English

• Immunogenic cell death (ICD), first observed in DOX-treated cells, is a unique type of regulatory cell death. Cells undergoing ICD are found to be immunogenic, which can activate immune responses in mice against the same kind of cells [1]. Follow up studies of ICD found that in addition to commonly used chemotherapy drugs, such as doxorubicin, oxaliplatin, and mitoxantrone, ICD can also be elicited by repeated low-dose X-rays, sonodynamic therapy (SDT), hyperthermia (HT), and photodynamic therapy (PDT) [2-6]. In the process of ICD, cells expose calreticulin (CRT), secrete ATP, and release high mobility group box 1 (HMGB1), etc. These emitted damage-related molecular patterns (DAMPs) act as danger signals and immune adjuvants, activating innate and adaptive immunity and enhancing anti-tumor immune responses [7,8]. Due to the vaccine-like function of immunogenic dying cancer cells, ICD has been extensively used for cancer immunotherapy [9,10].

  PDT uses specific wavelengths of light to irradiate the photosensitizer and generates cytotoxic reactive oxygen species (ROS), which in turn boosts the ICD in cells. To date, PDT to induce ICD to activate tumor immunity has attracted growing interest in cancer treatment [11,12]. Unfortunately, conventional PDT was inadequate to massively evoking ICD to afford a desirable immune activation effect. In an example of chlorin e6 (Ce6)-PDT-vaccine experiment, an additional ICD inducer, mitoxantrone, was required to compensate for the deficiency of single PDT in eliciting ICD, in order to develop a potent vaccine for cancer inhibition [13]. What is more, exogenous immune adjuvants or immune checkpoint inhibitors have also been needed to enhance the immunotherapeutic effect of PDT against cancer [14,15]. Under PDT-mediated cell damage, CRT, a highly conversed 46 kDa protein in the lumen of the endoplasmic reticulum (ER), was observed to translocate from the ER to the membrane of dying cells, characteristics that promote the identification, uptake and presentation of tumor antigens by dendritic cells (DCs) and subsequently induce immunological rejection to cancer cells [16-18]. Photodynamic generation of ROS in the ER can increase the exposure of CRT on cells and enhance the effect of PDT in eliciting ICD [19]. Therefore, several ER-targeted photosensitizers have been developed to induce severe ER stress and greatly increase the exposure of CRT on cells, which could effectively activate immune responses and inhibit tumor growth [20-22]. Mitochondria and ER are closely linked to maintain cell homeostasis. Photosensitizers that induce targeted stress in mitochondria could also cause ER stress and then induce large-scale CRT translocating from the ER to the cell membrane [23-25].

  The photosensitizer is an indispensable and key component of PDT, and its ability to generate ROS affects the ability of PDT to boost ICD. Traditional photosensitizers (such as Ce6, temoporfin, and pheophorbide A) exhibit typical aggregation-caused quenching (ACQ) effects. Their poor anti-photobleaching ability in the molecular state and reduced ability to generate ROS in the aggregated state limit their efficacy in PDT [26,27]. As a new type of photosensitizer, aggregation-induced emission (AIE) photosensitizers, with tunable optical properties, good photobleaching resistance, bright fluorescence, and strong ROS generation ability in the aggregated state, are increasingly favored in PDT [28-31]. Since targeted mitochondrial or ER oxidative stress can effectively induce ICD, researchers have developed several ER- or mitochondria-targeted AIEgen photosensitizers to induce ICD for the preparation of prophylactic tumor vaccines [32-34]. AIEgen-based PDT in the ER or mitochondria was effective in boosting ICD, and the results also demonstrated that the ability of photosensitizers to induce ICD was closely related to their ability to generate ROS. For example, TPE-PR-FFKDEL and hypericin can target the ER. TPE-PR-FFKDEL has a stronger ability to produce ROS than hypericin, and TPE-PR-FFKDEL-based PDT is more potent in eliciting ICD in cells [32]. In addition, TPE-DPA-TCyp and DPA-TCyp can target mitochondria. Due of the more twisted molecular structure of TPE-DPA-TCyp, TPE-DPA-TCyp produced more ROS, and TPE-DPA-TCyp-based PDT boosted more large-scale ICD in cells [34]. In short, the cellular distribution of the photosensitizer and its ability to generate ROS are important for the efficacy of photosensitizer in boosting ICD.

  Mitochondria and ER-targeted AIE molecules can effectively induce ICD. However, cells have a complex physicochemical environment, and small AIE molecules cannot maintain a stable aggregation state when used for PDT. The unstable aggregation state of AIE molecules not only affects the fluorescence intensity of the photosensitizer, but also affects the generation of ROS, which in turn weakens the effect of photosensitizer in evoking ICD. AIEgen-based nanomicelles, which maintain the stable micellar structure, can avoid the defect of AIE molecules in PDT applications. To date, few AIEgen-based nanomicelles targeting the mitochondria have been developed for ICD induction. Therefore, in this contribution, a mitochondria-targeted AIE nanophotosensitizer was developed. The stable micellar structure not only maintains the stable fluorescence properties of the photosensitizer, but also ensures strong ROS generation. The micelles were used as a highly potent ICD inducer for cancer immunotherapy.

  To prepare nano-AIEgens, a salicylaldazine-based amphiphilic polymer (C₁₆-AIE-PEG₂₀₀₀, AIE-1) was synthesized as shown in Figs. 1A and B. Briefly, hydrophobic and hydrophilic segments of AIE-1 were obtained by modification of 1-bromohexadecane and methoxypolyethylene glycols (mPEG₂₀₀₀) with 2,4-dihydroxybenzaldehyde. Conjugation of the hydrophobic and hydrophilic segments with hydrazine afforded AIE-1 (Scheme S1 in Supporting information). C₁₈-PEG₂₀₀₀-TPP (CPT) was also synthesized. TPP-C₄-COOH was synthesized by a one-pot method. TPP-C₄-COOH was then reacted with C₁₈-PEG₂₀₀₀-OH (CPO) to give CPT (Scheme S2 in Supporting information). The ¹H nuclear magnetic resonance (NMR) spectra of these intermediates and the target compounds were characterized to confirm their identity (Figs. S1–S7 in Supporting information).

  Figure 1

  

  Figure 1.  (A) Schematic illustration of the prepared micelles. (B) Chemical structure of C₁₆-AIE-PEG₂₀₀₀ (AIE-1). (C) Fluorescence intensity at 525 nm of AIE-1 (40 µmol/L) in DMSO/H₂O mixture. (D) Determination of the critical micelle concentration (CMC) of AIE-1. (E) Absorption spectra of micelles in water (the concentration of AIE-1 is 40 µmol/L). (F) Fluorescence spectra of micelles in water (the concentration of AIE-1 is 40 µmol/L). (G) Zeta potential and (H) size distribution of A-M and A/C-M in water. (I, J) Fluorescence intensity at 525 nm of solution containing different concentrations of micelles. (K, L) TEM images of A-M and A/C-M. (M, N) Fluorescence spectra of DCFH solution containing A-M or A/C-M under UV irradiation for a period of time. (O, P) Fluorescence spectra of SOSG solution containing A-M or A/C-M under UV irradiation for a period of time. (Q, R) Fluorescence spectra of TA solution containing A-M or A/C-M under UV irradiation for a period of time. 365 nm, 10 mW/cm². Scale bar: 50 nm.

  DownLoad: Full-Size Img PowerPoint

  AIE-Micelles (A-M), mainly dispersed in the cytoplasm, were constructed by the self-assembly of AIE-1, which was used as a control nanophotosensitizer for ICD induction. Incorporation of CPT into A-M resulted in mitochondria-targeted AIE/CPT-Micelles (A/C-M). We then investigated their ability to generate ROS and induce ICD in cells.

  The amphiphilic AIE-1 can self-assembly into micelles (A-M) in water, and the optical properties of A-M (the concentration of AIE-1 was 40 µmol/L) were investigated. As shown in Fig. 1E, an ultraviolet (UV) absorption peak was observed between 300 nm and 400 nm, and the maximum absorption was identified at 360 nm. Under 365 nm UV irradiation, the maximum emission at 525 nm in the fluorescence emission spectrum was observed (Fig 1F). As shown in Fig. 1C, AIE-1 exhibited typical AIE characteristics. It emits weakly at 525 nm in DMSO and becomes highly emissive in DMSO solutions containing more than 20% water. A solution containing positively charged micelles (A/C-M) was also prepared. A/C-M showed stronger UV absorption at wavelengths between 200 nm and 300 nm and a slight decrease in fluorescence intensity at 525 nm compared to A-M, which can be attributed to the incorporation of CPT into the micelles. The CMC of AIE-1 was 2.68 µmol/L (Fig. 1D). What is more, faint fluorescence at 525 nm of the micellar solution was observed when the concentration of AIE-1 was as low as 1 µmol/L, indicating that micelles can be formed even at very low concentrations (Figs. 1I and J). AIE nanomicelles have a good stability in dilute solution and a good self-assembly ability in aqueous solution, making them suitable for PDT applications. The average hydrodynamic sizes of A-M and A/C-M were 18 and 20 nm, respectively (Fig. 1H). The transmission electron microscopy (TEM) images revealed the spherical morphology of the micelles (Figs. 1K and L). The zeta potentials of A-M and A/C-M were −17.8 mV and 6.3 mV, respectively (Fig. 1G). The positive charges of A/C-M allow them to accumulate in the mitochondria. In short, these micelles, with ideal stability and similar optical properties, are suitable for biomedical applications.

  Dichloro-dihydro-fluorescein diacetate (DCFH-DA), singlet oxygen sensor green (SOSG) and terephthalic acid (TA) were used to detect the generation of ROS, singlet oxygen (¹O₂), and hydroxyl radicals (·OH). As shown in Figs. 1M and N, faint green fluorescence at 525 nm was observed by simply mixing DCFH with A-M or A/C-M. After UV irradiation, enhanced fluorescence of DCF was observed. The fluorescence intensity of the solution containing SOSG and micelles showed similar trends: the green fluorescence became stronger with increasing illumination time (Figs. 1O and P), whereas the fluorescence of the solution containing TA and micelles did not change (Figs. 1Q and R). These results indicate that the micelles have a similar ability to generate ¹O₂, which is suitable for further PDT of cells.

  The mitochondrial targeting abilities of the positively charged A/C-M and negatively charged A-M were assessed. As shown in Fig. 2A, the uptake of A/C-M by cells increased over time, and the green fluorescence of A/C-M was observed to overlap with the red fluorescence of Mito-Tracker, indicating the ability of A/C-M to target the mitochondria. As a control, A-M was also evaluated and showed lower cellular uptake and negligible mitochondrial accumulation. In A/C-M-treated cells, the fluorescence intensity profiles of Mito-Tracker were consistent with that of A/C-M, indicating a good overlap between A/C-M and the mitochondria (Figs. 2D and E). The overlap effect was enhanced by prolonged incubation time. The Pearson correlation coefficients were 0.65 and 0.82 at 8 h and 12 h, respectively, indicating that A/C-M were mainly located in the mitochondria after 12 h incubation. A-M showed a weak mitochondrial targeting ability (Figs. 2B and C), the Pearson correlation coefficients were 0.49 and 0.52 at 8 h and 12 h, respectively. A-M and A/C-M, with different cellular distributions, may induce oxidative damage at different sites. We then explored the efficacy of nanophotosensitizers in boosting ICD.

  Figure 2

  

  Figure 2.  (A) CLSM images of 4T1 cells treated with A-M or A/C-M (green fluorescence) for 8 h and 12 h, which were co-stained with Mito-Tracker Red. (B, C) Fluorescence intensity profile of Mito-Tracker and A-M when cells were treated with A-M for 8 h (B) or 12 h (C) based on typical images in (A). (D, E) Fluorescence intensity profile of Mito-Tracker and A/C-M when cells were treated with A/C-M for 8 h (D) or 12 h (E) based on typical images in (A). (F) The mean fluorescence intensity and (G) the quantity of CRT-positive cells when cells were treated with A/C-M for 12 h, irradiated with UV for a period of time and then incubated for 8, 12 or 24 h. (H) CLSM image of CRT (green fluorescence) on 4T1 cells when cells were irradiated with UV for a period of time and then incubated for 12 h. (I) Flow cytometry analysis of ROS in cells after cells were treated with micelles for 12 h and then irradiated with UV for a period of time. (J) Flow cytometry analysis of CRT on 4T1 cells after cells were treated with micelles for 12 h, irradiated with UV for a period of time and then incubated for 12 h. (K) Surface exposed CRT (green fluorescence), (L) released ATP, (M) and secreted HMGB1 (red fluorescence) of cells after PDT treatment. PDT was performed with 365 nm light (10 mW/cm²). Scale bar: 20 µm.

  DownLoad: Full-Size Img PowerPoint

  Under PDT-induced cell stress, CRT, a hallmark of ICD, was observed to translocate from the ER to the membrane of dying cells. We then investigated the ability of A/C-M (the concentration of AIE-1 is 40 µmol/L) to enhance surface exposure of CRT. As shown in Fig. 2F, significantly enhanced green fluorescence of the CRT antibody was detected when cells were treated with mitochondria-targeted PDT, and the strongest green fluorescence emission was observed after cells were exposed to UV for 6 min and then incubated for 12 h. Moreover, the percentage of cells with enhanced CRT exposure significantly increased when cells were exposed to UV irradiation. The most significant increase in the CRT-positive population was also observed when the cells were exposed to UV radiation for 6 min and then incubated for 12 h (Fig. 2G). The number of CRT-positive cells increased from 2.58% to 45.76% (Fig. S8 in Supporting information). The expression of CRT was also observed by confocal laser scanning microscopy (CLSM), and the observation of cells was consistent with the results of flow cytometry analyses (Fig. 2H).

  We then investigated the ability of A-M to induce ICD. We quantitatively detected ROS in cells by flow cytometry, and found that A-M irradiated for 7 min generated a similar amount of ROS as A/C-M irradiated for 6 min, which was attributed to the effective uptake of positively charged A/C-M by the cells (Fig. 2I). At the same level of ROS, A-M induced a positive cell ratio of only 29.85%, whereas that of A/C-M treated cells was 47.26% (Fig. 2J). Furthermore, the positive cell ratio of A-M-treated cells did not exceed 30% even after increasing the illumination time. These results suggest that A/C-M, which targets the mitochondria, was superior to A-M in inducing CRT exposure. A/C-M can powerfully induce CRT exposure, which can be used as a highly effective ICD inducer for eliciting ICD.

  During PDT, ICD intensity was positively correlated with ROS levels, but only within a certain range. These results suggest that the amount of ROS generated should be considered when evaluating the efficacy of photosensitizer-based ICD inducers.

  The results of the flow cytometry analysis suggested that A/C-M, which targets the mitochondria, was effective in eliciting CRT exposure. CRT exposure of the cells was then observed by CLSM. Stronger green fluorescence of CRT was found in A/C-M + L treated cells, which is consistent with the flow cytometry analysis (Fig. 2K). The secretion of ATP and HMGB1 from the cells was also examined. As shown in Fig. 2L, high levels of ATP were detected in the cell supernatant of the A/C-M+ L group. The concentrations of ATP in the A/C-M+ L group were 0.48-fold and 0.88-fold higher than those in the A-M+L and phosphate buffer solution (PBS) groups. The red fluorescence of HMGB1 in cells was observed by CLSM, and mitochondria-targeted PDT induced the most enhanced release of HMGB1 from the nucleus to the cytoplasm (Fig. 2M). These results highlight the efficacy of A/C-M in inducing ICD, which is promising for the further exploration of the immune activating capacity of mitochondrial PDT vaccines.

  Cell death was assessed by calcein/propidium iodide (PI) cell viability/cytotoxicity, Annexin V-fluorescein isothiocyanate (FITC)/PI apoptosis detection, and MTT assay. As shown in Fig. S9 (Supporting information), PDT-treated cells exhibited weak green fluorescence and strong red fluorescence, indicating a low intracellular esterase activity and a loss of plasma membrane integrity. In addition, in the Annexin V-FITC/PI apoptosis detection assay, PDT-treated cells showed obvious apoptosis (Fig. S10 in Supporting information). In the MTT assay (Figs. S11 and S12 in Supporting information), no obvious cell death was observed in the dark at a concentration of AIE-1 up to 40 µmol/L, whereas the growth of UV-irradiated cells was significantly inhibited. These results indicate that micelle-mediated PDT can effectively induce cell death. Furthermore, more pronounced cell killing effects were observed in A/C-M + L-treated cells, which could be attributed to targeted oxidative damage of the mitochondria by A/C-M.

  After PDT, JC-1 was used to analyze the mitochondrial membrane potential of the cells. Cells treated with PBS, light or micelles emitted strong red fluorescence due to the aggregation of JC-1, whereas those treated with PDT showed weak red fluorescence, demonstrating mitochondrial damage after UV-mediated PDT. The most severe mitochondrial damage was observed in mitochondria-targeted PDT-treated cells, with negligible red fluorescence and highly emissive green fluorescence (Fig. S13 in Supporting information). Flow cytometry analysis showed a similar result that mitochondria-targeted PDT caused more severe mitochondrial damage in cells (Fig. S14 in Supporting information). Western blot analysis showed much higher levels of cleaved caspase 3 in PDT-treated cells, suggesting typical mitochondria-dependent apoptosis (Fig. S15 in Supporting information). Protein kinase R-like endoplasmic reticulum kinase (PERK), located in the ER, plays an important role in regulating cell stress and maintaining calcium homeostasis. Cells, treated with A/C-M + L, experienced more severe mitochondrial oxidative stress, which strongly induced ER stress, and then expressed higher levels of p-PERK, amplifying CRT exposure from the ER to the cell membrane.

  Inactivation of cells with X-ray provides vaccines. Vaccines prepared by different treatments were recorded as the PBS, A-M (A), A/C-M (AC), A-M + L (AL), and A/C-M + L (ACL) groups. No vaccine was added in the control (Ctrl) group. We investigated the immunostimulatory activity of the vaccines in bone marrow dendritic cells (BMDCs) in vitro. The expression of co-stimulatory molecules (CD86 and CD80) on DCs were assessed by flow cytometry. The percentage of cells expressing CD80 and CD86 after co-culture with the vaccines was higher than that in the Ctrl group. The highest level of DCs maturation was observed in the ACL group. The percentage of mature DCs in the AL and ACL groups was 0.74-fold and 1.21-fold higher than that in the Ctrl group, respectively (Figs. 3A and B).

  Figure 3

  

  Figure 3.  (A) Representative flow cytometric plots of the percentage of CD80⁺CD86⁺ DCs after in vitro incubation with different vaccines. (B) Corresponding percentage of DCs maturation after in vitro incubation with different vaccines (n = 3). (C) The level of IL-6 and (D) the level of TNF-α in the supernatant of DCs after treatment with different vaccines (n = 3). (E) Representative flow cytometric plots of the percentage of CD80⁺CD86⁺ DCs in the lymph node of mice (gated by CD11c). (F) Corresponding percentage of DCs maturation in the lymph node of mice (n = 3). (G) The level of IL-6, (H) the level of IFN-γ and (I) the level of TNF-α in the blood serum of mice after treatment with different vaccines (n = 3). (J) t-SNE visualization of clustering of representative markers of T cells from the spleen of vaccinated mice by flow cytometry. Data are presented as the mean ± standard deviation (SD). ***P < 0.001, **P < 0.01, P < 0.05, and ns indicates P > 0.05.

  DownLoad: Full-Size Img PowerPoint

  Mature DCs produce interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α), which activate the immune responses. As shown in Figs. 3C and D, secreted levels of IL-6 and TNF-α were measured. DCs treated with the vaccines secreted increased levels of IL-6 and TNF-α compared to 11 pg/mL of IL-6 and 43 pg/mL of TNF-α in the Ctrl group. The highest levels of IL-6 and TNF-α were observed in the ACL group. The concentrations of IL-6 and TNF-α in the ACL group were 372 and 458 pg/mL, respectively, which were higher than those in the AL group. These results demonstrate that mitochondrial PDT vaccines were more effective in activating DCs, and activated DCs produced more cytokines to induce potent anti-tumor immune responses.

  Mice were immunized with the vaccines and the ability of the vaccines to activate the immune responses in vivo was assessed. All mice were cared according to the guideline of the Care and Use of Laboratory Animals, and the procedures were approved by the China Animal Care and Use Committee of Wuhan University. CD80⁺CD86⁺ DCs in the lymph node were measured by flow cytometry. AL and ACL vaccines significantly enhanced the maturation of DCs (Figs. 3E and F). The potent DC maturation capacity of ACL in vivo was consistent with the in vitro immune activation assay. IL-6 and TNF-α play important roles in T cell activation. The levels of IL-6 and TNF-α in the blood were measured. Mice treated with the vaccines secreted increased levels of IL-6 and TNF-α compared to 64 pg/mL of IL-6 and 63 pg/mL of TNF-α in the Ctrl group. The concentrations of IL-6 and TNF-α in the ACL group were 130 and 170 pg/mL, respectively, which were higher than the 105 pg/mL of IL-6 and 126 pg/mL of TNF-α in the AL group (Figs. 3G and I). Moreover, interferon-γ (IFN-γ) levels were also tested and higher levels of IFN-γ were observed in the AL and ACL groups (Fig. 3H). T cells in the spleen of immunized mice were analyzed, and the results of t-SNE analysis showed an increase in CD8⁺IFNγ⁺ T cells, CD4⁺IFNγ⁺ T cells, and T memory cells. What's more, the ratio of CD8⁺ T cells to CD4⁺ T cells was increased. These results suggest that mice have a better immune response to inhibit tumor growth and metastasis (Fig. 3J and Fig. S16 in Supporting information).

  To investigate the anti-tumor metastasis ability of immunized mice, 5 × 10⁵ 4T1 cells were intravenously injected into the mice. Metastasis nodules of the tumor in the lung and changes in the structure and morphology of the lung were observed 15 days later. ACL vaccines are most effective in activating in vivo immune responses, and the most improved anti-tumor metastasis results are found in mice in the ACL group, with few tumor colonies in appearance and unnoticeable metastatic nodules in the hematoxylin and eosin (H&E) staining result (Fig. 4A).

  Figure 4

  

  Figure 4.  (A) Photograph of tumor metastasis nodules in the lung and H&E staining of the lung on day 15 after tail intravenous injection of 4T1 cells to vaccinated mice. (B) The volume of tumors in vaccinated mice after tumor inoculation (n = 10). (C) Survival curve of vaccinated mice after tumor inoculation (n = 7). (D) Weight of tumors in vaccinated mice on day 34 after tumor inoculation (n = 3). (E) Tumor inhibition rate of the vaccines (n = 10). (F) Photograph of tumors in vaccinated mice. (G) H&E staining and (H) immunofluorescence staining of tumors on day 34 after tumor inoculation in vaccinated mice. Scale bar: 200 µm. Data are presented as the mean ± SD. ***P < 0.001, P < 0.05, and ns indicates P > 0.05.

  DownLoad: Full-Size Img PowerPoint

  Tumor cells were subcutaneously injected into the right back of mice to access their anti-tumor activity. As shown in Fig. 4B, rapid tumor growth was observed in mice in the Ctrl group, with the average tumor volume reaching 1953 mm³ on day 34 of the experiment. Insignificant anti-tumor effects were observed in mice treated with vaccines without ICD induction, and the mean tumor volumes in the PBS, A, and AC groups were 1819, 1676 and 1656 mm³, respectively. A modest antitumor effect was observed in mice treated with the AL vaccines, with a mean tumor volume of 1223 mm³ and a significant inhibition of tumor growth was observed in mice treated with the ACL vaccines, with a mean volume of 544 mm³. In addition, two of the seven mice in the AL group and all the mice in the ACL group survived 44 days after live 4T1 inoculation, whereas the mice in the other groups all died. On day 51, all mice in the AL group had died and six of the seven mice in the ACL group were still alive. These results confirmed that ACL vaccines, which activated potent in vivo immune responses, can significantly delay tumor growth and prolong the survival of mice (Fig. 4C). The tumor inhibition rates in the PBS, A, AC, AL, ACL groups were 6.86%, 14.14%, 15.16%, 37.35% and 72.13%, respectively (Fig. 4E). Weight and photograph (Figs. 4D and F) of the corresponding tumors were recorded. Significant reductions in tumor size and weight were observed in the AL and ACL groups compared with the other groups, with ACL exerting the most significant inhibition of tumor growth. H&E staining revealed similar results: tumor in the ACL group showed the largest necrotic areas with prominent cell separation and nuclear fragmentation (Fig. 4G). These results demonstrate that the ACL vaccines were effective in inhibiting tumor growth and tumor metastasis.

  Immune cells in the tumor tissue were analyzed by immunofluorescence staining and flow cytometry. As shown in Fig. 4H, mice vaccinated with the PDT vaccines showed increased infiltration of CD8⁺ T cells into the tumor. Cytotoxic T cells induce apoptosis of tumor cells by inducing caspase 3 activation. The enhanced expression levels of caspase 3 correlate well with the increased infiltration of CD8⁺ T cells in the tumor of mice treated with AL and ACL vaccines, indicating better tumor inhibitory effects. As shown in Fig. S17 (Supporting information), the number of tumor infiltrating regulatory T cells (Tregs, CD25⁺Foxp3⁺), effector memory T cells (Tems, CD44⁺CD62L⁻) and cytotoxic T lymphocytes (CTLs, CD8⁺IFNγ⁺) was examined. Decreased proportions of Tregs and increased proportions of Tems and CTLs were observed in the PDT vaccine groups. The high proportion of CTL in the ACL group was consistent with the results of immunofluorescence staining. What's more, the ratio of CD8⁺ T cells to CD4⁺ T cells was increased. These results indicate that PDT vaccines were beneficial for enhancing immune responses in the tumor (Fig. S18 in Supporting information). CD80⁺CD86⁺ DCs in the lymph node were measured, and ACL vaccines significantly enhanced the maturation of DCs, with 24.6% of DCs activated, which was twice as many as that in the Ctrl group.

  Throughout the course of anti-tumor treatment, no significant changes in the body weight of mice were observed (Fig. S20 in Supporting information). Routine blood tests showed that the blood components of the mice were at normal levels. These results suggest that the vaccines do not cause any apparent systemic toxicity (Fig. S21 in Supporting information). The heart, liver, spleen, lung, and kidney of the mice were collected for pathological observation, and no pathological lesions were found (Fig. S19 in Supporting information). These results demonstrate that the vaccines are safe and suitable for in vivo application.

  In summary, we designed a mitochondria-targeted nanophotosensitizer as a potent ICD inducer for vaccine preparation and tumor prevention. Under UV irradiation, singlet oxygen generated in the mitochondria by A/C-M evoked massive ICD in cells. The strongest induction of ICD was found when cells were treated with A/C-M, irradiated with UV for 6 min and then incubated for 12 h. Further inactivation of cells with X-ray afforded vaccines. We found mitochondrial PDT vaccines were effective in activating immune responses, and thus significantly inhibited tumor growth and metastasis. In short, A/C-M is a potent ICD inducer capable of inducing large-scale ICD in cells, and A/C-M-developed vaccines are effective in activating immune responses to inhibit tumor growth and metastasis.

  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.

  Acknowledgment

  This study was supported by the National Natural Science Foundation of China (Nos. 52173137 and 51873163).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2023.108882.

  
  
• 1. [1]

     N. Casares, M.O. Pequignot, A. Tesniere, et al., J. Exp. Med. 202 (2005) 1691–1701. doi: 10.1084/jem.20050915

  2. [2]

     M. Jiang, J. Zeng, L. Zhao, et al., Nanoscale 13 (2021) 17218–17235. doi: 10.1039/d1nr05512g

  3. [3]

     Q. Chen, J. Chen, Z. Yang, et al., Adv. Mater. 31 (2019) 1802228. doi: 10.1002/adma.201802228

  4. [4]

     S. Jia, Z. Gao, Z. Wu, et al., CCS Chem. 4 (2022) 501–514. doi: 10.31635/ccschem.021.202101458

  5. [5]

     Z. Wang, F. Zhang, D. Shao, et al., Adv. Sci. 6 (2019) 1901690. doi: 10.1002/advs.201901690

  6. [6]

     D. Wei, Y. Chen, Y. Huang, et al., Nano Today 41 (2021) 101288. doi: 10.1016/j.nantod.2021.101288

  7. [7]

     A.D. Garg, A.M. Dudek-Peric, E. Romano, P. Agostinis, Int. J. Dev. Biol. 59 (2015) 131–140. doi: 10.1387/ijdb.150061pa

  8. [8]

     D.V. Krysko, A.D. Garg, A. Kaczmarek, et al., Nat. Rev. Cancer 12 (2012) 860–875. doi: 10.1038/nrc3380

  9. [9]

     L. Fu, X. Ma, Y. Liu, Z. Xu, Z. Sun, Chin. Chem. Lett. 33 (2022) 1718–1728. doi: 10.1016/j.cclet.2021.10.074

  10. [10]

      D. Ding, X. Jiang, Small Methods 7 (2023) 2300354. doi: 10.1002/smtd.202300354

  11. [11]

      B. Ji, M. Wei, B. Yang, Theranostics 12 (2022) 434–458. doi: 10.7150/thno.67300

  12. [12]

      M. Zhang, Y. Zhao, H. Ma, Y. Sun, J. Cao, Theranostics 12 (2022) 4629–4655. doi: 10.7150/thno.72465

  13. [13]

      M. Korbelik, J. Banath, K.M. Saw, W. Zhang, E. Ciplys, Front. Oncol. 5 (2015) 15.

  14. [14]

      J. Li, J. Dai, Z. Zhuang, et al., Biomaterials 291 (2022) 121899. doi: 10.1016/j.biomaterials.2022.121899

  15. [15]

      F. Jin, D. Liu, X. Xu, J. Ji, Y. Du, Int. J. Nanomed. 16 (2021) 4693–4712. doi: 10.2147/ijn.s314506

  16. [16]

      T. Panaretakis, O. Kepp, U. Brockmeier, et al., EMBO J. 28 (2009) 578–590. doi: 10.1038/emboj.2009.1

  17. [17]

      M. Obeid, A. Tesniere, F. Ghiringhelli, et al., Nat. Med. 13 (2007) 54–61. doi: 10.1038/nm1523

  18. [18]

      S.J. Gardai, K.A. McPhillips, S.C. Frasch, et al., Cell 123 (2015) 321–334.

  19. [19]

      O. Kepp, L. Menger, E. Vacchelli, et al., Cytokine Growth Factor Rev. 24 (2013) 311–318. doi: 10.1016/j.cytogfr.2013.05.001

  20. [20]

      X. Liu, Y. Liu, X. Li, et al., ACS Nano 16 (2022) 9240–9253. doi: 10.1021/acsnano.2c01669

  21. [21]

      H. Ma, Y. Lu, Z. Huang, et al., J. Am. Chem. Soc. 144 (2022) 3477–3486. doi: 10.1021/jacs.1c11886

  22. [22]

      H. Deng, Z. Zhou, W. Yang, et al., Nano Lett. 20 (2020) 1928–1933. doi: 10.1021/acs.nanolett.9b05210

  23. [23]

      P. Zheng, B. Ding, Z. Jiang, et al., Nano Lett. 21 (2021) 2088–2093. doi: 10.1021/acs.nanolett.0c04778

  24. [24]

      Q. Jiang, C. Zhang, H. Wang, et al., Front. Oncol. 9 (2019) 1196. doi: 10.3389/fonc.2019.01196

  25. [25]

      J. Zhang, D. Zhang, Q. Li, et al., ACS Appl. Mater. Interfaces 11 (2019) 42904–42916. doi: 10.1021/acsami.9b13556

  26. [26]

      K. Deng, H. Yu, J.M. Li, et al., Biomaterials 275 (2021) 120959. doi: 10.1016/j.biomaterials.2021.120959

  27. [27]

      H. Zhao, J. Xu, C. Feng, et al., Adv. Mater. 34 (2022) 2106390. doi: 10.1002/adma.202106390

  28. [28]

      J. Qi, H. Ou, Q. Liu, D. Ding, Aggregate 2 (2021) 95–113. doi: 10.1002/agt2.25

  29. [29]

      H. Wang, Z. He, X.A. Liu, et al., Small Struct. 3 (2022) 2200036. doi: 10.1002/sstr.202200036

  30. [30]

      Y. Wang, N. Niu, Y. Huang, et al., Small methods 6 (2022) 2200393. doi: 10.1002/smtd.202200393

  31. [31]

      H. Zhao, N. Li, C. Ma, et al., Chin. Chem. Lett. 34 (2023) 107699. doi: 10.1016/j.cclet.2022.07.042

  32. [32]

      J. Li, H. Gao, R. Liu, et al., Sci. China Chem. 63 (2020) 1428–1434. doi: 10.1007/s11426-020-9846-4

  33. [33]

      S. Wang, C. Chen, J. Wu, et al., Sci. China Chem. 65 (2022) 870–876. doi: 10.1007/s11426-021-1207-0

  34. [34]

      C. Chen, X. Ni, S. Jia, et al., Adv. Mater. 31 (2019) 1904914. doi: 10.1002/adma.201904914

• 

  Figure 1  (A) Schematic illustration of the prepared micelles. (B) Chemical structure of C₁₆-AIE-PEG₂₀₀₀ (AIE-1). (C) Fluorescence intensity at 525 nm of AIE-1 (40 µmol/L) in DMSO/H₂O mixture. (D) Determination of the critical micelle concentration (CMC) of AIE-1. (E) Absorption spectra of micelles in water (the concentration of AIE-1 is 40 µmol/L). (F) Fluorescence spectra of micelles in water (the concentration of AIE-1 is 40 µmol/L). (G) Zeta potential and (H) size distribution of A-M and A/C-M in water. (I, J) Fluorescence intensity at 525 nm of solution containing different concentrations of micelles. (K, L) TEM images of A-M and A/C-M. (M, N) Fluorescence spectra of DCFH solution containing A-M or A/C-M under UV irradiation for a period of time. (O, P) Fluorescence spectra of SOSG solution containing A-M or A/C-M under UV irradiation for a period of time. (Q, R) Fluorescence spectra of TA solution containing A-M or A/C-M under UV irradiation for a period of time. 365 nm, 10 mW/cm². Scale bar: 50 nm.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (A) CLSM images of 4T1 cells treated with A-M or A/C-M (green fluorescence) for 8 h and 12 h, which were co-stained with Mito-Tracker Red. (B, C) Fluorescence intensity profile of Mito-Tracker and A-M when cells were treated with A-M for 8 h (B) or 12 h (C) based on typical images in (A). (D, E) Fluorescence intensity profile of Mito-Tracker and A/C-M when cells were treated with A/C-M for 8 h (D) or 12 h (E) based on typical images in (A). (F) The mean fluorescence intensity and (G) the quantity of CRT-positive cells when cells were treated with A/C-M for 12 h, irradiated with UV for a period of time and then incubated for 8, 12 or 24 h. (H) CLSM image of CRT (green fluorescence) on 4T1 cells when cells were irradiated with UV for a period of time and then incubated for 12 h. (I) Flow cytometry analysis of ROS in cells after cells were treated with micelles for 12 h and then irradiated with UV for a period of time. (J) Flow cytometry analysis of CRT on 4T1 cells after cells were treated with micelles for 12 h, irradiated with UV for a period of time and then incubated for 12 h. (K) Surface exposed CRT (green fluorescence), (L) released ATP, (M) and secreted HMGB1 (red fluorescence) of cells after PDT treatment. PDT was performed with 365 nm light (10 mW/cm²). Scale bar: 20 µm.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (A) Representative flow cytometric plots of the percentage of CD80⁺CD86⁺ DCs after in vitro incubation with different vaccines. (B) Corresponding percentage of DCs maturation after in vitro incubation with different vaccines (n = 3). (C) The level of IL-6 and (D) the level of TNF-α in the supernatant of DCs after treatment with different vaccines (n = 3). (E) Representative flow cytometric plots of the percentage of CD80⁺CD86⁺ DCs in the lymph node of mice (gated by CD11c). (F) Corresponding percentage of DCs maturation in the lymph node of mice (n = 3). (G) The level of IL-6, (H) the level of IFN-γ and (I) the level of TNF-α in the blood serum of mice after treatment with different vaccines (n = 3). (J) t-SNE visualization of clustering of representative markers of T cells from the spleen of vaccinated mice by flow cytometry. Data are presented as the mean ± standard deviation (SD). ***P < 0.001, **P < 0.01, P < 0.05, and ns indicates P > 0.05.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (A) Photograph of tumor metastasis nodules in the lung and H&E staining of the lung on day 15 after tail intravenous injection of 4T1 cells to vaccinated mice. (B) The volume of tumors in vaccinated mice after tumor inoculation (n = 10). (C) Survival curve of vaccinated mice after tumor inoculation (n = 7). (D) Weight of tumors in vaccinated mice on day 34 after tumor inoculation (n = 3). (E) Tumor inhibition rate of the vaccines (n = 10). (F) Photograph of tumors in vaccinated mice. (G) H&E staining and (H) immunofluorescence staining of tumors on day 34 after tumor inoculation in vaccinated mice. Scale bar: 200 µm. Data are presented as the mean ± SD. ***P < 0.001, P < 0.05, and ns indicates P > 0.05.

  下载: 全尺寸图片 幻灯片
•