for effective PDT. In the absence of laser irradiation, all tested liposomes (0–10 μg/mL) showed negligible cytotoxicity, with cell viability remaining above 90% after 12 h of incubation (Fig. 4A). Under laser irradiation (+), each liposome exhibited concentration-dependent photocytotoxicity. Notably, at an equivalent low dose of 5-ALA (2.5 μg/mL), the viability of U87 cells was 92.87 ± 0.54% for the A@L (+) group and 87.81 ± 1.03% for the A@LS (+) group. This difference may be attributed to GSH-mediated rapid drug release, which facilitates faster and greater biosynthesis of PpIX. A consistent reduction in cell viability was observed for the AD@LS (+) group compared to the A@LS (+) group across all concentrations. This effect was particularly pronounced at the maximum dosage of 5-ALA (10 μg/mL), with cell viabilities of 21.40 ± 2.15% for AD@LS (+) group and 53.12 ± 2.54% for A@LS (+) group (Fig. 4B), indicating that DFO-mediated iron depletion significantly enhances the efficacy of 5-ALA-PDT. Moreover, compared with free 5-ALA + DFO, AD@LST exhibited significantly higher phototoxicity at shortened co-incubation times (3 h: 69.61 ± 0.41% vs 88.87 ± 1.47%; 6 h: 25.86 ± 0.63% vs 39.18 ± 2.75%), confirming the necessity of liposomal delivery for rapid intracellular drug availability (Figure S9). Flow cytometric analysis confirmed that the AD@LST (+) group induced the highest level of apoptosis (85.4%), which is attributed to the synergy of TfR-mediated iron targeting and iron chelation by DFO (Fig. 4C; Figure S10). We next assessed oxidative stress in U87 cells using 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA). After laser irradiation, the AD@LST (+) and AD@LS (+) groups produced elevated ROS levels (470.70 ± 37.42% and 457.96 ± 29.62%, respectively) compared to the A@LS (+) (328.61 ± 12.75%) and the A@L (+) (221.77 ± 12.02%) groups (Fig. 4D and E). These results indicate that DFO potentiates ROS generation in the process of 5-ALA-PDT.

Fig. 4.

AD@LST amplified PDT efficacy in vitro. (A) Dark toxicity and (B) phototoxicity of AD@LST in U87 cells. (C) Apoptosis analysis by Annexin V-FITC/PI staining in U87 cells treated with indicated liposomes with laser irradiation. (D) Representative fluorescence images of DCF (ROS), JC-1 (mitochondrial membrane potential), and γ-H2AX (DNA damage) in treated U87 cells. Semi-quantitative analysis of (E) DCF, (F) JC-1, and (G) γ-H2AX fluorescence intensity. The "+" symbol denotes laser irradiation (635 nm, 200 mW/cm², 3 min). Quantification by ImageJ software. DCF and γ-H2AX data were normalized to the control group (100%). Data are presented as mean ± SD (n = 3).

Mitochondrial depolarization and DNA damage are hallmarks of PDT-induced cytotoxicity. JC-1 ratiometric analysis revealed severe mitochondrial membrane potential (ΔΨm) collapse in the AD@LST (+) and AD@LS (+) treated groups (JC-1 aggregate/monomer ratio = 0.34 ± 0.04 and 0.40 ± 0.02), which was much lower than the A@LS (+) (0.79 ± 0.18) and A@L (+) (1.15 ± 0.01) groups (Fig. 4D–F). Parallel immunofluorescence analysis of DNA damage using phosphorylated histone (γ-H2AX) showed markedly higher fluorescence intensity in the AD@LST (+) and AD@LS (+) groups, increasing by approximately 4.76 (±0.44) and 4.79 (±0.48)-fold relative to the irradiated control group, respectively. In contrast, the A@LS (+) and A@L (+) groups showed milder increases of only about 2.84 (±0.17) -fold and 2.22 (±0.12)-fold (Fig. 4D–G). Without laser irradiation, baseline ROS levels, mitochondrial membrane potential, and DNA damage did not differ significantly among groups (Figure S11–13). These results confirmed the superior PDT antitumor efficacy of AD@LST, while demonstrating its favorable biosafety which may be attributed to the clinically validated components of the liposome.

2.5. AD@LST enhanced intratumoral PpIX accumulation in vivo

Improving the tumor-specific accumulation of PSs remains a central challenge in PDT. We first evaluated the TfR-targeting capacity of AD@LST in U87 tumor-bearing mice. Real-time fluorescence images demonstrated that the accumulation of IR780-labeled AD@LST at the tumor tissues was significantly higher than that of the non-Tf-decorated AD@LS group across multiple time points post-injection. The signal reached its maximum at 12 h, with the IR780-AD@LST group exhibiting a 1.26 (±0.14)-fold higher fluorescence intensity than the IR780-AD@LS group (Fig. 5A and B). This finding aligns with prior cellular evidence of TfR-promoted uptake (Fig. 2A and B). We next quantified the kinetics of PpIX biosynthesis in excised U87 tumors at predetermined time points to minimize interference from overlying tissue signals. Semi-quantitative analysis of PpIX fluorescence intensity revealed that AD@LST exhibited a dual pharmacokinetic advantage: superior initial tumor penetration (1.52 (±0.13)-fold over AD@LS and 1.81 (±0.16)-fold over A@LST at 3 h) and prolonged retention (1.82 (±0.18)-fold over AD@LS and 1.58 (±0.15)-fold over A@LST at 24 h; Fig. 5C and D). The metabolic kinetics of PpIX after AD@LST treatment not only increased PpIX concentration but also significantly extended the therapeutic window, far surpassing that achievable with free 5-ALA, which undergoes rapid systemic clearance with the PpIX fluorescence signal diminishing within 6 h [24].

Fig. 5.

In vivo tumor targeting and pharmacokinetics of PpIX monitored by fluorescence imaging and intravital cranial imaging. (A) Representative in vivo fluorescence images and (B) semi-quantitative analysis of subcutaneous U87 tumor-bearing mice at the indicated time points after intravenous injection of IR780-labeled AD@LST or AD@LS. Images were acquired using an IVIS spectrum imaging system. (C) Representative ex vivo PpIX fluorescence images of tumors and organs, and (D) semi-quantitative intensity analysis, in U87 tumor-bearing mice treated with A@LST, AD@LS, or AD@LST. (E) Representative in vivo PpIX fluorescence images and (F) time-dependent semi-quantification of fluorescence intensity in orthotopic U87 GBM-bearing mice after different treatments with A@LST or AD@LST, acquired using intravital microscopy (IVIM). Quantification by ImageJ software. Data are presented as mean ± SD (n = 3).

Notably, AD@LST enables BBB crossing via TfR-mediated transcytosis. We directly visualized this process via fluorescence imaging of Dil-labeled AD@LST in brain orthotopic U87 tumor-bearing mice through intravital cranial imaging (Figure S14). In orthotopic U87 tumor models, AD@LST boosted PpIX fluorescence signal by 2.54 (±0.28)-fold within 12 h relative to A@LST, and this enhanced fluorescence intensity was sustained and remained detectable over the following 24 h (Fig. 5E and F). The prolonged intratumoral retention of PpIX is attributed to DFO-mediated iron chelation, which disrupts heme biosynthesis and traps PpIX in a therapeutically active status. This prolonged, tumor-specific PpIX accumulation could be clinically significant, as it extends the therapeutic window for PpIX-guided surgical resection and multiple repeated PDT without requiring repeated administration.

2.6. Fluorescence imaging-guided repeated PDT for the eradication of GBM

Given the capacity of AD@LST to overcome critical neuro-oncological barriers, including inadequate BBB penetration and rapid PpIX clearance, we employed AD@LST for fluorescence imaging-guided repeated PDT [24,50].

In subcutaneous U87 tumor models, AD@LST with repeated laser irradiation (6, 12, and 24 h post-injection, Fig. 6A) achieved 98.09 ± 0.63% tumor suppression, significantly surpassing the efficacy of A@LST (58.78 ± 3.32%) and AD@LS (58.56 ± 6.01%) (Fig. 6B–E). This augmented therapeutic outcome likely correlates with sustained intratumoral PpIX retention and persistent oxidative damage. Histopathological analysis further supported this mechanism: AD@LST-treated tumors exhibited pronounced nuclear pyknosis (Hematoxylin and Eosin (H&E) staining images), extensive apoptosis (Terminal deoxynucleotidyl transferase dUTP Nick-End Labeling (TUNEL) assay), and minimal proliferative activity (Kiel-67 (Ki-67) immunohistochemistry) (Fig. 6F).

Fig. 6.

AD@LST enables iron reprogramming and TfR targeting to amplify the repeated PDT in GBM. (A) Schematic of the fluorescence-guided repeated PDT protocol. (B) Individual tumor growth curves, (C) tumor images, (D) mean tumor growth curves, and (E) excised tumor weights on day 14 from U87 xenograft models (n = 5). Statistical significance was assessed using two-tailed Student's t-test. (F) Histopathological and immunohistochemical analysis of tumor tissues: hematoxylin and eosin (H&E) staining, Ki-67 proliferation index (brown nuclear staining), and TUNEL apoptosis assay (green fluorescence signal). (G) Serum biochemical parameters, including alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), albumin (ALB), and total protein (TP), measured on day 14. Treatment groups: G1 (Control), G2 (A@LST), G3 (AD@LS), G4 (AD@LST), G5 (A@LST + laser), G6 (AD@LS + laser), G7 (AD@LST + laser). Laser parameters: 635 nm, 200 mW/cm², 20 min.

Systemic biocompatibility of AD@LST was evidenced by preserved organ histology (Figure S15), stable hepatic and renal function (Fig. 6G and Figure S16), and unchanged body weights (Figure S17). In addition, iron functions as a critical cofactor for heme synthesis and hemoglobin production [51]. Prolonged iron depletion can cause reduced mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), and mean corpuscular volume (MCV). To evaluate this, we performed blood counts in the AD@LST (+) and control groups. The data demonstrated that MCV, MCH, and MCHC levels in the AD@LST (+) group remained comparable to those of the control group (Figure S18). Together, these results demonstrate that AD@LST-mediated PDT is substantially more effective than conventional 5-ALA-PDT. The improvement benefits not only from DFO-driven elevation of tumoral PpIX, but also from enhanced tumor targeting and prolonged retention of AD@LST, which enable repeated PDT after a single administration. Furthermore, because both 5-ALA and DFO are clinically approved agents and the liposomal platform is widely validated for clinical use, this liposome provides a readily translatable therapeutic strategy for eradicating residual GBM cells through spatiotemporally controlled oxidative damage.

3. Conclusion

The standard treatment for GBM, consisting of surgery, radiotherapy, and chemotherapy, faces critical challenges due to high rates of tumor recurrence and low overall survival. Recent clinical efforts are exploring the potential of 5-ALA-PDT in GBM, extending its application beyond fluorescence-guided surgical resection [22,52]. Although clinical studies indicate a therapeutic effect of 5-ALA-PDT, a significant limitation remains the persistent incidence of tumor recurrence following treatment.

Dysregulated iron metabolism constitutes a recognized and targetable vulnerability in tumors [53,54]. In this study, we exploited two key features of GBM pathology: elevated iron metabolism driven by TfR overexpression, and the iron-dependent conversion of PpIX to heme. By designing a targeted strategy that combines TfR recognition with specific iron metabolism modulation, we achieved a multifunctional therapeutic effect, simultaneously enhancing PpIX biosynthesis/accumulation, alleviating tumor hypoxia, and disrupting redox homeostasis. This approach exemplifies a paradigm that actively reprograms tumor metabolism to improve PDT efficacy, providing a framework for future therapies tailored to specific metabolic features. For example, heterogeneity in tumor TfR expression may influence targeting efficiency and patient selection criteria.

A key translational strength of this strategy lies in its components. Both 5-ALA and DFO are clinically approved agents, and the liposomal delivery platform is widely validated for clinical use [[55], [56], [57]]. These attributes confer great translational potential on the AD@LST liposome. However, advancing to clinical application requires addressing practical challenges, including scalable manufacturing, reproducible quality control, and the development of adaptable light-delivery technologies for deep-tissue irradiation.

4. Experimental methods

4.1. Reagents and materials

5-Aminolevulinic Acid Hydrochloride (5-ALA·HCl) was purchased from Shanghai Titan Scientific Co., Ltd. (China). Deferoxamine (DFO), lecithin, FITC, acetic acid-sodium acetate buffer, ethyl acetoacetate, and FeCl₃·6H₂O were obtained from Shanghai Macklin Biochemical Co., Ltd. (China). Cholesterol (CH) was purchased from J&K Scientific Ltd. (China). DSPE-PEG2K, DSPE-SS-PEG2K, DSPE-PEG2K-Tf, and DSPE-SS-PEG2K-Tf were sourced from Xi'an Ruixi Biological Technology Co., Ltd. (China). FerroOrange and Mito-FerroGreen were purchased from Goryo Chemical, Inc. (Japan). DAPI, Cell Counting Kit-8 (CCK-8), Reactive Oxygen Species Assay Kit (DCFH-DA), Dil, Mitochondrial Membrane Potential Assay Kit, NADP⁺/NADPH Assay Kit, Annexin V-FITC/PI Assay Kit and DNA Damage Assay Kit (γ-H2AX Immunofluorescence, Rabbit Monoclonal Antibody, Red Fluorescence) were purchased from Beyotime Biotechnology (China). Chlorin e6 (Ce6, HY-13594) and CO probe-1 (HY-D1244) were purchased from MedChemExpress (USA). IR780 was obtained from Sigma-Aldrich (USA). Hoechst 33342 was obtained from Thermo Fisher Scientific Inc. (USA). Dulbecco's Modified Eagle's Medium (DMEM), fetal bovine serum (FBS), phosphate-buffered saline (PBS), penicillin-streptomycin, and 0.25% trypsin-EDTA were supplied by Gibco® (Thermo Fisher Scientific Inc., USA). Lactic Acid Assay Kit was procured from Nanjing Jiancheng Bioengineering Institute (China). CPOX Enzyme Activity ELISA Assay Kit was sourced from Jiangsu Meimian Industrial Co., Ltd. (China). Heme ELISA Assay Kit was procured from YUANJU Bio (China). Mitochondrial Stress Test Assay Kit was purchased from Agilent Technologies Inc. (USA). The U87, U251, A431 and 293T-GFP cell lines were revived from cryopreserved stocks under standard laboratory protocols.

4.2. Characterizations

Transmission electron microscopy (TEM; HT7700, Hitachi High-Technologies Corporation, Japan) was employed to characterize the morphological features of the synthesized liposomes. The hydrodynamic diameter, polydispersity index (PDI), and zeta potential of the liposomal formulations were measured using a Zetasizer Ultra instrument (Malvern Panalytical Ltd., UK).

4.3. Bioinformatics data acquisition

Comparative analysis of transferrin receptor (TfR) expression in normal versus tumor tissues, TfR-stratified survival analysis using TCGA‑GBMLGG datasets, and gene correlation studies were performed via the online platform GEPIA2 (http://gepia2.cancer-pku.cn/) and Sangerbox (http://sangerbox.com/home.html). Immunohistochemical staining data for TfR were retrieved from the publicly accessible Human Protein Atlas database (https://www.proteinatlas.org/).

4.4. Synthesis of AD@LST

AD@LST was synthesized via a film-ultrasonic dispersion technique. A lipid mixture containing lecithin, cholesterol, DSPE-SS-PEG2K, and DSPE-SS-PEG2K-Tf at a molar ratio of 65:25:7.5:2.5 was dissolved in 3 mL dichloromethane and evaporated under reduced pressure using a vacuum rotary evaporator to form a homogeneous thin film. The film was hydrated with deionized water containing 5-ALA and DFO at a mass ratio of lipids: 5-ALA: DFO = 2 : 1: 1, followed by ultrasonication to generate liposomes. The resulting suspension was extruded through a 0.22 μm polycarbonate membrane (Nuclepore™, Whatman, UK) using a liposome extruder for 10 cycles to obtain large unilamellar vesicles. Unencapsulated drugs were removed via centrifugation with a 30 kDa molecular weight cut-off ultrafiltration device (VIVASPIN® Turbo, Sartorius AG, Germany). Control liposomes were prepared using identical protocols, including four formulations: A@L (5-ALA-loaded with DSPE-PEG2K), A@LS (5-ALA-loaded with DSPE-SS-PEG2K modification), AD@LS (co-loaded with 5-ALA/DFO and DSPE-SS-PEG2K modification), and A@LST (5-ALA-loaded with DSPE-SS-PEG2K-Tf modification). For fluorescence dye labeling, Dil (5 μL), FITC (10 μg) or IR780 (1 mg) was introduced into 1 mL liposomal solution under light-protected conditions, followed by 30 min sonication and subsequent ultrafiltration to remove unbound fluorescence dyes.

4.5. Detection of drug encapsulation efficiency and loading efficiency

Standard calibration curves for 5-ALA and DFO were established to quantify encapsulation efficiency. For 5-ALA analysis, varying concentrations of 5-ALA standard solutions were prepared. Aliquots were mixed with ethyl acetoacetate (0.2 mL) and acetic acid-sodium acetate buffer (pH 5.8, 4 mL), heated at 100°C for 25 min, and subsequently reacted with 4 mL HCl (4 M) and 1 mL Fe³⁺ solution (1 g/L) at 100°C for 30 min. Absorbance was measured at 480 nm after cooling to ambient temperature. For DFO quantification, standard solutions were mixed with Fe³⁺ solution (1 mM), incubated at 37°C for 5 min, and analyzed by measuring absorbance at 430 nm using a microplate reader (Synergy HT; BioTek Instruments, Inc., USA).

The EE and LE for 5-ALA and DFO were calculated using the following equations:

EE = (m_{total5-ALA/DFO}- m_{unloaded5-ALA/DFO})/m_{total5-ALA/DFO}× 100%

LE = (m_{total5-ALA/DFO} - m_{unloaded5-ALA/DFO})/m_totalAD@LST × 100%

Where m_{total5-ALA/DFO} denotes the total mass of 5-ALA or DFO initially added during synthesis, m_{unloaded5-ALA/DFO} represents the mass of unencapsulated drugs quantified in the supernatant, and m_totalAD@LST corresponds to the total mass of the final AD@LST formulation, including encapsulated 5-ALA, DFO, and input lipid components. All parameters were derived from triplicate experimental measurements to ensure statistical reliability.

4.6. In vitro release profiles of 5-ALA and DFO from AD@LST

The release kinetics of 5-ALA and DFO from AD@LST were evaluated using a dialysis bag method under simulated physiological conditions. AD@LST solution (10 mg/mL, 1 mL) was loaded into dialysis bags (2000 Da) and immersed in 5 mL phosphate-buffered saline (PBS, pH 7.4) or PBS containing 10 mM GSH (pH 7.4) under gentle agitation (80 rpm, 37°C). At predetermined intervals (0.25, 0.5, 1, 2, 4, 8, 12, 24, 36, 48 h), 1 mL of release medium was collected and replaced with fresh PBS (with or without GSH) to maintain sink conditions.

4.7. Functional verification at the cellular level

All cellular experiments were conducted using cells cultured in DMEM supplemented with 10% FBS and 1% penicillin-streptomycin under standard conditions (37°C, 5% CO₂). Cells were seeded at densities of 1 × 10⁴ cells/well in 96-well plates, 1 × 10⁵ cells/well in 6-well plates, and 1 × 10⁶ cells/well in confocal dishes.

TfR-dependent cellular uptake was investigated using Dil-labeled AD@LST (Dil-AD@LST). U87 cells were pre-incubated with/without excess Tf (1 h) to saturate membrane-bound TfRs prior to treatment. Subsequent incubation with Dil-AD@LST was performed at 4°C for 1 h to allow surface receptor binding while inhibiting internalization. For nuclear visualization, cells were counterstained with Hoechst 33342 (15 min), and subcellular localization was analyzed using super-resolution microscopy (SIM-Supreme, Guangzhou CSR Biotechnology Co. Ltd., China).

Intracellular ferrous iron (Fe²⁺) levels were assessed using fluorescent probes. U87 cells were treated with DMEM, A@LST, or AD@LST (containing 10 μg/mL DFO) for 24 h, followed by incubation with FerroOrange (1 μM, for cytoplasmic Fe²⁺ detection) or Mito-FerroGreen (5 μM, for mitochondrial Fe²⁺ detection) for 30 min at 37°C in the dark. Fluorescence images were acquired using HCAS (Operetta CLS; PerkinElmer, Inc., USA, FerroOrange: Ex/Em = 561/570–620 nm; Mito-FerroGreen: Ex/Em = 488/500–550 nm).

PpIX fluorescence images were acquired using HCAS. For U87, 1 × 10⁴ cells per well were seeded in 96-well plates and incubated for 12 h. Various ratios of 5-ALA and DFO were then applied, and fluorescence signal was measured after a further 12 h incubation. For treatment groups (A@LST and AD@LST), an equivalent dose of 10 μg/mL 5-ALA was applied, and fluorescence images were recorded at designated time points.

For co-culture experiments, 293T-GFP and U87 cells were seeded together at 5 × 10³ cells per cell type per well in 96-well plates and incubated for 12 h. The cells were then treated with A@LST or AD@LST at a 5-ALA equivalent concentration of 10 μg/mL. After 12 h of incubation, PpIX fluorescence images were acquired using HCAS. The PpIX fluorescence intensity in U87 cells was selectively quantified by gating on GFP-negative regions, whereas 293T-GFP cells were identified by GFP-positive signals, allowing independent assessment of PpIX accumulation in each cell type.

Lactic acid production was quantified in U87 cells following 48 h treatment with designated treatments. Harvested cells (1 × 10⁶ cells/mL) were lysed through three freeze-thaw cycles, and lactate levels in supernatants were determined using a commercial enzymatic assay kit.

Heme concentration was quantified in U87 cells following 12 h treatment with designated treatments. Harvested cells (1 × 10⁶ cells/160 μL) were lysed through three freeze-thaw cycles, and heme concentration in supernatants were determined using a commercial enzymatic assay kit.

NADP⁺/NADPH was quantified in U87 cell lysates after 24 h with designated treatments. Cells were trypsinized, washed with PBS, and analyzed using NADP⁺/NADPH assay kits according to the manufacturer's protocols.

Mitochondrial respiration was evaluated using a Seahorse XFe24 Analyzer (XFe24; Agilent Technologies, Inc., USA). U87 cells were seeded in 24-well plates at 1 × 10⁴ cells/well and incubated for 24 h. Treatments with blank medium, A@LST, or AD@LST (equivalent to 10 μg/mL 5-ALA) were then administered for 24 h. Assays used Seahorse XF DMEM supplemented with 10 mM glucose, 1 mM sodium pyruvate, and 2 mM glutamine. Oxygen consumption rate (OCR) was measured at baseline and following sequential injection of 1.5 μM oligomycin, 2.0 μM FCCP, and 0.5 μM rotenone/antimycin A.

CPOX activity was assessed in U87 cells. Cells were seeded in culture dishes at 5 × 10⁵ cells/dish and incubated for 24 h. Following incubation, cells were treated for 72 h under hypoxia (3% O₂) with fresh medium containing A@LST or AD@LST (5-ALA-equivalent: 10 μg/mL). After treatment, cells were washed with PBS, harvested by trypsinization, and resuspended in PBS at 1 × 10⁶ cells/mL. Cell suspensions were sonicated (4°C), and supernatants were collected. CPOX activity was quantified in supernatants using a commercial enzymatic assay kit according to the manufacturer's protocol.

ROS detection in U87 cells: Cells were seeded in 96-well plates and incubated for 24 h. The medium was replaced with 100 μL fresh medium containing either blank medium (control) or designated formulations (A@L, A@LS, AD@LS, or AD@LST; 10 μg/mL 5-ALA), followed by a 12 h incubation. Cells incubated with 10 μM DCFH-DA in serum‑free medium at 37 °C for 15 min, and then either irradiated with a 635 nm laser (200 mW/cm², 3 min) or maintained in the dark. Intracellular ROS levels were detected by HCAS (Ex/Em = 470/535 nm).

Mitochondrial membrane potential was assessed in U87 cells seeded in 96-well plates (24 h incubation). Cells were treated with 100 μL fresh medium containing blank medium (control) or designated formulations (A@L, A@LS, AD@LS, or AD@LST; 10 μg/mL 5-ALA) for 12 h, followed by 635 nm laser irradiation (200 mW/cm², 3 min) or dark incubation. After 6 h, cells were incubated with JC-1 (1 μg/mL) and Hoechst 33342 (1 μM) in PBS (20 min, 37°C). Following PBS washes, fluorescence images were acquired via HCAS. JC-1 monomer fluorescence signal was detected at excitation/emission wavelengths of 490/530 nm, while JC-1 aggregate fluorescence signal was measured at 525/590 nm.

DNA damage was evaluated in U87 cells treated as above. Post-irradiation (6 h), cells were fixed and stained using a γ-H2AX immunofluorescence kit (anti-γ-H2AX rabbit monoclonal antibody, red-fluorescent conjugate) according to the manufacturer's protocol. Nuclei were counterstained with Hoechst 33342 (1 μM, 15 min). γ-H2AX fluorescence images were acquired using HCAS (Ex/Em = 550/570 nm).

For cytotoxicity assays, the cytotoxicity of free 5-ALA+DFO, A@L, A@LS, AD@LS, and AD@LST was evaluated via a CCK-8 assay. U87 cells seeded in 96-well plates were incubated for 24 h. The medium was discarded, and cells were treated with fresh medium containing serial concentrations of the formulations (equivalent to 0-10 μg/mL 5-ALA and corresponding DFO) for different durations (3, 6 and 12h). Cells were irradiated with a 635 nm laser (200 mW/cm²) for 3 min or incubated in the dark. After an additional 6 h, the medium was replaced with 100 μL fresh medium containing 10% CCK-8 solution. Absorbance at 450 nm was measured using a microplate reader after 1 h of incubation.

For apoptosis detection by flow cytometry, U87 cells were seeded in 6-well plates at a density of 1 × 10⁵ cells per well and incubated for 24 h. The culture medium was then replaced with 1000 μL of fresh medium containing either blank medium (control) or the indicated formulations (A@L, A@LS, AD@LS, or AD@LST; 10 μg/mL 5-ALA), followed by a 12 h incubation. Subsequently, cells were either irradiated with a 635 nm laser (200 mW/cm², 3 min) or kept in the dark. After a 2 h post-irradiation period, cells were harvested and stained with Annexin V-FITC/PI for 15 min in the dark. Apoptosis analysis was performed using flow cytometry (FACSAria IIIu; Becton, Dickinson and Company, USA).

For quantitative cellular uptake analysis, U87 cells were incubated with FITC-labeled AD@LST or AD@LS for different time intervals (0, 5, 10, 20, 30 min). After washing with PBS, cells were harvested, and FITC fluorescence intensity was analyzed by flow cytometry to evaluate the kinetics of TfR-mediated internalization.

Phototoxicity assessment of chlorin e6 (Ce6) with or without DFO was performed in U87 cells. Cells were seeded in 96-well plates and incubated for 24 h. The medium was replaced with fresh medium containing Ce6 alone or Ce6 plus DFO, followed by 12 h incubation. Cells were then irradiated with a 635 nm laser (100 mW/cm², 1 min). After an additional 6 h, cell viability was assessed by CCK-8 assay according to the manufacturer's protocol.

Carbon monoxide (CO) detection in U87 cells: Cells were seeded in 96-well plates and incubated for 24 h. The medium was then replaced with 100 μL of fresh medium containing either blank medium (control) or the designated formulations (A@LST or AD@LST, each supplying 30 μg/mL 5-ALA), followed by a further 12 h of incubation. Intracellular CO levels were assessed by incubating the cells with CO Probe-1 (1 μM) and PdCl₂ (1 μM) in serum-free medium at 37°C for 30 min. Fluorescence images were subsequently acquired by using HCAS (Ex/Em = 435/522 nm).

4.8. Western blotting

Whole-cell extracts were prepared by lysing cells in RIPA buffer (R0010, Solarbio, China). Protein concentrations were quantified using the bicinchoninic acid (BCA) assay. Immunoblotting was performed according to standard protocols using the following primary antibodies: FECH (1:1000, Novus Biologicals, NBP2-33413), ALAS1 (1:1000, Thermo Fisher Scientific, PA5-100995), CPOX (1:1000, Proteintech, 12211-1-AP), TfR1 (1:1000, Abcam, ab108985), HO-1 (1:1000, BioSci, 8604328) and GAPDH (1:1000, Cell Signaling Technology, 2118S). Protein expression levels were visualized using a FluorChem E system (FCE, ProteinSimple, USA).

4.9. Establishment of the mice model

All animal experiments were conducted in accordance with protocols approved by the Animal Ethics and Welfare Committee of Shenzhen University (AEWC-SZU). 5-7-week-old BALB/c-nude mice were obtained from the Guangdong Medical Laboratory Animal Center (Guangzhou, China) and housed under standardized conditions (temperature maintained at ∼22°C, humidity 40–70%, 12 h light-dark cycle) with ad libitum access to autoclaved food and water.

For the U87 subcutaneous tumor model, male BALB/c-nude mice were subcutaneously injected with U87 cells (5 × 10⁶ cells in 100 μL PBS) into the right flank.

Tumor volume was calculated using the formula:

Tumor Volume (mm³) = 0.5 × Tumor Length (mm) × (Tumor Width (mm))².

Intracranial U87-GBM xenograft model preparation.

Male BALB/c nude mice (5–7 weeks old) were used to establish an orthotopic intracranial U87-GBM xenograft model. All surgical instruments and materials were sterilized prior to procedures. Briefly, mice were anesthetized with 1.5% isoflurane (RWD Life Science, Shenzhen, China) and immobilized in a stereotaxic apparatus (68055; RWD Life Science Co., Ltd., China). The intracranial glioma implantation protocol followed methods described in prior studies, summarized as follows: The scalp was disinfected with Betadine and ethanol. A 5-mm midline incision was made along the sagittal suture, and a burr hole was drilled into the right hemisphere (2 mm anterior and 2 mm lateral to the bregma). A suspension solution of 2 × 10⁵ U87-GFP-Luc cells was stereotactically injected into the right cerebral hemisphere at a rate of 400 nL/min. Subsequently, a 3-mm-diameter cranial window was surgically implanted according to the method established by Holtmaat et al. [58]. Tumor growth was confirmed via bioluminescence imaging.

4.10. In vivo fluorescence (FL) imaging

For the in vivo biodistribution analysis of targeting specificity, near-infrared fluorescent dye IR780 was used to label the formulations. Tumor-bearing mice were intravenously administered IR780-AD@LST or IR780-AD@LS, and their distribution was monitored at pre-defined time intervals (2, 4, 8, 12, 24, 48 and 72 h) using an IVIS Spectrum imaging system (IVIS Spectrum; PerkinElmer, Inc., USA).

To analyze the real-time metabolism of 5-ALA-derived PpIX in vivo, mice bearing tumors of approximately 100 mm³ were randomized into A@LST, AD@LS, and AD@LST treatment groups. Following intravenous administration of drugs (equivalent to 20 mg/kg 5-ALA) at designated time points, mice were euthanized, and major organs (heart, liver, spleen, lungs, kidneys) along with tumors were harvested. The ex vivo PpIX fluorescence images were acquired by using the same imaging system (Ex/Em = 430/620 nm).

4.11. BBB penetration ability and PpIX metabolism of AD@LST in gliomas

To visualize AD@LST penetration across the BBB in an orthotopic intracranial GBM xenograft model, the brain vasculature was fluorescently labeled via intravenous injection of Alexa Fluor 555-conjugated anti-CD31 antibody (obtained from IVIM Technology) 0.5 h prior to imaging. Dil-labeled AD@LST was intravenously administered through a tail vein catheter during imaging. The trans-BBB passage of AD@LST was observed in real-time through a surgically implanted cranial window using a commercial intravital microscope (IVIM; IVM-CMS3, IVIM Technology, South Korea).

Similarly, A@LST and AD@LST (equivalent to 20 mg/kg 5-ALA) were intravenously injected through a tail vein catheter during imaging. PpIX fluorescence signal dynamics in tumor tissues were monitored using the same microscope at sequential time points to assess metabolic activity.

4.12. Evaluation of therapeutic effect in vivo

U87 subcutaneous GBM model.

To determine the synergistic effects of DFO and targeting on 5-ALA-PDT efficacy in repeated treatments, U87 GBM-bearing mice (∼70 mm³) were assigned to seven groups: (G1) Saline, (G2) A@LST, (G3) AD@LS, (G4) AD@LST, (G5) A@LST + laser (+), (G6) AD@LS + laser (+), (G7) AD@LST + laser (+). All groups except the saline control (G1) received an equivalent dose of 5-ALA (20 mg/kg) on both day 1 and day 8. Laser-treated groups were irradiated (635 nm, 200 mW/cm², 20 min) at 6, 12, and 24 h post-injection. Tumor volume and body weight were measured every 2 days. Tumor tissues were harvested on day 14 for Ki-67, TUNEL, and H&E staining. Blood samples from groups were collected via retro-orbital puncture under anesthesia on day 14 and analyzed for biochemical indices including ALT, AST, ALP, ALB, and TP (Wuhan Servicebio Technology Co., Ltd., China). Data are presented with n = 5 per group. In addition, serum levels of blood urea nitrogen (BUN) and creatinine (CREA), as well as blood counts parameters (mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), and mean corpuscular hemoglobin concentration (MCHC)), were measured as supplementary analyses. Data are presented with n = 3 per group.

4.12.1. Terminal tissue processing protocol

At study completion, all mice were euthanized by cervical dislocation. Tumors and major organs (heart, liver, spleen, lung, kidneys) were excised, fixed in 4% paraformaldehyde, and processed for histopathological evaluation. Tissue samples were submitted to Wuhan Servicebio Technology Co., Ltd. for comprehensive histochemical staining and digitization using a high-resolution slide scanner.

4.13. Statistical analysis

The data represent the means ± standard deviation. Statistical differences were determined by an unpaired two-tailed Student's t-test between two groups. A p value < 0.05 was considered to be statistically significant.

CRediT authorship contribution statement

Rong Wen: Writing – original draft, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Gang He: Writing – original draft, Investigation, Formal analysis, Data curation, Conceptualization. Yilin Wan: Writing – review & editing, Methodology, Investigation, Data curation, Conceptualization. Xian Guo: Writing – review & editing, Investigation, Data curation. Shengquan Xiong: Writing – review & editing, Methodology, Data curation. Yumeng Wu: Writing – review & editing, Visualization. Yishan Wei: Writing – review & editing, Investigation, Data curation. Dongdong Li: Writing – review & editing, Supervision. Jing Lin: Writing – review & editing, Validation, Supervision, Resources, Funding acquisition, Conceptualization. Peng Huang: Writing – review & editing, Validation, Supervision, Resources, Funding acquisition, Conceptualization.

Ethics approval and consent to participate

All animal experiments described in this study were conducted in accordance with protocols (AEWC-202300010) reviewed and approved by the Animal Ethics and Welfare Committee of Shenzhen University (AEWC-SZU).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article.