Chlorin e6-associated photodynamic therapy enhances abscopal antitumor effects via inhibition of PD-1/PD-L1 immune checkpoint

We hypothesized that photodynamic therapy (PDT) with Chlorin e6 (Ce6) enhances antitumor abscopal effects via inhibition of the programmed cell death-1/programmed death-ligand 1 (PD-1/PD-L1) immune checkpoint. By using syngeneic melanoma and pancreatic tumor mouse models, we studied the Ce6-PDT-induced immune responses in local and distant tumor microenvironments. In addition, the Ce6-PDT's target in the PD-1/PD-L1 interaction was analyzed in MC38-hPD-L1 colon cancer and PD-1 expressing Jurkat T cell coculture. The tumors in the irradiated and non-irradiated sites in the abscopal effective (Abseff) group of both mouse models were regressed, proving the abscopal effect. The immunogenic effect in the Abseff group was associated with an expansion of T cell and other immune cells infiltration without changes in the CD39+ population in either the right or left tumors compared to control group. Furthermore, the abscopal ineffective (Absineff) group demonstrated lesser increase of T cells, decreased immune cell infiltration, and increased CD39-expressing Treg cells without suppression of tumor growth. In the coculture with PD-1-expressing Jurkat T cell, Ce6-PDT efficiently suppressed the PD-1/PD-L1 interactions by increasing the proliferation and cytotoxic activity of CD8+ T cells while decreasing CD39-expressing Treg cells in a dose-dependent manner. Likewise, the inhibition of PD-1/PD-L1 interactions was also correlated with the increased production of IL-2 and Granzyme B. Our findings imply that Ce6-PDT is a promising immunotherapy with the potential to improve the abscopal effect.

www.nature.com/scientificreports/ laser irradiation to transform molecular oxygen into tumor-killing reactive oxygen species (ROS). As a result, ROS can directly kill tumor cells by triggering necrosis or apoptosis, and it can impair the tumor vasculature by depriving it of oxygen and nutrients 3,4 . Further, PDT may also affect the immune system's immunosuppressive or immunostimulatory functions 5 . For many years, the explanation for how PDT affected cancer was only through local effects 6 . However, a new research has demonstrated that PDT also has the potential to eliminate metastatic and disseminated tumor cells by promoting antitumor immunity. PDT induces an effective anticancer immune response by forcing tumor cells to die in an immunogenic manner 7,8 . These dying tumor cells often release or expose damage-associated molecule patterns (DAMPs) as 'immunogenic signals' , that attract antigenpresenting cells (APCs) and are subsequently presented to CD8 + T cells, inducing a long-lasting antitumor immune response 8 . Through PDT, antitumor responses to both primary and distant cancers are connected by immunogenic cell death (ICD), dramatically increasing the immunogenicity of dying cancer cells, and causing controlled immune system activation 9 . Such immune responses generate the abscopal effect, a phenomenon where local irradiation induces distant contralateral non-treated tumors to regress 10 . The abscopal effect has been observed in patients with metastatic cancer treated with PDT alone or in conjunction with other therapy; however, it is rare. In the phase II clinical trial, patients with intraperitoneal tumors that primarily originated from ovarian cancer responded favorably to PDT with Photofrin (2.5 mg/kg) treatment 11 . According to Mortano et al., local recurrence and development of lung metastasis in clinical cases of Feline Injection-Site Sarcoma were significantly decreased after administration of combined PDS (Photodynamic Surgery) and PDT with acridine orange, reflecting the occurrence of an abscopal effect 12 . Furthermore, Kabingu et al., reported that enhanced cytolytic activity of splenocytes and infiltration of CD8 + lymphocytes were responsible for the remission of tumors at the distant site induced by local treatment of Photofrin-mediated PDT 13,14 .
Some preclinical trials have demonstrated benefit in other malignancies; the abscopal effect appears to be more common in tumors that are more immunogenic, such as melanoma, breast, lung, and liver cancers 15 . Despite the fact that PDT has been licensed and used for many different cancer types, the melanin pigmentation of melanoma limits the therapeutic efficiency of many existing photosensitizers like Photofrin, as melanin absorbs light at wavelengths between 300 and 600 nm 16 . However, PDT with photosensitizers, like Chlorin e6 (Ce6), is an effective and well-tolerated method of melanoma treatment because its absorption peak lies in the long-wave spectral region of 660 nm, which increases its ability to penetrate in biological tissues 17,18 . Since T cell dysfunction serves as a means of cancer immune escape, targeting the expression of PD-1 and PD-L1 immune checkpoints is an essential tactic for enhancing immunity against cancers 19 . Therefore, tumor survival and development may be affected by the PD-1/PD-L1 signaling axis 20 . The transmembrane protein PD-1 is known to be expressed on the surface of a variety of immune cells, including activated T cells like CD4 + and CD8 + T cells (cytotoxic T lymphocytes, or CTLs), while PD-L1 is expressed in tumor cells, and its interaction with PD-1 prevents the immune system-mediated killing of cancer cells 21 . As a result, T cell activation in peripheral tissues and the tumor milieu is suppressed by the molecular receptor PD-1, which further contributes to immune evasion 22,23 . Immune checkpoint inhibitors (ICIs) authorized by the FDA such as pembrolizumab, nivolumab, and cemiplimab target PD-1 or PD-L1 to a number of human tumors and have recently been shown to be effective in immunotherapy 24,25 . As opposed to monoclonal antibodies, small molecule inhibitors like BMS-200 and BMS-202 also offer a number of advantages in addressing these issues 26,27 .
Therefore, we investigated the immunogenic variables that affect the abscopal effect of Ce6-PDT. A bilateral subcutaneous melanoma and pancreatic tumor mouse model was employed in this study to test Ce6-PDT for the treatment of both local and distant tumors. Higher CD8 + T cell and macrophage peritumoral infiltration are linked to the antitumor effects of Ce6-PDT. PD-1/PD-L1 checkpoint inhibition has not yet been explored for Ce6-PDT. Hence, these findings suggest Ce6-PDT as a new immunotherapeutic strategy that targets primary tumors and induces systemic antitumor immunity by targeting the PD-1/PD-L1 interactions.

Result
Ce6-PDT elicited the inhibition of tumor growth. To investigate the Ce6-PDT-mediated antitumor abscopal effect, two bilateral tumor mouse models were developed using B16F10 melanoma and the Panc02 pancreatic cancer cells. After injecting Ce6 (2.5 mg/kg) followed by local irradiation to right tumor, we tested the ability of Ce6-PDT to prevent tumor growth in both irradiated right and non-irradiated left tumors from day 11 up to day 28 (Figs. 1A,B, and 2A). All PDT-treated animals initially had tumor shrinkage in the irradiated region; however, some mice later developed tumor regrowth due to the aggressiveness of B16F10 tumors. The Abs eff and Abs ineff groups were separated based on the tumor-specific efficacy of PDT and abscopal effect (Fig. 1C). Our findings indicated that Ce6-PDT exerted a dramatic suppression on the irradiated tumor volume of the Abs eff group as compared to the Abs ineff group, in Fig. 1D,E. Compared to tumors in the control and Abs ineff group, the non-irradiated melanoma tumors of the Abs eff group also displayed much lower tumor weight (Fig. 1F). The Abs ineff group showed larger tumor weight on both the left and right sides as compared to the Abs eff group. A 2.5 mg/kg dosage of Ce6-PDT also reduced tumor volume and weights in both irradiated and non-irradiated pancreatic tumors in the Abs eff group (Fig. 2B,C,D) as compared to that in Abs ineff and control mice. The spleen weights of the pancreatic tumor-bearing mice of the Abs eff and Abs ineff groups were also assayed. Similar to the tumor weight changes, the spleen of the Abs eff group showed lesser weight than those of the Abs ineff group and control group (Fig. 2E). The current investigation suggested that in both the effective and ineffective groups, the tumor volume, weight of the tumor and spleen showed similar patterns. Thus, Ce6-PDT suppressed growth of both irradiated right and non-irradiated left tumors in the Abs eff group. www.nature.com/scientificreports/ sion of tumors to attain PDT and its abscopal effect 28 . We measured the flow-cytometric analyses of T cells and their subtypes in control, Abs eff , and Abs ineff tumors of the irradiated right and non-irradiated left regions. We assayed the percentages of tumor T-cell infiltration of CD3 + , CD45 + , CD25 + , CD103 + , and CD39 + (Figs. 3A-E), to evaluate the systemic immune response after the primary tumors were subjected to PDT. After concurrent with reduced tumor growth, the Ce6-PDT treatment increased the percentage of CD3 + CD4 + tumor-infiltrating T cells and CD45 + , a leukocyte common antigen, indicating an increased immune activation (Fig. 3A,B). Ce6-PDT also resulted in the decrease of the irradiated tumor volumes, suggesting stronger T cell responses. Allografts in the Abs eff group exhibited considerably lower frequencies of regulatory CD25 + T cells in both the irradiated and non-irradiated tumors compared to the control, which signifies abrogation of their immune inhibitory activity (Fig. 3C). In the Abs eff group, the irradiated right tumor showed a low frequency of CD8 + CD103 + and CD8 + CD39 + T cells, but higher levels in the non-irradiated tumor (Fig. 3D,E). However, in the Abs ineff group, a decreased infiltration of CD3 + , CD45 + T cells was noted, thus indicating decrease in T cell function and poor prognosis of the tumor. We have also provided the gating scheme to determine the T cells and their subtypes in the supplementary file (SI-S3). These findings show that Ce6-PDT elicits host immunological responses through an increase in number of T cells. However, the abscopal ineffective tumors may have a hindrance in the tumor microenvironment that does not allow immune cells to migrate into distant tumor areas and results in lower T cell activation 29,30 . Ce6-PDT promoted the immune cell distribution in the Abs eff tumors of the melanoma cancer mouse model. Macrophages and other immune cells were similarly explored using flow cytometry in the right and left tumors of the melanoma mouse model. We then evaluated the population and status of macrophages in the Abs eff tumors to provide additional insight into the activation mechanism of the immune response produced by Ce6-PDT in vivo. We analyzed the percentages of CD11c, F4/80 (macrophage-monocyte marker), NK 1.1, CD86, and CD206, which are necessary for the induction of adaptive immunity, in the Abs eff and Abs ineff groups ( Fig. 4A-E). We noted that CD11c + and F4/80 + macrophage populations were increased in the irradiated right tumors, and the increased patterns were persisted in the non-irradiated left tumors in the Abs eff group (Fig. 4A,B). Similarly, in the Abs eff group, NK cells were increased in the non-irradiated tumors as compared to the ineffective group (Fig. 4C). However, the percentages of CD11c + , F4/80 + macrophage, and NK cells in the Abs ineff group were decreased. A key component of tumor immunity is the expression of CD86 + on macrophages and dendritic cells, which are essential for antigen presentation to T cells. The percentage of CD86 + and CD206 + macrophages was significantly higher in the non-irradiated tumors compared to the irradiated tumor of www.nature.com/scientificreports/ the Abs eff group ( Fig. 4D and E). The total number of immune cells in the abscopal effective tumor was significantly higher, which is consistent with the functioning of T cells and the pronounced abscopal responses in the Abs eff group as compared to the Abs ineff group.

Ce6-PDT increased the intra-tumoral density of T cells in the pancreatic cancer mouse model.
We also focused on the following group of markers to investigate the infiltration of various T cell subpopulations such as CD3 + , CD4 + , CD8 + , CD25 + , CD103 + , and CD39 + in the pancreatic cancer model with Ce6-PDT treatment ( Fig. 5A-F). The pancreatic tumors in the irradiated right and non-irradiated left sides of the Abs eff group showed higher CD3 + T cells compared to the Abs ineff group. CD4 + helper and CD8 + cytotoxic T cells make up the majority of T cell populations in the tumors ( Fig. 5B and C). A significantly higher CD3 + T cell infiltration is found in the left tumors compared to the right tumors (Fig. 5A). However, lesser effects were detected in the Abs ineff group. We further analyzed CD25 + , CD103 + , and CD39 + levels in both Abs eff and Abs ineff groups ( Fig. 5D-F). The pancreatic tumors on the right side revealed a lower percentage of CD25 + , CD103 + , and CD39 + compared to the non-irradiated left side in the Abs eff group while the increased percentages of CD25 + , CD103 + , and CD39 + were observed in the Abs ineff group. Thus, similar to the melanoma mouse model, Ce6-PDT elicits a positive abscopal effect through the inhibition of Treg cells that resulted in the activation of CD4 + and CD8 + T cells. However, the population of T cells in the spleen of the Abs eff and Abs ineff group did not show any difference (SI-1, Fig. S1).
Ce6-PDT promoted the intratumoral immune cell distribution in the pancreatic cancer mouse model. Focusing on the role of innate immune system with Ce6-PDT treatment in C57BL/6 mice, we examined the percentages of macrophages, dendritic cell, and NK cell populations in the tumor tissues by FACS analysis (Fig. 6A-G). CD11c + (macrophage cells) were increased in the irradiated tumors and non-irradiated tumors in the Abs eff group compared to the Abs ineff group (Fig. 6A). The results showed that the percentage of F4/80 and NK1.1 cells increased in both the irradiated and non-irradiated tumors of the Abs eff group (Fig. 6B,C). Accordingly, an increased CD86 + population revealed increased maturation of dendritic cells, following the same pattern as Abs eff group of the melanoma model (Fig. 6D). In the Abs eff group, CD86 + was found to be higher in the non-irradiated tumors than in the irradiated tumors, whereas the Abs ineff group showed a different result. CD206 + and CD68 + were also increased, indicating that PDT treatment triggers macrophage infiltra- www.nature.com/scientificreports/ tion and inflammatory response in the distant tumor sites (Fig. 6E,F). The Ce6-PDT treatment significantly increased tumor-associated CD163 + M2 macrophage in the ineffective group, pointing out the poor prognosis of the tumor (Fig. 6G). Therefore, these data indicated that the percentages of myeloid cells were increased in the effective group but decreased in the ineffective group. In the spleens of effective and ineffective groups, we found similar patterns of immune cell increase for CD11c + , F4/80 + , NK + , CD86 + , and CD206 + as shown in the tumors (SI-2, Fig. S2). In order to detect the ability of Ce6-PDT to improve T cell activity, the secretion of IL-2 was analyzed in the coculture system (Fig. 7A,B). In the system with escalating dosages (1-4 μM) of Ce6 followed by PDT for 48 h in hPD-1 effector and hPD-L1/TCR CHO-K1 cells, released IL-2 was increased, suggesting T cells activation by Ce6-PDT. Ce6-PDT at 1 μΜ elevated the levels of IL-2 up to about twofold higher than the untreated group, which then decreased from 2 to 4 μΜ in 48 h. However, for 24 h after Ce6-PDT treatment, IL-2 secretion was not detectable. These results suggest that Ce6-PDT effectively enhanced T cell immune function by inhibiting the PD-1/PD-L1 axis in 48 h.

Ce6-PDT enhanced T cell-associated cytokines by inhibiting
Tumor-specific CD8 + T cells upregulated granzyme B. In determining the level of granzyme B, we also evaluated the degranulation of CD8 + T cells from the cocultures. The CD8 + T cells that were treated with 2 and 4 μM of Ce6 and then exposed to 660 nm light exhibited elevated levels of granzyme B in 24 and 48 h www.nature.com/scientificreports/ respectively, as compared to the untreated and the Ce6 only (4 μM)-treated group (Fig. 8A,B). Ce6-based PDT at doses of 1 and 2 μM did not differ in any way from the untreated control. The levels of granzyme B in the tumor-specific CD8 + T lymphocytes were lower during 24 h incubation but were found to be increased after 48 h of incubation. These findings suggest that Ce6-PDT can cause CD8 + T cells to release granzyme B, which further trigger the death of targeted tumor cells and increase immune responses.
Ce6-PDT blockade of PD-1/PD-L1 interaction enhanced CD8 + T cell activities and reduced CD39 + T cells. The ability of Ce6-PDT to enhance CD8 + T cell activation was assessed using a PD-1/PD-L1 blockade assay. In this study, Ce6 was utilized at concentrations ranging from 1 to 4 μΜ, and of these, 4 μΜ demonstrated the highest PD-1/PD-L1 blockage with enhanced CD8 + T cell functional activity. In both the 24 and 48 h of coculture, CD8 + T cell functional activity at 4 μΜ was significantly higher than at other Ce6 doses (Fig. 9A,B). When compared to the control, Ce6-PDT (4 μM) treatment for 48 h enhanced T cell functional activity by 2.2 fold. We also examined the expression of CD39 + T cells in Ce6-PDT-treated cells (Fig. 9C,D). PDT with Ce6 of 2 and 4 μΜ significantly decreased the level of CD39 + at 24 h coculture compared to untreated, Ce6 only, and light only-treated cells. However, at 48 h coculture, the level was significantly declined with Ce6 doses from 1 to 4 μΜ. The PD-1/PD-L1 blockage achieved by Ce6-PDT produced antitumor effects by increasing the cytotoxic function of CD8 + T cell while decreasing the infiltration of CD39 + T cells.

Discussion
PDT has been considered useful for managing and eliminating local lesions in several malignant cancers 32 . The systemic impact of the abscopal effect by PDT is becoming more and more of a focus 3 . To test the Ce6-PDTmediated abscopal effect, we first established two immunogenic melanoma and pancreatic cancer mouse models (B16F10 and Panc02 cells). We discovered that in these models, Ce6-PDT generated the abscopal effect. Multiple studies have demonstrated that PD-1/PD-L1 targeted therapy presents a possible immunotherapy strategy for cancer by suppressing various stages of the immunological response between T cells and tumors 9,33 . The current study found that Ce6-PDT induced T cell activation and antitumor immunity by blocking PD-1 and PD-L1. However, the inhibitory effects of Ce6-PDT on the immune checkpoint are currently unknown and hence need further investigation. According to Ebner et al. (2017), the abscopal effect is a naturally occurring ability that can be activated and amplified by therapies such as immune treatment or radiation and depend on the actions of specific but undiscovered host-tumor epigenetic factors 34 . This supports our study's discovery that Ce6-PDT-treated mice www.nature.com/scientificreports/ had different abscopal outcomes, which led to the division of these animals into two main groups: an abscopal effective group (tumors on both irradiated and non-irradiated sides decreased), and an abscopal ineffective group (tumors on both sides did not decrease). The noteworthy outcomes in the Abs eff group may be attributed to Ce6-PDT's ability to modulate the tumor surroundings, which caused irradiated tumor cells to produce tumor antigens and immunogenic cell death. These antigens might be picked up by antigen-presenting cells like dendritic cells, which is correlated with increased infiltration of CD11c + , F4/30 + , and co-stimulatory molecules like CD86 + in the tumor of the Abs eff group. It is well known that activated CD4 + T cells might assist CD8 + T cells and further support the CD8 + effector for its memory formation and maintenance. We found that these CD8 + effector memory T cells then produce molecules such as granzyme A, granzyme B, perforin, etc., which have cytotoxic effector functions and assist in slowing the growth of the distal tumor in the effective group 35,36 .
Our study suggests that the innate immune system was involved, which most likely made it easier for neutrophils, dendritic cells, and macrophages to gather. Several studies have shown that PDT stimulates the immune system in many ways, including the release of tumor-associated antigens (TAAs) and immune-stimulatory molecules from tumors, which may activate and trigger an anti-cancer immune response 37,38 . Although PDT has been shown to activate both humoral and adaptive cell-mediated immunity, CD8 + T cells are primarily responsible for PDT's immunological effects 39 . Our findings demonstrated that the infiltration of CD8 + T cells was enhanced by Ce6-PDT in the irradiated and non-irradiated tumors of the Abs eff group. However, T cells were exhausted in the non-irradiated tumor of the Abs ineff group and failed to elicit an anticancer immune response. In the non-irradiated left tumor, we observed a reduction in the percentage of CD3 + , CD4 + , and CD8 + T cells after PDT in the ineffective group. The depletion of CD8 + T cells in the ineffective group indicates antitumor immunity is insufficient, as proliferation and activation of CD8 + T cells are too low and lack persistence due to the aggressiveness of tumors. The inability of the host CD8 + T cells to localize to a tumor might be due to the absence of sufficient immunogenic antigens for T cell recognition. These tumors might not have a high mutation rate in order to generate immunogenic neoantigens capable of activating CD8 + T cells 40 . According to previous study, tumors may have barriers in their milieu that prevent responding immune cells from migrating and penetrating into non-irradiated tumor 39 . Among the ineffective group, the composition of immune cells in the irradiated right tumor was predominantly changed by PDT, as characterized by fewer dendritic cells (DCs). Tumor-infiltrating immune cells represent the actual conditions of the tumor immune microenvironment. The immune cells such as macrophages (CD11c + and F4/80 + ) were reduced in the ineffective group in comparison to the effective group, indicating that Ce6-PDT www.nature.com/scientificreports/ was unable to promote the sufficient release of tumor antigens (B16F10 and Panc02), which could not activate immune cells and further kill cancer cells. Therefore, in the ineffective group, decreased immune cells were associated with fewer T cells. Previous research has shown that ligands such as PD-L1, are overexpressed in melanoma and pancreatic cancers 41,42 . Using the humanized PD-1 mouse model with hPD-L1 knock-in and the MC38 tumor-bearing PD-1 mouse model, we evaluated the antitumor effects of Ce6-PDT. According to our findings, Ce6-PDT-induced suppression of the PD-1/PD-L1 axis could stimulate increased activation of CD8 + T cells and improve antitumor activity. It is well known that IL-2 is also a central mediator of Treg suppression during immune response 43 . Therefore, a 2 to threefold reduction in the number of Tregs during PD-1/PD-L1 blockade could increase IL-2 release and the pool size of effector cells. Many studies have shown that combined IL-2 therapy and PD-L1 blockade can produce synergistic effects in treating human chronic inflammation and cancers 44 . Granzyme B is also a marker for T cell activation, which is known to be the product of either CD4 + or CD8 + T cells 45 . This cytokine's increased secretion in the coculture corresponds to increased activation of T cell after Ce6-PDT treatment with inhibition of PD-L1 and PD-1 check-point. In the study, Ce6-PDT increased the CD8 + T cell activity through IL-2 production at 2 μM concentration of Ce6. We also tested Ce6-PDT's capacity to increase T cell functional activity by preventing PD-1 and PD-L1 interactions. The results showed that Ce6-PDT increased CD8 + T cells In conclusion, our findings for the first time suggest that Ce6-PDT can induce potent local and systemic antitumor immune responses. Antitumor effects of Ce6-PDT were achieved by inhibiting the PD-1/PD-L1 interaction, which significantly accelerated the abscopal effects. To our knowledge, this is also the first time the impact of Ce6-PDT on the abscopal effect has been evaluated by targeting PD-1 and PD-L1 interactions. The findings of this study may aid in the development of effective inhibitors that block PD-1 or PD-L1 immune checkpoints and improve the outcomes of Ce6-PDT. We believe that more research on Ce6-PDT in combination with other immune check-point inhibitors like pembrolizumab or nivolumab may provide a strategy to improve the abscopal effects and get a better prognosis.

Methods
Cell culture. B16F10 mouse melanoma cells procured from the Korean Cell Line Bank (KCLB; Seoul, Korea) were cultured in Dulbecco's Modified Eagle Medium (DMEM) with 10% heat-inactivated fetal bovine serum (FBS) and 1% Antimycin A. Panc02/Luciferase stable cell line, a mouse pancreatic epithelial cell line derived from an adenocarcinoma model was obtained from GenTarget (San Diego, CA, USA) and was cultured in DMEM supplemented with 10% FBS, at 37 °C, in a humidified incubator, under a 5% CO 2 atmosphere. Every two days, the media was changed and sub-cultured until they reached 80-90% confluency.  Murine tumor model. Male C57BL/6 mice (5-6 weeks old) were procured from Orient Bio (Seongnam, Korea). Mouse models of bilateral melanoma tumors or pancreatic cancer models were established by injecting B16F10 (2 × 10 5 ) or Panc02 (8 × 10 5 ) subcutaneously into the right flank of mice (irradiated tumor) and the left flank (non-irradiated tumor) respectively of C57BL/6 mice. Every three days, the tumor dimensions were measured with digital calipers, and the volume was determined using the formula: volume = (W 2 x L)/2 (W: short diameter; L: long diameter).
When the mean tumor volume reached 90 mm 3 , the mice were randomly assigned to various treatment groups (n ≥ 5 per group). The mice in the control group were administered normal saline (NS). The Ce6 solution was prepared by dissolving it in normal saline. In the irradiation group, PDT was conducted by injecting 100 μL of Ce6 into the tail vein for the choice of dose of 2.5 mg/kg, with a drug-PDT interval of 3 h and LED 660-nm illumination at 100 J/cm 2 . The explanation for the choice of a dose of 2.5 mg/kg Ce6 and the pharmacokinetic study for determining the drug interval time before light irradiation in PDT are given in the supplementary file (SI-S5, S6, Fig. S6). The vehicle-treated control, however, did not receive any light exposure (SI-S7, Fig. S7). After 28 days, all the mice were sacrificed, and their tumors were collected for further studies. All the images were taken by the first and second authors. Animal procedures were conducted in accordance with appropriate regulatory standards under the protocol IACUC #ds002205112-EUTO3, which was reviewed by the Institutional Animal Care and Use Committee of the Dongsung Cancer Center. These studies were carried out in compliance with the ARRIVE guidelines.

Murine splenocytes or tumor-infiltrating T cells (TILs) isolation. Tumors from control, Abs eff and
Abs ineff group were selected for tumor extraction. The single-cell suspension of mouse splenocytes was prepared  IL-2 measurement assay. A sandwich ELISA was used to measure the amount of IL-2 released by activated T cells in the cell coculture supernatants according to the manufacturer's instructions. Briefly, the plates (0.32 cm 2 , # 3590, Corning) were coated with an anti-mouse IL-2 monoclonal antibody with 0.1 M sodium carbonate (pH 9.5) and treated overnight at 4 °C. The plates were then rinsed in Phosphate buffered saline with Tween 20 (PBS-T) and were blocked for 1 h at room temperature with PBS containing 10% FBS. Each well was treated for 1 h at room temperature with biotinylated IL-2 antibody and streptavidin-horse radish peroxidase (HRP). The relative absorbance was measured at 450 nm using a SpectraMax i3 microplate reader (Molecular Devices, San Jose, CA, USA).
Granzyme measurement assay. According to the manufacturer's instructions, the amount of Granzyme B produced by activated T cells in cell coculture was determined using culture supernatants that were gathered, spun free of cells, and frozen at -20 °C. The wells were blocked for 1 h at room temperature using PBS, 0.05% Tween-20, and 0.1% bovine serum albumin (BSA) after two PBS washes. After adding the samples for 2 h at room temperature, wells were washed and further 1 g/mL Granzyme B-biotin was added for 1 h. Wells were washed and were incubated in 1 g/mL of streptavidin-HRP for 1 h. Wells were washed following the addition of streptavidin-HRP, and then TMB substrate was added for up to 20 min (Biolegend). After adding the TMB stop solution to terminate color development, relative absorbance was measured at 450 nm using a SpectraMax i3 microplate reader (Molecular Devices, San Jose, CA, USA).
Statistical analysis. The data is presented as mean ± standard error of the mean. The mean difference between the treatment and control groups was taken into account when determining statistical significance using one-way ANOVA with Tukey's post hoc test for multiple comparisons. All statistical analyses were performed using GraphPad Prism software (v5.02; La Jolla, CA, USA). A p-value < 0.05 was considered statistically significant.
Ethics approval and consent to participate. The study protocol and animal procedures were approved by the Animal Care and Use Committee of the Dongsung Cancer Center, Daegu, under the protocol IACUC #ds002205112-EUTO3.

Data availability
The data sets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request.