Abstract

   In the treatment of uveal melanoma (UVM), histone deacetylase
   inhibitors (HDACi) have emerged as a promising epigenetic therapy.
   However, their clinical efficacy is hindered by the suboptimal
   pharmacokinetics and the strong self‐rescue of tumor cells. To overcome
   these limitations, reactive oxygen species (ROS)‐responsive
   nanoparticles (NPs) are designed that encapsulate HDACi MS‐275 and the
   glutamine metabolism inhibitor V‐9302. Upon reaching the tumor
   microenvironment, these NPs can disintegrate, thereby releasing MS‐275
   to increase the level of ROS and V‐9302 to reduce the production of
   glutathione (GSH) related to self‐rescue. These synergistic effects
   lead to a lethal ROS storm and induce cell pyroptosis. When combined
   with programmed cell death protein 1 monoclonal antibodies (α‐PD‐1),
   these NPs facilitate immune cell infiltration, improving anti‐tumor
   immunity, converting “immune‐cold” tumors into “immune‐hot” tumors, and
   enhancing immune memory in mice. The findings present a nano‐delivery
   strategy for the co‐delivery of epigenetic therapeutics and metabolic
   inhibitors, which induces pyroptosis in tumors cells and improves the
   effectiveness of chemotherapy and immunotherapy.

   Keywords: epigenetic therapy, immunotherapy, metabolic intervention,
   pyroptosis, uveal melanoma
     __________________________________________________________________

   The co‐delivery of histone deacetylase inhibitor MS‐275 and glutamine
   metabolism inhibitor V‐9302 via nanomaterials surpasses traditional
   monotherapy limitations in treating uveal melanoma, inducing ROS storms
   and achieving unexpected pyroptosis. This study presents an effective
   nanomedicine that elicits a potent immune response through the
   induction of pyroptosis, resulting in ideal coordination capacity.

   graphic file with name ADVS-11-2404375-g004.jpg

1. Introduction

   Uveal melanoma (UVM), the most common intraocular malignancy in adults,
   presents a formidable clinical challenge due to its high mortality
   rate, 45% within 15‐years and its propensity for metastasis.^[ [42]^1
   ^] Standard treatments such as surgery and radiotherapy, while
   effective for local tumors during early stages, fails to treat
   micro‐metastases and may cause damage to the surrounding eye tissues.^[
   [43]^2 ^] Cancer immunotherapy has emerged as a transformative
   approach, engaging the body's host immune system to recognize and
   eliminate tumor cells precisely.^[ [44]^3 ^] However, the unique
   characteristics of UVM, including low mutational load, poor
   immunogenicity, and the immunosuppressive tumor microenvironment,
   compromise the effectiveness of immunotherapy.^[ [45]^4 , [46]^5 ^]
   Developing strategies that increase tumor immunogenicity and enhance
   the responses of UVM to immunotherapy is of great significance.

   Epigenetic therapies have emerged as potential options to enhance the
   immunogenicity of cancer cells and improve anti‐tumor immunity, with
   histone deacetylase inhibitors (HDACi) at the forefront.^[ [47]^6 ,
   [48]^7 , [49]^8 ^] Despite their promise, the performance of HDACi in
   treating solid tumors is often unsatisfactory.^[ [50]^9 , [51]^10 ,
   [52]^11 ^] Monotherapy of HDACi in UVM is limited due to
   dose‐dependency, cellular metabolism, and cellular acetylation flux of
   the cell.^[ [53]^12 , [54]^13 ^] Preclinical studies suggested that,
   the high reactive oxygen species (ROS) environment in tumor cells
   usually results in activation of the mTOR pathway, which promotes
   glutamine uptake to generate glutathione (GSH) in order to sufficiently
   keep the cellular redox state.^[ [55]^14 ^] Notably, HDACi can disrupt
   the mTOR pathway but hard to cope with the uptake of glutamine by tumor
   cells.^[ [56]^12 , [57]^15 , [58]^16 , [59]^17 , [60]^18 ^] Such
   GSH‐associated ROS detoxification substantially limits the efficacy
   HDACi. Considering glutamine is essential for the biosynthesis of GSH
   in cancer cells, we hypothesize that employing HDACi to interfere with
   the mTOR pathway and concurrently inhibiting glutamine uptake can
   induce an uncontrollable ROS storm beyond the limit of a single‐agent
   action.

   Pyroptosis, a pro‐inflammatory form of cell death, could serve as a
   linchpin in this strategy. Canonical pyroptosis is characterized by the
   cleavage of Gasdermin D (GSDMD) to form the N‐terminal of GSDMD
   (NT‐GSDMD) through cysteine asparaginase‐1, which perforates the cell
   membrane and releases pro‐inflammatory cytokines.^[ [61]^19 ^] It has
   been reported that ROS storm can contribute to both NT‐GSDMD
   oligomerization and pore‐forming activity, leading to enhanced
   pyroptosis.^[ [62]^20 , [63]^21 , [64]^22 , [65]^23 ^] Therefore,
   combining HDACi with metabolic inhibitors to simultaneously inhibit the
   mTOR activity and glutamine uptake can induce the ROS storm, resulting
   in cancer cell pyroptosis and promoting tumor immunogenicity.

   Herein, we have combined MS‐275, a commercial HDACi,^[ [66]^17 ^] with
   V‐9302, a glutamine uptake inhibitor,^[ [67]^24 ^] for the treatment of
   UVM (Figure [68] 1a). Both MS‐275 and V‐9302 show unsatisfactory
   pharmacokinetics, low water solubility, and poor tumor accumulation,
   limiting their in vivo applications.^[ [69]^12 , [70]^25 , [71]^26 ,
   [72]^27 ^] To address these challenges, we have developed
   ROS‐responsive nanoparticles (NPs) loaded with both drugs. In the tumor
   microenvironments with high ROS levels, NPs can be degraded, releasing
   both MS‐275 and V‐9302, which synergistically induce the ROS storm and
   tumor cell pyroptosis. After intravenous injection into mice bearing
   tumors, NPs could selectively accumulate in the tumor sites and
   effectively inhibit tumor progression. When combined with programmed
   cell death protein 1 monoclonal antibody (α‐PD‐1), treatment of NPs
   induced a long‐lasting immune memory and prevent tumor reoccurrence
   (Figure [73]1b). Our study unlocks the potential of combining
   epigenetic therapies with metabolic inhibitors, offering a promising
   option for the treatment of UVM.

Figure 1.

   Figure 1
   [74]Open in a new tab

   NPs induced pyroptosis in tumor cells, thereby augmenting the efficacy
   of tumor therapy when combined with α‐PD‐1. a) The strategic design of
   NPs to incite pyroptosis. b) NPs triggered pyroptosis, transformed
   “immune cold” tumors into “immune hot” tumors, thereby eliciting potent
   anti‐tumor immune responses in vivo. The combination of NPs and α‐PD‐1
   further optimized tumor suppression. c) The association between
   disease‐free survival and HDAC3. d) The expression of SLC1A5 in
   patients with UVM. e) The correlation at the mRNA level between the
   expression of HDAC3 and mTOR in tumor tissues of UVM. f) The
   correlation at the mRNA level between the expression of SLC1A5 and mTOR
   in tumor tissues of UVM. Data in (c–f) were analyzed using GEPIA
   ([75]http://gepia.cancer‐pku.cn/).

2. Results and Discussion

2.1. Expression and Correlation Analysis of HDAC and Metabolism‐Related Genes
in UVM

   To assess the potential of using HDACi and metabolic intervention
   drugs, we obtained clinical data of UVM patients from public databases
   and performed bioinformatics analysis. UVM patients with low expression
   levels of histone deacetylase (HDAC3) and the amino acid transporter
   SLC1A5 were found to have longer disease‐free survival
   (Figure [76]1c,d; Figure [77]S1a,b, Supporting Information). These
   results suggested that the inhibition of HDAC proteins with HDACi and
   the inhibition of amino acid transporter with metabolic intervention
   drugs have the potential to treat UVM. Previous studies have reported
   that the perturbation of mTOR protein could be achieved by interfering
   with HDAC enzymes.^[ [78]^17 ^] SLC1A5, as a key protein in the
   metabolic pathway, is also associated with mTOR expression.^[ [79]^28
   ^] Gene correlation analysis showed that the expression of HDAC3 was
   significantly positively correlated with mTOR (R = 0.82), whereas
   SLC1A5 was positively correlated with mTOR (R = 0.48). This suggested
   that the modulation of mTOR expression in UVM by interfering with HDAC3
   may be more potent compared to that of SLC1A5 (Figure [80]1e,f). After
   focusing on mTOR expression in UVM, it was found that UVM patients with
   low mTOR expression had longer overall survival (Figure [81]S1c,
   Supporting Information). The results of the bioinformatics analysis
   lend credence to our previous conjecture that it is possible to prolong
   the survival of UVM patients by interfering with the expression of HDAC
   and metabolism‐related proteins, which may have been achieved by
   inhibiting mTOR expression.

2.2. Preparation and Characterization of Nanoparticles

   The bioinformatics analysis validated the potential of combining HDACi
   and metabolic inhibitors. Therefore, the cytotoxicity of MS‐275,
   V‐9302, and their combination, was first evaluated with
   3‐(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyl‐2H‐tetrazolium bromide (MTT)
   assay. At an equivalent dosage, the combination treatment of MS‐275 and
   V‐9302 significantly outperformed each single treatment in isolation
   (Figure [82]S2, Supporting Information). Inspired by the synergistic
   effects of MS‐275 and V‐9302, we designed NPs that encapsulated both
   drugs. Given the high ROS levels in the tumor environment, we built the
   shell of the NPs with the previously described ROS‐sensitive polymer
   (P1) to achieve targeted release at the tumor site.^[ [83]^29 , [84]^30
   , [85]^31 ^]

   To prepare NPs, MS‐275 and V‐9302 were individually dissolved in DMSO
   and subsequently amalgamated to yield a mixture. NPs were synthesized
   via self‐assembly by mixing the MS‐275 and V‐9302 with P1 and
   PEG[2k]‐DSPE. Transmission electron microscopy (TEM) results showed
   that NPs had an average diameter of ≈150 nm with intact spherical
   structure (Figure [86] 2a). The mean particle dimensions of the NPs
   were 145.4 nm, with a polydispersity index (PDI) of 0.08
   (Figure [87]2b). The size of NPs remained stable throughout 15 days of
   storage in serum by DLS (Figure [88]S3a, Supporting Information). The
   shell of the NPs comprised the ROS‐responsive polymer P1, which,
   theoretically, would undergo cleavage in the high‐ROS environment of
   the tumor. Consequently, an in vitro simulation of a high‐ROS tumor
   environment was conducted. The results revealed that under high ROS
   conditions (10 mm H[2]O[2]), the mean particle size of the NPs was
   96.3 nm, and the PDI escalated to 0.18 (Figure [89]S3b, Supporting
   Information). This substantiated the capability of the NPs to undergo
   cleavage in a high ROS environment.

Figure 2.

   Figure 2
   [90]Open in a new tab

   Characterization and anti‐cancer efficacy of NPs in vitro. a) TEM
   images of NPs. Scale bar: 100 nm. b) Hydrodynamic diameters of NPs
   measured by dynamic light scattering (DLS). c) CLSM images of OCM‐1
   cells treated with NPs‐Cy5.5 for 1, 4, and 7 h, respectively. Scale
   bar: 20 µm. d) Fluorescence intensity in mice at different time points
   and major organs by IVIS. e) Relative cell viabilities of OCM‐1, MUM‐2B
   and 92.1 cells with 24 h treatment of NPs via MTT assay, respectively.
   f) Quantification of apoptotic ratio via flow cytometry (FCM). g) Cell
   cycle profiles and quantification of cell cycle ratio via FCM. All
   groups were treated with the following compound concentration for 24 h:
   MS‐275 (10 µm), V‐9302 (2 µm), MS‐275 (10 µm)+V‐9302 (2 µm) or NPs
   (10 µm). Data were from three independent experiments. Data are
   presented as mean ± SD. Statistical significance between all groups was
   calculated via one‐way ANOVA in (f and g). ^** p < 0. 01, ^*** p < 0.
   001, ns, not significant.

   The uptake of NPs by cancer cells was examined. To visualize the
   cellular uptake process of NPs, they were labeled with Cy5.5 dye (red)
   to form NPs‐Cy5.5. OCM‐1 cells were exposed to NPs‐Cy5.5 for varying
   durations (1, 4, 7 h). The intensity of intracellular red fluorescence
   progressively amplified with the extension of treatment time, as
   observed through a confocal laser scanning microscope (CLSM),
   indicating the rapid uptake of NPs by tumor cells (Figure [91]2c;
   Figure [92]S4, Supporting Information). To further scrutinize the
   biodistribution of NPs in vivo, the OCM‐1 tumor‐bearing BALB/c nude
   mouse model were constructed. Utilizing an in vivo imaging system
   (IVIS), the fluorescence signals of BALB/c nude mice intravenously
   injected with Cy7.5‐labeled NPs (NPs‐Cy7.5) were monitored
   (Figure [93]2d). The fluorescence intensity at the tumor site steadily
   escalated, peaking at ≈12 h with a value of 2.09 × 10^8 p/s/cm^2/sr.
   Over time, the fluorescence value gradually declined, with the total
   fluorescence value detected after 48 h being 2.00 × 10^8 p/s/cm^2/sr
   (Figure [94]S5a, Supporting Information). Following 48 h of compound
   administration, ex vivo imaging of major organs and tumors in mice was
   conducted. The average fluorescence value of tumors ranked second among
   all organs at 9.72 × 10^7 p/s/cm^2/sr, significantly surpassing that of
   organs such as the heart, spleen, lungs, and intestines (Figure
   [95]S5b, Supporting Information). These results indicated that NPs can
   rapidly accumulate in the tumor site after systemic injection, showing
   prolonged retention.

2.3. The Anti‐Cancer Activity of NPs In Vitro

   To evaluate the anti‐cancer activity of NPs in vitro, the cytotoxicity
   of NPs in UVM cell lines (including OCM‐1, MUM‐2B, 92.1) was performed
   by MTT assays. Compared to the other two cell lines, the OCM‐1 cell
   line exhibited a lower IC[50] of NPs (Figure [96]2e; Table [97]S1,
   Supporting Information). The two compounds were amalgamated at the
   ratio present in nanoparticles as the combination group (MS+V), and the
   IC[50] of MS+V could attain 6.29 µm in OCM‐1 after 48 h, which was
   lower than that of NPs for 24 h (Figure [98]S6 and Table [99]S2,
   Supporting Information). These results suggested that the enhanced
   efficacy of NPs may result from efficient cellular uptake,
   corroborating previous conjecture. Subsequently, cell apoptosis was
   analyzed under varying treatments by Annexin V‐FITC and propidium
   iodide (PI) double staining. Specifically, OCM‐1 cells treated with NPs
   exhibited an apoptosis rate of 33.3%, while the apoptosis rate of cells
   treated with an equivalent concentration of MS+V was 25.5%
   (Figure [100]2f; Figure [101]S7, Supporting Information). Next, as both
   MS‐275 and V‐9302 demonstrated blockage of cell‐related cycles in prior
   studies,^[ [102]^12 , [103]^32 , [104]^33 ^] the effects of different
   compounds on the cell cycle were observed via the kit. NPs, MS‐275, and
   MS+V all impeded the cell cycle at the G[0]/G[1] phase
   (Figure [105]2g). Specifically. NPs enter into the cells via
   endocytosis. In this way, the cellular uptake of drugs is much more
   efficient than molecule MS‐275 or V‐9302, while molecule V‐9302 is even
   insoluble in water. Additionally, NPs composed the optimized ratios of
   MS‐275 and V‐9302 while molecule MS‐275 or V‐9302 enter cells randomly.
   These findings suggested that NPs are more cytotoxic than merely
   combining two compounds, and that this effect was not attributable to
   enhancing the characteristics of a single ingredient.

2.4. Mechanism of Action of NPs via RNA‐seq Analysis

   To investigate the anti‐cancer mechanism of NPs and to explore the
   mechanism behind the enhanced efficacy post‐nanosizing, a genome‐wide
   RNA‐seq analysis was performed on OCM‐1 cells treated with different
   compounds. A total of 282185 genes were identified. 701 genes were
   differential expressed only in cells treated with NPs versus PBS, 733
   genes were differential expressed only in cells treated with MS+V
   versus PBS, and 27 genes were differential expressed only in cells
   treated with NPs versus MS+V (Figure [106] 3a). Meanwhile, the total
   number of differentially expressed genes in cells treated with either
   NPs or MS+V was similar, suggested that the nano‐delivery system did
   not cause significant difference compared to simple mixed drug
   delivery. The 27 differentially expressed genes specific to the NPs
   treatment and the MS+V treatment were involved in transmembrane
   proteins, mTOR pathway regulation, and glutamate transport, etc (Table
   [107]S3, Supporting Information). This suggested that the nanodelivery
   platform stimulates substance transport links such as vesicular
   translocation, and potentially having a stronger stimulation of the
   mTOR pathway and glutamate metabolism than MS+V, which was consistent
   with our previous results. Moreover, 6587 genes were up‐regulated
   (depicted as red dots) and 5491 genes were down‐regulated (depicted as
   blue dots) in NPs‐treated cells relative to PBS‐treated cells; 6642
   genes were up‐regulated and 5488 genes were down‐regulated in MS+V
   ‐treated cells relative to PBS‐treated cells; 930 genes were
   up‐regulated and 775 genes were down‐regulated in NPs‐treated cells
   relative to MS+V‐treated cells (Figure [108]3b‐[109]d). After
   treatment, more differential genes showed a tendency to be
   up‐regulated. The genes differentially expressed in the NPs and the PBS
   were then subjected to Kyoto Encyclopedia of Genes and Genomes (KEGG)
   enrichment analysis. NPs primarily impacted the pathways involved in
   the Cell cycle, DNA replication, and Chemical carcinogenesis‐reactive
   oxygen species (Figure [110]3e). Moreover, NPs significantly affected
   the expression of genes in multiple signaling pathways of
   differentially expressed genes in the NPs versus MS+V (Figure [111]3f).
   The preceding results indicated that NPs interfered with Cell cycle,
   DNA replication, and Chemical carcinogenesis‐reactive oxygen species;
   and the effect of nanosizing on anti‐cancer efficacy was primarily
   achieved by influencing various signaling pathways.

Figure 3.

   Figure 3
   [112]Open in a new tab

   Transcription analysis of OCM‐1 cells by RNA‐seq after various
   treatments. a) Differential Venn diagram revealed the number of genes
   transcribed in each treatment group. b–d) Volcano plots displayed the
   differentially expressed genes. e,f) KEGG pathway enrichment analysis.
   g) Heatmap of gene expressions in cells treated with PBS, NPs and MS+V.
   h) Differential variable splicing event statistics. i) GSEA analysis
   showed the gene sets of Cell cycle, Reactive oxygen species, mTOR
   pathway and Glutathione metabolism.

   To further explore the differences between the groups, the
   differentially expressed genes after treatment in each group were
   visualized by a heatmap (Figure [113]3g). Compared to PBS, cells
   treated with the two compounds exhibited an upsurge in the expression
   of some pivotal genes, such as ERRFI1 and H1‐2, which potentially
   inhibit tumor invasion and metastasis and influence the function of
   downstream histone proteins. The differences between NPs and MS+V were
   relatively modest. Consequently, alternative splicing analysis was
   employed to probe the causes of the variations between NPs and MS+V
   (Figure [114]3h). Alternative splicing is the process of generating
   diverse mRNA splicing isoforms from an mRNA precursor via different
   splicing methods (selecting varying combinations of splice sites).^[
   [115]^34 ^] It is a crucial mechanism for regulating gene expression
   and engendering proteomic diversity. The ratio of exon skip underwent
   significant alteration between NPs and MS+V, indicating that the
   performance difference between the two may be reflected through exon
   skip. Subsequently, according to Gene Set Enrichment Analysis (GSEA)
   analysis, NPs considerably down‐regulated the expression of genes
   related to cell cycle in OCM‐1 cells, whereas the expression of genes
   associated with reactive oxygen species was moderately changed. RNA
   expression in the mTOR pathway and glutathione metabolism was
   dramatically escalated by NPs as compared to MS+V (Figure [116]3i).
   These results collectively prove that NPs possess enhanced anti‐cancer
   properties by simultaneously influencing the mTOR pathway and glutamine
   metabolism.

2.5. NPs Formed a ROS Storm and Induced Pyroptosis In Vitro

   Previous genome‐wide RNA‐seq results suggested that the efficacy of NPs
   may be achieved by interfering with the mTOR pathway and glutathione
   metabolism, aligning with the characteristics of the two individual
   compounds. To detect the perturbation of the mTOR pathway instigated by
   NPs, the expression of pertinent proteins in cells post‐treatment was
   observed via western blotting (WB). The expression of Akt, Phospho‐Akt,
   mTOR, Phospho‐mTOR, and downstream proteins were down‐regulated in
   cells treated with NPs and MS+V (Figure [117]4a; Figure [118]S8,
   Supporting Information). Specifically, MS‐275 reduced the expression of
   Akt, and V‐9302 had a greater effect on the expression of Phospho‐mTOR.
   Relevant proteins in the mTOR pathway such as Akt, Phospho‐Akt,
   Phospho‐S6 were down‐regulated to a certain extent after treated with
   MS+V. NPs encapsulated both MS‐275 and V‐9302, which inherited the
   properties of the single compound in previous experiments. The effects
   of MS+V can even be amplified due to the efficient delivery
   capabilities of nanosystems. As a result, the relevant proteins of the
   mTOR pathway were significantly down‐regulated in the cells treated
   with NPs. Subsequently, the level of GSH was assessed, and the results
   indicated that NPs have the same strength as V‐9302 to reduce GSH
   content in cellular (Figure [119]4b). NPs effectively suppressed the
   mTOR pathway and led to a decrease in intracellular GSH levels, which
   could increase ROS production, and decrease ROS consumption.^[ [120]^35
   , [121]^36 ^] Therefore, NPs may cause a lethal ROS storm.

Figure 4.

   Figure 4
   [122]Open in a new tab

   NPs interfered with the mTOR pathway and deprivated glutamine to form
   ROS storms, which induced pyroptosis in vitro. a) Expression levels of
   mTOR proteins related in OCM‐1 cells after various treatments for 24 h
   by WB. β‐Actin was used as the internal reference protein. b) Relative
   GSH level of cells after various treatments. c) CLSM images of ROS in
   cells after various treatments. (blue, DAPI; green, ROS). d) CLSM
   images of mitoROS in cells. (blue, DAPI; red, MitoSOX). e) Pyroptosis
   cells under microscope treated with PBS, MS+V and NPs for 24 h. f)
   Relative LDH level in OCM‐1 cells supernatant of different groups. g)
   Quantification of IL‐18 in OCM‐1 cells supernatant of different groups.
   h) Expression levels of pyroptosis‐related proteins in cells after
   various treatments for 24 h by WB. β‐Actin was used as the internal
   reference protein. i) Relative expression of NT‐GSDMD. Data were from
   three independent experiments. Data were presented as mean ± SD.
   Statistical significance between all groups was calculated via one‐way
   ANOVA in (b, f, g, and i). ^* p < 0.05, ^** p < 0. 01, ^*** p < 0.001,
   ^**** p < 0.0001, ns, not significant.

   To observe the ROS storm more precisely, ROS probes were employed to
   detect intracellular ROS (green) via CLSM (Figure [123]4c; Figure
   [124]S9a, Supporting Information). It was observed that NPs induced a
   substantial amount of intracellular ROS production. The ROS generation
   was further quantified by FCM. Specifically, the ROS generated by NPs
   were 17‐fold and 1.7‐fold higher than that of PBS and MS+V,
   respectively (Figure [125]S9b,c, Supporting Information).
   Intracellularly, one of the major sources of ROS is the substrate end
   of the respiratory chain in the inner mitochondrial membrane. To
   explore the source of ROS induced by intervention, MitoSOX was used to
   specifically bind with mitochondrial ROS (red). The results
   demonstrated that NPs indeed induced a substantial release of
   mitochondrial ROS (Figure [126]4d). The above results suggested that
   NPs create an intracellular ROS storm in cells by simultaneously
   inhibiting the mTOR pathway and reducing GSH production. Meanwhile,
   mitochondrial ROS accounted for a proportion of ROS storm, which
   favored the formation of cellular pores during pyroptosis.^[ [127]^21
   ^]

   The ROS storm set the stage for pyroptosis. A distinct pyroptosis
   phenomenon was observed under the microscope (Figure [128]4e). The
   black arrows indicated the vesicular structures formed by the flow of
   contents out of the cell membrane after perforation when pyroptosis
   occurred. Following treatment with different compounds for 24 h,
   supernatant medium was collected to detect lactate dehydrogenase (LDH)
   and adenosine 5′‐triphosphate (ATP), indicating cell rupture. It was
   observed that NPs significantly elevated LDH and ATP levels in the cell
   supernatant (Figure [129]4f; Figure [130]S10a, Supporting Information).
   As cells undergo pyroptosis, they release interleukin‐1β (IL‐1β) and
   interleukin‐18 (IL‐18), recruiting additional inflammatory cells and
   amplifying the inflammatory response.^[ [131]^19 ^] IL‐18 and IL‐1β
   levels in supernatant media were detected by an Elisa kit.
   Specifically, the levels of IL‐18 and IL‐1β in the supernatant of NPs
   treated cells were 14.48 and 41.53 pg mL^−1, respectively, which were
   7.87 times and 1.60 times higher than the corresponding cytokines in
   untreated cells (Figure [132]4g; Figure [133]S10b, Supporting
   Information). Subsequently, the results of WB were performed to
   evaluate protein expression of the pyroptosis‐related pathway. The
   results revealed that NPs significantly upregulated the expression of
   NT‐GSDMD and cleaved caspase‐1 (Figure [134]4h). The expression of
   NT‐GSDMD generated by NPs was 1.97 times higher than that of the PBS,
   as indicated by the normalized statistics of the internal reference
   protein β‐Actin (Figure [135]4i; Figure [136]S11, Supporting
   Information). These findings suggested that NPs could induce a ROS
   storm and trigger pyroptosis within cancer cells.

2.6. Anti‐Tumor Effects of NPs in a BALB/C Nude Mouse Tumor Model

   The anti‐tumor efficacy of NPs was then examined in a BALB/C nude mouse
   model based on OCM‐1 cells (Figure [137] 5a). Fifteen days after tumor
   engraftment, animals were treated with saline, MS+V, cisplatin, or NPs.
   Considering that cisplatin is incorporated in the initial chemotherapy
   regimen for UVM, the administered dose of cisplatin was referenced.^[
   [138]^37 , [139]^38 ^] Over the course of the 16‐day treatment cycle,
   it was observed that the subcutaneous tumors in mice treated with
   saline were larger in size, whereas the tumors in all other groups that
   received the intervention were smaller (Figure [140]5b). There was no
   significant change in the weight of the animals during the course of
   treatment (Figure [141]5c). The measured tumor volumes of the mice were
   analyzed, and the results demonstrated that the tumors in the mice
   treated with cisplatin increased by 330.72 mm^3, while the tumors in
   the mice treated with NPs increased by 137.94 mm^3, with the former
   being 2.40 times larger than the latter. In contrast, the tumors in the
   mice after receiving a simple combination of the two individual
   compounds increased by 260.93 mm^3, which was 1.89 times larger than
   the tumors in the mice treated with NPs (Figure [142]5d). Statistical
   analysis revealed no significant difference between NPs and MS+V
   groups, but significant differences were observed between NPs versus
   Cisplatin (P < 0.05) and NPs versus Saline groups (P < 0.01).
   Concurrently, H&E staining was performed on major organs (heart, liver,
   spleen, lungs, and kidneys) of mice under a fluorescence microscope.
   The major organs of mice treated with NPs did not exhibit any
   noticeable morphological changes compared to the histomorphology of
   mice treated with saline, further attesting to the sufficient safety of
   NPs (Figure [143]S12, Supporting Information). Collectively, these
   findings suggested that NPs exhibit low toxicity and high efficiency.

Figure 5.

   Figure 5
   [144]Open in a new tab

   In vivo anti‐tumor effects of the NPs. a) Schematic illustration of in
   vivo treatment. b) Representative tumor image of different groups after
   16 days of treatment. c) Body weight changes and d) tumor growth
   inhibition curves of mice treated with saline, MS+V, cisplatin and NPs
   at 4 mg kg^−1 body weight. n = 5 mice per group. e) IFN‐γ level in
   serum of mice with various treatments. f) TNF‐α level in serum of mice
   with various treatments. g) Immunofluorescence imaging of the
   expression of Ki67 in OCM‐1 tumor after different treatment. Data are
   presented as mean ± SD. Statistical significance between all groups was
   calculated via two‐way ANOVA in (c and d), one‐way ANOVA in (e and f).
   ^* p < 0.05, ^** p < 0.01,^**** p < 0.0001, ns, not significant.

   The expression of interferon (IFN)‐γ and tumor necrosis factor (TNF)‐α
   in the serum of mice were evaluated by ELISA (Figure [145]5e,f).
   Following treatment with NPs, IFN‐γ in the serum of mice escalated to
   34.82 pg mL^−1, while TNF‐α increased to 71.45 pg mL^−1, both of which
   were significantly up‐regulated compared to the mice treated with
   saline. This suggested that NPs could instigate a systemic response. To
   observe the proliferation of tumor cells, the mouse tumor tissues were
   excised to stain them with Ki67 immunofluorescence. Ki‐67 protein is a
   nuclear protein associated with the cell cycle. Cells in the growth and
   proliferation phase will be labeled as Ki‐67 positive expression.^[
   [146]^37 ^] The proliferative capability of the tumor cells was likely
   inhibited post‐treatment with NPs, as evidenced by the significant
   reduction in Ki67‐positive cells in treated with NPs and the abundance
   of Ki67‐positive cells in treated with saline (Figure [147]5g). In situ
   Ki67 assays suggested that the fluorescence in the PBS‐treated mice was
   3.29‐fold stronger than that in the NPs‐treated mice (Figure [148]S13,
   Supporting Information). The findings indicated that NPs can
   successfully halt tumor growth and may potentially enhance the body's
   response against tumors.

2.7. NPs Elicited Anti‐Tumor Immune Responses and Long‐Term Immune Memory In
Vivo

   Previous findings suggested that NPs induced tumor cell pyroptosis,
   promoted the secretion of IL‐1β and IL‐18, and increased the serum
   levels of IFN‐γ and TNF‐α in mice. These cytokines can trigger positive
   immunological regulation when they stimulate the immune system.
   Meanwhile, the administration of NPs enhanced PD‐L1 expression in vitro
   (Figure [149]S14, Supporting Information). Previous studies have shown
   that the level of PD‐L1 expression in tumor tissues can serve as an
   indicator of the efficacy of PD‐L1/PD‐1 monoclonal antibody.^[ [150]^38
   ^] Researchers combined MS‐275 with α‐PD‐1 for the treatment of
   metastatic UVM in the reported PEMDAC phase 2 clinical trial, achieving
   long‐lasting therapeutic results.^[ [151]^39 ^] Consequently, a
   subcutaneous tumor model in the B16‐F10 cell line was established for
   group therapy to investigate whether NPs could induce an anti‐tumor
   immune response in vivo and enhance the therapeutic response rate of
   α‐PD‐1 (Figure [152] 6a). After monitoring tumor growth to ≈80 mm^3,
   five mice were randomly assigned per group to four different groups:
   Saline, MS‐275 (4 mg kg^−1 intravenously)+α‐PD‐1 (10 mg kg^−1
   intraperitoneally), NPs (4 mg kg^−1 intravenously), and NPs (4 mg kg^−1
   intravenously)+α‐PD‐1 (10 mg kg^−1 intraperitoneally). The body weights
   and tumor volumes of the mice were measured and recorded every other
   day starting from the day of treatment. The tumor volume recording
   curves in mice revealed that NPs+α‐PD‐1 had the most potent anti‐tumor
   efficacy (Figure [153]6b; Figure [154]S15, Supporting Information).
   Subsequently, tumors and tumor draining lymph nodes (TDLNs) from
   different groups were procured, and the relevant immunological
   indicators were correlated to further investigate the impact of NPs on
   the tumor immune microenvironment. Dendritic cells (DC cells) act as
   sentinels of the immune system and are able to rapidly ingest antigens
   and process them for presentation to immature T cells.^[ [155]^40 ^]
   Compared to saline, NPs+α‐PD‐1 dramatically increased the proportion of
   mature DC cells in TDLNs and tumors where the population of
   CD80^+CD86^+ cells was significantly increased from 13.83% to 41.83%,
   and from 10.24% to 47.97%, respectively (Figure [156]6c; Figures
   [157]S16 and [158]S17, Supporting Information). Cytotoxic T lymphocytes
   (CTL) are a crucial type of immunological surveillance cell.^[ [159]^41
   , [160]^42 ^] Higher percentages of CTL in patients tumors can help
   suppress tumor progression and even eliminate the tumor over long
   periods.^[ [161]^43 ^] Therefore, high percentages of CTL with
   cytotoxic function in tumor tissues is a positive prognostic indicator.
   The infiltration of CD8^+ T lymphocytes in the tumor tissues of groups
   were examined using FCM (Figure [162]6d; Figure [163]S18, Supporting
   Information). The percentage of CD8^+ T cells (CD3^+ CD8^+) in the
   tumor tissues of mice given NPs+α‐PD‐1 was ≈1.4 times more than that of
   the mice given MS‐275+α‐PD‐1. This suggested that coadministration of
   NPs and α‐PD‐1 may be more beneficial for cytotoxic T cells to
   function.

Figure 6.

   Figure 6
   [164]Open in a new tab

   NPs induced an anti‐tumor immune response and enhanced immune memory in
   combination with α‐PD‐1 in vivo. a) Schematic illustration of treatment
   schedules in vivo. b) Tumor growth inhibition of mice treated with
   various compounds. n = 5 mice per group. c) The percentages of DC cells
   (CD80^+CD86^+) within tumor in each group are presented as histograms.
   d) The percentages of CD8^+ T cells (CD3^+CD8^+) within tumor in each
   group are presented as histograms. e) The percentages of populations of
   NK cells (CD3^−NK1.1^+CD69^+KLRG1^+) within tumor in each group are
   presented as histograms. f) The percentages of populations of MDSCs
   (CD11b^+GR‐1^+) within tumor in each group are presented as histograms.
   g) Schematic illustration of treatment schedules in vivo. h)
   Representative FCM analysis images and i) populations of T[CM]
   (CD44^+CD62L^+) in spleen. j) The percentages of populations of T[EM]
   (CD44^+CD62L^−) in spleen. k) The percentages of populations of T[CM]
   (CD44^+CD62L^+) in TDLNs. l) Survival analysis of mice tumor up to day
   25. Data are presented as mean ± SD. Statistical significance between
   all groups was calculated via one‐way ANOVA in (c–f, i–k), and unpaired
   two‐sided t‐test in l). ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001,
   ^**** p < 0.0001. ns, not significant.

   Natural killer cells (NK cells), constitute the third major class of
   lymphocytes following T cells and B cells. They are the primary source
   of IFN‐γ release and possess the capability to recognize and attack
   viruses, tumor cells, and foreign bodies.^[ [165]^44 ^] NK cells were
   concerned in tumors since previous findings indicated an elevation in
   IFN‐γ expression in mice treated with NPs. Specifically, the percentage
   of NK cells following NPs+α‐PD‐1 therapy could reach 14.20%, which is
   nearly triple as high as the Saline. The proportions of NK cells in the
   NPs and MS‐275+α‐PD‐1 were 10.90% and 7.94%, respectively
   (Figure [166]6e; Figure [167]S19, Supporting Information). Such a trend
   suggested that the tumor may be more reliant on NPs to induce cellular
   pyroptosis, followed by NK cell aggregation. Myeloid‐derived suppressor
   cells (MDSCs) can inhibit the normal innate and adaptive immunological
   functions of the body's immune cells.^[ [168]^45 ^] Therefore, the
   proportion of MDSCs was assessed in tumor tissues (Figure [169]6f;
   Figure [170]S20, Supporting Information). Following treatment with
   NPs+α‐PD‐1, the proportion of MDSCs in tumor tissues was 23.49%, which
   was ≈47.65% of the Saline. These results suggested that the combination
   of NPs and α‐PD‐1 could effectively increase NK cells infiltration and
   mitigate the original immunosuppression in the tumor.

   In the immune responses against malignancies, T cells and other immune
   cells can become exhausted.^[ [171]^46 ^] The long‐term effects of NPs
   in conjunction with α‐PD‐1 were investigated in C57BL/6 mice model
   (Figure [172]6g). Two million B16‐F10 cells were injected into the
   right buttock subcutaneously of C57BL/6 mice. Four treatments were
   initiated once the tumor volume reached ≈80 mm^3. Two days after the
   completion of the fourth treatment, the tumor was surgically excised.
   After a two‐day rest period, an equal number of B16‐F10 cells were
   injected into the mice's left buttock subcutaneously, allowing for the
   tracking of the development of the second tumor. A portion of the mice
   in each group were sacrificed once their second tumor volume reached
   500 mm^3, and their spleens and TDLNs were removed to monitor the ratio
   of memory T cells (T[CM], CD3^+CD8^+CD44^+CD62L^+) and effector T cells
   (T[EM], CD3^+CD8^+CD44^+CD62L^−). Following NPs+α‐PD‐1, the percentage
   of T[CM] in the spleen increased to 18.47%, which is approximately
   three times higher than in those treated with saline (Figure [173]6h,i;
   Figure [174]S21, Supporting Information). In mice treated with
   NPs+α‐PD‐1, the percentage of T[EM] in spleen was 3.36%
   (Figure [175]6j). After receiving NPs+α‐PD‐1, the percentage of T[CM]
   in TDLNs was 19.20%, while it was only 10.86% in those treated with
   saline (Figure [176]6k; Figure [177]S22, Supporting Information). The
   development of the second tumor volume to 100 mm^3, which was recorded
   as state 1, was used to plot survival curves (Figure [178]6l). Mice
   treated with NPs+α‐PD‐1 did not achieve state 1 over the 25‐day
   observation period, and some did not even develop a second tumor. The
   findings described above suggested that NPs and α‐PD‐1 together help to
   improve mice's immunological memory and prevent tumor recurrence.

3. Conclusion

   In conclusion, we have developed an innovative ROS‐responsive
   nano‐delivery system (NPs), which co‐delivers a HDACi, MS‐275, and a
   metabolic inhibitor, V‐9302, for the treatment of UVM. This targeted
   approach exploits the enhanced permeability and etention effect,
   ensuring that our NPs accumulate at the tumor site. Upon arrival, these
   NPs respond to the high ROS levels within the tumor microenvironment,
   releasing their therapeutic payload. In mouse melanoma models, NPs
   effectively suppressed tumor progression and improved anti‐tumor
   immunity by recruiting DC cells, CD8^+ T cells, and NK cells,
   converting the “immune‐cold” tumor to “immune‐hot” tumor. Furthermore,
   when combined with α‐PD‐1, NPs effectively promote immunological memory
   and prevent tumor reoccurrence. By enhancing immunogenicity and
   immunotherapy responsiveness through combining HDACi with metabolic
   inhibitors, our study lays a groundwork for future clinical
   applications and provides a potential blueprint for combatting tumor
   resistance.

Conflict of Interest

   The authors declare no conflict of interest.

Supporting information

   Supporting Information
   [179]ADVS-11-2404375-s001.pdf^ (1.8MB, pdf)

Acknowledgements