Resiquimod-loaded MOF525 enables synergistic photodynamic therapy and immunotherapy for colorectal cancer.

Introduction
Metal-Organic Frameworks (MOFs) have recently emerged as a promising class of materials with extensive applications across biomedical fields [1, 2] including cancer therapy [3–5]. MOFs are crystalline particles with highly ordered structures composed of metal ions coordinated to organic ligands [6–9]. The specific rotation angle and geometry that result from the interaction between metal ions and organic ligands lead to a range of distinct topologies [9–11]. Additionally, external factors such as reaction temperature, reaction duration, and reagent ratio also comprehensively contributes to the final morphology, making MOFs a structurally versatile platform [10, 12]. High porosity is another advantageous feature of MOFs that highlights their potential as therapeutic platforms, as the empty pores enable effective loading of therapeutic molecules [13–15]. Among various organic ligands, a unique group of ligands exhibit photoactive properties that can be utilized to develop photo-responsive MOFs [16, 17]. Porphyrin-based ligands, in particular, have drawn significant attention in phototherapy due to their ability to produce reactive oxygen species (ROS) in response to laser irradiation [18]. This unique property allows porphyrin-based MOFs to directly function not only as a drug delivery platform, but also as an effective ROS-producing photosensitizer [19].
Owing to the advantages of porphyrinic MOFs as both a drug delivery and a photodynamic therapy (PDT) platform, their roles are particularly prominent in colorectal cancer (CRC) treatment. CRC is an example of a cancer in which PDT offers great advantages over conventional therapies since PDT can be endoscopically delivered to lesions [20, 21] without an open surgery or other invasive approaches. PDT allows selective tumor ablation, preserves adjacent healthy tissues, and enables repeated administration that underscores its significance against CRC [22, 23]. Unfortunately, despite localized tumor ablation, the efficacy of PDT is often limited by the immunosuppressive tumor microenvironment (TME) and tumor recurrence [20, 21]. To overcome this issue, numerous studies have attempted to combine PDT with other treatment modalities that can enhance its overall therapeutic efficacy [24, 25].
Resiquimod (R848) is a Toll-like receptor (TLR) agonist commonly used as an adjuvant in immunotherapy to activate innate immune cells and to promote adaptive immune responses [26–28]. By engaging TLR7/8 signaling, R848 induces dendritic cell (DC) maturation and bridges the innate and adaptive immunity against cancer [29–31]. Since R848 triggers immune responses that may overcome the immunosuppressive TME, its use in combination with PDT offers a promising strategy that can synergistically augment the therapeutic efficacy of traditional PDT [32–35]. Notably, tumor-associated antigens (TAAs) are critical components in adjuvant-mediated immunotherapy, as mature DCs utilize TAAs to orchestrate cancer-specific immune responses [36]. Consequently, immunotherapy is often suboptimal in poorly immunogenic tumors [37]. Considering that PDT can directly release immune-activating molecules such TAAs and danger-associated molecular patterns (DAMPs) via immunogenic cell death (ICD) [38], PDT may in turn also provide an immunologic environment favorable for the immune responses mediated by R848 [20, 39–42]. Combining PDT with R848 immunotherapy therefore offers several advantages compared to using each modality alone. Simultaneously employing both PDT and immunotherapy not only provides a more comprehensive approach to tumor eradication but also offers a promising strategy for synergistic cancer treatment.
Herein, this study reports the use of R848-loaded MOF525 (R848@MOF525) as a potential PDT platform against CRC. Among various porphyrin-based MOFs, MOF525 was selected for its high photo-reactivity and efficient ROS generation upon irradiation [3, 43–45]. MOF525 has also been reported to possess relatively higher surface area [46] suitable for greater drug loading, and greater singlet oxygen generation upon irradiation [47]. The high porosity of MOF525 enabled successful encapsulation of R848, which was sustainably released over a week in vitro. R848@MOF525 irradiated with 660 nm laser demonstrated high ROS generation and induction of ICD evaluated via ATP release in colon adenocarcinoma cells (CT26). Immunostimulatory properties assessed using bone marrow-derived dendritic cells (BMDCs) also revealed that R848@MOF525 and MOF525 both strongly promoted dendritic cell maturation in vitro, further highlighting their roles as immunostimulatory adjuvants. The therapeutic potential of R848@MOF525 was evaluated using a CT26 subcutaneous tumor model as illustrated in Scheme 1. When evaluated in vivo, PDT using R848@MOF525 was able to yield faster tumor regression than with MOF525. Furthermore, tumor rechallenge tests also showed no evident tumor growth up to 3 weeks, indicating that PDT using R848@MOF525 could induce a durable systemic antitumor immunity. Notably, under an extended treatment interval, PDT mediated by MOF525 failed to suppress tumor growth, whereas those using R848@MOF525 successfully maintained antitumor efficacy. This study collectively demonstrates that R848@MOF525 can significantly inhibit tumor growth and induce robust immune responses, highlighting its potential as a multifunctional therapeutic platform incorporating both PDT and immunotherapy.

Scheme 1.

Result and discussion

Characterization of MOF525 and R848@MOF525
In this study, MOF525, a porphyrin-based MOF, was developed as a photosensitizer and carrier of the immune adjuvant R848. MOF525, composed of zirconium (Zr) metal units and Tetrakis(4-carboxyphenyl)porphyrin (TCPP) organic ligands, was synthesized using a solvothermal method. Following MOF525 synthesis, R848 was loaded into the pores of MOF525 via immersion, resulting in the drug loaded R848@MOF525 (Fig. 1A). Scanning electron microscopy (SEM) images exhibited that fabricated MOF525 were cuboctahedral-shaped, as reported previously (Fig. 1B) [48]. SEM images further confirmed that this shape was not altered after R848 loading, although dynamic light scattering measurements showed a subtle increase in hydrodynamic diameter from 320.9 nm to 337 nm after loading R848 (Fig. 1B). The zeta potential of MOF525 or R848@MOF525 were each − 12.3 mV and − 9.5 mV, respectively (Fig. 1C). This slight increase in surface charge may be attributed to the loading of R848, as similarly observed in another study [49]. To determine the colloidal stability of MOF525 in a physiological condition, MOF525 was also dispersed in phosphate buffered saline (PBS) containing 5% (v/v) serum for five days. Both the hydrodynamic diameter and polydispersity index measurements remained stable up to three days, although these values slightly increased on Day 5 (Figure S1). These results demonstrate that MOF525 remains colloidally stable in physiologically relevant conditions without abrupt changes indicative of aggregation.
The fabricated MOFs showed a distinct X-ray diffraction (XRD) pattern unique to the structure and crystallinity of MOF525 reported previously (Fig. 1D) [50]. These XRD patterns were similarly maintained with R848@MOF525, indicating that R848 loading did not significantly alter the unique crystallinity of MOF525 (Fig. 1D). Next, N2 adsorption-desorption isotherms of MOF525 and R848@MOF525 were evaluated. The isotherm graphs of MOF525 showed a dominant presence of micropores whereas these were reduced after loading R848 (Fig. 1E). The Brunauer-Emmett-Teller (BET) specific surface area and Barrett-Joyner-Halenda (BJH) specific pore volume of MOF525 was each 1356 m2 g−1and 0.8044 cm³ g−1, respectively (Table S1). Incorporating R848 caused these values to each decrease to 80.18 m2 g−1 and 0.5597 cm2 g−1, confirming successful loading of R848 within MOF525. Although the average pore diameter of R848@MOF525 increased from 1.682 nm to 66.93 nm, such change does not necessarily indicate a structural expansion, but is rather likely due to the shift in pore distribution as shown in Fig. 1E.
Fourier transform infrared spectroscopy (FTIR) analysis of MOF525 showed distinct peaks at 1600 cm− 1 (asymmetric stretching of carboxylate ligand), 1410 cm− 1 (symmetric stretching of carboxylate ligand), and 660 cm− 1 (stretching mode of the metal-ligand bonds) indicative of Zr-porphyrin MOFs [51] (Fig. 1F). These characteristic peaks were also observed in R848@MOF525, with additional peaks emerging at 3112 cm− 1 (stretching vibration of = C-H bonds) and 2975 cm− 1 (stretching of -C-H bonds), which were attributable to R848. To further examine whether any chemical interactions were present between R848 and MOF525, X-ray photoelectron spectroscopy (XPS) analysis was performed in the Zr 3d region before and after R848 loading (Fig. 1G). MOF525 alone displayed characteristic peaks corresponding to the Zr–O coordination [52]. Interestingly, a new signal emerged following R848 loading, which was attributable to O–Zr–N bonding. These results suggest that the Lewis basic nitrogen atoms of R848 may have formed coordination bonds with unsaturated Lewis acidic Zr4+ centers within the MOF525 framework [53]. Considering that such interactions are reversible under aqueous or physiological conditions, the presence of coordination bonds between R848 and surface Zr sites presents an interaction suitable for in vivo drug release [54].

In vitro evaluation of antitumor efficacy
The release profile of R848 from R848@MOF525 was analyzed using a dialysis-based method by suspending the MOFs in pH 7.4 PBS under mild shaking. The drug loading capacity of R848@MOF525 was determined as approximately 44 µg R848 per 1 mg MOF via reverse phase high-performance liquid chromatography (HPLC). The drug release profile was observed over 12 days, revealing that the majority of the payload was sustainably released until Day 7 (Fig. 2A). Cell cytotoxicity of MOF525 and R848@MOF525 was assessed in vitro using L929 mouse fibroblast cells as recommended by ISO 10993-5 [55]. According to the ISO 10993-5:2009 guideline, MOF concentrations exhibiting cell viability greater than 70% compared to negative control was considered as non-toxic [56]. Both MOF525 and R848@MOF525 showed a dose-dependent cytotoxicity toward L929 (Fig. 2B) but were not toxic below 50 µg/mL. Subsequently, MOF concentration of 50 µg/mL was chosen for further in vitro and in vivo experiments.
To examine whether the porphyrin ligands of MOF525 or R848@MOF525 could generate ROS in response to laser irradiation, ROS production was quantified in vitro using the singlet oxygen sensor green (SOSG) assay. When MOFs were irradiated using a 660 nm laser (500 mW) up to 10 min, SOSG intensities of both MOF525 and R848@MOF525 showed a gradual increase, indicating effective ROS generation (Fig. 2C). In terms of signal intensity, the SOSG signal produced by R848@MOF525 was weaker than those of MOF525. However, this is likely due to the absolute difference in the amount of porphyrin ligands present in each group. Due to the increased mass following encapsulation of R848, identical weights of MOF525 and R848@MOF525 could have resulted in differing amount of porphyrin ligands. Nevertheless, R848@MOF525 also exhibited notable ROS production following irradiation, confirming its potential as a PDT platform.
Next, the anticancer efficacy of PDT using R848@MOF525 was evaluated against CT26 colon adenocarcinoma cells. The viability of CT26 cells treated with either MOF525 + Laser or R848@MOF525 + Laser were significantly reduced compared to non-treated control (Fig. 2D). Although cells treated with laser alone also exhibited a subtle decrease in cell viability, PDT mediated by MOF525 showed significantly amplified anticancer effects, presumably due to ROS generation. These findings suggest that MOF525 effectively functions as cytotoxic photosensitizers upon laser irradiation. To further evaluate whether PDT mediated by MOF525 could induce sufficient immunogenic cell death in tumor cells, extracellular adenosine triphosphate (ATP), a representative damage-associated molecular patterns (DAMP) marker, was quantified following PDT. Among treatment groups, MOF525 treatment with laser irradiation induced significantly greater ATP release (6.92 pmol/µL) compared to MOF525 treatment alone (6.11 pmol/µL) or control (5.96 pmol/µL) (Fig. 2E). Despite the slight reduction in CT26 viability (Fig. 2D), laser treatment alone induced ATP releases comparable to those of untreated control. To further assess potential thermal effects associated with laser irradiation, the photothermal characteristics of MOF525 were evaluated. Upon irradiation with a 660 nm laser (500mW) for 5 min, MOF525 exhibited a temperature increase of approximately 3.4 °C (Figure S2). Although this increase was greater than those observed for PBS, both the endpoint (28.9 °C) and change (3.4 °C) in temperature was substantially lower than those typically required for photothermal therapy or tumor ablation [57, 58]. While potential immunogenic contributions of mild hyperthermia cannot be completely excluded [59], its specific contribution to the overall therapeutic effect, particularly in relation to PDT, remains to be further elucidated.
In addition to the photodynamic properties of R848@MOF525, its potential as a synergistic immunotherapy platform was also evaluated in BMDCs using DC maturation markers as an indication of immune activation. To assess DC maturation induced by MOF525 and R848@MOF525, different concentrations of MOFs were co-incubated with BMDCs and analyzed for changes in surface CD86 and MHCII expression using flow cytometry (Fig. 3A). Here, CD86 and MHCII were selected as model indicators of immune stimulation as mature DCs present up-regulated expression of surface costimulatory molecules such as CD86 and MHCII compared to immature DCs [60–62]. The flow cytometry gating used to specify BMDCs is depicted in Figure S3. The expression of CD86 and MHCII induced by R848@MOF525 treatment were both significantly greater than those of the non-treated group, successfully demonstrating a dose-dependent immunostimulatory profile of MOFs (Fig. 3B). Although BMDCs treated with R848@MOF at concentrations below 10 µg/mL also exhibited a subtle increase in CD86 and MHCII expression, the changes were less significant than BMDCs treated with higher MOF concentrations (Figure S4).
Interestingly, the data also revealed that even MOF525 treatment alone was able to induce a notable change in both DC maturation markers without the adjuvant R848, suggesting a possible synergistic role of MOF525 as a self-standing immunostimulatory adjuvant. The specific mechanism behind the immunostimulatory properties of MOF525 have not yet been addressed, but this effect may have arisen from the adjuvant-like property of zirconium. Zr4+ ion is known to stimulate innate immune responses and promote DC maturation [63] while other studies have reported the potential role of zirconium salts as adjuvant-like materials via complement activation [64] or IgM antibody production [65]. As zirconium oxides have also been investigated as mild adjuvants [66], various active forms of zirconium released from MOF525 may have induced the observed changes in DC maturation markers. The expression of CD86 in BMDCs treated with either 50 µg/mL MOF525 (56.03 ± 1.58%) or R848@MOF525 (58.04 ± 7.33%) was both elevated by more than 3.6 folds compared to non-treated control (15.4 ± 1.10%). Similarly, the expression of MHCII in BMDCs treated with either 50 µg/mL MOF525 (57.08 ± 1.49%) or R848@MOF525 (59.66 ± 7.25%) was also over 2 folds greater than those of non-treated control (27.57 ± 2.84%) (Fig. 3C). These immunostimulatory effects were also notably greater when compared to BMDCs treated with free R848 (CD86: 21.63 ± 2.11%; MHCII: 20.76 ± 1.42%) or lipopolysaccharide (CD86: 37.93 ± 2.53%; MHCII: 28.89 ± 3.94%).
Another notable observation was that the expression of DC maturation markers in cells treated with either MOF525 or R848@MOF525 was comparable, which did not clearly delineate the additional contribution of the loaded drug. Although DCs treated with R848@MOF525 exhibited higher expression levels of CD86 and MHCII compared to those treated with MOF525 alone, these differences were not statistically significant. This observed phenomenon may be attributed to the release profile of R848 (Fig. 2A). Considering the 24 h treatment duration used in this in vitro study, the amount of R848 released during this period can be estimated to be approximately 0.66 µg/mL. While 2 µg/mL of free R848 was included as a positive control resembling the total R848 content within R848@MOF525, free R848 treatment also did not induce significant changes in expression of CD86 or MHCII (Fig. 3C). Despite the well-established role of R848 as a TLR7/8 agonist and its efficacy in cancer immunotherapy [67], several studies have also similarly reported limited changes in CD86 expression following R848 treatment in myeloid DCs and BMDCs under in vitro conditions [68, 69]. Collectively, the observed increase in CD86 and MHCII expression following treatment with either MOF525 or R848@MOF525 suggest that these MOFs possess inherent immunostimulatory properties. At the same time, these findings denotes the possibility of in vitro assays not fully depicting the immunotherapeutic contribution of R848, necessitating further in vivo evaluations to comprehensively assess its contributory role in this study.

In vivo evaluation of antitumor efficacy
The antitumor efficacy of R848@MOF525 was further verified in vivo using a subcutaneous colon carcinoma model. Mice were subcutaneously inoculated with CT26 colon adenocarcinoma cells, then subjected to various treatments (PBS; R848; MOF525; R848@MOF525; MOF525 + Laser; R8482MOF525 + Laser) once tumors reached an average volume of approximately 100 mm3 (Fig. 4A). Treatment groups without laser irradiation were intratumorally injected with appropriate solutions while PDT groups received additional irradiation using a 660 nm laser (500 mW) for a total duration of 10 min. To induce PDT while avoiding localized heating, laser irradiation was divided into two 5 min treatments.
As depicted in Fig. 4B, treatment of either MOF525 or R848 alone was not able to induce significant regressions in tumor size. Interestingly, despite comparable immunostimulant properties between MOF525 and R848@MOF525 in vitro (Fig. 3C), the presence of R848 was able to elicit significantly greater tumor regressions in vivo, presumably owing to the sustained release of R848 (Fig. 4B-D). In groups that received PDT, both MOF525 + Laser and R848@MOF525 + Laser treatments demonstrated significant anti-tumor effects, inducing complete tumor regression by Day 40 (Fig. 4C). Such results highlight the significance of porphyrin-based MOF525 as an effective platform for PDT against tumors. However, in terms of tumor regression rates, R848@MOF525 combined with PDT was able to induce a more rapid regression, as all treated mice exhibited complete tumor regression even before Day 25 (Fig. 4B). In PDT groups that were treated with MOF525, complete tumor regression across all mice was observed on Day 35 without any significant changes in body weight (Fig. 4E). To investigate whether treatment of R848@MOF525 combined with PDT may also induce long-term immunity, mice that exhibited complete tumor regression were re-challenged with CT26 cells on the opposite flank on Day 60. Considering the growth rate of CT26 tumors observed after primary inoculation (Fig. 4B) and monitoring periods previously reported in similar studies [70–72], tumor growth was monitored for 3 weeks following re-challenge. Even after 3 weeks, no signs of tumor growth or nodule formation was observed in the re-challenged site, indicating that the initial PDT treatment not only ablates the primary tumor, but may also establish a robust systemic immune response (Figure S5).
To further validate the synergistic contribution of R848, an additional in vivo study was performed employing an extended treatment interval (Figure S6). The treatment interval was extended from the original 3 to 5 days to evaluate whether PDT using MOF525 or R848@MOF525 could also induce sufficient tumor regression with less frequent treatments. Unlike the results from treatments using a 3-day interval (Fig. 4B), PDT treatments using MOF525 was not able to induce tumor regression under the 5-day treatment interval, suggesting that the current treatment interval was insufficient to overcome the immunosuppressive TME (Fig. 4F, S7). On the contrary, mice treated with PDT using R848@MOF525 showed significantly greater tumor regression, particularly after the third treatment. These findings across different treatment schedules collectively suggest that although the antitumor effects of PDT cannot be overlooked, delivery of R848 via R848@MOF525 can significantly accelerate tumor regression and allow suboptimal PDT treatments to overcome the immunosuppressive TME. Such results suggests that PDT using R848@MOF525 may offer therapeutic advantages over MOF525 in tumors with higher proliferation rates or stronger immunosuppressive environments. Under these circumstances, R848@MOF525 presents a promising platform that can synergistically amplify the therapeutic efficacy of porphyrin-based MOFs used in PDT.

Histopathologic evaluation of antitumor efficacy
As PDT-treated mice exhibited complete tumor regression (Fig. 4B), histopathological analyses were not conducted in these groups. Instead, tumors remaining in mice treated with either PBS, R848, MOF525 or R848@MOF525 were collected and fixed in 10% neutralized buffered formalin for histopathological evaluations. Tumor samples were subjected to a general paraffin embedding procedure, followed by hematoxylin and eosin (H&E) staining. To assess the overall size and cellular composition of the tumor, whole mount image was collected (Fig. 5A). Remaining tumors, in terms of area, were found to be larger in the order of control, MOF525, R848, and R848@MOF525 groups (Fig. 5A, B). Necrotic lesions were present in the central area of all treatment groups, although tumors treated with MOF525 had a wider necrotic lesion compared to those treated with R848. The size of tumors treated with R848@MOF525 was smaller overall, with smaller foci of necrosis found throughout the tumors in addition to the central necrotic lesions.
When observed at a higher magnification, necrotic lesions composed of amorphous necrotic material were observed in all groups, with a large number of inflammatory cells infiltrating around these necrotic lesions (Fig. 5C). Furthermore, while the inflammatory cell infiltration was not prominent in the proliferating region of tumor in the control group, tumors from other treatment groups (R848, MOF525, and R848@MOF525) had varying degrees of inflammatory cell infiltration. Tumors treated with R848@MOF525, in particular, showed the formation of small necrotic foci with scattered inflammatory cell infiltration. Compared to tumors treated with MOF525, more mononuclear cell infiltration was observed in the middle and periphery lesion of those treated with R848. Quantitative analysis of necrotic lesions further showed that the relative necrotic lesion (necrotic lesion: tumor area) of tumors treated with either MOF525 or R848@MOF525 was both significantly greater than those of the control group (Fig. 5D). No statistical significance was observed between tumors treated with either MOF525 or R848. Among experimental groups, tumors treated with R848@MOF525 had the largest relative necrotic lesions, although this may be due to their relatively smaller tumor size as well as the inclusion of both central and sporadic necrotic lesions, collectively leading to an increase in relative necrotic foci.
Furthermore, TUNEL staining was performed to evaluate the extent of apoptotic cell death in the tumor. Central necrosis was observed in all groups, characterized by the presence of multiple TUNEL positive cells (Orange to brownish color) (Fig. 5E). In particular, multiple TUNEL positive apoptotic cell death foci in the area between the central necrosis and the periphery was observed in tumors treated with either R848 or R848@MOF525. Quantitative analysis also showed that the TUNEL positive areas of tumors treated with R848@MOF525 were significantly higher than all other treatment groups (Fig. 5F). Taken together, these histopathological analyses indicate that although R848@MOF525 treatment did not yield complete tumor regression, R848 released from R848@MOF525 was still able to induce stronger necrosis, apoptotic cell death, and infiltration of inflammatory cells in tumors compared to those treated with bare MOF525.

In vivo biosafety evaluation
The systemic toxicity of R848@MOF525-mediated PDT therapy was evaluated by conducting a toxico-pathological analysis across major vital organs (e.g. Heart, Lung, Liver, Kidney, Spleen) harvested 3 weeks after the tumor rechallenge test (Figure S8). Following PDT treatment using either MOF525 or R848@MOF525, a mild proliferation of pulmonary interstitial tissue, a mild increase in the number of reactive hepatocytes in the liver, and a mild increase in the size of splenic nodules were observed in some individual specimens. However, these changes were likely mild side effects of immunologic responses to the repetitive tumor inoculations and subsequent tumor cell therapies. No significant damage to vital organs were observed in any of the evaluated mice. These results confirm that PDT utilizing R848@MOF525 exerts no systemic toxicity.

Conclusion
In this study, a porphyrin-based metal–organic framework platform loaded with R848, namely R848@MOF525, was fabricated to integrate PDT with immunotherapy for colorectal cancer treatment. MOF525, the parent of R848@MOF525, provided a highly porous structure capable of efficiently encapsulating the Toll-like receptor agonist R848 while simultaneously functioning as a photosensitizer owing to the presence of porphyrin ligands. Structural and spectroscopic analyses confirmed that R848 loading did not alter the crystallinity nor the morphology of MOF525, and drug release studies demonstrated sustained release of R848 up to seven days. In vitro experiments showed that R848@MOF525 effectively generated ROS under 660 nm laser irradiation, leading to PDT-like cytotoxic effects in CT26 colon adenocarcinoma cells and inducing immunogenic cell death, as evidenced by increased extracellular ATP release. Furthermore, MOF525 and R848@MOF525 were both able to promote notable DC maturation as indicated by elevated expression of CD86 and MHCII in BMDCs, suggesting that MOF525 itself could also stimulate immune activation in addition to mediating PDT.
The therapeutic potential of R848@MOF525 was further validated in vivo using a CT26 tumor-bearing mouse model. While MOF525-mediated PDT alone also effectively suppressed tumor growth, the incorporation of R848 significantly accelerated tumor regression and demonstrated persistent antitumor effects even under extended treatment intervals. Notably, mice that achieved complete tumor regression following R848@MOF525-mediated PDT showed no tumor growth in tumor rechallenge tests, indicating that R848@MOF525-mediated PDT could induce a durable systemic antitumor immunity. Histopathological analyses further supported these findings by revealing enhanced necrosis, apoptosis, and inflammatory cell infiltration in tumors treated with R848@MOF525. While the present study demonstrates promising therapeutic efficacy and suggests the induction of antitumor responses, further in vivo immunological characterization could enhance mechanistic understanding. In particular, comprehensive temporal evaluation of tumor-infiltrating lymphocytes may delineate the immunological mechanisms underlying therapeutic efficacy, as well as to further elucidate the immunological impact of prolonged R848 release. Such investigations represent a valuable direction for future studies. Nevertheless, these results collectively demonstrate that R848@MOF525 represents a promising photoimmunotherapeutic platform capable of synergistically enhancing PDT efficacy and overcoming immunosuppressive tumor environments via immunotherapy.

Materials and methods

Materials
Benzoic acid (assay ≥ 99.0%), TCPP (assay ≥ 97.0%) and R848 (assay ≥ 98.0%) were purchased from TCI (Japan). Dimethylformamide (DMF, assay ≥ 99.5%) was purchased from DAEJUNG (Korea). Zirconyl chloride octahydrate (ZrOCl2 8H2O) and LPS from E. coli (O55:B5) were purchased from Sigma-Aldrich (USA). Dulbecco’s Modified Eagle Medium (DMEM), RPMI 1640 Medium (RPMI) and Phosphate buffered saline (pH 7.4) were purchased from Thermo Fisher Scientific (USA).

Preparation of MOFs
MOF525 was synthesized following a previously reported protocol with slight modifications [73]. Briefly, TCPP (235 mg, 0.297 mmol) and benzoic acid (6750 mg, 55 mmol) were dissolved in 40 mL DMF and sonicated for 30 min. ZrOCl2 8H2O (525 mg, 1.62 mmol) was added to the TCPP solution and subsequently dissolved for 30 min. The resulting solution was heated in an oven at 100 ℃ for 24 h. The resultant products were collected by centrifugation and washed thrice with DMF and ethanol to remove excess reactants. The resulting particles were dried at 80 ℃ for 6 h to achieve dry powders. Subsequently, MOF525 was dried under vacuum at 150 ℃ to clear the pores and remove residual solvent. To prepare R848@MOF525, 20 mg R848 was first dissolved in 2 mL ethanol and sonicated, to which 100 mg of activated MOF525 were added. The suspension was kept at RT for 24 h. Then, R848-loaded MOF525 was filtered using a 200 nm PVDF membrane filter, washed with ethanol thrice, and dried at 70 ℃ to achieve R848@MOF525.

Characterizations
Images of the morphology and size of MOF525 and R848@MOF525 were obtained using a SEM (Dimension ICON, Bruker, USA). The hydrodynamic size and zeta potential were measured using a Zetasizer (Zetasizer pro, Malvern Panalytical, UK) while the serum stability of MOFs were evaluated after incubation in PBS containing 5% FBS. PXRD patterns were recorded using an X-ray diffractometer (SmartLab, Rigaku, Japan) at 3 kW with CuKα radiation. A surface area and porosity analyzer (Autosorb IQ, Quanmtachrome, USA) at -196℃ was used to measure the surface area, pore size, and pore volume of MOFs. FTIR spectra of MOF525 and R848@MOF525 were analyzed using a FTIR spectrophotometer (Cary670 (Main bench) + Cary620 (microsocope)), Agilent, USA) over the range from 4000 to 500 cm− 1 at room temperature. The XPS spectra of MOF525 and R848@MOF525 were obtained using an X-Ray Photoelectron Spectroscopy System (PHI Quantera-Ⅱ, Ulvac-PHI, Japan).

R848 release profile in vitro
The release profile of R848 form R848@MOF525 was evaluated using a dialysis-based method. 5 mg of R848@MOF525 was dispersed in 5 mL of PBS and placed in the inner compartment of the SnakeSkin® dialysis tubing (MWCO: 3500, Thermo Fisher Scienticfic, USA). The tube was dialyzed against 20 mL of the same buffer at 37 ℃ under mild shaking. At each predetermined time points, the outer solution was recovered and replaced with equivalent volume of fresh PBS. The obtained samples were centrifuged at 17,000 rpm for 10 min and the supernatant was used for subsequent analyses. R848 content of the supernatant samples was quantified using reverse phase high-performance liquid chromatography (1260 Infinity Ⅱ, Agilent, USA). The mobile phase consisted of 60% 20 mM PBS containing phosphoric acid (pH 2.5) and 40% acetonitrile, with a flow rate of 1.0 mL/min. R848 was detected at 249 nm, with the injection volume of 10 µL. Maximum R848 loading was separately quantified by dispersing R848@MOF525 in PBS at a concentration of 0.5 mg/mL and sonicating for 3 h to fully extract the encapsulated drug.

Cell culture
L929 mouse fibroblast cells were cultured with DMEM containing 10% (v/v) fetal bovine serum (FBS) and 1% (v/v) Penicillin-Streptomycin (Gibco, Life Technologies, USA) in a cell incubator (5% CO2, 37 °C). CT26 mouse colon cancer cells (ATCC, USA) were cultured with RPMI containing 10% (v/v) FBS and 1% (v/v) Penicillin-Streptomycin in a cell incubator (5% CO2, 37 °C).

Cytotoxicity evaluation in vitro
Cytotoxicity of MOF525 and R848@MOF525 was evaluated in vitro using a water-soluble tetrazolium salt-based cell viability assay (EZ-Cytox; Daeil Lab Service, Korea) against L929 cells. L929 cells were seeded at a density of 0.9 × 104 cells/well in a 96-well cell culture plate and incubated for 24 h to ensure complete cell attachment. Next, the medium was completely removed and replaced with fresh medium containing either MOF525 or R848@MOF525 at varying concentrations (50, 25, 10 and 1 µg/mL) and incubated for another 24 h. Cells were washed thrice with PBS, then incubated for 1 h in 10% (v/v) EZ-Cytox solution diluted in medium. The absorbance at 450 nm was measured with a reference wavelength at 600 nm using a microplate reader (Synergy H1, BioTek, USA). Cell viability was calculated by using the following equation:

Assessment of ROS generated by laser irradiation in vitro
ROS generation was evaluated in vitro using Singlet Oxygen Sensor Green (SOSG; Thermo Fisher Scientific, USA). Briefly, 100 µL of either MOF525 or R848@MOF525 suspended in medium (25 µg/mL) was added in black 96-well plates. Subsequently, 100 µL of 10 µM SOSG dissolved in medium was added to the wells and mixed via pipetting. The resulting mixture was irradiated with a 660 nm laser (500 mW, 10 min). The SOSG intensity at each timepoint was measured using a microplate reader at 500 nm with a reference wavelength at 530 nm.

Assessment of PDT-induced cytotoxicity in vitro
PDT-induced cytotoxicity of MOF525 and R848@MOF525 was assessed using CT26 cells. Cells were seeded at a density of 1.0 × 104 cells/well in a 96-well cell culture plate. After 24 h, the medium was removed and replaced with medium with either MOF525 (50 µg/mL) or R848@MOF525 (50 µg/mL). Cells were irradiated with a 660 nm laser (500 mW) for 5 min, rested for 1 min, then re-irradiated for 5 min (total 10 min irradiation) After 24 h, cell viability was evaluated using EZ-Cytox, as described above.

Assessment of ATP release in vitro
ATP released after MOF525 treatment and laser irradiation was evaluated using a commercial PicoSensTM ATP Assay Kit (BIOMAX, Korea). CT26 cells were seeded at a density of 1 × 106 cells/well in a 24-well cell culture plate and incubated for 24 h. The cells were treated with MOF525 (50 µg/mL) dispersed in medium, then irradiated with a 660 nm laser (500 mW) for 10 min (5 min irradiation, 1 min rest, 5 min irradiation). After 24 h, ATP measurement was performed according to the instructions provided by the manufacturer.

Assessment of photothermal properties of MOF525
Temperatures changes of MOF525 following laser irradiation was evaluated using a thermal imaging camera (MobIR Air, GUIDE, China). Either MOF525 (50 µg/mL) or PBS in 1.5 mL centrifuge tubes were irradiated with a 660 nm laser (500 mW) for 5 min. Solution temperatures were measured and recorded at predetermined time points (0, 1, 2, 3, 4, 5 min).

Culture of Bone marrow-derived dendritic cells (BMDCs)
Mouse bone marrow progenitor cells were isolated and differentiated to dendritic cells following a previous study with slight modifications [74]. In brief, bone marrow progenitor cells were flushed from the femurs and tibias of BALB/c mice using a syringe. After removing red blood cells with ACK Lysing Buffer (Gibco, USA), purified bone marrow cells were cultured in 100 mm petri dishes using RPMI with 20 ng/mL of granulocyte-macrophage colony-stimulating factor (GM-CSF, Biolegend, USA). On day 6, non-adherent and loosely adherent cell clusters were considered as proliferating immature DCs and were harvested using 2 mM ethylene-diamine-tetraacetic acid (EDTA) and used accordingly.

In vitro stimulation of BMDCs
BMDCs were seeded in non-treated 96-well cell culture plates at 1.0 × 105 cells/well and treated with either MOF525 (50, 25, and 10 µg/mL) or R848@MOF525 (50, 25, and 10 µg/mL). LPS (100 ng/mL) or R848 (2 µg/mL) was used as a positive control for BMDC activation. After 24 h, expression of BMDC activation markers were evaluated using a flow cytometer (Cytoflex, Beckman Coulter, USA). The following antibodies were used in this study: Zombie Violet™ Viability Dye, CD11c (PerCP-Cy5.5), CD86 (PE), and MHCII (FITC) (Biolegend, USA).

Animals
All animal experiments were conducted in accordance with the protocols approved by the Institutional Animal Care and Use Committee (IACUC) of Sungkyunkwan University College of Medicine (IACUC No. SKKU IACUC 2024-02-40-1). BALB/c mice (female, 7 weeks old) were purchased from Orient Bio (Seongnam, Korea). All animals were housed under a 12:12 light–dark cycle and acclimatized for 7 days before any experiments.

Evaluation of in vivo antitumor efficacy
The antitumor efficacy of R848@MOF525 treatment was evaluated using a subcutaneous CT26 tumor model resembling CRC. BALB/c mice were subcutaneously inoculated with 5 × 105 CT26 cells suspended in 100 µL of PBS in the right flank. Tumor sizes were measured using a digital caliper and volumes were calculated using the following formula: []. Once tumor volumes reached approximately 100 mm3, mice were randomly divided into one of the following treatment groups: PBS (100 µL), R848 (100 µL, 0.2 mg/mL in PBS), MOF525 (100 µL, 1 mg/mL in PBS), R848@MOF525 (100 µL, 1 mg/mL in PBS), MOF525 (100 µL, 1 mg/mL in PBS) + Laser, or R848@MOF525 (100 µL, 1 mg/mL in PBS) + Laser. Groups including laser treatment were irradiated with a 660 nm laser (500 mW) for a total of 10 min (5 min irradiation, 1 min rest, 5 min irradiation). Treatments were provided as either 3-day intervals (Day 16, Day 19, Day 22, Day 25 post-inoculation) or 5-day intervals (Day 13, Day 18, Day 23, Day 28 post-inoculation). Tumor sizes and body weights were measured after each treatment. Tumor tissues from PBS, R848, MOF525 and R848@MOF525 treatment groups were extracted 40 days post-inoculation for histopathological evaluations. Since MOF525 + Laser and R848@MOF525 + Laser groups exhibited complete tumor regression, a tumor re-challenge test was conducted instead at Day 60 post-inoculation. For tumor re-challenge tests, cell volumes identical to those of the primary inoculation was injected to the opposite flank and tumor growth was monitored for 3 additional weeks. Vital organs from these mice were then extracted for biosafety evaluations. All procedures, including tumor inoculation, were performed under isoflurane inhalation anesthesia.

Histopathological analysis
Both tumor samples and vital organs were collected and stored in 10% neutralized buffered formalin prior to histological analyses. Samples were embedded in paraffin blocks and sectioned (4 μm thickness) for either Hematoxylin and Eosin (H&E) staining or TUNEL staining. Images of H&E stained samples were obtained using an inverted light microscope (Amscope, USA), and quantitative analysis of tumor size and necrotic lesions was performed using the ImageJ software. TUNEL staining was performed according to the manufacturer’s guidelines (Abcam, USA). In brief, slides were deparaffinized/hydrated, removed of endogenous peroxidase using H2O2 in methanol, then treated for pressurized heat antigen retrieval in citrate buffer prior to staining. TUNEL staining images were obtained using an inverted light microscopy, evaluated using the ImageJ software.

Statistical analysis
Statistical analyses were performed using the GraphPad Prism 8 software (GraphPad Software, USA). Data are presented as mean ± standard deviation unless stated otherwise. The precise number of replicates or samples for each data is specified in the figure legend. Statistical significance between groups was assessed using one-way ANOVA followed by Tukey’s post hoc multiple comparison test unless stated otherwise. P-values of *p < 0.0332, **p < 0.0021, ***p < 0.0002, and ****p < 0.0001 were considered statistically significant.

Supplementary Information
Below is the link to the electronic supplementary material.