Enzyme‐Directed and Organelle‐Specific Sphere‐to‐Fiber Nanotransformation Enhances Photodynamic Therapy in Cancer Cells

Abstract

Employing responsive nanoplatforms as carriers for photosensitizers represents an effective strategy to overcome the challenges associated with photodynamic therapy (PDT), including poor solubility, low bioavailability, and high systemic toxicity. Drawing inspiration from the morphology transitions in biological systems, a general approach to enhance PDT that utilizes enzyme‐responsive nanoplatforms is developed. The transformation of phosphopeptide/photosensitizer co‐assembled nanoparticles is first demonstrated into nanofibers when exposed to cytoplasmic enzyme alkaline phosphatase. This transition is primarily driven by alkaline phosphatase‐induced changes of the nanoparticles in the hydrophilic and hydrophobic balance, and intermolecular electrostatic interactions within the nanoparticles. The resulting nanofibers exhibit improved ability of generating reactive oxygen species (ROS), intracellular accumulation, and retention in cancer cells. Furthermore, the enzyme‐responsive nanoplatform is expanded to selectively target mitochondria by mitochondria‐specific enzyme sirtuin 5 (SIRT5). Under the catalysis of SIRT5, the succinylated peptide/photosensitizer co‐assembled nanoparticles can be transformed into nanofibers specifically within the mitochondria. The resulting nanofibers exhibit excellent capability of modulating mitochondrial activity, enhanced ROS formation, and significant anticancer efficacy via PDT. Consequently, the enzyme‐instructed in situ fibrillar transformation of peptide/photosensitizers co‐assembled nanoparticles provides an efficient pathway to address the challenges associated with photosensitizers. It is envisaged that this approach will further expand the toolbox for enzyme‐responsive biomaterials for cancer therapy.

A general approach utilizing enzyme‐responsive nanoplatforms to enhance photodynamic therapy is developed. When exposed to the cytoplasmic enzyme alkaline phosphatase or the mitochondrial enzyme sirtuin 5, the co‐assembled nanoparticles can be transformed into nanofibers in the cytoplasm or specifically within the mitochondria. The resulting nanofibers exhibit enhanced ROS formation and significant anticancer efficacy via photodynamic therapy.

1. Introduction

Photodynamic therapy (PDT) is a widely used method for treating various malignant cancers due to its non‐invasive feature.^{[ 1} , ^{2 ]} Photosensitizers are transported to the tumor site and activated under light irradiation, producing reactive oxygen species (ROS) or singlet oxygen, which leads to the destruction of the tumor cells.^{[ 3 ]} Significant progress has been made in the development of highly efficient photosensitizers,^{[ 4} , ^{5 ]} particularly the porphyrin family and its derivatives, because of their high quantum yield in ¹O₂ generation. However, the application of porphyrins in PDT is often limited by poor water solubility, low bioavailability, strong dark cytotoxicity, and off‐target accumulation.^{[ 6 ]} To overcome these limitations, various nanoplatforms for photosensitizer delivery have been designed. Recently, the covalent photosensitizer‐peptide self‐assembly strategy has been utilized in cancer photoacoustic imaging,^{[ 7} , ⁸ , ⁹ , ¹⁰ , ^{11 ]} photothermal therapy,^{[ 7} , ¹¹ , ^{12 ]} and PDT treatment.^{[ 13} , ¹⁴ , ^{15 ]} For example, Wang and co‐workers designed a panel of photosensitizer‐peptide conjugates that can self‐assemble at the tumor sites for cancer imaging and therapy.^{[ 7} , ⁸ , ⁹ , ¹⁰ , ¹¹ , ^{12 ]} Furthermore, the co‐assembly strategy of peptide/photosensitizer was employed as another tactic for PDT in cancer therapy.^{[ 16} , ¹⁷ , ^{18 ]} Yan's group developed some short peptide‐based co‐assembled supramolecular nanodrugs for PDT, exhibiting excellent effects on different cancer treatments.^{[ 16} , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ^{21 ]} Notably, they reported that the fibril‐transformable strategy can greatly prolong the retention time in tumors and enhance tumor suppression efficacy remarkably.^{[ 14 ]}

Enzyme‐instructed self‐assembly (EISA), pioneered by Xu's group, is a strategy in which enzymes can convert soluble monomers into nanoassemblies.^{[ 22 ]} This approach has undergone significant development and found diverse applications, such as tumor therapy,^{[ 23} , ²⁴ , ^{25 ]} construction of 3D cell spheroids,^{[ 26} , ²⁷ , ²⁸ , ²⁹ , ^{30 ]} membrane imaging,^{[ 31 ]} intracellular enzyme,^{[ 32 ]} and ATP sequestration.^{[ 33 ]} Combining EISA with photosensitizer/peptide co‐assembly could provide an efficient approach to overcome the aforementioned limitations of photosensitizers. In addition, it circumvents the need for tedious chemical synthesis and enables the nano‐system to accumulate in the tumor site, making it possible to design stimuli‐responsive nanomaterials with improved anticancer efficiency. Despite significant progress in EISA, no reports have shown the transformation of co‐assembled peptides and photosensitizers into nano‐systems within cancer cells, particularly in sub‐organelles, to enhance PDT efficacy. The co‐assembly approach eliminates the need for tedious synthesis, and the enzymes of interest provide cancer and organelle‐targeting capability. Consequently, designing a nano‐system that can activate or enhance photosensitizing capacity through specific cancer‐associated events, such as in situ transformation by enzymes,^{[ 34} , ³⁵ , ^{36 ]} is of great interest for achieving high therapeutic efficacy in PDT.

In this study, we successfully developed a nanoplatform for enhancing PDT efficacy using an enzyme‐responsive peptide and photosensitizer co‐assembly strategy. The platform undergoes sphere‐to‐fiber shape transformation triggered by the enzymes, leading to enhanced ROS generation efficiency. The peptide and photosensitizer co‐assemble to form nanoparticles (NPs) in water due to the hydrophobic interactions between the photosensitizers and peptides. Once exposed to the enzymes, the sphere‐to‐fiber switch is activated, resulting in the transformation to nanofibers. This transformation facilitates the intersystem crossing process and improves ROS generation efficiency.^{[ 14} , ³⁷ , ³⁸ , ^{39 ]} In addition, this transformation also prolongs retention time in the cancer cells, affording excellent cancer cell imaging and high photodynamic therapeutic efficacy with negligible side effects to normal cells.^{[ 11} , ⁴⁰ , ⁴¹ , ^{42 ]}

As a proof‐of‐concept, we first developed peptide/PpIX co‐assembled NPs that can respond to alkaline phosphatase (ALP) and transform into nanofibers in the cytoplasm. This transformation has led to a significant improvement in PDT of killing cancer cells. To further validate our design, we employed SIRT5 as a sub‐organelle trigger for transforming the co‐assembled NPs into nanofibers, specifically in the mitochondria.^{[ 43 ]} The in situ transformation under SIRT5 triggering greatly increased ROS generation inside the mitochondria, changed mitochondrial membrane potentials, and induced apoptosis of cancer cells (Scheme 1 ).

Scheme 1.

Schematic illustration of in situ transformation of photosensitizer/peptide co‐assembled nanoparticles (co‐assembled NPs) into nanofibers and its application in photodynamic therapy of cancer cells.

2. Results and Discussion

2.1. In Vitro Co‐Assembly of ALP‐Responsive NPs and Fibrillar Transformation

We initially investigated the self‐assembly behavior of phosphopeptide and photosensitizer with and without an enzyme trigger. The ALP‐responsive peptide, Fmoc (fluorenylmethoxycarbonyl)‐FFKY_p, mainly consists of three parts: 1) a hydrophobic core Fmoc group and 2) a self‐assembled fragment (phenylalanine‐rich peptide), both of which provide strong π─π stacking and hydrophobic interaction for co‐assembly, 3) a phosphorylated tyrosine with good solubility and electrostatic repulsion, acting as an ALP enzyme responsive “sphere‐to‐fiber switch”. A peptide, Fmoc‐FFKY without the responsive unit was included as a control. The peptides were synthesized via the standard Fmoc‐solid phase peptide synthesis (SPPS) method (Scheme S1 and Figure S1, Supporting Information). Subsequently, the ALP‐responsive co‐assembled NPs were prepared using the solvent‐shifting method. The co‐assembled NPs showed good stability in DMEM for 48 h, as analyzed by DLS (Figure S2, Supporting Information). As depicted in Figure 1B, a homogeneous dispersion of peptide 1/PpIX co‐assembled NPs was observed. The morphology of the peptide 1/PpIX co‐assembled NPs was further characterized using scanning electron microscopy (SEM) (Figure S3, Supporting Information) and transmission electron microscopy (TEM) (Figure 1C). The images showed that the co‐assembled NPs were in a spherical conformation with sizes of hundreds of nanometers, which was consistent with the DLS results (Figure S4A, Supporting Information). The co‐assembly of peptides and PpIX into NPs is attributed to several factors, including hydrophobic effects, electrostatic interactions, hydrogen bonds, etc. In contrast, the non‐enzymatic responsive co‐assembly of peptide Fmoc‐FFKY and PpIX prefers to form nanofibers due to decreased solubility and electrostatic repulsion (Figure S5, Supporting Information).

Figure 1.

A) Structures of the photosensitizers (PpIX) and peptide 1: Fmoc‐FFKY_p, and scheme of the morphology transformation under enzyme catalysis. B) Hydrogelation behavior of peptide 1/PpIX co‐assembled NPs w/o ALP addition. C) TEM images of peptide 1/PpIX co‐assembled NPs w/o ALP addition. D) CLSM images of HeLa cells incubated with peptide 1/PpIX co‐assembled NPs (with 1 µg mL⁻¹ of PpIX and 88.4 µm of peptide 1) at different time points (2, 4, and 8 h). Free PpIX (1 µg mL⁻¹) was used for comparison study. E) Photoactivity of peptide 1/PpIX NFs, NPs and free PpIX in vitro. DCFH‐DA was used as ROS indicator. Ex: 488 nm, Em: 500–600 nm. F) CLSM images of intracellular ROS generation in HeLa cells treated with peptide 1/PpIX co‐assembled NPs (with 1 µg mL⁻¹ of PpIX and 88.4 µm of peptide 1) w/o 650 nm LED light irradiation (60 mW cm⁻²). For inhibition experiments, cells were treated with ALP inhibitors DQB and NPPS respectively. Free PpIX (1 µg mL⁻¹) in H₂O was used for comparison study. Scale bar = 20 µm. G) Quantitative analysis of the green fluorescence in Figure F (1. Co‐assembled NPs (Light), 2. Co‐assembled NPs (Dark), 3. Co‐assembled NPs with NPPS (Light), 4. Co‐assembled NPs with NPPS (Dark), 5. Co‐assembled NPs with DQB (Light), 6. Co‐assembled NPs with DQB (Dark), 7. Free PpIX (Light), 8. Free PpIX (Dark)). The fluorescent intensity was quantified by ImageJ software. Column 1 (Co‐assembled NPs (Light)) versus each column, * p < 0.05; ** p < 0.01; *** p < 0.001 (mean ± SD, n = 5). H) TEM analysis of cytosol fractions from HeLa cells treated with peptide 1/PpIX co‐assembled NPs (with 1 µg mL⁻¹ of PpIX and 88.4 µm of peptide 1) or free PpIX (1 µg mL⁻¹) in H₂O. I) In vitro dark cytotoxicity and phototoxicity of HeLa cells incubated with different concentrations of peptide 1/PpIX co‐assembled NPs or PpIX w/o 650 nm LED light irradiation (60 mW cm⁻²) (indicated concentrations were the concentrations of PpIX in the diluted co‐assembled NPs), determined by CCK‐8 assay. Co‐ assembled NPs (light) versus Free PpIX (light) * p < 0.05; ** p < 0.01; *** p < 0.001 (mean ± SD, n = 3).

After addition of ALP, the spherical structure of the peptide 1/PpIX co‐assembled NPs was gradually destroyed (Figure S6, Supporting Information), and a small number of particulate nanofibrillar structures appeared, as shown in Figure S7 (Supporting Information). With increasing incubation time, more nanofibrillar structures were detected (Figure S7, Supporting Information), while no significant changes were observed in the control group without ALP addition (Figure S6, Supporting Information). After 24 h, the initial liquid solution transformed into a solid‐like gel (Figure 1B), which is a characteristic signature of nanoscale fiber network formation. SEM (Figure S3, Supporting Information) and TEM (Figure 1C) revealed a network of entangled nanofibers with diameters ≈10–50 nm and lengths of several micrometers after 24 h of incubation. The transformation from NPs to nanofibers was further confirmed in solution by DLS analysis (Figure S4B, Supporting Information). These results provide strong evidence that ALP is critical in the conversion of co‐assembled NPs into nanofibers. The phosphate group of the peptide induces strong electrostatic repulsion between peptides and photosensitizer, as well as increased solubility. These factors contribute to the formation of NPs rather than nanofibers. Upon ALP treatment, the phosphate group of the “sphere‐to‐fiber switch” is cleaved from the peptide, leading to weakened electrostatic repulsion and increased hydrophobic interactions. These modulated interactions subsequently stimulate the formation of peptide/PS nanofibers.

2.2. Cellular Uptake and ALP‐Triggered Fibrillar Transformation in Living Cells

After demonstrating the enzyme‐induced morphology transition in solution, cellular uptake and intracellular localization of the peptide 1/PpIX co‐assembled NPs were investigated by confocal laser scanning microscopy (CLSM). In Figure 1D, only weak red fluorescence was observed in HeLa cells treated with photosensitizer PpIX, even after 8 h incubation. This can be attributed to the large aggregate formation of the PpIX in the cell medium (Figure S8, Supporting Information) and the electrostatic repulsion between PpIX and the negatively charged cell membranes. In contrast, for HeLa cell treated with peptide 1/PpIX co‐assembled NPs, a stronger and larger area of red fluorescence was observed after 8 h incubation. These results indicated that the cellular internalization of photosensitizers is significantly improved after co‐assembly with phosphorylated peptide. Additionally, we confirmed that the cellular internalization mechanism of the co‐assembled NPs is mainly through phagocytosis (Figure S9, Supporting Information). Once uptaken by the cancer cells, the co‐assembled NPs will undergo in situ transformation to form nanofibers in cells after dephosphorylation due to the high expression levels of ALP in HeLa cells. The dephosphorylated product of peptide 1 after self‐assembly in HeLa cells was further confirmed through HPLC analysis. Simultaneously, two hydrophilic peptides unable to undergo self‐assembly were not detected inside cells (Figure S10, Supporting Information). These results suggest that the formation of nanofibers will endow peptide/photosensitizer co‐assembly with prolonged retention time in the cancer cells. It resulted in stronger red fluorescence in HeLa cells treated with peptide 1/PpIX co‐assembled NPs when compared with the cells treated with free photosensitizer or non‐enzymatic responsive control (Fmoc‐FFKY/PpIX co‐assembly) (Figure S11A, Supporting Information).

2.3. Transformation of NPs to Nanofibers by ALP in Living Cells

To explore whether co‐assembled NPs are spatially transformed into nanofiber within cells, we conducted cell fractionation and TEM imaging experiments to study their transformation behavior. Briefly, HeLa cells were first incubated with the prepared peptide 1/PpIX co‐assembled NPs for 8 h, followed by the nucleus and cytosol fraction isolation via the protocol described in the literature.^{[ 43 ]} Further TEM visualization clearly showed that nanofibers were present in the cytosol fraction of the co‐assembled NPs‐treated cells, but no nanofiber was found in the cytosol of free PpIX‐treated cells (Figure 1H). The size of the transformed nanofibers also agreed with that shown in the solution experiment (Figure 1C). These data firmly proved that nanofibers can be effectively transformed through peptide1/PpIX co‐assembled NPs in HeLa cells overexpressing ALP.

2.4. ALP‐Triggered Fibrillar Transformation Enhances Photodynamic Effects in Solution and in Living Cells

We hypothesized that the conversion of co‐assembled NPs into nanofibers, triggered by enzymes, would amplify the photodynamic effect. To investigate this hypothesis, we evaluated the photoactivity of peptide 1/PpIX co‐assembled NPs by measuring the ROS generation capacity using 2′,7′‐dichlorodihydrofluorescein diacetate (DCFH‐DA). DCFH generated by DCFH‐DA after hydrolysis, can be rapidly oxidized into dichlorofluorescein (DCF) and produce fluorescence in the presence of ROS, along with enhanced green fluorescence. As depicted in Figure 1E, the fluorescence intensity of DCFH treated with the peptide 1/PpIX co‐assembled NPs was consistently and significantly higher than that of the free PpIX group, when irradiated with a 650 nm light continuously. The low ROS generation in the free PpIX group can be attributed to the aggregation and low solubility of the photosensitizer. In addition, compared to the peptide 1/PpIX co‐assembled NPs, the increased fluorescence intensity in ALP‐treated peptide 1/PpIX co‐assembled NPs group was remarkably higher, indicating an enhanced ROS generation efficiency through transformation from NPs to nanofibers triggered by enzymes. This phenomenon can be attributed to the amplified intersystem crossing resulting from the enzyme‐triggered nanofibrillar transformation, which is consistent with the reported literatures.^{[ 14} , ³⁷ , ³⁸ , ^{39 ]}

Next, we investigated whether nanofiber transformation could enhance the photodynamic effect in living cells. The ROS indicator DCFH‐DA was used to detect the ability of ROS generation in live cells. HeLa cells were treated with peptide 1/PpIX co‐assembled NPs and free photosensitizer respectively, and intracellular ROS production was examined by CLSM. Bright fluorescence from DCF in HeLa cells was readily observed after the treatment of peptide 1/PpIX co‐assembled NPs and light irradiation for 5 min, suggesting significant ROS generation. Conversely, weak fluorescence was observed in HeLa cells treated with free PpIX after irradiation, indicating weak ROS generation due to PpIX aggregation in H₂O. Furthermore, the control groups with the treatment of ALP inhibitors (1‐Naphthyl phosphate potassium salt (NPPS) and2,5‐dimethoxy‐N‐(quinolin‐3‐yl)benzenesulfonamide (DQB)) displayed reduced fluorescence (Figure 1F,G), indicating that the ALP‐triggered morphology transformation contributed to the enhancement of intracellular ROS generation. The cells treated with non‐enzymatic responsive control (Fmoc‐FFKY/PpIX co‐assembly) only displayed weak fluorescence (Figure S12A, Supporting Information), underscoring the superiority of the enzymatic‐mediated transformation. In addition, we employed lung cancer cell A549, which has lower ALP expression than HeLa cell,^{[ 44} , ⁴⁵ , ^{46 ]} as a negative control in our intracellular ROS generation study. As shown in Figure S13 (Supporting Information), the results were consistent with the inhibitor treatment experiments, displaying much weaker ROS fluorescence. Scattered fluorescence dots were found in both peptide 1/PpIX co‐assembled NPs and free PpIX treated cells, regardless of light irradiation. These experiments collectively suggest that the ALP‐triggered in situ nanofiber transformation plays a crucial role in enhancing intracellular ROS generation.

Subsequently, we evaluated the dark cytotoxicity and phototoxicity of the co‐assembly strategy toward cancer cells using CCK‐8 assays. As shown in Figure 1I, no significant dark cytotoxicity was observed in HeLa cells treated with peptide 1/PpIX co‐assembled NPs (4 µg mL⁻¹ of photosensitizer and 353.6 µm of peptide 1), indicating their excellent biocompatibility. Additionally, peptide 1 alone showed low cytotoxicity even at high concentrations (Figure S14, Supporting Information). Under light irradiation for 5 min, both peptide 1/PpIX co‐assembled NPs and free PpIX demonstrated remarkable anticancer efficacy, with the viability of HeLa cells decreasing when photosensitizer concentration increases. Notably, the cytotoxicity of the peptide 1/PpIX co‐assembled NPs was significantly higher than that of free PpIX or non‐enzymatic responsive control (Fmoc‐FFKY/PpIX co‐assembly) (Figure S15A, Supporting Information) under the same conditions. The enhanced cytotoxicity observed in peptide 1/PpIX co‐assembled NPs‐treated group can be attributed to the enzyme responsive co‐assembly strategy, which facilitates increased cellular uptake of the photosensitizer. Subsequently, this nanoassembly undergoes an in situ enzyme‐induced morphological transition, enhanced ROS generation, and toxicity against cancer cells.

2.5. SIRT5‐Triggered Fibrillar Transformation

To further prove that our co‐assembled nanoplatform is a facile strategy that can be also used to target sub‐organelles, we further established another in situ nanofiber transformation specifically within mitochondria. SIRT5 is known as a mitochondria‐specific enzyme and recognizes succinylated substrates.^{[ 43 ]} The designed succinylated peptides, Fmoc‐FFFGK_succG, was mainly composed of three parts: 1) a Fmoc group and 2) a phenylalanine‐rich peptide fragment, acting as a hydrophobic core and an assembling unit, and 3) a succinylased lysine (K_succ) module acting as a SIRT5‐responsive “sphere‐to‐fiber switch” for specific enzyme‐triggering nanofiber transformation. A peptide, Fmoc‐FFFGKG without the responsive unit was included as a negative control. The peptides were synthesized using SPPS methods (Scheme S1 and Figure S1, Supporting Information).^{[ 43 ]} Following that, peptide 2/CPTP co‐assembled NPs were prepared (Figure 2A). These NPs exhibited excellent stability after 48 h of incubation in DMEM (Figure S2, Supporting Information). Since peptide 2 is more hydrophilic than peptide 1, we chose a more hydrophobic photosensitizer, CPTP, for co‐assembly study. As shown in Figure 2B, after incubation with 0.0025 equiv. SIRT5 and 2 equiv. NAD⁺, the initial liquid solution was transformed into a solid‐like gel due to the desuccinylation of peptide 2 (Figure S16, Supporting Information). Further TEM images clearly revealed a network of entangled fibers with diameters within nanometers after the treatment of SIRT5 and NAD⁺ (Figure 2C). In a control experiment performed under the same conditions, TEM images showed that the co‐assembled NPs did not form nanofibers in the absence of SIRT5 enzyme or NAD⁺ (Figure 2C). And the non‐enzymatic responsive control (Fmoc‐FFFGKG/CPTP co‐assembly) preferred to form nanofibers spontaneously (Figure S5, Supporting Information). These results clearly demonstrate that the transition of the nanofibers from their NP state can be attributed to the enzymatic interaction between SIRT5 and peptide 2.

Figure 2.

A) Structures of the photosensitizer (CPTP) and peptide 2: Fmoc‐FFFGK_succG, and scheme of the morphology transformation under enzyme catalysis. B) Hydrogelation behavior of peptide 2/CPTP co‐assembled NPs under SIRT5 catalysis. C) TEM images of peptide 2/CPTP NPs (top, left), NAD⁺ treated NPs (bottom, left), SIRT5 treated NPs (bottom, right), SIRT5 and NAD⁺ treated NPs (top, right). D) Photoactivity of the co‐assembled nanostructures in vitro. Fluorescence of DCFH‐DA (ROS indicator) sensitized by irradiating peptide 2/CPTP co‐assembled NPs/NFs at 650 nm, measured by the increase in the fluorescence of DCF as a function of time. Pure CPTP and DCFH‐DA were used as the control group. All samples were measured under irradiation for 180 s. E) Cell imaging study of peptide 2/CPTP co‐assembled NPs (with 1 µg mL⁻¹ of CPTP and 156.4 µm of peptide 2) w/o inhibitors, or free CPTP (1 µg mL⁻¹) in HeLa cells. Scale bar = 20 µm. F) TEM analysis of mitochondria fractions after incubating HeLa cells with peptide 2/CPTP co‐assembled NPs (with 1 µg mL⁻¹ of CPTP and 156.4 µM of peptide 2) or free CPTP (1 µg mL⁻¹) in H₂O.

2.6. Mitochondria‐Confined Fibrillar Transformation Triggered by SIRT5 in Living Cells

We hypothesized that the co‐assembled NPs can efficiently traverse cell membrane, interact with SIRT5 in the mitochondrion, and undergo in situ transformation into nanofibers in living cells. To explore this hypothesis, we selected HeLa cell expressing SIRT5 enzyme in this study. Briefly, the prepared peptide 2/CPTP co‐assembled NPs and free CPTP were respectively incubated with HeLa cells at a suitable concentration (1 µg mL⁻¹ of photosensitizer, Figure S17, Supporting Information). The fluorescence intensity of the cells was monitored using confocal fluorescence microscopy, as depicted in Figure 2E. Cells incubated with co‐assembled NPs for 24 h showed significant intracellular red fluorescence. In contrast, cells incubated with free CPTP displayed weak intracellular red fluorescence. The observed red fluorescence inside the cells is mainly due to the presence of CPTP. The manifestation of significant intracellular red fluorescence inside cells incubated with co‐assembled NPs signified that the prepared NPs enhanced the cellular uptake of the photosensitizer. Moreover, the in situ transformation into nanofiber may also prolong the retention time of photosensitizer. Conversely, for the cells incubated with free CPTP, the hydrophobic photosensitizer in the solution will form aggregation easily (Figure S18, Supporting Information) which may hinder the efficacy of cellular uptake. Consequently, no significant red fluorescence was observed.

Next, we investigated the precise localization of the co‐assembled NPs in living cells. To achieve this, we conducted co‐localization experiments by incubating the peptide 2/CPTP co‐assembled NPs and Mito‐Tracker Green, a dye used to specifically label mitochondria. As shown in Figure 2E, a substantial overlap was observed between the red fluorescence emitted by the photosensitizer and the green fluorescence of Mito‐Tracker Green, confirming that the transformed nanofibers were indeed localized in the mitochondrial region. The line scan profile further proved the excellent alignment between the fluorescence signals from the nanofibers and Mito‐Tracker Green. Through further quantitative analysis, the Pearson correlation coefficient was determined as 0.87 (Figure S19, Supporting Information). In contrast, the fluorescence overlay between the red fluorescence of the photosensitizer and the green fluorescence of Mito‐Tracker Green was notably weaker in the CPTP only groups. Furthermore, in the inhibition experiments with the SIRT5 inhibitor suramin, the co‐localization between the fluorescence of the nanofibers and Mito‐Tracker Green fluorescence was also reduced. Additionally, in the non‐enzymatic responsive control (Fmoc‐FFFGKG/CPTP co‐assembly), both the red fluorescence of photosensitizer and mitochondria co‐localization effect were notably weaker (Figure S11B, Supporting Information). These data together indicate a high level of co‐localization between the co‐assembled NPs and Mito‐Tracker dye, indicating that the co‐assembled NPs are likely to interact with SIRT5 in the mitochondria and undergo nanofiber transformation through SIRT5 catalysis.

Furthermore, we conducted cell fractionation and TEM imaging experiments to investigate whether the nanofibers were indeed formed in the mitochondria. Briefly, we incubated HeLa cells with the co‐assembled NPs, followed by isolating the nucleus, mitochondria, and cytosol fractions via the protocol described in the literature.^{[ 43 ]} Further TEM visualization clearly showed that the co‐assembled transformed nanofibers were present in the mitochondrial fraction, as shown in Figure 2F, but not in other fractions, such as the nucleus and cytosol (Figure S20, Supporting Information). And the size also agreed with that shown in Figure 2C. In addition, control cells incubated without co‐assembled NPs showed only amorphous materials (Figure 2F). The combined fluorescence and TEM imaging data clearly indicated that supramolecular nanostructures can be selectively formed in the mitochondrial area of live cells.

2.7. SIRT5‐Triggered In Situ Fibrillar Transformation Enhanced Photodynamic Effects in Solution and in Living Cells

We next investigated the influence of SIRT5 on the photodynamic effect of peptide 2/CPTP co‐assembled NPs in vitro. As shown in Figure 2D, the fluorescence intensity from DCF in the peptide 2/CPTP co‐assembled NPs group was significantly higher than that in the free CPTP and blank group under a 650 nm light. The inherent strong hydrophobicity of CPTP caused its aggregation in water, leading to low ROS generation. Furthermore, the fluorescence increase was significantly higher in the SIRT5‐treated peptide 2/CPTP co‐assembled NPs groups. In this group, the co‐assembled NPs transformed into nanofibers under SIRT5 catalysis, demonstrating improved ROS generation efficacy after the transformation from NPs to nanofibers.

We next investigated the ability of the in situ transformable nanosystem to enhance ROS generation in living cells under light irradiation. HeLa cells were treated with peptide 2/CPTP co‐assembled NPs and free CPTP, intracellular ROS generation was then examined by CLSM. After light irradiation, cells treated with the peptide 2/CPTP co‐assembled NPs exhibited strong green fluorescence (Figure 3A), indicating significant ROS generation, while cells without light irradiation showed much weaker fluorescence. In contrast, free CPTP groups with or without light irradiation, displayed weak fluorescence due to the aggregation of CPTP in H₂O. In addition, the free peptide‐incubated groups with or without light irradiation also displayed an extremely weak fluorescence signal. Furthermore, we investigated the effect of inhibiting SIRT5 on ROS generation in HeLa cells. A significant decrease of intracellular ROS fluorescence was observed after incubating HeLa cells with SIRT5 inhibitor suramin (Figure 3A) and inhibitor NRD 167 (Figure S21, Supporting Information). The cells treated with non‐enzymatic responsive control (Fmoc‐FFFGKG/CPTP co‐assembly) also displayed weak fluorescence (Figure S12B, Supporting Information), indicating a weaker generation of ROS. These findings together demonstrated that the in situ morphology transformation into nanofibers is SIRT5‐dependent and can indeed enhance the photodynamic efficacy of photosensitizer in living cells.

Figure 3.

A) CLSM images of intracellular ROS generation in HeLa cells treated with peptide 2/CPTP co‐assembled NPs (consists of 1 µg mL⁻¹ of CPTP and 156.4 µm of peptide 2) w/o SIRT5 inhibitor suramin, free CPTP (1 µg mL⁻¹) in H₂O or peptide only (156.4 µm) w/o 650 nm LED light irradiation (60 mW cm⁻²), Scale bar = 20 µm. B) In vitro dark cytotoxicity and phototoxicity toward HeLa cells and C) 3T3 cells treated with peptide 2/CPTP co‐assembled NPs and free CPTP in H₂O w/o 650 nm LED light irradiation (60 mW cm⁻²) (indicated concentrations were the concentrations of CPTP in the diluted co‐assembled NPs), determined by CCK‐8 assay. The standard deviation for each data point was averaged over three samples (n = 3). Co‐ assembled NPs (light) versus Free CPTP (light) ** p < 0.01; *** p < 0.001 (mean ± SD, n = 3).

We next investigated the effect of peptide 2/CPTP co‐assembled NPs on mitochondria under light irradiation. HeLa cells were incubated with the peptide 2/CPTP co‐assembled NPs (1 µg/mL of CPTP) for 24 h, followed by irradiation with 650 nm light for 5 min and staining with JC‐1 for 20 min before imaging using CLSM. As shown in Figure 4A, after treating with co‐assembled NPs and light irradiation, a significant fluorescence decrease in the red channel was observed, accompanied by a corresponding fluorescence increase in the green channel. These experiments demonstrated that most JC‐1 existed in increased monomer forms and few in aggregation forms, suggesting that our nanosystem depolarized and decreased the mitochondria membrane potentials under light irradiation. In contrast, there was no significant change in cells treated with free CPTP with or without light irradiation. Furthermore, cell imaging experiments with Mito‐SOX were performed to examine mitochondrial ROS generation. Figure 4B shows a clear and strong fluorescence signal in the Mito‐SOX channel in cells treated with co‐assembled NPs and light irradiation, compared to the control cells, indicating efficient mitochondrial ROS generation in cancer cells after incubation with co‐assembled nanostructures and irradiation. The results from the cell‐based experiments confirm that SIRT5‐induced in situ transformation of nanofibers in the mitochondria can significantly enhance mitochondrial ROS generation and depolarize the mitochondrial membrane potentials with light irradiation.

Figure 4.

A) Fluorescence images of JC‐1 staining in HeLa cells exposed to peptide 2/CPTP co‐assembled NPs (consists of 1 µg mL⁻¹ of CPTP and 156.4 µm of peptide 2) or free CPTP (1 µg mL⁻¹) w/o 650 nm LED light irradiation (60 mW cm⁻²), respectively. JC‐1 dye was used as an indicator of mitochondrial potential. Scale bar = 20 µm. B) CLSM images of intracellular ROS generation in the mitochondria of HeLa cells treated with peptide 2/CPTP co‐assembled NPs (consists of 1 µg mL⁻¹ of CPTP and 156.4 µm of peptide 2) or free CPTP (1 µg mL⁻¹) in H₂O w/o 650 nm LED light irradiation (60 mW cm⁻²). Mito‐SOX was used as an indicator of ROS generation in the mitochondria. Scale bar = 20 µm.

The cell viability study of our co‐assembled strategy was evaluated by CCK‐8 assays to verify the PDT efficacy in cancer therapy. As shown in Figure 3B, no obvious dark cytotoxicity was observed in HeLa cells treated with peptide 2/CPTP co‐assembled NPs (even up to 1.5 µg mL⁻¹ of photosensitizer), indicating good biocompatibility of the co‐assembled nano‐system. After 5 min irradiation by a 650 nm light, the co‐assembled NPs treated group demonstrated remarkable phototoxicity compared with the controls. Moreover, the viability of HeLa cells decreased with the increase of the CPTP concentration in the co‐assembled NPs. And the IC₅₀ value of the co‐assembled NPs‐treated group (0.92 µg mL⁻¹) was significantly lower than that of the free CPTP or peptide‐treated group. There was no dark and photocytotoxicity observed in the free CPTP or peptide‐treated groups, likely due to the aggregation of photosensitizer or good compatibility of the peptide (Figure S22, Supporting Information). Similarly, no dark cytotoxicity was observed in the non‐enzymatic responsive control group (Fmoc‐FFFGKG/CPTP co‐assembly), while cytotoxicity was only observed at relatively high concentration of Fmoc‐FFFGKG/CPTP co‐assembly under light irradiation (Figure S15B, Supporting Information). Furthermore, HepG2 cells which expresses high levels of SIRT5, exhibited substantial photocytotoxicity when treated with peptide 2/CPTP co‐assembled NPs (Figure S23, Supporting Information). However, the observed cytotoxicity was lower than that in HeLa cells, and this difference may be attributed to the elevated levels of catalase and glutathione reductase in HepG2 cells.^{[ 47} , ⁴⁸ , ⁴⁹ , ⁵⁰ , ^{51 ]} Additionally, a SIRT5 low‐expressing cell line NIH/3T3 was utilized for cytotoxicity studies. Negligible dark cytotoxicity was observed in NIH/3T3 cells treated with peptide 2/CPTP co‐assembled NPs, with only minimal photocytotoxicity detected under light exposure (Figure 3C). The IC₅₀ value of co‐assembled NPs under light irradiation was significantly higher than that in HeLa/HepG2 cells. These results indicate that our approach selectively targets and eliminates cancer cells expressing high SIRT5 levels more efficiently than normal cells.

We further investigated the phototoxicity mechanism of peptide 2/CPTP co‐assembled NPs using flow cytometric analysis with fluorescently labeled Annexin V‐FITC/propidium iodide (PI). As depicted in Figure 5A, the proportion of cell apoptosis induced by peptide 2/CPTP co‐assembled NPs during PDT was significantly higher than in the absence of light irradiation. Conversely, cells treated with free CPTP or peptides w/o light exposure showed negligible apoptosis. Moreover, live/dead cell imaging revealed that (Figure 5B), under light irradiation, the majority of HeLa cells treated with peptide 2/CPTP co‐assembled NPs (1 µg mL⁻¹ of CPTP and 156.4 µm of peptide 2) were stained with red fluorescence, indicating significant phototoxicity. In contrast, HeLa cells treated with free CPTP (± light) and free peptide (± light) predominantly exhibited green fluorescence, signifying the survival of most cells. These results were consistent with those obtained from CLSM and cytotoxicity assays, demonstrating the efficacy of our platform in effectively targeting and inducing apoptosis in cancer cells.

Figure 5.

A) Cell apoptosis analysis with Annexin V‐FITC/PI assay for different samples, including blank, free CPTP (1 µg mL⁻¹), free peptide (156.4 µM), and peptide 2/CPTP co‐assembled NPs (consists of 1 µg mL⁻¹ of CPTP and 156.4 µm of peptide 2) w/o 650 nm LED light irradiation (60 mW cm⁻²). B) Live/dead staining images of HeLa cells under various treatments. Live cells were stained with green fluorescence, while dead cells were labeled with red fluorescence. Ex: 488 nm, Em: 500–550 nm for Calcein AM; Ex: 543 nm, Em: 600–650 nm for PI. Scale bar = 20 µm.

3. Conclusion

In summary, we have successfully developed a general strategy of nanotransformation for PDT. The co‐assembled NPs, composed of photosensitizers and enzyme‐responsive peptide, can undergo transformation into nanofibers in the cytoplasm and mitochondria, catalyzed by ALP and SIRT5, respectively. Morphology transformation experiments proved that the enzymes can trigger the transformation of NPs into nanofibers efficiently. Cell‐based experiments demonstrated that ALP and SIRT5 can selectively catalyze the transformation of co‐assembled nanoplatform into nanofibers within live cells or specifically within mitochondria, as evidenced by the formation of supramolecular nanostructures in a spatially selective manner. Moreover, these transformed nanofibers showed great enhancement of ROS generation and excellent anticancer activity when employed for PDT. Our study provides new insights into the design of transformable peptide‐based co‐assembled nanostructures, overcoming the limitations of photosensitizer and endowing them with high bioavailability, and improved cancer cell killing efficacy.

Conflict of Interest

The authors declare no conflict of interest.

Supporting information

Supporting Information

Gan S., Yang L., Heng Y., Chen Q., Wang D., Zhang J., Wei W., Liu Z., Njoku D. I., Chen J. L., Hu Y., Sun H., Enzyme‐Directed and Organelle‐Specific Sphere‐to‐Fiber Nanotransformation Enhances Photodynamic Therapy in Cancer Cells. Small Methods 2024, 8, 2301551. 10.1002/smtd.202301551